Synergistic (positive) interactions

Nematodes may modify the internal environment within the host to make it more suitable for a second species, with the consequence that the second species stands a greater chance of survival in the presence of the first. This may arise because the first species weakens the host's defenses in order to facilitate its own survival, but thereby generates an opportunity for other species both to colonize the host and/or to survive longer than they otherwise would. This type of interaction was termed ‘interactive protection from expulsion’ by Mitchell (1979) and would be expected to lead to positive associations between species in wild host populations.

Antagonistic (negative) interactions

Parasites may also interact with each other antagonistically, leading to reduction in the abundance of the second species in animals harbouring the first species. Theoretically, antagonistic interactions between parasites may occur for a variety of reasons. Non-immunologically-mediated interactions may include competition for attachment sites or for limited nutritional resources within the host, the action of toxic metabolic products from one species against another, and physical displacement as a consequence of a bulky biomass (e.g. adult cestodes, Haukisalmi & Henttonen 1993b).

Immunologically-mediated antagonism may arise through specific or non-specific cross immunity. If two parasite species share antigens, a specific immune response elicited by one may be effective against the second. A negative interaction may also follow when the host response to the first species generates a hostile environment in which the second species cannot survive adequately. This type of interaction was termed ‘interactive intestinal expulsion’ by Mitchell (1979).

Non-immunologically-mediated interactions

There is little direct experimental evidence for non-immunological synergistic interactions in the intestine between rodent nematodes of the type that have been reported in other host/parasite combinations (Holmes, 1973). Although nematode infections profoundly alter the intestinal environment this usually has no effect on other species present concurrently or is deleterious. For example, no evidence was found for altered growth rates or effects on fecundity during concurrent infections with Trichinella spiralis and Nippostrongylus brasiliensis in mice, other than in the phase when mice responded to either parasite with intestinal inflammation (Kennedy, 1980). In rats exposed concurrently to N. brasiliensis and Strongyloides ratti, the establishment of both nematodes was much the same as in single infection controls (Nawa & Korenaga, 1983). Similarly, in three-way interactions between Hymenolepis diminuta, Moniliformis moniliformis and N. brasiliensis no evidence was found for synergistic interactions (Holland, 1987).

Species of nematodes found in concurrent infections often occur in different regions of the intestine, their distributions presumably determined by differences in the qualities of the niche occupied. This spatial separation may not favour non-immunological synergistic interactions and may in fact represent the consequences of antagonistic interactions (see below).

Interactions mediated through the host's immune system

Following a primary exposure the trichostrongyloid nematode Heligmosomoides polygyrus establishes long-lasting infections in mice. Experiments with the murine subspecies H. p. bakeri have shown that in most strains of laboratory mice infections last for 8–10 months (Robinson et al. 1989). The parasite survives because it immunomodulates the intestinal environment, secreting immunomodulatory factors (IMF) that interfere with the T cell- and cytokine-mediated regulation of the intestinal inflammatory response (Behnke, 1987); Monroy & Enriquez, 1992); Telford et al. 1998). In particular, H. polygyrus infection impairs the mucosal mast cell response (Dehlawi, Wakelin & Behnke 1987), although little is known about the molecular mechanisms involved. As might be expected, host protective immune responses against other species of intestinal nematodes, which depend on a functional mast cell response, are downregulated in concurrent infections, enabling the heterologous species to survive much longer than they otherwise would. For example, infections with Trichinella spiralis, which are normally terminated in 2–3 weeks, persist for more than 6 weeks when H. polygyrus is present (Behnke, Wakelin & Wilson 1978). Similarly infections with Trichuris muris are also prolonged (Jenkins & Behnke, 1977), enabling even moderately intense worm burdens to survive to patency (Behnke, Ali & Jenkins 1984). This non-specific interactive protection from expulsion even extends to tapeworms such as Hymenolepis diminuta and H. citelli (Hopkins, 1980; Alghali, Hagan & Robinson, 1985).

Experimental work involving concurrent infections with H. polygyrus and T. spiralis has shown that production of cytokines such as IL-9 and IL-10, which are required to elicit and sustain mucosal mastocytosis in response to infection with T. spiralis, is significantly inhibited in concurrent infections. The outcome is a less intense acute response, and a slower loss of T. spiralis (Behnke et al. 1993).

Trichuris muris is also believed to enhance its survival by host immunomodulation. During a primary infection Th₂-mediated host-protective immunity operates primarily against larval stages, being considerably less effective against adult worms (Grencis, 1997). If the host response is delayed and the worms develop to the pre-adult and adult stages, chronic infections result. These stages are believed to secrete IMFs that polarize the host response towards Th₁. Since such responses down-regulate Th₂ activity, the parasite burden is then not expelled. This dependency of immunity against T. muris upon functional Th₂ responses may well explain earlier observations that infections were prolonged in mice exposed to protozoan infections, where Th₁ responses predominate (Phillips, Selby & Wakelin, 1974; Phillips & Wakelin, 1976). Similar data were obtained in concurrent infections between T. spiralis and Eimeria vermiformis (Rose, Wakelin & Hesketh, 1994). Phillips & Wakelin (1976) used concurrent infections involving Babesia microti, a species that is prevalent in wild rodent populations and therefore may exert an influence on their nematode infections. Effects upon the protozoan infection were not reported, but it is interesting to note that, more recently, in similar experiments involving H. polygyrus, it was demonstrated that the time-course of Babesia microti infections was not affected (Behnke, Sinski & Wakelin, 1999).

The possibility that nematode-induced immune modulation influencing concurrent infections may be more widespread than so far studied experimentally is reinforced by data showing that a number of species release excretory/secretory factors capable of down-regulating host T cell proliferation (Allen & MacDonald, 1998). It is also interesting to note that helminth-induced upregulation of Th₂ responses has been proposed as a dominant factor in the pathogenesis of AIDS in Africa, such responses, and the corresponding downregulation of Th₁ responses, making the host more susceptible to infection with HIV (Bentwich, Kalinkovich & Weisman, 1995).

Non-immunologically-mediated interactions

The niches occupied by adult nematodes in the intestine vary both longitudinally and radially. Longitudinally-distinct niches range from the oesophagus to the rectum, whilst radially-distinct niches include the lumen, the mucosal surface, the various layers of the mucosa itself and the serosa (Crompton, 1973). Niche selection begins at the establishment stage and is reflected in the choices made by incoming larvae (Holmes, 1961), 1973); Sommerville, 1963). Some have argued that the narrow location specificity of nematodes in the intestine is a reflection of past competitive interactions, leading to selective segregation and preferential selection of non-overlapping niches to minimize competition between species in order to achieve maximum fitness (Schad, 1963; Holmes, 1973). However, others have pointed out that since many niches in vertebrate hosts are unfilled (Price, 1980) niche segregation may be primarily driven by the need for parasites to aggregate for mating and reproduction (Rohde, 1979). Competitive interactions among helminth parasites in general have been thoroughly summarized by Christensen et al. (1987), and their long-term evolutionary consequences (i.e. extinction, and interactive or selective niche segregation) have been reviewed (Holmes, 1973), discussed (Rohde, 1979); Price, 1980) and modeled (Dobson, 1985).

Interactions mediated through the host's immune system

Non-specific inflammatory effector mechanisms. Intestinal inflammatory responses to infection with nematodes are controlled largely by Th lymphocytes through their release of cytokines that regulate the various effector arms (reviewed Behnke et al. 2000, Else & Finkelman, 1999). Whilst there is general agreement that host-protective intestinal responses require functional Th₂ cell activity, there is still controversy over what it is that actually causes worms to leave the host. One idea that has been current for many years is that the effectors released at the height of the acute response create an intestinal environment that is detrimental to worm survival. Early work using concurrent infections of T. spiralis and other parasites, suggested that the effector phase appeared to act non-specifically to bring about expulsion of heterologous species resident in the intestine at the time of the acute response to T. spiralis. Thus T. muris and Nippostrongylus brasiliensis were expelled prematurely when the host response to T. spiralis was maximal (Bruce & Wakelin, 1977); Kennedy, 1980) and T. spiralis itself was lost prematurely when the response to N. brasiliensis was maximal (Kennedy, 1980). Protozoan parasites such as Giardia muris (Roberts-Thomson et al. 1976), Eimeria nieschulzi (Stewart, Reddington & Hamilton, 1980) and E. pragensis (Rose et al. 1994) were also affected by inflammatory responses elicited by T. spiralis. Similarly, tapeworms such as H. diminuta and Rodentolepis nana were lost from T. spiralis-infected mice and rats (Behnke, Bland & Wakelin, 1977); Christie, Wakelin & Wilson, 1979); Ferretti et al. 1984). Rodentolepis (= Hymenolepis) microstoma suffered severe impairment to its growth but because of its relatively sheltered attachment position in the bile duct escaped expulsion from the intestine (Howard et al. 1978). Thus the picture which emerged from these studies was one in which an antigen-specific immune response elicited wide ranging and non-specific effector mechanisms, which removed other parasites out of the intestine at the time of, or immediately following, the acute response to the inducing species.

This picture was radically altered when Japanese workers published an elegant and informative set of experiments involving concurrent infections (reviewed Nawa et al. 1994). By exploiting nude (athymic) mice as a immunologically-inert host environment, they showed that injecting IL-4 (to induce a mastocytosis) into mice carrying both N. brasiliensis and Strongyloides ratti, resulted in the removal only of the latter species. This raised the concept of selectivity in the effectiveness of effector mechanisms, i.e. intestinal mucosal mastocytosis was essential for the elimination of S. ratti, but was a redundant effector in the case of N. brasiliensis. Instead, the latter succumbs to the infection-induced mucus response even though the products of activated mast cells are released during the course of infection (Nawa & Korenaga, 1983). Later studies (Khan et al. 1993) showed that mastocytosis was also a key effector against Strongyloides venezuelensis, suggesting a genus-specific susceptibility.

Effect of non-specifically-elicited effectors on chronic nematode species. Experiments involving concurrent infections have been very informative in helping to understand how chronic nematode species survive in their hosts. When mice carrying low intensity infections with H. polygyrus (i.e. levels which minimize the immunomodulatory effects) were challenged with T. spiralis, the adult H. polygyrus were lost in proportion to the dose of T. spiralis administered (Behnke, Cabaj & Wakelin 1992). Since it is known that there is a positive correlation between the dose of T. spiralis and the intensity of the mucosal response induced (Dehlawi & Wakelin, 1995) this result indicates that H. polygyrus is susceptible in a dose-dependent manner to the non-specific effectors of the mucosal response. Therefore, the survival strategy of H. polygyrus may depend more on preventing intestinal responses from occurring rather than on tolerating the hostile intestinal environment once elicited (Smith & Bryant, 1986).

In contrast to H. polygyrus, hookworms were not eliminated during the course of a concurrent infection with T. spiralis (Behnke, Rose & Little, 1994). The adults of neither Necator americanus nor Ancylostoma ceylanicum were lost when hamsters rejected superimposed infections with T. spiralis, indicating that in these species resilience to the non-specific components of inflammatory responses are key features of the host-parasite relationship. How hookworms manage to survive in experimental model systems (e.g. hamsters) or in their canine and human hosts is still largely unknown, but current interest centres on hookworm products that alter the environment locally where the mouth-parts embed in the mucosa (Hotez & Pritchard, 1995). The combination of firm, but not permanent attachment (hookworms move feeding stations at regular intervals, Kalkofen, 1970), bioactive chemicals and oxygen radical scavenging enzymes (Brophy & Pritchard, 1992) may endow hookworms with sufficient resilience to survive acute responses to other parasites.

Immunologically-specific interactions

Cross immunity. There is very considerable antigenic cross-reactivity between parasitic nematodes, a fact that bedevils immunodiagnosis. Such cross-reactivity may also extend to antigens involved in host-protective immune responses. There have been several records of cross-resistance between intestinal nematodes although an antigenic basis for the resistance has not always been demonstrated. One example where this has been shown concerns Trichinella spiralis and Trichuris muris, members of the same taxonomic group. Mice exposed to primary infection with T. spiralis or T. muris showed enhanced resistance to a subsequent challenge with the other species (Lee, Grencis & Wakelin, 1982) and this resistance could be induced both by active immunization with an antigen preparation of each species and by adoptive transfer of immune lymphocytes.

Enhancement of Th₂-based responsiveness. Just as the polarization of host immunity towards Th₁ responses may explain some examples of synergistic interactions (see above), the converse may provide a theoretical basis for antagonistic interactions. Experimental studies using T. muris, where the susceptibility of certain inbred mouse strains reflects the dominance of Th₁ responses, have shown that susceptibility can be overridden by concurrent infections with parasites that induce strong specific Th₂ responses. This has been demonstrated using T. spiralis and T. muris (Hermanek, Goyal & Wakelin, 1994) and Schistosoma mansoni and T. muris (Curry et al. 1995). In both cases the Th changes were reflected in altered antibody isotype and cytokine profiles.

Evidence from the field – helminth communities in wild rodents

In host populations where total parasite species richness is high and where several core (dominant) species each show high prevalence, polyparasitism is likely to be common even in the absence of any synergistic interactions; indeed this is well recognized to be the case. Among randomly-associating helminth species, purely on grounds of probability, we would expect a preponderance of positive associations in component communities characterized by many common (prevalence>90%) species and negative associations when rare (prevalence<10%) species predominate (Lotz & Font, 1994).

Kisielewska (1970b) was among the first to address the question of whether such co-occurrences include interactions between species. Her studies revealed that some species of nematodes showed antagonistic associations (Capillaria murissylvatici with Mastophorus muris, Heligmosomum halli [now = H. mixtum, Tenora & Meszaros, 1971)] with Heligmosomoides glareoli) whilst others showed synergistic associations e.g. H. mixtum with Syphacia obvelata [now = S. petrusewiczi, Tenora & Meszaros, 1975)], S. petrusewiczi with C. murissylvatici, S. petrusewiczi with the cestode Catenotaenia pusilla (probably C. henttoneni, See Haukisalmi & Tenora, 1993), and H. mixtum with C. henttoneni. However, Kisielewska (1970b) employed simple statistics in her analyses and re-examination of her results failed to substantiate the validity of these associations other than that involving H. mixtum and C. pusilla (Hobbs, 1980).

Employing considerably more powerful statistical tests, Haukisalmi & Henttonen (1993a) came to the conclusion that significant associations of helminths were rare in Clethrionomys glareolus in Finland. Of the 12 pairs of helminths infecting their bank vole populations only 4 combinations (H. mixtum with Heligmosomoides glareoli, C. murissylvatici with M. muris, H. mixtum with Catenotaenia sp., and H. mixtum with the cestode Paranoplocephala gracilis), all positive, were identified in presence/absence data and only three showed significant quantitative associations (negative – H. mixtum with Catenotaenia sp., H. glareoli with P. gracilis and positive – C. murissylvatici with M. muris). Only two of these combinations interacted consistently (H. mixtum with Heligmosomoides glareoli, C. murissylvatici with M. muris), the other associations being weak and unpredictable with respect to the intrinsic and extrinsic factors that also influenced worm burdens.

Collectively these studies, and others, suggest that whilst helminths co-occur commonly in nature, associations (presence/absence data and quantitative) are rare and inconsistent when detected, often dependent on a particular season, site, or occurring only in specific years. Mostly, they are generated through factors other than interactions between the species concerned. Montgomery & Montgomery (1990), who studied nine species of helminths in wood mice (Apodemus sylvaticus) in Northern Ireland, arrived at much the same conclusion.

Unequivocal demonstration that synergistic or antagonistic interactions between parasites occur in naturally infected animals is fraught with statistical and logistic difficulties. Co-occurrence, and moreover, associations of particular pairs of species, do not necessarily imply that interactions between them exist as species of parasites may aggregate independently in particular hosts for a variety of reasons. To give just two examples, parasite transmission stages may selectively infect hosts of a particular sex and age (e.g. because they forage more extensively, as males do in the breeding season) and infective stages may be aggregated in a particular microenvironment into which some hosts wander more often than others. Even quantitative associations (as reflected in statistically significant positive/negative correlations) may arise simply because hosts show patterns of behaviour that predispose them to infection with both of a pair of species of parasites.

Taking the field systems into the laboratory

The laboratory strain H. polygyrus bakeri is widely used as a model of chronic intestinal nematode infections (Monroy & Enriquez, 1992). However, the fact that it was originally isolated from an abnormal host and has been passaged through various mouse strains for almost half a century (Behnke, Keymer & Lewis, 1991), suggests that this parasite is probably now quite different from that found in wild animals in the field. In Europe, wood mice (Apodemus sylvaticus) carry a closely related species H. polygyrus polygyrus (Lewis 1968a; Montgomery & Montgomery, 1988). This species has been isolated recently and experiments in wood mice have shown that infections last some 12 weeks or so, i.e. less than H. p. bakeri in laboratory mice (Gregory, Keymer & Clarke, 1990) but considerably longer than species such as T. spiralis and N. brasiliensis. When this wild isolate was given to laboratory mice virtually no larvae survived the initial week of the tissue stage of infection (Quinnell, Behnke & Keymer, 1991). Whilst, intuitively we might suppose H. p. polygyrus to employ a similar, albeit perhaps somewhat weaker, immunomodulatory strategy to H. p. bakeri, to our knowledge this has not been investigated.

In Europe, wild voles carry nematodes of the related genus Heligmosomum (Tenora & Meszaros, 1971; Meszaros, 1977); Tenora & Stanek, 1995). In Poland, common voles, Microtus arvalis, show a high prevalence of infection with Heligmosomum costellatum (Bajer et al. unpublished). We have established both a breeding colony of the voles and cultures of the nematode to test hypotheses concerning immunomodulation as a survival strategy in this host-parasite combination. Preliminary data suggest that there are synergistic interactions between H. costellatum and the intestinal protozoan Cryptosporidium parvum, a species that infects domestic animals and is capable of causing severe enteritis in humans (James, 1997); Bednarska, Bajer & Sinski, 1998). Common voles have proved to be competent reservoir hosts of C. parvum and individuals infected with adult H. costellatum have been found to excrete significantly more oocysts than worm free animals from day 17 post infection with C. parvum until the end of the observation period (day 42) (Bajer et al. 2000).

Field tests of laboratory-generated hypotheses

While earlier quantitative survey data on helminth parasites of wild wood mice and voles did not provide strong support for interactions between species, laboratory experiments with H. p. bakeri would suggest that synergistic interactions should be found between the related Heligmosominae (H. p. polygyrus, H. costellatum and H. mixtum) and other parasites. In a recent re-analysis of such data Behnke et al. (unpublished) found that whilst H. p. polygyrus was associated positively with other species in terms of categorical data (presence/absence), only one species, the cestode C. pusilla, showed a quantitative positive interaction, the number of C. pusilla increasing with increasing burdens of H. p. polygyrus. Perhaps surprisingly, after taking account of differences between sites, seasons and host intrinsic factors such as age and sex, there was little evidence of positive interactions between H. p. polygyrus and Syphacia stroma, two species that occupy a similar intestinal site and represent the dominant helminths of wood mice in the U.K.