Review series on helminths, immune modulation and the hygiene hypothesis: Immunity against helminths and immunological phenomena in modern human populations: coevolutionary legacies?

Abstract

Although the molecules and cells involved in triggering immune responses against parasitic worms (helminths) remain enigmatic, research has continued to implicate expansions of T-helper type 2 (Th2) cells and regulatory T-helper (T_reg) cells as a characteristic response to these organisms. An intimate association has also emerged between Th2 responses and wound-healing functions. As helminth infections in humans are associated with a strong Th2/T_reg immunoregulatory footprint (often termed a ‘modified Th2’ response), plausible links have been made to increased susceptibility to microbial pathogens in helminth-infected populations in the tropics and to the breakdowns in immunological control (allergy and autoimmunity) that are increasing in frequency in helminth-free developed countries. Removal of helminths and their anti-inflammatory influence may also have hazards for populations exposed to infectious agents, such as malaria and influenza, whose worst effects are mediated by excessive inflammatory reactions. The patterns seen in the control of helminth immunity are discussed from an evolutionary perspective. Whilst an inability to correctly regulate the immune system in the absence of helminth infection might seem highly counter-adaptive, the very ancient and pervasive relationship between vertebrates and helminths supports a view that immunological control networks have been selected to function within the context of a modified Th2 environment. The absence of immunoregulatory stimuli from helminths may therefore uncover maladaptations that were not previously exposed to selection.

Overview

A general consensus has emerged that effector T-helper cell type 2 (Th2) mechanisms evolved to counter infections with helminths (and perhaps other large metazoan parasites) and that these defences are generally deployed against such pathogens within the context of a significant regulatory T (T_reg) cell response.¹ There is also a growing appreciation that immune responses against helminths share some of the same signals and mediators as the body's damage repair systems.² As the interaction between the vertebrate immune system and at least some of the major helminth groups extends back hundreds of millions of years,^3–5 near to a time when the adaptive immunity first evolved^6,7 (see Fig. 1), reciprocal co-adaptation may have fundamentally shaped the characteristics of both. Below we consider the mechanism of anti-helminth immune responses and how ancient adaptations to deal with helminth infection may have left their mark on the way the immune system is structured and controlled. We discuss how this relates to adverse immunological phenomena seen in modern helminth-free human populations, such as allergy, autoimmunity and responses to immunopathological infectious agents.

Figure 1.

The immune system evolved in the presence of helminths. The earliest dates of colonization by helminth lineages are mapped onto a phylogenetic history of vertebrates. Also mapped are major putative events in immune system evolution and the appearance of mediators of importance in anti-helminth responses. Divergence times of the major vertebrate groups from the mammalian lineage are indicated (Mya, million years ago).⁹⁶ Macrophage-,⁹⁷ eosinophil-⁹⁸ and mast-⁹⁹ like cells occur in Agnathans (lampreys) predating the appearance of classical adaptive immunity. The evolution of classical adaptive immunity was possibly preceded¹⁰⁰ by two rounds of whole genome, or large-scale partial genome, duplication (4n and then 8n). Insertion of recombination activating (RAG) genes¹⁰¹ into the genome then allowed the evolution of the recombinatorial receptor system underpinning T-cell and B-cell functions. This occurred at some point between the divergences of the Agnatha (lampreys and hagfishes) and the Chondricthyes (sharks and rays). Colonization by platyhelminths occurred within this period, with all extant parasitic groups believed to be derived from a single common parasitic ancestor.⁴ The earliest evidence of Th2 cytokines^102–104 and alternatively (aaMΦ) versus classically (caMΦ) activated macrophages^105,106 is found in the Actinopterygii (bony fishes). Colonization of land probably allowed the invasion of primitive tetrapods by parasitic nematodes^3,5 (molecular phylogenetic analyses suggest at least four independent invasions).¹⁰⁷ The immunoglobulin E (IgE) isotype, important in anti-helminth responses, is a recent innovation in the mammalian lineage. Ig, immunoglobulin; Mφ, macrophage.¹⁰⁸

Helminths

‘Helminth’ is a term used to group ecologically similar worm-like parasites actually derived from three unrelated phyla: the platyhelminths (tapeworms and flukes), the nematodes (roundworms) and the acanthocephalans (thorny-headed worms). A breathtaking diversity of species occur in vertebrates, the global fauna in an individual vertebrate host species typically numbering tens or even hundreds of different taxa. Life cycles may be direct, with transmission from one definitive host to another, or complex, involving distinct life-history stages cycling through different host species. Invading stages often undertake complex migrations through different organ systems and tissues once inside the definitive host. Adult helminths are large in comparison to microbial pathogens, individuals varying in size from hundreds of microns to tens of metres long. Generally, they produce propagules that are dispersed away from the existing host, sometimes by an arthropod vector. The relatively slow maturation time of individuals (a consequence of their size) and the requirement for host–host dispersal to complete the life cycle mean that the potential for the exponential population growth seen in microbes that reproduce in situ is very limited. Herein lies a fundamental difference. Whilst microbes represent an imminent threat and are usually countered by rapid, relatively violent inflammatory responses, helminths seem to be handled with gentler immune responses, with a strong regulatory component, that may take weeks, months or even years to reach their peak effectiveness.

Any survey of a vertebrate wildlife population will generally yield an array of different helminths (see e.g. ref. 8), with many hosts infected by tens, hundreds or even thousands of individual worms. It also seems reasonable to assume the same for pre-industrial human populations, given the rates of helminth parasitism seen in the modern-day tropics. Almost every organ system and tissue in the human body (Fig. 2) is a potential target for infection, and well over 100 species have been recovered from humans.⁹ However, in terms of current global significance, a smaller number of species might be highlighted, all of which are primarily distributed in the tropics. In blood flukes of the genus Schistosoma, water-borne larvae penetrate the host's skin before establishing in the blood vascular system, producing eggs that migrate through tissues before release to the exterior. Larvae of the filarial nematode Onchocerca volvulus are transmitted by blood-feeding blackflies and establish in subcutaneous nodules, from which adult females release millions of microfilaria larvae into the dermis of the host. The ‘geohelminth’ soil-transmitted nematodes Ancylostoma duodenale, Necator americanus, Ascaris lumbricoides and Trichuris trichiura infect the host's gut, and from there eggs are passed to the exterior in faeces. Ancylostoma duodenaleand N. americanus (hookworms) invade the host by larval penetration of the skin, migrating through the lungs before arriving in the intestine. Ascaris lumbricoides (roundworm) and T. trichiura (whipworm) are transmitted by an oral route, the former undergoing a complex migration through the lungs, and the latter developing entirely in a mucosal site. Worm infection tends to be chronic, with maximum individual life-spans of months or a few years (the gastrointestinal nematodes) to decades (schistosomes). Pathology mediated by these helminths is relatively limited: overt symptoms are usually rare in human populations although adverse outcomes may be measurable, for example, in reduced growth or cognitive development. That this limited level of pathogenicity is a result of adaptation seems likely. For example, outbreaks in humans in the Philippines of an intestinal capillarid usually infecting fish-eating birds resulted in severe disease and a 6% mortality rate from > 1000 confirmed cases between 1967 and 1990.¹⁰

Figure 2.

The wide range of organ systems targeted by a selection of the helminth parasites occurring in humans.

A stereotypical immune response

The development of resistance to gastrointestinal nematodes is generally associated with a Th2 response in laboratory models^11–13 (Fig. 3a). Findings consistent with this have been reported in epidemiological studies of hookworm,¹⁴ roundworm^15–17 and whipworm^15–17 in humans. The triggering of the Th2 response programme by intestinal nematodes, orchestrated by cytokines including interleukin (IL)-4,¹² IL-5,¹⁸ IL-9,¹⁹ IL-13,²⁰ IL-21²¹ and IL-25²² and IL-33,²³ leads to a stereotyped cascade of effector mechanisms, potentially including immunoglobulin E (IgE) isotype-switched B-cell responses, mast cells, eosinophils, basophils and increased permeability, smooth muscle contractility and mucus production in the gut. Increases in IL-13-driven intestinal epithelial cell (IEC) turnover²⁴ and production of the protein resistin-like molecule beta (RELMβ),²⁵ which interferes with nematode chemotaxis, may also be involved in protection. However, only a subset of the recruited mediators²⁶ may be important effectors in resistance against any individual gastrointestinal nematode species. For example, in mouse model systems, mucosal mast cell responses are critical for worm expulsion in Trichuris muris and Trichinella spiralis infection^19,27,28 but not in Heligmosomoides bakeri^29,30 or Nippostrongylus brasiliensis infection.³¹ Th2 responses are also triggered by tissue-dwelling (histozoic) species such as the schistosomes³² and filarial nematodes,³³ although it is not clear-cut how important Th2-driven effector mechanisms are in protecting the host.^34–38 In these cases Th1 mechanisms can be implicated in resistance and are perhaps especially important in killing incoming larval stages^38,39 migrating through tissues, but once the definitive site is reached the adult parasite may survive for many years. Fibrotic responses can be a prominent feature of the host reaction against histozoic species: Th2-driven granulomas are focussed upon eggs in Schistosoma spp.³⁸ and fibrosis in Onchocerca infection can be associated with microfilarial migrations and occur around the onchocercoma nodules enclosing adults.⁴⁰

Figure 3.

Cells and molecules involved in the generation of immune responses in helminth infections. Solid arrows indicate signalling influences on cells (with important signalling molecules indicated alongside), and dashed arrows indicate the developmental trajectory of an individual cell type. (a) A widely accepted hypothesis for the function of mediators and effectors involved in the modified Th2 response against helminths. Pathogen-associated molecular patterns (PAMPs) from helminths may stimulate the differentiation of dendritic cells (DCs) that bias T-helper (Th) cell polarization towards Th2 or regulatory T (T_reg) cell subsets. Th2 cells produce a range of cytokines driving effector responses, including: eosinophil responses, mast cell responses, isotype-switched B-cell responses, increases in intestinal permeability, smooth muscle contractility and mucus production, and the differentiation of alternatively activated macrophages (aaMΦ). T_reg cells induced in the periphery and producing suppressive cytokines [interleukin (IL)-10 and transforming growth factor (TGF)-β] dampen levels of innate and adaptive immune activation. (b) A model incorporating alternative hypotheses for early events in the initiation of anti-helminth responses. Helminth PAMPs might stimulate signalling activity in other ‘front-line’ cell types (eosinophils, basophils, mast cells or intestinal epithelial cells) which, in turn, influences DCs or even naïve Th (Th0) cells. Damage-associated molecular patterns (DAMPs) released by dying cells, or signals resulting from apoptosis, might also influence either DCs or Th0 cells. EDN, eosinophil-derived neurotoxin; TSLP, thymic stromal lymphopoetin.

The other recurring immunological characteristic of helminth infections is that despite significant responses, for example vigorous antibody responses and eosinophilia, which may or may not be effective in resistance, there is often a marked systemic suppression of innate and adaptive immunity⁴¹ in the host. This may be linked to expansions of regulatory CD4⁺ CD25⁺ T cells producing the suppressive cytokines transforming growth factor (TGF)-β and/or IL-10^42–44 and perhaps other suppressor T-cell subsets,⁴⁵ including CD8⁺ T cells with cytokine-independent suppressive activity.⁴⁶ Immunoepidemiological studies in humans have strongly associated regulatory cytokine activity and/or immunological hyporesponsiveness with increasing infection intensity in filarial nematodes,^37,47Schistosoma flukes^48,49 and intestinal nematodes.⁵⁰ A further important dimension in helminth infections is the differentiation of alternatively activated macrophages (aaMΦs)^2,51 under the influence of IL-4, IL-13 and IL-21. Whilst aaMΦs recruit to the site of infection⁵² and have been implicated as functional effectors,⁵³ they also have strong anti-inflammatory properties, secreting IL-10 and TGF-β, and strongly expressing genes such as Arginase-1, Ym1 (chitinase-3-like protein 3) and FIZZ1 (resistin-like molecule alpha)⁵⁴ whose functions relate to the repair of extracellular matrices, wound healing and fibrosis.

The overall outcome of a helminth infection may therefore be an environment with down-regulated proinflammatory responsiveness, activated damage repair mechanisms and a tightly controlled development of Th2 anti-parasite effector responses. Although exposure to helminths in exposed human populations typically begins in early childhood, delayed age-intensity peaks⁵⁵ and substantial reinfection rates⁵⁶ are a testament to the slow development of immunity.

Evolution of the Th2 arm: home win or score draw?

From a ‘worm's view’ the nature of the host response is such that primary worm infections are often successful, but secondary waves of invasion are likely to encounter progressively increasing levels of resistance. The apparently evolutionarily fixed nature of the stereotypical anti-helminth Th2/T_reg response⁵⁷ may stem from the fact that, within the constraints acting on both protagonists, it maximizes fitness in both the host and some individual parasites.⁵⁸ For the host, the highly regulated Th2 anti-worm response and initial toleration of infection may lead to greater fitness than the generation of violent, energetically costly and potentially immunopathological responses to a low or intermediate threat. For the parasites in primary infections, relative fitness may be very high because they can survive and reproduce whilst later invaders succumb to the developing immune responses. There might therefore be strong host-mediated selection for parasite genotypes that co-operate with, rather than subvert, the Th2/T_reg phenotype, perhaps even to the extent that there could have been synergistic coevolution of complementary pathogen-associated molecular pattern (PAMP)–host innate receptor recognition systems. Stabilizing features of such an interaction might be that the parasite cannot easily evolve mechanisms for incoming stages to evade or subvert a fully developed memory response in a well-adapted host and that primary infections that fail to stimulate a Th2 response might benefit late-arriving competitors, penalizing the fitness of early colonizers.

As mentioned above, Th2 mediators are important in defence against metazoan parasites, but also for damage repair,⁵⁷ and it is intuitive that the two roles might be evolutionarily linked. Worm infections are likely to produce high levels of local tissue damage associated with the feeding and migratory activities of individual parasites. There may also sometimes be a requirement for the host to physically contain (‘wall off’) dangerous motile stages by fibrotic reactions within tissues. It thus seems plausible that, during evolution, anti-worm mediators may have been ‘grafted’ onto the control networks that regulate wound repair and tissue remodelling. Mechanisms involved in wound healing and responses to helminths are not totally overlapping, of course, as the latter also involve genuine effector mechanisms (IgE, mast cells and esoinophils) not usually implicated in wound repair, and themselves capable of generating immunopathology. However, the intimacy of damage repair and anti-helminth responses may give clues as to the enigmatic signals that initiate dominant immune responses in worm infections.

The initiation and control of immune responses in worm infections

Whilst the exact signalling pathways by which Th2 responses are initiated remain obscure, helminth products seem able to stimulate the development of partially activated DCs,⁵⁹ with suppressed expression of Toll-like receptors (TLRs) and activation markers,^60,61 that promote Th2 and/or T_reg phenotypes.^59,62,63 From an analogy with the very well-known pathways by which Th1 cells are induced, the simplest model for the initiation of Th2 responses in helminth infections would be an interaction between a PAMP derived from a worm and a pattern recognition receptor (PRR) expressed by a DC. Worm PAMPs might act via one or more of the TLR family, or perhaps other classes of innate receptor^64,65 such as C-type lectins, nucleotide-binding oligomerization domain-containing protein (NOD)-like receptors or protease-activated receptors. Prospecting of helminth excretory-secretory (ES) products for molecules with TLR-mediated immunomodulatory effects, mainly focussing on DCs, has yielded a number of examples that might be consistent with such a model. Synthetic lacto-N-fucopentanose III, equivalent to a molecule occurring naturally in soluble egg antigen (SEA), has been reported to interact with TLR4 to produce Th2 polarizing DCs⁶⁶ and schistosome lyso-phosphatidylserine may interact with TLR2 to produce T_reg polarizing DCs.⁶⁷ ES-62 from the filarial nematode Acanthocheilonema viteae is reported to exert immunomodulatory effects on macrophages⁶⁸ and DCs⁶⁹ by a TLR4-dependent mechanism.⁷⁰ However, the interaction of ES-62 with TLR4 on mast cells, which results in the inhibition of Th1 but not Th2 cytokine production, does not involve classical TLR signalling.⁷¹ Also, the Th2-inducing effects of SEA on DCs have recently been shown to be independent of MyD88, TLR2 and TLR4.⁷² A further consideration is that PRRs on cell types other than DCs may be important (Fig. 3b), as responses that favour a Th2 environment have been reported in basophils,^73,74 mast cells,⁷¹ eosinophils⁷⁵ and IECs.⁷⁶ In the case of IECs, Trichuris muris infection stimulates inhibitor of kappa-B kinase beta (IKKβ)-dependent expression of the cytokine thymic stromal lymphopoetin (TSLP), which can stimulate the differentiation of Th2-inducing DCs. Innate responses by eosinophils include secretion of eosinophil-derived neurotoxin (EDN), which mediates TLR2/Myd88-dependent activation of DCs that drive in vivo antigen-specific adaptive responses towards a Th2 phenotype. A wide range of other alarmin-type signalling molecules can be released during infection (e.g. defensins, cathelicidins and high-mobility group 1 box proteins), including some with known pro-inflammatory influences,⁷⁷ but perhaps others with undiscovered Th2-inducing activities. Even if DCs are pivotal in initiating expansions of Th2 cells, pattern recognition by these cells themselves might not therefore be necessary, as cytokines or alarmins released following pattern recognition by more ‘frontline’ cells, such as IEEs (Fig. 3b), could be sufficient for maturation of Th2-inducing DC phenotypes. It is even possible that some Th2-inducing signals might bypass DCs altogether and act directly on T lymphocytes prior to acquisition of a terminally differentiated phenotype.⁵⁹

Apart from an exploration of cell types other than DCs, the overlap between effector and tissue repair functions in anti-helminth immune responses suggests that the search for innate ‘pattern recognition’ signals initiating anti-worm responses might need to consider endogenous host molecules that convey information about cellular damage (Fig. 3b) and that trigger anti-inflammatory programmes appropriate for wound healing. During helminth infection such damage-associated molecular patterns (DAMPs; here used to denote endogenous signals only) might be sensed by elements of the innate immune system, perhaps as combinatorial signals with PAMPs.⁵⁷ A range of endogenous molecules have been shown to interact with receptors of the TLR2, TLR3, TLR4, TLR7, TLR8 and TLR9 families, including some that are released on cell damage and that can exert either stimulatory or anti-inflammatory effects.⁷⁸ For example, suppressive IL-10-inducing ligands for TLR9 have been characterized,⁷⁹ some of the most potent of which contain TTAGGG multimers characteristic of mammalian telomeres.⁸⁰ Regulatory T cells may also be induced by other endogenous mechanisms: interactions with apoptotic cells, or phosphatidylserine from apoptotic cells, causes differentiation of immature DCs with tolerogenic properties favouring anti-inflammatory Th phenotypes.^81–84 It even seems a possibility that much of the T_reg activity in a worm infection could be initiated by signalling from endogenous molecules. In the case of anti-helminth Th2 responses, however, cross-talk between worm PAMPs and the innate immune system would be expected to be important, as many anti-helminth effectors (e.g. IgE and mast cells) are recruited that are apparently unconnected to wound healing.

Worm versus man: coevolutionary legacies?

An ideal organism would be able to perfectly control its immune system under any combination of infection challenges and antigenic stimuli. However, organisms are not perfect and have been selected to perform optimally within the set of conditions encountered by their ancestors. In humans, one of these past conditions is arguably exposure to a diverse community of helminth parasites. The relationship between helminths and vertebrates is ancient and can be traced back to when nematodes invaded primitive tetrapods³ and the ancestral platyhelminths⁴ first appeared in fishes prior to the colonization of land (Fig. 1). In the intervening time, individual helminth lineages may have diverged, switched between host lineages, or become extinct many times. However, if the pattern of infection seen in present-day wildlife populations can be extrapolated into the past, then almost every natural vertebrate population would have been exposed to a varied assemblage of worm-like pathogens. Because helminth infection is and (probably) has been so ubiquitous, mammalian immune systems may have been selected to anticipate that they would suffer chronic exposure to large-bodied Th2/T_reg-inducing pathogens. In other words, there would have been selection for immunoregulatory systems that work well in the context of worm infection. However, there would rarely have been selection for immunoregulatory systems that work optimally when worms are absent (the situation seen in modern human populations). The absence of parasitic worms and their strong Th2/T_reg-inducing influences may therefore uncover flaws in immunoregulatory networks that have previously been hidden from selection.^85,86 Such a scenario might explain the dramatic increases in allergy and autoimmune conditions in Western countries and is supported by the protective effect of helminth infection, or helminth-derived products, in laboratory models of allergy^43,87 and automimmunity.^85,88

The future

The strong Th2/T_reg-inducing influence of worms, and the possibility that the immune system may malfunction in the absence of such stimuli, has global public health implications: both for the parts of the world population that are helminth-free and for the parts that remain exposed. In the first case, a search for molecules that mimic the immunoregulatory stimuli seen in worm infections may one day yield pharmaceuticals that can be used routinely to induce healthier, better regulated immune systems. Given the intimate link between immune responses to worms and tissue repair, it might be necessary to go beyond a search for PAMPs that interact with DCs and to extend this to include other cell types⁶⁵ and other molecular patterns⁵⁷ including DAMPs. For helminth-exposed populations it has been demonstrated that the immunosuppressive activities of worm infections may make individuals less responsive to microbial pathogens and to vaccination. This is a major concern because wherever helminth infection is prevalent so too are dangerous microbial infectious agents, such as human immunodeficiency virus (HIV), Mycobacterium tuberculi and malarial parasites. Anthelmintic treatment programmes might result in significant health benefits from the increased immunological responsiveness against such pathogens. However, a ‘flip-side’ to this coin might be foreseen in the context of highly immunopathogenic coinfections such as malaria or influenza, whose worst effects are caused by excessive immune responses. The potential realism of this scenario is supported by laboratory model studies suggesting that gastrointestinal nematode infections suppress immunopathology during influenza.⁸⁹ It is also at least hinted at by the contradictory body of epidemiological^90,91 and laboratory model^92–95 data for interactions in malaria–helminth coinfections. The consequences of de-worming for susceptibility to the worst symptoms of important immunopathogenic infections are therefore of urgent interest for further laboratory model studies and should be carefully monitored in the field.⁵⁸