The Immunomodulatory Role of Epidermal Hepcidin in Infectious Condition

Université Paris Descartes

École doctorale : Bio Sorbonne Paris Cité Institut Cochin, Hypoxia and iron homeostasis team The immunomodulatory role of epidermal hepcidin in infectious condition

Par Mariangela MALERBA

Thèse de doctorat de Physiopathologie

Dirigée par Carole PEYSSONNAUX

Présentée et soutenue publiquement le 25/09/2018

Devant un jury composé de : Dr. Carole PEYSSONNAUX Directeur de thèse Pr. Sylvie CHOLLET-MARTIN Rapporteur Dr. François CANONNE-HERGAUX Rapporteur Pr. Selim ARACTINGI Examinateur Dr. Agnès FOUET Membre invité

Contents ABSTRACT ...... 6 INTRODUCTION ...... 8 HEPCIDIN ...... 9 1.1. Hepcidin: synthesis and structure ...... 9 1.2. Hepcidin antimicrobial function ...... 11 1.3. Hepcidin: iron regulatory hormone ...... 12 1.3.1. Iron metabolism ...... 12 1.3.2. Regulation of the iron homeostasis ...... 16 1.4. Hepcidin/infection axis ...... 17 1.5. Hepcidin regulation ...... 21 1.6. Extra-hepatic hepcidin and other roles...... 25 2. SKIN ...... 29 2.1. Mammalian skin architecture ...... 29 2.2. Skin Immunity ...... 35 3. NEUTROPHIL ...... 47 3.1. Neutrophil production and bone marrow egress ...... 47 3.2. Neutrophils in tissue ...... 50 3.3. Chemoattractant molecules: the chemokine family ...... 53 3.4. Role of neutrophils in fighting infection ...... 55 4. GAS ...... 63 4.1. Epidemiology of GAS infection ...... 63 4.2. M1T1 : GAS invasive strain ...... 66 4.3. GAS-phagocyte fighting ...... 70 4.4. GAS-keratinocyte fighting ...... 73 4.5. GAS iron metabolism ...... 74 THE AIM OF THE STUDY ...... 75 RESULTS ...... 77 1. Hepcidin shares structural similarities with IL8 and displays neutrophil chemoattractant properties. 78 Hepcidin is expressed in the epidermis ...... 85 Epidermal hepcidin has a protective role against GAS infection ...... 88 Hepcidin has no antimicrobial effect against GAS in vitro and does not influence keratinocyte antimicrobial functions...... 92 Hamp1 Δker mice do not display epidermal iron accumulation ...... 93 Hepcidin modulates keratinocyte immune response ...... 96 Hepcidin- mediated CXCL1 production protects against GAS infection...... 99 2

Hepcidin is resistant to SpyCEP cleavage ...... 101 Hepcidin subcutaneous injections have a therapeutic potential against GAS infection ...... 104 SUPPLEMENTARY FIGURES ...... 106 MATERIALS AND METHODS ...... 110 DISCUSSION ...... 117 New chemoattractant role of hepcidin...... 118 Hepcidin is expressed in the skin in pathophysiological conditions ...... 120 Keratinocyte- derived hepcidin has an immunomodulatory role during infection ...... 122 Hepcidin as a therapeutic agent ...... 127 Conclusion and highlights ...... 130 BIBLIOGRAPHY ...... 131 ANNEXES ...... 158 Epidermal hepcidin prevents systemic bacterial spreading in a Group A streptococcus-induced necrotizing soft tissue infection model ...... 159 Methods for treating gram positive bacterial infection ...... 176

TABLE OF ABBREVIATIONS

Aa amino acid IRP iron regulatory protein RLR RIG-like helicases AI anaemia of inflammation JAK Janus activated kinases ROS reactive oxygen species AMP antimicrobial peptide KRT keratin SC subcutaneous APC antigen presenting cell LF lactoferrin Scl1 streptococcus collagen-like ARF acute rheumatic fever LFA1 lymphocyte function- ScpA streptococcal C5a peptidase associated antigen1 A BM basement membrane LIP labile iron pool Sda1 Sterptococcal DNAse 1 BMP bone morphogenetic protein LPS lipopolysaccharide Shp streptococcal heme binding CLR C-type lectin receptor protein LTA lipoteicoic acid CXCL CXC-chemokine ligand Shr streptococcal heme protein MMP8/9 metrix metalloprotease receptor CXCR CXC-chemokine receptor 8/9 Ska1 Streptokinase 1 DAMP damage associated MPO myeloperoxidase molecular pattern SLO streptolysin O NE Neutrophil elastase DC dendritic cell SLS streptolysin S NET neutrophil extracellular trap dcytB duodenal cytochrome B sodA superoxide-detoxifying NLR NOD like receptor DMT1 Divalent metal transporter enzyme A 1 NoxA NADPH oxidase A SpeA/B Streptococcal Pyrogenic EPO erythropoiesis NRAMP1/2 natural resistance Exotoxin A/B associated macrophage protein Spy CEP streptococcus pyogenes fMLP N-Formylmethionyl- 1/2 leucyl-phenylalanine cell envelope protease NTBI non transferrin bound iron FPN ferroportin Sse Streptococcal esterase OH heme oxygenase GAS Group A streptococcus STAT signal transducers and PAMP pathogen associated activators of transcription GCPR G protein coupled receptor molecular pattern STSS streptococcal toxic shock G-CSF Granulocyte colony PDA 4 protein arginine deiminase syndrome stimulating factor type 4 Tf transferrin GM-CSF Granulocyte- PGN peptidoglycan macrophage colony-stimulating TfR1/2 transferrin receptor 1/2 PMN polymorphonuclear TJ tight junction HA hyaluronan capsule leukocyte TLR toll-like receptor HCP1 heme carrier protein PRR pattern recognition receptor TMPRSS6 matriptase HF hair follicle PSGL1 P-selectin glycoprotein HH hereditary hemochromatosis ligand 1 WT wild type HIFs hypoxia inducible factors PSGL1 P selectin glycoprotein ligand 1 IFE interfollicular epidermis PSGN poststreptococcal ILC innate lymphoid cell glomerulonephritis IRE iron responsive element RHD rheumatic heart disease

ABSTRACT Hepcidin, first discovered as an antimicrobial peptide, is now considered as the key iron regulatory hormone, inhibiting duodenal iron absorption and iron recycling by macrophages. Hepcidin expression is induced by iron accumulation and infection/inflammation while diminished in situations of iron needs. Hepcidin, as suggested by its name, is mainly expressed in the liver (Hep- for Hepatic) but also at low levels by many other cells. Our team has demonstrated that hepatic hepcidin is sufficient to ensure the systemic iron homeostasis in basal conditions; suggesting that extra-hepatic hepcidin could play local roles in pathophysiological conditions. Hepcidin contains 4 disulfide bonds with β-sheet fold and positively charged residues at the surface, showing close structural similarity to β-defensins, a subfamily of antimicrobial peptides (AMP).

AMPs are small proteins with antimicrobial activities but also immunomodulatory functions (chemotaxis, …). Because hepcidin possesses disulfide bridges, resembling the intramolecular disulfide bonds critical for the function of chemokine proteins, we investigated the direct immunomodulatory potential of hepcidin. We observed that human hepcidin shares structural similarities with the human CXCL1 homologue. Taking advantage of in vivo bioluminescence assay, we observed that subcutaneous (SC) injection of hepcidin causes myeloperoxidase activity neutrophil in vivo. We demonstrated that hepcidin triggers neutrophil migration in vitro with a bell-shaped dose response curve, characteristic of chemokines. These results highlight a new immunomodulatory role of this AMP. Epithelia, such as intestine and skin, are the main source of AMPs. However, expression of hepcidin and its functional role as AMP in these tissues has never been studied. Therefore, I specifically investigated this aspect in the skin, using a model of group A streptococcal (GAS) SC infection. These gram-positive bacteria commonly colonize skin and are responsible for a wide range of both skin and invasive infections causing more than 500,000 deaths per year. The first host immune effectors encountered by GAS are endogenous AMP produced by epithelial cells and by circulating immune cells (neutrophils and macrophages) so GAS represents a powerful model to study hepcidin function as antimicrobial peptide in the skin. We specifically deleted hepcidin in keratinocytes (Hepc KOker) and in myeloid cells (Hepc KOmyel). Interestingly, while Hepc KOmyel and WT littermates showed the same number of bacteria at local and systemic level, we

ABSTRACT found that Hepc KOker mice present a worse outcome after infection than WT littermates, with a significantly higher number of bacteria at the lesion site and a stronger systemic spread of infection. We observed that hepcidin has neither bacteriostatic nor bacteriolytic effect against GAS in vitro and that both WT and Hepc KOker primary keratinocytes displayed the same bactericidal activity. Interestingly, we found that hepcidin promotes keratinocyte production of CXCL1, a key neutrophil chemokine, in vitro. Consistently, both the amount of CXCL1 and the number of neutrophils recruited at the site of infection was lower in Hepc KOker mice compared to WT littermates. Moreover, CXCL1 SC injections rescue the phenotype, confirming that the invasiveness observed in Hepc KOker is specifically caused by the absence of hepcidin-mediated CXCL1 production. Finally, we demonstrated that injections of exogenous hepcidin prevents the infection spread. These results highlight a new protective role of epidermal hepcidin against GAS infection, suggesting that hepcidin could represent a novel approach for therapy of GAS infections.

INTRODUCTION

HEPCIDIN Hepcidin is a 25aa peptide (2,7kDa) mainly produced by the liver. This hepatic peptide was originally isolated from human blood (Krause et al., 2000) and urine (Park et al., 2001a) and described as a antimicrobial peptide (AMP). First named LEAP-1 (Liver Expressed Antimicrobial Peptide), it was next renamed “hepcidin” recapitulating its two key properties: hepc- stands for hepatic and -cidin is related to its antimicrobial properties. Then in 2001, a third team observed an upregulation of hepcidin in the liver of iron overloaded mice, suggesting an additional role of this hepatic peptide besides its antimicrobial properties described earlier (Pigeon et al., 2001). Finally, in 2001 the team of Sophie Vaulont elucidated hepcidin iron regulatory function. Studying USF2-/- animals, they accidentally deleted hepcidin. They observed that, in contrast to wild-type mice, hepcidin knock out animals display a strong age- dependent multivisceral iron overload and iron depletion in splenic macrophages, showing that the lack of this peptide determines an abnormal high iron absorption and iron recycling from reticuloendothelial macrophages (G. Nicolas et al., 2001). Hepcidin deficiency has been next associated with hereditary hemochromatosis, a common recessive disorder characterized by progressive iron overload, which may lead to severe clinical complications and multiorgan dysfunctions. In regard to our study, I will present in this first chapter this peptide, its regulation and its key functions: the antimicrobial one, the iron regulatory one and the hepcidin/infection axis.

1.1. Hepcidin: synthesis and structure Human hepcidin is encoded by a three exons gene on chromosome 19 (Park et al., 2001a). In mouse there are two hepcidin genes (hamp1 and hamp2), but only hamp1 has the iron regulatory function (Lou et al., 2004). Hepcidin gene encodes a prepropeptide of 84 amino acids. At the

NH2-terminal end, preprohepcidin has a signal sequence which is cleaved to give rise to the 60 aa intermediate cellular prohepcidin. This prohormone contains a consensus sequence for the furin convertase. Furin cleaves prohepcidin to form active hepcidin-25 that is then secreted from cells (Figure 1) (Krause et al., 2000; Park et al., 2001a; Valore & Ganz, 2008).

INTRODUCTION

Figure 1: Hepcidin processing. Schematic representation of preprohepcidin and its maturation. Hepcidin is translated as 84aa protein, which is cleaved to the prohepcidin by eliminating the N-terminal signal peptide. The convertase furin (red arrow) is responsible for the formation of the mature 25aa peptide. Dashed arrows represent the plausible cleaving sites of hepcidin-20 and -22.

The mature form of hepcidin contains 8 cysteine residues, highly conserved throughout the evolution (Figure 2), forming 4 disulfide bonds and cationic residues that confer its positive charge (Krause et al., 2000; Park et al., 2001a) (Figure 2). Although the 25aa-hepcidin is the main form circulating in the bloodstream, other hepcidin isoforms have been isolated (Park et al., 2001a). Indeed, the mature hormone can be further processed to yield 20- and 22- aa hepcidin forms. Whether these isoforms are the results of hepcidin catabolism or are actively produced by an active 25aa-hepcidin cleavage is still under debate (Campostrini et al., 2012). Moreover, since the N-terminus of hepcidin is essential for its iron regulatory role (Clark et al., 2011; Nemeth et al., 2006), these two N-terminal truncated isoforms loose this function. Hepcidin-20 and -22 physiological roles are still unclear. Numerous studies have observed antimicrobial effects of these hepcidin isoforms in vitro, suggesting a specific antimicrobial residual function of the ancestral bactericidal property of hepcidin present in lower vertebrate species. However, the microbicidal effect is described exclusively in vitro.

INTRODUCTION

Figure 2: Hepcidin 3D structure. Amino acid sequence of hepcidin-25 among different species. The 8 cysteine residues highly conserved are highlighted (on the left). Tridimensional structure and disulfure bonds (in yellow) of hepcidin, adapted from Clark R J, et al., 2011.

1.2. Hepcidin antimicrobial function The amphipathic structure and disulfide bonds of hepcidin-25 are common characteristics of antimicrobial and antifungal peptides. As previously discussed, hepcidin was first described as an AMP. In a first study, the authors demonstrated, using the radial diffusing assay in vitro, an antibacterial role of this peptide against a broad range of gram-positive bacteria, such as B. Megaterium, B. Subtilis, M. Luteus, (Krause et al., 2000). This antimicrobial effect was further confirmed by a second study that demonstrated the bactericidal action of hepcidin-25 against E. Coli and S. Aureus. Curiously, this latest study highlighted that one shorter form of hepcidin, hepcidin-20, presented a stronger antimicrobial activity than hepc-25. This shorter form was able to kill not only E. coli, S. Aureus but also S. Epidermis, group B streptococcus and C. Albicans (Park et al., 2001b). Hepcidin microbicidal function, described in these early studies (Krause et al., 2000; Park et al., 2001a), is today largely challenged. These studies have showed the hepcidin antimicrobial role in vitro, using doses that are strongly higher than the physiological ones. Furthermore, hepcidin antimicrobial activity has been never demonstrated in vivo in higher vertebrates. Hepcidin shows a strong structural similarity to AMPs, and in particular to a group of specialized microbicidal peptides: the defensins. Moreover, in some fish species hepcidin-like peptides have antimicrobial functions in vitro and their expression can be induced in the liver or in other tissues in response to infectious challenges in vivo (Bao, et al. 2005; Douglas et al., 2003; Hu et al., 2017; Shike et al., 2002; Shike et al., 2004). Interestingly, in some fishes hepcidin-like peptides show only antimicrobial properties while, other species display both antimicrobial and iron regulatory functions (J. Shi & Camus, 2006). These evidences suggest that hepcidin may have evolved in early vertebrates as an antimicrobial peptide that became secondarily involved in iron homeostasis. In addition, in higher vertebrates, hepcidin expression is induced by infectious/inflammatory stimuli (Nicolas et al. 2002) (see below, hepcidin regulation). Taken together, these facts suggest that hepcidin could be an ancestral link between iron homeostasis and innate immunity.

INTRODUCTION

1.3. Hepcidin: iron regulatory hormone Today, hepcidin is rather defined as the key iron regulatory hormone. Iron is indispensable for almost all the organisms, required for many fundamental biological processes as oxygen transport and cellular respiration thanks to its electron-exchanging properties. However iron excess can be dangerous since it can promote the production of Reactive Oxygen Species (ROS), e.g. through the Fenton reaction. Therefore iron homeostasis must be tightly regulated in order to ensure adequate iron supply but not iron excess. To date, there are no known active mechanisms of iron excretion and organisms regulate their iron homeostasis controlling the iron absorption, recycling and store. Hepcidin is the hormone implicated in this regulation, since it manages the main processes involved in the supply of plasma iron: (i) the duodenal iron absorption, (ii) the macrophage-mediated iron recycling and (iii) the iron mobilization from the iron storing cells (Figure 3).

Figure 3. Hormonal hepcidin function.

1.3.1. Iron metabolism 1.3.1.1. Iron absorption The principal forms of dietary iron are heme iron, ferritin and inorganic iron (Ganz, 2013). All these forms require different uptake transporters at the apical membrane of enterocytes (Figure 4).

Heme iron is more easily absorbed compared to inorganic iron. In 2005, a study has identified a heme carrier protein (HCP1) expressed at the brush-border membrane of duodenal enterocytes

INTRODUCTION

(Shayeghi et al., 2005). Today, this transporter is rather known to be involved in the folate absorption, thus the contribution of HCP1 to heme transport remains to be clarified (Qiu et al., 2006). Once absorbed, heme is degraded by the heme oxygenase (HO) and the ferrous iron (Fe2+) is released into the cytoplasm of enterocytes. (Raffin et al., 1974).

Inorganic iron, principally released from vegetables and cereals, is mainly present in form of ferric iron (Fe3+). Divalent metal transporter 1/Natural resistant associated macrophages protein 2 (DMT1/NRAMP2), is a Fe2+/H+ co-transporter mainly expressed at the brush border of enterocytes (Canonne-Hergaux et al. 1999). DMT1 is the key carrier required for inorganic iron uptake through apical membrane of duodenal enterocytes (M. D. Fleming et al., 1998; Gruenheid et al., 1999; Gunshin et al., 1997). Since DMT1 transports Fe2+ ions, to fulfil its function this transporter must be associated to a reductase activity. Mckie and colleagues have identified the principal reductase required for this process, duodenal cytochrome B dcytB (McKie et al., 2001).

Iron released into the enterocyte cytoplasm can be used for enzyme synthesis, stored in ferritin or directly transported across the basolateral membrane of the enterocytes through ferroportin, the only known iron exporter (see below).

Figure 4. Iron absorption. Alimentary iron is absorbed by enterocytes. Heme iron crosses the apical membrane through a specific transporter, still uncertain. Then HO degrades heme allowing the release of iron. DcytB reduces inorganic iron, allowing its absorption via DMT1. In the cytosol iron can be used for various cellular function, stored in ferritin or exported through FPN. Dcytb: duodenal cytochrome B; DMT1: divalent metal transporter1; HO: heme oxygenase; FPN: ferroportin; Tf: transferrin.

INTRODUCTION

1.3.1.2. Iron recycling The other important mechanism that supports plasma iron amount per day is the macrophage- mediated iron recycling. Macrophages can recycle iron from two different sources: via phagocytosis of senescent red blood cells (erythro phagocytosis EP), (Bratosin et al., 1998) or via endocytosis of heme- and haemoglobin- associated proteins (Ganz, 2012). Heme is released and catabolized by heme oxygenases (HO-1 and/or HO-2) leading to ferrous iron, carbon monoxide and biliverdin production (Ryter, Alam, & Choi, 2006). The intracellular compartments involved in these different steps of iron recycling are still not totally certain (Figure 5).

Similarly to enterocytes, once Fe2+ is released into the cytoplasm, macrophages can use it, store it in ferritin or export it through ferroportin.

Figure 5: Iron recycling. Macrophages phagocyte aged red blood cells and degrade them in the phagosome, represented as here a circle inside the cell. Heme is released from hemoglobin and translocated into the cytoplasm, via a mechanism involving the heme transporter HRG1 (White et al., 2013). Heme is catabolized by HO and the iron produced via this process is either excreted from macrophages, via FPN, or stored intracellularly, by ferritin. FPN: ferroportin; Tf: transferrin.

1.3.1.3. Iron export and extracellular transport In order to achieve iron absorption and recycling, macrophage and enterocyte require another iron carrier to ensure Fe2+ transfer to plasma. To date, the only known iron exporter is Ferroportin (FPN). This protein is a metal transporter, expressed both at the basolateral membrane of enterocytes and at the macrophage membrane (Canonne-Hergaux et al. 2006; Donovan et al. 2000). FPN allows the entry of iron to systemic circulation, therefore its regulation is essential to ensure the right plasma iron level. Actually, FPN expression is regulated at the transcriptional (e.g. HIF2α, Nrf2), post-transcriptional (by the IRE/IRPs system) and post-translational level (hepcidin/FPN interaction), (see regulation of iron

INTRODUCTION metabolism). FPN-mediated iron export is physiologically associated with a ferroxidase activity, because of Fe3+ affinity of the iron extracellular carrier (see below). This oxidation is catalysed by a family of copper containing proteins, that include ceruloplasmin and hephaestin (Cherukuri et al., 2005; Harris et al., 1999; Vulpe et al., 1999). Once exported and oxidized, iron must be transported, though systemic circulation, to the other tissues. This transport is mediated by a plasma iron transporter Transferrin (Tf) which can bind two atoms of iron, forming the holotransferrin (holoTf). This binding is indispensable for maintaining ferric iron in a redox-inert state, preventing the generation of toxic radicals. Tf iron affinity is pH- dependent: higher for pH of 7,4 and lower in acidified environment (Gkouvatsos et al. 2012).

1.3.1.4. Cellular iron uptake Except for some specific cells, such as enterocytes or macrophages, other cells acquire iron through the TfR1/holoTf system (Figure 6). Transferrin receptor 1 (TfR1) is a transmembrane glycoprotein that interacts with holo-Tf. This binding triggers clathrin coated pits-mediated endocytosis of TfR1-holoTF (Ponka & Lok, 1999). The acidification of the endosome allows Fe3+ release from Tf. Once liberated, Fe3+ is reduced by a ferric reductase and exported in the cytoplasm through DMT1 (Fleming et al. 1998; Gruenheid et al. 1999), Finally the Tf/TfR1 complex is recycled back to the cell membrane, where the Tf is released to the bloodstream. Once acquired, iron fate depends on the cell type and on the iron requirement of the cell for its biological processes. Cellular iron that is not immediately required can be sequestered in cellular iron store complexes.

Figure 6: Cellular iron uptake. After binding to TfR1 (depicted in black), the holotransferrin (blue) is endocytosed in cells. In the endosome the acid pH favorizes the Fe3+ release. Then ferric iron is

INTRODUCTION

reduced and exported to the cytosol through DMT1. Finally, the endosome merges the cellular membrane allowing the TfR1 recycling and the Tf release. DMT1: divalent metal transporter 1.

1.3.1.5. Cellular iron store Once absorbed, ferrous iron reaches the cellular pool of free iron, i.e. the labile iron pool (LIP). Intracellular iron trafficking is still not well characterized; however, it is well established that iron not used for cellular functions must be stored to avoid free radical generation and cellular damages. Ferritin, a globular protein highly conserved throughout evolution and ubiquitously expressed in the cytosol of cells, is the intracellular iron storage molecule in both eukaryote and prokaryote organisms. Ferritin has 24 subunits of H- and L- chains that form a shell-like cage with a large cavity able to store up to 4500 iron atoms. Free intracellular ferrous iron is transported to ferritin, by a specific chaperone, PCBP1 (H. Shi et al., 2008). Iron release and mobilization occur through gated pores (Theil et al. 2006) or though ferritin degradation (Aisen et al. 2001). How iron is transported to the iron exporter (ferroportin) or to other cellular compartment for iron usage, is still unknown.

1.3.2. Regulation of the iron homeostasis Iron homeostasis presents a transcriptional (HIF 2α, Nrf2), post-transcriptional (IRE/IRP system) and post-translational regulation. In some cases, these three levels of regulation crosstalk between themselves, giving rise to a highly tuned “multilayered” regulation of iron homeostasis. I will focus here on the post-translation regulation by the hepcidin/ferroportin axis.

1.3.2.1. The hepcidin/ferroportin axis and the systemic iron homeostasis Despite transcriptional and post-transcriptional regulation at the cellular level, systemic iron homeostasis is mainly controlled by the hepcidin/ferroportin axis (Figure 7). Hepcidin ensures the post-translation regulation of FPN, regulating the iron export in the bloodstream and therefore the plasma iron level (Nemeth et al. 2004). The N-terminus end of hepcidin (Clark et al., 2011; Nemeth et al., 2006) interacts with an extracellular loop of ferroportin (Drakesmith et al., 2005; Fernandes et al., 2009), triggering the iron exporter ubiquitination (Qiao et al., 2012), internalization and degradation (Nemeth et al. 2004). A recent study describes hepcidin- mediated ferroportin occlusion, as a new mechanism to avoid iron export. Taking advantage of FPN gain-of-function mutants, resistant to ligand-induced endocytosis, the authors demonstrated that the ubiquitination/endocytosis-mediated degradation of ferroportin is not the only mechanism of action of hepcidin (Aschemeyer et al., 2018). Therefore, by inducing FPN

INTRODUCTION degradation and/or occlusion, hepcidin blocks the only known iron exporter avoiding duodenal iron absorption, macrophage iron recycling and iron mobilization from iron storing cells, such as hepatocytes; therefore hepcidin by decreasing the serum iron level, works as the hyposidermic hormone.

Figure 7: The hepcidin/FPN axis. When required, iron (here represented as grey balls) is absorbed, released after recycling and from intracellular store (left). On the contrary in case of iron overload, inflammation or infection hepatocyte-derived hepcidin mediates FPN degradation, thus the decrease in serum iron level. FPN: ferroportin; hepc: hepcidin.

1.4. Hepcidin/infection axis Although the antimicrobial function of hepcidin is today largely debated, this hormone is tightly associated to the infection, via its iron regulatory role. All forms of life use similar kind and quantities of organic and inorganic components, taking part to a nutritional battle (Weinberg, 1975). As iron is required for various host and pathogen important biological processes, such as carrying or storing oxygen, catalysing metabolic, signaling-related, and antimicrobial redox reactions (Ganz, 2013; Schaible & Kaufmann, 2004), multicellular organisms and microbial invaders strongly compete for it (Drakesmith & Prentice, 2012). In this battle, both host and pathogens have developed different strategies to overcome iron acquisition of adversary. Pathogens have evolved a wide range of mechanisms to scavenge iron from iron-binding proteins of the hosts. Principally bacteria produce strain-specific iron chelators, called siderophores, which remove the iron from the host sources. These bacterial iron chelators have a strong affinity to iron, that allows the competition with host transferrin. Otherwise bacteria, especially gram positive one, have developed some heme-uptake system to free haemoglobin from erythrocytes and use heme as a nutrient source. Finally, some bacteria present some host binding-iron proteins uptake system, such as transferrin and/or lactoferrin uptake systems, that allow a direct acquisition of iron from key host iron-binding proteins (Figure 8). The high

INTRODUCTION number of bacterial genes upregulated in case of iron restriction highlights the relevance of iron for pathogen survival (Ratledge & Dover, 2000).

Figure 8. bacterial strategies for iron acquisition. On the left, a gram-positive organism can obtain iron using hemoglobin or heme, that are shuttled through the cell wall (in ochre) and then transported across the cell membrane via ABC-type transporter. Otherwise these bacteria secrete siderophore that capture iron and then reenter the cell. On the right a Gram negative, showing siderophore, heme/hemoglobin receptor and a transferrin/lactoferrin transporter. TF: transferrin; LF: lactoferrin. Adapted from Cassat J E, 2013 (Cassat & Skaar, 2013)

On the other hand, to limit iron availability for pathogens, hosts tightly bind iron to some chaperones. The main iron-binding proteins are transferrin (TF), lactoferrin (LF) and siderocalin (or lipocalin-2). The first two proteins belong to the transferrin family of proteins. As described above, the transferrin iron binding ensures a safe iron carrying in the organism and facilitates cellular iron uptake. Lactoferrin, highly present in mucosal secretions and in neutrophil granules binds iron with high affinity; (see chapter 3 ) (Masson et al. 1969). Lactoferrin iron-chelation has, as main aim, an innate immune protection in case of infection, trauma or tissue injury (Baker et al. 2012; Masson et al. 1969). Lactoferrin, unlike transferrin, maintains its iron-binding capacity even at low pH, and is therefore a strong iron scavenger in

INTRODUCTION acid infectious site (Baker et al. 2012). Siderocalin prevents the microbe iron acquisition by binding microbial iron-loaded siderophores. A study demonstrated that siderocalin binds a Escherichia coli siderophore, protecting host from infection (Flo et al., 2004).

Despite the iron-binding efficacy of these chelators, in response to many infections hosts usually reduce serum iron level, decreasing iron absorption and increasing iron store in macrophages. This general strategy, defined as hypoferremia of infection, is an ancestral host- defense mechanism, based on the host withholding of iron to fight against pathogens (Cartwright et al. 1946). As a key hyposidermic hormone, hepcidin has a role in this host defense process. Common bacterial and viral infections, such as streptococcus pneumoniae and influenza A virus, as well as pathogen-derived TLR agonists, (HKLM, Pam3CSK4, LPS, flagellin) rapidly up-regulate hepatic hepcidin expression, through a IL-6 signaling pathway (see section 1.5.2.) (Ganz & Nemeth, 2015; Gaël Nicolas et al., 2002; Rodriguez et al., 2014). There is also another molecular mechanism, hepcidin-independent, playing a central role in hypoferremia of infection. Actually, various works demonstrate that in response to TLR ligands, such as LPS, FSL1, PAM3CSK4, a rapid transcriptional down regulation of Fpn occurs in liver and in spleen and that this fast down regulation is sufficient to ensure plasma iron decrease (Deschemin and Vaulont 2013; Guida et al. 2015). Whatever the molecular mechanism regulating it, hypoferremia of infection is defined as general feature of antimicrobial host response.

Numerous studies have shown a protective effect of hepcidin-induced hypoferremia in mice infected with Vibrio vulnificus, Yersinia enterocolitica and Klebsiella pneumoniae. These pathogens are extracellular siderophilic bacteria, whose growth require or is facilitated by free iron. Michels and colleagues have shown that hepcidin-induced hypoferremia protects against Klebsiella pneumoniae infection. They first show that increased iron availability worsens infection outcome. Then comparing the virulence of infection in WT and hepcidin knock out mice, the authors found that the lack of hepcidin triggers a more severe infection. To discriminate if the susceptibility depends on the hepcidin deficiency or on the associated iron overload, the authors injected WT mice with iron. They observed a higher infection dissemination, despite an increased hepcidin level in plasma, thus they conclude that the elevated plasma iron, and not hepcidin deletion, enhances bacterial spread (Michels et al., 2017). Similarly, Stefanova and colleagues showed that hepcidin mediates resistance to

INTRODUCTION

Yersinia enterocolitica infection, by reducing the circulating iron concentration, however they observed that both Staphylococcus aureus and Mycobacterium tuberculosis infection were insensitive to plasma iron variations. This work describes hepcidin protective role against siderophilic extracellular bacteria by a clearance of non-transfrerrin-bound iron (NTBI) underlining the importance of NTBI only for certain pathogen growth (Stefanova et al., 2017a). Consistently, early studies have described an higher susceptibility to siderophilic bacterial infection, such as Vibrio vulnificus, Yersinia enterocolitica in hemochromatosis patient (Bergmann et al. 2001; Gerhard et al. 2001; Höpfner et al. 2001).

Although common, hypoferremia does not always occur and its contribution to host resistance is infection-dependent. For example, during hepatitis C virus (HCV) and hepatitis B virus (HBV) acute infection, hepcidin is not induced (Armitage et al. 2014). Contrary to what was previously published, that is a rapid increase of hepcidin and hypoferremia of infection, this study highlights, for the first time, that hosts do not always adopt hypoferramia as a defensive mechanism.

Even if the hypoferremia of infection is defined as a “general antimicrobial host defense”, this process is not always induced (Armitage et al. 2014; Girelli et al. 2009; Lin et al. 2013), or can be useless, as shown by Stefanova and collegues for Staphylococcus aureus and Mycobacterium tuberculosis (Stefanova et al., 2017b) and can even be counter-productive against infection. Depending on pathogen tropism, extracellular iron availability affects more or less infection development. Indeed, hypoferremia induces plasma iron decrease but an increase in macrophage iron retention. Therefore, this strategy is a double-edged sword, because macrophage intracellular bacteria, such as Salmonella, Leishmania, and Mycobacterium, will take advantage of intracellular iron store (Chlosta et al., 2006; Olakanmi et al., 2006). In this kind of infection, hosts must modify the iron intracellular localization rather than promote the macrophage intracellular iron-store. Nramp1, Natural Resistance Associated Macrophages Protein 1 (Vidal et al. 1995), has an important role during this kind of infection. This protein is an iron transporter present at the phagosome membrane of macrophages. The key function of this transporter is to reduce ions availability in endosome, limiting pathogen growth in this compartment (Canonne-Hergaux et al. 1999; Jabado et al. 2000). Mice with a loss of function mutation in Nramp1 gene have a major susceptibility to specific intracellular bacteria (Canonne-Hergaux et al. 1999). Moreover, in case of systemic hypoferremia, iron loaded

INTRODUCTION macrophages can have an impaired potential to kill some bacteria and viruses (Theurl et al., 2005); therefore, hypoferremia can furthermore inhibit immunological macrophage functions. In conclusion there is a delicate balance between iron availability and risk of infection. The regulation of hepcidin expression integrates bacterial ability to scavenge iron from the host iron- binding protein, intra- or extracellular bacterial tropisms, bacterial cell type and anatomical niches and iron-mediated modulation of immune cells. Therefore, host control of hepcidin expression can be variable between different infections but must be always finely regulated in order to control iron availability and localization.

1.5. Hepcidin regulation The main level of hepcidin regulation is at the transcriptional level. Like other hormones, hepcidin is feedback-regulated by the substance it regulates. Beyond the iron level, inflammation upregulates hepcidin (Gaël Nicolas et al., 2002), while hypoxia (Gaël Nicolas et al., 2002) and erythropoietic drive (Kautz et al., 2014) suppress the expression of this hormone (Figure 9).

Figure 9. Hepcidin regulation. The iron overload and inflammation promote the expression of this hormone (green arrows), while the hypoxia and the erythropoiesis (red arrows) repress its expression.

1.5.1. Iron-mediated hepcidin expression Hepatic hepcidin expression is positively regulated in response to circulating iron. This regulation requires extracellular iron sensors and is quite different from the tissue iron-store regulation (Corradini et al., 2011a; Kautz et al., 2009). The molecules involved in this regulation are TFR2 and HFE (Figure 10).

INTRODUCTION

HFE is an atypical major histocompatibility (MHC) class I-like molecule while TFR2 an homolog gene of TFR1, mainly expressed in the liver. Both TFR2 and HFE were initially described as molecules mutated in hereditary hemochromatosis (HH) patients (Camaschella et al., 2000; J. N. Feder et al., 1996). Later, the pathological iron overload associated to these mutations, was explained by a down regulation of hepatic hepcidin expression in both human (Bridle et al., 2003; Nemeth et al., 2005) and mice (R. E. Fleming et al., 2002; Kawabata et al., 2005; Wallace et al., 2009). Although the molecular mechanism of action is still largely debated, HFE and TFR2 are involved in hepcidin modulation in response to circulating iron increase. Ramos and colleagues demonstrated that HFE and TFR2 knock out animals are totally able to suitably respond to chronic iron challenge, while they are not able to upregulate hepcidin expression after an acute dietary iron challenge, (Ramos et al., 2011).

Iron increase, either in the circulation or in hepatic intracellular stores, induces hepcidin expression via the bone marrow morphogenetic (BMP)/SMAD pathway (Figure 10).

While in vitro, different BMP ligands, such as BMP2, BMP4, BMP9, are able to trigger hepcidin transcription (Babitt et al., 2007; Truksa et al., 2006; Wang et al., 2005), various studies demonstrate that only BMP6 is necessary and sufficient for the BMP-mediated hepatic hepcidin induction in vivo. First of all, iron rich diet up regulates BMP6 mRNA transcription and induces the functional activation of BMP/SMAD pathway (Kautz et al. 2008). Moreover, mice lacking BMP6, show iron overload and a strongly decreased hepcidin expression. These evidences highlight the fact that, in mice, BMP6 plays an essential and non-redundant role in iron homeostasis regulation (Andriopoulos Jr et al., 2009; Meynard et al., 2009). BMP6 binds to BMP receptors, allowing the phosphorylation of SMAD1/5/8 that in in turn bind to SMAD4. This SMAD complex translocates into the nucleus where it mediates hepcidin transcription (Wang et al., 2005) by binding to BMPs responsive element in HAMP promoter (Casanovas et al. 2009; Truksa et al. 2009; Verga Falzacappa et al. 2008).

At the plasma membrane of hepatocytes there is a complex molecular platform that allows BMP/SMAD signaling and hepcidin expression. This molecular platform includes: bone morphogenetic proteins receptors (BMPRs), hemojuvelin (HJV), matriptase-2 and noegenin.

The BMP-6 receptors involved in this signaling are Alk2 and Alk3 (Steinbicker et al., 2011). These receptors interact with a co-receptor, the hemojuvelin (HJV). The formation of this

INTRODUCTION complex is pivotal for the expression of hepcidin by the BMP/SMAD signaling pathway. Indeed, mutations of HJV determine a decreased hepcidin expression (Babitt et al., 2006, 2007) allowing to the development of the juvenile hemochromatosis (Papanikolaou, et al. 2004). While HJV positively regulates this signaling pathway, the matriptase 2 is the main negative regulator of this platform. Indeed, the loss of this protein causes an inappropriate high hepcidin expression, causing FPN degradation and inhibition of iron absorption (Du et al., 2008). Neogenin, is another protein present at the BMPR complex, interacting with HJV and matriptase-2; its exact role in hepcidin regulation remains incompletely understood.

In conclusion, increase in circulating and/or store iron triggers a complex and not completely clarified network of signaling pathways that increase hepcidin expression and prevent iron overload. This feedback loop is modified in case of infection and inflammation.

1.5.2. Inflammation-mediated hepcidin upregulation Many studies have shown an hepcidin induction in response to inflammatory agents such as turpentine oil, TLR4 (Toll-like receptor 4) agonist LPS, pathogen and pathogen-derived molecules (Lee et al., 2004; Gaël Nicolas et al., 2002; Pigeon et al., 2001; Rodriguez et al., 2014). IL-6 is the main cytokine required to trigger hepcidin expression in response to inflammatory/infectious stimuli both in vitro and in vivo (Nemeth et al., 2003, 2004). The ligand binds IL-6 α subunit and gp130.9 receptor complex, activating associated Janus kinases (JAKs) that phosphorylate Signal Transducers and Activators of Transcription (STAT) proteins (Heinrich et al. 2003). Two different groups described STAT3 binding to hepcidin promoter, integrating the inflammatory IL-6 mediated molecular signaling involved in hepcidin expression (Verga Falzacappa et al. 2007; Wrighting et al. 2006) (Figure 10). Consecutive studies find out a IL-22 mediated JAK/STAT3 dependent hepcidin upregulation in hepatoma cell line Hep3B and HepG2 (Andrew E Armitage et al., 2011) and in liver (Smith et al., 2013). Although IL-6/JAK/STAT signaling has a central role in inflammatory regulation of hepcidin, other cytokines, such as IL-1β and activin-B, can promote hepcidin expression in inflammatory conditions via the BMP/SMAD signaling pathway (Besson-Fournier et al. 2012; Canali et al. 2016; Shanmugam et al. 2015). Tissue injury that, usually occurs in case of infection or inflammation, determines endoplasmatic reticulum (ER) stress. In 2009 it was described hepcidin induction in response to ER stress (Vecchi et al., 2009). Therefore inflammation, infection and tissue damage promote hepatic hepcidin expression, leading to hypoferremia and

INTRODUCTION even, when this condition becomes chronic, to anaemia of inflammation (AI). AI is a multifactorial acquired disorder of iron homeostasis characterized by low plasma iron levels and high iron retention in macrophages. Despite a normal level of erythropoietin (EPO), a key cytokine required for the erythropoiesis, the AI-determined serum iron restriction, inevitably, reduces red blood cells production. The main role of hepcidin induction in the development of AI is demonstrated by the fact that in hepcidin knock out animals, model of AI results in a mild anemia, reasoning that hepcidin is not the only but nevertheless an important agent that trigger AI (Kim et al., 2015).

1.5.3. Negative regulation: Hypoxia and erythropoiesis Hepcidin transcription is strongly reduced in case of iron deficiency, hypoxia and erythropoietic drive. Iron deficiency suppresses hepcidin expression in hepatocytes. The key negative regulator of BMP6/SMAD signaling is the matriptase 2, but the precise molecular mechanism of action is still largely debated. Moreover, in response to iron deprivation Bmp6 expression is decreased, suggesting another negative regulatory mechanism (Kautz et al., 2008). Hypoxia suppresses hepcidin expression (Nicolas et al. 2002). Hypoxia is indissolubly associated with erythropoietic drive. In 2012, it has been reported that HIF-2-mediated hepcidin downregulation required an erythropoietin-induced erythropoiesis (Liu et al., 2012; Mastrogiannaki et al., 2012). A recent study has identified a new key bone marrow derived hormone, erythroferrone (ERFE), acting in erythroid-mediated hepcidin regulation. This hormone is induced by EPO treatment and inhibits hepcidin in vitro (Kautz et al. 2014). Despite this study, to date little is known about the “erythroid regulator” pathway (Figure 10).

INTRODUCTION

Figure 10. Signaling pathway involved in regulation of hepcidin expression. Iron accumulation either in the serum and in hepatocytes induces hepcidin expression via BMP/SMAD pathway. TFR2 and HFE have a pivotal role in circulating iron-mediated hepcidin regulation. Pathogens are recognized, triggering the release of inflammatory cytokines, IL6 and IL22. These cytokines determine the up regulation of hepcidin via STAT3. Otherwise other inflammatory cytokines, IL1 β and activin B, promote hepcidin expression by BMP/SMADs pathway. The molecular signaling pathways of hepcidin down regulation are, to date, less characterized. The lack of the positive regulator, BMP6, may contribute to hepcidin down regulation. In case of iron deficiency, matriptase, TMPRSS6, is the main negative regulator involved in the suppression of hepcidin. Hypoxia and erythropoiesis are tightly associated and both are involved in the suppression of hepatic hepcidin expression, but the signaling pathways are still not totally elucidated.

1.6. Extra-hepatic hepcidin and other roles. As suggested by its name, hepcidin is mainly produced in the liver, by hepatocytes (Park et al., 2001b; Pigeon et al., 2001). However, experiments of Escherichia coli subcutaneous pouch infection model in mice revealed hepcidin expression at the infection site, suggesting for the first time, that hepcidin could be directly express by others cells different from hepatocytes, such as inflammatory cells recruited to the site of infection (Motley et al., 2004). Starting from this first observation, more and more studies have revealed hepcidin expression in heart, lung, leukocytes, kidney in physiological and pathological conditions (Chen et al., 2014; Houamel et al., 2016; Lakhal-Littleton et al., 2016; Merle et al., 2007; Moulouel et al., 2013; Peyssonnaux

INTRODUCTION et al., 2006). In order to discriminate the importance of the extra hepatic hepcidin contribution to systemic iron homeostasis, our laboratory developed a new mouse model in which the hepcidin gene can be conditionally inactivated: mice with floxed hamp1 alleles (with loxP sequences flanking exons 2 and 3 of the gene) allowing tissue-specific gene deletion with the classical Cre-loxP strategy.

A liver-specific knockout mouse model has been first compared with total KO mice. This recent study demonstrated that hepatic hepcidin is sufficient in physiological conditions to ensure systemic iron homeostasis suggesting that extra hepatic hepcidin could have some local roles, especially in pathological conditions (Zumerle et al., 2014) (Figure 11).

Figure 11. Local role of extra hepatic hepcidin. The liver- derived hepcidin is sufficient for ensuring the systemic iron homeostasis and the other organs-derived hepcidin do not contribute to the reservoir of systemic hepcidin (dashed lines). These evidences suggest a local role of this extra hepatic hepcidin.

1.6.1. Leucocytes The production of hepcidin at the site of subcutaneous infection (Motley et al., 2004) and the hepcidin antimicrobial ancestral origin prompted Peyssonnaux and colleagues to investigate if leukocytes could expressed hepcidin in response to infection (Peyssonnaux et al., 2006). The authors demonstrated that isolated murine neutrophils and macrophages upregulate hepcidin

INTRODUCTION expression after Group A streptococci (GAS) and Pseudomonas Aeruginosa (PA) infection in a TLR4 dependent fashion in vitro and in vivo (Peyssonnaux et al., 2006). The lack of TLR4 affects the myeloid but not the liver expression of hepcidin in infected mice, suggesting myeloid specific transcriptional regulation of hepcidin in response to these pathogens. The authors suggest that leukocyte hepcidin production may alter iron availability for pathogens at the infection foci.

1.6.2. Heart The central role of iron in cardiac pathophysiology prompts to investigate hepcidin regulation in this organ. First Merle and colleagues described the expression of hepcidin in the rat heart, showing that this expression was induced by hypoxia and turpentine. While the inflammatory agent (turpentine) shows the same positive hepcidin regulation in both liver and heart, hypoxia displays a mirror effect on hepcidin regulation that is negative in the liver and positive in the hart. The authors suggest a cardiac hepcidin involvement in heart diseases (Merle et al., 2007). Then, the hepcidin-FPN axis involvement in heart pathophysiology was studied taking advantages of two different mouse models: the cardiomyocyte hepcidin knock-out mice and the hepcidin-resistant ferroportin knock-in model. Both murine models display no abnormalities at the systemic iron level, as expected (Zumerle et al., 2014), but develop a fatal heart contractile dysfunction caused by myocardiocyte iron deficiency. This abnormality is not observed in total hepcidin knock out mice because of iron overload characteristic of these animals. Therefore this team underlined for the first time a cell autonomous role of the hepcidin-FPN axis (Lakhal- Littleton et al., 2015).

1.6.3. Lung The lung epithelial cells express hepcidin and induce it after IL-6 and IFN γ stimulation in vitro (Frazier et al., 2011). This evidence prompts laboratories to investigate the role of this peptide during lung injury and infection. Chen and colleagues observed that the airway epithelial cell- derived hepcidin was upregulated during a murine model of polymicrobial sepsis and that deleting airway epithelial hepcidin by short hairpin RNA injection increased the lung injury and pulmonary bacterial infection, augmenting the mortality. The authors observed an alteration in intracellular iron metabolism after hepcidin silencing, reasoning that the lower macrophage retention of iron could affect the phagocyte activity and explain this phenotype (Chen et al., 2014).

INTRODUCTION

1.6.2. Kidney Since hepcidin was originally isolated from human urine (Park et al., 2001b), different studies investigated hepcidin expression in kidney. Hepcidin is expressed in human, mouse and rat kidney with a similar pattern. The hormone is expressed with a specific pattern along renal tubules, showing higher expression in cortical and medullary thick ascending limb (D. Houamel et al., 2016; Kulaksiz et al., 2005; Moulouel et al., 2013). The renal expression of hepcidin suggests a role of this peptide in physiological renal tubules system, controlling the expression of iron transporters to prevent renal iron overload (Kulaksiz et al., 2005; Moulouel et al., 2013) and in pathological renal antibacterial defense (D. Houamel et al., 2016). In regard to this latter function, the authors suggested hepcidin pleiotropic effects against uropathogenic Escherichia coli, since hepcidin modulates iron content and inflammatory response in kidneys of infected mice. This work is based on comparison between wild type and total hepcidin knock out mice, therefore it is difficult to discriminate the effect of local hepcidin lack and systemic iron overload.

Hepcidin is structurally close to β defensins, a subgroup of antimicrobial peptides (AMPs). The skin together with the intestine are the principal sources of antimicrobial peptides. While studies have shown that other epithelia -such as respiratory and renal (Frazier et al., 2011; Houamel et al., 2016) epithelia- express hepcidin in response to infectious challenges, expression and putative role of hepcidin in the skin has never been investigated.

INTRODUCTION

2. SKIN

The skin has various roles in mammals. Skin is a tough, elastic, waterproof body cover protecting the organisms against dehydration, injury and infection. To fulfill all these functions, the skin presents a specific architecture and comprises different cell types.

2.1. Mammalian skin architecture Mammalian skin is composed mainly by two different elements: the epidermis and the dermis that are tightly associated through the basement membrane.

Associated to this organ there are different appendages, such as hair, nail, sebaceous and sweat glands (Figure 12). The distribution and the abundance of these appendages differ among mammal species and in the same organisms between different body sites. Hairs and nails are epidermal derivate composed by tough keratin. The hairs cover body, more or less extensively according to the species, ensuring protection against environmental factors. Nails protect the ends of the digits against wear and tear. The sebaceous glands are associated to hair follicles. These glands produce and secrete, through a holocrine pathway, a lipid matter, named sebum. Through cell eruption sebum is released, coating skin surface and ensuring its waterproof feature. Finally, the sweat glands include the eccrines and apocrines glands. The eccrine sweat glands participate to body temperature regulation, through evaporation of sweat; while the apocrine ones are responsible for body odor (Richards, Bereiter-Hahn, & Matoltsy, 1984).

Skin appendages have specific developmental processes responsible for their histological structures that allow their functions. Despite their importance for homoeostasis and function of the skin, epithelial appendages will be not examined in this work.

INTRODUCTION

Figure 12. Mammalian skin architecture. Representation of the skin and its appendages. The squamous stratified epithelium, epidermis. The underlying connective tissue, dermis, is characterized by numerous fibers, blood vessel and mechanoreceptor. The sebaceous gland is associated to the hair follicle, while the sweet gland crosses all the tissue (here in green).

2.1.1. Dermis The dermis is the connective tissue associated to epidermis. The main dermal cells are fibroblasts. These cells are immersed in a gel-like matrix, highly vascularized and characterized by a meshwork of fibrils. Dermis represents both the nutritional- and mechanical-support of epidermis.

Indeed, the epidermis is not vascularized and receives all nutritional supports from dermis by passive diffusion. Moreover, cohesiveness, flexibility and extensibility of the dermal fibrous meshwork determine the mechanical properties of the skin and ensure the encase and stabilization of epidermal appendages (Montagna & Parakkal, 1974). The dermis is organized in three different parties: papillary, reticular and hypodermis. These three layers present different histological features. The papillary dermis is the closest to epidermis and has the higher number of fibroblasts. Beneath papillary dermis there is the reticular dermis, rich in collagen (mainly type I and III) and elastic fibers. This layer accounts for the skin elasticity and strength. Finally, the lower dermal component is the hypodermis, manly composed of white adipocytes (Watt & Fujiwara, 2014). Although fibroblasts are the main abundant cells, there

INTRODUCTION are a lot of immune cells associated to the dermis (see paragraph 2.1). Therefore the dermis fulfills not only nourishment and mechanical skin functions but has also a role in skin immunity.

2.1.2. The Epidermis 2.1.2.1. Epidermis Structure Epidermis is the outermost component of the skin, composed of the interfollicular epidermis (IFE) and the associated hair follicles (HFs) (Freedberg, Eisen, Wolff, Austen, & Goldsmith, 2003). Following the cyclic hair growth, HF is a very dynamic structure that is cyclically submitted to three different phases: an active growing phase (anagen), then a regression and shortening phase (catagen) and finally a resting phase (telogen) (Hardy, 1992). This cyclic and continuous process is ensured by some adult multipotent stem cells, located in a portion of HF, named bulge. Being multipotent, these stem cells contribute not only to HF but also to IFE homeostasis in adult skin in response to penetrating wound (Oshima, Rochat, Kedzia, Kobayashi, & Barrandon, 2001; Taylor, Lehrer, Jensen, Sun, & Lavker, 2000). Moreover, the HF is one of the main skin niches where occurs the dialogue between skin microbiota and host immune cells. Indeed, it was suggested that HF simplifies this interaction, favoring the catch of microbial metabolites by dermal innate immune cells (Belkaid & Segre, 2014). Therefore this specific epidermis structure mediates both the skin homeostasis and a proper response in case of infection and/or alteration of commensal flora, allowing commensal microbiota-resident immune cells dialogue. Despite the hair follicle importance for the skin homoeostasis and functions, this dynamic and essential structure will be not examined in this work. I will focus only on IFE description; to simplify I will write “epidermis” to refer to the interfollicular epidermis.

The epidermis is a squamous stratified epithelium made by keratinocytes, organized in various cell layers. In particular, this epithelium consists of one inner layer of undifferentiated cells and three upper layers of progressively more and more differentiated keratinocytes (Figure 13). Going from the innermost to the outermost epidermis layer, there are the basal, spinous, granular and cornified layers; these names reflect either the layer position in the epidermis or a structural property of the cells composing the layer. In each layer, keratinocytes are tightly interacting each other, through adherent junctions. These kind of intercellular junctions form extracellular E-cadherin bridges between the actin cytoskeleton of neighboring cells (Yap, Brieher, & Gumbiner, 1997). Except the basal layer adhesion to the basement membrane, all

INTRODUCTION the others keratinocyte sheets are bond each other through desmosomes. In contrast to adherent junctions, desmosomes associate with the intermediated filaments of keratinocyte cytoskeleton, keratin intermediate filaments (Kowalczyk, Bornslaeger, Norvell, Palka, & Green, 1999). Another specific junction characteristic of epidermis is the tight junction (TJ). These junctions have an extracellular component, made by transmembrane proteins occludin, claudin and junctional adhesion molecules, scaffolded by cytoplasmatic zonula occludens (ZO) proteins (Niessen, 2007). TJs are specifically present at the apical side of granular keratinocytes, forming a belt that prevents water loss and pathogen infection.

Figure 13. Epidermis structure and keratinocyte junctions. This epithelium is made by keratinocytes organized in various layers: the basal, the spinous, the granular and the cornified layer. The epithelium is associated to the underlying dermis through the basement membrane. Each layer is characterized by the expression of specific marker: keratin 14, 5 in the basal, keratin 1,10 and involucrin in the spinous and granular, filaggrin and loricrin for the granular and cornified layers (on the left). On the right, a schematic representation of the main keratinocyte junctions. In each layer, keratinocytes are associated thought adherent junctions (here represented in blue), while keratinocytes from two distinct layers are linked by desmosome junction (in yellow). KRT14: keratin 14; KRT5: keratin 5; KRT1: keratin 1; KRT10: keratin 10; INV: involucrin; FLG: filaggrin; LOR: loricrin; BM: basement membrane; AJ: adherent junction.

Epidermis is defined as a squamous stratified epithelium because moving from the basal to cornified layer, the cells undergo a dramatic structural remodeling, becoming, at the end of the

INTRODUCTION differentiation, flattened, lipid-covered, flake-like died cells, named corneocytes. This dramatic process of differentiation, that entails final cellular death, is necessary to develop an outermost tough and waterproof epidermis layer, allowing skin protective functions. The commitment to differentiation implies the growth arrest (Jetten & Harvat, 1997), therefore, contrary to the undifferentiated proliferative basal cells, keratinocytes in the upper layers are post-mitotics. Keratinocytes, undergoing their terminal differentiation, display different structures and properties among the different layers (Freedberg et al., 2003; Montagna & Parakkal, 1974).

Keratins are a heterogeneous group of proteins forming intermediate filaments. These proteins are the main component of the cytoskeleton of epidermis cells, so named keratinocytes (W. G. Nelson & Sun, 1983). These proteins are responsible of epidermis toughness. There are different kinds of keratins, differentially expressed according the differentiation state of the cells, therefore these molecules act as marker to distinguish the different epidermis layers.

The basal layer is composed by cuboid keratinocytes, mitotically and metabolically active. These keratinocytes express two different kind of keratins: keratin 14 (KRT14) and keratin 5 (KRT5) (E. Fuchs, 1990a; W. G. Nelson & Sun, 1983). These small keratins, respectively 50 kDa and 58 kDa, ensure the steady attachment of the epidermis to the dermis, allowing keratinocyte-basement membrane interaction.

In the spinous layer, the cells increase their size and mainly synthetize heavy 67kDa-keratin 1 (KRT1) and 56kDa-keratin 10 (KRT10) to form thin bundles (Einchner et al. 1986; Fuchs 1990). Moreover, these cells start the synthesis of envelope proteins: glutamine- and lysine- rich proteins, such as the involucrine, synthetized and stored inside keratinocytes of the spinous layer. These proteins will, once cells have migrated in the granular layer be released in the extracellular space (Rice & Green, 1979).

As spinous cells reach the granular layer, they stop generating keratin, and make their final protein synthesis, to achieve the synthesis of cornified envelope proteins. Therefore, they produce mainly filaggrin and loricrin (Fleckman et al.1985; Mehrel et al. 1990). These proteins are organized in granules, explaining the name of this compartment. Once reached the cornified layer, cells are dead, flatted, without any intracellular organelles but filled with keratin filaments (E. Fuchs, 1990a). The macromolecules produced in spinous and granular layers, are released and cross-linked to develop a seal, ensuring impermeable, insoluble, and highly

INTRODUCTION protective fortress of the external skin, keeping microorganisms out and essential body fluids in. Human and murine skin are almost close with some structural difference others each other. The murine epidermis is thinner than the human one and is mainly composed by HF, while human epidermis mainly characterized by IFE. As we will discuss in the paragraph 2.2, there are also some difference between murine and human skin immune cells (Pasparakis, Haase, & Nestle, 2014).

2.1.2.2. Keratinocyte Differentiation As discussed above, the mechanism of stratification of mammalian epidermis, tightly linked to keratinocyte differentiation, is indispensable to ensure skin structure and function.

The main extracellular factor involved in keratinocyte differentiation process is the extracellular calcium concentration. Indeed an extracellular gradient of calcium, increasing from the basal to the cornified layer, regulates keratinization and epidermal cell differentiation and stratification (E. Fuchs, 1990a). Although extracellular calcium gradient is required to modulate keratinocyte differentiation, it does not trigger basal cells commitment to differentiation. In 2005, Lechler and colleagues demonstrated, for the first time, that keratinocyte differentiation is triggered by the asymmetrical division of basal cells. They observed that at the basal layer, cellular divisions occur perpendicularly to the basement membrane. This division fashion is essential for the epidermis stratification, since mutant that made lateral division at the basal layer show stratification defects (Lechler & Fuchs, 2005). Moreover, they elucidate the mechanism by which the perpendicular spindle orientations favor stratification and epidermis differentiation. Actually, after a perpendicular division one cell daughter maintains and the other loses basement membrane contact. This kind of division provides a natural mechanism for an asymmetrical division, allowing cellular cycle arrest and commitment to differentiation for the daughter cell that moves and mitosis activity for the daughter cell that remains attached to BM (Lechler & Fuchs, 2005). Today, it is believed that the integrins at the BM have an essential role in allowing this asymmetrical division(Simpson, Patel, & Green, 2011). Although these studies highlight the cellular mechanism required to trigger cell commitment to differentiation, the model to explain the lineage of basal cells is still largely discussed. The two main models proposed are the hierarchical and the stochastic models. According to the hierarchical model, epidermal stem cells, give rise to several secondary proliferative cells, named trans amplifying (TA) cells. After few divisions TA are committed to differentiation, leaving cell cycle,

INTRODUCTION becoming post mitotically and moving through the other layers (Mackenzie, 1970; Potten, 1974). While, according to the stochastic model, basal cells are all the same and each division can yield three different outcomes: one differentiated daughter, and one progenitor or two basal progenitors or two differentiated daughters. Although the fate choices and mitotic spindle orientations are random, the probabilities of the three different outcomes are similar, so that the generation of differentiated cells and the maintenance of progenitor pools are balanced at the population level and long-term homeostasis is ensured (Clayton et al., 2007; Lim et al., 2013).

2.2. Skin Immunity As already described, the skin is the outermost barrier of the body. For a long time, skin has been defined as an inert physical barrier, today it is well accepted that skin ensures protection in an active way, through an interplay of three different barriers: the microbioma, the chemical- physical and the immune barrier (Figure 14). Each barrier has specific functions and interacts with the other barriers, ensuring host protection and, at the same time, skin integrity. Keratinocytes, as main epidermis cell type, have central role in constituting and coordinating all these barriers.

Figure 14. Level and components of skin barrier. Skin assures host protection through various barriers. The skin commensal microbes constitute the first defence against pathogens. In the cornified layer, keratinocytes avoid pathogen growth secreting lipids and decreasing the pH. Tight junctions (here represented as black sticks), characteristic of the granular layer, avoid to pathogens to invade the deeper tissues, developing a physical barrier. Finally, the immune barrier is composed by the plethora of innate and adaptive immune cells resident in the skin, such as keratinocytes, dendritic cells, macrophages…TJ: tight junction; DC: dendritic cell; ILC: innate lymphoid cell.

INTRODUCTION

2.2.2. Physical/chemical barrier The physical barrier consists of all the intercellular junctions between epithelial cells, especially tight junction (TJ) (Figure 14). Being specifically localized at the level of the granular layer, TJs make a quite exterior shield against pathogen invasion. The integrity of the physical barrier is mainly influenced by the skin microbiota (see paragraph 2.2.3) and by the chemical barrier (Eyerich, Eyerich, Traidl-Hoffmann, & Biedermann, 2018).

The chemical barrier consists of all the molecules that acidify the epidermis surface. This barrier includes amino acids and their derivatives resulting from the proteolysis of epidermal filaggrin (Eyerich et al., 2018). Actually, the chemical barrier works keeping acid the pH of outermost layers of epidermis. This low pH environment allows a better antimicrobial peptide activity (see section 2.2.2.3.1.) and barrier integrity. Several studies have illustrated that acid pH is required for a correct TJ assembly, underlining the indissoluble association between epidermis chemical and physical barriers (J. P. Hachem et al. 2005; Jean Pierre Hachem et al. 2003; Rippke et al. 2002).

Epidermal TJs form the physical barrier and keratinocyte secreted molecules constitute the chemical barrier. Therefore keratinocytes are essential for both physical and chemical barriers. Furthermore, keratinocytes are an “active” constituent of the immune barrier of the skin.

2.2.2. Immune Barrier Skin is equipped of a large range of resident immune cells that contribute to skin immunity. Human and murine skin immune barriers present similarities: both consist of dendritic cells (DCs), including epidermal Langerhans and dermal dendritic cells (dDCs), macrophages, mast cells, innate lymphoid cells (ILCs), skin-resident T cells. However, there are some differences: in murine skin the most prevalent epidermis immune cell type is a category of non-conventional T cells, named γδ T cells, while in human epidermis the most prominent populations are Langerhans cells and CD8+ T cells. (Pasparakis et al., 2014). These cells can be classified either according to their skin localization (epidermis and dermis associate immune cells) or according to their role in skin immunity (innate or adaptive).

2.2.2.1. Skin Adaptive Immunity Adaptive immune responses occur later in infection and are characterized by the highly antigen- specific activation of lymphocytes and contribute to the immunological memory.

INTRODUCTION

T cells are highly abundant in the skin, e.g. in normal human skin reside twice the number of circulating T cells (Pasparakis et al., 2014). These cells have a key role in skin immunity and are involved in some inflammatory skin diseases, such as psoriasis. After the interaction with some antigen presenting cells (APCs) in skin lymph node, naïve T cells give rise to effector memory cells (TEM) and central memory cells (TCM). During maturation TEM acquire, the skin homing receptors, CLA and CCR4. These cells migrate into the skin, especially at the infection site, ensuring pathogen clearance. These cells constitute long-lived sessile skin resident T cells that can, in case of re-exposure, ensure again a rapid pathogen clearance. At the same time, TEM expressing molecules for peripheral organs homing, migrate in non-cutaneous lymph node, such as intestinal, pulmonary and other peripheral lymph nodes. In this way skin immunization provide systemic immune protection. TCM cells remain in the skin draining lymph node for giving rise to a new population of skin homing TEM following a pathogen re exposure (Mann, 2014; Pasparakis et al., 2014).

2.2.2.2. Skin Innate Immunity Skin innate immune cells consist of macrophages, ILCs, mast cells, DCs and keratinocytes. Contrary to adaptive immune cells, all these cells rapidly sense invading pathogens, such as bacteria, virus, fungi, and act, through different mechanisms, as a first line of defense of the body.

To accomplish the role of sentinels, innate immune cells have some extra- and intra-cellular sensing systems, named Pattern Recognition Receptors (PRRs) that ensure microbial identification. The main PRRs include different classes: the Toll-like receptors (TLRs), the NOD-like receptors (NLRs), the C-type lectin receptors (CLRs) and the RIG-like helicases (RLRs). In particular, PRRs recognize some pathogen-associated molecular patterns (PAMPs), that are components of given classes of microorganisms (Pivarcsi, 2005). Among PAMPs there is, e.g. LPS, a cell wall component of gram-negative bacteria; mannan and zymosan, constituents of yeast cell wall; the peptidoglycan (PGN) or lipoteicoic acid (LTA) constituents of gram-positive bacteria. Each family of receptors includes various members recognizing the different PAMPs. Although a certain PAMP is present in a high number of microorganisms, innate immune response to a certain microorganism is highly specific and finely tuned, giving rise to a suitable immune response. This is possible because microorganisms have numerous PAMPs, so the response of a certain immune cell is the result

INTRODUCTION of the crosstalk between different PRRs-mediated signaling pathways of the same immune cell. Moreover in response to a specific microorganism, multiple innate immune cells, from the same or from different skin compartments, interact for killing microbes. Interacting, the different cells mutually shape and influence the ‘cytokine milieu’, giving rise to a unique possible outcome. Therefore the PRRs and cellular combinations give rise to a ‘combinative’ immune sensing, allowing the right outcome (Volz, Kaesler, & Biedermann, 2012). This fine regulation is very important to ensure the skin commensal microbiota development and maintenance (see below).

Macrophages reside in the adult skin, accomplishing the clearance of infectious agents via phagocytosis (reviewed in Pasparakis, Haase, and Nestle 2014). In early phase of infection, resident macrophages can also produce chemokines, recruiting leukocytes to the infection site. A study has shown the key role of skin resident macrophages in neutrophil recruitment during S. aureus infection (Abtin et al., 2010). In a second phase of infection, circulating macrophages can be recruited to the infection site, reaching the pool of resident leukocytes. This recruitment is essential for the infection outcome and for restoring tissue integrity. The role of recruited macrophages in the second phase of infection can be highly variable according to infection and its development; I will discuss further the role of recruited macrophages in the case of GAS subcutaneous infection model (see chapter 4).

Mast cells are granulocytes that, after interaction with antigen, release their granules with a large plethora of inflammatory mediators. Mast cells, like macrophages can have both pro- and anti- inflammatory activity in the skin, according to cytokine environment.

Innate lymphoid cells (ILCs) are a family of different innate immune cells, classifiable according to their cytokine expression. Group 1 ILCs are defined by their capacity to produce IFNγ; group 2 ILCs are able to produce T helper 2 associated cytokine, such as IL-5 and IL-13 and the last group, named group 3 ILCs include all the ILCs secreting T helper 17 associated cytokines, like IL-17 and IL-22 (Spits et al., 2013).

Skin dendritic cells (DC) include epidermal Langherans cells and dermal DC. These cells have a key role in the transition of innate immune signals into adaptive immunity, modulating naïve T cell polarization. After maturation, DCs migrate in skin lymph node, where they prime naïve T cells, triggering a process of polarization into various specialized subtypes of T cells,

INTRODUCTION such as Th1, Th17, Treg and Th2. The polarization is highly influenced by the cytokines present during DC-T cells interaction (Volz et al., 2012).

Last but not least, keratinocytes sense pathogens, through PRR/PAMPs. After microbial interaction and recognition, keratinocytes can phagocyte the microorganism and, more commonly, produce antimicrobial peptides (AMPs), cytokines and/or chemokines, guaranteeing a (I) direct or (II) indirect protection against microbes (Nestle, Di Meglio, Qin, & Nickoloff, 2009). The immune response of keratinocyte will be described in detail below.

2.2.2.3. Keratinocyte as Immune Sentinel Since 1970, it has been suggested that keratinocytes are able to initiate immune responses in the skin, contrary to the “passive” mechanism of protection already acknowledged for these epithelial cells (McKenzie & Sauder, 1990). Today it is well accepted that these cells have a key role in triggering the cascade of immune events that will ensure host protection. This keratinocyte-initiating role in immunity is, first of all, associated to their topological position, since they are the first cells that external pathogens meet at the skin surface.

To carry out this role of epidermis sentinel, keratinocytes have, as the other innate immune cells, different kind of PRRs (Pivarcsi et al., 2005). Therefore, these cells are directly involved in the “sensing of microorganisms”. Furthermore, keratinocyte immune activity is also initiated in response to immune mediators produced by neighboring activated immune cells (Naik et al., 2015). This reveals a powerful mutual crosstalk between all the cells forming the immune barrier of the skin. Once activated, either by direct microorganism recognition or by cytokine stimulation, keratinocytes make a specific response, adapted to the pathogen. Basically, epidermal cells can secrete AMPs for fighting pathogens, or/and cytokines or chemokines in order to establish a dialogue with and transmit the ‘message’ to closest immune cells.

2.2.2.3.1. Antimicrobial peptides Biological function of the antimicrobial peptides

AMPs are ancestral antimicrobial molecules with amphipathic structure that ensure the first line of defense in microorganisms, plants and animals. Microbial AMPs, as eukaryotic AMPs, limit the growth of other microorganisms, although their different structure. Commensal AMP production is one of the key microbiota-mediated defence of the host against pathogens (see skin microbiome). Moreover, some microbial AMPs are largely used as antibiotic drugs.

INTRODUCTION

AMPs fulfill their biological role thanks to a wide plethora of mechanisms; these peptides can either (i) directly kill microbes and/or (ii) modulate host immunity (Figure 15). Their ability to directly kill microorganisms depends mainly on AMP capacity to interact and bind microbe membranes and cell walls. Almost all the AMPs display a positive charge that allows an electrostatic interaction with some microbial membrane components. After interaction, some AMPs, like LL-37, can form pores in the microbial membrane, destroying its integrity, through a non-enzymatic process (Zhang et al., 2016). There are also some AMPs, like lysozyme, that mediates enzymatic hydrolysis of peptidoglycan of bacterial cell walls. Nonetheless some AMPs are able to enter the bacterium for inhibiting its intracellular functions. Finally, there are some AMPs that avoid biofilm formation and/or destroy the already existing biofilm (see below S.Epidermis vs S.Aureus). All these different processes allow AMPs to kill microbes directly through AMP/targeted microbes physical interaction (review in Zhang et al. 2016).

Despite the direct killing action, AMPs are potent pleiotropic molecules, showing also different indirect multiple defense strategies. Indeed, AMPs can ensure host protection, modulating its immunity. AMPs themselves can have chemotactic properties, triggering the recruitment and the activation of immune cells, as shown both in vitro and in vivo for CRAMP (Kurosaka, Chen, Yarovinsky, Oppenheim, & Yang, 2005).

Otherwise some AMPs can promote chemokines and cytokines production by resident cells, triggering the recruitment and/or the activation of both innate immune cells, such as neutrophils, macrophages, DCs, and adaptive immune cells, like T lymphocytes. For example, it was demonstrated that LL-37 triggers the secretion of IL8 in alveolar epithelial cells, and the cathelecidin-mediated production of this chemokine enhances the amount of IL8 produced in case of E. Coli infection (Scott et al., 2002). Moreover, during infections and tissue injuries nucleic acids (ssRNA and dsDNA) and PAMPs are released. Different studies, reviewed by Lee and colleagues, have demonstrated that AMPs can bind these components, forming AMP/immune ligand complexes that modulate PRR-mediated signaling pathways and facilitate the immune cells responses (E. Y. Lee, Lee, & Wong, 2018).

INTRODUCTION

Figure 15. Antimicrobial peptide mechanisms of action. After elcetrostatic interaction with bacteria, AMPs can form pores or can inhibits bacterial intracellular processes, fulfilling to their antimicrobial function. Otherwise, AMPs can assure host protection modulating the host immunity. These peptides can both directly chemoattract various immune cells, such as dendritic cells, neutrophils and macrophages, and, forming complex with DAMP (DNA or RNA), trigger cyto- and chemokine production in immune cells.

Skin antimicrobial peptide

Considering their essential biological roles, AMPs expression must be tightly regulated. These peptides are mainly expressed in the epithelia interacting with external environment (Figure 16). In the skin some AMPs are constitutively expressed, like β-defensin 2 in human skin, to ensure the tissue homoeostasis and the correct development of the skin-associated microbiota. However almost all AMPs, e.g. cathelicidin peptides (see below), are induced in the skin during infection or in response to inflammation and tissue injury.

INTRODUCTION

Figure 16. AMPs expression in human skin. Representative but not complete scheme of AMPs express in the human skin in physiological condition. Different skin compartment, such as hair follicle, epidermis and sebaceous gland produce different AMPs according the inflammation/infectious state of the skin, e.g. cathelicidin is mainly express in the sebaceous gland in physiological condition while during some infection is preferentially expressed in the epidermis.

In mammals there are different families of antimicrobial peptides expressed in different host- environment interaction surfaces, such as skin, respiratory, urinary and gastrointestinal tracts by both the epithelial and the epithelial-associated immune cells. I will focus on cationic antimicrobial peptide (CAMPs) families, cathelicidin and β-defensins (Gallo & Hooper, 2012), since these are the AMPs mainly expressed in the skin.

Cationic AMPs (CAMPs) typically are 10 to 50 amino acid peptides positively charged. These peptides usually contain basic amino acids and hydrophobic residues, allowing three dimensional arrangement in a hydro soluble, positively charged and hydrophobic molecules (Zhang et al., 2016). Depending on their secondary structure CAMPs can be further distinguished in two wide families: the α-helical and β-sheet peptides. The latter group consists basically of one wide family named β-defensins. In mammals β-defensins family consists of around 90 different peptides. Despite the high number of peptides and their large sequence variability, all mammalian defensins have six cysteines, allowing intramolecular disulfide bindings and consequently the formation of two or more β-sheet structures. Hepcidin shows a similar secondary conformation and, like cationic AMPs, is small and positively charged. As discussed above, hepcidin structural affinity to CAMPs, in particular to β-defensins, is one of

INTRODUCTION the key reason supporting the idea that the iron regulatory hormone has, at least, ancestral antimicrobial peptide origin. On the other hand, α-helical peptide family consists of cathelicidins. These peptides have been described in both invertebrate and vertebrate species. The cathelicidins are considered to be a one gene family, in human the only cathelicidin gene is CAMP, orthologous to the murine Camp. These genes encode for preproproteins that undergo post-translational processing giving rise to, respectively, LL-37 in human and CRAMP in mice (Lai & Gallo, 2009). Once mature, these peptides exhibit a weak secondary structure in aqueous solution but adopt a α-helical secondary structure in non-polar environment, such as bacterial membrane.

Cathelicidin peptides were originally identified in both human (Gudmundsson et al., 1996) and murine myeloid cells (Gallo et al., 1997), showing a strong antimicrobial activity in vitro. This antimicrobial peptide is largely expressed in the skin, in different histological zones according the state of the skin. At the steady state this peptide is barely expressed in epidermis and mainly presents in sweat and expressed in eccrine gland epithelial cells (Murakami et al., 2002). However, its expression is strongly increased and induced in interfollicular keratinocytes in response to Group A streptococci (GAS)-mediated infection wound (Dorschner et al., 2001a). This model of subcutaneous infection mimics human necrotizing fasciitis, determining skin and underlying soft tissue necrosis in mouse. Interestingly the upregulation of cathelicidin according to the infectious/inflammatory skin state, have suggested a key role of this AMP in skin protection. In 2001, the team of Richard Gallo demonstrated an antimicrobial role of CRAMP in a model of skin infection in vivo. Taking advantage of transgenic Cramp knockout animals and cramp-resistant GAS bacterial strain, the authors demonstrate the protective role of CRAMP against invasive infections in skin in vivo. Indeed they observed a drastically worst infection outcome in mice lacking this peptide and, at the same time, they described that CRAMP -resistant GAS give rise to a more aggressive infection compared to wild type GAS in wild type mice. Although this study did not allow to discriminate the neutrophil- or keratinocyte-derived CRAMP contribution, it was the first study that highlights the importance of cathelicidin in host defense against invasive skin infection in vivo (Nizet et al., 2001). The Gallo team next discriminated keratinocyte-derived CRAMP role in the same infection model. After neutrophil depletion, CRAMP knock out-mice still display a worst infection outcome,

INTRODUCTION highlighting the relevance of epidermal cramp expression in host defense in vivo (Braff, Zaiou, Fierer, Nizet, & Gallo, 2005a).

2.2.2.3.2. Cytokines and chemokines Although one of the key keratinocyte-mediated responses to infection consists in AMPs production, the epithelial cells can secrete chemokines, for recruiting immune cells to the infection/inflammation site, and cytokines to establish an immune dialogue with the resident- or recruited- immune cells. Various studies have demonstrated that microbes and microorganism-derived molecules trigger cytokine secretion both in human and murine keratinocyte. The cytokine array secreted in response to pathogen infection has a proinflammatory profile. PAMP/PRR interaction usually activates NF-KB signaling pathway, determining, according to the infection stimuli and timing, the expression of the cytokines required for keratinocyte-immune cell dialogue, including IL-1, IL-6, IL-18 and TNF (Nestle et al., 2009).

For example SA infection mediates GM-CSF production, that is required for Langerhans and dendritic cell maturation and differentiation (Matsubara, Harada, Manabe, & Hasegawa, 2004). Similarly CpG, a component of bacterial DNA, triggers a keratinocyte IL-1α, GM-CSF and TNF-α secretion activating Langerhans cells (Sugita et al., 2006). Moreover various “sterile” inflammatory stimuli, such as ultraviolet light, trauma, tissue injury and, especially, cytokines, stimulate the keratinocyte production of wide of immune modulators both in vitro and in vivo, defining keratinocytes as key agents contributing to fine tuning the inflammatory state of the skin in different pathophysiological conditions (Albanesi, Scarponi, Giustizieri, & Girolomoni, 2005).

Keratinocytes are also an important source of chemokines and mediate the attraction of different cell types into the skin in response to infection/inflammation. It is well known that the early recruitment of leukocytes, such as neutrophils and macrophages, to the site of infection is crucial for eliminating pathogens and avoiding systemic infection spread. It is well accepted that subcutaneous infection determine the increase of the chemokine expression in the skin of infected animals in vivo (Kurupati et al., 2010a).

Various studies have demonstrated that different pathogens or PAMPs, like S. aureus, C. Albicans, PGN/LTA, M. Tuberculosis, Group A streptococcal M1, trigger keratinocyte

INTRODUCTION expression of macrophage and neutrophil chemokines, such as RANTES/CCL5, CXCL-1, CXCL-2 and IL-8 (Matsubara et al., 2004; Mempel et al., 2003; Olaru & Jensen, 2010; Persson, Wilk, Mörgelin, & Herwald, 2015; Pivarcsi et al., 2003). All these studies in vitro show that upon stimulation the chemokine up regulation occurs at transcriptional level, through NF-KB signaling pathway activation. Although also other skin cells can show some chemoattractant potential (Mishalian et al., 2011a), all these studies in vitro underline the role of keratinocytes in immune cells recruitment during infection. Leukocyte recruitment in response to infection will be better described in the specific case of GAS subcutaneous infection (see below, chapter 4)

2.2.3. Skin Microbioma The skin microbioma is the earliest protective barrier that prevents pathogen invasion though both its own protective function and shaping/influencing the other barriers.

Various bacteria, virus and fungi commensal families inhabit skin. Structural organization of epidermis, notably the abundance of skin appendages, affects its physical characteristics, such as pH, temperature, moisture and sebaceous amount, giving rise to distinct niches and allowing skin microbiota topographical diversity associated to specific characteristic of the skin sites (E. A. Grice et al., 2009; Elizabeth A. Grice & Segre, 2011).

Once established at the neonatal stage (Scharschmidt et al., 2015), skin microbiota forms the prime protection against pathogens. First, commensals limit pathogens by competing for nutrients and environmental niches. Beyond this “competitive” mechanism, commensals can (I) directly fight against pathogen and/or (II) modulate innate or adaptive immune cells of the skin.

For example, Staphylococcus epidermis (SE), the most abundant skin commensal, secretes a specific protease inhibiting Staphylococcus Aureus (SA) biofilm formation and allow host protection against this infection (Iwase et al., 2010; Sugimoto et al., 2013). Moreover, different studies have shown that commensal bacteria can produce specific antibiotics and/or antimicrobial peptides, against specific pathogens. As other antimicrobial peptides (see above), these commensal-derived antimicrobial molecules can either directly kill bacteria or modulate host immune cells. Cogen and colleagues described a commensal-derived molecule promoting neutrophil extracellular traps (NETs) formation, therefore promoting Group A streptococcus

INTRODUCTION

(GAS) killing (for details about NET see below) (Cogen et al., 2010). Beyond the AMP- mediated immunomodulation, commensal bacteria can directly regulate host immunity. Although less characterized compared to the gut, skin microbiota immunomodulation is more and more studied. A study of Chehoud and colleagues is an excellent example of host immunity and skin microbial ecosystem reciprocal regulation. The authors demonstrate that alteration of the host complement-signaling pathway determines a skin dysbiosis. Moreover, comparing germ free vs conventional mice they demonstrate that skin microbiota regulates the complement gene expression in the skin. These evidences defined a host-microbiota mutual regulation in the skin (Chehoud et al., 2013). Microbiome can regulate skin immunity by a dialogue with immune cells. In 2015 Naik and colleagues observed that S. Epidermis, activate resident dermal dendritic cells. These commensal-activated dendritic cells, in turn, active naïve T cells that modulate keratinocyte innate immune functions. Epidermal cell immunity actions ensure defense against pathogens and skin microbiota homeostasis (Naik et al., 2015).

Similarly, another recent study shows that commensal bacteria activate dendritic cells that, in turn, migrate to the lymph node where they activate resident T cells. The authors demonstrate that this commensal-activated T cells promotes the protection against pathogens and the wound healing (Linehan et al. 2018). This recent study demonstrates a different microbiota-mediated immune cells regulation and emphasizes the immune cell role in ensuring tissue integrity in the skin. The microbiota has a key role also in the crosstalk with the other protective barriers of the skin. Chemical/physical barriers and microbiota interact and mutually shape themselves. Indeed, microbiota alteration can directly influences tight junction development and stability, affecting therefore, the physical barrier both in vitro and in vivo (Ohnemus et al., 2008).

As discussed before, the chemical/physical, immune and microbiome barriers are carefully orchestrated to stabilize cutaneous homeostasis. In case of pathogen infection all these barriers are modified and try to mount measures of defense. One of the key process that occurs earliest during infection is the recruitment of neutrophils to the infection site.

INTRODUCTION

3. NEUTROPHIL

Neutrophils are the most abundant type of granulocytes in most mammals. These cells are generated in the bone marrow, through a process named granulocytopoiesis, and circulate in the blood as resting cells. However, in case of infection and/or inflammation, neutrophils are rapidly activated and recruited to the infection/inflammation sites. Despite neutrophil involvement in inflammatory conditions (like tissue injury-healing or chronic inflammatory diseases), I will specifically focus on neutrophil role in infection, as host’s primary line of defence against invading pathogens, like bacteria.

Upon infection, the endothelial cells capture circulating neutrophils and guide them through the endothelial cell lining, mediating the process of extravasation. Once in the tissue, neutrophils migrate towards microbes, following chemoattractant gradients. Through different mechanisms, such as phagocytosis, degranulation, respiratory burst and neutrophil extracellular traps (NETs) production, neutrophils clear pathogen infection. For the host, neutrophil defence mechanisms represent a double-edged sword, because these processes also allow the digestion of host tissues. For this reason, neutrophil homeostasis must be tightly regulated, to ensure adequate protection against microbial pathogens while minimizing damage to host tissues. This homeostasis is achieved through a balance of neutrophil production, neutrophil recruitment and clearance from the peripheral tissues.

3.1. Neutrophil production and bone marrow egress 3.1.1. Granulocytopoiesis Neutrophils have a short life span; thus, these immune sentinels must be continuously re- established. The granulocyte colony stimulating factor (G-CSF) plays a major role in homeostatic regulatory loop responsible for the “steady-state” neutrophil production and it is the major cytokine orchestrating the de novo generation of neutrophils in response to infection or acute inflammation (“emergency” granulocytopoiesis) (Lieschke et al., 1994). The G-CSF acts as the final and decisive factor that modulates neutrophil production, integrating various upstream signals. As a matter of fact, the granulocytopoiesis must be finely tuned for ensuring the transient elevation of neutrophils during infections and allowing the following return to the physiological neutrophil baseline (Borregaard, 2010). One key signal modulating neutrophil production in the bone marrow is the number of apoptotic neutrophils in the peripheral tissues.

INTRODUCTION

Basically, a higher number of apoptotic neutrophils in tissues modifies the macrophage and dendritic cell cytokine production, allowing a decreased production of IL-23, and the consecutive IL-17A production. Since this interleukin is an important stimulus for G-CSF production, its decrease provokes the reduction of G-CSF hence a diminished neutrophil production. This negative feedback loop contributes to the decrease in the transient elevated number of neutrophils reached during infection. Therefore, this cytokine-cascade signalling avoids the progression of the acute inflammation towards a chronic inflammatory state (Borregaard, 2010).

In steady-state, neutrophils are produced and matured in the bone marrow. During their maturation, neutrophil nucleus undergoes a process of segmentation and the neutrophil cytoplasm fills up with granules and secretory vesicles. The secretory vesicles are rich in integrins, as Mac-1, and chemoattractant receptors, like fMLP receptor. This kind of molecules are essential for neutrophil extravasation and migration towards the site of infection in peripheral tissues (Borregaard, 2010) (for more details see below). Neutrophil granules are very heterogeneous and can be classified according different parameters, like their protein contents. There are two different groups of granules: the peroxidase positive and peroxidase negative granules. The first group is characterized by the presence of the myeloperoxidase (MPO) among the various granule proteins. MPO, a 150 kDa haemoprotein, contributes to the production of reactive oxygen species with strong antimicrobial features (see below). Otherwise peroxidase- negative granules can be further classified into specific and gelatinase granules. The specific granules contain a wide range of antimicrobial molecules, such as lactoferrin, lipocalin-2, cathelicidin peptide, lysozyme etc. Overall these molecules can either kill bacteria, as lysozyme or cathelicidins, or avoid the bacterial growth, e.g. chelating essential nutrient like lactoferrin and lipocalin-2 (for more details see section 3.3). The gelatinase granules contain mainly matrix degrading enzymes, such as heparanase and gelatinase, and membrane receptors involved in neutrophil extravasation, chemotaxis and activation, like fMPL receptor, Mac-1 etc. It is the mobilisation of vesicles and granules that makes possible the switch between “resting neutrophil” in “potent actor” of innate immunity (for more details see section 3.3.1.3). Therefore, at the end of their maturation, neutrophils are already equipped with the proteins required to kill microorganisms, but carefully segregated in these intracellular structures.

INTRODUCTION

3.1.2. The neutrophil egress from bone marrow Being circulating innate immune cells, at the end of their maturation, neutrophils have to move from the bone marrow to the circulation. The number and the rate of neutrophil exit from the bone marrow are finely regulated by the balance act between CXCR4 and CXCR2. These two receptors are expressed on neutrophils membrane and are responsible, respectively, of bone marrow homing and of the egress from bone marrow to reach the circulation. The receptor CXCR4 binds CXCL12, which is expressed by bone marrow stromal cells, especially osteoblasts. CXCR4/CXCL12 binding is essential for the homing of mature neutrophils into the bone marrow (Eash, Means, White, & Link, 2009). While the exit from the bone marrow is mediated by the CXCR2 binding to its ligands CXCL1 and CXCL2, expressed by bone marrow endothelial cells (Eash, Greenbaum, Gopalan, & Link, 2010). In steady-state the CXCR4/CXCL12 axis is favoured, therefore neutrophils preferentially remain in the bone marrow; in case of infection or in response to other stimuli, G-CSF promotes CXCL1 and CXCL2 expression in bone marrow endothelial cells, thus tilting the CXCR4-CXCR2 signalling in favour of CXCR2 and neutrophil release (Eash et al. 2010) (Figure 17).

Figure 17. Neutrophils egress from the bone marrow. At the end of their maturation, neutrophils remain in the bone marrow, constituting a reservoir (on the right). The osteoblasts produce CXCL12 that binding the neutrophil receptor CXCR4 (represented in orange), mediates the bone marrow homing. On the other hand, in case of inflammation/infection, bone marrow endothelial cells produce the chemokine CXCL1/2, that binding CXCR2 (purple) prompt neutrophil massive egress (on the left). G-CSF: Granulocyte-colony stimulating factor.

INTRODUCTION

3.2. Neutrophils in tissue 3.2.1. Neutrophil extravasation Once out of the bone marrow, neutrophils are in the blood stream. In case of infection, circulating neutrophils must cross the vascular wall, to reach the site of infection and ensure bacterial clearance. This passage takes place at the postcapillary venules. These vessels have a thin wall and their diameter is small enough to allow neutrophil contact with endothelial cells, but sufficiently large to avoid the occlusion of the vessel when neutrophils arrest and make a firm contact with the endothelium. The process of extravasation, or neutrophil exit from the vessels to the tissues, consists of different phases: the rolling, the firm adhesion and the transmigration. The process is initiated at the infection site, where resident cells have an essential role in transmitting this information to endothelial cells. Then the “activated” endothelium expresses some molecules required to slow the neutrophil rolling and permit neutrophil firm adhesion and consecutive transmigration (Figure 18).

Upon infection, tissue resident cells, like macrophages, dendritic, mast and epithelial cells, release pro-inflammatory mediators, triggering the process of neutrophil recruitment to the infection site. Indeed, these soluble pro-inflammatory factors, activate the endothelial cells. First of all, the inflamed endothelium express P- and E-selectins on the luminal surface of endothelial cells. These endothelial molecules interact with the P-selectin glycoprotein ligand 1 (PSGL-1), which is constitutively expressed on neutrophil microvilli. This interaction mediates bonds that form and disassemble sequentially, yielding the slow rolling of neutrophils on inflamed endothelium (Ley, Laudanna, Cybulsky, & Nourshargh, 2007). Furthermore, in response to the pro-inflammatory stimuli, endothelial cells produce chemoattractant molecules, like CXCL8/IL8 or LTB4 (see below). These classical chemoattractant molecules are the most powerful mediators of neutrophil firm adhesion to endothelium (Borregaard, 2010). Actually, in vitro and in vivo studies have established that neutrophil firm arrest during the slow rolling is due to the interaction of neutrophil integrins, especially LFA-1 with the endothelial intercellular adhesion molecule, ICAM1. Neutrophil integrins can be defined as “activatable” receptors, since they require some stimuli to increase their ligand-binding capacity. The chemoattractant molecules produced by the inflamed endothelium, are potent activators of the neutrophil integrins: these chemoattractant molecules improve the affinity of LFA-1 for ICAM1 and increase the density of LFA-1 per area of neutrophil plasma membrane, enhancing the valency of this binding (Campbell, Fledrick, Zlstnik, & Siani, 1998; Ley et al., 2007; Seo,

INTRODUCTION

McIntire, & Smith, 2001). The chemoattractant molecules activate also another neutrophil integrin, named Mac-1, mainly involved in the next step of extravasation. Upon the firm adhesion, activated-Mac-1 interacts with ICAM1, yielding the extension of neutrophil membrane protrusions into the endothelial cell-body or cell-junction. This process, named crawling, triggers endothelial intracellular cytoskeleton rearrangements that lead enhanced endothelial cell contraction and hence promote neutrophil migration through endothelial junctions (paracellular road) or through the body of the endothelium (transcellular road). Then, neutrophils cross the endothelial basement membrane moving through gaps between adjacent pericytes and digesting extracellular matrix through the activity of the enzymes neutrophil elastase (NE) and matrix metalloprotease (MMP8 or MMP9) (Ley et al., 2007).

Figure 18. Neutrophil extravasation. In response to infection/inflammation, immune cells produce cytokines which activate endothelial cells prompting the expression of selectin at the luminal membrane of these cells. The selectins interact with PSGL1, forming transient binding and allowing the slow rolling of neutrophil above inflamed endothelium. The cytokines also trigger the endothelial secretion of chemokines, such as CXC- chemokines. These molecules active the neutrophil integrin, LFA1, causing the firm adhesion and the following crawling. Finally neutrophil migrates in the tissue, crossing endothelial junctions (transcellular) or crossing endothelial cell (paracellular). PSGL1: P-selectin glycoprotein ligand 1; LFA1: lymphocyte function-associated antigen 1.

3.2.2. Neutrophil migration Infection generates a wide range of signals that establish chemoattractant gradients throughout tissues. Once in the tissue, after the extravasation, neutrophils feel and move towards the infection site (Figure 19). In order to fulfil this forward migration, neutrophils present a wide number of different receptors that allow the pro-inflammatory and infectious sensing, yielding

INTRODUCTION the neutrophil migration. Among these receptors are the G protein-coupled receptor (GPCRs), cytokine receptors, PRRs and adhesion receptors. Basically, the binding of microbial components and/or chemoattractant molecules to their specific receptors switch on different kind of downstream chemoattractant signalling pathways, responsible of actomyosin dynamics, thus neutrophil migration.

At the site of infection, there are high amount of PAMPs and Damage-Associated Molecular Patterns (DAMPs), released by damaged and necrotic cells. This kind of molecules mediate the first neutrophil migration through (i) direct and/or (ii) indirect mechanisms. Neutrophil receptors can recognize and directly bind DAMPs and microbial components. As mentioned above, these interactions can trigger the neutrophil chemoattraction towards microbes, e.g. the bacterial N-Formylmethionine-leucyl-phenylalanine tripeptide (fMLP) is one of the strongest chemoattractant molecules for neutrophils. Otherwise, PAMPs and DAMPs can trigger the production of long-term and long-range signals, that are chemoattractant molecules, to sustain neutrophil recruitment. There are two main families of molecules involved in this spatial- and temporal-extension: (i) the CXCL8 family and (ii) lipid mediators (see below). These molecules are produced in the affected tissue, in particular by endothelial, epithelial and immune cells.

The expression of these molecules, especially the chemokines, can be modulated at the transcriptional level, through NF-KB or AP-1 activation (De Oliveira, Rosowski, & Huttenlocher, 2016). However, pre-made stores of chemokines can be rapidly released from endothelial cells activated by DAMPs (Oynebraten et al., 2005). Therefore, the production of chemoattractant molecules can be regulated at different levels according to the cell type and the stimuli, ensuring a fine tuning of neutrophil recruitment in peripheral tissues. As described before, the chemoattractant molecules are sensed by neutrophils through specific receptors, activating downstream signalling pathways and allowing neutrophil migration.

Once recruited at the infection site, neutrophils activate tissue and tissue resident cells, magnifying the production of chemoattractant molecules and causing ‘neutrophil swarming’ in the tissue. This amplification is ensured by different mechanisms. Activated neutrophils can (i) produce chemokine and lipid mediators, (ii) establish immune crosstalk with resident or early recruited cells (as endothelial cells, macrophages and fibroblasts) prompting their chemokine expression. Finally, the death of the initially recruited cells, give rise to further DAMPs,

INTRODUCTION contributing to secondary neutrophil swarming. The mechanisms that emphasize the neutrophil swarming are especially pronounced during infection (Figure 19).

Compared to sterile damaged tissues, infected tissues present much wider range of immune cells participating in and modulating neutrophil recruitment, like macrophages (Abtin et al., 2010) and dendritic cells (Sacramento et al., 2015). Furthermore, many dead neutrophils are present at the site of infection, therefore the secondary neutrophil infiltration in response to cell- death is very pronounced. Since the neutrophil recruitment at the infection site is highly pronounced, pathogens try, with a plethora of mechanisms, to avoid this recruitment. This dynamic battle involves microbial virulence factors that, directly or indirectly, alter neutrophil recruitment by the host. The strategies are pathogen-specific, and I will discuss more in detail the GAS strategies in the following chapter.

Figure 19. Neutrophil migration in the tissue. Once extravasated in the tissue, neutrophils migrate towards the site of infection to ensure pathogen clearance. During infection abundant PAMPs and DAMPs are produced. These molecules directly chemoattract neutrophil and promote the secretion of chemokine and other chemoattractant molecules by tissue resident cells, e.g. resident myeloid cell. Then, recruited neutrophils secrete chemokine, prompt chemokine production by other immune cells and finally die, producing DAMPs. These processes boost the neutrophil recruitment, causing a next neutrophil swarming to the site of infection.

3.3. Chemoattractant molecules: the chemokine family As described before, the chemoattractant molecules have a key role in recruiting in immune cells, such as monocytes, macrophages, neutrophils, Th1, Th2, NK, eosinophils, basophils,

INTRODUCTION etc… There are many molecules involved in this function: chemokines, lipid mediators, complement fraction C5a, fMLP and platelet activator factor (PAF).

In general, the chemokines, or chemoattractant cytokines, are small molecules (8-15 kDa) which control leukocytes recruitment and activation under basal, infectious and inflammatory states. In mammals the chemokine system includes almost 50 ligands, expressed in a constitutive or inducible manner, and 20 receptors with 7 transmembrane domains, belonging to the GPCRs (Russo, Garcia, Teixeira, & Amaral, 2014). Considering the aim of this study, I will especially introduce the CXCL8/IL8 family that is especially involved in neutrophil chemoattraction.

As CXCL8/IL8, all these chemokines present the glutamic acid-leucine-arginine (ELR) motif, thus are defined as ELR+ chemokines. This family of chemokines is also named CXC- chemokines because they have four cysteines, of which the first two are separated by one amino acid, forming the motif CXC. These chemokines bind the GPCRs CXC-chemokine receptors, CXCR1 and CXCR2, inducing migration and activation of neutrophils (De Oliveira et al., 2016).

CXCL8/IL8, the prototypical member of this family, was originally isolated and described in 1987 as a monocyte-derived neutrophil-specific chemotactic factor (Yoshimura et al., 1987).

This chemokine can be produced by immune cells, such as neutrophils, macrophages and T cells, or by non-leukocyte cells, such as epithelial and endothelial cells. Actually this chemokine has pleiotropic roles, modulating both the recruitment and various activities of neutrophils (Baggiolini, Walz, & Kunkel, 1989; Peveri et al., 1988; Thelen et al., 1988; Yoshimura et al., 1987). As described above, CXCL8/IL8 can be secreted by stimulated endothelial cells, mediating the firm arrest during the slow rolling of neutrophils, thus ensuring extravasation (DiVietro et al., 2001; Huber, Kunkel, Todd, & Weiss, 1991; Middleton et al., 1997; Seo et al., 2001). Moreover, IL8 binding to its receptors activates downstream signalling pathways responsible for the asymmetric actomyosin dynamic and consequent neutrophil migration. The importance of IL8/CXCR2 signalling in neutrophil migration was largely studied and confirmed both in vitro and in vivo. In an early study, Yoshimura and colleagues have already described the powerful ability of IL8 to induce neutrophil migration in vitro (Yoshimura et al., 1987). Basically, using the LPS-stimulated monocyte-derived IL8, the authors observed a strong

INTRODUCTION neutrophil migration, finding that 50% of cells migrated in response to this chemokine. Interestingly, they observed that this protein was not chemotactic for monocytes, suggesting the specificity of IL8 for neutrophil recruitment (Yoshimura et al., 1987). Then, it was demonstrated that the intradermal injection of the human recombinant IL8 triggers a neutrophil accumulation to the site of injection in rabbit, highlighting for the first time the same chemotactic property in vivo (Colditz, Zwahlen, Dewald, & Baggiolini, 1989). Once the gene encoding for the IL8 receptor has been discovered (Holmes, Lee, Kuang, Rice, & Wood, 1991), the murine homolog gene CXCR2 was next cloned (Cacalano et al. 1994). CXCR2 knock-out animals display impaired intraperitoneally (i.p.) neutrophil recruitment in response to i.p. thioglycolate injection demonstrating that this receptor was essential for neutrophil recruitment in vivo (Cacalano et al., 1994). As mentioned above, IL8 enhances different neutrophil functions, such as respiratory burst and degranulation (Peveri et al., 1988). As it will be discussed below, these are key mechanisms for neutrophil-mediated bacterial clearance. Given the important roles of IL8/CXCR2 axis in neutrophil extravasation, recruitment and bacterial killing functions, the role of this signalling was investigated in various animal models of infection. Although mouse and rat do not have the IL8 homolog gene, different studies have shown that the mutations of CXC-chemokines/CXCR2 signalling pathway, determine a worst outcome of different kind of infection models (Cai, Batra, Lira, Kolls, & Jeyaseelan, 2007; Nagarkar et al., 2009; Ritzman et al., 2010). One of the first study investigated the role of the principal murine CXC-chemokines, CXCL1 and CXCL2, in Pseudomonas Aeruginosa infection. The depletion of each of the chemokines resulted in a modest decrease in neutrophil recruitment, while blocking CXCR2 decreased the bacterial clearance and severely affected the survival. This study highlighted the importance of CXCR2 in host protection against Pseudomonas Aeruginosa pneumonia (Tsai et al., 2000). However, a more recent study, on pulmonary Klebsiella infection, replaces CXCL1 as central player of the host defence. In this model of infection, CXCL1 is important for neutrophil recruitment and furthermore essential for the expression of cell adhesion molecules and other neutrophil chemoattractants. These multiple functions of CXCL1 emphasize its role in infection resolution (Cai et al., 2007).

3.4. Role of neutrophils in fighting infection First of all, neutrophils recognize pathogens through a wide range of receptors: PRRs, such as TLRs, and opsonic receptors, like Fc receptors. The latest category of receptors recognizes the opsonins (as complement or antibodies) that are deposited on the pathogen’s surface. After

INTRODUCTION pathogen recognition/interaction neutrophils can carry out a wide plethora of antimicrobial activities that can be classified in intracellular and extracellular antimicrobial mechanisms.

3.4.1. Intracellular antimicrobial mechanism 3.4.1.1. Phagocytosis The phagocytosis is a specific form of endocytosis characteristic of some innate immune cells, such as, neutrophils and macrophages. The plasma membrane of these phagocytic cells surrounds the microorganism, through an actin-myosin contractile system, enclosing the pathogen inside a vacuole, known as phagosome (W. L. Lee, Harrison, & Grinstein, 2003). When the phagosome is formed around a particle, it is not equipped for killing microbes. However, the process of phagocytosis triggers the mobilization of cytoplasmatic secretory vesicles and granules. In particular these intracellular organelles fuse with the phagosome, changing the phagosomal contents and membrane and allowing the “phagosome maturation” (W. L. Lee et al., 2003). The “mature phagosome” gets antimicrobial peptides (e.g. defensins, lactoferrin, cathelicidin) proteolytic enzyme (e.g. lysozyme) and enzymes required for the generation of high toxic ROS (myeloperoxidase and NADPH oxidase) (El-Benna et al., 2016).

Therefore, after phagosome maturation, internalized pathogens can be destroyed by a wide range of antimicrobial mechanisms, classifiable according to their way of action in (i) oxidative and (ii) non-oxidative killing processes (Figure 20). The first class demands ROS production, while the second is independent of ROS generation.

3.4.1.2. Non-oxidative mechanisms The granules fuse with the phagosome, discharging antimicrobial peptides and enzymes. These antimicrobial molecules include (i) AMPs, (ii) nutrient chelators and (iii) proteases. The main AMPs released are α-defensins, bactericidal/permeability-increasing proteins (BPI) and LL-37. As described before, AMPs are able to directly kill bacteria by binding to and damaging microbial membranes. Furthermore, nutrient chelators such as lactoferrin and lipocalin-2 are released in the phagosome, leading to a limited iron availability for microbes. Similarly, Nramp1 (Natural resistance-associated macrophage protein 1), is translocated to phagosomal membrane after degranulation. Nramp1, as Nramp2, is a transporter of divalent cations and can play an essential role for depriving microorganism in the phagosome of metals such as Fe2+, Mn2+ and Zn2+ (François Canonne-Hergaux et al., 2002). Finally, the last category of molecules

INTRODUCTION involved in the non-oxidative killing are different kind of proteases, like the lysozyme, that contribute to bacterial killing, mediating an enzymatic digestion of bacterial cell wall.

Overall these molecules can either kill bacteria and/or avoid the bacterial growth, assuring a broad range of antimicrobial non-oxidative killing mechanisms.

3.4.1.3. Oxidative mechanism: oxidative burst Despite the importance of non-oxidative killing processes, the ROS production constitutes the main antimicrobial function of neutrophils. Indeed, the ROS are highly bactericidal and some downstream effects of their production may ultimately be responsible for other bactericidal activity of neutrophils, like NETosis (see below) (Nguyen, Green, & Mecsas, 2017).

.- Superoxide anion (O2 ) is the key initiator of oxidative killing process, properly referred as ‘oxidative burst’. This ROS allows, in turn, the generation of other reactive oxygen species, such as hydrogen peroxide, hydroxyl radical and hypochlorous acid, in the phagosome (El- Benna et al., 2016). All the ROS produced contribute to the death and the destruction of bacteria.

Figure 20. Intracellular antimicrobial mechanisms. After phagocytosis, neutrophils have different mechanisms for killing pathogens. Granules fuse with the phagosome, releasing AMPs, portease and nutrient chelators, allowing a non- oxidative killing (on the left). Otherwise, after phagocythosis the NADPH complex is assembled at the phagocyte membrane,

- allowing the production of superoxide anion (O2 ). Azurophilic granule (here depicted in light blue) fuses and releases MPO, allowing the HOCl production. AMP: antimicrobial peptide; MPO: myeloperoxidase.

.- The generation of this O2 ‘spark’, requires the assembly and activation of a complex enzymatic system, named NADPH oxidase. This complex system is composed by a (i) membrane-

INTRODUCTION

catalysing component, named flavocytochrome b558 (cytb558), and (ii) four soluble-regulatory proteins, p47phox, p67phox, p40phox and the small G-proteins Rac 1/2. The cytb558 catalyses the transfer of electrons from NADPH to molecular oxygen, generating the superoxide anion .- inside the phagosome. The O2 produced, can react with protons to generate hydrogen peroxide

(H2O2), which is used by the MPO to produce hypochlorous acid (HOCl). Moreover, in 2+ 2+ .- - presence of metals (Fe or Cu ) the O2 can react with H2O2 to produce hydroxyl radical (OH

). In resting neutrophils, cytb558 components are segregated in intracellular vesicles, specific- and gelatinase-granules. The MPO, enzyme essential for HOCl production is stored in peroxidase positive granules, also named azurophilic granules. Moreover, the trimeric complex of regulatory proteins (p47phox, p40phox and p67phox), resides in the cytosol of resting cells. When PMNs are activated, the NADPH components (catalysing and regulatory) assemble at the phagosome membrane and azurophilc granules fuse, releasing MPO in the phagosome. This activation can occur directly, like for fMLP-mediated oxidative burst (El-Benna et al., 2016; Nguyen et al., 2017). However, most of the time this process of activation occurs as “multistep” process. Actually, it is possible to distinguish three different state: (i) resting, (ii) priming and (iii) activated state (Figure 21).

Figure 21. The NADPH oxidase in resting, primed and activated neutrophil. In the resting phase, the catalysing and regulator components are separated. Then during priming, the regulatory component moves close to the enzymatic one and various phosphorylation and conformational changes occur. Finally, activation induces the assembly of the complex and the production of superoxide anion. Adapted from El Benna J, 2016.

INTRODUCTION

The hypothesis of a multilayer activation of oxidative burst was originally suggested by the fact that several agents do not induce activation per se, but potentiate the activation induced by a secondary stimulus (El-Benna et al., 2016; Hallett & Lloyds, 1995). The intermediate state of activation, defined as priming state, prepares the NADPH oxidase for a stronger activation. There are different agents of priming. According the literature, it is possible to distinguish three classes: (i) the chemoattractant, including fMLP, C5a, LTB4 and PAF; (ii) cytokines, such as TNFα and GM-CSF; (iii) TLR agonist, like LPS (El-Benna et al., 2016). Each priming agent works in a different way. Overall, these molecules may upregulate different pathways converging on NADPH oxidase activation, promoting the partial phosphorylation, the redistribution and the conformational change of NADPH oxidase components (El-Benna et al., 2016; Sheppard et al., 2005).

3.4.2. Extracellular killing methods Neutrophils have different mechanisms for killing bacteria that are independent from the phagocytosis and that occur outside these cells.

3.4.2.1. Degranulation The degranulation is defined as the extracellular release of these granule- and vesicle-associated molecules, through a process of exocytosis (Faurschou & Borregaard, 2003). Neutrophils secretory vesicles have the higher propensity for degranulation, followed by gelatinase and specific granules. The vesicle degranulation has a key role in neutrophil extravasation and migration, since these vesicles are particularly rich in integrins, as Mac-1, chemoattractant receptors, like fMLP receptor and matrix degrading enzymes, such as metalloproteases. On the other hand, degranulation of the gelatinase- and specific- granules ensures that the arsenal of antimicrobial substances could be released at the exterior of the cells, to allow extracellular killing.

3.4.2.2. Neutrophil extracellular traps: NETs Brinkmann and colleagues described, for the first time, another mechanism of extracellular killing. The authors observed that human neutrophils stimulated with IL-8 or phorbol myristate acetate (PMA), show membrane protrusions, and more interestingly form some extracellular structures. Through electron microscopy and immunofluorescence experiments, the authors describe the composition of these extracellular structures, named neutrophil extracellular traps

INTRODUCTION

(NETs). These fibers are composed of chromatin and granule- and nuclei-derived proteins, such as neutrophil elastase and histones (Brinkmann et al., 2004).

The authors have demonstrated that the formation of these meshworks allow to trap and kill some gram-positive and -negative bacteria. Moreover, they have observed high abundance of NETs in acute inflammatory models in vivo, providing the proof that NETs are novel extracellular antimicrobial mechanisms of neutrophils (Brinkmann et al., 2004).

The formation of NETs, can occurs through two distinct pathways: (i) the classical NETosis and (ii) the non-lytic NETosis (Papayannopoulos, 2018). (Figure 22)

Figure 22. NET formation. Adapted from Papayannopoulous, 2018.

The classical NETosis is an alternative death by necrosis or apoptosis. During NETosis, nucleus reshapes and the eu- and heterochromatin are merged. Then the intracellular membranes are degraded, allowing the mix of decondensed chromatin with cytoplasmatic and granule

INTRODUCTION components. Finally the NETs emerge from the cells as the cytoplasmatic membrane is ruptured through an active cell death, different from apoptosis and necrosis (T. A. Fuchs et al., 2007).

Otherwise, some studies have demonstrated that Staphylococcus Aureus can trigger a rapid release of NETs in absence of cell death. This mechanism occurs in two distinct steps: first nuclear chromatin is extracellularly released; then the proteins, discharged through degranulation, assemble with the chromatin, allowing an extracellular building of NETs. Moreover, this alternative non lytic NETosis leaves behind a phagocytic cytoplast still active, thus able to ingest microorganisms (Papayannopoulos, 2018).

Regardless of the mechanism of formation, it is well accepted today that NETs ensure a high local concentration of antimicrobial agents, such as neutrophil elastase, MPO, cathelicidins and defensins, to degrade virulence factors and kill bacteria (Brinkmann et al., 2004). Moreover, NETs avoid bacterial systemic spread from a specific locus of infection (Walker et al., 2007).

Although NETs do not kill bacteria through the ROS toxic function, this antimicrobial mechanism is strongly associated to the ROS production. Indeed, the chemical inhibition of NADPH oxidase avoids the NETs production and treatment with glucose oxidase, (trigger of ROS production), rescues the NETosis (T. A. Fuchs et al., 2007). Since NETosis is a quite recent discovery in innate immune defence, the underlying mechanisms that lead to the formation of NETs are not totally understood. The MPO and the neutrophil elastase (NE) seem to have key roles for the cellular depolarization and chromatin decondensation. Indeed, NADPH oxidase stimulates MPO to trigger the activation and the translocation of NE from azurophilc granules to the cytoplasm, for actin degradation and successively to the nucleus, where the enzyme catalyses the histone processing and the chromatin decondensation. Another enzyme involved in this step of ROS-mediated NETosis is the protein arginine deiminase type 4 (PDA4). PDA4 catalyses the citrullination of histones, contributing to the chromatin decondensation (Papayannopoulos, 2018). This molecular model is considerably debated, and even less is known about the first step of nuclear delobulation and degradation of the nuclear envelope. A study has suggested that this essential step, that separates NETosis form classical cell death, could require LL-37 activity. This AMP damages the bacterial membranes, and the authors found that LL-37 treatment increased the NETs formation in

INTRODUCTION human neutrophils in vitro, leading to the destruction of nuclear membranes (Neumann et al., 2014).

Considering the variance and efficiency of neutrophil antimicrobial mechanisms, bacteria, in turn, have developed a number of strategies to resist to the phagocytosis, killing by ROS and by NETs. Of course, these strategies are specific of the mechanisms by which the bacteria try to resist and are species- and strain- specific. In the next chapter, I will develop the strategies that GAS has evolved to resist to neutrophils.

INTRODUCTION

4. GAS

Streptococcus pyogenes, also known as group A streptococcus (GAS), is a gram-positive, catalase-negative, anaerobic coccus. GAS is an important human pathogen, responsible for a broad spectrum of human diseases. S. pyogenes preferentially colonizes human epithelia, especially throat and skin epithelium (Walker et al., 2014).

GAS provokes superficial infection of these epithelia, such as impetigo (skin), pharyngitis and scarlet fever (throat). However, GAS can also cause invasive infection, spreading bacteria in sterile sites far from the infected epithelium. The invasive infection can be further classified in non-severe, like bacteremia and cellulitis; and severe invasive infection, as necrotizing fasciitis (NF) and streptococcal toxic shock syndrome (STSS). Finally, GAS infection can trigger serious post-infectious immune-mediated disorders, or immune sequelae, like the poststreptococcal glomerulonephritis (PSGN), the acute rheumatic fever (ARF) and the rheumatic heart disease (RHD) (Carapetis et al. 2005; Walker et al., 2014). In 2005 the World Health Organization (WHO) ranks GAS as the ninth leading infectious cause of human mortality, with the majority of death attributable to invasive infections and RHD (around 500000 deaths/year) (Carapetis et al., 2005; G. E. Nelson et al., 2016). Epidemiological studies have been crucial to monitor changes in the worldwide appearance and distribution of the GAS- mediated diseases (Steer et al. 2009). In relation to our study, I will discuss more in details in the next chapter the severe invasive diseases caused by the clone M1T1

4.1. Epidemiology of GAS infection 4.1.1. Superficial GAS infection Pharyngitis: GAS is the most common pathogen responsible for pharyngitis, responsible for over 600 million cases each year. The clinical manifestation of GAS-mediated pharyngitis is the inflammation of the pharynx and tonsils. The disease spreads through oral transmission, via nasal secretion or saliva droplets from carriers or infected subjects. In developed countries around 15% of schoolchildren and 4 to 10% of adults may suffer of one episode of pharyngitis each year. The incidence is 5 to 10 times higher in developing countries (Carapetis et al., 2005).

Scarlet fever: Occasionally GAS-mediated pharyngitis can be associated to scarlet fever, that manifests with sore throat, fever and the characteristic deep red, papular erythematous rash on the tongue, indicated as “strawberry tongue”. This pathology is due to infection with a GAS

INTRODUCTION strain that secrete exotoxins, especially SpeA (see below) (Walker et al., 2014). The morbidity and mortality of scarlatina significantly decreased in the last 200 years.

Impetigo: GAS superficial cutaneous infection, forming pustules that gradually expand and break. The infection spreads by direct skin contact and is frequent in children living in tropical and subtropical climes, with poor hygiene. This cutaneous manifestation counts 111 million of cases per year (Carapetis et al., 2005).

4.1.2. Immune sequelae Acute rheumatic fever: a symptomatic disorder that may occur following untreated GAS pharyngitis. The main clinical manifestations are inflammation (i) of the joints, i.e. arthritis (60 to 80% of cases), (ii) of the heart, i.e. carditis (30 to 45% of cases) and (iii) less frequent neurological symptoms (10% of cases). ARF is one of the main cause of morbidity and mortality, since this inflammatory disease usually results in rheumatic heart disease, RHD. This complication is quite frequent and highly harmful, with 282000 new cases and 233000 deaths each years (Carapetis et al., 2005).

Acute post-streptococcal glomerulonephritis: is an inflammatory disease affecting kidneys. The clinical symptoms include oedema, hypertension, urinary sediment abnormalities and lower levels of complement components in the bloodstream. Globally there are 470000 cases per year and 5000 deaths worldwide (Carapetis et al., 2005). Poor hygiene and poverty are among the principal risk factors of this clinical manifestation of GAS infection (Walker et al., 2014).

4.1.3. Invasive disease Some GAS strains are able to break the physical barrier of the epithelia and to bypass the immune defence, giving rise to invasive infections. The ability to invade uninfected tissues is influenced by different parameters, including bacterial factors, as GAS strain and subtype (see below) and host factors, like the resistance and efficacy of host innate immune system. Different epidemiological studies have shown that GAS invasive infections have a higher occurrence rate in elders worldwide (Lamagni et al., 2008; G. E. Nelson et al., 2016; O’Grady et al., 2007). The principal predisposing factor for invasive infection is an acute skin breakdown, including skin lesions, burns, blunt trauma and surgical wounds (Lamagni et al., 2008; G. E. Nelson et al., 2016). About 20% of invasive cases of S. Pyogenes infection die within 7 days after diagnosis (Lamagni et al., 2008; G. E. Nelson et al., 2016). Invasive infections include the (i) less severe infections (bacteraemia and cellulitis), and the (ii) more severe (necrotizing fasciitis and

INTRODUCTION streptococcus toxic shock syndrome). Overall, the degree of severity is associated to the rate of mortality.

The bacteraemia is defined as the presence of GAS in the bloodstream and it is usually associated with fever, nausea, as result of a burst of proinflammatory cytokine response. S. pyogenes can (i) directly colonizes the bloodstream, during the childbirth (puerpal sepsis) or after wounding; or (ii) subsequently reaches the blood flow after superficial infection of skin or throat (Walker et al., 2014). Cellulitis is a GAS infection affecting the inner layer of the skin, in particular the dermis and hypodermis. This cutaneous infection is characterized by red, inflamed, swelled and painful skin.

Cellulitis and bacteraemia are the less severe and the more frequent manifestations of invasive GAS infection, representing 20 to 40% of invasive disease cases.

The necrotizing fasciitis, also known as flash-eating disease, occurs when GAS skin infection extends to and degrades fascia and underlying muscle, causing tissues necrosis (gangrene). Actually, different multiple aerobic and anaerobic microorganisms can be responsible of NF, however I will focus my attention on necrotizing fasciitis caused by S. pyogenes. Compared to bacteraemia and cellulitis, necrotizing fasciitis is less frequent, 6 to 8% of case patients (Lamagni et al., 2008; G. E. Nelson et al., 2016; Walker et al., 2014). GAS gangrene begins at a site of trivial or even unapparent skin injury. At the beginning, lesion looks like an area of mild erythema, but undergoes a rapid evolution over the next 24–72 hours. Skin inflammation becomes more pronounced and extensive. By the fourth to fifth day, the affected skin is thoroughly changed: skin histological structure is completely loss and tissues are totally necrotised. The process may advance inexorably over large body areas unless measures are taken to stop it. The patient with streptococcal gangrene appears perilously ill, with high fever and extreme prostration (Stevens, 1992). NF develops rapidly and violently, therefore it is associated to a highest rate of mortality. It was reported that the 32% (Europe) and 24% (United states) of cases of S. Pyogenes-mediated necrotizing fasciitis, die within 7 days after infection (Lamagni et al., 2008; G. E. Nelson et al., 2016); and this rate is even worst in developing countries (Walker et al., 2014). Despite superficial infections, that are usually treatable with antibiotics, invasive infections, especially NF, are more complicated to treat and often require surgical intervention to remove necrotized sections.

INTRODUCTION

Streptococcus Toxic Shock Syndrome (STSS) is characterized by high fever, rapid-onset hypotension and accelerated multiorgan failure. This rapidly progressing illness is due to super antigen production and occurs in conjunction with invasive GAS infection. Considering the violence and the symptoms of this disease, 23 to 81% of patients with STSS die within one week post-infection (Walker et al., 2014). Necrotizing fasciitis is the invasive infection that more commonly give rise to and is linked to STSS (Lamagni et al., 2008).

The variability of clinical outcomes in response to GAS infection is, at least in part, due to the variability of GAS types or serotypes. Therefore, an important part of epidemiological surveillance for GAS disease has been the typing of collected bacterial isolates. The archaic method used to classify GAS was typing the bacterial isolates according to the M protein, a dominant surface antigen. Today GAS bacteria are classified according the sequence of 5’ variable region of the emm gene, encoding for M proteins. To date, more than 200 emm sequence types have been described (Cole et al. 2011). The emm geographical distribution is highly variable, presenting a significant difference between high-income countries, Africa and Pacific region (Steer et al., 2009). Furthermore, emm types and subtypes show temporal variability, with prevalent emm types in a particular year being substituted by other different emm types in successive years. Moreover, within an individual emm type, the subtypes can be deeply different. All these evidences suggest that emm type and subtype present a prevalence pattern driven by a dynamic feedback between host and GAS, that implies emm natural selection by host immunity. This variable prevalence pattern explains the resurgence of severe invasive streptococcal diseases in the 1980s.

In the 19th century, GAS infections were associated with invasive and often fatal diseases, as the scarlet fever in United States and Great Britain. This invasive trend extended up to 1920s, when the severity of illnesses strongly dropped. However, at the end of 1980s it was observed the recrudescence of several and fatal forms of invasive GAS infection. The return of severe invasive group A streptococcal infections was mainly due to the appearance and global dissemination of a new subtype of the serotype M1, named M1T1 (Aziz & Kotb, 2008).

4.2. M1T1 : GAS invasive strain M1T1, commonly isolated from both non-invasive and invasive infection cases, is most frequently associated with severe invasive infection, like NF and STSS. M1T1 was originally described in a study of 1992. The genomic DNA of 19 isolates from severe invasive cases and

INTRODUCTION of 48 isolates from uncomplicated pharyngitis or superficial skin infections were submitted to restriction-enzyme digestion and electrophoresis. Clearly and colleagues have described two main restriction-fragment profiles, that they named “invasive” and “non-invasive” profiles. They observed that 90% of the isolates from cases with serious disease have the same “invasive” profile, suggesting that the emergence of one high virulent variant of pre-existing strains could explain the increase of severe invasive GAS infections registered in the late 1980s (Cleary et al., 1992). Then, this clone was better characterized and described as possessing 70 kb of phage DNA. Two different phage sequences constitutes the majority of the genetic difference with other close M1 strains (Cleary et al. 1998). These two prophage sequences encode for SpeA2 and Sda1 (Aziz et al., 2005); the introduction of these prophages into the M1T1 clonal strain has contributed to the critical difference in epidemiologic property and invasiveness of this strain compared to the ancestral M1 clone. The acquisition of prophages took away certain genes and introduced new ones; this process usually contributes to the emergence of new strains.

SpeA2 is an important and potent streptococcal superantigen. The supernatigens (SAgs) are a family of secreted heat-stable exotoxins (Peterson, McCormick, & Schlievert, 2006; Spaulding et al., 2013). In contrast to conventional antigens, SAgs bind as unprocessed molecules to major histocompatibility (MHC) class II molecules. This binding triggers excessive stimulation T lymphocytes. This leads to a massive release of T cell mediators and pro-inflammatory cytokines contributing to toxic shock syndrome (Peterson et al., 2006). All GAS strains have a wide plethora of SAgs, however each strain has a different combination of these exotoxins (Aziz & Kotb, 2008). Interestingly, SpeA encoding gene complies with the periods of higher invasiveness of GAS. Indeed, this gene was seen in M1 isolates recovered in 19th and early 20th, while SpeA encoding gene had almost disappeared between 1920s and 1980s, period of decrease of GAS invasiveness. Likewise, the reintroduction of SpeA gene, by SpeA2 allele appearance, occurs simultaneously with the resurgence of GAS-mediated invasive diseases both in Europe and in the United states (Aziz & Kotb, 2008).

Sda1, is a potent streptococcal nuclease. The streptococcal nucleases, or streptodornases, are secreted extracellular nucleases, guaranteeing bacterial virulence. The importance of nucleases for GAS virulence and aggressiveness is known since 1950. An early study described the key role of streptococcal deoxyribonucleases in digesting the purulent materials from pyogenic

INTRODUCTION infection (Tillett, Sherry, & Christensen, 1947). Despite this seminal hypothesis that suggests the involvement of bacterial DNAse in liquifying pus, more recently Sumby and colleagues have demonstrated that DNAse activity is essential for the virulence of GAS infection, affecting neutrophil-mediated extracellular killing in vitro. Referring to the work of Brinkmann and colleagues (Brinkmann et al., 2004), the authors suggest that DNAses ensure more invasive infection by triggering the NET degradation (P. Sumby et al., 2005). Finally, it was demonstrated that Sda1, the M1T1 DNAse promotes resistance to neutrophil killing through NETs degradation both in vivo and in vitro (Buchanan et al., 2006). Interestingly overexpression of Sda1 in other non-invasive GAS serotype, like serotype M49 strongly increase the bacterial degradation of NET in vitro and the invasiveness of the serotype. These evidences highlight that Sda1, unlike other nucleases, appears to have a driving role in virulence, directly increasing GAS resistance to NETs killing (Buchanan et al., 2006) and promoting infection systemic spread in vivo (Walker et al., 2007).

However, host predisposition and M1T1 virulence factors cannot totally explain the outcome variability, since this emm serotype gives rise to both superficial pharyngitis and NF. Different studies have highlighted phenotypical (e.g. hyaluronic acid capsule) and transcriptional (speB expression) differences between M1T1 isolates obtained from NF or from superficial infection cases (Chatellier et al., 2000; Kansal et al., 2000; Levin & Wessels, 1998). Engleberg and colleagues elucidated the mutations responsible for these differences: mutations of genes encoding for the two-component regulatory system CovRS (Engleberg et al., 2001). The covS is a membrane associated kinase that, in response to environmental conditions, phosphorylates and thus modulates the activity of the regulatory subunity covR, a DNA-binding protein. Once activated, covR modulates target gene expression. The authors demonstrate that M1T1 virulence is instigated by spontaneous mutation of the covRS (Engleberg et al., 2001); furthermore it was demonstrated that these spontaneous mutations occurring during infection are the driving force for the phenotypical switch from a “pharyngitis” to “invasive” transcriptome profile (Sumby et al., 2006). The principal genes that undergo different expressions after covRS mutations are summarized in the table 1.

INTRODUCTION

Gene expression level of covRS-mutants Gene Protein and function relative to wild type M1T1 sic Streptococcal inhibitor of complement + Has Hyaluronan synthase operon (hyaluronic ++ ABC acid capsule synthesis) SpyCEP CXC-chemokine protease ++ sda1 Extracellular streptodornase D, a DNAse + Emm M protein No change SagA Streptolysin S precursor - sagB Production of streptolysin S - sagC Production of streptolysin S - speA Exotoxin type A, GAS superanitgen + slo Streptolysin O ++ speB Exotoxin B, cysteine protease - ska Streptokinase + spd Dnase - Table 1. Genes with altered transcriptional profile following covRS mutation. ++ upregulated more than tenfold, + upregulated between twofold and tenfold; - downregulated between twofold and tenfold. Data from (Paul Sumby et al., 2006)

All these factors have key different roles in the invasiveness of GAS, ensuring the tissue degradation and/or the bypass of innate immune cell defence mechanisms. GAS-phagocyte interactions will be described in the following paragraph. Of particular interest is the regulation of SpeB. This protein is a secreted cysteine protease, expressed by almost all GAS isolates (Cole et al., 2011). SpeB contributes to and enables GAS adhesion and epithelial colonization, since it degrades extracellular matrix components, host antimicrobial peptides and cytokine precursors. However, this protease cleaves GAS virulence factors too (like M1, various SAgs, streptokinase (Ska) and Sda1) (see below) (Cole et al., 2011). Despite this paradoxical double function of SpeB, degrading both host and pathogen protein, the host pathogen-interaction “naturally” selects mutants for SpeB. Although the loss of SpeB prevents the degradation of host tissue and immune modulators, the same alteration allows a stabilization of the strongest invasive factors of M1T1, Ska and Sda1; and this advantage wins on the loss of tissue

INTRODUCTION degradation. As described before, Sda1 has a key role in GAS invasiveness promoting systemic bacterial dissemination in murine infection model (Walker et al., 2007). Another advantage associated to the loss of SpeB is the stabilization of Streptokinase (Ska). This enzyme catalyses the digestion of plasminogen in plasmin, that degrades fibrin clots supporting invasiveness (Cole et al., 2006). Since SpeB is, in turn, regulated by covRS system, this protease represents the essential link between the transcriptional and the post-transcriptional regulation of gene expression in GAS M1T1. The loss of SpeB synergizes with the transcriptional up regulation of Sda1 and Ska1, boosting their transcriptional up regulation.

4.3. GAS-phagocyte fighting As already discussed, to give rise to invasive infection, GAS must be able to bypass the innate immune cells recruited to the site of infection. These phagocytes are essential innate immune “body-guards”. Invasive pyogenic bacterial skin and soft tissue infections constitute a formidable challenge for neutrophils and macrophages, since GAS has evolved innumerable evasion mechanisms to avoid phagocyte killing (Figure 23).

4.3.1. Recruitment First of all, GAS presents strategies to avoid and or limit phagocyte recruitment to the infection site. To date, no one GAS factor that specifically cleaves macrophage chemoattractants has been described. We can speculate that the multiple GAS proteases could ensure this function, but there are no evidences of this function. On the other hand, GAS-mediated neutrophil recruitment hindrance is well described. The chemokines involved in neutrophil recruitment belong to the CXC chemokine family. GAS has a serine peptidase, named SpyCEP Streptococcus Pyogenes Cell Envelope Protease, that cleaves human CXCL1/Groα, CXCL2/Groβ and CXCL8/IL8 and the murine CXCL1/KC and CXCL2/MIP-2 in the same conserved site (Hidalgo-Grass et al., 2006; Kurupati et al., 2010b). This bacterial protease was originally described in 2005, when Edwards and colleagues demonstrated that GAS isolated from patients with invasive infections express a protease able to cleave human IL8 in vitro; the authors demonstrate that the cleavage determines the loss of the chemoattractant properties in vitro. Since this protease was isolated from more invasive strains and displays this chemokine proteolytic function, the authors suggest the SpyCEP involvement in GAS invasiveness during infection. Then Hidalgo-Grass and colleagues demonstrated that SpyCEP mutated strains determine less severe outcome in murine infection models. This less aggressive phenotype is

INTRODUCTION due to a major abundance of chemokines at the site of infection, and a consecutive major neutrophil recruitment. The authors demonstrate that SpyCEP critically impairs the PMN innate immune response in vivo (Hidalgo-Grass et al., 2006). Similarly, Streptococcal esterase (Sse) and streptococcal C5a peptidase A (Scp A) inhibit neutrophil recruitment, degrading respectively the chemotactic platelet activating factor and the complement factor C5a (Döhrmann, Cole, & Nizet, 2016).

4.3.2. Phagocytosis GAS has two different strategies to avoid phagocytosis: (i) GAS modulates bacterial surface composition in order to enhance the resistance to the phagocytosis or (ii) GAS kill phagocytes, avoiding the following engulfment. For the first strategy, the M protein and GAS hyaluronan (HA) capsule are essential (Moses et al., 1997). Indeed, the surface-attached M protein recruits (i) inhibitory regulators of complement system or (ii) binds fibrinogen, limiting the opsonization of bacterial surface (Oehmcke et al., 2010). The HA capsule inhibits the binding of antibodies and complement through molecular mimicry, since its composition imitates the host glycosaminoglycan (Dale et al., 1996). Otherwise GAS produces streptolysin S and O (SLS; SLO), exotoxins showing neutrophil and macrophage killing activity (Datta et al., 2005; Timmer et al., 2009). Indeed, compared to the WT strain, SLS-/- bacteria give rise to a mild infection in vivo and are less resistant to phagocytosis in whole blood (Datta et al., 2005). Similarly, mutation in SLO strongly impairs GAS-mediated apoptosis by phagocytic cells (Timmer et al., 2009). Even if these mechanisms, are efficient, GAS is not always able to escape the phagocytosis. However, different studies have shown that GAS is able to survive inside the phagocytes escaping from the phagosome (Medina et al., 2003; O’Neill et al., 2016). A recent study interprets this phenomenon as a strategy to improve GAS dissemination, from the infection site to sterile tissues, contributing to the development of invasive infections (Mishalian et al., 2011b).

4.3.3. Oxidative burst Upon phagocytosis, neutrophil and macrophages produce an oxidative burst resulting in the rapid release of ROS in the phagosome, allowing bacteria killing. The GAS SLO suppresses ROS generation in phagocytes. Uchiyama and colleagues, taking advantage of SLO-/- bacteria, demonstrate that SLO avoids neutrophil oxidative burst by modulating NADPH oxidase function (Uchiyama et al., 2015). Moreover, another study underlines the importance of SLO

INTRODUCTION for GAS survival after macrophage phagocytosis. The SLO produces pores in phagosomal membrane avoiding the acidification of this intracellular compartment. The H+ gradient is required to generate some reactive oxygen species but also for other bactericidal function, thus SLO impairs the intra-phagosomal killing of GAS in macrophages (Bastiat-sempe et al., 2014). Furthermore, GAS possesses also different indirect mechanisms that allow the resistance to ROS production; GAS has multiple enzymes that can catalyse superoxide detoxification (like SodA or NoxA) and some metal ion binding proteins that, chelating Fe2+, limit the formation of hydroxyl radical. Moreover, if ROS-mediated damages occur, GAS possesses different enzymes that ensure DNA and proteins repairs (Henningham et al., 2015).

4.3.4. Extracellular killing strategies: degranulation and NETs As discussed in the previous chapter, neutrophils present also some extracellular mechanisms of killing, NETosis. Evidently, GAS has evolved strategies to bypass also this defence mechanisms of the host. GAS has different strategies to escape from NETs; GAS can (i) avoid NETs formation, or (ii) can destroy the formed NETs. GAS prevents NETs formation, acting at different levels. As described in the original work describing NETs, IL8 promotes the NETs formation (Brinkmann et al., 2004); a brilliant study describes an alternative function of SpyCEP in avoiding NETs formation by degrading IL8 (Zinkernagel et al., 2008). As already discussed, the key actors involved, at the molecular point of view, in NETosis are MPO and NE (Papayannopoulos, 2018). It was demonstrated that streptococcus collagen-like (Scl1) protein prevents the MPO release suppressing NETs formation (Do¨hrmann et al., 2014). LL-37 is present in NETs and represents one of the harmful AMPs for GAS (Nizet et al., 2001). It was suggested that M1 protein binds to the antimicrobial peptide LL-37 to prevent NETs formation (Neumann et al., 2014) (Larock et al., 2015). Even if NETosis takes place, GAS has also different mechanisms to destroy NETs. Basically, NETs are made by DNA and granular protein and the GAS virulence factor mainly involved in NETs destruction is the DNAse Sda1. This DNAse has a key role in promoting GAS invasiveness in vivo (Buchanan et al., 2006),

INTRODUCTION suggesting that NETs digestion is a driving event in GAS invasive disease progression (see before) (Walker et al., 2007).

Figure 23. Neutrophil GAS fighting. Schematic representation of the neutrophil antimicrobial mechanisms and GAS counterattacks. Neutrophil, as described in the chapter 3 ensures host protection mainly through phagocytosis, oxidative burst, NETs (reported in black). GAS have developed various counterattacks (here written in violet, in grey boxes) to bypass neutrophil strategies. HA: hyaluronan capsule; SpyCEP: Streptococcus Pyogenes Cell Envelope Protease; Scl-1: streptococcus collagen like 1. Adapted from Dormann,2016.

4.4. GAS-keratinocyte fighting Despite the importance of phagocytes in fighting GAS infection, invasive infections require the tissue barrier degradation. GAS infection starts in epithelia and our study is based on subcutaneous (SC) model of infection; thus, the interaction taking place between Streptococcus Pyogenes and keratinocytes are of interest. As described in the second chapter, keratinocytes are active actors of immune response against various bacteria. In response to GAS infection, keratinocytes produce (i) AMPs, especially LL-37 (Braff et al.,2005), (ii) and pro inflammatory mediators, particularly the chemokine IL-8 (Persson et al., 2015). Despite well-characterized functions of LL-37 and IL-8, the signalling pathways regulating their expression in keratinocytes in response to GAS are poorly known. Persson and colleagues observed that M1 protein specifically interacts with TLR2, switching on an inflammatory intracellular signalling

INTRODUCTION cascade consisting of ERK, p38 and JNK. This signalling determines the subsequent activation and nucleus migration of the transcription factor NF-KB and AP-1. The authors demonstrate that IL-8 release is specifically associated to M1 and does not occur after M4, M5, M22 in vitro stimulation (Persson et al., 2015).

- Keratinocytes have also been shown to produce O2 in response to various infectious and inflammatory stimuli. A recent study has demonstrated that keratinocytes produce superoxide anions in response to GAS infection, inhibiting bacterial growth in vitro (Regnier et al., 2016). However, the ROS production has a side effect, causing keratinocyte cell death. GAS try to escape keratinocyte-mediated host protection, boosting cell death through SLO exotoxin (Cywes-Bentley et al., 2005).

4.5. GAS iron metabolism With the exception of the Lyme disease pathogen, Borrelia Burgdoferi, almost all microorganisms require iron for their biological function. S. Pyogenes, β-haemolytic bacteria, cannot live in absence of iron (Ge et al., 2009). GAS present three different iron acquisition system: (i) ferrous iron transporter, named MtsABC; (ii) heme transporter, named SiaABC; (iii) siderophore transporter, FtsABCD (Ge et al., 2009). In S. Pyogenes, ten genes constitute the operon sia, encoding for heme transporter; this genic investment suggests the central role of these system in iron metabolism of GAS. Shr is the Streptococcal heme protein receptor; Shp is the Streptococcal heme binding protein and the siaABC system consists of the heme-binding lipoprotein, the membrane permease and the ATPase. Since iron is indispensable for bacterial growth, a study investigated the importance of Shr in the virulence of the M1T1 group A streptococci. The authors observed that the survival of the mutant strain is significant lower compared to wild-type bacteria in whole blood, however both strains display exactly the same resistance to neutrophil killing. There is no difference neither in their resistance to LL-37 killing nor in their adherence to epithelial cells. When the authors use Shr-/- and wild-type bacteria in SC murine infection model, they found a larger lesion at the infection site, however they did not show data about the bacterial dissemination, which is the readout of the virulence of M1T1(Dahesh et al., 2012).

THE AIM OF THE STUDY

Hepcidin, known to be the key iron regulatory hormone, was originally described as an antimicrobial peptide (AMP). Today, the direct bactericidal activity of hepcidin is debated and considered to be weak compared to other specialized AMPs. However, the link between hepcidin and infectious conditions is largely recognized and investigated. Indeed, iron has a particular role in mediating host-pathogen interactions, since this metal is at the center of a nutritional battle between host and pathogens. Iron is the only micronutrient, whose hormonal regulation depends on both the nutritional and infection status. For all these reasons, during the last 2 decades, a lot of studies have focused their attention on the hyposideremia of infection, investigating the importance of hepatic hepcidin during various infections.

Actually, the antimicrobial peptides ensure host protection through two different mechanisms: (i) the direct killing of pathogens (bactericidal function) and/or (ii) the modulation of host immune cells (immune modulatory function). Despite the pleiotropic roles of AMPs, to date the only aspect investigated for hepcidin antimicrobial activity was the direct killing activity.

During my thesis, I first investigated the immunomodulatory potential of hepcidin. Unpublished data from our laboratory highlighted structural similarities between human hepcidin and IL8. As described before, IL8 is the key chemokine required for neutrophil recruitment and activation. To test if this structural similarity has a functional meaning, we assessed the hepcidin chemoattractant potential in vivo by performing subcutaneous injections of exogenous hepcidin and measuring the MPO activity in vivo by bioluminescence. We then carried out a migration assay in vitro, to confirm the direct effect of hepcidin in neutrophil recruitment.

In agreement with their functions, AMPs are mainly expressed in the epithelia, such as skin and intestine. Thus, I next investigated hepcidin expression and functional role as AMP in the skin, a major epithelium in host defense. We challenged mice with SC injection of the M1T1 5448 GAS serotype. This model of infection is particularly attractive because it reproduces the core features of necrotizing fasciitis disease: (i) skin and soft tissue lesion with high bacterial proliferation, (ii) extensive recruitment of innate immune cells, of which the majority are neutrophils, and (iii) systemic dissemination of infection four days after infection. In order to discriminate the possible contribution of myeloid cells and keratinocytes, we assessed the

THE AIM OF THE STUDY outcomes of infection (lesion size, CFU at the lesion site and in peripheric organs) in hepcidin keratinocyte- and myeloid-specific KO animals, comparing them to WT littermates.

RESULTS

1. Hepcidin shares structural similarities with IL8 and displays neutrophil chemoattractant properties. Hepcidin contains four disulfide bridges, highly conserved among different vertebrates. This kind of functional group is recurrent within AMPs, peptides that ensure host protection mainly through antimicrobial function but also through the modulation of host immunity (chemotaxis, immunomodulatory effect…) (Zhang et al., 2016). Curiously, these disulfide bridges resemble the intramolecular disulfide bonds critical for the function of chemokine proteins (C, CC, CXC,

CX3C). We asked whether hepcidin could share structural similarities with chemokines. In collaboration, Synsight, a computer-aided drug design company, performed a 1-D and 3-D sequence alignment of human hepcidin with a list of human known chemokines (CCL5, CCL11, CCL20, CCL27, CXCL11, CXCL12, CXCL14, XCL1, IL7 and CXCL8/IL8) for a wide range of immune cells: monocytes, macrophages, neutrophils, Th1, Th2, NK, eosinophils, basophils, etc…Among all the alignment performed, hepcidin displayed the highest number of primary and structural similarities with IL8, the functional mouse CXCL1/2 homologue (Figure 24a). The whole crystal structures of hepcidin and IL8 are represented in Figure 24b depicted in green and in indigo, respectively. The sequence alignment outcomes between hepcidin and IL8 showed two common motifs broadly aligned. In particular, the first motif, that aligns ten aa of hepcidin with IL8, is here depicted in red and salmon. The beta sheet and the beta-hairpin which include disulfide bridges, represent a common 3D motif between these two proteins, suggesting a functional similarity of hepcidin with interleukin IL8.

Given the well-known pleiotropic roles of IL8 in neutrophil recruitment and activation, it was tempting to speculate similar roles for hepcidin.

The myeloperoxidase (MPO) is one of the key components of the neutrophil defence armamentarium and the activity of this enzyme constitutes a faithful marker of neutrophils. The activity of this oxidase follows the neutrophil activation state: while in a resting neutrophil MPO is inactivated and segregated in the azurophilic granules, in an activated neutrophil this enzyme is released in the phagosome to produce ROS (see section 3.3.1.3). Neutrophil recruitment and activation are tightly associated. Thus, the measure of the MPO activity in a tissue mirrors both the activation state and the abundance of these phagocytes. Luminol, a redox-sensitive compound, emits luminescence when exposed to appropriate oxidizing agents, such as ROS produced through MPO activity during the phagocytic oxidative burst (Figure 24c). Upon

RESULTS systemic administration (i.p. injection), luminol enables non-invasive bioluminescence imaging of MPO activity in vivo (Gross et al., 2009). Taking advantage of this powerful tool, we first measured the ability of hepcidin to recruit neutrophils by measuring MPO activity in vivo. Mice were SC injected with hepcidin (right paw) and its vehicle (left paw) (group I). The bioluminescence was measured at different time points after SC injection (3, 6 hours) as schematized in Figure 24d. Topical application of PMA onto the earlobes of mice induces acute inflammation of the skin, associated with local swelling, erythema and neutrophil infiltration. We used this well-established model as positive control in our bioluminescence measurements. The bioluminescence fold-change of PMA compared to its vehicle gives a measure of the luminol efficacity, thus allowing to normalize the bioluminescence measurement and to compare the values between different experiments. Here, I show some representative bioluminescence imaging photos, 3 hours after the SC injection. Interestingly, we observed that the injection of hepcidin induced a stronger emission of bioluminescence (Figure 24f) compared to the injection of vehicle (Figure 24e). The bioluminescence emitted from the site of injection of hepcidin was 3.6-fold over the background level of the vehicle-injected paw (Figure 24i). In order to determine if this induction was biologically relevant, we examined the emission of bioluminescence after a SC injection of CXCL1, the key neutrophil chemokine (group II). As expected, the injection of CXCL1 induced a stronger emission of bioluminescence (Figure 24h) [3.3-fold change] compared to the injection of the vehicle (Figure 24g). Interestingly, this level of induction was surprisingly comparable to that mediated by hepcidin (Figure 24i). These data demonstrate that hepcidin induces MPO activity in vivo at similar level than CXCL1, a key neutrophil chemokine.

RESULTS

Figure 24. Hepcidin aligns with IL8 and induces MPO activity in vivo. (a) 1D sequence alignments and overlay 3D structures between hepcidin and IL8. The 3-D structures of hepcidin (PDB code: 2KEF

RESULTS

(Jordan et al., 2009) ) have been used as reference and compared to IL8 (PDB code: 5D14). The conservation line output are represented by three characters: “*” indicates positions which have a single, fully conserved residue - “:” indicates that one of the following 'strong' groups is fully conserved: STA, NEQK, NHQK, NDEQ, QHRK, MILV, MILF, HY, and FYW. - “.”indicates that one of the following 'weaker' groups is fully conserved: CSA, ATV, SAG, STNK, STPA, SGND, SNDEQK, NDEQHK, NEQHRK, FVLIM, and HFY These are all the positively scoring groups that occur in the Gonnet Pam250 matrix (Pearson, 1990). The strong and weak groups are defined as strong score > 0.5 and weak score =< 0.5 respectively. For profile alignments, secondary structure and gap penalty masks are displayed above the sequences, if any data is found in the profile input file. (b) On the left, 3D representations of secondary structures of hepcidin (in top) and interleukin IL8 (bottom), respectively colored in green and indigo. On the right, representation of 3D alignments of superimposed structures (in top) and individual (bottom) hepcidin HEP in green and interleukin IL8 in blue. Staining of the 1D alignments of the blocks of sequences: Block 1 (HEP, red, IL8, salmon) and Block 2 (HEP, blue, IL8, purple). The sequence alignment outcomes between hepcidin and IL8 showed a first common motif colored in red and salmon. (c) Schematic representation of the biochemical basis of the luminol bioluminescence during phagocyte respiratory burst. The luminol produces light reacting with HOCl, a product of MPO activity. Otherwise MPO mediates luminol oxidization, catalyzing the reaction of this compound with others ROS, such as O2 .-, generating light. (d) Schematic representation of the protocol adopted for noninvasive bioluminescence imaging. C57BL6 mice were SC injected with hepcidin or CXCL1 (right paw) and their vehicles (PBS or PBS/BSA 0.1%) (left paw). As positive control dermatitis are generated by topical application of PMA on the right ear lobe, the left ears served as vehicle (acetone) controls. One, three and six hours after the SC injections and PMA application, luminol was administered (i.p.) and mice are imaged. At the end animals are sacrificed and sites of SC injections are recovered. (e-h) two representative mice for each treatment, n=4 for each group. (i) Luminol bioluminescence (mean +/- s.e.m.) quantified as fold increase over vehicle treated paw for each mouse at three hours after SC.

The measurement of MPO activity can be influenced by the abundance and/or by the activation state of neutrophils. We were then interested in discriminating if hepcidin-mediated MPO activity was really due to a recruitment of these phagocytes. For this reason, at the end of the kinetic (Figure 24d), we sacrificed the mice and recovered the murine skin at the site of

RESULTS injection to perform a polymorphonucleocytes (PMN) staining. We did not observe any intradermal neutrophil infiltration in response to vehicle injection as illustrated in the figure 2a (top panel, on the left). On the contrary, hepcidin subcutaneous injection induced an active extravasation and migration of neutrophils in the dermal compartment. (Figure 25a). Indeed, we detected neutrophils (i) close to the vessels, immediately after extravasation (figure 25a, on the right) and (ii) in the intradermal space (Figure 25a, on the middle). Skin is characterized by an ‘immune barrier’, consisting in resident immune cells, that broadly reside in the dermis and epidermis. However, neutrophils do not reside in this tissue in physiological conditions and are rather actively recruited in response to suitable chemoattractant stimuli (Eyerich and al., 2018). Thus, even if we observed a moderate number of neutrophils in the hepcidin-injected group, the presence of these phagocytes clarified that the MPO activity measurement in vivo is due to a chemoattractant potential of hepcidin. We next aimed to investigate whether the chemoattractant role of hepcidin was direct or indirect. Therefore, we assessed the ability of hepcidin-25 to attract purified neutrophils in a classical migration assay in vitro with the Boyden chamber method in presence of increasing concentrations of hepcidin into the lower wells. Hepcidin increased the number of recruited neutrophils to the lower wells up to 2-fold (figure 25b). CXCL1 triggered a more potent neutrophil migration peaking at approximately 8-fold the control migration value (figure 25g). Nevertheless, hepcidin induced a dose-dependent migration of neutrophils in the typical bimodal manner (De et al., 2000; Grigat et al., 2007; Rohrl et al., 2010), revealing that hepcidin possesses chemoattractant properties. This bimodal pattern is characteristic of chemokine-mediated neutrophil migration in vitro (De Yang et al., 2000; Grigat et al., 2007; Rohrl et al., 2010). The peak response was observed at 36 nM, a concentration at which various antimicrobial peptides exert their maximal chemoattractant action (Grigat et al., 2007; Rohrl et al., 2010). This effect was lost when hepcidin was added in the upper well, ruling out a random migration (Figure 25d). The murine mini hepcidin, a shorter form containing the first 9 N-terminal amino acids of hepcidin (Preza et al., 2011), was not able to trigger neutrophil migration, highlighting that the 3D structure is essential for this new immunomodulator function of hepcidin (Figure 25c). To investigate the signalling pathways involved in hepcidin-mediated neutrophil migration, we used a pharmacological approach. We first assessed whether this effect was dependent on the G protein-coupled receptors. The pertussis toxin (PTX) prevents the interaction of G proteins with their coupled receptors on the cell membrane. Pre-treatment of neutrophils (Figure 25e) with PTX inhibited the hepcidin-

RESULTS mediated chemotaxis suggesting the involvement of G-proteins. We next tried to individuate the upstream receptor required for the hepcidin-mediated recruitment of neutrophils. Considering the structural affinity between hepcidin and CXCL1 (figure 24a, b) it was tempting to test whether these two molecules shared the same receptor. Thus, among the different receptors coupled to G protein, we chose to investigate the role of CXCR2, the CXCL1 receptor. We observed that the chemical inhibition (Figure 25e) or the neutralization (Figure 25f) of CXCR2 prevented neutrophil migration, suggesting that hepcidin-mediated chemotaxis requires this receptor. Both in vitro and in vivo data are coherent and support a new chemoattractant function of hepcidin. However, while CXCL1 is a more potent neutrophil chemoattractant than hepcidin in vitro, the level of hepcidin-mediated MPO activity was similar to the one induced by CXCL1 in vivo. Therefore, hepcidin could not only directly recruit neutrophils, as shown by the Boyden chamber assay but also immunomodulate other cells that in turn could contribute to the neutrophil recruitment. Of course, this putative indirect action would be lost in the in vitro assay, giving rise to a plausible explanation for the strength difference between in vitro-in vivo phenomena.

Altogether these results highlighted a new immunomodulatory role of hepcidin. Thus, our interest shifted on these new actions rather than on the, already studied and largely discussed, antimicrobial role of this AMP.

RESULTS

Figure 25. Hepcidin induces neutrophil migration in vivo and in vitro. (a) Histological analysis of the region of the skin injected showing polymorphonucleocyte recruitment (pointed out by white arrows)

RESULTS

6 hours after hepcidin injection. Neutrophils are stained in brown, samples are counterstained with hematoxylin. (b) Migration assay of neutrophils with the Boyden chamber method in presence of increasing concentrations of hepcidin into the lower wells. n>3 experiments (with n=3 per group); mean +/- s.e.m. Statistical analysis by a one-way analysis of variance (ANOVA) followed by a Bonferroni posttest. **p< 0.01.; ****p< 0.0001. (c) Migration assay of neutrophils with the Boyden chamber method in presence of increasing concentrations of mini-hepcidin into the lower wells. n=3 per group; mean +/- s.e.m. (d) Migration assay with the Boyden chamber method in presence of hepcidin (36 nM) added or not (-) to the upper (u.w) or lower (l.w.) wells. n=3; mean +/- s.e.m. (e) Hepcidin was added to the lower wells in presence of PTX or the chemical CXCR2 inhibitor Sb2250052. n=3 per group; mean +/- s.e.m. Statistical analysis by a one-way analysis of variance (ANOVA) followed by a Bonferroni posttest. **p< 0.01; ****p< 0.0001. (f) Hepcidin was added to the lower wells in presence or a neutralizing CXCR2 antibody. n=3 per group; mean +/- s.e.m. Statistical analysis by a one-way analysis of variance (ANOVA) followed by a Bonferroni posttest. **p< 0.01.; ***p< 0.001; ****p< 0.0001. (g) Migration assay with the Boyden chamber methods in presence of increasing concentration of CXCL1 into the lower wells. n=3 per group; mean +/- s.e.m. Statistical analysis by a one-way analysis of variance (ANOVA) followed by a Bonferroni posttest. **p< 0.01.; ***p< 0.001; ****p< 0.0001.

Hepcidin is expressed in the epidermis To determine the physiological relevance of this new hepcidin immunomodulatory activity, we decided to explore its expression and its role in the skin epithelium. To date some studies have described hepcidin expression in some epithelia, such as renal and gastric epithelium, both in basal and in infectious conditions. However, hepcidin expression in the skin and more specifically in keratinocytes, the main epidermal source of antimicrobial peptides and of immunomodulatory mediators, has never been investigated.

We first examined the expression of hepcidin using a human immortalized keratinocyte cell line, HaCaT (Figure 26a). These cells were cultured for one week in low or high calcium media.

The high calcium medium (supplemented with 2,8 mM CaCl2) mimics the physiological calcium gradient within the epidermis, required for the majority of the differentiation processes (E. Fuchs, 1990b). This protocol allows to mimic in vitro the physiological keratinocyte differentiation. As expected, the keratinocyte differentiation markers, keratin 1 and keratin 10, were significantly induced by higher concentration of calcium. Interestingly hepcidin

RESULTS expression followed the keratin 1 and keratin 10 expression pattern, showing a significantly higher expression in high calcium conditions (Figure 26a). These results suggest that hepcidin, as other AMPs, is preferentially expressed in more differentiated, [thus more superficial and close to external environment] keratinocytes. Although very simple to manage, being an immortalized cell line, HaCaT cells are not the more suitable model to study the keratinocyte differentiation process in vitro, since physiologically this differentiation program entails and requires end cellular death. For this reason, I optimized a protocol for culturing primary keratinocytes isolated from epidermis of newborn mice. This protocol allows to obtain 8-10 million of cells per epidermis. As shown by immunofluorescence in Figure 26b, around 95% of isolated cells are positive for the keratin 14 staining and can be considered as keratinocytes. These primary cells are a powerful tool to study the gene expression throughout keratinocyte differentiation. The high calcium medium triggered the morphological changes of cellular layers: 48 hours after the addition of calcium, keratinocytes looked more enlarged and flattened, undergoing squamous cell differentiation (Figure 26c, right panel). Moreover, “differentiated” primary keratinocytes expressed the keratin 10, as shown by immunofluorescence (Figure 26d) and by western blot (Figure 26e). Thus, this model recapitulates the keratinocyte differentiation in vitro. Then, we investigated hepcidin expression throughout primary keratinocyte differentiation. In agreement with the data from the HaCaT cell line, hepcidin expression was strongly upregulated in differentiated keratinocytes (Figure 26f). Interestingly, hepcidin expression pattern is close to that of cramp (Figure 26f). This latest gene encodes the murine cathelicidin, essential AMP involved in the host response to Group A streptococci in a murine infection model (Braff et al., 2005; Nizet et al., 2001).

RESULTS

Figure 26. Hepcidin is expressed in differentiated-like keratinocytes. (a) KRT1, KRT10 and HAMP expression by Q-PCR in HaCaT cells cultured for one week in low (1.2 mM of Ca2+) or high (2.8 mM of Ca2+) calcium media. n>3 experiments (with n=3 per group); mean +/- s.e.m. Statistical analysis by a

RESULTS

one-way analysis of variance (ANOVA) followed by a Bonferroni posttest. ***p<0.001.; ****p< 0.0001. (b) Kerartin14 staining in cells obtained from wild-type newborn skin. Around 95% of cells are positive. Negative control (right) corresponds to the omission of the primary antibody. Scale bar 50 μm. (c) Bright field microscopy images of murine primary keratinocytes growth in low (left) or high (right) calcium media. (Leica DMI3000B microscope, Leica DFC310FX camera [Leica LAS Core software]). (d) Immunofluorescence for keratin 10 in murine primary keratinocyte growth in high Ca2+ medium for four days, the negative control (left) corresponds to the omission of the primary antibody for Keratin10, nuclei are counterstained by Hoechst solution (Leica DMI3000B microscope, Leica DFC310FX camera [Leica LAS Core software]). (e) Keratin 10 expression in wild type primary keratinocyte cells cultured in low Ca2+ (line 1 and 3) and in high Ca2+ (line 2 and 4) by western blotting. (f) Hamp1and Cramp expression by Q-PCR in murine primary keratinocytes cultured for one week in low or high calcium media. n=3 per condition; mean +/- s.e.m. Statistical analysis by a one-way analysis of variance (ANOVA) followed by a Bonferroni posttest. ***p<0.001.; ****p< 0.0001.

Epidermal hepcidin has a protective role against GAS infection We decided to explore the role of hepcidin during Group A Streptococcal (GAS) infection, for different reasons: (1) hepcidin showed the same expression pattern than cramp (Figure 26f) (2) some unpublished observations from our group have described a strong increase of hepcidin expression in the skin of GAS-infected mice and (3) this infection is defined as a pyogenic disease, since it is associated with a massive infiltration of leukocytes, mainly neutrophils, to the site of infection. Thus, this infection model was ideal for investigating the function of epidermal hepcidin as AMP and, in particular, the hepcidin-mediated recruitment of neutrophils.

GAS, a beta haemolytic gram positive bacteria, preferentially colonizes human epithelia, giving rise to numerous diseases with various clinical manifestations (Cole et al., 2011). Necrotizing fasciitis (NF), also known as “flesh eating-bacterial infection” is one of the most violent and invasive manifestations of GAS infection. We had the possibility to investigate hepcidin expression in 5 skin biopsies of NF patients, with 4 out of 5 cases caused by Group A streptococcus (detailed in Table 2). As shown in the Figure 27a, the nature of the biopsies and the histology of the tissues were quite different among the 5 cases, however we detected hepcidin immunoreactivity in all the specimens (staining in brown). Interestingly we observed

RESULTS a strong expression of hepcidin in the keratinocytes at the basal layer, close to the basement membrane and interacting directly with the dermal compartment (Figure 27a). These evidences seemed in contradiction with the previous data showing the expression of hepcidin in differentiated keratinocytes (Figure 26). However, the delocalization according to the infectious/inflammatory skin state occurs usually for antimicrobial peptides, like LL-37 (Dorschner et al., 2001a; Murakami et al., 2002). All these evidences prompted us to investigate the role of epidermal hepcidin during this infection. Even if GAS preferentially give rise to human infection, some murine models of subcutaneous GAS infection have been developed (Datta et al., 2005; Nizet et al., 2001; Peyssonnaux et al., 2005) . The GAS inoculum was introduced subcutaneously (SC) into a shaved area on the flank of mice, allowing the development of a skin and soft tissue lesion that recapitulates NF lesion (Figure 27b). We found that hepcidin was expressed in the skin adjacent to the necrotic lesion. In particular, we observed, by immunohistochemistry (staining in blue, Figure 27b), that hepcidin was not only expressed in myeloid cells recruited to the site of infection, as previously reported (Peyssonnaux et al., 2006) but was also strongly detected in keratinocytes (Figure 27b). This pattern mirrors the expression of cramp in GAS-infected mice, with both epidermal (Braff et al., 2005b) and myeloid (Dorschner et al., 2001b; Nizet et al., 2001) cells producing this antimicrobial peptide. Therefore, we have been interested in studying the role and in discriminating the contribution of myeloid- and epidermal-derived hepcidin during S. pyogenes infection. We generated two murine models with a specific ablation of hamp1 in (i) myeloid cells (Figure 27c) and (ii) epidermis (Figure 27e). Hamp1 lox/lox mice were crossed with transgenic animals expressing the Cre recombinase under the control of the murine promotors LysM (Hamp1lox/lox/LysM Cre = Hamp1 Δmyeloid) or KRT14 (Hamp1lox/lox/KRT14 Cre = Hamp1 Δker). As expected deletion of the floxed hamp1 allele was observed in the epidermis of Hamp1Δker mice and in the bone marrow of Hamp1Δmyeloid mice (Supplementary figure 1a). The deletion efficiency of hepcidin in epidermis was ~80% in Hamp1 Δker while, as expected, there was no deletion in the liver and in the bone marrow of these mice (Supplementary figure 1b). The efficacity of deletion was ~87% in the macrophages and ~83% in the neutrophils derived from the bone marrow of Hamp1Δmyeloid mice, as determined by quantitative PCR on mouse genomic DNA (Supplementary figure 1c). Iron parameters (such as plasma iron, ferritin, and transferrin) and hematological parameters (such as RBC, Hct, Hgb, CCMH and TCMH) were unchanged between Hamp1lox/lox and Hamp1Δker mice (Supplementary figure 2a) and between

RESULTS

Hamp1lox/lox and Hamp1Δmyeloid (Supplementary figure 2b), confirming that the absence of extrahepatic hepcidin did not influence the systemic iron homeostasis (Zumerle et al., 2014).

Mice were injected with a GAS inoculum of 107 CFUs/animal and were daily monitored (weight variation and lesion size). After four days mice were scarified; skin ulcer, spleen and blood were recovered for counting the bacteria (colony forming units, CFUs). As shown in the Figure 27d, the size of lesions developed was exactly the same for Hamp1Δmyel and Hamp1lox/lox littermates; moreover, the two groups presented the same number of bacteria in the skin and in peripheric organs, such as the blood and the spleen. Therefore, we concluded that the deletion of hepcidin in myeloid cells did not influence neither invasiveness nor the severity of GAS infection in this model. Surprisingly, we found that Hamp1Δker mice presented with a significantly smaller lesion compared to Hamp1lox/lox littermates (Figure 27f). We wondered if the specific deletion of hepcidin in keratinocytes could provoke a defect in healing. We found that the observed phenotype concerning the size of ulcers was specifically associated to the infection, since there was no difference between hamp1lox/lox and hamp1Δker in the healing of a sterile wound (Supplementary figure 3b). Concerning local bacteria growth and their systemic dissemination, we found that Hamp1Δker presented with a significantly higher number of CFUs at the site of infection and at the systemic level (blood and spleen) (Figure 27f).

These results underscore that epidermal hepcidin has a key role during GAS infection, limiting the local bacterial growth and, more interestingly, restricting the systemic spread of bacterial infection. Usually, AMPs have pleotropic functions and modify their expression and function according to the infectious/inflammatory status of the host. Even if we had previously demonstrated that synthetic hepcidin had a chemoattractant potential in vivo (Figure 24f, 2a) and in vitro (Figure 25b), we could not exclude that epidermal hepcidin fulfil host protection against GAS infection through other different mechanisms. Therefore, hepcidin could ensure host protection against S. Pyogenes through several possible ways.

RESULTS

Figure 27. Epidermal hepcidin protects against GAS infection. (a) Hepcidin immunohistochemistry on sections of cutaneous human biospies of group A streptococal Necrotizing Fascitiis (NF) patients. The staining is visualized using DAB IHC detection kit. The specimens are counterstained with

RESULTS

hematoxylin. The negative control consists in the omission of primary antibody. Scale bars, 5000 μm. (HCX PL Fluotar L 403/0.6 PH2). (b) Immunohistochemistry for hepcidin in the skin of WT mice challenged with GAS. The staining is visualized by alkaline phosphatase substrate kit that produce a blue reaction product. The samples are counterstained with fast red (Leica DMI-3000B microscope, Leica DFC 310FX digital camera, LAS V3.8 Core software.) (e) Generation of HepcΔkerat and (c) Hepc Δmyeloid mice. Hamp1 floxed mice were breed with LysM-CRE+ and with KRT14-CRE+ transgenic animals. (d) Hepdicin deletion in myeloid cells does not influence mice susceptibility to GAS infection. Lesion size, bacterial count in skin, blood and spleen of WT and HepcΔmyeloid mice four days after injection with GAS. At least eight mice in each group were tested in two independent experiments. (e) Hepdicin deletion in keratinocytes renders mice more susceptible to GAS infection. Lesion size, bacterial count in skin, blood and spleen of WT vs HepcΔkerat mice four days after injection with GAS. At least seven mice in each group were tested in three independent experiments. Statistical analysis was performed using unpaired Student’s t-test; *p<0.05. ** p<0.01.

Table 2. Clinical features of NF patients.

Hepcidin has no antimicrobial effect against GAS in vitro and does not influence keratinocyte antimicrobial functions. The first studies describing hepcidin demonstrated that this peptide was able to inhibit the growth of many Gram positive bacteria (such as Bacillus Megaterium, Bacillus Subtilis, Micrococcus Luteus) in vitro (Krause et al., 2000) and that hepcidin had bactericidal activity against certain pathogens, such as E. coli, S. Aureus (Park, et al. 2001). These ancestral properties of hepcidin, which are today are largely debated, prompt us to assess if epidermal

RESULTS hepcidin fulfills protection against this pathogen through these antimicrobial activities. Thus, we first tested if hepcidin had bactericidal and/or bacteriostatic effect against GAS in vitro. We incubated S. pyogenes in presence of PBS, penicillin G and two different doses of hepcidin or LL-37 and we evaluated bacterial growth in these different conditions measuring the absorbance at 600 nm. As shown in Figure 28a, in contrast to penicillin G and to LL-37, known to be an efficient AMP against GAS (Nizet et al., 2001), hepcidin did not influence GAS growth, even at doses broadly higher than minimal functional doses of efficient AMPs (Nizet et al., 2001). Thus, we concluded that hepcidin has no bacteriostatic properties against S. Pyogenes. Next, we tested if hepcidin displayed a bactericidal activity against GAS. Bacteria were exposed to LL-37 and hepcidin for one hour at 37° C. While the positive control, cathelicidin, killed 60% of bacteria in 1 hour, hepcidin do not kill bacteria (Figure 28b). These results revealed that hepcidin did not have any direct antimicrobial activity in vitro against GAS and prompted us to discard this hypothesis to explain our phenotype.

Keratinocytes produce a broad arsenal of antimicrobial molecules and immune modulators to protect host from infection (Pivarcsi et al., 2005). Even if hepcidin did not show any antimicrobial actions against GAS, we next wondered if the expression of this peptide could, in turn, influence the antimicrobial activity of keratinocyte against S. Pyogenes. Murine primary keratinocytes isolated from Hamp1lox/lox and Hamp1Δker newborn pups were cultured and differentiated in vitro. Then, these cells were infected with log phase of GAS. Wild-type and hepcidin knock out keratinocytes showed the same antibacterial properties in vitro. Indeed, both the WT and hepcidin KO keratinocyte monolayers killed bacteria, reducing the CFUs/ml compared to the control condition, where GAS was grown under identical conditions (figure 28c). Thus, we found that primary keratinocytes have a hepcidin-independent antimicrobial activity against GAS. This negative result encourages us to move towards other hypotheses.

Hamp1 Δker mice do not display epidermal iron accumulation Since hepcidin is the key iron regulatory hormone and iron is an essential nutrient for the growth of almost all bacteria, we next wondered if epidermal hepcidin could protect the host during GAS infection, by controlling the local iron availability for this pathogen. We did not find any differences in intracellular iron amount in the epidermis of Hamp1lox/lox and Hamp1Δker at the steady state, as demonstrated by the transcriptional regulation of TfR1 (Supplementary figure 3a). However, the regulation of the iron metabolism is deeply influenced by infectious

RESULTS conditions. Therefore, we investigated if the deletion of hepcidin in keratinocytes could determine an extracellular iron accumulation in the infected skin of Hamp1Δker mice. To answer to this question, we realized a Perl’s Prussian blue coloration on biopsies of infected skins (Figure 28e, lower panel). This staining is commonly used in histopathology to detect the presence of iron in biopsy specimens. As expected, spleen specimens were positive for the coloration (Figure 28e, lower panel, right), however infected skin specimens from both Hamp1lox/lox and Hamp1Δker mice did not show any positive staining (Figure 28e, lower panel). Since the sensitivity of this staining is pretty low, we tried to enhance the efficiency of this iron detection method by performing a DAB-Perl’s staining (Figure 28e, top panel). Using DAB, it is possible to enhance the histochemical detection of iron in paraffin sections by Perl’s Prussian staining. Again, the spleen showed a positive staining (brown staining, figure 28e upper panel, right) but no iron deposits could be detected in the wild-type or in the Hamp 1Δker skin biopsies. Even if there was a difference in iron deposition and subcellular localization, we did not detect it by these classical staining assays. However we observed that, GAS growth was similar in a iron restricted or in a medium supplemented with ferric chloride, Figure 28d.

RESULTS

Figure 28. Hepcidin has not antimicrobial properties against GAS and does not influence keratinocyte antimicrobial actions. (a) Inhibitory effect of hepcidin, LL-37 and penicillin G against

5 GAS in vitro. An early log phase of bacteria is diluted in THY medium, to obtain A600=0.001(~ 2x10

RESULTS

CFU). Bacterial growth was monitored with A600 measurement every 5 minutes, though overnight incubation at 37°C. Data points represent the mean +/- s.e.m. and are representative of 3 independent experiments (with n=3 per group). Statistical analysis by a one-way analysis of variance (ANOVA) followed by a Bonferroni posttest. ****p< 0.0001. (b) Bacterial killing kinetics of 32Μm of hepcidin, LL-37 and PBS versus S. pyogenes. Bacteria are exposed to the different agent for one hour at 37°C, the dilutions are plated at time points from 3 to 60 minutes to established CFUs. Data points represent the mean +/- s.e.m , with n=3 per group. (c) Wild type and hepcidin knock out primary keratinocytes are infected with an exponential phase GAS, MOI=10. Bacterial are recovered 2 and 4 h following the infection. GAS control wells consist in bacteria grown incubated with bacteria grown under identical conditions, without keratinocytes. n= 3 experiments, with n=3 per group; mean +/- s.e.m. Statistical analysis by a one-way analysis of variance (ANOVA) followed by a Bonferroni posttest. ****p< 0.0001. ns=not significative. (d) M1T1 GAS growth is assessed in iron depleted medium (control in blue) and media complemented with 250 μM and 500 μM FeCl3 by measuring the absorbance at 600 nm overnight. Data demote mean +/- s.e.m. of two independent experiment (n=3 for group). (e) . Perls’s blue staining (bottom) and DAB-Perl’s coloration (on the top) of the skin of infected Hamp1lox/lox and Hamp1Δkerat. Spleen specimens are the positive control for these colorations (Leica DMI3000B microscope, Leica DFC310FX camera [Leica LAS Core software])

Hepcidin modulates keratinocyte immune response

The AMPs accomplish their functions mainly by two distinct strategies: these small peptides can (i) directly kill bacteria or (ii) modulate the host immunity. Since hepcidin neither showed antimicrobial functions against GAS nor influenced the direct keratinocyte antimicrobial ability (Figure 28c), we next wondered if this peptide could take part and/or contribute to the immune dialogue between keratinocyte and host immune cells. For this purpose, we mimicked the GAS- mediated hepcidin upregulation, by incubating murine primary keratinocytes with different doses of hepcidin (0.36 μM and 3.6 μM) at different time points (1h and 3h) and realized a cytoplex on the cell culture supernatants. As illustrated in figure 29a, among different inflammatory cytokines tested (IL-1β, IFN γ, IL-10, CXCL1, IL-2, IL-4, IL-5, IL-6, TNFα), hepcidin specifically induced the production of CXCL1, the key neutrophil chemokine, in a dose dependent manner. Thus, hepcidin exhibits a specific immunomodulatory function on keratinocytes: triggering CXCL1 production. Curiously, mini hepcidin treatment did not show this immunomodulation (supplementary figure 4), suggesting that, once again, the 3D structure of the peptide is essential for the immunomodulatory function of this AMP. This

RESULTS unexpected result prompted us to better investigate the signaling pathway involved in this immunomodulation. Usually, chemokine production is regulated at the transcriptional level, through the activation of signaling pathways, such as NF-KB or AP-1 (De Oliveira et al., 2016). However, unexpectedly, we found that hepcidin did not induce the transcription of Cxcl1 (figure 29b). As illustrated in Figure 29c, pretreatment of keratinocytes with cycloheximide, a translation inhibitor, prevented hepcidin-mediated CXCL1 production, while pretreatment with actinomycin D, a transcription inhibitory agent, had no effect. Therefore, hepcidin promoted keratinocyte CXCL1 production at the translational level. CXCL1, also known as Keratinocyte- derived Chemokine (KC), is anyway expressed by differentiated keratinocytes in vitro, as proved by the detection of this chemokine in the control keratinocyte supernatant (Figure 29a) and by relative mRNA expression of Cxcl1 at basal level (Figure 29b). Our results demonstrate that hepcidin triggers the translation of CXCL1 in keratinocytes, rather than to determine a de novo transcription of the Cxcl1 gene. This kind of regulation looks like an attempt to rapidly increase the keratinocyte CXCL1 secretion, synergizing with the canonical NF-KB transcriptional regulation, and to boost the CXCL1 production, especially in case of infection.

RESULTS

Figure 29. Hepcidin promotes CXCL1 production at the translational level. (a) IFNγ, IL10, IL1β, IL2, IL4, IL5, IL6, TNFα and CXCL1 levels measured by ELISA in the culture supernantant of murine primary keratinocytes stimulated for 1 or 3 hours with hepcidin vs PBS. n>4; mean +/- s.e.m. (b) Expression by Q-PCR of CXCL1 in WT primary keratinocyte incubated for 1 and 3 hours with 0,36 μM, 3,6 μM hepcidin or PBS. n>4; mean +/- s.e.m. (C) CXCL1 levels measured by ELISA in the culture supernantant of murine primary keratinocytes stimulated for with 3,6 μM hepcidin in presence of solvent/ actinomycin/ cycloheximide/brefeldinA. Representative of three independent experiment (with n= 3 per group). Statistical analysis was performed using unpaired Student’s t-test; * p<0.05.

Hepcidin- mediated CXCL1 production protects against GAS infection. These results encouraged us to hypothesize that the absence of hepcidin-mediated CXCL1 translation could be responsible for the defect of Hamp1Δker mice in GAS fighting. Therefore we investigated, by immunohistochemistry, CXCL1 expression in specimens of skin adjacent to ulcer lesions 96 hours after infection. In agreement with the in vitro results, we observed a stronger CXCL1 immunoreactivity in the epidermis of Hamp1lox/lox infected animals compared to Hamp1 Δker (Figure 30a). Because of the critical role of CXCL1 in neutrophil recruitment during infectious conditions, hepcidin keratinocyte-specific knock out mice presented less polymorphonuclear (PMN) leukocytes at the site of infection compared to Hamp1lox/lox littermates, as illustrated by the PMN immunostaining in Figure 30b (brown staining). In order to verify our hypothesis, we injected wild-type and Hamp1Δkerinfected mice with recombinant CXCL1 (Figure 30c). Mice were daily monitored and sacrificed four days after infection. The injection of CXCL1 completely rescued the phenotype, since the Hamp1Δker mice injected with CXCL1 presented exactly the same number of bacteria than the Hamp1lox/lox mice. These results confirm that the susceptibility of the Hamp1Δker mice to Streptococcus pyogenes infection is due to a specific defect in CXCL1 production. All these evidences highlight the importance of epidermal hepcidin as player of keratinocyte immune activity, mediating the neutrophil recruitment during infection.

RESULTS

100

RESULTS

Figure 30. The reduced hepcidin-mediated CXCL1 production is responsible for the higher sensibility to GAS infection. (a) Immunohistochemistry using CXCL1 (b) or anti polimorphonulcear leucocytes (PMN) antibodies on skin of WT and HepcΔkerat mice challenged with GAS. (c-d) CXCL1 supply totally rescue the phenotype. WT and HepcΔkerat infected mice were SC injected with CXCL1 or its vehicle. Colony forming units amounts in the skin 96 hours after the infection. Data represent mean +/- s.e.m.; n=4. Statistical analysis was performed using unpaired Student’s t-test; * p<0.05.

Hepcidin is resistant to SpyCEP cleavage As previously discussed, GAS have developed throughout the evolution a wide plethora of mechanisms to bypass innate immune host protection. One of the key strategy that GAS have developed to divert host defense is to avoid the neutrophil recruitment to the site of infection, through the activity of the S. Pyogenes Cell Envelop Protease SpyCEP (Edwards et al., 2005; Hidalgo-Grass et al., 2006). Despite the strong 1D and 3D similarity with IL8 (Figure 24a), hepcidin does not present the site, conserved among different ELR+ chemokines of different species that is recognized by SpyCEP (Figure 31a). Since the activity of this bacterial protease is one of the principal mechanisms of GAS immune-escape, we wondered if hepcidin could be sensitive or resistant to SpyCEP cleavage. We incubated the ELR+ chemokine CXCL1 and synthetic hepcidin with purified SpyCEP overnight and examined the products of digestion by mass spectrometry analyses (Figure 31). As shown by the MALDI spectrum (figure 31b) in presence of the bacterial protease, CXCL1 was cleaved in a big and a small fragment, respectively of 6 and 2 kDa, that were not observed in the control (figure 31c) This cleavage is responsible for the loss of function of CXCL1. As showed in the figure 31f, after incubation with SpyCEP the chemokine lost its ability to trigger the neutrophil migration in vitro. The protease efficacy depended of the quantity of its target. Indeed, when SpyCEP was incubated with 36 nM of CXCL1, some neutrophils were still able to migrate, therefore the chemoattractant function was, at least in part, conserved. While, when there was 10 times less substrate, the protease was more efficient, totally preventing neutrophil migration (figure 31f). This result suggested that, by increasing the quantity of CXCL1, the enzymatic activity of the SpyCEP become less and less efficient, to reach the saturation of the reaction. Contrary to CXCL1, hepcidin was resistant to the activity of the bacterial protease, as shown by the comparison of the mass spectrometry analysis of hepcidin incubated with (figure 31d) or

101

RESULTS without (Figure 31e) SpyCEP. Therefore we speculated that, being resistant to the activity of this protease, hepcidin could boost CXCL1 secretion during infection and that this over production could contribute to the exhaustion of the activity of the bacterial protease. In conclusion, we hypothesized that epidermal hepcidin has a double function during host fight against GAS infection. Hepcidin could, at least in part, contribute to the neutrophil recruitment by its chemoattractant potential but also prompts the rapid translation and release of CXCL1 by keratinocytes. This latest function contributes to restore the CXCL1 amount at the site of infection, opposing the SPyCEP-mediated attempts to prevent neutrophil recruitment.

102

RESULTS

103

RESULTS

Figure 31. Hepcidin is resistant to SpyCEP cleaving, contrary to CXCL1. (a) Site of SpyCEP-mediated cleavage of various chemokines determined in previously studies: arrowhead. The position of the a- helix in this region is indicated by a grey shaded box. Cleavage Mass spectroscopy analysis of cleaved (b) and uncleaved (c) CXCL1 using electrospray ionization generated a series of multiply charged ions (indicated as m/z; mass-to-charge ratio) from which the average molecular mass (m) of each was deduced. (d-e) Mass spectroscopy analysis of hepcidin incubated overnight (c) with or (d) without SpyCEP using electrospray ionization generated a series of multiply charged ions (indicated as m/z; mass-to-charge ratio) from which the average molecular mass (m) of each was deduced. (f) Migration assay with the Boyden chamber method in presence of CXCL1 (3,6 nM or 36 nM) incubated overnight at 37°C with or without spyCEP. n=3; mean +/- s.e.m. Statistical analysis by a one-way analysis of variance (ANOVA) followed by a Bonferroni posttest. *p<0.1.; ****p< 0.0001.

Hepcidin subcutaneous injections have a therapeutic potential against GAS infection The critical aspect of GAS infection, that determines the high severity of infection and mortality in ~20% of the cases (Carapetis et al., 2005), is the ability of GAS to escape to the physical and to the innate immune defense of the host. Considering the new roles of hepcidin that we highlighted in this work, we hypothesized that SC administration of synthetic hepcidin could provide a protection against the bacterial dissemination. Wild-type animals were infected and daily injected with synthetic hepcidin, as shown in the scheme of the figure 32a. The number of bacteria at the local level was exactly the same between the control group and the treated group (figure 32c). However, 7 out of 9 mice treated with hepcidin did not have bacterial dissemination in peripheric organs, such as the spleen (Figure 32d). During the systemic dissemination of the infection, animals lost weight. This is an important readout of the invasiveness of infection. In agreement with the hepcidin-mediated hindrance of systemic dissemination, while mice infected and injected with the vehicle lost around 5% of their initial weight, animals treated with the peptide totally recovered their initial weight (Figure 32b).

104

RESULTS

Altogether, we found that SC injections of hepcidin did not reduce the development of the local infection burden but strongly limit the systemic dissemination and clearly improve the outcome of infection. These striking results suggest that the use of hepcidin could represent a future therapeutic treatment against group A streptococcus infection, especially in preventing the bacterial dissemination, principal cause of mortality linked to this pathogen.

Figure 32. Local injection of hepcidin prevents systemic spread of infection. (a) GAS infected mice were daily SC injected with synthetic hepcidin. (b)Weight variation in infected mice treated with 3 subcutaneous injection of hepcidin (circle) or PBS (square) during 96 hours of infection. (c) Bacterial counting in the skin of infected mice. (d) GAS dissemination in the spleen of GAS infected mice 96 hours after infection. Ten mice in each group were tested in two experiments. Statistical analysis was performed using unpaired Student’s t-test; *p<0.05.

105

RESULTS

SUPPLEMENTARY FIGURES

Supplementary figure 1. Generation of Hamp1 Δ myeloid and Hamp1 Δ keratinocyte animals. (a) Recombination of the floxed hamp1 allele by genomic PCR in the epidermis of the HepcΔkerat and WT mice (left) and in bone marrow derived macrophages of the HepcΔmyeloid and WT mice (right). (b) Quantification of the relative expression of hamp1 by qPCR on mouse genomic DNA from the epidermis, bone marrow and liver of the HepcΔkerat and WT mice. (c) quantification of the deletion rate for hamp1in the genomic DNA isolated from bone marrow derivate macrophages and neutrophils of the HepcΔmyeloid and WT mice. n>3, statistical analysis was performed using unpaired Student ‘s t-test; ***p<0.001.; *p<0.05.

106

RESULTS

107

RESULTS

Supplementary figure 2. Blood and iron parameters are unchanged between Hamp1 Δ myeloid and WT littermate and between Hamp1 Δ keratinocyte and littermate. Plasma iron, ferritin, transferrin, red blood cells, haematocrit, haemoglobin TCMH and CCMH in hamp1 lox/lox and hamp1Δker (a) and in hamp1 lox/lox and hamp1Δmyeloid animals (b). All the values are shown as mean +/- s.e.m. n>4. Statistical analysis was performed using unpaired student’s t test. ns: not significant.

Supplementary figure 3. Regulation of intracellular iron is the same between hamp1 lox/lox and hamp1keratinocytes. Sterile wound healing is the same between the hamp1 lox/lox and hamp1 Δkerat

control

50 IL10 200 CXCL1 mini hepc 3,6 mM 50 IFN γ 40 40 150

l 30 30 m

/ 100 g

p 20 20 50 10 10

0 0 0 0 1 2 3 0 1 2 3 0 1 2 3

50 IL 2 50 IL 4 50 IL 1 β 40 40 40

l 30 30 30

m /

g 20 p 20 20

10 10 10

0 0 0 0 1 2 3 0 1 2 3 0 1 2 3

200 IL 6 200 TNF a 50 IL 5 40 150 150

l 30 m

/ 100 100 g

p 20 50 50 10

0 0 0 0 1 2 3 0 1 2 3 0 1 2 3 time (hours) time (hours) time (hours)

108

RESULTS

mice. (a) Transcriptional response to intracellular iron level in epidermis of Hamp1lox/lox and Hamp1Δker animals in basal condition. The data represent mean +/- s.e.m., n=4. Statistical analysis was performed using unpaired student’s t test. ns: not significant. (b)Wound healing assay in WT and keratinocyte- specific hepcidin knock out animals. Analysis of wound size change in both the genotype. The data present mean+/- s.e.m., representative of two independent experience, with n= 3 per group.

Supplementary Figure 4. Cytoplex on murine primary keratinocytes stimulated with hepcidin. IFN γ, IL-10, TNF-α, IL-2, IL-4, IL-5, IL-1β and IL-6 levels measured in the supernatant of WT murine primary keratinocytes incubated for 1 or 3 hours with 3,6 μM minihepcidin or PBS, with a cytoplex proinflammatory panel kit.

109

MATERIALS AND METHODS

Mice

Hamp1lox/lox mice (Zumerle et al., 2014) were bred with (i) K14-Cre (Jonkers et al., 2001) or with (ii) LySM-Cre (Clausen B.E et al., 2001) transgenic mice, in which the Cre recombinase is respectively under the control of the murine K14 and LysM promotor. Studies were performed in a mixed C57BL6/genetic background, using 7-10 week old male littermates. The animal studies described here were reviewed and approved (Agreement n° CEEA34.CP.003.13) by the “Président du Comité d'Ethique pour l'Expérimentation Animale Paris Descartes” and are in accordance with the principles and guidelines established by the European Convention for the Protection of Laboratory Animals (Council of Europe, ETS 123, 1991).

1-D and 3-D Sequence alignment method 1-D and 3-D sequence alignment has been performed by Synsight. Clustal W/X version 2.0 (Larkin et al., 2007) has been used to carry out the 1-D sequence alignment. Clustal programs incorporate a position-specific scoring scheme and a weighting scheme for down weighting over-represented sequence groups. All pairs of sequences to be aligned are compared by pair- wise alignment and a score matrix of similarity has been produced, indicating the divergence or similarities of each pair. The “quality score” represents the main method of alignment and has been performed with default parameters (Thompson et al., 1997). TM-align algorithm (Zhang and Skolnick, 2005) has been used to perform 3-D structural alignment between protein pairs. This method combines the TM-score rotation matrix (Zhang and Skolnick, 2004) and dynamic programming (Needleman and Wunsch, 1970). TM-align only employs the backbone C-alpha coordinates of the given protein structures. TM-align process combines two methods that are initial alignments and heuristic iterative algorithm. Initial alignments are exploited using dynamic programming and then heuristic iteration using TM-score rotation.

Bioluminescence imaging of myeloperoxydase activity In Vivo follow-up of myeloperoxydase activity was performed using bioluminescence imaging of Luminol as described in (Gross et al., 2009). Briefly, mice were anesthetized using isoflurane in air (4% for induction and 2% during imaging) and were injected IP with 200 mg/Kg Luminol

111

MATERIALS AND METHODS

(alfa Aesar). Bioluminescence activity was then measured over 10 minutes using a dedicated in vivo optical imager (Photon Imager RT, Biospacelab, France). Acquisitions were repeated for each side of the animal, for better photon collection. Quantification was performed using M3Vision Software (BiospaceLab, France), measuring photon flux in ROIs and expressed as photon per seconds (ph/s).

Primary mouse bone marrow-derived neutrophils Bone marrows from WT mice were flushed from femurs and tibias with PBS using a 25 G needle. Tissues were homogenized through a 18 G needle and the suspension was passed through 70 μm and 40 μm cell strainers. Cells were centrifuged at 400 g for 10 min and resuspended in cold MACS buffer (PBS, BSA 0.5%, EDTA 2mM, pH 7.4). After counting, cells were incubated on ice with FcBlock (Miltenyi Biotec) for 10 min, anti-Ly6C for 15 min and anti-biotin antibody microbeads for 15 min, following the manufacturer’s instructions. Cells were washed with MACS buffer and centrifuged at 400 g for 10 min. Neutrophils were immuno-magnetically separated following filtration of the cell suspension through an LS column (Miltenyi Biotec) according to manufacturer’s instructions.

Migration assay Neutrophil migration was assessed using transwell (3 μm; BD-Falcon) coated with 10 μg/ml fibronectin (Sigma-Aldrich). The cells (5 × 105/well) were diluted in 700 μL DMEM+ 10% SVF and migration towards hepcidin (Peptide International) in the bottom well was allowed for 2 hours at 37°C. The resulting migrated cells recovered from the bottom well were counted using hemocytometer. In some experiments, 200 ng/ml PTX (Sigma-Aldrich), 250 nM SB 225002 (Sigma-Aldrich) or 10 µg/ml neutralizing anti-CXCR2 antibody (R&D Systems) was added 30 minutes before the addition of hepcidin.

Isolation and treatment of murine primary keratinocytes.

Keratinocytes were isolated from the skin of 2-day-old WT or Hamp1 Δkerat mice, based on modifications from a previously described protocol (Lichti et al., 2008). In brief, dermis were easily separated from epidermis after floating newborn mice skin on freshly thawed cold trypsin 0,25% (thermo fisher 15050065) overnight at 4°C with the dermis side down. Cells were

112

MATERIALS AND METHODS isolated in stop trypsin medium (DMEM, high glucose, no glutamine, no calcium (thermo fisher 21068028) containing 20% heat inactivated and Chelex (Bio rad 1421253) treated- FBS, 1% GlutaMAX Supplement (thermo fisher 35050061), 1% Sodium Pyruvate (thermo fisher 11360070), 100U/ml Penicillin, 100 μg/ml Streptomycin (thermo fisher), 2,5 μg/ml Amphotericin B (Gibco)). Keratinocytes were obtained by epidermis mechanical dissociation through shaking at 350 rpm for 1 hour at 37 °C. The resulting cell suspension was filtered through 100 m filter, centrifuged at 250 g for 10 min at 4°C, plated on fibronectin and bovine collagen I coated tissue culture plate and incubated at 37°C in a humidified 5% CO2 atmosphere. The cells were cultured in 0,07 mM calcium medium (Low Ca2+) containing EGF (10 ng/ml) and cholera toxin (10-10mol/L). Once confluence reached, cells were switched in 1,2 mM calcium medium (High Ca2+) in order to assure their differentiation. After three days in High Ca2+ cells were differentiated and suitable for different experiments.

Reverse transcription and real-time quantitative PCR RNA extraction, reverse transcription and quantitative PCR have been performed as previously described (Zumerle et al., 2014). All samples were normalized to the threshold cycle value for cyclophilin-A. The primers used are presented in the table below.

Mouse model of GAS infection A well-established model of necrotizing soft tissue infection was adapted for this study 9,10. In brief, 100 μl of a mid logarithmic growth phase (approximately 107 c.f.u.) of M1 GAS was mixed with an equal volume of sterile Cytodex beads (Sigma-Aldrich) and injected subcutaneously into the flank of Hamp1lox/lox and Hamp1Δker mice male littermates, previously shaved followed by treatment with hair removal cream. Weight loss and lesion size of mice were daily estimated. After 96 hours, mice were sacrificed, skin necrotic ulcers, spleen, and blood (via retro orbital bleeding) were recovered and homogenized in 1:2 mg/μl of sterile PBS. Serial dilutions of tissue homogenate were stretched on THA plates, in order to enumerate

113

MATERIALS AND METHODS c.f.u. For the rescue experiment, WT and Hamp1Δker littermates were subcutaneously infected with 107 c.f.u. of GAS with PBS or 1 μg/mouse of rCXCL1 (R&D system, 1395-KC-025/CF). Mice were intralesionally injected again at 24, 48 and 72 h with PBS or 1 μg of rCXCL1. For the therapeutic treatment, C57BL6 mice were infected as described above. One day after infection, animals were subcutaneously injected with PBS (control) or 1μg of hepcidin (peptide international). Mice were treated again at 48 and 72 hours with 500 ng of hepc-25. After 96 hours mice were sacrificed, spleens were taken, homogenized for c.f.u. enumeration.

Human patients

Informed consent to the protocol was obtained for all subjects and was approved by the Institutional Review Board and the regional ethics committee Paris IV (IRB 2016/40NICB), n° IRB 00003835. The collection of personal data was approved by the Commission Nationale de l'Informatique et des Libertés (CNIL, https://www.cnil.fr).

Antimicrobial assays

As previously described, (Nizet et al., 2001) GAS M1T1 5448 were grown in Todd Hewitt Broth (THB, Sigma) plus 0,3% yeast extract to early log phase (absorbance at 600 nm, A600=0,2). Bacteria were then diluted in THB toA600=0,001 (~2x105 colony forming units (c.f.u.) ml-1). This bacterial suspension was kept growing overnight at 37°C in presence of PBS, LL-37, hepcidin (peptide international) or penicillin G (sigma P7794). For antimicrobial killing kinetics, exponential phase GAS (A600=0,4) was exposed to PBS or LL-37 or hepcidin and incubated at 37°C. At different time point serial dilutions of these bacterial suspensions were plated on THA plate for counting c.f.u.

Keratinocyte killing assay

Keratinocytes from WT and Hamp1 Δkerat newborn mice were isolated and cultured as above described. Their antimicrobial function was tested with log-phase GAS culture.

114

MATERIALS AND METHODS

CO2 for different time points. At each time point, well contents were harvested and spread on THA plates for enumerating surviving c.f.u.

Immunostaining Tissues were fixed in 10% formaldehyde and embedded in paraffin. Sections were cut at a thickness of 3 μm. Following dewaxing and rehydratation, sections were subjected to antigen retrieval in a pressure cooker for 15 min at 95°C with citrate buffer and permeabilization with 0,5% Triton in phosphate buffered saline for 30 minutes at room temperature, then incubated overnight at 4°C with rabbit polyclonal anti-CXCL1 primary antibody (1/100, NBP1-51188 Novus). A solution of protease K diluted at 1/30 in phosphate buffered saline for 5 min was necessary for unmasking sections before incubating with rat monoclonal anti-PMN primary antibody (1/200, sc-71674 Santa Cruz). Sections were incubated with HRP labeled anti-rabbit antibody (A6154, Sigma) or anti-rat antibody (ab97057, Abcam) for 1 hour before washing and substrate development according to the manufacturer’s instructions (Impact Novared SK48-05, Vector). Sections were counterstained with Mayer haemalun.

Quantification of cytokines

Keratinocytes, previously starved of FBS for 1 hour, were treated with 1 or 10 g/ml hepcidin or minihepcidin (Peptide International) and incubated at different time points. In some experiments, inhibitors (Brefaldin A (sigma B7651), cycloheximide (sigma C7698), actinomycin D (sigma A1410)) were added 30 minutes before hepcidin treatment during FBS starvations. The cells were lysed with Tri reagent (sigma) while supernatant cytokines were measured with the V-PLEX Proinflammatory Panel1 kit (Meso Scale Discovery), according to the manufacturer’s instruction.

Mass spectrometry and analysis MALDI

Peptide dilutions series were mixed 1:1 on a MALDI target plate with 5mg/ml of alpha-cyano- 4-hydroxycinnamic acid matrix (Laser Biolabs) dissolved in of 70%ACN 0.1% TFA 30%milliQ H2O, with 3µM Glu-fibrinopeptide B as internal standard and let to dry on a 96- wells OptiTOF MALDI plate target (ABSciex). Mass spectra were acquired and processed at the 3P5 proteomics facility of the Université Paris Descartes with a 4800 MALDI-TOF/TOF

115

MATERIALS AND METHODS analyser (ABSciex) with a Nd:YAG pulsed LASER (355nm) at 200Hz shooting frequency in positive reflectron mode, fixed fluency, low mass gate and delayed extraction. 500 to 1000 spectra were summed with the 4000series explorer software (ABSciex) by steps of 50 spectra in the range of 1300 to 9000 M/Z (CXCL1) or 500 to 4000 M/Z (HepC). Monoisotopic values were labeled from isotopes clusters, when accessible, with a minimal s/n ratio of 1000. Internal calibration was performed on Glu-fibrinopeptide B at M/Z 1570.68.

Statistical analysis Statistical analyses were performed using GraphPad Prism 5.0. The significance of experimental differences was evaluated by a two-tailed unpaired Student's t-test for comparisons between two groups. A regular one-way analysis of variance (ANOVA) followed by a Bonferroni posttest was performed to compare more than two groups and a two-way analysis of variance (ANOVA) followed by a Bonferroni posttest to examine the influence of two different independent variables. A Log-Rank (Mantel-Cox) test was performed to compare survival curves.

116

DISCUSSION

Hepcidin, the key iron regulatory hormone, was originally described as a functional antimicrobial peptide produced by the liver (Krause et al., 2000; Park et al., 2001a). In the last 20 years, different groups have focused their attention on its hormonal regulatory activity. Even if in various fish species, hepcidin and hepcidin-like peptides conserve antimicrobial functions, in higher vertebrates these properties are today largely challenged. However, it is broadly recognized and accepted that hepcidin structure is highly similar to that of cationic antimicrobial peptides, since this hormone is an amphipathic, positively charged peptide with four disulfide bridges and beta sheets. Although AMPs have not only antimicrobial but also immunomodulatory functions, to date the hepcidin-mediated modulation of host immune cells has been poorly investigated.

During my PhD I investigated the immunomodulatory role of hepcidin and its activity as a functional AMP in the skin, which is one of the principal sources of these innate immune molecules.

New chemoattractant role of hepcidin.

We observed a structural affinity between hepcidin and IL8, a pleiotropic interleukin essential both for the neutrophil recruitment and activity, prompting respectively the extravasation and the release of proteases and ROS for microbial killing in peripheric tissues. Thus, we wondered if this structural affinity entails similar properties for hepcidin. We chose to first test if hepcidin can trigger the MPO activity of neutrophils in vivo, because this readout reflects both the number and the activation state of these phagocytes. We demonstrated that hepcidin promotes the MPO activity in vivo just like CXCL1 and that this activity was due to the active recruitment of neutrophils. Similarly to hepcidin, other antimicrobial peptides, such as LL-37, α-defensins and β-defensins, have been shown to possess chemoattractant properties We further demonstrated that minihepcidin, the shorter form of this peptide, does not exhibit chemoattractant properties, at least in vitro. This truncated hepcidin presents the 9 N-terminal amino acids of hepcidin-25, thus conserves the FPN binding function. This evidence suggests that the chemoattractant function of hepcidin requires its complete 3D structure and evokes the hypothesis that this new function could be independent of the canonical target of hepcidin, ferroportin. Indeed, the fact that minihepcidin does not provoke the migration could imply that (i) ferroportin is not the receptor required for the chemoattraction or that (ii) the minihepcidin-

118

DISCUSSION and hepcidin-binding to FPN induce two different downstream signaling pathways in neutrophils. In particular, it would be interesting to investigate the ability of hepcidin to recruit neutrophils mutated for FPN-hepcidin interaction. Such mouse line with a hepcidin-resistant FPN (FPN C326S) mutation has been generated and prevents hepcidin from binding to FPN. Phagocytes isolated from this mouse line will have a functional iron exporter and will not present abnormalities in the intracellular iron accumulation, that could affect the phagocyte activities (de Sousa, 1989). Thus, these mutants could help us to investigate exclusively the importance of FPN-hepcidin binding during the chemoattractant process. Moreover we observed, using two different tools, that hepcidin chemoattraction requires CXCR2. In the literature, there are also other examples of AMP-chemokine receptor binding. Similarly to what we have observed for hepcidin, β-defensin2 has been shown to be able to bind a chemokine receptor. It binds to CCR6, thereby acting as a chemoattractant for immature dendritic cells and memory T cells (D. Yang et al., 1999). Anyway, it will be important in the future to perform a protein-protein interaction assay, e.g. FRET assay, to determine whether hepcidin physically interacts with CXCR2. The leukocyte chemotaxis hinges on the interaction between a chemokine and its receptor. CXCR2 is also expressed in other immune cells like mast cells, monocytes and macrophages (Stadtmann & Zarbock, 2012). Therefore in the future, it will be interesting to evaluate if hepcidin can accomplish, via this receptor, the chemoattraction of these immune cells. Each of these immune cells has a specific function and lead to a distinct immune response during the infection/inflammation. Thus if this hypothesis is confirmed, hepcidin could mediate a wide range of not redundant immune processes, distinct from neutrophil-mediated bacterial clearance, by recruiting other immune cells all along an infectious/inflammatory condition. Our data suggest that hepcidin and CXCL1 share the same receptor and demonstrate that these two molecules mediate the same function, triggering the neutrophil recruitment. These evidences insinuate a possible competition between these two molecules. We could postulate that the binding of CXCL1 and hepcidin to CXCR2 may be differentially regulated at the spatiotemporal level, avoiding the competition for this receptor. Both human and murine neutrophils have two distinct chemokine receptors: CXCR1 and CXCR2. In human, IL8 binds to both receptors. CXCL1, the murine functional homologue of IL8, binds specifically and strongly to CXCR2. For this reason, we first chose to investigate the role of this receptor. Contrary to CXCR2, understanding the biological role of CXCR1 was

119

DISCUSSION hampered by difficulties in identifying and characterizing mouse CXCR1. About 10 years ago the murine homologue for CXCR1 has been identified (Moepps, Nuesseler, Braun, & Gierschik, 2006), and was described as the CXCL5 receptor. A recent study described a key role of this receptor during Candida Albicans infection in activating rather than recruiting neutrophils (Swamydas et al., 2016). It will be informative to investigate if and which effects hepcidin could have on neutrophil functions thorough CXCR1.

Using murine models of attenuated (hfe -/-) or ablated (hamp1-/-) hepcidin expression and systemic iron overload, two precedent studies have already described a defect in MPO activity and neutrophil recruitment to the peripheric tissue in response to respectively LPS-mediated systemic inflammation and urinary infection (Benesova et al., 2012; Dounia Houamel et al., 2016). Since the two teams have used models of iron overload and the link between iron and inflammation is well established, the authors explain their phenotype by an impaired inflammatory response. Therefore, our work is the first demonstration of the chemoattractant properties of hepcidin.

Hepcidin is expressed in the skin in pathophysiological conditions

For a long time, the researchers have focused their attention on the expression, regulation and function of hepcidin in the liver. During the last decade, several groups described low expression of hepcidin in other organs; then it was demonstrated that extra hepatic hepcidin does not influence systemic iron homeostasis (Zumerle et al., 2014), evoking the idea that extra hepatic hepcidin could have local roles. Here, we demonstrated that hepcidin is expressed in the skin, in particular by keratinocytes, in physiological conditions and during Streptococcus pyogenes infection both in human and in mouse models. To our knowledge, this study is the first description of an epidermal hepcidin expression. We observed that this peptide presents an expression pattern expected for a molecule involved in the host defense. AMPs can be constitutively expressed in the skin, such as human β-defensin 1 (Ganz, 2003), or can be expressed at low level and strongly up regulated in response to infection, as cathelicidin and β-defensin 2 (Nizet et al., 2001). Similarly to the cathelicidin cramp, hepcidin is strongly induced in response to GAS subcutaneous infection. In a previous study, Peyssonnaux and colleagues have demonstrated that GAS trigger hepcidin expression in

120

DISCUSSION macrophages via TLR4 (Peyssonnaux et al., 2006). We tried to characterize the signaling pathway required for the up regulation of hepcidin in keratinocytes in response to GAS. We infected murine primary keratinocytes isolated from wild type mice in vitro. Although keratinocytes have a broad plethora of PRRs to quickly recognize pathogens, GAS did not increase hepcidin expression compared to the not-infected control in vitro, neither at mRNA nor at the protein level (data not shown). This negative result, in contradiction with what we observed in vivo, supports the idea that other resident or recruited cells signal to keratinocytes to promote the production of hepcidin during GAS infection in vivo. We first hypothesized that IL6 could trigger the keratinocyte expression of hepcidin. Indeed, IL6 is the main cytokine involved in the hepcidin expression in response to inflammation (Nemeth et al., 2004) and Streptococcus Pyogenes infection causes the expression of this cytokine in the tissue (Carey et al., 2016). However, upon bacterial or viral infection, IL6 is synthesized by a variety of cell types, including keratinocytes (Turksen et al., 1992; Yoshizaki et al., 1990). For this last reason we discarded the “IL-6 hypothesis”. To date, we think that the best candidate for this function would be IL-22 for different reasons. First of all, this cytokine triggers hepcidin expression in response to inflammatory stimuli (Andrew E Armitage et al., 2011; Smith et al., 2013). IL22 has a key role in host defense against microbes, in particular it was demonstrated that Streptococcus Pyogenes strongly induces this cytokine in infected skin (Carey et al., 2016). Moreover, IL-22 is essential for host defense at mucosal surfaces against extracellular pathogens, promoting the expression of antimicrobial molecules and inducing neutrophil recruitment (Valeri & Raffatellu, 2016). For all these reasons, to test this hypothesis we will perform the same experiment of GAS infection in vitro in presence or in absence of IL 22. The previous work of Peyssonnaux and colleagues focused on the macrophage-derived hepcidin (Peyssonnaux et al., 2006). In agreement with this study, we observed hepcidin expression in the recruited myeloid cells but also in the keratinocytes. Taking advantage of myeloid- and keratinocyte- specific hamp1 knock out murine models, our study allows us to elucidate the importance of these two contributions, at least in our protocol of infection. While myeloid- derived hepcidin does not influence the infection outcome, epidermal hepcidin is essential to protect against the local and, more interestingly, the systemic dissemination of Group A streptococcus infection. The absence of phenotype in myeloid-specific hepcidin knock out mice can be explained by a timing issue: our model consists in four days of infection and the macrophages are principally recruited starting from day 3 of infection. In order to evaluate the

121

DISCUSSION putative role of myeloid-derived hepcidin, it will be useful to realize some infection experiments with a longer kinetic in WT and Hamp1Δmyeloid mice. During a subcutaneous infection, keratinocytes are among the first cells interacting with bacteria. We demonstrated that the epidermal-derived hepcidin has a key role in fighting against the infection.

Keratinocyte- derived hepcidin has an immunomodulatory role during infection

Although early studies have described a mild antimicrobial effect of hepcidin against various bacteria in vitro, we did not find neither a bacteriostatic nor a bacteriolytic effect of hepcidin- 25 against GAS in vitro. Hepcidin antimicrobial function is mainly influenced by variation of pH (Maisetta et al., 2013). Of course, our data do not exclude that at the lesion site in vivo, hypoxia and variation of pH can favor the hepcidin antimicrobial function. However, we have tested hepcidin antimicrobial properties in vitro at different pH, without finding any difference. Even if the activity of each AMP can be deeply influenced by various parameters, LL-37 in the same conditions is able to prevent the growth and to kill this pathogen. These evidences suggest that hepcidin does not likely have an antimicrobial effect against GAS during mouse infection. As previously illustrated, there are also shorter forms of hepcidin, hepcidin-20 and hepcidin- 22. We don’t know which form of hepcidin is produced in the skin during GAS infection and it has been demonstrated that these shorter forms have stronger bactericidal properties compared to hep-25. Therefore, it will be informative to test the bactericidal activity of these shorter forms against Streptococcus Pyogenes. Then, we considered the hypothesis that the virulence of GAS infection and the bacterial growth could be influenced by a different local iron availability in the skin of wild type and hamp1Δker animals. We did not observe any extracellular iron accumulation in the skin specimens by Perl’s Prussian and DAB Perl’s staining. Despite the low sensitivity of these assays, these colorations allow to visualize in which cells iron is stored and its subcellular distribution. Since we were interested in this information we chose these methods rather than the iron dosage in the whole infected tissues. Anyway, it is important to remember that GAS cannot use lactoferrin or transferrin as source of iron (Dahesh et al., 2012). Streptococcus pyogenes has compensated for this inability, by developing a broad plethora of cell lysis strategies, e.g. producing streptolysin

122

DISCUSSION

O, streptolysin S, protease, superantigens which allow to recover iron from intracellular iron compounds, including heme, heme-containing proteins, myoglobin and ferritin. Accordingly, GAS growth in iron restricted medium was not different from the bacterial growth in a medium supplemented with ferric chloride. The ability of acquiring iron from the heme-containing proteins, represents an important evolutive advantage of some bacteria, such as Streptococcus Pyogenes or Staphylococcus Aureus, that makes these bacteria independent from the free extracellular iron in contrast to siderophilic bacteria (Stefanova et al., 2017b).

To accomplish the host immune protection, keratinocytes have mainly two mechanisms: (i) the AMP-mediated microbicidal activity and (ii) the expression of chemokines and cytokines, to establish an “immune-dialogue” with closest immune cells. We demonstrated that the bactericidal activity of murine primary keratinocytes against GAS is independent from hepcidin. This negative result prompted us to test if keratinocyte-derived hepcidin can fulfill its protective role by modulating the production of immune-mediators, such as cyto- and chemokines, in these cells.

We observed that while hepcidin-mediated MPO activity is comparable to that triggered by CXCL1, the efficacity of hepcidin-mediated chemiotaxis in vitro is significantly lower compared to that of CXCL1. This incoherence may be explained by the fact that hepcidin, like others antimicrobial peptide (Zhang et al., 2016), can induce the production of chemokines, such as CXCL1/2 in resident or recruited cells. In support of this hypothesis we observed that exogenous hepcidin trigger the secretion of CXCL1 in murine primary keratinocytes . This “double-function” property was already described for other AMPs, such as defensins or LL-37( G Scott et al., 2002). These peptides are able to induce directly chemotaxis and to promote the chemokine expression in some epithelial cells, according to the various inflammatory/infectious conditions (Yang et al., 2001). To discriminate the contribution of these two functions of hepcidin, it would be interesting to realize the same bioluminescence experiment in mice injected with hepcidin and neutralizing antibodies for CXCL1 and CXCL2.

To our knowledge, this study is the first description of a direct immunomodulatory role of hepcidin. Like for the chemotaxis, we found that minihepcidin is not able to give rise to this immunomodulation. Accordingly, we observed, by quantification of TfR1 transcriptional expression, that cells stimulated with hepcidin do not present an intracellular iron overload (data

123

DISCUSSION not shown). These data suggest that hepcidin immunomodulation of keratinocytes is independent of the canonical iron regulatory function of this hormone. However, preliminary data showed that pretreatment of murine primary epidermal cells with a chemical inhibitor of FPN (Ross et al., 2017) partially inhibits the hepcidin-mediated CXCL1 production. The discrepancy between the inability of minihepcidin to produce CXCL1 and the putative FPN- dependent hepcidin-mediated CXCL1 secretion may be explained by the fact that the hepcidin- and minihepcidin-binding to FPN can trigger two distinct signaling pathways in keratinocytes. Our current proposition is that hepcidin-mediated CXCL1 induction depends on the 3D structure of the mature 25aa-peptide and may involve an interaction with FPN. Like for the chemotaxis, it will be necessary to confirm this hypothesis and use murine primary keratinocytes isolated from mice mutated for the FPN-hepcidin interaction (Altamura et al., 2014), to discriminate definitively if this new role requires this binding. An interesting finding of our work is the observation that hepcidin regulates keratinocyte CXCL1 expression at the translational but not at the transcriptional level. The expression of this chemokine presents transcriptional, post-transcriptional and post- translational mechanisms of regulation. Various studies have described a post-translational regulation of CXCL1 expression. This chemokine is synthetized and stored in endothelial cells (Hol et al., 2010; Øynebråten et al., 2004). In response to various “secretagogue” molecules this chemoattractant is rapidly released in order to efficiently allow the neutrophil extravasation. Contrary to this fast regulation observed in the endothelial cells, in keratinocytes, the main regulation of CXCL1/IL8 expression, in response to various infectious stimuli, occurs at the transcriptional level, though the TLR/NF-KB signalling pathway (Pivarcsi et al., 2005). The post-transcriptional mechanisms that modify mRNA stability and/or translation provide a more flexible control of gene expression and are essential in coordinating the initiation of inflammation (Anderson, 2010). CXCL1 presents a putative HuR binding sites (AU Rich Elements, ARE) in the 3’ untranslated region (Novotny et al., 2005). HuR is a RNA Binding Protein (RBP) that binds to various inflammatory transcripts, prompting their stability or their translation (Panganiban et al., 2014). Although this kind of post transcriptional regulation has been never described for CXCL1 in the keratinocytes, this modulation was described in various cell types, including epithelial cells (Fan et al., 2011; Novotny et al., 2005). Therefore, it will be interesting to test this hypothesis, e.g. measuring the decay rates of CXCL1 mRNA over time in presence of hepcidin. We have already measured, by qPCR, the CXCL1 mRNA 3 hours

124

DISCUSSION after hepcidin stimulation, but to investigate the stability of a certain transcript, it is necessary to do an actinomycin D pulse chase assay. Functional related genes are usually subjected to a similar regulation at the post transcriptional level in response to a certain stimulus. Therefore, in the future, we will investigate whether hepcidin can promote the expression of other chemokines and immune mediators, or if this immunomodulation is exclusively restricted to CXCL1. During infection, the interaction of GAS with keratinocytes, via TLR, triggers the transcriptional expression of a large plethora of proinflammatory molecules, including CXCL1. At the same time, hepcidin quickly promotes the CXCL1 translation and secretion. We speculate that during GAS infection there is an integration between transcriptional and post transcriptional regulation of CXCL1, strongly boosting the keratinocyte-mediated CXCL1 production. We will test this hypothesis by performing some experiments on keratinocytes infected by GAS in vitro with or without hepcidin.

We demonstrated that in our subcutaneous model of infection, epidermal-derived hepcidin ensures host protection through its new immune modulatory role, since hamp1Δker mice present less CXCL1 and a reduced neutrophil infiltration to the site of infection compared to wild type littermates. Moreover, the injection of CXCL1 completely rescues the phenotype. Therefore, despite the hepcidin chemoattractant potential that we have observed, during GAS infection in vivo, CXCL1 and hepcidin are not redundant. It is evident that in this context the hepcidin-mediated CXCL1 expression, and the consecutive neutrophil recruitment represents the main contribution to the resolution of the infection. We know that during this infection, the neutrophils recruitment has a pivotal role in the host defense (Döhrmann et al., 2016). Considering that the intrinsic chemoattractant efficiency of hepcidin is sensibly lower compared to that of CXCL1 (figure 25) it was expected that the CXCL1 chemotaxis exceeds the direct hepcidin chemiotaxis, thus the main role of this peptide during this infection is to trigger the CXCL1 expression. It is important to highlight that even if indirect, the function of hepcidin during GAS infection is crucial. S. pyogenes produces a protease (SpyCEP) that cleaves ELR+ chemokines, including CXCL1, to avoid neutrophil recruitment at the site of infection (Edwards et al., 2005; Hidalgo-Grass et al., 2006). As expected the activity of this enzyme is saturable. Moreover, our data demonstrate that hepcidin is resistant to this cleavage. First of all, hepcidin has not the sequence canonically recognized by this bacterial protease. Moreover, hepcidin has

125

DISCUSSION four disulfide bridges while CXCL1 has only two. This makes hepcidin-25 aa more stabilized and highly compact, thus less accessible to any proteases. Since we demonstrated that hepcidin is resistant to the activity of SpyCEP, we speculated that hepcidin-mediated CXCL1 expression is one of the key process ensuring the maintaining of a suitable quantity of this chemokine, counteracting the GAS attempt to circumvent host defense. The interaction between host and pathogen determines a co-evolution throughout the time. Indeed, when a host develop new strategies to stop a pathogen attack, among the pathogens occurs a “natural” selection that allow them to acquire a new more aggressive strategy of attack in order to fight the enhanced host defenses. In response, the host will develop new defense to cope with the new aggressive infection, and the cycle continues. Among the host genes, antimicrobial peptides, cytokines and chemokines present the higher rate of genetic variance appearance, because of this continuously host pathogen co-evolution. Interestingly, CXCL1 is phylogenetically ancient and is expressed in Dictostelium Discoideum, an amoeba belonging to the phylum Amoebozoa (Tettamanti et al., 2006). While hepcidin-like peptides appear more recently during evolution, at the teleost fish level. Therefore, hepcidin is evolutionary more recent than CXCL1 (Figure 33). Considering the theory of the co-evolution driven by host- pathogen interaction, we speculate that GAS still did not evolve yet a virulence factor that can cope with the hepcidin resistance to SpyCEP.

Figure 33. Host-pathogen co evolution. Eukaryotic phylogenetic tree showing CXCL1 (in pink) end hepcidin (in green) apparition during evolution. During the co-evolution host-pathogen, GAS have

126

DISCUSSION

developed a counterattack against CXCL1, SpyCEP. Since hepcidin appears later, these bacteria have not yet developed a strategy to bypass hepcidin-mediated host protection.

Although precedent studies have already investigated the expression and the role of this peptide in various epithelia during infectious conditions (Dounia Houamel et al., 2016; Schwarz et al., 2012), our study is the first description of the role of epidermal hepcidin during cutaneous infection. More interestingly, we demonstrated that epidermal hepcidin has a key role in the resolution of this infection, modulating the immune activity of keratinocytes. The ablation of hepcidin in keratinocytes determines a reduced CXCL1 production and a restricted number of neutrophils recruited to the site of infection causing the local bacterial burst and the systemic spread of infection (Figure 34).

Figure 34. Model for the role of epidermal-derived hepcidin during GAS infection. In response to [1] GAS infection keratinocytes ensure host protection, upregulating [2] CXCL1 (green triangle) and [2] hepcidin. Keratinocyte-derived hepcidin contributes, at least in part, to the neutrophil recruitment and promotes epidermal CXCL1 production. This last function counterbalances to the SpyCEP cleaving activity (here depicted as red scissors), ensuring the maintenance of a suitable amount of this chemokine, thus an efficient neutrophil recruitment [3] and host protection [4] (schema on the left). Whereas in absence of hepcidin (schema on the right) CXCL1-derived from keratinocytes is extensively cleaved, reducing the [3] neutrophil recruitment and the [4] host defense against GAS and allowing infection development [5].

Hepcidin as a therapeutic agent

Our data and the growing problem of resistance and overuse of conventional antibiotics have stimulated our curiosity to test if hepcidin could be used as a therapeutic agent. The advantage

127

DISCUSSION of AMPs compared to classical antibiotics is that these peptides, besides their antimicrobial properties, can also give rise to a specific immunomodulation. Our study has demonstrated a possible therapeutic use of hepcidin to cure GAS invasive infections. We observed that hepcidin subcutaneous administration avoids the systemic bacterial dissemination, limiting the principal cause of mortality associated to this pathogen. Curiously we did not observe a local restriction of infection in response to hepcidin injection. Thus adding hepcidin in excess did not further reduce the number of bacteria at the infection site. This may be explained considering that neutrophil recruitment in peripheric tissue reaches a plateau from a certain concentration of chemoattractant. An interesting finding of our work is that hepcidin treatment almost totally prevents the bacterial dissemination, in contrast to CXCL1 treatment (data not shown). This finding suggests that daily administration of exogenous hepcidin can trigger another mechanism of protection, different from the promotion of the neutrophil recruitment (Figure 35).

Figure 35. Hepcidin treatment prevents bacterial systemic dissemination. During infection GAS acquire various mutations that allow the bacterial invasion and the systemic dissemination of this pathogen (on the left). Daily hepcidin SC injections prevent this phenomenon.

During invasive infection, GAS undergo several mutations that render bacteria more aggressive. The two principal virulence agents that are upregulated during the transition from local to systemic infection and that characterize most of the invasive Streptococcus pyogenes are the bacterial DNAse Sda1 and the bacterial streptokinase Ska1. The Sda1 ensures NETs digestion, allowing bacterial escape from the extracellular traps. Ska1 is an enzyme secreted by GAS responsible for the conversion of plasminogen into plasmin, which once activated, catalyzes the fibrogenolysis, degrading blood clots and ECM components, and activates

128

DISCUSSION metalloproteinases (Cole et al., 2011). During a bacterial infection there are four different steps: bacteria must (i) adhere, (ii) colonize, (iii) circumvent the host immune and physical defenses and (iv) finally move into the bloodstream to systemically disseminate. Preventing the coagulation and the formation of blood clots is, therefore, essential for allowing the bacterial dissemination. Since we have observed that hepcidin limits systemic dissemination, we will in the future investigate the possibility that hepcidin might avoid one of these two mechanisms. It has already been described in the literature that LL-37, via its ability to bind and damage the membrane of neutrophils, could facilitate the NETs coming out (Neumann et al., 2014). Considering this similarity, it would be interesting to test if hepcidin can promote the formation of neutrophil extracellular traps. To test this hypothesis, we will test if synthetic hepcidin triggers the production of NETs, at least in vitro. However, proteins-derived granules contained in the NETs, kill bacteria. These structures are defined as an extracellular killing strategy of neutrophils. Since we didn’t observe a difference in the local bacterial load, we don’t favor this hypothesis (Figure 36). On the other hand, in the literature there are some recent studies describing a direct correlation between hepcidin and fibrinogen (Ellingsen et al., 2018; Pechlaner et al., 2016). This is a glycoprotein circulating in the blood, which has an essential role in the formation of fibrin-based blood clot. Fibrinogen is essential in the process of healing and represents one of the principal targets of activated plasmin. Thus we wonder if the hepcidin triggered-inhibition of GAS dissemination can depend on a higher expression of fibrinogen, attempting to cope with bacterial plasmin formation (Figure 36).

Figure 36. Hepcidin avoids GAS systemic spread. Schematic representation of the two main hypothesis for explaining the hepcidin-mediated bacterial dissemination inhibition.

129

DISCUSSION

Conclusion and highlights Despite the initial interest in the antimicrobial function of hepcidin, this study highlights the immunomodulatory potential of this peptide. We have studied its expression and its role as an antimicrobial peptide in one of the key sources of immune regulators, the skin. We demonstrated (i) direct chemoattractant and (ii) indirect immunomodulatory functions of hepcidin. We chose the GAS subcutaneous infection model that triggers a massive infiltration of neutrophils to the site of infection, to better investigate the chemoattractant potential of hepcidin. Unexpectedly, in this context of infection, the hepcidin indirect function is the main agent for the resolution of infection. Moreover we described a therapeutic application of hepcidin for limiting GAS systemic spread. Nevertheless a better characterization of all the molecular players involved remains to be determined. Altogether, these results offer new avenues for therapeutic approaches in a wide array of inflammatory and infectious diseases.

130

BIBLIOGRAPHY

BIBLIOGRAPHY Abtin, A., Jain, R., Mitchell, A. J., Roediger, B., Brzoska, A. J., Tikoo, S., … Weninger, W. (2010). Perivascular macrophages mediate neutrophil recruitment during bacterial skin infection, 14(4), 384–399. https://doi.org/10.1080/10810730902873927.Testing Aisen, P., Enns, C., & Wessling-Resnick, M. (2001). Chemistry and biology of eukaryotic iron metabolism. International Journal of Biochemistry and Cell Biology, 33(10), 940– 959. https://doi.org/10.1016/S1357-2725(01)00063-2 Albanesi, C., Scarponi, C., Giustizieri, M. L., & Girolomoni, G. (2005). Keratinocytes in inflammatory skin diseases. Current Drug Targets. Inflammation and Allergy, 4(3), 329– 334. https://doi.org/10.2174/1568010054022033 Altamura, S., Kessler, R., Gröne, H.-J., Gretz, N., Hentze, M. W., Galy, B., & Muckenthaler, M. U. (2014). Resistance of ferroportin to hepcidin binding causes exocrine pancreatic failure and fatal iron overload. Cell Metabolism, 20(2), 359–367. https://doi.org/10.1016/j.cmet.2014.07.007 Anderson, P. (2010). Post-transcriptional regulons coordinate the initiation and resolution of inflammation. Nature Reviews Immunology, 10(1), 24–35. https://doi.org/10.1038/nri2685 Andriopoulos Jr, B., Corradini, E., Xia, Y., Faasse, S. A., Chen, S., Grgurevic, L., … Babitt, J. L. (2009). BMP6 is a key endogenous regulator of hepcidin expression and iron metabolism. Nature Genetics, 41(4), 482–486. https://doi.org/10.1038/ng.335 Armitage, A. E., Eddowes, L. A., Gileadi, U., Cole, S., Spottiswoode, N., Selvakumar, T. A., … Drakesmith, H. (2011). Hepcidin regulation by innate immune and infectious stimuli. Blood, 118(15), 4129–4139. https://doi.org/10.1182/blood-2011-04-351957 Armitage, A. E., Stacey, A. R., Giannoulatou, E., Marshall, E., Sturges, P., Chatha, K., … Borrow, P. (2014). Distinct patterns of hepcidin and iron regulation during HIV-1, HBV, and HCV infections. Proceedings of the National Academy of Sciences, 111(33), 12187– 12192. https://doi.org/10.1073/pnas.1402351111 Aschemeyer, S., Qiao, B., Stefanova, D., Valore, E. V., Sek, A. C., Alex Ruwe, T., … Nemeth, E. (2018). Structure-function analysis of ferroportin defines the binding site and an alternative mechanism of action of hepcidin. Blood, 131(8), 899–910. https://doi.org/10.1182/blood-2017-05-786590 Aziz, R. K., Edwards, R. A., Taylor, W. W., Low, D. E., McGeer, A., & Kotb, M. (2005). Mosaic prophages with horizontally acquired genes account for the emergence and diversification of the globally disseminated M1t1 clone of Streptococcus pyogenes. Journal of Bacteriology, 187(10), 3311–3318. https://doi.org/10.1128/JB.187.10.3311- 3318.2005 Aziz, R. K., & Kotb, M. (2008). Rise and persistence of global M1T1 clone of Streptococcus pyogenes. Emerging Infectious Diseases, 14(10), 1511–1517. https://doi.org/10.3201/eid1410.071660 Babitt, J. L., Andrews, N. C., Lin, H. Y., Huang, F. W., Xia, Y., & Sidis, Y. (2007).

131

BIBLIOGRAPHY

Modulation of bone morphogenetic protein signaling in vivo regulates systemic iron balance. J Clin Invest, 117(7), 1933–1939. https://doi.org/10.1172/JCI31342 Babitt, J. L., Huang, F. W., Wrighting, D. M., Xia, Y., Sidis, Y., Samad, T. A., … Lin, H. Y. (2006). Bone morphogenetic protein signaling by hemojuvelin regulates hepcidin expression. Nature Genetics, 38(5), 531–539. https://doi.org/10.1038/ng1777 Baggiolini, M., Walz, A., & Kunkel, S. L. (1989). Neutrophil-activating peptide-1/interleukin 8, a novel cytokine that activates neutrophils. J Clin Invest, 34(1), 126–133. Baker, H. M., & Baker, E. N. (2012). A structural perspective on lactoferrin function. Biochemistry and Cell Biology = Biochimie et Biologie Cellulaire, 90(3), 320–328. https://doi.org/10.1139/o11-071 Bao, B., Peatman, E., Li, P., He, C., & Liu, Z. (2005). Catfish hepcidin gene is expressed in a wide range of tissues and exhibits tissue-specific upregulation after bacterial infection. Developmental and Comparative Immunology, 29(11), 939–950. https://doi.org/10.1016/j.dci.2005.03.006 Bastiat-sempe, B., Love, J. F., Lomayesva, N., & Wessels, M. R. (2014). Streptolysin O and NAD-Glycohydrolase Prevent Phagolysosome Acidification and Promote Group A Streptococcus Survival in Macrophages. American Society for Microbiology, 5(5), 1–11. https://doi.org/10.1128/mBio.01690-14.Editor Belkaid, Y., & Segre, J. A. (2014). Dialogue between skin microbiota and immunity. Science. https://doi.org/10.1126/science.1254229 Benesova, K., Spasić, M. V., Schaefer, S. M., Stolte, J., Baehr-Ivacevic, T., Waldow, K., … Muckenthaler, M. U. (2012). Hfe deficiency impairs pulmonary neutrophil recruitment in response to inflammation. PLoS ONE, 7(6), 39363. https://doi.org/10.1371/journal.pone.0039363 Bergmann, T. K., Vinding, K., & Hey, H. (2001). Multiple hepatic abscesses due to Yersinia enterocolitica infection secondary to primary haemochromatosis. Scandinavian Journal of Gastroenterology, 36(8), 891–895. https://doi.org/10.1080/003655201750313450 Besson-Fournier, C., Latour, C., Kautz, L., Bertrand, J., Ganz, T., Roth, M. P., & Coppin, H. (2012). Induction of activin B by inflammatory stimuli up-regulates expression of the iron-regulatory peptide hepcidin through Smad1/5/8 signaling. Blood, 120(2), 431–439. https://doi.org/10.1182/blood-2012-02-411470 Borregaard, N. (2010). Neutrophils, from Marrow to Microbes. Immunity, 33(5), 657–670. https://doi.org/10.1016/j.immuni.2010.11.011 Braff, M. H., Zaiou, M., Fierer, J., Nizet, V., & Gallo, R. L. (2005a). Keratinocyte production of cathelicidin provides direct activity against bacterial skin pathogens. Infection and Immunity, 73(10), 6771–6781. https://doi.org/10.1128/IAI.73.10.6771-6781.2005 Braff, M. H., Zaiou, M., Fierer, J., Nizet, V., & Gallo, R. L. (2005b). Keratinocyte production of cathelicidin provides direct activity against bacterial skin pathogens. Infection and Immunity. https://doi.org/10.1128/IAI.73.10.6771-6781.2005 Braff, M. H., Zaiou, M., Fierer, J., Nizet, V., & Gallo, R. L. (2005c). Keratinocyte production

132

BIBLIOGRAPHY

of cathelicidin provides direct activity against bacterial skin pathogens. Infection and Immunity, 73(10), 6771–6781. https://doi.org/10.1128/IAI.73.10.6771-6781.2005 Bratosin, D., Mazurier, J., Tissier, J. P., Estaquier, J., Huart, J. J., Ameisen, J. C., … Montreuil, J. (1998). Cellular and molecular mechanisms of senescent erythrocyte phagocytosis by macrophages. A review. Biochimie, 80(2), 173–195. https://doi.org/10.1016/S0300-9084(98)80024-2 Bridle, K. R., Frazer, D. M., Wilkins, S. J., Dixon, J. L., Purdie, D. M., Crawford, D. H. G., … Ramm, G. A. (2003). Disrupted hepcidin regulation in HFE-associated haemochromatosis and the liver as a regulator of body iron homoeostasis. Lancet, 361(9358), 669–673. https://doi.org/10.1016/S0140-6736(03)12602-5 Brinkmann, V., Reichard, U., Goosmann, C., Fauler, B., Uhlemann, Y., Weiss, D. S., … Zychlinsky, A. (2004). Neutrophil Extracellular Traps Kill Bacteria. Science (New York, N.Y.), 303(5663), 1532–1535. https://doi.org/10.1126/science.1092385 Buchanan, J. T., Simpson, A. J., Aziz, R. K., Liu, G. Y., Kristian, S. A., Kotb, M., … Nizet, V. (2006). DNase expression allows the pathogen group A Streptococcus to escape killing in neutrophil extracellular traps. Current Biology, 16(4), 396–400. https://doi.org/10.1016/j.cub.2005.12.039 Cacalano, G., Lee, J., Kikly, K., Ryan, A. M., Pitts-meek, S., Hultgren, B., … Moore, M. W. (1994). Neutrophil and B Cell Expansion in Mice That Lack the Murine IL-8 Receptor Homolog. Science, 265(5172), 682–684. Cai, S., Batra, S., Lira, S. A., Kolls, J. K., & Jeyaseelan, S. (2007). CXCL1 Regulates Pulmonary Host Defense to Klebsiella Infection via CXCL2 , CXCL5, NF-κB and MAPKs, 86(3), 573–579. https://doi.org/10.4049/jimmunol.0903843.CXCL1 Camaschella C, Roetto A, Calì A, De Gobbi M, Garozzo G, Carella M, Majorano N, Totaro A, G. P. (2000). The gene TFR2 is mutated in a new type of haemochromatosis mapping to 7q22. Nature Genetics, 25(1), 14–15. Campbell, J., Fledrick, J., Zlstnik, A., & Siani, N. A. (1998). Chemokines and the arrest of lymphocytes rolling under flow conditions. Science, 9–12. Campostrini, N., Traglia, M., Martinelli, N., Corbella, M., Cocca, M., Manna, D., … Girelli, D. (2012). Serum levels of the hepcidin-20 isoform in a large general population: The Val Borbera study. Journal of Proteomics, 76, 28–35. https://doi.org/10.1016/j.jprot.2012.08.006 Canali, S., Core, A. B., Zumbrennen-Bullough, K. B., Merkulova, M., Wang, C. Y., Schneyer, A. L., … Babitt, J. L. (2016). Activin B induces noncanonical SMAD1/5/8 signaling via BMP type I receptors in hepatocytes: Evidence for a role in hepcidin induction by inflammation in male mice. Endocrinology, 157(3), 1146–1162. https://doi.org/10.1210/en.2015-1747 Canonne-Hergaux, F., Calafat, J., Richer, E., Cellier, M., Grinstein, S., Borregaard, N., & Gros, P. (2002). Expression and subcellular localization of NRAMP1 in human neutrophil granules. Blood, 100(1), 268–275. https://doi.org/10.1182/blood.V100.1.268

133

BIBLIOGRAPHY

Canonne-Hergaux, F., Donovan, A., Delaby, C., Wang, H., & Gros, P. (2006). Comparative studies of duodenal and macrophage ferroportin proteins. American Journal of Physiology. Gastrointestinal and Liver Physiology, 290(1), G156-63. https://doi.org/10.1152/ajpgi.00227.2005 Canonne-Hergaux, F., Gruenheid, S., Govoni, G., & Gros, P. (1999). The Nramp1 protein and its role in resistance to infection and macrophage function. Proceedings of the Association of American Physicians. https://doi.org/10.1046/j.1525-1381.1999.99236.x Canonne-Hergaux, F., Gruenheid, S., Ponka, P., & Gros, P. (1999). Cellular and subcellular localization of the Nramp2 iron transporter in the intestinal brush border and regulation by dietary iron. Blood, 93(12), 4406–4417. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10361139 Carapetis, J. R., Steer, A. C., Mulholland, E. K., & Weber, M. (2005). Review The global burden of group A streptococcal diseases. Lancet, 5(November), 685–694. Carey, A. J., Weinberg, J. B., Dawid, S. R., Venturini, C., Lam, A. K., Nizet, V., … Ulett, G. C. (2016). Interleukin-17A contributes to the control of streptococcus pyogenes colonization and inflammation of the female genital tract. Scientific Reports, 6(February), 1–12. https://doi.org/10.1038/srep26836 CARTWRIGHT, G. E., & LAURETSEN, M. A. (1946). The anemia associated with chronic infection. Science (New York, N.Y.), 103(2664), 72. Retrieved from http://science.sciencemag.org.gate2.inist.fr/content/sci/103/2664/72.full.pdf Casanovas, G., Mleczko-Sanecka, K., Altamura, S., Hentze, M. W., & Muckenthaler, M. U. (2009). Bone morphogenetic protein (BMP)-responsive elements located in the proximal and distal hepcidin promoter are critical for its response to HJV/BMP/SMAD. Journal of Molecular Medicine, 87(5), 471–480. https://doi.org/10.1007/s00109-009-0447-2 Cassat, J. E., & Skaar, E. P. (2013). Iron in infection and immunity. Cell Host and Microbe. https://doi.org/10.1016/j.chom.2013.04.010 Chatellier, S., Ihendyane, N., Kansal, R. G., Khambaty, F., Basma, H., Norrby-Teglund, A., … Kotb, M. (2000). Genetic relatedness and superantigen expression in group A streptococcus serotype M1 isolates from patients with severe and nonsevere invasive diseases. Infection and Immunity, 68(6), 3523–3534. https://doi.org/10.1128/IAI.68.6.3523-3534.2000 Chehoud, C., Rafail, S., Tyldsley, A. S., Seykora, J. T., Lambris, J. D., & Grice, E. A. (2013). Complement modulates the cutaneous microbiome and inflammatory milieu. Proceedings of the National Academy of Sciences, 110(37), 15061–15066. https://doi.org/10.1073/pnas.1307855110 Chen, Q. X., Song, S. W., Chen, Q. H., Zeng, C. L., Zheng, X., Wang, J. L., & Fang, X. M. (2014). Silencing airway epithelial cell-derived hepcidin exacerbates sepsis induced acute lung injury. Critical Care (London, England), 18(4), 470. https://doi.org/10.1186/s13054-014-0470-8 Cherukuri, S., Potla, R., Sarkar, J., Nurko, S., Harris, Z. L., & Fox, P. L. (2005). Unexpected role of ceruloplasmin in intestinal iron absorption. Cell Metabolism, 2(5), 309–319.

134

BIBLIOGRAPHY

https://doi.org/10.1016/j.cmet.2005.10.003 Chlosta, S., Fishman, D. S., Harrington, L., Johnson, E. E., Knutson, M. D., Wessling- Resnick, M., & Cherayil, B. J. (2006). The iron efflux protein ferroportin regulates the intracellular growth of Salmonella enterica. Infection and Immunity, 74(5), 3065–3067. https://doi.org/10.1128/IAI.74.5.3065-3067.2006 Clark, R. J., Tan, C. C., Preza, G. C., Nemeth, E., Ganz, T., & Craik, D. J. (2011). Understanding the structure/activity relationships of the iron regulatory peptide hepcidin. Chemistry and Biology, 18(3), 336–343. https://doi.org/10.1016/j.chembiol.2010.12.009 Clausen B.E, C.Burkhardt, Reith, W., Renkawitz, R., & Förster I. (2001). Conditional gene targeting in macrophage and granulocytes using LysMcre mice. Transgenic Research, 96(October), 317–330. https://doi.org/10.1023/A Clayton, E., Doupé, D. P., Klein, A. M., Winton, D. J., Simons, B. D., & Jones, P. H. (2007). A single type of progenitor cell maintains normal epidermis. Nature, 446(7132), 185– 189. https://doi.org/10.1038/nature05574 Cleary, P. P., LaPenta, D., Vessela, R., Lam, H., & Cue, D. (1998). A globally disseminated M1 subclone of group A streptococci differs from other subclones by 70 kilobases of prophage DNA and capacity for high- frequency intracellular invasion. Infection and Immunity, 66(11), 5592–5597. Cleary, P. P., Schlievert, P. M., Handley, J. P., Kim, M. H., Hauser, A. R., Kaplan, E. L., & Wlazlo, A. (1992). Clonal basis for resurgence of serious Streptococcus pyogenes disease in the 1980s. The Lancet, 339(8792), 518–521. https://doi.org/10.1016/0140- 6736(92)90339-5 Cogen, A. L., Yamasaki, K., Muto, J., Sanchez, K. M., Alexander, L. C., Tanios, J., … Gallo, R. L. (2010). Staphylococcus epidermidis antimicrobial δ-toxin (phenol-soluble modulin- γ) cooperates with host antimicrobial peptides to kill group A Streptococcus. PLoS ONE, 5(1), 1–7. https://doi.org/10.1371/journal.pone.0008557 Colditz, I., Zwahlen, R., Dewald, B., & Baggiolini, M. (1989). In vivo inflammatory activity of neutrophil-activating factor, a novel chemotactic peptide derived from human monocytes. The American Journal of Pathology, 134(4), 755–760. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1879797&tool=pmcentrez&r endertype=abstract Cole, J. N., Barnett, T. C., Nizet, V., & Walker, M. J. (2011). Molecular insight into invasive group A streptococcal disease. Nature Reviews Microbiology, 9(10), 724–736. https://doi.org/10.1038/nrmicro2648 Cole, J. N., McArthur, J. D., McKay, F. C., Sanderson-Smith, M. L., Cork, A. J., Ranson, M., … Walker, M. J. (2006). Trigger for group A streptococcal M1T1 invasive disease. The FASEB Journal, 20(10), 1745–1747. https://doi.org/10.1096/fj.06-5804fje Corradini, E., Meynard, D., Wu, Q., Chen, S., Ventura, P., Pietrangelo, A., & Babitt, J. L. (2011a). Serum and liver iron differently regulate the bone morphogenetic protein 6 (BMP6)-SMAD signaling pathway in mice. Hepatology, 6(1), 247–253. https://doi.org/10.1111/j.1743-6109.2008.01122.x.Endothelial

135

BIBLIOGRAPHY

Corradini, E., Meynard, D., Wu, Q., Chen, S., Ventura, P., Pietrangelo, A., & Babitt, J. L. (2011b). Serum and liver iron differently regulate the bone morphogenetic protein 6 (BMP6)-SMAD signaling pathway in mice. Hepatology, 54(1), 273–284. https://doi.org/10.1002/hep.24359 Cywes-Bentley, C., Hakansson, A., Christianson, J., & Wessels, M. R. (2005). Extracellular group A Streptococcus induces keratinocyte apoptosis by dysregulating calcium signalling. Cellular Microbiology, 7(7), 945–955. https://doi.org/10.1111/j.1462- 5822.2005.00525.x D. Yang, O. Chertov, S. N. Bykovskaia, Q. Chen, M. J. Buffo, J. Shogan, … J. J. Oppenheim. (1999). β-Defensins: Linking Innate and Adaptive Immunity Through Dendritic and T Cell CCR6. Science. Retrieved from http://science.sciencemag.org.gate2.inist.fr/content/sci/286/5439/525.full.pdf Dahesh, S., Nizet, V., & Cole, J. N. (2012). Study of streptococcal hemoprotein receptor (Shr) in iron acquisition and virulence of M1T1 group A streptococcus. Virulence, 3(7). https://doi.org/10.4161/viru.21933 Dale, J. B., Washburn, R. G., Marques, M. B., & Wessels, M. R. (1996). Hyaluronate capsule and surface M protein in resistance to opsonization of group A streptococci. Infection and Immunity, 64(5), 1495–1501. Datta, V., Myskowski, S. M., Kwinn, L. A., Chiem, D. N., Varki, N., Kansal, R. G., … Nizet, V. (2005). Mutational analysis of the group A streptococcal operon encoding streptolysin S and its virulence role in invasive infection. Molecular Microbiology, 56(3), 681–695. https://doi.org/10.1111/j.1365-2958.2005.04583.x De Oliveira, S., Rosowski, E. E., & Huttenlocher, A. (2016). Neutrophil migration in infection and wound repair: Going forward in reverse. Nature Reviews Immunology, 16(6), 378–391. https://doi.org/10.1038/nri.2016.49 de Sousa, M. (1989). Immune cell functions in iron overload. Clinical and Experimental Immunology, 75(1), 1–6. De Yang, Chen, Q., Schmidt, A. P., Anderson, G. M., Wang, J. M., Wooters, J., … Chertov, O. (2000). Ll-37, the Neutrophil Granule–And Epithelial Cell–Derived Cathelicidin, Utilizes Formyl Peptide Receptor–Like 1 (Fprl1) as a Receptor to Chemoattract Human Peripheral Blood Neutrophils, Monocytes, and T Cells. The Journal of Experimental Medicine, 192(7), 1069–1074. https://doi.org/10.1084/jem.192.7.1069 Deschemin, J. C., & Vaulont, S. (2013). Role of Hepcidin in the Setting of Hypoferremia during Acute Inflammation. PLoS ONE, 8(4). https://doi.org/10.1371/journal.pone.0061050 DiVietro, J. A., Smith, M. J., Smith, B. R. E., Petruzzelli, L., Larson, R. S., & Lawrence, M. B. (2001). Immobilized IL-8 Triggers Progressive Activation of Neutrophils Rolling In Vitro on P-Selectin and Intercellular Adhesion Molecule-1. The Journal of Immunology. https://doi.org/10.4049/jimmunol.167.7.4017 Do¨hrmann, S., Anik, S., Olson, J., Anderson, E. L., Etesami, N., No, H., … Okumura, C. Y. M. (2014). Role for streptococcal collagen-like protein 1 in M1T1 group A

136

BIBLIOGRAPHY

Streptococcus resistance to neutrophil extracellular traps. Infection and Immunity, 82(10), 4011–4020. https://doi.org/10.1128/IAI.01921-14 Döhrmann, S., Cole, J. N., & Nizet, V. (2016). Conquering Neutrophils. PLoS Pathogens. https://doi.org/10.1371/journal.ppat.1005682 Donovan, A., Brownlie, A., Zhou, Y., Shepard, J., Pratt, S. J., Moynihan, J., … Leonard I, Z. (2000). Positional cloning of zebrafish ferroportin1 identifies a conserved vertebrate iron exporter. Nature, 403(6771), 776–781. https://doi.org/10.1038/35001596 Dorschner, R. A., Pestonjamasp, V. K., Tamakuwala, S., Ohtake, T., Rudisill, J., Nizet, V., … Gallo, R. L. (2001a). Cutaneous injury induces the release of cathelicidin anti-microbial peptides active against group A streptococcus. Journal of Investigative Dermatology. https://doi.org/10.1046/j.1523-1747.2001.01340.x Dorschner, R. A., Pestonjamasp, V. K., Tamakuwala, S., Ohtake, T., Rudisill, J., Nizet, V., … Gallo, R. L. (2001b). Cutaneous injury induces the release of cathelicidin anti-microbial peptides active against group A streptococcus. Journal of Investigative Dermatology, 117(1), 91–97. https://doi.org/10.1046/j.1523-1747.2001.01340.x Douglas, S. E., Gallant, J. W., Liebscher, R. S., Dacanay, A., & Tsoi, S. C. M. (2003). Identification and expression analysis of hepcidin-like antimicrobial peptides in bony fish. Developmental and Comparative Immunology, 27(6–7), 589–601. https://doi.org/10.1016/S0145-305X(03)00036-3 Drakesmith, H., & Prentice, A. M. (2012). Hepcidin and the iron-infection axis. Science. https://doi.org/10.1126/science.1224577 Drakesmith, H., Schimanski, L. M., Ormerod, E., Merryweather-Clarke, A. T., Viprakasit, V., Edwards, J. P., … Townsend, A. R. M. (2005). Resistance to hepcidin is conferred by hemochromatosis-associated mutations of ferroportin. Blood, 106(3), 1092–1097. https://doi.org/10.1182/blood-2005-02-0561 Du X, She E, Gelbart T, Truksa J, Lee P, Xia Y, Khovananth K, Mudd S, Mann N, Moresco EM, Beutler E, B. B. (2008). The Serine Protease TMPRSS6 Is Required to Sense Iron Deficiency. Science, 320. Eash, K. J., Greenbaum, A. M., Gopalan, P. K., & Link, D. C. (2010). CXCR2 and CXCR4 antagonistically regulate neutrophil trafficking from murine bone marrow. Journal of Clinical Investigation, 120(7), 2423–2431. https://doi.org/10.1172/JCI41649 Eash, K. J., Means, J. M., White, D. W., & Link, D. C. (2009). CXCR4 is a key regulator of neutrophil release from the bone marrow under basal and stress granulopoiesis conditions. Blood, 113(19), 4711–4719. https://doi.org/10.1182/blood-2008-09-177287 Edwards, R. J., Taylor, G. W., Ferguson, M., Murray, S., Rendell, N., Wrigley, A., … Sriskandan, S. (2005). Specific C‐Terminal Cleavage and Inactivation of Interleukin‐8 by Invasive Disease Isolates of Streptococcus pyogenes. The Journal of Infectious Diseases, 192(5), 783–790. https://doi.org/10.1086/432485 Einchner, R., Sun, T.-T., & Aebi, U. (1986). The role of keratin subfamilies and keratin pairs in the formation of human epidermal intermediated filamants. The Journ, 102(May),

137

BIBLIOGRAPHY

1767–1777. El-Benna, J., Hurtado-Nedelec, M., Marzaioli, V., Marie, J. C., Gougerot-Pocidalo, M. A., & Dang, P. M. C. (2016). Priming of the neutrophil respiratory burst: role in host defense and inflammation. Immunological Reviews, 273(1), 180–193. https://doi.org/10.1111/imr.12447 Ellingsen, T. S., Lappegård, J., Ueland, T., Aukrust, P., Brækkan, S. K., & Hansen, J.-B. (2018). Plasma hepcidin is associated with future risk of venous thromboembolism. Blood Advances, 2(11), 1191–1197. https://doi.org/10.1182/bloodadvances.2018018465 Engleberg, N. C., Heath, A., Miller, A., Rivera, C., & DiRita, V. J. (2001). Spontaneous Mutations in the CsrRS Two‐Component Regulatory System of Streptococcus pyogenes Result in Enhanced Virulence in a Murine Model of Skin and Soft Tissue Infection. The Journal of Infectious Diseases, 183(7), 1043–1054. https://doi.org/10.1086/319291 Eyerich, S., Eyerich, K., Traidl-Hoffmann, C., & Biedermann, T. (2018). Cutaneous Barriers and Skin Immunity: Differentiating A Connected Network. Trends in Immunology, 39(4), 315–327. https://doi.org/10.1016/j.it.2018.02.004 Fan, J., Ishmael, F. T., Fang, X., Myers, A., Cheadle, C., Huang, S.-K., … Stellato, C. (2011). Chemokine Transcripts as Targets of the RNA-Binding Protein HuR in Human Airway Epithelium HHS Public Access. J Immunol. J Immunol. February, 15(1864). https://doi.org/10.4049/jimmunol.0903634 Faurschou, M., & Borregaard, N. (2003). Neutrophil granules and secretory vesicles in inflammation. Microbes and Infection, 5(14), 1317–1327. https://doi.org/10.1016/j.micinf.2003.09.008 Feder, J. N., Gnirke, A., Thomas, W., Tsuchihashi, Z., Ruddy, D. A., Basava, A., … Wolff, R. K. (1996). A novel MHC class I-like gene is mutated in patients with hereditary haemochromatosis. Nature Genetics. https://doi.org/10.1038/ng0896-399 Fernandes, A., Preza, G. C., Phung, Y., Domenico, I. De, Kaplan, J., Ganz, T., … Ganz, T. (2009). The molecular basis of hepcidin-resistant hereditary hemochromatosis The molecular basis of hepcidin-resistant hereditary hemochromatosis. Blood, 114(2), 437– 443. https://doi.org/10.1182/blood-2008-03-146134 Fleckman, P., Dale, B. A., & Holbrook, K. A. (1985). Profilaggrin, a high-molecular-weight precursor of filaggrin in human epidermis and cultured keratinocytes. Journal of Investigative Dermatology, 85(6), 507–512. https://doi.org/10.1111/1523- 1747.ep12277306 Fleming, M. D., Romano, M. A., Su, M. A., Garrick, L. M., Garrick, M. D., & Andrews, N. C. (1998). Nramp2 is mutated in the anemic Belgrade (b) rat: Evidence of a role for Nramp2 in endosomal iron transport. Proceedings of the Association of American Physicians, 95, 1148–1153. https://doi.org/10.1073/pnas.95.3.1148 Fleming, R. E., Ahmann, J. R., Migas, M. C., Waheed, A., Koeffler, H. P., Kawabata, H., … Sly, W. S. (2002). Targeted mutagenesis of the murine transferrin receptor-2 gene produces hemochromatosis. Proceedings of the National Academy of Sciences of the United States of America, 99(16), 10653–10658. https://doi.org/10.1073/pnas.162360699

138

BIBLIOGRAPHY

Flo, T. H. T., Smith, K. D. K., Sato, S., Rodriguez, D. J. D. J., Holmes, M. A., Strong, R. K., … Aderem, A. (2004). Lipocalin 2 mediates an innate immune response to bacterial infection by sequestrating iron. Nature, 432(7019), 917–921. https://doi.org/10.1038/nature03021.Published Folgueras, A. R., De Lara, F. M., Pendás, A. M., Garabaya, C., Rodríguez, F., Astudillo, A., … Velasco, G. (2008). Membrane-bound serine protease matriptase-2 (Tmprss6) is an essential regulator of iron homeostasis. Blood, 112(6), 2539–2545. https://doi.org/10.1182/blood-2008-04-149773 Frazier, M. D., Mamo, L. B., Ghio, A. J., & Turi, J. L. (2011). Hepcidin expression in human airway epithelial cells is regulated by interferon-g. Respiratory Research, 12, 100. https://doi.org/10.1186/1465-9921-12-100 Freedberg, I. M., Eisen, A. Z., Wolff, K., Austen, K. F., & Goldsmith, L. a. (2003). Dermatology In General Medicine. General Medicine. Fuchs, E. (1990a). Epidermal Differentiation : The Bare Essentials. The Journal of Cell Biology, 111(6), 2807–2814. https://doi.org/10.1083/jcb.111.6.2807 Fuchs, E. (1990b). The Bare Essentials Epidermal Differentiation : The Journal of Cell Biology, 111(6), 2807–2814. https://doi.org/10.1083/jcb.111.6.2807 Fuchs, T. A., Abed, U., Goosmann, C., Hurwitz, R., Schulze, I., Wahn, V., … Zychlinsky, A. (2007). Novel cell death program leads to neutrophil extracellular traps. Journal of Cell Biology, 176(2), 231–241. https://doi.org/10.1083/jcb.200606027 Gallo, R. L., & Hooper, L. V. (2012). Epithelial antimicrobial defence of the skin and intestine. Nature Reviews Immunology. https://doi.org/10.1038/nri3228 Gallo, R. L., Kim, K. J., Kozak, C. A., Merluzzi, L., Gennaro, R., Bernfield, M., & Zanetti, M. (1997). Identification of CRAMP , a Mouse Identification of CRAMP , a Cathelin- related Antimicrobial Peptide Expressed in the Embryonic and Adult Mouse *, 272(20), 13088–13093. https://doi.org/10.1074/jbc.272.20.13088 Ganz, T. (2003). Defensins: Antimicrobial peptides of innate immunity. Nature Reviews Immunology. https://doi.org/10.1038/nri1180 Ganz, T. (2012). Macrophages and systemic iron homeostasis. Journal of Innate Immunity, 4(5–6), 446–453. https://doi.org/10.1159/000336423 Ganz, T. (2013). Systemic Iron Homeostasis. Physiological Reviews, 93(4), 1721–1741. https://doi.org/10.1152/physrev.00008.2013 Ganz, T., & Nemeth, E. (2015). Iron homeostasis in host defence and inflammation. Nature Reviews Immunology. https://doi.org/10.1038/nri3863 Ge, R., Sun, X., & He, Q. (2009). Iron acquisition by Streptococcus species: An updated review. Frontiers of Biology in China, 4(4), 392–401. https://doi.org/10.1007/s11515- 009-0035-4 George Papanikolaou, Mark E Samuels, Erwin H Ludwig, Marcia L E MacDonald2, P. L. F., Marie-Pierre Dubé, Lisa Andres, Julie MacFarlane, Nikos Sakellaropoulos, M. P.,

139

BIBLIOGRAPHY

Elizabeta Nemeth, Jay Thompson, Jenni K Risler, Catherine Zaborowska, R. B., Christopher C Radomski, Terry D Pape, Owen Davidas, John Christakis5, Pierre Brissot, G. L., & Tomas Ganz, Michael R Hayden, & Y. P. G. (2004). Mutations in HFE2 cause iron overload in chromosome 1q–linked juvenile hemochromatosis. Nature Genetics, 36(1). https://doi.org/10.1038/ng1274 Gerhard, G. S., Levin, K. A., Goldstein, J. P., Wojnar, M. M., Chorney, M. J., & Belchis, D. A. (2001). Vibrio vulnificus septicemia in a patient with the hemochromatosis HFE C282Y mutation. Archives of Pathology and Laboratory Medicine, 125(8), 1107–1109. https://doi.org/10.1043/0003-9985(2001)125<1107:VVSIAP>2.0.CO;2 Girelli, D., Pasino, M., Goodnough, J. B., Nemeth, E., Guido, M., Castagna, A., … Fattovich, G. (2009). Reduced serum hepcidin levels in patients with chronic hepatitis C. Journal of Hepatology, 51(5), 845–852. https://doi.org/10.1016/j.jhep.2009.06.027 Gkouvatsos, K., Papanikolaou, G., & Pantopoulos, K. (2012). Regulation of iron transport and the role of transferrin. Biochimica et Biophysica Acta - General Subjects. https://doi.org/10.1016/j.bbagen.2011.10.013 Grice, E. A., Kong, H. H., Conlan, S., Deming, C. B., Davis, J., Young, A. C., … Segre, J. A. (2009). Topographical and Temporal Diversity of the Human Skin Microbiome. Science, 324(5931), 1190–1192. https://doi.org/10.1126/science.1171700 Grice, E. A., & Segre, J. A. (2011). The skin microbiome. Nature Reviews Microbiology, 9(4), 244–253. https://doi.org/10.1038/nrmicro2537 Grigat, J., Soruri, A., Forssmann, U., Riggert, J., & Zwirner, J. (2007). Chemoattraction of Macrophages, T Lymphocytes, and Mast Cells Is Evolutionarily Conserved within the Human -Defensin Family. The Journal of Immunology, 179(6), 3958–3965. https://doi.org/10.4049/jimmunol.179.6.3958 Gross, S., Gammon, S. T., Moss, B. L., Rauch, D., Harding, J., Heinecke, J. W., … Piwnica- Worms, D. (2009). Bioluminescence imaging of myeloperoxidase activity in vivo. Nature Medicine, 15(4), 455–461. https://doi.org/10.1038/nm.1886 Gruenheid, S., Canonne-Hergaux, F., Gauthier, S., Hackam, D. J., Grinstein, S., & Gros, P. (1999). The Iron Transport Protein NRAMP2 Is an Integral Membrane Glycoprotein That Colocalizes with Transferrin in Recycling Endosomes. The Journal of Experimental Medicine, 189(5), 831–841. https://doi.org/10.1084/jem.189.5.831 Gudmundsson, G. H., Agerberth, B., Odeberg, J., Bergman, T., Olsson, B., & Salcedo, R. (1996). The human gene FALL39 and processing of the cathelin precursor to the antibacterial peptide LL-37 in granulocytes. European Journal of Biochemistry / FEBS, 238(2), 325–332. https://doi.org/10.1371/journal.pone.0017755 Guida C, Altamura S, Klein FA, Galy B, Boutros M, Ulmer AJ, Hentze MW, M. M. (2015). A novel inflammatory pathway mediating rapid hepcidin-independent hypoferremia. Blood, 125(14). https://doi.org/10.1182/blood Gunshin, H., Mackenzie, B., Berger, U. V., Gunshin, Y., Romero, M. F., Boron, W. F., … Hediger, M. A. (1997). Cloning and characterization of a mammalian proton-coupled metal-ion transporter. Nature, 388(6641), 482–488. https://doi.org/10.1038/41343

140

BIBLIOGRAPHY

Hachem, J. P., Behne, M., Aronchik, I., Demerjian, M., Feingold, K. R., Elias, P. M., & Mauro, T. M. (2005). Extracellular pH controls NHE1 expression in epidermis and keratinocytes: Implications for barrier repair. Journal of Investigative Dermatology, 125(4), 790–797. https://doi.org/10.1111/j.0022-202X.2005.23836.x Hachem, J. P., Crumrine, D., Fluhr, J., Brown, B. E., Feingold, K. R., & Elias, P. M. (2003). pH directly regulates epidermal permeability barrier homeostasis, and stratum corneum integrity/cohesion. Journal of Investigative Dermatology, 121(2), 345–353. https://doi.org/10.1046/j.1523-1747.2003.12365.x Hallett, M. B., & Lloyds, D. (1995). Neutrophil priming: the cellular signals that say ‘amber’ but not ‘green.’ Immunology Today, 16(6), 264–268. https://doi.org/10.1016/0167- 5699(95)80178-2 Hardy, M. H. (1992). The secret life of the hair follicle. Trends in Genetics, 8(2), 55–61. https://doi.org/10.1016/0168-9525(92)90350-D Harris, Z. L., Durley, A. P., Man, T. K., & Gitlin, J. D. (1999). Targeted gene disruption reveals an essential role for ceruloplasmin in cellular iron efflux. Proceedings of the National Academy of Sciences of the United States of America, 96(19), 10812–10817. https://doi.org/10.1073/pnas.96.19.10812 HEINRICH, P. C., BEHRMANN, I., HAAN, S., HERMANNS, H. M., MÜLLER-NEWEN, G., & SCHAPER, F. (2003). Principles of interleukin (IL)-6-type cytokine signalling and its regulation. Biochemical Journal, 374(1), 1–20. https://doi.org/10.1042/bj20030407 Henningham, A., Döhrmann, S., Nizet, V., & Cole, J. N. (2015). Mechanisms of group A Streptococcus resistance to reactive oxygen species. FEMS Microbiology Reviews, 39(4), 488–508. https://doi.org/10.1093/femsre/fuu009 Hidalgo-Grass, C., Mishalian, I., Dan-Goor, M., Belotserkovsky, I., Eran, Y., Nizet, V., … Hanski, E. (2006). A streptococcal protease that degrades CXC chemokines and impairs bacterial clearance from infected tissues. EMBO Journal, 25(19), 4628–4637. https://doi.org/10.1038/sj.emboj.7601327 Hol, J., Wilhelmsen, L., & Haraldsen, G. (2010). The murine IL-8 homologues KC, MIP-2, and LIX are found in endothelial cytoplasmic granules but not in Weibel-Palade bodies. Journal of Leukocyte Biology, 87(3), 501–508. https://doi.org/10.1189/jlb.0809532 Holmes, W. E., Lee, J., Kuang, W. J., Rice, G. C., & Wood, W. I. (1991). Structure and functional expression of a human interleukin-8 receptor. Science (New York, N.Y.), 253(5025), 1278–1280. https://doi.org/10.1126/science.1840701 Höpfner, M., Nitsche, R., Rohr, A., Harms, D., Schubert, S., & Fölsch, U. R. (2001). Yersinia enterocolitica infection with multiple liver abscesses uncovering a primary hemochromatosis. Scandinavian Journal of Gastroenterology, 36(2), 220–224. https://doi.org/10.1080/003655201750066004 Houamel, D., Ducrot, N., Lefebvre, T., Daher, R., Moulouel, B., Sari, M.-A., … Karim, Z. (2016). Hepcidin as a Major Component of Renal Antibacterial Defenses against Uropathogenic Escherichia coli. Journal of the American Society of Nephrology, 27(3), 835–846. https://doi.org/10.1681/ASN.2014101035

141

BIBLIOGRAPHY

Houamel, D., Ducrot, N., Lefebvre, T., Daher, R., Moulouel, B., Sari, M.-A., … Karim, Z. (2016). Hepcidin as a Major Component of Renal Antibacterial Defenses against Uropathogenic Escherichia coli. Journal of the American Society of Nephrology, 27(3), 835–846. https://doi.org/10.1681/ASN.2014101035 Hu, Y., Li, A., Xu, Y., Jiang, B., Lu, G., & Luo, X. (2017). Transcriptomic variation of locally-infected skin of Epinephelus coioides reveals the mucosal immune mechanism against Cryptocaryon irritans. Fish and Shellfish Immunology, 66, 398–410. https://doi.org/10.1016/j.fsi.2017.05.042 Huber, A. R., Kunkel, S. L., Todd, R. F., & Weiss, S. J. (1991). Regulation of transendothelial migration by endogenous interleukin 8. Science, 254(15), 99–102. Iwase, T., Uehara, Y., Shinji, H., Tajima, A., Seo, H., Takada, K., … Mizunoe, Y. (2010). Staphylococcus epidermidis Esp inhibits Staphylococcus aureus biofilm formation and nasal colonization. Nature, 465(7296), 346–349. https://doi.org/10.1038/nature09074 Jabado, N., Jankowski, A., Dougaparsad, S., Picard, V., Grinstein, S., & Gros, P. (2000). Natural resistance to intracellular infections: natural resistance-associated macrophage protein 1 (Nramp1) functions as a pH-dependent manganese transporter at the phagosomal membrane. The Journal of Experimental Medicine, 192(9), 1237–1248. https://doi.org/10.1084/jem.192.9.1237 Jetten, A. M., & Harvat, B. L. (1997). Epidermal differentiation and squamous metaplasia: from stem cell to cell death. J.Dermatol., 24(0385–2407 (Print)), 711–725. Kansal, R. G., McGeer, A., Low, D. E., Norrby-Teglund, A., & Kotb, M. (2000). GAS serotype. Infection and Immunity, 68(11), 6362–6369. https://doi.org/10.1128/IAI.68.11.6362-6369.2000 Kautz, L., Jung, G., Nemeth, E., & Ganz, T. (2014). Erythroferrone contributes to recovery from anemia of inflammation. Blood, 124(16). https://doi.org/10.1182/blood-2014-06- 584607 Kautz, L., Jung, G., Valore, E. V, Rivella, S., Nemeth, E., & Ganz, T. (2014). Identification of erythroferrone as an erythroid regulator of iron metabolism. Nature Genetics, 46(7), 678–684. https://doi.org/10.1038/ng.2996 Kautz, L., Meynard, D., Besson-Fournier, C., Darnaud, V., Al Saati, T., Coppin, H., & Roth, M. P. (2009). BMP/Smad signaling is not enhanced in Hfe-deficient mice despite increased Bmp6 expression. Blood, 114(12), 2515–2520. https://doi.org/10.1182/blood- 2009-02-206771 Kautz, L., Meynard, D., Monnier, A., Darnaud, V., Bouvet, R., Wang, R. H., … Roth, M. P. (2008). Iron regulates phosphorylation of Smad1/5/8 and gene expression of Bmp6, Smad7, Id1, and Atoh8 in the mouse liver. Blood, 112(4), 1503–1509. https://doi.org/10.1182/blood-2008-03-143354 Kawabata, H., Fleming, R. E., Gui, D., Moon, S. Y., Saitoh, T., O’Kelly, J., … Koeffler, H. P. (2005). Expression of hepcidin is down-regulated in TfR2 mutant mice manifesting a phenotype of hereditary hemochromatosis. Blood, 105(1), 376–381. https://doi.org/10.1182/blood-2004-04-1416

142

BIBLIOGRAPHY

Kim, A., Fung, E., Parikh, S. G., Valore, E. V, Gabayan, V., Nemeth, E., & Ganz, T. (2015). A mouse model of anemia of inflammation : complex pathogenesis with partial dependence on hepcidin, 123(8), 1129–1137. https://doi.org/10.1182/blood-2013-08- 521419.A.K. Kowalczyk, A. P., Bornslaeger, E. A., Norvell, S. M., Palka, H. L., & Green, K. J. (1999). Desmosomes: intercellular adhesive junctions specialized for attachment of intermediate filaments. Int Rev Cytol, 185, 237–302. https://doi.org/10.1016/S0074-7696(08)60153-9 Krause, A., Neitz, S., Schulz, A., Forssmann, W., Schulz-knappe, P., & Adermann, K. (2000). LEAP-1 , a novel highly disul ¢ de-bonded human peptide , exhibits antimicrobial activity 1. FEBS Letters, 480, 147–150. https://doi.org/10.1016/S0014-5793(00)01920-7 Kulaksiz, H., Theilig, F., Bachmann, S., Gehrke, S. G., Rost, D., Janetzko, A., … Stremmel, W. (2005). The iron-regulatory peptide hormone hepcidin: Expression and cellular localization in the mammalian kidney. Journal of Endocrinology, 184(2), 361–370. https://doi.org/10.1677/joe.1.05729 Kurosaka, K., Chen, Q., Yarovinsky, F., Oppenheim, J. J., & Yang, D. (2005). Mouse Cathelin-Related Antimicrobial Peptide Chemoattracts Leukocytes Using Formyl Peptide Receptor-Like 1/Mouse Formyl Peptide Receptor-Like 2 as the Receptor and Acts as an Immune Adjuvant. The Journal of Immunology, 174(10), 6257–6265. https://doi.org/10.4049/jimmunol.174.10.6257 Kurupati, P., Turner, C. E., Tziona, I., Lawrenson, R. A., Alam, F. M., Nohadani, M., … Sriskandan, S. (2010a). Chemokine-cleaving Streptococcus pyogenes protease SpyCEP is necessary and sufficient for bacterial dissemination within soft tissues and the respiratory tract. Molecular Microbiology. https://doi.org/10.1111/j.1365- 2958.2010.07065.x Kurupati, P., Turner, C. E., Tziona, I., Lawrenson, R. A., Alam, F. M., Nohadani, M., … Sriskandan, S. (2010b). Chemokine-cleaving Streptococcus pyogenes protease SpyCEP is necessary and sufficient for bacterial dissemination within soft tissues and the respiratory tract. Molecular Microbiology, 76(6), 1387–1397. https://doi.org/10.1111/j.1365-2958.2010.07065.x Lai, Y., & Gallo, R. L. (2009). AMPed up immunity: how antimicrobial peptides have multiple roles in immune defense. Trends in Immunology. https://doi.org/10.1016/j.it.2008.12.003 Lakhal-Littleton, S., Wolna, M., Carr, C. A., Miller, J. J. J., Christian, H. C., Ball, V., … Robbins, P. A. (2015). Cardiac ferroportin regulates cellular iron homeostasis and is important for cardiac function. Proceedings of the National Academy of Sciences, 112(10), 3164–3169. https://doi.org/10.1073/pnas.1422373112 Lakhal-Littleton, S., Wolna, M., Chung, Y. J., Christian, H. C., Heather, L. C., Brescia, M., … Robbins, P. A. (2016). An essential cell-autonomous role for hepcidin in cardiac iron homeostasis. ELife, 5(NOVEMBER2016). https://doi.org/10.7554/eLife.19804 Lamagni, T. L., Darenberg, J., Luca-Harari, B., Siljander, T., Efstratiou, A., Henriques- Normark, B., … Neal, S. (2008). Epidemiology of severe Streptococcus pyogenes

143

BIBLIOGRAPHY

disease in Europe. Journal of Clinical Microbiology, 46(7), 2359–2367. https://doi.org/10.1128/JCM.00422-08 Larock, C. N., Döhrmann, S., Todd, J., Corriden, R., Olson, J., Johannssen, T., … Nizet, V. (2015). Group A Streptococcal M1 Protein Sequesters Cathelicidin to Evade Innate Immune Killing HHS Public Access. Cell Host Microbe, 18(4), 471–477. https://doi.org/10.1016/j.chom.2015.09.004 Lechler, T., & Fuchs, E. (2005). Asymmetric cell divisions promote stratification and differentiation of mammalian skin. Nature, 437(7056), 275–280. https://doi.org/10.1038/nature03922 Lee, E. Y., Lee, M. W., & Wong, G. C. L. (2018). Modulation of toll-like receptor signaling by antimicrobial peptides. Seminars in Cell and Developmental Biology. https://doi.org/10.1016/j.semcdb.2018.02.002 Lee, P., Peng, H., Gelbart, T., & Beutler, E. (2004). The IL-6- and lipopolysaccharide-induced transcription of hepcidin in HFE-, transferrin receptor 2-, and beta 2-microglobulin- deficient hepatocytes. Proceedings of the National Academy of Sciences of the United States of America, 101(25), 9263–9265. https://doi.org/10.1073/pnas.0403108101 Lee, W. L., Harrison, R. E., & Grinstein, S. (2003). Phagocytosis by neutrophils. Microbes and Infection, 5(14), 1299–1306. https://doi.org/10.1016/j.micinf.2003.09.014 Leung, P. S., Srai, S. K., Mascarenhas, M., Churchill, L. J., & Debnam, E. S. (2005). Increased duodenal iron uptake and transfer in a rat model of chronic hypoxia is accompanied by reduced hepcidin expression. Gut, 54(10), 1391–1395. https://doi.org/10.1136/gut.2004.062083 Levin, J. C., & Wessels, M. R. (1998). Identification of csrR/csrS, a genetic locus that regulates hyaluronic acid capsule synthesis in group A streptococcus. Molecular Microbiology, 30(1), 209–219. https://doi.org/10.1046/j.1365-2958.1998.01057.x Ley, K., Laudanna, C., Cybulsky, M. I., & Nourshargh, S. (2007, September 1). Getting to the site of inflammation: The leukocyte adhesion cascade updated. Nature Reviews Immunology. Nature Publishing Group. https://doi.org/10.1038/nri2156 Lichti, U., Anders, J., & Yuspa, S. H. (2008). Isolation and short-term culture of primary keratinocytes, hair follicle populations and dermal cells from newborn mice and keratinocytes from adult mice for in vitro analysis and for grafting to immunodeficient mice. Nature Protocols, 3(5), 799–810. https://doi.org/10.1038/nprot.2008.50 Lieschke, G. J., Grail, D., Hodgson, G., Metcalf, D., Stanley, E., Cheers, C., … Dunn, a R. (1994). Mice lacking granulocyte colony-stimulating factor have chronic neutropenia, granulocyte and macrophage progenitor cell deficiency, and impaired neutrophil mobilization. Blood, 84(6), 1737–1746. Lim, X., Tan, S. H., Lian, W., Koh, C., Man, R., Chau, W., … Nusse, R. (2013). Interfollicular epidermal stem cell self-renew cia autocrine Wnt signaling. Science, 342(6163), 1226–1230. https://doi.org/10.1126/science.1239730.Interfollicular Lin, D., Ding, J., Liu, J. Y., He, Y. F., Dai, Z., Chen, C. Z., … Wang, X. (2013). Decreased

144

BIBLIOGRAPHY

Serum Hepcidin Concentration Correlates with Brain Iron Deposition in Patients with HBV-Related Cirrhosis. PLoS ONE, 8(6), 1–9. https://doi.org/10.1371/journal.pone.0065551 Linehan, J. L., Harrison, O. J., Han, S. J., Byrd, A. L., Vujkovic-Cvijin, I., Villarino, A. V., … Belkaid, Y. (2018). Non-classical Immunity Controls Microbiota Impact on Skin Immunity and Tissue Repair. Cell, 172(4), 784–796.e18. https://doi.org/10.1016/j.cell.2017.12.033 Liu, Q., Davidoff, O., Niss, K., & Haase, V. H. (2012). Hypoxia-inducible factor regulates hepcidin via erythropoietin-induced erythropoiesis. Journal of Clinical Investigation, 122(12), 4635–4644. https://doi.org/10.1172/JCI63924 Lou, D. Q., Nicolas, G., Lesbordes, J. C., Viatte, L., Grimber, G., Szajnert, M. F., … Vaulont, S. (2004). Functional differences between hepcidin 1 and 2 in transgenic mice. Blood, 103(7), 2816–2821. https://doi.org/10.1182/blood-2003-07-2524 Mackenzie, I. C. (1970). Relationship between mitosis and the ordered structure of the stratum corneum in mouse epidermis. Nature, 226(5246), 653–655. https://doi.org/10.1038/226653a0 Maisetta, G., Vitali, A., Scorciapino, M. A., Rinaldi, A. C., Petruzzelli, R., Brancatisano, F. L., … Batoni, G. (2013). PH-dependent disruption of Escherichia coli ATCC 25922 and model membranes by the human antimicrobial peptides hepcidin 20 and 25. FEBS Journal, 280(12), 2842–2854. https://doi.org/10.1111/febs.12288 Mann, E. R. (2014). Review: Skin and the Immune System. Journal of Clinical & Experimental Dermatology Research, 4(3). https://doi.org/10.4172/2155-9554.S2-003 Masson, P. L., Heremans, J. F., & Schonne, E. (1969). Lactoferrin, an iron-binding protein in neutrophilic leukocytes. The Journal of Experimental Medicine, 130(3), 643–658. https://doi.org/10.1084/jem.130.3.643 Mastrogiannaki, M., Matak, P., Mathieu, J. R. R., Delga, S., Mayeux, P., Vaulont, S., & Peyssonnaux, C. (2012). Hepatic hypoxia-inducible factor-2 down-regulates hepcidin expression in mice through an erythropoietin-mediated increase in erythropoiesis. Haematologica, 97(6), 827–834. https://doi.org/10.3324/haematol.2011.056119 Matsubara, M., Harada, D., Manabe, H., & Hasegawa, K. (2004). Staphylococcus aureus peptidoglycan stimulates granulocyte macrophage colony-stimulating factor production from human epidermal keratinocytes via mitogen-activated protein kinases. FEBS Letters, 566(1–3), 195–200. https://doi.org/10.1016/j.febslet.2004.04.028 McKenzie, R. C., & Sauder, D. N. (1990). The role of keratinocyte cytokines in inflammation and immunity. The Journal of Investigative Dermatology. https://doi.org/10.1111/1523- 1747.ep12874955 McKie, A. T., Barrow, D., Latunde-Dada, G. O., Rolfs, A., Sager, G., Mudaly, E., … Simpson, R. J. (2001). An iron-regulated ferric reductase associated with the absorption of dietary iron. Science, 291(5509), 1755–1759. https://doi.org/10.1126/science.1057206 Medina, E., Rohde, M., & Chhatwal, G. S. (2003). Intracellular Survival of Streptococcus

145

BIBLIOGRAPHY

pyogenes in Polymorphonuclear Cells Results in Increased Bacterial Virulence. INFECTION AND IMMUNITY, 71(9), 5376–5380. https://doi.org/10.1128/IAI.71.9.5376 Mehrel, T., Hohl, D., Rothnagel, J. A., Longley, M. A., Bundman, D., Cheng, C., … Roop, D. R. (1990). Identification of a major keratinocyte cell envelope protein, loricrin. Cell, 61(6), 1103–1112. https://doi.org/10.1016/0092-8674(90)90073-N Mempel, M., Voelcker, V., Köllisch, G., Plank, C., Rad, R., Gerhard, M., … Ollert, M. (2003). Toll-Like Receptor Expression in Human Keratinocytes: Nuclear Factor κB Controlled Gene Activation by Staphylococcus aureus is Toll-Like Receptor 2 but Not Toll-Like Receptor 4 or Platelet Activating Factor Receptor Dependent. Journal of Investigative Dermatology, 121(6), 1389–1396. https://doi.org/10.1111/j.1523- 1747.2003.12630.x Merle, U., Fein, E., Gehrke, S. G., Stremmel, W., & Kulaksiz, H. (2007). The iron regulatory peptide hepcidin is expressed in the heart and regulated by hypoxia and inflammation. Endocrinology, 148(6), 2663–2668. https://doi.org/10.1210/en.2006-1331 Meynard, D., Kautz, L., Darnaud, V., Canonne-Hergaux, F., Coppin, H., & Roth, M.-P. (2009). Lack of the bone morphogenetic protein BMP6 induces massive iron overload. Nature Genetics, 41(4), 478–481. https://doi.org/10.1038/ng.320 Michels, K. R., Zhang, Z., Bettina, A. M., Cagnina, R. E., Stefanova, D., Burdick, M. D., … Mehrad, B. (2017). Hepcidin-mediated iron sequestration protects against bacterial dissemination during pneumonia, 2(6). https://doi.org/10.1172/jci.insight.92002 Middleton, J., Neil, S., Wintle, J., Clark-Lewis, I., Moore, H., Charles, L., … Antal, R. (1997). Transcytosis and surface presentation of IL-8 by venular endothelial cells. Cell, 91(3), 385–395. https://doi.org/10.1016/S0092-8674(00)80422-5 Mishalian, I., Ordan, M., Peled, A., Maly, A., Eichenbaum, M. B., Ravins, M., … Hanski, E. (2011a). Recruited Macrophages Control Dissemination of Group A Streptococcus from Infected Soft Tissues. The Journal of Immunology. https://doi.org/10.4049/jimmunol.1101385 Mishalian, I., Ordan, M., Peled, A., Maly, A., Eichenbaum, M. B., Ravins, M., … Hanski, E. (2011b). Recruited Macrophages Control Dissemination of Group A Streptococcus from Infected Soft Tissues. The Journal of Immunology, 187(11), 6022–6031. https://doi.org/10.4049/jimmunol.1101385 Moepps, B., Nuesseler, E., Braun, M., & Gierschik, P. (2006). A homolog of the human chemokine receptor CXCR1 is expressed in the mouse. Molecular Immunology, 43, 897– 914. https://doi.org/10.1016/j.molimm.2005.06.043 Montagna, W., & Parakkal, P. F. (1974). The structure and function of skin. Academic Press. Retrieved from https://www.sciencedirect.com/science/book/9780125052634 Moses, A. E., Wessels, M. R., Zalcman, K., Albertí, S., Natanson-Yaron, S., Menes, T., & Hanski, E. (1997). Relative contributions of hyaluronic acid capsule and M protein to virulence in a mucoid strain of the group A Streptococcus. Infection and Immunity, 65(1), 64–71.

146

BIBLIOGRAPHY

Motley, S. T., Morrow, B. J., Liu, X., Dodge, I. L., Vitiello, A., Ward, C. K., & Shaw, K. J. (2004). Simultaneous analysis of host and pathogen interactions during an in vivo infection reveals local induction of host acute phase response proteins, a novel bacterial stress response, and evidence of a host-imposed metal ion limited environment. Cellular Microbiology, 6(9), 849–865. https://doi.org/10.1111/j.1462-5822.2004.00407.x Moulouel, B., Houamel, D., Delaby, C., Tchernitchko, D., Vaulont, S., Letteron, P., … Karim, Z. (2013). Hepcidin regulates intrarenal iron handling at the distal nephron. Kidney International, 84(4), 756–766. https://doi.org/10.1038/ki.2013.142 Murakami, M., Ohtake, T., Dorschner, R. A., Schittek, B., Garbe, C., & Gallo, R. L. (2002). Cathelicidin anti-microbial peptide expression in sweat, an innate defense system for the skin. Journal of Investigative Dermatology, 119(5), 1090–1095. https://doi.org/10.1046/j.1523-1747.2002.19507.x Nagarkar, D. R., Wang, Q., Shim, J., Zhao, Y., Tsai, W. C., Lukacs, N. W., … Hershenson, M. B. (2009). CXCR2 Is Required for Neutrophilic Airway Inflammation and Hyperresponsiveness in a Mouse Model of Human Rhinovirus Infection. The Journal of Immunology, 183(10), 6698–6707. https://doi.org/10.4049/jimmunol.0900298 Naik, S., Bouladoux, N., Linehan, J. L., Han, S. J., Harrison, O. J., Wilhelm, C., … Belkaid, Y. (2015). Commensal-dendritic-cell interaction specifies a unique protective skin immune signature. Nature, 520(7545), 104–108. https://doi.org/10.1038/nature14052 Nelson, G. E., Pondo, T., Toews, K.-A., Farley, M. M., Lindegren, M. Lou, Lynfield, R., … Van Beneden, C. A. (2016). Epidemiology of Invasive Group A Streptococcal Infections in the United States, 2005–2012. Clinical Infectious Diseases, 63(4), 478–486. https://doi.org/10.1093/cid/ciw248 Nelson, W. G., & Sun, T. T. (1983). The 50- and 580 kdalton keratin classes as molecular markers for stratified squamous epithilia: cell culture studies. J. Cell Biol., 97(July), 244–251. https://doi.org/10.1083/jcb.97.1.244 Nemeth E, Tuttle MS, Powelson J, Vaughn MB, Donovan A, Ward DM, Ganz T, K. J. (2004). Hepcidin Regulates Cellular Iron Efflux by Binding to Ferroportin and Inducing Its Internalization. Science, 306(5704), 2087–2090. https://doi.org/10.1126/science.1103786 Nemeth, E., Preza, G. C., Jung, C. L., Kaplan, J., Waring, A. J., & Ganz, T. (2006). The N- terminus of hepcidin is essential for its interaction with ferroportin: Structure-function study. Blood, 107(1), 328–333. https://doi.org/10.1182/blood-2005-05-2049 Nemeth, E., Rivera, S., Gabayan, V., Keller, C., Taudorf, S., Pedersen, B. K., & Ganz, T. (2004). IL-6 mediates hypoferremia of inflammation by inducing the synthesis of the iron regulatory hormone hepcidin. Journal of Clinical Investigation, 113(9), 1271–1276. https://doi.org/10.1172/JCI200420945 Nemeth, E., Roetto, A., Garozzo, G., Ganz, T., & Camaschella, C. (2005). Hepcidin is decreased in TFR2 hemochromatosis. Blood, 105(4), 1803–1806. https://doi.org/10.1182/blood-2004-08-3042 Nemeth, E., Valore, E. V, Territo, M., Schiller, G., Lichtenstein, A., & Ganz, T. (2003). Hepcidin, a putative mediator of anemia of inflammation, is a type II acute-phase

147

BIBLIOGRAPHY

protein. Blood, 101(7), 2461–2463. https://doi.org/10.1182/blood-2002-10-3235 Nestle, F. O., Di Meglio, P., Qin, J.-Z., & Nickoloff, B. J. (2009). Skin immune sentinels in health and disease. Nature Reviews Immunology, 9(10), 679–691. https://doi.org/10.1038/nri2622 Neumann, A., Berends, E. T. M., Nerlich, A., Molhoek, E. M., Gallo, R. L., Meerloo, T., … von Köckritz-Blickwede, M. (2014). The antimicrobial peptide LL-37 facilitates the formation of neutrophil extracellular traps. Biochemical Journal, 464(1), 3–11. https://doi.org/10.1042/BJ20140778 Nguyen, G. T., Green, E. R., & Mecsas, J. (2017). Neutrophils to the ROScue: Mechanisms of NADPH Oxidase Activation and Bacterial Resistance. Frontiers in Cellular and Infection Microbiology, 7(August). https://doi.org/10.3389/fcimb.2017.00373 Nicolas, G., Bennoun, M., Devaux, I., Beaumont, C., Grandchamp, B., Kahn, A., & Vaulont, S. (2001). Lack of hepcidin gene expression and severe tissue iron overload in upstream stimulatory factor 2 (USF2) knockout mice. Proceedings of the National Academy of Sciences, 98(15), 8780–8785. https://doi.org/10.1073/pnas.151179498 Nicolas, G., Chauvet, C., Viatte, L., Danan, J. L., Bigard, X., Devaux, I., … Vaulont, S. (2002). The gene encoding the iron regulatory peptide hepcidin is regulated by anemia, hypoxia, and inflammation. Journal of Clinical Investigation, 110(7), 1037–1044. https://doi.org/10.1172/JCI200215686 Niessen, C. M. (2007). Tight junctions/adherens junctions: Basic structure and function. Journal of Investigative Dermatology, 127(11), 2525–2532. https://doi.org/10.1038/sj.jid.5700865 Nizet, V., Ohtake, T., Lauth, X., Trowbridge, J., Rudisill, J., Dorschner, R. A., … Gallo, R. L. (2001). Innate antimicrobial peptide protects the skin from invasive bacterial infection. Nature, 414(6862), 454–457. https://doi.org/10.1038/35106587 Novotny, M., Datta, S., Biswas, R., & Hamilton, T. (2005). Functionally independent AU-rich sequence motifs regulate KC (CXCL1) mRNA. Journal of Biological Chemistry, 280(34), 30166–30174. https://doi.org/10.1074/jbc.M502280200 O’Grady, K. A., Kelpie, L., Andrews, R. M., Curtis, N., Nolan, T. M., Selvaraj, G., … Carapetis, J. R. (2007). The epidemiology of invasive group A streptococcal disease in Victoria, Australia. Med J Aust, 186(11), 565–569. https://doi.org/ogr10917_fm [pii] O’Neill, A. M., Thurston, T. L., & Holden, D. W. (2016). Cytosolic Replication of Group A Streptococcus in Human. MBio, 7(2), 1–12. https://doi.org/10.1128/mBio.00020- 16.Invited Oehmcke, S., Shannon, O., Mörgelin, M., & Herwald, H. (2010). Streptococcal M proteins and their role as virulence determinants. Clinica Chimica Acta, 411(17–18), 1172–1180. https://doi.org/10.1016/j.cca.2010.04.032 Ohnemus, U., Kohrmeyer, K., Houdek, P., Rohde, H., Wladykowski, E., Vidal, S., … Brandner, J. M. (2008). Regulation of epidermal tight-junctions (TJ) during infection with exfoliative toxin-negative Staphylococcus strains. Journal of Investigative

148

BIBLIOGRAPHY

Dermatology, 128(4), 906–916. https://doi.org/10.1038/sj.jid.5701070 Olakanmi, O., Schlesinger, L. S., & Britigan, B. E. (2006). Hereditary hemochromatosis results in decreased iron acquisition and growth by Mycobacterium tuberculosis within human macrophages. Journal of Leukocyte Biology, 81(1), 195–204. https://doi.org/10.1189/jlb.0606405 Olaru, F., & Jensen, L. E. (2010). Staphylococcus aureus stimulates neutrophil targeting chemokine expression in keratinocytes through an autocrine IL-1α signaling loop. Journal of Investigative Dermatology, 130(7), 1866–1876. https://doi.org/10.1038/jid.2010.37 Oshima, H., Rochat, A., Kedzia, C., Kobayashi, K., & Barrandon, Y. (2001). Morphogenesis and renewal of hair follicles from adult multipotent stem cells. Cell, 104(2), 233–245. https://doi.org/10.1016/S0092-8674(01)00208-2 Øynebråten, I., Bakke, O., Brandtzaeg, P., Johansen, F.-E., & Haraldsen, G. (2004). Rapid chemokine secretion from endothelial cells originates from 2 distinct compartments. Blood, 104(2), 314–320. https://doi.org/10.1182/blood-2003-08-2891 Oynebraten, I., Barois, N., Hagelsteen, K., Johansen, F.-E., Bakke, O., & Haraldsen, G. (2005). Characterization of a Novel Chemokine-Containing Storage Granule in Endothelial Cells: Evidence for Preferential Exocytosis Mediated by Protein Kinase A and Diacylglycerol. The Journal of Immunology, 175(8), 5358–5369. https://doi.org/10.4049/jimmunol.175.8.5358 Panganiban, R. P., Vonakis, B. M., Ishmael, F. T., & Stellato, C. (2014). Coordinated Post- Transcriptional Regulation of the Chemokine System: Messages from CCL2. Journal of Interferon & Cytokine Research, 34(4), 255–266. https://doi.org/10.1089/jir.2013.0149 Papayannopoulos, V. (2018). Neutrophil extracellular traps in immunity and disease. Nature Reviews Immunology, 18(2), 134–147. https://doi.org/10.1038/nri.2017.105 Park, C. H., Valore, E. V., Waring, A. J., & Ganz, T. (2001a). Hepcidin, a Urinary Antimicrobial Peptide Synthesized in the Liver. Journal of Biological Chemistry, 276(11), 7806–7810. https://doi.org/10.1074/jbc.M008922200 Park, C. H., Valore, E. V, Waring, A. J., & Ganz, T. (2001b). Hepcidin, a Urinary Antimicrobial Peptide Synthesized in the Liver. Journal of Biological Chemistry, 276(11), 7806–7810. https://doi.org/10.1074/jbc.M008922200 Pasparakis, M., Haase, I., & Nestle, F. O. (2014). Mechanisms regulating skin immunity and inflammation. Nature Reviews Immunology. https://doi.org/10.1038/nri3646 Pechlaner, R., Kiechl, S., Mayr, M., Santer, P., Weger, S., Haschka, D., … Weiss, G. (2016). Correlates of serum hepcidin levels and its association with cardiovascular disease in an elderly general population. Clinical Chemistry and Laboratory Medicine, 54(1), 151– 161. https://doi.org/10.1515/cclm-2015-0068 Persson, S. T., Wilk, L., Mörgelin, M., & Herwald, H. (2015). Vigilant keratinocytes trigger pathogen-associated molecular pattern signaling in response to streptococcal M1 protein. Infection and Immunity. https://doi.org/10.1128/IAI.00887-15

149

BIBLIOGRAPHY

Peterson, M. L., McCormick, J. K., & Schlievert, P. M. (2006). Toxins and Superantigens of Group A Streptococci. In Gram-Positive Pathogens, Second Edition (pp. 47–58). American Society of Microbiology. https://doi.org/10.1128/9781555816513.ch5 Peveri, P., Walz, A., Dewald, B., Baggiolini, M., Baggiolini, M., Institut, T. K., … Bern, C.-. (1988). A NOVEL NEUTROPHIL-ACTIVATING FACTOR PRODUCED BY HUMAN MONONUCLEAR PHAGOCYTES. J. Exp. Med, 167(May). Peyssonnaux, C., Datta, V., Cramer, T., Doedens, A., Theodorakis, E. A., Gallo, R. L., … Johnson, R. S. (2005). HIF-1α expression regulates the bactericidal capacity of phagocytes. Journal of Clinical Investigation, 115(7), 1806–1815. https://doi.org/10.1172/JCI23865 Peyssonnaux, C., Zinkernagel, A. S., Datta, V., Lauth, X., Johnson, R. S., & Nizet, V. (2006). TLR4-dependent hepcidin expression by myeloid cells in response to bacterial pathogens. Blood, 107(9), 3727–3732. https://doi.org/10.1182/blood-2005-06-2259 Pigeon, C., Ilyin, G., Courselaud, B., Leroyer, P., Turlin, B., Brissot, P., & Loréal, O. (2001). A New Mouse Liver-specific Gene, Encoding a Protein Homologous to Human Antimicrobial Peptide Hepcidin, Is Overexpressed during Iron Overload. Journal of Biological Chemistry, 276(11), 7811–7819. https://doi.org/10.1074/jbc.M008923200 Piperno, A., Stefania Galimberti, Raffaella Mariani, Sara Pelucchi, Giulia Ravasi, Carolina Lombardi, Grzegorz Bilo, Miriam Revera, Andrea Giuliano, Andrea Faini, Veronica Mainini, MarkWesterman, Tomas Ganz, Maria Grazia Valsecchi, Giuseppe Mancia, and G. P., & 1Department. (2011). Modulation of hepcidin production during hypoxia- induced erythropoiesis in humans in vivo: data from the HIGHCARE project. Blood, 117(10), 2953–2959. https://doi.org/10.1182/blood Pivarcsi, A., Bodai, L., Réthi, B., Kenderessy-Szabó, A., Koreck, A., Széll, M., … Kemény, L. (2003). Expression and function of Toll-like receptors 2 and 4 in human keratinocytes. International Immunology, 15(6), 721–730. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12750356 Pivarcsi, A., Nagy, I., & Kemeny, L. (2005). Innate Immunity in the Skin: How Keratinocytes Fight Against Pathogens. Current Immunology Reviews, 1(1), 29–42. https://doi.org/10.2174/1573395052952941 Ponka, P., & Lok, C. N. (1999). The transferrin receptor: role in health and disease. The International Journal of Biochemistry & Cell Biology, 31(10), 1111–1137. https://doi.org/10.1016/S1357-2725(99)00070-9 Potten, C. S. (1974). The epidermal proliferative unit: the possible role of the central basal cell. Cell Tissue Kinet., 77–88. Preza, G. C., Ruchala, P., Pinon, R., Ramos, E., Qiao, B., Peralta, M. A., … Nemeth, E. (2011). Minihepcidins are rationally designed small peptides that mimic hepcidin activity in mice and may be useful for the treatment of iron overload. Journal of Clinical Investigation, 121(12), 4880–4888. https://doi.org/10.1172/JCI57693 Qiao, B., Sugianto, P., Fung, E., Del-Castillo-Rueda, A., Moran-Jimenez, M. J., Ganz, T., & Nemeth, E. (2012). Hepcidin-induced endocytosis of ferroportin is dependent on

150

BIBLIOGRAPHY

ferroportin ubiquitination. Cell Metabolism, 15(6), 918–924. https://doi.org/10.1016/j.cmet.2012.03.018 Qiu, A., Jansen, M., Sakaris, A., Min, S. H., Chattopadhyay, S., Tsai, E., … Goldman, I. D. (2006). Identification of an intestinal folate transporter and the molecular basis for hereditary folate malabsorption. Cell, 127(5), 917–928. https://doi.org/10.1016/j.cell.2006.09.041 Raffin, S. B., Woo, C. H., Roost, K. T., Price, D. C., & Schmid, R. (1974). Intestinal absorption of hemoglobin iron-heme cleavage by mucosal heme oxygenase. Journal of Clinical Investigation, 54(6), 1344–1352. https://doi.org/10.1172/JCI107881 Ramos, E., Kautz, L., Rodriguez, R., Hansen, M., Gabayan, V., Ginzburg, Y., … Ganz, T. (2011). Evidence for distinct pathways of hepcidin regulation by acute and chronic iron loading. Hepatology, 53(4), 1333–1341. https://doi.org/10.1002/hep.24178.Evidence Ratledge, C., & Dover, L. G. (2000). Iron Metabolism in Pathogenic Bacteria. Annual Review of Microbiology, 54(1), 881–941. https://doi.org/10.1146/annurev.micro.54.1.881 Regnier, E., Grange, P. A., Ollagnier, G., Crickx, E., Elie, L., Chouzenoux, S., … Dupin, N. (2016). Superoxide anions produced by Streptococcus pyogenes group A-stimulated keratinocytes are responsible for cellular necrosis and bacterial growth inhibition. Innate Immunity, 22(2), 113–123. https://doi.org/10.1177/1753425915619476 Rice, R. H., & Green, H. (1979). Presence in human epidermal cells of a soluble protein precursor of the cross-linked envelope: Activation of the cross-linking by calcium ions. Cell, 18(3), 681–694. https://doi.org/10.1016/0092-8674(79)90123-5 Richards, K. S., Bereiter-Hahn, J., & Matoltsy, A. G. (1984). Biology of the integument 2 Vertebrates. Comparative Biochemistry and Physiology Part A: Physiology (Vol. 86). https://doi.org/10.1016/0300-9629(87)90650-5 Rippke, F., Schreiner, V., & Schwanitz, H.-J. (2002). The acidic milieu of the horny layer: new findings on the physiology and pathophysiology of skin pH. American Journal of Clinical Dermatology, 3(4), 261–272. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12010071 Ritzman, A. M., Hughes-Hanks, J. M., Blaho, V. A., Wax, L. E., Mitchell, W. J., & Brown, C. R. (2010). The chemokine receptor CXCR2 ligand KC (CXCL1) mediates neutrophil recruitment and is critical for development of experimental lyme arthritis and carditis. Infection and Immunity, 78(11), 4593–4600. https://doi.org/10.1128/IAI.00798-10 Rodriguez, R., Jung, C. L., Gabayan, V., Deng, J. C., Ganz, T., Nemeth, E., & Bulut, Y. (2014). Hepcidin induction by pathogens and pathogen-derived molecules is strongly dependent on interleukin-6. Infection and Immunity, 82(2), 745–752. https://doi.org/10.1128/IAI.00983-13 Rohrl, J., Yang, D., Oppenheim, J. J., & Hehlgans, T. (2010). Human -Defensin 2 and 3 and Their Mouse Orthologs Induce Chemotaxis through Interaction with CCR2. The Journal of Immunology, 184(12), 6688–6694. https://doi.org/10.4049/jimmunol.0903984 Ross, S. L., Biswas, K., Rottman, J., Allen, J. R., Long, J., Miranda, L. P., … Arvedson, T. L.

151

BIBLIOGRAPHY

(2017). Identification of antibody and small molecule antagonists of ferroportin-hepcidin interaction. Frontiers in Pharmacology, 8(NOV). https://doi.org/10.3389/fphar.2017.00838 Russo, R. C., Garcia, C. C., Teixeira, M. M., & Amaral, F. A. (2014). The CXCL8/IL-8 chemokine family and its receptors in inflammatory diseases. Expert Review of Clinical Immunology, 10(5), 593–619. https://doi.org/10.1586/1744666X.2014.894886 Ryter, S. W., Alam, J., & Choi, A. M. K. (2006). Heme Oxygenase-1 / Carbon Monoxide : From Basic Science to Therapeutic Applications. Physiol Rev, 86(2), 583–650. https://doi.org/10.1152/physrev.00011.2005. Sacramento, L., Trevelin, S. C., Nascimento, M. S., Lima-Jùnior, D. S., Costa, D. L., Almeida, R. P., … Carregaro, V. (2015). Toll-like receptor 9 signaling in dendritic cells regulates neutrophil recruitment to inflammatory foci following Leishmania infantum infection. Infection and Immunity, 83(12), 4604–4616. https://doi.org/10.1128/IAI.00975-15 Schaible, U. E., & Kaufmann, S. H. E. (2004). Iron and microbial infection. Nature Reviews Microbiology. https://doi.org/10.1038/nrmicro1046 Scharschmidt, T. C., Vasquez, K. S., Truong, H. A., Gearty, S. V., Pauli, M. L., Nosbaum, A., … Rosenblum, M. D. (2015). A Wave of Regulatory T Cells into Neonatal Skin Mediates Tolerance to Commensal Microbes. Immunity, 43(5), 1011–1021. https://doi.org/10.1016/j.immuni.2015.10.016 Schwarz, P., Kübler, J. A. M., Strnad, P., Müller, K., Barth, T. F. E., Gerloff, A., … Kulaksiz, H. (2012). Hepcidin is localised in gastric parietal cells, regulates acid secretion and is induced by Helicobacter pylori infection. Gut, 61(2), 193–201. https://doi.org/10.1136/gut.2011.241208 Scott, M. G., Davidson, D. J., Gold, M. R., Bowdish, D., & Hancock, R. E. W. (2002). The Human Antimicrobial Peptide LL-37 Is a Multifunctional Modulator of Innate Immune Responses. The Journal of Immunology, 169(7), 3883–3891. https://doi.org/10.4049/jimmunol.169.7.3883 Seo, S. M., McIntire, L. V, & Smith, C. W. (2001). Effects of IL-8, Gro-alpha, and LTB(4) on the adhesive kinetics of LFA-1 and Mac-1 on human neutrophils. American Journal of Physiology. Cell Physiology, 281, C1568–C1578. Shanmugam, N. K. N., Chen, K., & Cherayil, B. J. (2015). Commensal Bacteria-induced interleukin 1beta (IL-1beta) secreted by macrophages up-regulates hepcidin expression in hepatocytes by activating the bone morphogenetic protein signaling pathway. Journal of Biological Chemistry, 290(51), 30637–30647. https://doi.org/10.1074/jbc.M115.689190 Shayeghi, M., Latunde-Dada, G. O., Oakhill, J. S., Laftah, A. H., Takeuchi, K., Halliday, N., … McKie, A. T. (2005). Identification of an intestinal heme transporter. Cell, 122(5), 789–801. https://doi.org/10.1016/j.cell.2005.06.025 Sheppard, F. R., Kelher, M. R., Moore, E. E., McLaughlin, N. J. D., Banerjee, A., & Silliman, C. C. (2005). Structural organization of the neutrophil NADPH oxidase: phosphorylation

152

BIBLIOGRAPHY

and translocation during priming and activation. Journal of Leukocyte Biology, 78(5), 1025–1042. https://doi.org/10.1189/jlb.0804442 Shi, H., Bencze, K. Z., Stemmler, T. L., & Philpott, C. C. (2008). A cytosolic iron chaperone that delivers iron to ferritin. Science, 320(5880), 1207–1210. https://doi.org/10.1126/science.1157643 Shi, J., & Camus, A. C. (2006). Hepcidins in amphibians and fishes: Antimicrobial peptides or iron-regulatory hormones? Developmental and Comparative Immunology, 30(9), 746– 755. https://doi.org/10.1016/j.dci.2005.10.009 Shike, H., Lauth, X., Westerman, M. E., Ostland, V. E., Carlberg, J. M., Van Olst, J. C., … Burns, J. C. (2002). Bass hepcidin is a novel antimicrobial peptide induced by bacterial challenge. European Journal of Biochemistry, 269(8), 2232–2237. https://doi.org/10.1046/j.1432-1033.2002.02881.x Shike, H., Shimizu, C., Lauth, X., & Burns, J. C. (2004). Organization and expression analysis of the zebrafish hepcidin gene, an antimicrobial peptide gene conserved among vertebrates. Developmental and Comparative Immunology, 28(7–8), 747–754. https://doi.org/10.1016/j.dci.2003.11.009 Simpson, C. L., Patel, D. M., & Green, K. J. (2011). Deconstructing the skin: Cytoarchitectural determinants of epidermal morphogenesis. Nature Reviews Molecular Cell Biology, 12(9), 565–580. https://doi.org/10.1038/nrm3175 Smith, C. L., Arvedson, T. L., Cooke, K. S., Dickmann, L. J., Forte, C., Li, H., … Maxwell, J. R. (2013). IL-22 regulates iron availability in vivo through the induction of hepcidin. Journal of Immunology (Baltimore, Md. : 1950), 191(4), 1845–1855. https://doi.org/10.4049/jimmunol.1202716 Spaulding, A. R., Salgado-Pabón, W., Kohler, P. L., Horswill, A. R., Leung, D. Y. M., & Schlievert, P. M. (2013). Staphylococcal and streptococcal superantigen exotoxins. Clinical Microbiology Reviews, 26(3), 422–447. https://doi.org/10.1128/CMR.00104-12 Spits, H., Artis, D., Colonna, M., Diefenbach, A., Di Santo, J. P., Eberl, G., … Vivier, E. (2013). Innate lymphoid cells-a proposal for uniform nomenclature. Nature Reviews Immunology, 13(2), 145–149. https://doi.org/10.1038/nri3365 Stadtmann, A., & Zarbock, A. (2012). CXCR2: From bench to bedside. Frontiers in Immunology, 3(AUG), 1–12. https://doi.org/10.3389/fimmu.2012.00263 Steer, A. C., Law, I., Matatolu, L., Beall, B. W., & Carapetis, J. R. (2009). Global emm type distribution of group A streptococci: systematic review and implications for vaccine development. The Lancet Infectious Diseases, 9(10), 611–616. https://doi.org/10.1016/S1473-3099(09)70178-1 Stefanova, D., Raychev, A., Arezes, J., Ruchala, P., Gabayan, V., Skurnik, M., … Nemeth, E. (2017a). Endogenous hepcidin and its agonist mediate resistance to selected infections by clearing non–transferrin-bound iron. Blood, 130(3), 245–257. https://doi.org/10.1182/blood-2017-03-772715 Stefanova, D., Raychev, A., Arezes, J., Ruchala, P., Gabayan, V., Skurnik, M., … Nemeth, E.

153

BIBLIOGRAPHY

(2017b). Endogenous hepcidin and its agonist mediate resistance to selected infections by clearing non–transferrin-bound iron. Blood, 130(3), 245–257. https://doi.org/10.1182/blood-2017-03-772715 Steinbicker, A. U., Bartnikas, T. B., Lohmeyer, L. K., Leyton, P., Mayeur, C., Kao, S. M., … Bloch, K. D. (2011). Perturbation of hepcidin expression by BMP type I receptor deletion induces iron overload in mice. Blood, 118(15). https://doi.org/10.1182/blood- 2011-03-339952 Stevens, D. L. (1992). Invasive group A Streptococcus infections. Clin Infect Dis, 14(1), 2– 11. https://doi.org/10.2307/4456232 Sugimoto, S., Iwamoto, T., Takada, K., Okuda, K. I., Tajima, A., Iwase, T., & Mizunoe, Y. (2013). Staphylococcus epidermidis Esp degrades specific proteins associatedwith staphylococcus aureus biofilm formation and host-pathogen interaction. Journal of Bacteriology, 195(8), 1645–1655. https://doi.org/10.1128/JB.01672-12 Sugita, K., Kabashima, K., Atarashi, K., Shimauchi, T., Kobayashi, M., & Tokura, Y. (2006). Innate immunity mediated by epidermal keratinocytes promotes acquired immunity involving Langerhans cells and T cells in the skin. Clinical and Experimental Immunology, 147(1), 176–183. https://doi.org/10.1111/j.1365-2249.2006.03258.x Sumby, P., Barbian, K. D., Gardner, D. J., Whitney, A. R., Welty, D. M., Long, R. D., … Musser, J. M. (2005). Extracellular deoxyribonuclease made by group A Streptococcus assists pathogenesis by enhancing evasion of the innate immune response. Proceedings of the National Academy of Sciences, 102(5), 1679–1684. https://doi.org/10.1073/pnas.0406641102 Sumby, P., Whitney, A. R., Graviss, E. A., DeLeo, F. R., & Musser, J. M. (2006). Genome- wide analysis of group A streptococci reveals a mutation that modulates global phenotype and disease specificity. PLoS Pathogens, 2(1), 0041–0049. https://doi.org/10.1371/journal.ppat.0020005 Swamydas, M., Gao, J. L., Break, T. J., Johnson, M. D., Jaeger, M., Rodriguez, C. A., … Lionakis, M. S. (2016). CXCR1-mediated neutrophil degranulation and fungal killing promote Candida clearance and host survival. Science Translational Medicine, 8(322). https://doi.org/10.1126/scitranslmed.aac7718 Taylor, G., Lehrer, M. S., Jensen, P. J., Sun, T.-T., & Lavker, R. M. (2000). Involvement of Follicular Stem Cells in Forming Not Only the Follicle but Also the Epidermis. Cell, 102(4), 451–461. https://doi.org/10.1016/S0092-8674(00)00050-7 Tettamanti, G., Malagoli, D., Benelli, R., Albini, A., Grimaldi, A., Perletti, G., … Ottaviani, E. (2006). Growth factors and chemokines: a comparative functional approach between invertebrates and vertebrates. Current Medicinal Chemistry, 13(23), 2737–2750. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/17073625 Theil, E. C., Matzapetakis, M., & Liu, X. (2006). Ferritins: Iron/oxygen biominerals in protein nanocages. Journal of Biological Inorganic Chemistry, 11(7), 803–810. https://doi.org/10.1007/s00775-006-0125-6 Thelen, M., Peveri, P., Kernen, P., von Tscharner, V., Walz, A., & Baggiolini, M. (1988).

154

BIBLIOGRAPHY

Mechanism of neutrophil activation by NAF, a novel monocyte-derived peptide agonist. The FASEB Journal : Official Publication of the Federation of American Societies for Experimental Biology, 2(11), 2702–2706. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2840318 Theurl, I., Fritsche, G., Ludwiczek, S., Garimorth, K., Bellmann-Weiler, R., & Weiss, G. (2005). The macrophage: A cellular factory at the interphase between iron and immunity for the control of infections. BioMetals, 18(4), 359–367. https://doi.org/10.1007/s10534- 005-3710-1 Tillett, W. S., Sherry, S., & Christensen, L. R. (1947). Streptococcal Desoxyribonuclease: Significance in Lysis of Purulent Exudates and Production by Strains of Hemolytic Streptococci. Proc Soc Exp Biol Med., 184–188. Timmer, A. M., Timmer, J. C., Pence, M. A., Hsu, L. C., Ghochani, M., Frey, T. G., … Nizet, V. (2009). Streptolysin O promotes group A streptococcus immune evasion by accelerated macrophage apoptosis. Journal of Biological Chemistry, 284(2), 862–871. https://doi.org/10.1074/jbc.M804632200 Truksa, J., Lee, P., & Beutler, E. (2009). Two BMPresponsive elements, STAT, and bZIP/HNF4/COUP motifs of the hepcidin promoter are critical for BMP, SMAD1, and HJV responsiveness. Blood, 113(3), 688–695. https://doi.org/10.1182/blood-2008-05- 160184 Truksa, J., Peng, H., Lee, P., & Beutler, E. (2006). Bone morphogenetic proteins 2, 4, and 9 stimulate murine hepcidin 1 expression independently of Hfe, transferrin receptor 2 (Tfr2), and IL-6. Proceedings of the National Academy of Sciences of the United States of America, 103(27). Retrieved from http://www.pnas.org/content/pnas/103/27/10289.full.pdf Tsai, W. C., Strieter, R. M., Mehrad, B., Newstead, M. W., Zeng, X., & Standiford, T. J. (2000). CXC chemokine receptor CXCR2 is essential for protective innate host response in murine Pseudomonas aeruginosa pneumonia. Infection and Immunity, 68(7), 4289– 4296. https://doi.org/10.1128/IAI.68.7.4289-4296.2000 Turksen, K., Kupper, T., Degenstein, L., Williams, I., & Fuchs, E. (1992). Interleukin 6: insights to its function in skin by overexpression in transgenic mice. Proc Natl Acad Sci U S A, 89(11), 5068–5072. Retrieved from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citat ion&list_uids=1375756 Uchiyama, S., Döhrmann, S., Timmer, A. M., Dixit, N., Ghochani, M., Bhandari, T., … Nizet, V. (2015). Streptolysin O rapidly impairs neutrophil oxidative burst and antibacterial responses to Group A Streptococcus. Frontiers in Immunology. https://doi.org/10.3389/fimmu.2015.00581 Valeri, M., & Raffatellu, M. (2016). Cytokines IL-17 and IL-22 in the host response to infection, 74, 111. https://doi.org/10.1093/femspd/ftw111 Valore, E. V., & Ganz, T. (2008). Posttranslational processing of hepcidin in human hepatocytes is mediated by the prohormone convertase furin. Blood Cells, Molecules,

155

BIBLIOGRAPHY

and Diseases, 40(1), 132–138. https://doi.org/10.1016/j.bcmd.2007.07.009 Vecchi, C., Montosi, G., Zhang, K., Lamberti, I., Duncan, S. A., Kaufman, R. J., & Pietrangelo, A. (2009). ER stress controls iron metabolism through induction of hepcidin. Science, 325(5942), 877–880. https://doi.org/10.1126/science.1176639 Verga Falzacappa, M. V., Casanovas, G., Hentze, M. W., & Muckenthaler, M. U. (2008). A bone morphogenetic protein (BMP)-responsive element in the hepcidin promoter controls HFE2-mediated hepatic hepcidin expression and its response to IL-6 in cultured cells. Journal of Molecular Medicine, 86(5), 531–540. https://doi.org/10.1007/s00109- 008-0313-7 Verga Falzacappa, M. V., Spasic, M. V., Kessler, R., Stolte, J., Hentze, M. W., & Muckenthaler, M. U. (2007). STAT3 mediates hepatic hepcidin expression and its inflammatory stimulation. Blood, 109(1), 353–358. https://doi.org/10.1182/blood-2006- 07-033969 Vidal, S. (1995). The Ity/Lsh/Bcg locus: natural resistance to infection with intracellular parasites is abrogated by disruption of the Nramp1 gene. Journal of Experimental Medicine, 182(3), 655–666. https://doi.org/10.1084/jem.182.3.655 Volz, T., Kaesler, S., & Biedermann, T. (2012). Innate immune sensing 2.0 - from linear activation pathways to fine tuned and regulated innate immune networks. Experimental Dermatology, 21(1), 61–69. https://doi.org/10.1111/j.1600-0625.2011.01393.x Vulpe, C. D., Kuo, Y. M., Murphy, T. L., Cowley, L., Askwith, C., Libina, N., … Anderson, G. I. (1999). Hephaestin, a ceruloplasmin homologue implicated in intestinal iron transport, is defective in the sla mouse. Nature Genetics, 21(2), 195–199. https://doi.org/10.1038/5979 Walker, M. J., Barnett, T. C., McArthur, J. D., Cole, J. N., Gillen, C. M., Henningham, A., … Nizet, V. (2014). Disease manifestations and pathogenic mechanisms of group A Streptococcus. Clinical Microbiology Reviews, 27(2), 264–301. https://doi.org/10.1128/CMR.00101-13 Walker, M. J., Hollands, A., Sanderson-Smith, M. L., Cole, J. N., Kirk, J. K., Henningham, A., … Nizet, V. (2007). DNase Sda1 provides selection pressure for a switch to invasive group A streptococcal infection. Nature Medicine, 13(8), 981–985. https://doi.org/10.1038/nm1612 Wallace, D. F., Summerville, L., Crampton, E. M., Frazer, D. M., Anderson, G. J., & Subramaniam, V. N. (2009). Combined deletion of Hfe and transferrin receptor 2 in mice leads to marked dysregulation of hepcidin and iron overload. Hepatology, 50(6), 1992– 2000. https://doi.org/10.1002/hep.23198 Wang, R.-H., Li, C., Xu, X., Zheng, Y., Xiao, C., Zerfas, P., … Deng, C.-X. (2005). A role of SMAD4 in iron metabolism through the positive regulation of hepcidin expression. Cell Metabolism, 2. https://doi.org/10.1016/j.cmet.2005.10.010 Watt, F. M., & Fujiwara, H. (2014). Cell-Extracellular Matrix Interactions in Normal and Diseased Skin Cell-Extracellular Matrix Interactions in Normal and Diseased Skin, (December), 1–15. https://doi.org/10.1101/cshperspect.a005124

156

BIBLIOGRAPHY

Weinberg, E. D. (1975). Nutritional Immunity: Host’s Attempt to Withhold Iron From Microbial Invaders. JAMA: The Journal of the American Medical Association, 231(1), 39–41. https://doi.org/10.1001/jama.1975.03240130021018 White, C., Yuan, X., Schmidt, P. J., Bresciani, E., Samuel, T. K., Campagna, D., … Hamza, I. (2013). HRG1 is essential for heme transport from the phagolysosome of macrophages during erythrophagocytosis. Cell Metabolism, 17(2), 261–270. https://doi.org/10.1016/j.cmet.2013.01.005 Wrighting, D. M., & Andrews, N. C. (2006). Interleukin-6 induces hepcidin expression through STAT3. Blood, 108(9), 3204–3209. https://doi.org/10.1182/blood-2006-06- 027631 Yang, D., Chertov, O., & Oppenheim, J. J. (2001). Participation of mammalian defensins and cathelicidins in anti-microbial immunity: receptors and activities of human defensins and cathelicidin (LL-37). Journal of Leukocyte Biology, 69(5), 691–697. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11358975 Yap, A. S., Brieher, W. M., & Gumbiner, B. M. (1997). Molecular and Functional Analysis of Cadherin-Based Adherens Junctions. Annual Review of Cell and Developmental Biology, 13(1), 119–146. https://doi.org/10.1146/annurev.cellbio.13.1.119 Yoshimura, T., Matsushima, K., Tanaka, S., Robinson, E. a, Appella, E., Oppenheim, J. J., & Leonard, E. J. (1987). Purification of a human monocyte-derived neutrophil chemotactic factor that has peptide sequence similarity to other host defense cytokines. Proceedings of the National Academy of Sciences of the United States of America, 84(24), 9233–9237. https://doi.org/10.1073/pnas.84.24.9233 Yoshizaki, K., Nishimoto, N., Matsumoto, K., Tagoh, H., Taga, T., Deguchi, Y., … et al. (1990). Interleukin 6 and expression of its receptor on epidermal keratinocytes. Cytokine, 2(5), 381–387. https://doi.org/10.1016/1043-4666(90)90069-6 Zhang, L., Gallo, R. L., Cullen, T. W., Schofield, W. B., Barry, N. A., Putnam, E. E., … Zasloff, M. (2016). Antimicrobial peptides. Current Biology : CB, 26(1), R14-9. https://doi.org/10.1016/j.cub.2015.11.017 Zinkernagel, A. S., Timmer, A. M., Pence, M. A., Locke, J. B., Buchanan, J. T., Turner, C. E., … Nizet, V. (2008). The IL-8 Protease SpyCEP/ScpC of Group A Streptococcus Promotes Resistance to Neutrophil Killing. Cell Host and Microbe, 4(2), 170–178. https://doi.org/10.1016/j.chom.2008.07.002 Zumerle, S., R Mathieu, J. R., ephanie Delga, S., ene Heinis, M., Viatte, L., Vaulont, S., & Peyssonnaux, C. (2014). Targeted disruption of hepcidin in the liver recapitulates the hemochromatotic phenotype. Blood, 123(23). https://doi.org/10.1182/blood-2014-01

157

ANNEXES

1. Article in preparation: Epidermal hepcidin prevents systemic bacterial spreading in a Group A streptococcus-induced necrotizing soft tissue infection model

2. Brevet PCT/FR2018/051133: Methods for treating gram positive bacterial infection

158

ANNEXES

Epidermal hepcidin prevents systemic bacterial spreading in a Group A streptococcus-induced necrotizing soft tissue infection model

Mariangela Malerba1,2,3,4, Sabine Louis1,2,3,4, Sylvain Cuvellier1,2,3,4, Andrea Tedeschi1,2,3,4, Agnès

Fouet1,2,3, Annelies Zinkernagel5, Camille Hua6, Camille Gomart6, Nicolas Ortonne7, Claire Poyart1,2,3,

Jean-Winoc Decousser6, Jacques RR Mathieu1,2,3,4, Carole Peyssonnaux1,2,3,4*

1 INSERM, U1016, Institut Cochin, Paris, France

2 CNRS, UMR8104, Paris, France

3 Université Paris Descartes, Sorbonne Paris Cité, Paris, France

4 Laboratory of Excellence GR-Ex, Paris, France

5 Division of Infectious Diseases and Hospital Epidemiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland.

6 Assistance Publique-Hôpitaux de Paris, Hôpitaux Universitaires Henri Mondor, Laboratoire de Bactériologie Hygiène et Equipe Opérationnelle d’Hygiène

EA 7380 Dynamyc Université Paris-Est Créteil (UPEC), Ecole nationale vétérinaire d'Alfort (EnvA), Faculté de Médecine de Créteil, Créteil, FRANCE.

7Département de pathologie, hôpital Henri-Mondor, Créteil, France.

* Corresponding author: Carole Peyssonnaux, PhD Institut Cochin, Department Endocrinology Metabolism and Diabetes INSERM U1016, CNRS UMR8104 24 rue du Faubourg Saint Jacques 75014 Paris FRANCE e-mail : [email protected] Phone : (33)1 44 41 24 71

Fax : (33)1 44 41 24 21

Abstract

159

ANNEXES

Necrotizing fasciitis (NF) is a life threatening destructive soft tissue infection. Despite systemic antibiotics and radical debridement of necrotic tissue lethality remains high. We report here that mice lacking hepcidin in keratinocytes failed to restrict bacterial systemic spreading in a GAS- induced necrotizing soft tissue infection model, due to a defect in CXCL1 production. Accordingly, injection of hepcidin prevented bacterial systemic spreading and is a potential NF therapeutic.

NF, also known as gangrene, was first described by Hippocrates in the 5th century BC 1. NF is characterized by widespread necrosis of the skin, subcutaneous tissues and fascia. The standard treatment of NF consists of broad-spectrum antibiotics, wide surgical debridement, and supportive care.

Even with current state of the art treatment, NF is associated with a fulminant course and high mortality rate, ranging from 25% to 35% in recent series 2. Group A Streptococcus (GAS) is considered the most common cause of necrotizing fasciitis and is associated with bacteremia and shock. The incidence of invasive GAS disease is increasing worldwide3. Upon detection of these Gram-positive pyogenic bacteria, the immune system launches a complex response which critically depends on the recruitment and activation of neutrophils 4.

Hepcidin is a 25 amino acid antimicrobial peptide (AMP), mainly produced by the liver, demonstrated to be the key iron regulatory hormone in the organism 5. Hepcidin limits iron fluxes to the bloodstream by promoting degradation of the iron exporter ferroportin in target cells 6. Although hepcidin is mainly produced by the liver, an autocrine/paracrine role of hepcidin in other organs such as the lung and kidney has been suggested during infection 7,8. The putative expression and local role of hepcidin in the skin, major site of AMP production, has never been investigated.

Hepcidin immunoreactivity was detected in the skin of patients suffering from streptococcal NF

(detailed in Supplementary table 1), especially in keratinocytes (Fig. 1a). To investigate its role in the etiology of NF, we used an established animal infection model of GAS induced necrotizing soft tissue infection 9,10, where the inoculum is introduced subcutaneously into a shaved area on the flank of mice.

160

ANNEXES

In skin biopsies of infected lesions of WT mice, hepcidin was not only expressed in myeloid cells recruited to the site of infection, as previously reported 11 but also in keratinocytes (Fig. 1b).

To probe the functional significance of keratinocyte-derived hepcidin in vivo, we developed a novel mouse model of keratinocyte-specific hepcidin deficiency (Hamp1ker) by crossing Hamp1lox/lox mice with K14cre+ mice (Fig. 1c). We observed an efficient truncation of the floxed hamp1 allele in the epidermis of the Hamp1ker mice (Supplementary Fig. 1). Iron parameters were unchanged between

Hamp1lox/lox and Hamp1ker mice (Supplementary Fig. 2), confirming our previous study 12 showing that that extra-hepatic tissues do not contribute to systemic iron homeostasis. Hamp1lox/lox and Hamp1ker mice were infected with GAS resulting in a necrotizing soft tissue infection. Four days post-infection,

Hamp1ker mice had a significantly higher number of bacteria than the Hamplox/lox littermates at the lesion site (6.107 vs 9. 105 CFU/mg) but also in the blood (104 vs 9.102 CFU/ml) and in peripheral organs such as the spleen (9.102 vs 20 CFU/mg) (Fig. 1d). Hamp1ker mice also presented with a greater weight loss than the Hamplox/lox mice further underlining the increased morbidity in these mice. These data indicate that keratinocyte production of hepcidin is important in limiting the ability of GAS to replicate within the necrotic skin tissues and to disseminate from the initial focus of infection into the bloodstream and systemic organs. We next asked whether local injection of hepcidin could, accordingly, prevent the systemic spread of bacteria. Twenty four hours after GAS infection, 1 g of synthetic hepcidin/PBS was subcutaneously injected at the inoculation site, followed by two injections of 500 ng of hepcidin for 2 consecutive days (Fig. 1e). In contrast to the mock-injected mice, that exhibited systemic signs of infection including weight loss (Fig. 1e), rough hair coat and hunched posture (data not shown), hepcidin-treated mice did not present any signs of systemic disease and accordingly did not loose weight.

Remarkably, whereas all the control mice presented with systemic bacterial dissemination (as shown by the number of bacteria in the spleen), 7 out of 9 hepcidin-treated mice showed absolutely no dissemination (Fig. 1e).

161

ANNEXES

To investigate the mechanisms by which epidermal hepcidin restrains the infection, we first determined the bacteriostatic and bactericidal effects of hepcidin against GAS. While the antimicrobial peptide LL-

37 presented with both bacteriostatic and bactericidal activity against GAS, hepcidin had no effect

(Supplementary Figs. 3a and 3b). Moreover, primary keratinocytes derived from WT and Hamp1ker mice displayed the same bactericidal activity against this pathogen (Supplementary Fig. 3c).

We next asked whether hepcidin could have a direct immunomodulatory role on keratinocytes. A cytoplex was performed on murine primary keratinocytes incubated with 0.36 M and 3.6 M hepcidin resulting in an unexpectedly induced a dose dependent increase of the key neutrophil chemokine CXCL1

(Fig. 2a) but not of other tested inflammatory cytokines (Supplementary Fig. 4). Hepcidin did not activate CXCL1 gene transcription (Fig. 2b). Cycloheximide, but not actinomycin D, significantly reduced hepcidin-induced CXCL1 production, suggesting a possible role for hepcidin on CXCL1 mRNA translation (Fig. 2c). CXCL1 translational regulation by hepcidin could provide several advantages for the host, including rapid onset of the response. In corroboration with the in vitro results, the expression of CXCL1 was decreased in the keratinocytes of GAS infected-Hamp1ker mice compared to WT littermates, showing that keratinocyte-derived hepcidin contributes to CXCL1 production in vivo (Fig. 2d). As a consequence of the decrease in CXCL1 production, a significant decrease in neutrophil recruitment (Fig. 2e) was observed in the skin of Hamp1ker mice as compared to the control littermates. This defect in keratinocyte-derived hepcidin ability to recruit neutrophils at the site of infection was translated in a decrease in the necrotic skin lesion size of the Hamp1ker mice as compared to controls (Fig. 2f). Importantly, subcutaneous injection of CXCL1 to Hamp1ker mice decreased the number of bacteria at the level of Hamp1lox/lox mice, confirming that the lack of CXCL1 production was responsible for the susceptibility of mutant mice to infection in the Hamp1ker mice

(Fig. 2g). In conclusion, our results show that hepcidin is unexpectedly critical to regulate CXCL1 production in the keratinocytes and tune the magnitude of the neutrophil response. Importantly, we show

162

ANNEXES

here a putative curative role of hepcidin in NF, where there is still today no treatment to avoid dissemination of the bacteria.

Figure legends

Figure 1 Hepcidin expression in epidermis restricts GAS systemic spread

(a-b) Immunohistochemistry with or without primary antibody detecting hepcidin on sections of cutaneous human biospies of group A streptococal Necrotizing Fascitiis (NF) patients (a) and on WT mice challenged or not with GAS (b). PerkinElmer's LaminaTM multilabel slide scanner Panoramic

Viewer software. (c) Generation of Hamp1ker mice. (d) Bacterial count in skin, blood and spleen of

Hamp1lox/lox and Hamp1ker mice four days after injection with GAS. N=10 per group. (e) Therapeutic protocol (left). Weight variation in infected mice treated with 3 subcutaneous injection of hepcidin (n=9, red circle) or PBS (n=9, black square) during 96 hours. GAS dissemination in the spleen of GAS infected mice 96 hours after infection (right).

All values are shown as mean ± s.e.m. Statistical analysis was performed using unpaired Student’s t-test except for Fig 2e: a two-way analysis of variance (ANOVA). * p<0.05. ** p<0.01.

Figure 2 Hepcidin is required for CXCL1 production and neutrophils recruitment

(a) CXCL1 levels measured by ELISA in the culture supernatant of murine primary keratinocytes stimulated for 1 or 3 hours with hepcidin or PBS. N=3 per group. Representative of 3 independent experiments (b) Expression by Q-PCR of CXCL1 gene in WT primary keratinocyte incubated for 1 and

3 hours with 360 nM, 3.6 M hepcidin or PBS. N=3 per group. Representative of 3 independent experiments (c) CXCL1 levels in the culture supernatant of murine primary keratinocytes stimulated for with 3.6 M hepcidin in the presence of solvent / actinomycin / cycloheximide. N=3 per group.

Representative of 3 independent experiments. Immunohistochemistry using (d) anti-CXCL1 or (e) anti

163

ANNEXES

polimorphonulcear leucocytes (PMN) antibodies on skin of WT and Hamp1ker mice challenged with

GAS. (f) Area of necrotic ulcers in skin of WT and Hamp1ker mice four days after GAS injection. (g)

Bacterial count in the skin of WT and Hamp1ker mice injected daily with CXCL1 or PBS, four days after injection with GAS. All values are shown as mean ± s.e.m. Statistical analysis was performed using unpaired Student’s t-test except for Fig 2f : a two-way analysis of variance (ANOVA). * p<0.05. ** p<0.01. *** p<0.001.

REFERENCES

1. Descamps, V., Aitken, J. & Lee, M.G. Hippocrates on necrotising fasciitis. Lancet 344, 556 (1994). 2. Hakkarainen, T.W., Kopari, N.M., Pham, T.N. & Evans, H.L. Necrotizing soft tissue infections: review and current concepts in treatment, systems of care, and outcomes. Curr Probl Surg 51, 344-362 (2014). 3. Sims Sanyahumbi, A., Colquhoun, S., Wyber, R. & Carapetis, J.R. Global Disease Burden of Group A Streptococcus. in Streptococcus pyogenes : Basic Biology to Clinical Manifestations (eds. Ferretti, J.J., Stevens, D.L. & Fischetti, V.A.) (Oklahoma City (OK), 2016). 4. Walker, M.J., et al. Disease manifestations and pathogenic mechanisms of Group A Streptococcus. Clin Microbiol Rev 27, 264-301 (2014). 5. Nicolas, G., et al. Lack of hepcidin gene expression and severe tissue iron overload in upstream stimulatory factor 2 (USF2) knockout mice. Proc Natl Acad Sci U S A 98, 8780-8785 (2001). 6. Nemeth, E., et al. Hepcidin regulates cellular iron efflux by binding to ferroportin and inducing its internalization. Science 306, 2090-2093 (2004). 7. Chen, Q.X., et al. Silencing airway epithelial cell-derived hepcidin exacerbates sepsis induced acute lung injury. Crit Care 18, 470 (2014). 8. Houamel, D., et al. Hepcidin as a Major Component of Renal Antibacterial Defenses against Uropathogenic Escherichia coli. J Am Soc Nephrol 27, 835-846 (2016). 9. Datta, V., et al. Mutational analysis of the group A streptococcal operon encoding streptolysin S and its virulence role in invasive infection. Mol Microbiol 56, 681- 695 (2005). 10. Peyssonnaux, C., et al. HIF-1alpha expression regulates the bactericidal capacity of phagocytes. J Clin Invest 115, 1806-1815 (2005). 11. Peyssonnaux, C., et al. TLR4-dependent hepcidin expression by myeloid cells in response to bacterial pathogens. Blood 107, 3727-3732 (2006). 12. Zumerle, S., et al. Targeted disruption of hepcidin in the liver recapitulates the hemochromatotic phenotype. Blood 123, 3646-3650 (2014). 13. Jonkers, J., et al. Synergistic tumor suppressor activity of BRCA2 and p53 in a conditional mouse model for breast cancer. Nat Genet 29, 418-425 (2001).

164

ANNEXES

14. Lichti, U., Anders, J. & Yuspa, S.H. Isolation and short-term culture of primary keratinocytes, hair follicle populations and dermal cells from newborn mice and keratinocytes from adult mice for in vitro analysis and for grafting to immunodeficient mice. Nat Protoc 3, 799-810 (2008). 15. Nizet, V., et al. Innate antimicrobial peptide protects the skin from invasive bacterial infection. Nature 414, 454-457 (2001).

Methods

Mice

Hamp1lox/lox mice 12 were bred with K14-Cre transgenic mice 13, in which the Cre recombinase is under the control of the murine K14 promoter. Studies were performed in a C57BL6 background, using 8-12 week old male littermates. The animal studies described here were reviewed and approved (Agreement n° CEEA34.CP.003.13) by the “Président du Comité d'Ethique pour l'Expérimentation Animale Paris

Descartes” and are in accordance with the principles and guidelines established by the European

Convention for the Protection of Laboratory Animals (Council of Europe, ETS 123, 1991).

Mouse model of GAS infection

A well-established model of necrotizing soft tissue infection was adapted for this study 9,10. In brief, 100

μl of a mid logarithmic growth phase (approximately 107 c.f.u.) of M1 GAS was mixed with an equal volume of sterile Cytodex beads (Sigma-Aldrich) and injected subcutaneously into the flank of

Hamp1lox/lox and Hamp1ker mice male littermates, previously shaved followed by treatment with hair removal cream. Weight loss and lesion size of mice were daily estimated. After 96 hours, mice were sacrificed, skin necrotic ulcers, spleen, and blood (via retro orbital bleeding) were recovered and

165

ANNEXES

homogenized in 1:2 mg/μl of sterile PBS. Serial dilutions of tissue homogenate were stretched on THA plates, in order to enumerate c.f.u.

For the rescue experiment, WT and Hamp1ker littermates were subcutaneously infected with 107 c.f.u. of GAS with PBS or 1 μg/mouse of rCXCL1 (R&D system, 1395-KC-025/CF). Mice were intralesionally injected again at 24, 48 and 72 h with PBS or 1 μg of rCXCL1.

For the therapeutic treatment, C57BL6 mice were infected as described above. One day after infection, animals were subcutaneously injected with PBS (control) or 1μg of hepcidin (peptide international).

Mice were treated again at 48 and 72 hours with 500 ng of hepc-25. After 96 hours mice were sacrificed, spleens were taken, homogenized for c.f.u. enumeration.

NF patients

Informed consent to the protocol was obtained for all subjects and was approved by the Institutional

Review Board and the regional ethics committee Paris IV (IRB 2016/40NICB), n° IRB 00003835. The collection of personal data was approved by the Commission Nationale de l'Informatique et des Libertés

(CNIL, https://www.cnil.fr).

Immunohistochemistries

Tissues were fixed in 10% formaldehyde and embedded in paraffin. Sections were cut at a thickness of

3 m. Following dewaxing and rehydratation, sections were subjected to antigen retrieval in a pressure cooker for 15 min at 95°C with citrate buffer and permeabilization with 0,5% Triton in phosphate buffered saline for 30 minutes at room temperature, then incubated overnight at 4°C with rabbit polyclonal anti-CXCL1 primary antibody (1/100, NBP1-51188 Novus). A solution of protease K diluted at 1/30 in phosphate buffered saline for 5 min was necessary for unmasking sections before incubating with rat monoclonal anti-PMN primary antibody (1/200, sc-71674 Santa Cruz).

166

ANNEXES

Sections were incubated with HRP labeled anti-rabbit antibody (A6154, Sigma) or anti-rat antibody

(ab97057, Abcam) for 1 hour before washing and substrate development according to the manufacturer’s instructions (Impact Novared SK48-05, Vector). Sections were counterstained with Mayer haemalun.

Human tissues were stained with a rabbit polyclonal to Hepcidin-25 (1/50, Ab30760 Abcam) and mouse tissues with a rabbit polyclonal anti mouse Hepcidin (1/100, Diagnostic, HEPC11-A Alpha). Tissue sections were stained with the ABC-peroxydase technique.

Isolation and treatment of murine primary keratinocytes.

Keratinocytes were isolated from the skin of 2-day-old WT or Hamp1ker mice, based on modifications from a previously described protocol. 14. In brief, dermis were separated from epidermis after floating newborn mice skin on freshly thawed cold trypsin 0,25% (thermo fisher 15050065) overnight at 4°C with the dermis side down. Cells were isolated in stop trypsin medium (DMEM, high glucose, no glutamine, no calcium (thermo fisher 21068028) containing 20% heat inactivated and Chelex (Bio rad

1421253) treated- FBS, 1% GlutaMAXSupplement (thermo fisher 35050061), 1% Sodium Pyruvate

(thermo fisher 11360070), 100U/ml Penicillin, 100 g/ml Streptomycin (thermo fisher), 2,5 g/ml

Amphotericin B (Gibco)).

Keratinocytes were obtained by epidermis mechanical dissociation through shaking at 350 rpm for 1 hour at 37 °C. The resulting cell suspension was filtered through 100 m filter, centrifuged at 250 g for

10 min at 4°C, plated on fibronectin and bovine collagen I coated tissue culture plate and incubated at

37°C in a humidified 5% CO2 atmosphere. The cells were cultured in 0.07 mM calcium medium (Low

Ca2+) containing EGF (10 ng/ml) and cholera toxin (10-10mol/L). Once confluence reached, cells were switched in 1.2 mM calcium medium (High Ca2+) in order to ensure their differentiation. After three days in High Ca2+ cells were differentiated and suitable for different experiments.

Reverse transcription and real-time quantitative PCR

RNA extraction, reverse transcription and quantitative PCR have been performed as previously described 12. All samples were normalized to the threshold cycle value for cyclophilin-A.

167

ANNEXES

Quantification of cytokines

Keratinocytes, previously starved of FBS for 1 hour, were treated with 1 or 10 g/ml hepcidin (Peptide

International) and incubated at different time points. In some experiments, inhibitors (Brefaldin A

(sigma B7651), cycloheximide (sigma C7698), actinomycin D (sigma A1410)) were added 30 minutes before hepcidin treatment during FBS starvations. The cells were lysed with Tri reagent (sigma) while supernatant cytokines were measured with the V-PLEX Proinflammatory Panel1 kit (Meso Scale

Discovery), according to the manufacturer’s instruction.

Antimicrobial assays

As previously described, 15 GAS M1T15448 were grown in Todd Hewitt Broth (THB, Sigma) plus 0,3% yeast extract to early log phase (absorbance at 600 nm, OD600=0,2). Bacteria were then diluted in THB

5 -1 to OD600=0,001 (≈2x10 colony forming units (c.f.u.). ml ). This bacterial suspension was grown for

16h at 37°C in presence of PBS, LL-37, hepcidin (peptide international) or penicillin G (sigma P7794) and the OD600 followed. For antimicrobial killing kinetics, exponential phase GAS (A600=0,4) was exposed to PBS, LL-37 or hepcidin and incubated at 37°C. At different times point serial dilutions of these bacterial suspensions were plated on THA plate and the number of c.f.u. determined after 24h.

Keratinocyte killing assay

Keratinocytes from Hamp1lox/lox and Hamp1ker newborn mice were isolated and cultured as described above. Their antimicrobial function was tested with exponential phase GAS culture. Cells were previously antibiotic starved for 2 hours and infected with a multiplicity of infection (MOI) of 10 bacteria per cell. GAS were centrifuged onto keratinocytes at 350 g for 10 min, in order to ensure bacteria-cell interaction. Infected cells were incubated at 37°C/5% CO2 for different time points. At each time point, well contents were harvested and spread on THA plates for enumerating surviving c.f.u.

168

ANNEXES

ACKNOWLEDGEMENTS

This work was also supported by a funding from the European Research Council under the European

Community’s Seventh Framework Program (FP7/2011-2015 Grant agreement no 261296), the

“Fondation pour la Recherche Médicale” (DEQ20160334903), the Laboratory of Excellence GR-Ex, reference ANR-11-LABX-0051, funded by the program “Investissements d’avenir” of the French

National Research Agency, reference ANR-11-IDEX-0005-02.

We thank the Henri Mondor Hospital Necrotizing Fasciitis Group (Sbidian Emilie, Bosc Romain,

Chosidow Olivier, de Prost Nicolas, Tomberli Francoise, Decousser Jean Winoc, Woerther Paul Louis,

Gomart Camille, Lepeule Raphael, Luciani Alain, Nakad Lionel, de Angelis Nicola, Champy Cecile),

Antonin Weckel and Marthe Rizk for hepflul technical advices, and Sophie Vaulont for discussions and critical reading of the manuscript.

AUTHOR CONTRIBUTIONS

M.M., S.L., A.T., S.C., designed the experiments, carried out experiments and performed data analysis.

A.F. AZ, C.H., C.G., C.P. and J-W. D. provided essential reagents and scientific advice. J.M.RR and

C.P. designed experiments and supervised the project. C.P. wrote the manuscript.

COMPETING FINANCIAL INTERESTS

The authors declare no competing financial interests.

169

ANNEXES

170

ANNEXES

171

ANNEXES

172

ANNEXES

173

ANNEXES

174

ANNEXES

175

ANNEXES

Methods for treating gram positive bacterial infection

FIELD OF THE INVENTION

The present invention relates to method for treating gram-positive bacterial infection with mature hepcidin. More specifically, it concerns method for treating with mature form of hepcidin gram- positive bacterial infection such as Group A streptococcus infection associated with Necrotizing fasciitis (NF).

BACKGROUND OF THE INVENTION

Necrotizing fasciitis (NF), also known as gangrene, was first described by Hippocrates in the 5th century BC 1. NF is characterized by widespread necrosis of the skin, subcutaneous tissues, and fascia. Even with optimal treatment, NF is associated with a fulminant course and high mortality rate, ranging from 25% to 35% in recent series 2. Group A streptococcus (GAS), considered as the most common cause of necrotizing fasciitis, is associated with bacteremia and shock. The incidence of invasive S.pyogenes disease is increasing worldwide3. Upon detection of these Gram-positive pyogenic bacteria, the immune system launches a complex response which critically depends on the recruitment and activation of neutrophils4.

Hepcidin is the key iron regulatory hormone is mainly produced by the hepatocytes. Inventors previously demonstrated that hepatic hepcidin is sufficient to ensure systemic iron homeostasis in physiological conditions suggesting that production of hepcidin by extra-hepatic tissues may have local roles 5. Hepcidin first described as an antimicrobial peptide (AMP), but its putative expression and role in the infected skin, major source of AMPs have never been investigated.

Several recent studies (such Michels KR et al JCI Insight. 2017;2(6):e92002 and Stefanova D. et al Blood. 2017 Jul 20;130(3):245-257.) show that hepcidin analogs (minihepcidins) can have an beneficial effect on Gram negative infection by limiting extracellular iron availability. But property of mature form of hepcidin on Gram-positive infection has never been described.

SUMMARY OF THE INVENTION

Hepatic hepcidin is known as being the key hormone in iron homeostasis; it is able to decrease plasma iron levels by blocking iron absorption in the duodenum and iron release from macrophages thus targeting the two entrance gates for iron in the circulation. Several stimuli have been shown to be involved in hepcidin regulation: iron, hypoxia, erythropoietic demand and inflammation.

176

ANNEXES

177

ANNEXES

The present invention relies on the discovery that, surprisingly, when it is produced locally in the epidermis, mature form of hepcidin (hepc-25) is able to directly initiate the recruitment of neutrophils by increasing the CXCL1 production in keratinocytes. While hepcidin had no direct antimicrobial activity against GAS, injection of hepcidin, at the site of infection prevented GAS systemic spread. Hepcidin agonists may represent a novel therapeutic to prevent life-threatening bacteremia not only in streptococcal NF but also in complicated infections due compromised host immunity. Accordingly, the present invention relates to a method of treating gram-positive bacterial infection in a patient in need thereof, comprising administering to the patient a polypeptide comprising a sequence of 20 amino acids having - at least 50% identity or at least 60% identity with the sequence SEQ ID NO: 1

- cysteine residues at positions 2, 5, 6, 8, 9, 14, 17, and 18 or of a nucleic acid encoding said polypeptide

More specifically, it concerns the discovery of the unexpected property of the hepcidin on CXCL1 production in keratinocytes, useful for the treatment of gram-positive bacterial infections and also useful to prevent (or treat) harmful consequence of such gram-positive bacterial infections, which could be associated with necrotizing fasciitis (NF).

DETAILED DESCRIPTION OF THE INVENTION

In the present study, inventors demonstrate that expression of hepcidin was unexpectedly increased in the skin of streptococcal NF patients compared to healthy controls. To address its role, mice lacking hepcidin in keratinocytes (Hepc kerat) were submitted to a GAS-induced necrotizing soft tissue infection model. Hepc kerat mice failed to restrict systemic spread and showed a defect in neutrophil recruitment at the initial focus, as well as a decrease in the expression of CXCL1, a key neutrophil chemokine in the response to GAS. While hepcidin (hepc-25) had no direct antimicrobial activity against GAS, it unexpectedly increased the translation of CXCL1 in keratinocytes. Finally, injection of hepcidin, at the site of infection prevented GAS systemic spread.

The present invention relates to a method of treating gram-positive bacterial infection in a patient in need thereof, comprising administering to the patient a polypeptide comprising a sequence of 20 amino acids having - at least 50% identity or at least 60% identity with the sequence SEQ ID NO: 1

- cysteine residues at positions 2, 5, 6, 8, 9, 14, 17, and 18

178

ANNEXES

179

ANNEXES

or a nucleic acid encoding said polypeptide.

In particular embodiment of the invention, the Gram-positive bacteria infection is a skin Gram- positive bacteria infection. Hepcidin was first identified as a liver-derived antimicrobial peptide. The human hepcidin gene encodes an 84-residue prepropeptide that contains a 24-residue N-terminal signal peptide that is subsequently cleaved to produce pro-hepcidin. Pro-hepcidin is then processed to produce a mature 25-amino acid hepcidin (Jordan et al., The Journal of Biological chemistry, 2009,284, 36, 24155- 24167). It is now well established that hepatic hepcidin is the regulator of iron homeostasis (as disclosed by Lesbordes-Brion et al., Blood, 2006, 108, 1402–1405 and Nicolas et al., Proc. Natl. Acad. Sci. U.S.A.,2002, 99, 4596-4601). Hepcidin is a hypoferremic hormone; it binds to and degrades ferroportin (FPN, the only iron exporter known to date), thereby decreasing the iron levels in the circulation. Since its expression can be induced by inflammation (Nicolas et al., J. Clin. Invest., 2002, 110, 1037–1044), hepcidin has been proposed as an important mediator of Anemia of Chronic disease (ACD), also known as Anemia of Inflammation (AI).

Preferred polypeptides or nucleic acids for the method according to the invention are the mature forms of human hepcidin, represented for instance by a polypeptide of 20 amino-acids having the sequence Ile Cys Ile Phe Cys Cys Gly Cys Cys His Arg Ser Lys Cys Gly Met Cys Cys Lys Thr (SEQ ID NO: 1), or by a polypeptide of 22 amino-acids having the sequence:

Phe Pro Ile Cys Ile Phe Cys Cys Gly Cys Cys His Arg Ser Lys Cys Gly Met Cys Cys Lys Thr (SEQ ID NO: 2). or by a polypeptide of 25 amino-acids having the sequence:

Asp Thr His Phe Pro Ile Cys Ile Phe Cys Cys Gly Cys Cys His Arg Ser Lys Cys Gly Met Cys Cys Lys Thr (SEQ ID NO: 3), or nucleic acids encoding said polypeptides

The term “hepcidin” should be understood broadly, it encompasses the mature forms of hepidin, variants and fragments thereof having same biological activity: increasing the CXCL1 production in keratinocytes. Biological activity of mature form of hepcidin (CXCL1 production by keratinocyte) can be measured for example by supernatant cytokines quantification using Elisa (see experimental

180

ANNEXES

section and figure 2). or with the V-PLEX Proinflammatory Panel1 kit Meso Scale Discovery). Precursors of said mature forms of hepcidin, i.e. prohepcidin and preprohepcidin and nucleic acids encoding said precursors can also be used. Other examples of polypeptides or nucleic acids suitable for use according to the invention are vertebrate, preferably mammalian, homologous of mature forms of human hepcidin or precursors thereof, or nucleic acids encoding said polypeptides. Known vertebrate homologous of human hepcidin include for instance rat hepcidin, mouse hepcidin, trout hepcidin. Preferably, if the patient is human, the hepcidin is a mature form of human hepcidin (UniProtKB - P81172). The protein sequence of said human hepcidin, may be found in NCBI database with the following access numbers: mRNA NM_ NM_021175, and protein_id: NP_ NP_066998 (hepcidin preproprotein)

Chimeric polypeptides, comprising the sequence of a mature form of hepcidin, can also be used. The invention also encompasses the use of functional equivalents of the above-defined polypeptides. Functional equivalents are herein defined as peptide variants, having the same functional biological activity as the mature forms of hepcidin. All these polypeptides and nucleic acids can be obtained by classical methods known in themselves. For instance, the 20 amino-acids and 25 amino-acids forms of hepcidin can be obtained from plasma or from urine, as disclosed by KRAUSE et al. ( FEBS Lett., 2000, vol. 480, 147-150) ; or PARK et al. (J. Biol. Chem., 2001, vol. 276,7806-7810). Alternatively, they can be obtained by culturing cells expressing hepcidin, and recovering said polypeptide from the cell culture. According to a particular embodiment, said cells are host cells transformed by a nucleic acid encoding one of the polypeptides defined above. Methods for producing hepcidin are well known. For example, hepcidin can be obtained by classical sequential procedures for hepcidin purification from total human plasma. Starting from human plasma, ADDO L et al (Int J Hematol. 2016 Jan;103(1):34-43) purified 3 isoform of human hepcidin by using liquid chromatography-tandem mass spectrometry (LC-tandem MS). By using liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS), a highly specific and quantitative serum hepcidin method (isoform 25) was also describe (Lefebvre T et al. Clin Chem Lab Med. 2015 Sep 1;53(10):1557-67 or Abbas et al , Anal Bioanal Chem. 2018 Apr 18. doi: 10.1007/s00216-018-1056-0.). By using surface-enhanced laser-desorption/ionization time-of-flight mass spectrometry (SELDI-TOF MS),

181

ANNEXES

isoform of human hepcidin detection in serum and urine was also describe (Kemna et al., Clin Chem. 2007 Apr;53(4):620-8.). Alternatively, hepcidin can be a recombinant hepcidin as described in Gagliardo et al., FEBS J. 2008 Aug;275(15):3793-803.. Chemical synthesis can also be used, in particular in the case of the peptide derivatives.

A nucleic acid encoding hepcidin can for instance be obtained from a genomic or cDNA library of a vertebrate, using suitable primers able to hybridize selectively with said nucleic acids. It can also be obtained by the classical techniques of polynucleotide synthesis. Typically a variant of hepcidin has at least 50%, preferably, at least 60%, preferably, at least 70%, preferably, at least 80%, preferably, at least 85% more preferably at least 90%, more preferably at least 95% and even more preferably at least 99% identity with human mature form of hepcidin.

Typically, identity may be determined by BLAST or FASTA algorithms.

The term "gram-positive" bacterial infection refers to a local or systemic infection with gram-positive bacteria. Gram-positive bacteria are bacteria that give a positive result in the Gram stain test, which is traditionally used to quickly classify bacteria into two broad categories according to their cell wall. Gram-positive bacteria take up the crystal violet stain used in the test, and then appear to be purple- coloured when seen through a microscope. This is because the thick peptidoglycan layer in the bacterial cell wall retains the stain after it is washed away from the rest of the sample, in the decolorization stage of the test. In the classical sense, six Gram-positive genera are typically pathogenic in humans. Two of these, Streptococcus and Staphylococcus, are cocci (sphere-shaped). The remaining organisms are bacilli (rod-shaped) and can be subdivided based on their ability to form spores. The non-spore formers are Corynebacterium and Listeria (a coccobacillus), whereas Bacillus and Clostridium produce spores (Gladwin, et al (2007). Miami, Florida: MedMaster. pp. 4–5. ISBN 978-0-940780-81-1). In particular embodiment of the invention the Gram-positive bacteria according to the invention are hyperinvasive bacteria The term “hyperinvasive” for Gram-positive bacteria means that bacteria are resistant to innate immune clearance. Invasive bacterial infections occur when the bacteria invade the living tissue especially parts of the body where bacteria are not normally present ( such as the bloodstream, soft tissues like muscle or fat, and the meninges) and evade the immune system of the person who is infected. This may occur when a person has sores or other breaks in the

182

ANNEXES

skin that allow the bacteria to get into the tissue, or when the person’s ability to fight off the infection is decreased because of chronic illness or an illness that affects the immune system. Main example of Invasive Gram positive Bacterial Diseases are Streptococcal Disease, Group A Invasive or Streptococcal TSS Streptococcal Disease, Invasive Group B Streptococcus pneumoniae (pneumococcus). In particular embodiment of the invention, the Gram-positive bacteria according to the invention are selected from the group consisting of Streptococcus, Staphylococcus, Clostridium, Listeria, Baccilus and Corynebacterium. In another particular embodiment of the invention, the Gram-positive bacteria according to the invention is Streptococcus selected from the group consisting of beta-hemolytic (Group A or Group B), alpha hemolytic (pneumonia) or gamma haemolytic (Enterococcus).

In more particular embodiment, the Gram-positive bacteria according to the invention is Group A beta-hemolytic Streptococcus (GAS). In more particular embodiment, the Gram-positive bacteria according to the invention is a clone of hyperinvasive serogroup M1T1 group A beta-hemolytic Streptococcus (GAS). Group A streptococcus (GAS) also called “Streptococcus pyogenes” refers to specific species of Gram- positive bacteria. These bacteria are aerotolerant and an extracellular bacterium, made up of non- motile and non-sporing cocci. As expected with a streptococci, it is clinically important in human illness. It is an infrequent, but usually pathogenic, part of the skin microbiota. It is the predominant species harboring the Lancefield group A antigen, and is often called group A streptococcus (GAS). Group A streptococci when grown on blood agar typically produces small zones of beta-hemolysis, a complete destruction of red blood cells. (A zone size of 2–3 mm is typical.) It is thus also called group A (beta-hemolytic) streptococcus (GABHS), and can make colonies greater than 5 mm in size. An estimated 700 million GAS infections occur worldwide each year. While the overall mortality rate for these infections is 0.1%, over 650,000 of the cases are severe and invasive, and have a mortality rate of 25% (Aziz RK, et al (2010). PLoS ONE. 5 (4): e9798). Early recognition and treatment are critical; diagnostic failure can result in sepsis and death. As described in the experimental data (figure 2F), the treatment with mature form of hepcidin of gram- positive infection, such as Group A streptococcus (GAS) infection allows to prevent (or to treat) harmful consequence of gram positive bacterial infection such as Necrotizing fasciitis (NF).

183

ANNEXES

Accordingly, the invention also provides a method of treating gram-positive bacterial infection associated with Necrotizing fasciitis (NF) in a patient in need thereof with mature form of hepcidin according to the invention. As used herein, the term “patient” denotes a mammal, such as a rodent, a feline, a canine, and a primate. Preferably, a patient according to the invention is a human. In a particular embodiment of the invention, said patient is suffering from gram-positive bacterial infections. In a more particular embodiment of the invention, said patient is suffering from a skin gram-positive bacterial infection. In a specific embodiment, the patient is at risk to develop a gram positive bacterial infections, such as a patient before a medical or surgical intervention, very young and very old subject (infants and seniors), people with chronic or serious illnesses (including diabetes and cancer), and immunocompromised patients.

In a specific embodiment, very young subject (infant), means subject with less than 5, 4, 3, 2, 1 years old. In a specific embodiment, very old subject (senior), means subject more than ,70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80; 81; 82; 83; 84; 85; 86; 87; 88; 89; 90; 91; 92; 93; 94; 95; 96; 97; 98; 99 or 100 years old. The present invention also relates to a method for treating gram-positive infection in a patient at risk to develop such infection, such method involving the step of administering to a patient in need thereof a therapeutically effective amount of mature form of hepcidin. This gram-positive bacterial infection could be associated with Necrotizing fasciitis (NF). In particular embodiment of the invention, the Gram-positive bacterial infection is a skin Gram-positive bacterial infection. By a "therapeutically effective amount" is meant a sufficient amount to be effective, at a reasonable benefit/risk ratio applicable to any medical treatment. It will be understood, however, that the total daily usage will be decided by the attending physician within the scope of sound medical judgment. The specific therapeutically effective dose level for any particular patient in need thereof will depend upon a variety of factors including the age, body weight, general health, sex and diet of the patient, the time of administration, route of administration, the duration of the treatment; drugs used in combination or coincidental with the and like factors well known in the medical arts. For example, it is well known within the skill of the art to start doses of the compound at levels lower than those required to achieve the desired therapeutic effect and to gradually increase the dosage until the desired effect is achieved.

184

ANNEXES

The hepcidin can be administered in a suitable formulation to humans and animals by topical or systemic administration, including oral, rectal, nasal, buccal, ocular, sublingual, transdermal, rectal, topical, vaginal, parenteral (including subcutaneous, intra-arterial, intramuscular, intravenous, intradermal, intrathecal and epidural), intracisternal and intraperitoneal. It will be appreciated that the preferred route may vary with for example the condition of the recipient. In a preferred embodiment hepcidin is administered by parenteral way (including subcutaneous, intra-arterial, intramuscular, intravenous, intradermal, intrathecal and epidural). In the context of the invention, the term "treating" or "treatment", as used herein, means reversing, alleviating, inhibiting the progress of, or preventing the disorder or condition to which such term applies, or reversing, alleviating, inhibiting the progress of, or preventing one or more symptoms of the disorder or condition to which such term applies.

As used herein, the term "treatment" or "treat" refer to both prophylactic or preventive treatment as well as curative or disease modifying treatment, including treatment of subjects at risk of contracting the disease or suspected to have contracted the disease as well as subjects who are ill or have been diagnosed as suffering from a disease or medical condition, and includes suppression of clinical relapse. The treatment may be administered to a subject having a medical disorder or who ultimately may acquire the disorder, in order to prevent, cure, delay the onset of, reduce the severity of, or ameliorate one or more symptoms of a disorder or recurring disorder, or in order to prolong the survival of a subject beyond that expected in the absence of such treatment. By "therapeutic regimen" is meant the pattern of treatment of an illness, e.g., the pattern of dosing used during therapy. A therapeutic regimen may include an induction regimen and a maintenance regimen. The phrase "induction regimen" or "induction period" refers to a therapeutic regimen (or the portion of a therapeutic regimen) that is used for the initial treatment of a disease. The general goal of an induction regimen is to provide a high level of drug to a subject during the initial period of a treatment regimen. An induction regimen may employ (in part or in whole) a "loading regimen", which may include administering a greater dose of the drug than a physician would employ during a maintenance regimen, administering a drug more frequently than a physician would administer the drug during a maintenance regimen, or both. The phrase "maintenance regimen" or "maintenance period" refers to a therapeutic regimen (or the portion of a therapeutic regimen) that is used for the maintenance of a subject during treatment of an illness, e.g., to keep the subject in remission for long periods of time (months or years). A maintenance regimen may employ continuous therapy (e.g., administering a drug at a regular

185

ANNEXES

intervals, e.g., weekly, monthly, yearly, etc.) or intermittent therapy (e.g., interrupted treatment, intermittent treatment, treatment at relapse, or treatment upon achievement of a particular predetermined criteria [e.g., disease manifestation, etc.]) Typically medicaments according to the invention comprise a pharmaceutically-acceptable carrier. A person skilled in the art will be aware of suitable carriers. Suitable formulations for administration by any desired route may be prepared by standard methods, for example by reference to well-known text such as Remington; The Science and Practice of Pharmacy.

FIGURE LEGENDS

Figure 1. (A) Generation of Hepc kerat mice. Recombination of the floxed hamp1 allele by genomic PCR in the epidermis of the Hepc kerat14 and WT mice. (B) Hepdicin deletion in keratinocytes renders mice more susceptible to GAS infection. Bacterial count in skin, blood and spleen of WT vs Hepc kerat14 mice four days after injection with GAS. At least seven mice in each group were tested in three independent experiments. Statistical analysis was performed using unpaired Student’s t-test; *p<0.05. ** p<0.01.

Figure 2 –

(A-C) Hepdicin promotes CXCL1 production at the translational level. CXCL1 levels measured by ELISA in the culture supernatant of murine primary keratinocytes stimulated for

1 or 3 hours with hepcidin vs PBS. (B) Expression by Q-PCR of CXCL1 in WT primary keratinocyte incubated for 1 and 3 hours with 360 nM, 3,6 µM hepcidin or PBS. n>4; mean

+/- s.e.m. (C) CXCL1 levels measured by ELISA in the culture supernatant of murine primary keratinocytes stimulated for with 3,6 µM hepcidin in presence of solvent/ actinomycinD/ cycloheximide/ brefeldinA. Statistical analysis was performed using unpaired Student’s t-test;

* p<0.05. (D) Area of necrotic ulcers in skin of WT and Hepc kerat14 mice four days after injection with GAS. (E) Bacterial count in the skin of WT and Hepc kerat14 mice +/- injected daily with CXCL1 four days after injection with GAS. (F) Local injection of hepcidin prevents systemic spread of infection. Weight variation in infected mice treated with 3 subcutaneous injection of hepcidin (circle) or PBS (square) during 96 hours (left). GAS dissemination in the spleen of GAS-infected mice 96 hours after infection (right). Ten mice in each group were tested in two experiments. Statistical analysis was performed using unpaired Student’s t-test; *p<0.05. ** p<0.01.

Figure 3. Hepdicin has no antimicrobial activity against GAS in vitro and does not influence keratinocyte antimicrobial activity. (A) GAS growth curve in presence of

186

ANNEXES

pennicilin G, LL-37 and hepcidin. (B) GAS killing kinetics of 32 μM of LL-37 and hepcidin.

(C) Bacterial recovering at 2h and 4h following incubation of log-phase GAS with murine primary keratinocytes from WT and Hamp1 ker mice. Data are representative of two independent experiments preformed in triplicate. Statistical analysis was performed using unpaired Student’s t-test; ** p<0.01. Figure 4 Mini-hepdicin does bot promote CXCL1 production in keratinocytes. CXCL1 levels measured by ELISA in the culture supernatant of murine primary keratinocytes stimulated for 1 or 3 hours with mini-hepcidin or PBS.

EXAMPLE

MATERIALS

Mice

Hamp1lox/lox mice 5 were bred with K14-Cre transgenic mice (Jonkers et al., 2001 Nat Genet. 2001;29:418–425), in which the Cre recombinase is under the control of the murine K14 promoter. Studies were performed in a C57BL6/genetic background, using 8-12 week old male littermates.

The animal studies described here were reviewed and approved (Agreement n° CEEA34.CP.003.13) by the “Président du Comité d'Ethique pour l'Expérimentation Animale Paris Descartes” and are in accordance with the principles and guidelines established by the European Convention for the Protection of Laboratory Animals (Council of Europe, ETS 123, 1991).

Mouse model of GAS infection

A well-established model of necrotizing soft tissue infection was adapted for this study (7,9). In brief,

100 μl of a mid logarithmic growth phase (approximately 107 c.f.u.) of GAS was mixed with an equal volume of sterile Cytodex beads (Sigma-Aldrich) and injected subcutaneously into the flank of 8- to 12-week-old WT and Hepc kerat14 male littermates, previously shaved (veet hair removal cream). Weight loss and lesion size of mice were daily estimated. After 96 hours mice were sacrificed, skin necrotic ulcers, spleen, and blood (via retro orbital bleeding) were recovered and homogenized in 1:2 mg/μl of sterile PBS. Serial dilutions of tissue homogenate were stretched on THA plates, in order to enumerate living c.f.u.

For rescue experiment WT and Hepc kerat littermates were subcutaneously infected with 107 c.f.u. of GAS with mock or 1 μg/mouse of rCXCL1 (R&D system, 1395-KC-025/CF). Mice were intralesionally injected again at 24, 48 and 72 h with mock or 1 μg of rCXCL1.

187

ANNEXES

For therapeutic treatment C57BL6 mice were infected as described above. One day after infection, animals were subcutaneously injected with PBS (control) or 1μg of hepcidin (peptide international). Mice were treated again at 48 and 72 hours with 500 ng of hepc-25. After 96 hours mice were sacrificed, spleens were taken, homogenized for c.f.u. enumeration.

Human patients

Informed consent to the protocol was obtained for all subjects. Biopsies were fixed in 10% formalin and embedded in paraffin.

Isolation and treatment of murine primary keratinocytes.

Keratinocytes were isolated from the skin of 2-day-old WT or Hepc kerat mice, based on modifications from a previously described protocol8. In brief, dermis were easily separated from epidermis after floating newborn mice skin on freshly thawed cold trypsin 0,25% (thermo fisher 15050065) overnight at 4°C with the dermis side down. Cells were isolated in stop trypsin medium (DMEM, high glucose, no glutamine, no calcium (thermo fisher 21068028) containing 20% heat inactivated and Chelex (Bio rad 1421253) treated- FBS, 1% GlutaMAXSupplement (thermo fisher 35050061), 1% Sodium Pyruvate (thermo fisher 11360070), 100U/ml Penicillin, 100 g/ml Streptomycin (thermo fisher), 2,5 g/ml Amphotericin B (Gibco)).

Keratinocytes were obtained by epidermis mechanical dissociation through shaking at 350 rpm for 1 hour at 37 °C. The resulting cell suspension was filtered through 100 m filter, centrifuged at 250 g for 10 min at 4°C, plated on fibronectin and bovine collagen I coated tissue culture plate and incubated at 37°C in a humidified 5% CO2 atmosphere. The cells were cultured in 0,07 mM calcium medium (low Ca2+) containing EGF (10 ng/ml) and cholera toxin (10-10mol/L). Once confluence reached, cells were switched in 1,2 mM calcium medium (high Ca2+) in order to assure their differentiation.

After three days in high Ca2+ cells were differentiated and suitable for different experiments.

Reverse transcription and real-time quantitative PCR

RNA extraction, reverse transcription and quantitative PCR have been performed as previously described {Zumerle, 2014 #46}. All samples were normalized to the threshold cycle value for cyclophilin-A.

Quantification of cytokines

Keratinocytes, previously starved of FBS for 1 hour, were treated with 1 or 10 µg/ml hepcidin (Peptide International) and incubated for different time points. In some experiments, inhibitors were added 30 minutes before hepcidin treatment during FBS starving: brefaldin A (sigma B7651), cycloheximide (sigma C7698), actinomycinD (sigma A1410). The cells were

188

ANNEXES

lysed with Tri reagent (sigma) while supernatant cytokines were measured with the V-PLEX Proinflammatory Panel1 kit (Meso Scale Discovery), according to the manufacturer’s instruction.

Antimicrobial assays

As previously described, 9 GAS NZ131were grown in Todd Hewitt Broth (THB, Sigma) plus 0,3% yeast extract to early log phase (absorbance at 600 nm, A600=0,2). Bacteria were then diluted in THB toA600=0,001 (≈2x105 colony forming units (c.f.u.) ml-1). This bacterial suspension was kept growing overnight at 37°C in presence of PBS, LL-37, hepcidin (peptide international) or penicillin G (sigma P7794). For antimicrobial killing kinetics, exponential phase GAS (A600=0,4) was exposed to PBS or LL-37 or hepcidin and incubated at 37°C. At different time point serial dilutions of these bacterial suspensions were plated on THA plate for counting surviving c.f.u.

Keratinocyte killing assay

Keratinocytes from WT and Hepc kerat newborn mice were isolated and cultured as above described. Their antimicrobial function was tested with log-phase GAS culture.

Cells were previously starved of antibiotics for 2 hours then infected with a multiplicity of infection (MOI) of 10 bacteria per cell. GAS were centrifuged onto keratinocytes at 350 g for 10 min, in order to assure bacteria-cell interaction. Infected cells were incubated at 37°C/5% CO2 for different time points. At each time point, well contents were harvested and spread on THA plates for enumerating surviving c.f.u.

RESULTS

We examined hepcidin expression, by immunohistochemical staining, in the skin of healthy subjects and patients suffering from streptococcal necrotizing fasciitis (NF) (detailed in table 1). Hepcidin immunoreactivity was increased in NF suffering patients compared to healthy controls. To investigate the role of hepcidin in the etiology of NF, we chose an established animal infection model of GAS induced necrotizing soft tissue infection6,7. The GAS inoculum was introduced subcutaneously (SC) into a shaved area on the flank of WT mice. In skin biopsies of the infected lesions, hepcidin was not only expressed in myeloid cells recruited to the site of infection, as previously reported10 but was also strongly detected in keratinocytes.

To probe the functional significance of keratinocyte-derived hepcidin in vivo, we developed a novel mouse model of keratinocyte-specific (Hepc kerat) by crossing Hepclox/lox mice with K14cre mice. We observed an efficient truncation of the floxed hamp1 allele in the

189

ANNEXES

epidermis of the Hepc kerat mice, with deletion of the hamp1 allele in the total hepcidin KO mice as a positive control (Fig. 1A). Hepclox/lox and Hepc kerat mice were submitted to the GAS induced necrotizing soft tissue infection model. Quantitative bacterial cultures were performed 4 days post- infection in the skin ulcer, blood, and spleen. Hepc kerat mice had a significantly higher number of bacteria than the WT littermates at the lesion site (6.107 vs 9. 105 CFU/mg) but also in the blood (104 vs 9.102 CFU/ml) and in peripheral organs such as the spleen (9.102 vs 20 CFU/mg) indicating that keratinocyte production of hepcidin was important in limiting the ability of GAS to replicate within the necrotic skin tissues and to disseminate from the initial focus of infection into the bloodstream and systemic organs.

To investigate the mechanisms by which epidermal hepcidin restrains the infection, we first determined the bacteriostatic and bactericidal effects of hepcidin against GAS. The activity of hepcidin was evaluated by its capacity to inhibit GAS growth as compared to LL-

37. Hepcidin had no direct bacteriostatic effect against GAS in vitro, while, at the same concentration, the antimicrobial peptide LL-37 was able to prevent the growth of the bacteria (Fig. 3A). The bactericidal activity was assessed by adding hepcidin to the culture medium at 4 μg/ml. While bactericidal activity corresponded to a significant reduction in bacterial titers after LL-37 exposure as compared to control, addition of hepcidin had no effect (Fig. 3B).

Moreover, primary keratinocytes derived from WT and Hepc kerat mice displayed the same bactericidal activity against this pathogen (Fig. 3C).

We therefore asked whether hepcidin could have a direct immunomodulatory role on keratinocytes. A cytoplex was performed on murine primary keratinocytes, incubated with

360 nM and 3.6 M hepcidin. Hepcidin induced a dose dependent increase of CXCL1 production (Fig. 2A) but not of other tested inflammatory cytokines. CXCL1 levels in the cell culture supernatant reached respectively 150 and 50 pg/ml at 3 hours at the dose of 360 nM and 3.6 M (Fig. 2A). Intriguingly, hepcidin, did not increase CXCL1 expression at the mRNA level. (Fig. 2B). Cycloheximide, but not actinomycin D, significantly reduced hepcidin-induced CXCL1 production, suggesting a possible role for hepcidin on CXCL1 mRNA translation (Fig. 2C). In contrast, a murine mini-hepcidin {Preza, 2011 #180}, consisting in the 9 N-terminal amino acids of hepcidin was not able to trigger CXCL1 production. These data show that, while mini-hepcidin is sufficient to degrade ferroportin and regulate iron homeostasis, it is not sufficient to display this new immunomodulary role (Fig. 4).

190

ANNEXES

In corroboration with the in vitro results, the expression of CXCL1 was decreased in the keratinocytes of GAS infected-Hepc kerat mice compared to WT littermates, showing that keratinocyte-derived hepcidin contributes to CXCL1 production in vivo.

As a consequence of the decrease in CXCL1 production, a significant decrease in neutrophil recruitment was observed in the skin of Hepc kerat mice as compared to the control littermates. This defect in keratinocyte-derived hepcidin to recruit neutrophils at the site of infection was translated in a decrease in the necrotic skin lesion size of the Hepc kerat mice as compared to controls (Fig. 2D). Subcutaneous injection of CXCL1 to Hepc kerat mice decreased the number of bacteria at the level of WT mice, confirming that the lack of CXCL1 production was responsible for the susceptibility of mutant mice to infection.

PMNs fulfill a key role in preventing the transition of GAS soft tissue infections from local to rapidly disseminating life-threatening diseases 4. We asked whether injection of hepcidin could have a therapeutic action in the GAS induced necrotizing soft tissue infection model. 24 hours after GAS infection, 1 g of hepcidin/mock was subcutaneously injected at the inoculation site, followed by two injections of 500 ng of hepcidin for 2 consecutive days (Fig. 2F). At the end of the treatment, the weight was measured as a marker of morbidity. In contrast to mock-injected mice, that exhibited weight loss (Fig. 2F) and signs of morbidity such as rough hair coat, hunched posture (data not shown), hepcidin-treated mice didn’t present any signs of morbidity and did not loose weight. Remarkably, whereas all the control mice presented with systemic bacterial dissemination (as shown by the number of bacteria in the spleen), 7 out of 9 hepcidin-treated mice showed no dissemination (Fig. 2F).

The standard treatment of NF consists of broad-spectrum antibiotics, wide surgical debridement, and supportive care. Even with optimal treatment, NF portend significant morbidity and have high mortality rates 2. Hepcidin agonists may represent a novel therapeutic to prevent life- threatening bacteremia not only in streptococcal NF but also in complicated infections due to compromised host immunity.

191

ANNEXES

Table 1

Patient Age Site Underlying Preceedin Type of Sepsis Outcome No. (y) disease g trauma infection

1 57 Leg None No Group A No Survived Streptococcus

2 51 Leg None Yes Group A Yes Survived Streptococcus

3 80 Leg Prostate No Streptococcus Yes Died adenocarcima dysgalactiae

4 27 Thum None Yes Group A No Survived b Streptococcus

5 72 Leg None No Group A Yes Survived Streptococcus MRSA

REFERENCES

Throughout this application, various references describe the state of the art to which the invention pertains. The disclosures of these references are hereby incorporated by reference into the present disclosure. 1. Descamps, V., Aitken, J. & Lee, M.G. Hippocrates on necrotising fasciitis. Lancet 344, 556 (1994). 2. Hakkarainen, T.W., Kopari, N.M., Pham, T.N. & Evans, H.L. Necrotizing soft tissue infections: review and current concepts in treatment, systems of care, and outcomes. Curr Probl Surg 51, 344-362 (2014). 3. Sims Sanyahumbi, A., Colquhoun, S., Wyber, R. & Carapetis, J.R. Global Disease Burden of Group A Streptococcus. in Streptococcus pyogenes : Basic Biology to Clinical Manifestations (eds. Ferretti, J.J., Stevens, D.L. & Fischetti, V.A.) (Oklahoma City (OK), 2016). 4. Walker, M.J., et al. Disease manifestations and pathogenic mechanisms of Group A Streptococcus. Clin Microbiol Rev 27, 264-301 (2014). 5. Zumerle, S., et al. Targeted disruption of hepcidin in the liver recapitulates the hemochromatotic phenotype. Blood 123, 3646-3650 (2014). 6. Datta, V., et al. Mutational analysis of the group A streptococcal operon encoding streptolysin S and its virulence role in invasive infection. Mol Microbiol 56, 681-695 (2005). 7. Peyssonnaux, C., et al. HIF-1alpha expression regulates the bactericidal capacity of phagocytes. J Clin Invest 115, 1806-1815 (2005). 8. Lichti, U., Anders, J. & Yuspa, S.H. Isolation and short-term culture of primary keratinocytes, hair follicle populations and dermal cells from newborn mice and keratinocytes from adult mice for in vitro analysis and for grafting to immunodeficient mice. Nat Protoc 3, 799-810 (2008). 9. Nizet, V., et al. Innate antimicrobial peptide protects the skin from invasive bacterial infection. Nature 414, 454-457 (2001). 10. Peyssonnaux, C. et al. TLR4-dependent hepcidin expression by myeloid cells in response to bacterial pathogens. Blood (2006). doi:10.1182/blood-2005-06-2259

192

ANNEXES

CLAIMS

1. Method for treating Gram positive bacterial infection in a patient in need thereof, comprising administering to the patient a polypeptide comprising a sequence of 20 amino acids having - at least 50% identity or at least 60% identity with the sequence SEQ ID NO: 1 - cysteine residues at positions 2, 5, 6, 8, 9, 14, 17, and 18 or a nucleic acid encoding said polypeptide.

2. A method according to claim 1, wherein said polypeptide wherein said polypeptide is a vertebrate homologue of a mature form of human hepcidin.

3. The method according to claim 2 wherein said vertebrate homologue is a mammalian homologue.

4. A method according to any one of claims 1 to 3; wherein the polypeptide is a polypeptide of 25 amino-acids having the sequence SEQ ID NO: 3.

5. A method according to any one of claims 1 to 4 wherein the gram positive bacteria is hyperinvasive bacteria.

6. A method according to any one of claims 1 to 5 wherein the gram positive bacteria is selected from the group consisting of Streptococcus, Staphylococcus, Clostridium, Listeria, Baccilus and Corynebacterium.

7. The method according to claim 6 wherein the gram positive bacteria is Streptococcus selected from the group consisting of beta-hemolytic (Group A or Group B), alpha hemolytic (pneumonia) or gamma haemolytic (Enterococcus).

8. The method according to claim 7 wherein the gram positive bacteria is Group A beta-hemolytic Streptococcus (GAS).

9. The method according to any one of claims 1 to 8, wherein Gram positive bacterial infection is associated with Necrotizing fasciitis (NF).

193

ANNEXES

- 17 -

ABSTRACT OF THE INVENTION

METHODS FOR TREATING GRAM POSITIVE BACTERIAL INFECTION

5 The present invention relies on the discovery that, surprisingly, when it is produced locally in the epidermis, hepcidin is able to directly initiate the recruitment of neutrophils by increasing the CXCL1 production in keratinocytes. While hepcidin had no direct antimicrobial activity against Group A streptococcus (GAS), injection of hepcidin, at the site of infection prevented

GAS systemic spread. Hepcidin agonists may represent a novel therapeutic to prevent life

10 threatening bacteremia not only in streptococcal NF but also in complicated infections due to compromised host immunity. The present invention relates to a method for treating Gram- positive bacterial infection in a patient in need thereof, comprising administering to the patient hepcidin polypetide. More specifically, it concerns method for treating with mature form of hepcidin, gram-positive bacterial infection such as Group A streptococcus infection 15 which could be associated with Necrotizing fasciitis (NF).

194

ANNEXES

195

ANNEXES

196

ANNEXES

197

ANNEXES

198

ANNEXES

199

ANNEXES

200

ANNEXES

201

ANNEXES

202