Contribution of myeloid cell subsets to liver fibrosis in parasite infection

Authors

  • Alain Beschin,

    Corresponding author
    1. Myeloid Cell Immunology Laboratory, Brussels, Belgium
    2. Cellular and Molecular Immunology Unit, Vrije Universiteit Brussel, Brussels, Belgium
    • Correspondence to: A Beschin, Cellular and Molecular Immunology Unit, Vrije Universiteit, Building E, Floor 8, Pleinlaan 2, 1050 Brussels, Belgium. e-mail: abeschin@vub.ac.be

    Search for more papers by this author
  • Patrick De Baetselier,

    1. Myeloid Cell Immunology Laboratory, Brussels, Belgium
    2. Cellular and Molecular Immunology Unit, Vrije Universiteit Brussel, Brussels, Belgium
    Search for more papers by this author
  • Jo A Van Ginderachter

    1. Myeloid Cell Immunology Laboratory, Brussels, Belgium
    2. Cellular and Molecular Immunology Unit, Vrije Universiteit Brussel, Brussels, Belgium
    Search for more papers by this author

  • No conflicts of interest were declared.

Abstract

Accumulation of extracellular matrix components secreted by fibroblasts is a normal feature of wound healing during acute inflammation. However, during most chronic/persistent inflammatory diseases, this tissue repair mechanism is incorrectly regulated and results in irreversible fibrosis in various organs. Fibrosis that severely affects tissue architecture and can cause organ failure is a major cause of death in developed countries. Organ-recruited lymphoid (mainly T cells) and myeloid cells (eosinophils, basophils, macrophages and DCs) have long been recognized in their participation to the development of fibrosis. In particular, a central role for recruited monocyte-derived macrophages in this excessive connective tissue deposit is more and more appreciated. Moreover, the polarization of monocyte-derived macrophages in classically activated (IFNγ-dependent) M1 cells or alternatively activated (IL-4/IL-13) M2 cells, that mirrors the Th1/Th2 polarization of T cells, is also documented to contribute differentially to the fibrotic process. Here, we review the current understanding of how myeloid cell subpopulations affect the development of fibrosis in parasite infections. Copyright © 2012 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Introduction

The restoration of homeostasis following injury requires a strictly timed interplay of cells and molecules during the healing process. Myeloid cells, and in particular distinct subsets of monocytes and macrophages, have been reported to critically orchestrate this process in, for example, the wounded skin [1], myocardium and kidney [1-3]. However, when the insult turns chronic, deregulated and continuous repair processes ensue and lead to the development of fibrosis. Fibrosis is characterized by an excessive accumulation of extracellular matrix (ECM) components, such as collagen and fibronectin, in and around the wound, leading to tissue remodelling, permanent scarring and, eventually, potentially lethal organ failure [4]. Activated myofibroblasts, that transdifferentiate, for instance, from vitamin A-storing quiescent hepatic stellate cells (HSCs), are the major ECM producers and, consequently, the main drivers of the fibrogenic process, irrespective of the fibrotic trigger [5]. Although the pathways leading to myofibroblast activation can differ between different pathogenic systems, the expansion/recruitment of myeloid cells secreting inflammatory mediators in areas of tissue injury is a recurrent theme tightly regulating the production of ECM and the recruitment of new myofibroblasts [6]. Moreover, the pro- or anti-fibrogenic activity of Th2 (like IL-13) or Th1 (IFNγ, IL-12p40) cytokines, as well as the pro-fibrotic activity of myeloid cell-derived but anti-fibrotic function of CD4+ regulatory T cell (Treg)-secreted TGFβ are by now well described [4, 7, 8].

The general principles of fibrogenesis apply to parasites causing chronic infection [6]. Here, we provide an overview of the mechanisms by which clinically important parasites – such as Schistosoma mansoni and S. japonicum, and hydatid cysts from Echinococcus multilocularis that localize primarily in the liver – drive liver fibrogenesis following their interaction with myeloid cell subsets. Other parasites, such as Trypanosoma cruzi, Taenia solium, Nippostrongylus brasiliensis and Schistosoma haematobium, inducing fibrosis in the heart, brain, lungs and bladder, respectively [9-15], and liver flukes, such as Clonorchis sinensis and Opisthorchis viverrini, causing bile periductal fibrosis, which increases the risk for cholangiocarcinoma [16-20], are not covered here, due to the marginal information on the phenotype, function and activation status of myeloid cells in these pathogenic processes.

Schistosomiasis

Schistosomal infections plague more than 250 million people worldwide. The main human-infecting species are S. mansoni (in Africa and South America) and S. japonicum (in South and East Asia), causing intestinal and hepatosplenic schistosomiasis, and S. haematobium (in Africa and Middle East), causing urinary schistosomiasis. The latter species accounts for nearly half of that number, but little is known about the mechanisms underlying its pathophysiology, primarily due to the lack of an experimentally tractable mouse model [15]. These parasites are responsible for the most common fibrotic disease that arises due to chronic inflammation and the scarring of (liver or bladder) tissue. Schistosomiasis also affects the health burden of other diseases, including AIDS and viral hepatitis. Hence, identifying the immune processes and the role of different cell types involved in schistosomiasis-induced pathogenicity is essential to develop novel therapeutic settings that limit the fibrotic response. The basic immune mechanisms associated with the development of granuloma and fibrosis are similar in S. mansoni and S. japonicum infection [21-23].

Host immune response determines the pathogenicity of schistosomiasis

Upon schistosome infection, adult worms migrate to mesenteric veins, where they begin laying eggs 5–6 weeks postinfection. The eggs that cannot travel against the blood flow into the intestine for excretion embolize in the liver vasculature, where they initiate an inflammatory granulomatous reaction. The granuloma, composed of recruited lymphoid and myeloid cells surrounded by a collagen matrix deposited mainly by IL-4/IL-13-activated fibroblasts and hepatic stellate cells, sequesters toxic egg antigens [22]. Most infected individuals develop a mild intestinal schistosomiasis, but in a small proportion the lesions do not resolve and progress to life-threatening hepatosplenic schistosomiasis associated with chronic hepatic fibrosis and portal hypertension. Similarly, several experimental mouse models develop a pronounced acute hepatic inflammation and acute death, while others develop milder lesions allowing the progression of the disease into a chronic stage, which is eventually detrimental, mainly due to hepatic fibrosis. This variation in pathogenicity is largely dependent on the host genetic background [24] and, as summarized below, on the immune response.

In hosts developing mild schistosomiasis, the immune response in draining and non-draining lymph nodes, as well as in hepatic and intestinal tissues, switches rapidly to a Th2 response dominated by IL-4, IL-5, IL-13 and IL-21 and the expansion of Th2-associated myeloid cells [eosinophils, basophils, mast cells, alternatively activated (M2) macrophages and dendritic cells (DCs)]. In the acute stage, adult worms slightly bias towards a Th1 response (IFNγ) and M1-type myeloid cells, which can kill the parasites in a TNF-R1 and iNOS-dependent way [25, 26] and restrict the granuloma size and collagen deposit due to their IL-12- and iNOS-mediated anti-fibrotic activity [27-29]. Subsequently, IL-4Rα+ M2, induced by IL-4 and IL-13, are required to suppress the IL-12/Th1/M1 and IL-23/Th17 response, thereby allowing the emergence of regulatory T cells that promote the production of IL-10 and TGFβ. These cytokines limit the IL-17-mediated recruitment of neutrophils and promote the repair/wound healing of the intestinal wall as worms cross from mesenteric blood to the lumen for expulsion, avoiding destruction of liver and gut barrier integrity as well as exposure of the host to lethal septicaemia in acute infection [7–9 weeks post-infection (wkpi)]. IFNγ also regulates the Th17 response and thereby favours the transition from an acute to a chronic infection [30-46]. Yet, at this stage, despite granuloma involution as eggs are killed and the development of hyporesponsiveness in Th2 cells [47], the Th2/M2 inflammatory response elicits tissue remodelling, including induction of matrix metalloproteinases (MMPs), tissue inhibitor of metalloproteinases (TIMPs) and collagens, that culminates in life-threatening liver and intestinal fibrosis, portal hypertension and haemorrhages (16–20 wkpi) [28, 38, 41, 48-50]. IL-13, via the IL-13Rα1/IL-4Rα receptor, is the most important fibrotic agent in schistosomiasis, with additive effects of IL-4, IL-5, IL-10 and IL-21 but not of TGFβ [45, 50-53]. Consequently, distinct collaborative IL-13 regulatory pathways that suppress the Th2/M2 activation and hamper fibrosis progression have been identified: the high-affinity IL-13 decoy receptor (IL-13Rα2) expressed by stromal cells; IL-12/IL-23p40 produced by macrophages and DCs driving the Th1/Th17 response; and IL-10 produced by M2 and regulatory T cells [46, 54-59]. Of note, there is no evidence that mice that lack IL-10 alone develop more extensive fibrosis during acute or chronic schistosomiasis [60, 61]. Rather, neutralization of IL-10 signalling in the chronic infection precipitates the death of infected mice without affecting the Th2 response and the fibrotic process, but by inducing liver portal hypertension and the resulting portal-systemic shunting of egg parasites to the heart and the lungs [62].

The innate immune response and initiation/induction of a Th2 response in skin imprint hepatic inflammation

Schistosome larvae penetrate the host via a percutaneous route as free-swimming cercariae. The invading larvae release excretory/secretory molecules that aid penetration of the skin [63, 64]. Larval antigens are found within 3 h in skin GR-1 (which comprises the Ly6G and Ly6C molecules)+MHCII neutrophils, followed by F4/80+MHCIILo macrophages and CD11c+MHCIIHi DCs within 24 h. Antigen-carrying macrophages and DCs are later detected (within 40 h) in the skin-draining lymph nodes [65, 66]. These and other myeloid cells, including eosinophils, basophils and mast cells, sense the infection and initially promote innate parasite killing by producing ROI/RNI, by releasing lytic enzymes and through antibody-dependent cytotoxicity [67]. This killing initiates resistance that reduces the pathogen burden in infected hosts. However, resistance inevitably induces collateral tissue damage (pathogenicity), which is caused by the pathogen itself or by the anti-pathogen immune response and influences host fitness [68-70]. Such degradation of host tissues generates Th2-polarizing alarmins, including thymic stromal lymphopoietin (TSLP), IL-25 and IL-33, by basophils, epithelial cells and/or innate immune cells such as nuocytes [71, 72]. However, the role of alarmins in schistosome infection is likely not prominent, with at best a transient TSLP-mediated effect on the development of IL-4/IL-13-dependent acute granulomatous inflammation and no effect on IL-13-dependent fibrosis in chronic infection [73, 74]. On the other hand, since canonical Th2 cytokines, including IL-4 and IL-13, are secreted rapidly by distinct innate cells (see below) during helminth infection, DCs and macrophages become very rapidly exposed to these cytokines, which will impact on the functional capability of these antigen-presenting cells (APCs) to induce a CD4+ Th2 response, the hallmark of helminth infections.

CD11cHiMHCII+ DCs, but not Siglec-F+GR-1Int eosinophils or CD19Siglec-FFcϵR1α+ basophils, were shown to be crucial initiators of Th2 immune responses during schistosome infection [75-80]. In this respect, from their penetration in the skin to the chronic stage of infection, schistosomes condition DCs towards a ‘modulated/tolerogenic’ phenotype by suppressing the activity of Th1-inducing (PAMPs), by inducing T-cell intrinsic hyporesponsiveness/anergy, and by inducing co-stimulatory molecules (eg CD40, OX40L and ICOSL) that bias the response towards Th2 [22, 47, 71, 72, 81-83]. Similarly, CD11b+F4/80+GR-1+ myeloid cells have been proposed to act as myeloid-derived suppressor cells (MDSCs) upon encounter with schistosome antigens, down-regulating Th1 proliferation and differentiation while inducing a Th2 response [71, 84-88].

Although schistosome egg antigens (SEA) recognized by TLRs (eg TLR4) or CLRs (eg DC-SIGN, MGL, MMR) are the more commonly accepted Th2 triggers [71, 84, 87, 89-92], it is clear that an antigen-specific Th2 response already develops during prepatent infection, before egg deposition begins. Indeed, in mice exposed percutaneously to S. mansoni, skin F4/80+MHCIILoLy6CHiLy6GSiglec-FCD11cLo myeloid cells rapidly switch to an IL4Rα-dependent M2 activation status (expressing Arg1, Chi3li3 and Il10, but not Relma), particularly after multiple exposures of the host to schistosomes, induce hypoproliferation/anergy in CD4+ T cells from draining and distal lymph nodes and impair the Th1/Th2 cytokine response (IFNγ, IL-4, IL-10, IL-5) to parasites prior to egg-laying [65]. Interestingly, these myeloid cells suppress the antigen-presenting capacity of skin F4/80+/−MHCIIHiLy6CLoSiglec-FCD11cHi DCs. Thus, at the infection site, cercarial products could skew recruited monocyte-derived myeloid cells to an alternatively activated suppressive/regulatory phenotype and the production of IL-10, paralleled by a diminished pro-inflammatory activity of DCs, which together limit IL-12- and NO-driven skin damage and contribute to wound healing [64, 66, 93-95]. The induction of tolerogenic DCs as well as suppressor, M2-orientated myeloid cells with a phenotype reminiscent of monocytic MDSCs, both at extra- and intrahepatic sites, may represent a feedback mechanism whereby activated DCs and M2 limit the Th2 response that triggered their expression, thereby regulating granuloma inflammation and fibrosis in the liver [65, 66, 96-101].

Although egg-induced granulomatous inflammation is often considered a major driver of liver fibrosis in schistosome infection [22], granuloma size, intensity of infection (expressed as worm pairs or total liver eggs) and fibrosis are often dissociated [102]. Thus, in schistosomiasis as in other parasitic diseases, resistance of the host (which reduces pathogen burden but induces tissue pathogenicity) and tolerance (which limits the negative impact of the pathogenicity but favours host fitness without directly affecting the pathogen burden) are dissociated phenomena [68-70]. In the next paragraphs, we will focus on the role of myeloid cell subsets in the fibrotic process during schistosomiasis.

Macrophages and fibrosis

The pro-fibrotic role of macrophages as a stereotyped response to chronic injury and multiple mechanisms by which macrophages cause fibrosis have been proposed [4, 103, 104]. It has long been assumed that recruited macrophages activated by IL-13 via IL-4Rα/IL-13Rα1 receptor signalling are the major triggers of schistosome-induced fibrosis. However, while fibrosis decreased in infected mice lacking complete IL-4Rα signalling or in IL-4/IL-13−/− mice, this process developed normally in strains deprived of M2 macrophages, such as LysMcreIl4ra−/lox or iLckCreIl4rα−/lox mice with specific depletion of IL-4Rα in macrophages and neutrophils or in all T cells, respectively. Notably, all these mouse strains succumbed in acute infection, due to excessive egg-induced Th1/M1 responses resulting in severe liver damage, reduced intestinal barrier integrity and endotoxaemia, suggesting that M2 macrophages are not required for the development of fibrosis but to avert acute mortality and morbidity. This protective effect of M2 macrophages is Th2- and IL-10-independent [31-34, 39, 54]. However, as fibrosis is not fully developed in the acute schistosome infection, a role for M2 macrophages in this chronic phase process cannot be overlooked. In this regard, M2 macrophages, through their arginase-1 activity, have been described during schistosome infection as down-regulators of inflammatory responses and facilitators of tissue remodelling, by competing with inducible NO synthase-2 (NOS-2) for l-arginine and by initiating the synthesis of proline and polyamines [28]. Using mice with a conditional deletion of arginase-1 (LysMCreArg1−/lox or Tie2CreArg1−/lox), macrophage arginase-1 activity was shown to limit the Th2 expansion that normally drives lethal liver fibrosis and portal hypertension by depleting l-arginine required for T cell proliferation (Figure 1). The expression of M2 genes characteristic for schistosome infection (Chi3li3, Relma) and Treg development were not altered upon arginase-1 deletion in macrophages [101, 105]. In addition, using chimeric mice lacking arginase-1 in bone marrow-derived cells, it was shown that the anti-proliferative activity of arginase-1 was also at work during acute infection, limiting Th17 differentiation, promoting TGFβ production and Foxp3 expression and thereby preventing lethal neutrophilia and endotoxaemia [35].

Figure 1.

Cellular interactions in the chronic fibrotic process induced by parasites triggering a polarized Th2 immune response. Circulating bone marrow-derived Ly6CHiCCR2Hi inflammatory monocytes are recruited and differentiate into M2-type monocyte-derived macrophages or DCs. DCs can drive Th2 cells and Treg differentiation. The production of profibrotic cytokines by these cells can modulate the activation of hepatic stellate cells and their differentiation into extracellular matrix component-secreting myofibroblasts. M2-type macrophages, in a negative feedback mechanism involving Arg1 and Relma, can restrict the proliferation of Th2 cells and hereby protect the liver from fibrosis development. M2-type macrophages and IL-10 and TGFβ-secreting Tregs can also block the infiltration of neutrophils that trigger liver damage through the production of ROS and RNS. Infiltration of the liver by basophils and eosinophils can also promote the differentiation of M2-type macrophages. It remains speculative (dashed line) whether DCs and macrophages derived from Ly6CLoCX3CR1Hi monocytes contribute to liver remodelling and fibrosis during parasite infection; and whether resident macrophages and DCs influence liver remodelling and proliferate upon IL-4 stimulation.

Relma is another IL-4/IL-13-induced protein that is expressed by granuloma SSCHiCD11b+MHCIIHi F4/80+GR-1Siglec-FIL4Rα+ macrophages in S. mansoni infection [54]. Using KO mice, Relma was shown, similar to arginase-1, to limit Th2-associated inflammation, thereby regulating IL-13-dependent fibrosis [101, 105]. In contrast, a Th2 promoting role of Relma produced by DCs was recently reported in S. mansoni-infected mice (Figure 1) [106]. These CD11cHiMHCIIHiF4/80FSCLo DCs exhibited an IL-4Rα-dependent alternative activation state, reflected by increased expression of Relma, Chi3li3 and Mrc1, but not Arg1. Mice treated with (SEA)-pulsed Relma−/− DCs produced lower levels of IL-10 and IL-13, increased levels of IFNγ and unaffected IL-4 production. Since Relma can be produced by several cell types in schistosome-induced granulomas (including eosinophils as the major source, and epithelial cells [65, 101]), their relative contribution to fibrosis deserves further investigation.

Regarding their origin, organ-associated myeloid cells constitute a heterogeneous population, including 'tissue-resident' cells likely derived from the yolk sac and 'tissue-recruited' cells derived from bone marrow-released circulating precursors/monocytes [107]. During induction of fibrosis in different pathogenic situations, bone marrow CCR2HiLy6CHi inflammatory (classical) monocytes are selectively recruited to the injured organ, where they can differentiate into CCR2HiLy6CLo monocyte-derived macrophages. These cells exhibit a M2 activation state under the influence of the Th2-dominated response and could interact directly with hepatic stellate cells via TGFβ to favour myofibroblast transdifferentiation and collagen production. These monocyte-derived macrophages are thought to be the major contributors to fibrosis [2, 3, 108-114]. In schistosome-infected hosts, based on the perivascular location of most of the granulomas, it is assumed that liver and granuloma macrophages derive from circulating cells and are larger in number than resident macrophages (Figure 1). In agreement, the absence of CCR2 signalling results in decreased fibrosis elicited by S. mansoni eggs [115]. Granuloma macrophages were identified as SSCHiCD11b+MHCIIHiF4/80+GR-1Siglec-FIL4Rα+ cells that also expressed the macrophage-specific markers CD204, Dectin 1 and CD68, as well as the M2 markers Mrc1 and Chi3li3. Interestingly, a minor population of these myeloid cells express GR-1 [54]. Whether the latter cells represent CCR2HiLy6CHi inflammatory monocytes recruited to the granuloma of S. mansoni-infected mice and are the precursors of GR-1 macrophages, as described in other inflammatory models, deserves further investigation. Moreover, based on their M2 gene expression profile, two subpopulations of granuloma macrophages were described: one located in close proximity to the eggs, in which the expression of Mrc1 and Chi3li3 was dependent on the expression of IL-4Rα in macrophages; and one located at the periphery of the granuloma, in which the expression of Mrc1 and Chi3li3 was dependent on IL-10 but independent of IL-4Rα signalling [54]. Since schistosome components are ligands of the C-type lectin MMR encoded by Mrc1 [116], the macrophages located in the centres of granulomas may exert a regulatory function on T cells. However, how these populations of macrophages with distinct location in the granuloma contribute to the control of fibrosis is currently unknown.

When analysing the cellular composition of the liver from chronic S. mansoni-infected mice, a population of infiltrating F4/80+Ly6cHiCD11b+MHCIIHi monocyte-derived macrophages was found to accumulate over time and to secrete IL-10 in response to immune complexes, reminiscent of regulatory macrophages generated by immune complex stimulation [117]. These IgG1+ macrophages prevented the development of severe portal hypertension and the portosystemic shunt of eggs to the heart and the lungs. However, in contrast to what is usually thought, this portosystemic shunting was not directly linked to an increased IL-13 production and associated hepatic fibrosis [62].

Although DCs were identified as major actors in the Th2 response induction in schistosome infection at the onset of egg laying [76], their impact on the fibrotic process that develops several weeks after exposure to eggs in chronic infection remains unknown.

Granulocytes and fibrosis

Granulocytes comprise different cell types, such as eosinophils, neutrophils and basophils. Having the ability to sense PAMPs, to phagocytose and to act as (APCs), these cells can promote innate parasite killing [67, 118]. By secreting a number of molecules, they are generally viewed as amplifiers of the adaptive Th2/M2 immunity in helminth infection (Figure 1). The phenotypic characterization of granulocytes that are recruited to the infection site or granuloma of schistosome-infected mice identified eosinophils as SSCHiCD11b+MHCIIF4/80+GR-1IntLy6GLoSiglec-FHiCD11cLoIL4Rα cells, neutrophils as SSCHiC D11b+MHCIIF4/80GR-1HiLy6GHiSiglec-FCD11cLoIL4Rα cells and basophils as SSCLoCD11b+MHCII+FcϵR1α+C-KitCD200R3+CD131/cytokine receptor common β-chain subunit (βc)+IL-3Rβ+Siglec-FCD117 cells [54, 75-77, 119].

Eosinophils

Eosinophils possess intrinsic characteristics that could be relevant during schistosome infections. In contrast to skin macrophages, skin eosinophils did not exert a suppressive activity on the capacity of DCs to trigger T cell proliferation. They can exert immune regulatory roles in the fibrotic process by driving M2 macrophage activation in the skin and/or the liver, being a predominant source of IL-4, IL-13 and Relma, and as mediators of tissue injury or tissue remodelling and debris clearance following tissue injury [65, 67, 78, 89, 101, 120-122]. However, despite constituting the major cell population in hepatic granulomas (∼50%), the function of eosinophils during infection largely remains elusive. Using several eosinophil-deficient mouse models (TgPHIL mice, in which eosinophil peroxidase drives the expression of the diphtheria toxin, or ΔdblGATA mice, having a deletion of a GATA1 binding sequence), it was observed that eosinophils seemingly had no impact on the Th2 response, worm burden, egg deposition, granuloma number and size, hepatocellular damage or fibrosis at weeks 8 or 12 of S. mansoni infection [79].

Neutrophils

During S. mansoni infection, neutrophils were thought to have little effect, since their depletion does not influence the severity of the disease [33]. In contrast to skin macrophages and similar to skin eosinophils, skin neutrophils did not exert a suppressive activity on the capacity of DCs to trigger T cell proliferation [65, 66]. However, in models where IL-17 was associated with the pathogenicity of schistosomiasis, Th17 responses driven by an IL-23p19/IL-1β pathway favoured neutrophil recruitment/accumulation and degranulation, thereby exacerbating egg-induced tissue damage and causing mortality in acute infection [35, 40, 42, 123-127]. Considering the death in acute infection of mice developing IL-17-mediated pathogenicity, the contribution of neutrophils to fibrosis was not clearly addressed. A role in this process cannot be excluded, considering that neutrophils can adopt an alternative activation status [128-131]. Moreover, since neutrophils and not macrophages or eosinophils are the source of arginase-1 activity in humans [132, 133], it will be important for future studies to determine whether human neutrophils express other markers of M2 activation, such as Relma, that could be important for fibrosis development.

Although the basic immune mechanisms associated with the development of granuloma and fibrosis are similar in S. mansoni and S. japonicum infection [21-23], more severe hepatic lesions and more vigorous Th17/IL-17-dependent neutrophil hepatic granulomatous inflammation developed in S. japonicum acute infection [123, 127, 134]. Although IL-4/IL-13 signalling suppressed the excess of neutrophil accumulation in both schistosome infections, acute hepatic granuloma inflammation in mice lacking IL-4/IL-13 signalling was impaired in mansoni schistosomiasis, while being more severe in japonicum schistosomiasis, suggesting that a moderate neutrophil infiltration is required to limit the hepatic lesion in acute S. japonicum-infected mice [126]. In addition, unlike in S. mansoni infection, the mortality rate of infected mice did not increase in the absence of IL-4/ IL-13 [32, 33, 126, 135]. Whether the Th17/neutrophil response contributes to prolonged survival and influences the development of fibrosis in S. japonicum-infected mice deserves investigation.

Basophils

Although an APC function and TLSP, IL-4 and IL-13 secreting function were attributed to basophils, these cells play little, if any, role in type 2 immune responses during schistosome infection [72, 75-77]. Interestingly, basophils become less responsive (evaluated by their CD200R expression and IL-4 production) to IgE-mediated stimulation in mice chronically (as compared to acutely) infected with S. mansoni, primarily due to IL-10 [119]. However, their role in induction of M2, and thus in fibrosis development, was not investigated. The suppression of basophil responsiveness may be a mechanism by which chronic helminth infections protect against the development of allergic diseases.

Echinococcosis

Echinococcosis is caused by the larval stages of cestodes belonging to the genus Echinococcus. Hepatic echinococcosis comprises cystic and alveolar forms, associated with E. granulosus and E. multilocularis infections, respectively. Both are life-threatening diseases establishing a periparasitic granuloma in the liver and irreversible fibrosis which results in bile duct and vessel obstruction and ultimately in liver failure.

Remarkably, similar to S. mansoni infection, hosts generate distinct immune responses during different stages of alveolar echinococcosis, whereby the sequential cytokine profiles appear crucial for prolonged metacestode growth and survival, and for partial protection of the host [136]. During experimental alveolar echinococcosis, three main stages of cytokine production can be discerned: an early, acute-stage production of Th1 cytokine (mainly IFNγ) associated with a slow parasite growth; a second stage with a mixed Th1/Th2 profile associated with rapid parasite growth; and a final stage of strong immunosuppression ablating both lymphocyte proliferation and cytokine secretion [137].

DCs and macrophages are among the first cells encountered by the metacestodes, recognizing parasite-associated molecules through TLRs and CLRs. The M1-orientated early activation of these innate cells, especially macrophages, turns them into anti-parasitic effectors protecting the host [138, 139]. The M1 activation status is paralleled and amplified by an initial elevation of pro-inflammatory Th1 cytokines. In this context, pretreatment of mice with IL-12 or IFNα (but to a lesser extent IFNγ) prevents lesion formation [140, 141]. Interestingly, the protective effect of IFNα is associated with a decreased IL-13 and increased IFNγ production [141], suggesting that a gradual shift toward Th2 immunity terminates the early cytotoxic response. No information is available as yet on the potential involvement of Th17 cytokines in the acute phase of alveolar echinococcosis.

The chronic stage of infection, both in patients and experimental mouse models, is characterized by a predominant Th2 response, illustrated by high IL-4, IL-13 and IL-5 levels and relatively low IFNγ [141]. The mechanisms driving this shift are unknown, but it seems likely that parasite-derived molecules induce functional changes in DCs and macrophages. As a matter of fact, macrophages from E. multilocularis-infected mice exhibit a reduced antigen-presenting capacity, possibly through the effect of a parasite-derived macrophage modifying factor [142]. In addition, crude non-fractionated E. mutilocularis antigen lowers DC maturation [143], leading to altered interactions with T cells. Another aspect potentially contributing to redirecting the immune response to metacestodes is the gradually increasing production of the anti-inflammatory/regulatory cytokines IL-10 and TGFβ. IL-10 is spontaneously secreted by PBMCs in patients with progressing alveolar echinococcosis lesions [144], but the highest concentration of IL-10 might be found in the periparasitic granuloma, where it is produced by T cells closest to the parasitic vesicles [145]. Similarly, TGFβ-secreting cells are present in the granuloma, although their identity is presently not clear [146].

The ultimate hallmark outcome of alveolar echinococcosis is fibrosis, which destroys the liver parenchyma and consequently is responsible for most of the pathological manifestations. However, in the case of E multilocularis, the embedding of parasitic vesicles in heavily crosslinked collagens also leads to metacestode death. Accordingly, it has been demonstrated that mouse strains with higher fibrogenic activity are more resistant to the disease [147, 148]. Of note, the periparasitic fibrotic lesions are very dense and cannot rupture. Consequently, the firm barrier between parasites and immune cells prevents parasite destruction and may also be responsible for the low frequency of anaphylactic reactions in alveolar echinococcosis patients [149]. Moreover, antiparasitic drugs cannot reach the lesions. Extensive extracellular matrix crosslinking may be mediated by a parasite-associated transglutaminase, whose expression is particularly strong at the border of the parasitic vesicles and was shown to crosslink human collagen in vitro [150]. In addition, more diffusible mediators appear to mediate fibrogenesis, since fibrotic areas arise even far from the parasitic vesicles [151, 152].

It is likely, as in Schistosoma infection, that recruited macrophages and DCs are central effectors cell in liver fibrosis caused by Ecchinococus infection. However, the exact surface phenotype and origin of these cells has not been established.

Perspectives

Among myeloid cells, M2-type macrophages derived from bone marrow-recruited CCR2HiLy6CHiCX3CR1Lo monocytes in the skin, liver and other tissues are considered to be major actors in fibrosis during schistosomiasis and other parasitic infections. However, defining the identities, dynamic and relative contribution of other M2-type myeloid cell subsets in this process – including resident macrophages such as Kupffer cells, macrophages differentiating from recruited GR-1LoCCR2LoCX3CR1Hi monocytes for which a role in other fibrotic process has been amply documented, yolk sac versus fetal liver/bone marrow-derived macrophages, resident conventional DCs versus recruited monocyte-derived DCs – remains challenging (Figure 1) [2, 114, 153-162]. In particular, considering the finding of IL-4-triggered resident macrophages proliferation during helminth infections [163, 164], it remains to be tested whether locally proliferating M2-like resident macrophages contribute to the outcome of parasite infection as a tissue repair mechanism. In addition, a comparison of the surface marker phenotype and a better definition of the M2 gene subsets that are expressed by myeloid cells elicited in distinct organs, including skin, liver, intestine and lungs, in parasite-infected hosts is required. Moreover, since schistosome products interact with TLRs and CLRs [116, 165-167], it will be instructive to determine whether differences in expression of TLRs and CLRs by macrophages, DCs and eosinophils contribute to their distinct function.

The differentiation and polarization of myeloid cells can be regulated by dynamic changes in transcription factors and miRNA expression and by epigenetic regulation, including DNA methylation and histone modifications by methyltransferases, demethylases, acetyltransferases and deacetylases [107, 168-174]. Thus, it will be important to investigate whether Th2 priming by DCs and M2 subsets development are dependent on these processes and which genes are targeted, with the aim of identifying specific mechanisms promoting the development of fibrosis in parasite infection. In this context, the function of understudied M2/MDSC-associated genes, such as chitinase family proteins [175, 176], should be elucidated. Of note, granulocytes become sensitized to produce IL-4 and IL-13 and to express M2 genes in response to worm antigens well in advance of egg deposition, warranting further investigations into their role in fibrosis initiation.

Overall, there is an unmet medical need to develop new modalities for the therapy of potentially lethal chronic fibrotic disorders that affect millions worldwide. Identifying the common and unique inflammatory fibrogenic mechanisms in distinct organs and in different pathologies (parasite infection, asthma, idiopathic pulmonary fibrosis, liver fibrosis caused by chronic hepatitis infection, alcohol abuse and non-alcoholic steatohepatitis (NASH), systemic sclerosis, ulcerative colitis, etc.) will be essential to evaluate novel therapies.

Acknowledgements

This study was supported by grants from an Interuniversity Attraction Pole Programme and the Fund for Scientific Research Flanders (FWO-Vlaanderen).

Abbreviations

ECM, extracellular matrix; HSCs, hepatic stellate cells; MDSCs, myeloid-derived suppressor cells; Tregs, regulatory T cells; TSLP, thymic stromal lymphopoietin.

Author contributions

All authors were involved in writing the paper and had final approval of the submitted manuscript.

Ancillary