Correale and Farez observed that multiple sclerosis (MS) does not relapse in patients with eosinophilia from parasitic infection (1) and that eradication of parasites led to relapse of MS and loss of eosinophilia (2). The effect was demonstrated to be due to increased CD4⁺CD25⁺ T regulatory cells (Treg) (1) and regulatory B cells (3). Administration of parasites to modulate MS (4, 5) and other autoimmune diseases (6) has been trialed with limited effect (7). In these MS studies, the animal-derived parasites do not survive in humans so any effect of the parasites on the immune system was transient. There are major ethical issues with use of parasites that are human pathogens. This has led to attempts to identify antigens or chemicals in parasites that impede immunity and would not require infection with a pathogenic parasite, reviewed in Ref. (7, 8).

Hosts infected with parasitic worms can have reduced immune responses to autoantigens administered to induce autoimmunity, reviewed in Ref. (7) as well as vaccines (9) and infections, such as tuberculosis (10), malaria (11), HIV (12), and cholera (13). Thus, parasites of different species induce an immune inhibitory response to a variety of different antigens and pathogenic immune responses.

Parasitic infections induce a type-2 response that is initiated by innate lymphoid cells 2 (ILC2)-producing interleukin (IL)-5, IL-9, and IL-13 (14, 15), which induces CD4⁺ T helper 2 (Th2) cells that secrete IL-4, IL-5, and IL-13 (15). ILC2 and Th2 cells promote parasite expulsion (16). Parasites also induce Treg that can prevent immune-mediated elimination of parasites leading to chronic parasitic infections (9).

Diversion of T cell responses to a Th2 phenotype has been associated with a tolerogenic and less inflammatory immune response, including inhibition of pathogenic Th1 (17, 18) and Th17 (19) responses. Inflammatory responses to parasitic worms produce IL-5, which induces eosinophilia that in some models promotes parasite elimination (20) and in others prevent effective immunity (21, 22). The Th2 response to parasites also expands CD4⁺CD25⁺FOXP3⁺ Treg (23, 24) and other immune regulatory mechanisms, reviewed in Ref. (7, 8).

Experimental autoimmune encephalomyelitis (EAE), the animal model of MS, can be induced by immunization with autoantigen, which activates Th1 (25) and Th17 (26) effector responses, but a Th2 response is also induced and produces IL-4, IL-5, and IL-13 (25, 27, 28). Spontaneous recovery from EAE is associated with activation of Treg that promote recovery (29, 30). Parasitic infections reduce severity of EAE, including Th17 response and T cell infiltration to central nervous system (CNS) (31, 32). Parasitic infections that reduce severity of EAE and induce Treg include Trypanosoma cruzi (33), Plasmodium chaubaudi (34), Trichinella spiralis (35), and Nippostrongylus brasiliensis (36). Recently, it was reported IL-5 and IL-33 may mediate the protective effect of parasite infection on autoimmunity (36).

Our hypothesis is based on our observations about antigen-specific CD4⁺CD25⁺FOXP3⁺ Treg that mediate immune tolerance, reviewed in Ref. (37–39). These studies led us to recognize that antigen-specific Treg depend upon cytokines other than IL-2 and IL-4 (40–43) and that one such cytokine is IL-5 (43). This hypothesis is illustrated in Figure 1. The key finding related to this study is naïve CD4⁺CD25⁺FOXP3⁺ Treg activated with an autoantigen or alloantigen and rIL-4, not IL-2, are induced to express IL-5Rα, the specific receptor for IL-5 (19, 44).

Our hypothesis is supported by a number of findings in allo- and autoimmunity. First, we observed that treatment with IL-5 inhibits allograft rejection (45) and experimental autoimmune neuritis (19). Treatment with rIL-5 to suppress autoimmunity and allograft rejection depends upon host IL-4 (19, 45) and CD25⁺ T cells (19). The treatment with rIL-5 inhibits Th1 (19, 45) and Th17 responses (19). Second, we observed that animals tolerant to an allograft have CD4⁺CD25⁺FOXP3⁺ Treg that express IL-5Rα and respond to specific alloantigen in the presence of rIL-5 (46). Third, we observed that the CD4⁺ T cells from tolerant animals lose the capacity to transfer tolerance in culture unless they are activated by specific donor alloantigen and cytokines in supernatant from Con A-activated lymphocytes (40, 42). Cytokines that maintain these antigen-specific CD4⁺ tolerance transferring cells are IL-5 (43) and IFN-γ (42), whereas IL-2 (40) and IL-4 (41) do not.

Taken together, these findings show that naïve CD4⁺CD25⁺FOXP3⁺ Treg activated by a specific antigen can be activated down separate pathways by different cytokines. This is a two-step process. With type-1 responses, IL-2 and later IFN-γ or IL-12 activate these cells (44, 47). In type-2 responses, first IL-4 and later IL-5 activate antigen-activated Treg. The second step is blocked by the original activating cytokine, thus IL-2 (40) or IL-4 (41) do not maintain antigen-specific Treg and may inhibit them. These findings suggested IL-5 may promote these antigen-specific Treg. Thus, we reasoned that IL-5 produced by the inflammatory response to parasite may also expand autoantigen-specific Treg that have been first activated by IL-4 and the immunizing autoantigen.

In rats, N. brasiliensis infection induces eosinophilia and Th2 cytokines IL-4 and IL-5 between 7 and 14 days, after which these parameters return to normal by 28 days due to expulsion of the parasites (48). This study examined if the inflammatory response activated by infection with N. brasiliensis (49) produced IL-5 that tipped the balance to further expand antigen-specific activated Treg that had already been activated by autoantigen and IL-4, see Figure 1. Lewis rats were infected with N. brasiliensis (50) 3 days prior to immunization with myelin basic protein (MBP) and complete Freund’s adjuvant (CFA) to induce EAE. These parasite-infected hosts were treated with a monoclonal antibody (mAb) that inhibits IL-5 at the time of peak IL-5 production, which blocked the beneficial effects of parasite infections and reduced induction of antigen-specific Treg. To demonstrate the key role of IL-5, we examined whether therapy with recombinant (r) IL-5 inhibited EAE and induced antigen-specific Treg.

The beneficial effects of environmental acquired parasitic infections on MS and other autoimmune diseases have been associated with eosinophilia. Induction of eosinophilia is due to Type-2 cytokines IL-4 and IL-5 produced by Th2 and ILC2 cells (14, 15) in response to the parasitic infection. These clinical observations led to studies where patients with MS or inflammatory bowel disease are infected with parasites (6, 61–65). Deliberate infection of patients with a parasite has significant ethical issues, as parasites that infect humans can induce morbidity and mortality. In clinical studies to date, non-human parasites were given to patients, but these are rapidly eliminated and do not produce a prolonged Type-2 response and eosinophilia (6, 61–65). This may explain why they failed to show a therapeutic benefit. As a consequence, there are many investigations of molecules produced by parasites that alone may induce the Type-2 response and eosinophilia, reviewed in Ref. (7, 8). Administration of enriched parasitic molecules that induce Type-2 responses and eosinophilia may be therapeutic without the risks associated from persistent parasitic infection.

As tissue invasive parasitic infection induces type-2 responses that produce IL-4, IL-5, and IL-13 (66, 67), and IL-5 induces eosinophilia (21), these observations led us to the hypothesis examined in this study. We previously observed that IL-5 promotes induction of antigen-specific CD4⁺CD25⁺ Treg that prevent autoimmune-mediated demyelination in experimental autoimmune neuritis (19). We also demonstrated that naïve CD4⁺CD25⁺FOXP3⁺ Treg activated by a specific antigen and the Type-2 cytokine IL-4, but not the usual Treg cytokine IL-2, are induced to express IL-5Rα (44) and that IL-5 further activates these antigen-specific Treg (44). Antigen-specific Treg that express Il5ra are further expanded by IL-5 and stimulation with specific antigen both in vitro (46, 56) and in vivo (19). We proposed that tTreg activated by a specific antigen and IL-4 require IL-5 to promote their survival (43) and their inhibition of auto- and allo-immunity (19, 43, 44), as illustrated in Figure 1.

In this study, we examined if the effects of parasitic infection could be due to increased IL-5, which not only induce eosinophilia but also as a bystander effect further expands antigen-specific CD4⁺CD25⁺ Treg activated by autoantigen and IL-4. The observations that supported our hypothesis were as follows: (i) infection with the parasitic nematode N. brasiliensis reduced the severity of EAE and induced eosinophilia, (ii) treatment with an anti-IL-5 mAb blocked this protective effect of parasitic infection and partially reduced the level of eosinophilia; (iii) compared to EAE controls and anti-IL5 mAb-treated parasite-infected EAE hosts, parasite-infected hosts with EAE had increased numbers of CD4⁺CD25⁺ T cells that responded to the specific-autoantigen MBP, but not to a third-party autoantigen; (iv) depletion of CD25⁺ cells prior to induction of EAE prevented parasitic infection reducing the severity of EAE. Treatment with an anti-CD25 mAb that depletes CD25⁺ cells is a standard method of showing CD4⁺CD25⁺ Treg are required to mediate the effect.

To test Koch’s postulate, we showed treatment with IL-5 reduced the severity of EAE to a degree comparable to parasite infection. We also showed IL-5 treatment, as with parasitic infection, increased the number of CD4⁺CD25⁺ Treg that expressed Il5ra.

These findings show parasitic infection mediates part of its effect of reducing the severity of EAE via Th2 and/or ILC2 cells (14, 15) activated to produce high levels of IL-5. IL-4 produced early in the response to parasitic infection may have contributed to the initial activation of antigen-activated Treg to induce expression of il5ra and subsequent dependence on IL-5.

When first described, Th2 cells were considered anti-inflammatory and tolerance-promoting (67). Nasal therapy with MBP peptides combined with IL-4 reduces severity of EAE in Lewis rats (68). Combining rIL-4 and anti-CD3 mAb therapy reduces severity of EAE and inhibits Type-1 responses, but spares type-2 (25). rIL-4 therapy prevents autoimmune nephritis (17). IL-10 therapy inhibits EAE, but its effect is abrogated by IL-4 (69). In IL-5 knock out mice, the severity and time of onset of EAE was similar to wild-type controls, but type-2 responses were reduced (70). The effects of parasite infection on EAE in IL-5 knock out animals are not known. On the other hand, ERK1 knockout animals have enhanced Th1 responses and reduced IL-5 production, and develop more severe EAE (71) consistent with type-2 responses modifying EAE. These studies do not prove type-2 cytokines directly inhibit autoimmunity, however.

Recently, Finlay et al. described that immunization with Fasciola hepatica secretory/excretory products inhibited EAE and reduced Th1 and Th17 responses (36). This effect is associated with type-2 responses, but is not dependent on IL-4, IL-10, or Treg. They identified that IL-33 with IL-5, produced by ILC2, mediates the effect by induction of eosinophilia (36). In this study, blocking IL-5 also reduced parasite-induced protection against EAE (36). IL-33 has been shown to reduce EAE by inhibition of Th1/Th17 response, as well as enhanced Type-2 responses and M2 macrophages (72). IL-33 enhances both Th2 effector responses and activation of Treg by Type-2 cytokines (73, 74), thus these findings may complement those described here. Other immune regulatory mechanisms may be induced by type-2 responses, including alternate activated macrophages (M2), which are induced when parasite infection reduces severity of EAE (32).

Although the induction of EAE is by Th1 and Th17 responses, there is induction of a Th2 response (25, 27). The recovery from EAE is driven by induction of antigen-specific CD4⁺CD25⁺ Treg and is dependent upon IFN-γ (75). This is the Th1 cytokine-induced pathway for activation of antigen-specific Treg (44, 47) that is the major contributor to the spontaneous recovery from EAE in controls (75).

The detection of the slow and small proliferation of antigen-specific Treg requires culture techniques that eliminate non-specific background proliferation, as we have previously described (19, 44, 46, 56). In this study, CD4⁺CD25⁺FOXP3⁺ Treg proliferation to the specific antigen MBP was greater before recovery from EAE (day 16) in parasite-infected hosts, than EAE controls and anti-IL-5 mAb-treated parasite-infected EAE hosts. This response was only to specific-antigen MBP and not to RTA or unprimed self-stimulator cells, confirming this was an antigen-specific Treg response. This confirmed that parasite-infected hosts had more MBP-specific CD4⁺CD25⁺ Treg than EAE controls and anti-IL-5 mAb-treated EAE animals. Further CD4⁺CD25⁺ T cells from parasite-infected and IL-5 treated hosts expressed more Il5ra than CD4⁺CD25⁺ T cells from EAE controls or anti-IL-5 mAb-treated parasite-infected EAE animals.

Our interpretation of the effects of parasitic infection is that production of IL-4 and IL-5 by the activated Th2 cells and ILC2 accelerates the induction of autoantigen-specific Treg that express Il5ra, which in turn, reduced the severity of EAE. The source of the IL-5 in our hosts was not identified but could be Th2 or ILC2 cells that have been activated in response to parasitic infection (36). Late treatment with an anti-IL-4 mAb did not prevent the effects of parasites on EAE. We think that by the time anti-IL-4 mAb was given, the host ILC2 and Th2 cells had already produced sufficient IL-4 to activate naïve CD4⁺CD25⁺FOXP3⁺ Treg with TCR specific for MBP to express Il5ra. In an allograft tolerance model, we showed IL-4 does not sustain the alloantigen-specific Treg that mediate tolerance (41), but IL-5 does (43). In other models, early blocking of IL-4 prevents induction of il5ra expressing Treg and rIL-5 therapy is ineffective (19, 45). With parasitic infection and rIL-5 therapy, these autoantigen and IL-4 activated Treg were further expanded, respectively, by parasite infection-induced host type-2 response producing IL-5 or by administered IL-5.

Whether or not the benefits of chronic parasitic infection where numbers of CD4⁺CD25⁺ T cells are increased in patients with MS (1, 2) is mediated in part or mainly by IL-5 requires investigation.

Historically, IL-5 has been considered pro-inflammatory in allergic diseases including asthma. IL-5Rα is not expressed by T cells other than IL-4 and antigen-activated CD4⁺CD25⁺FOXP3⁺ Treg (19, 44). IL-5Rα is mainly expressed on eosinophils, mast cells, and basophils, as well as a subpopulation of naïve B cells. Thus, IL-5 cannot activate T effector cells. Trials with an anti-IL-5 mAb in asthma showed blocking IL-5 had limited or no effect, except in patients with high eosinophil counts, reviewed in Ref. (76). Many patients, including those with parasitic infection have continued production of IL-5, which does not enhance autoimmunity or allergic diseases. The safety issues for IL-5 or IL-5 analogs in therapy will require careful examination, but may be acceptable. For treatment of chronic diseases such as MS, continued treatment may be required as the antigen-specific Treg require exposure to IL-5 to maintain their function.

The hygiene hypothesis related to autoimmunity and allergy can in part be explained by lack of parasitic infections that can reduce unwanted immunity, including allergies and autoimmunity (77). This study suggests IL-5 produced in response to parasite infection may mediate a major part of the protection from autoimmunity and allergy by activation of antigen-specific Treg that have been activated by IL-4.

High levels of IL-5 and the resultant eosinophilia are protective against some parasite species, but not against others and may even be counterproductive (21, 22). The immune response that eliminates parasites is complex and varies depending on the species of parasites, as well as the type-2 response and eosinophilia (78).

Our hypothesis is illustrated in Figure 1. Parasitic infection initially activates ILC2 cells producing IL-4, IL-5, and IL-13 (14, 15), which direct effector cells to differentiate into Th2 cells. Th2 cells initially produce IL-4 and later produce IL-5 and IL-13. In parallel, naïve tTreg activated by an antigen in the presence of high IL-4 are activated to express Il5ra. Treg activated by the parasite will suppress effector responses directed at the parasite. In our experiments, there is a second immune response to MBP, which activates Th17, Th1, and Th2 responses. Normally, the Th17 and Th1 response dominate, and IL-2 induces naïve CD4⁺CD25⁺ Treg activated by MBP to express Ifngr and Il12rb2. This produces potent antigen-specific Treg that control the effector response to MBP so clinical disease abates and animals recover from paralysis. This response is normally driven by IFN-γ (75). However, immunization with MBP/CFA also activates some effector Th2 cells that initially produce IL-4 (25, 27) and this promotes activation of antigen-activated Treg to express Il5ra. In the parasite-infected host, the IL-4 from the early type-2 response to parasites would increase the number of MBP-specific Treg that express Il5ra and these would be further expanded by IL-5 produced by the late type-2 response to the parasite. These MBP-activated Treg suppressed the response to MBP and reduced the severity of EAE.

The results as a whole are consistent with IL-5 that is produced in excess by the type-2 response to the parasites, promoting earlier expansion of autoantigen-specific Treg that limit the severity of immune injury in EAE. The identification of IL-5 as a key molecule in reducing immunity to autoantigens, some infections, vaccines, and exacerbating responses to allergens, makes IL-5 a target to block and restore protective or less destructive immunity.

GT designed and conducted experiments and contributed to analysis, PW conducted and analyzed experiments, LD produced and inoculated parasites and contributed to analysis, NV conducted experiments and contributed to analysis. CR conducted and analyzed experiments, NC conducted some control EAE experiments, BH designed and analyzed experiments, SH designed and analyzed experiments. All authors contributed to the writing of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.