Immunological relationships during primary infection with Heligmosomoides polygyrus (Nematospiroides dubius): dose-dependent expulsion of adult worms

Summary The survival of Heligmosomoides polygyrus was monitored during primary infections in female C57Bl10, NIH and BALB/c mice at low and high intensities of infection. Survivorship curves were fitted for each data set and analysed. C57Bl10 mice, given either low or high intensities of infection, harboured parasites for 28–37 weeks, heavier infections surviving marginally but significantly longer. Essentially the survivorship curves of H. polygyrus in C57Bl10 mice could be accounted for by senility, the increased probability of worms with a longer life-span occurring at high infection intensities and, possibly, by a contribution from host-protective immune mechanisms in the terminal stages of infection. The pattern of survivorship was different in NIH and BALB/c mice. NIH mice showed weak but significant density-dependent suppression of parasite loss and infections in this strain did not exceed 27·5 weeks in duration. Primary infections in BALB/c mice were briefer still and showed marked dependence on parasite density. Thus low-level infections lasted 10–15 weeks whereas heavier infections survived for 21–34 weeks. The data suggested that both strains developed host-protective responses to adult H. polygyrus and that parasite survival was curtailed earlier than would be expected if senility alone was involved. The hybrid strains (C57Bl10 × NIH)F1 and (B10G × NIH)F1 both expelled H. polygyrus in a dose-dependent manner, worm loss commencing within 10 weeks of infection. In some experiments worm loss was clearly evident by weeks 4 and 6. These hybrid strains showed gene complementation in that adult worms were cleared considerably earlier than in parental strains.


INTRODUCTION
The longevity of Heligmosomoides polygyrus in many commonly used laboratory mouse strains is well established. Since the early work of Ehrenford (1954) experiments in various laboratories have confirmed that adult worms may survive in mice for many months following a single pulse administration of infective larvae (Liu, 1966;Day, Howard, Prowse, Chapman & Mitchell, 1979;Behnke & Robinson, 1985;Keymer & Hiorns, 1986). The inability of mice to mount rapid host-protective responses to primary infection with H. polygyrus contrasts markedly with the brevity of infections with Nippostrongylus brasiliensis and Trichinella spiralis in the same host. The chronic infections with H. polygyrus have been attributed to the immunodepressive activity of the adult worm population influencing immunoregulatory mechanisms locally in the intestinal environment (Dobson & Cayzer, 1982;Dehlawi & Wakelin, 1988; reviewed by Behnke, 1987).
Despite the general agreement that H. polygyrus is a long-lived parasite in the mouse, there is evidence that the pattern of survival (survivorship curve) may be different among the mouse strains and influenced by the intensity of infection (Liu, 1966;Prowse, Mitchell, Ey & Jenkins, 1979;Dobson, Sitepu & Brindley, 1985). Detailed data for survivorship, based on worm counts are rare in the literature, most studies of longevity being based on faecal egg counts (Liu, 1966;Behnke & Robinson, 1985). Dobson et al. (1985) reported that high-intensity infections lasted longer than low-intensity infections in Quackenbush mice, a finding confirmed by monitoring faecal egg counts in CFLP mice (Behnke, Williams, Hannah & Pritchard, 1987). The results from both studies are consistent with the view that adult worms survive through exerting an immunodepressive influence on the host. Heavy infections with larger numbers of adult parasites would exert a stronger immunodepressive effect on the host than less intense infections. These studies, however, contrast with the report by Keymer & Hiorns (1986) who did not find a density-dependent effect on parasite survival in the randomly bred MFi strain.
The involvement and interaction of genetic and immunological factors in determining the duration of primary infection with H. polygyrus is still a poorly understood aspect of the host-parasite relationship, contrasting with the wealth of information on acquired resistance to this species (Behnke, 1987). Nevertheless, it is during primary exposure that the adult parasite's mechanism for evading host immunity is most successful; resistance to reinfection is readily inducible even in strains which tolerate chronic primary infections (Behnke & Robinson, 1985;Prowse et al. 1979;Behnke & Wakelin, 1977;Mitchell & Munoz, 1983). The B lo and BALB congenic series of mouse strains provide a convenient means through which the effect of H-2 linked and background genes on resistance may be investigated, but a detailed knowledge of parasite survival in the original reference strains (C 57 B1 1O and BALB/c) is first required to enable economic experimentation. In this paper we report the results of experiments in which the survivorship curves of H. polygyrus were determined under low and high intensities of infection in C 57 B1 1O and BALB/c mice. We also present data for NIH and (C 57 B1 1O x NIH)F1 hybrid strains.

Animals
Syngeneic NIH, C 57 B1 1O and BALB/c mice were originally purchased from Harlan Olac Ltd and, together with the hybrid strains, were bred in the departmental animal house. The mice were housed under conventional animal house conditions with access to food and water ad libitum.

Parasite
The methods used to maintain H. polygyrus, infect mice and recover worms at autopsy have all been described previously (Jenkins & Behnke, 1977).

Statistical analysis of results
In order to allow a direct comparison between the rate of worm loss in mice given high and low intensities of infection, all experiments included groups of mice killed 14 days post-infection. The worm burdens for all the mice on day 14 and for all subsequent autopsies were then transformed into a percentage of the 14-day mean worm burden for each experimental group (mean percentage worm recovery, MPWR). Some data were presented as the mean worm recovery ( M W R ) ± S . E . M . using nontransformed, raw data.
Survival curves were fitted using the function

P=a + br t ,
where P = worm burden expressed as a percentage of the day 14 mean worm recovery, t = weeks after infection and a, b and r are constants describing the best fit for this function to each data set (see Keymer & Hiorns, 1986). The best fit was obtained using maximum likelihood methods (the package MLP: Ross, 1980) because it enabled us to assume Poisson-distributed errors essential for these data. Two curves were assessed as significantly different using a likelihood ratio test (model deviance, distributed as chi-squared), by fitting the curves individually and then a single curve to all points. The maximum likelihood for the curve with Poisson-distributed errors is:- where m i = a + br x t .
Data for particular time points post-infection were also compared using the Mann-Whitney U test (Sokal & Rohlf, 1969).

Experimental design
Nine experiments were carried out and the composition of each, including the strain and sex of mouse and the intensity of infection administered, is presented in Table 1. The number of parasites establishing in each experimental group as reflected in worm burdens 14 days post-infection, is also given. Low intensities of infection were established by giving mice 50-100 L 3 and, as can be seen from Table 1, the MWR 14 days post-infection varied from 15-0 + 1-7 (Exp. la) to 6 7 0 ± 8 0 (Exp. 4a), but there were no significant differences (Mann-Whitney U test) in parasite establishment between groups within any particular experiment. High-intensity infections were achieved by giving 250-300 L 3 and the range of MWR on day 14 was from 156-5 + 6-1 (Exp. 9a) to 280-2 + 11-2 (Exp. 6b). Again, no significant differences were detected between groups within any one experiment.

Survival of Heligmosomoides polygyrus in C i7 Bl 10 mice at low infection intensity
The survival of H. polygyrus in C 57 B1 1O mice was studied in two experiments and the results are presented in Fig. 1 (Exp. 1) and Fig. 2 (Exp. 2). Survivorship curves were fitted to the data and the values for the constants are given in Table 2. Estimated maximum survival times for worms were high, being 33-9 and .28-3 weeks respectively.

Survival of Heligmosomoides polygyrus in (C 57 £/ 10 x NIH)F\ and (B\0G x NIH)Fl hybrid mice
Five experiments were carried out in which the survival of H. polygyrus was studied in F l hybrid mice. In Exp. 1, female (C 57 B1 1O x NIH)F1 and both parental strains were given a low intensity infection. The data are summarized in Fig. 3 and it is apparent that the worms were rejected by F l mice significantly earlier than by both parental strains. In week 10, F l mice had a MPWR of 56±2-9 whereas in N I H and C 57 B1 1O mice the MPWR was 71-7 ±7-8 and 111-7 + 6-3 respectively.

The influence of the intensity of infection on survival of Heligmosomoides polygyrus
The influence of the intensity of infection on the survival of H. polygyrus in C 57 B1 1O mice was examined in Exp. 1 (Fig. 1). Analysis of the survivorship curves which were fitted to the data (Table 2) revealed that there was a significant difference between these curves (P<? 0-001). In addition, a direct comparison of the MPWR in week 35 is also significant (Mann-Whitney U-test, P = 0004).
Three experiments compared the survival of H. polygyrus in NIH mice given low or high-intensity infections. The results of Exp. 4 (Fig. 4) and Exp. 7 (Table 3) did not reveal a significant difference between the groups given low or high intensities of infection over a period of 17 weeks. T h e survivor-  ship curves fitted to the data from Exp. 2 and subsequent analysis (Table 2) showed a significant difference. No significant difference was detected when the MPWR from NIH mice killed in weeks 11, 18 and 25, given low and high-intensity infections, were compared directly using the Mann-Whitney U-test.
A very obvious and marked influence of infection intensity on parasite survival was recorded for BALB/c and Fl mice. The results from Exps 2, 8 and 9, in which low and high-intensity infections were compared in BALB/c mice, are presented in Fig. 6 and Table 3. Again the curves are highly significantly different. The intensity of infection was also a major factor determining the onset of expulsion in Fl mice. The results from Exp. 1 (male Fl, Fig. 3), Exp. 4 (female Fl, Fig. 4), Exp. 6 (male Fl, Fig. 5) and Exp. 7 [female (B10G x NIH)F1, Table  4] all concur. In each case low-intensity infections were rejected earlier than high-intensity infections.

Heligmosomoides polygyrus
The majority of our experiments were carried out in female mice but an opportunity was available to include male animals for comparison when Exp. 1 (Fl) and Exp. 9 (BALB/c) were initiated. In Exp. 1 ( Fig. la and b; a direct comparison is not illustrated) there was a significant difference in the MWR in week 10, female mice rejecting worms earlier than males (Mann-Whitney U-test, P = 0038). Likewise, male BALB/c mice were slower than female mice at rejecting H. polygyrus. A significant difference was observed in week 10 (P = 0-003) and week 15 (P = 0-0034) and also through a comparison of the entire data sets for both sexes by maximum likelihood analysis (Table 2).  Week 17 14-2 + 2-3 (5-3) 166-8 ±9-3 1-6 + 0-9 (900) 76-3 + 47-6 (630) All groups contained 6 mice.

DISCUSSION
Although H. polygyrus is recognized as a long-lived parasite, little attention has been given in the past to determining accurately the pattern of survival in syngeneic strains of mice. In previous reports it has been considered that C 57 Bl l0 mice were weak responders to H. polygyrus developing acquired immunity only after prolonged and repeated stimulation (Behnke & Robinson, 1985;Cypess, Lucia, Zidian & Rivera-Ortiz, 1977;Cypess & Zidian, 1975). The survival characteristics of primary infection adult worms in C 57 B1 1O match closely patterns observed in mouse strains where infections were recognized as chronic (Liu, 1966;Keymer & Hiorns, 1986;Behnke et al. 1987). Worm burdens at low and high infection intensities remained stable for 10 weeks before declining, slowly at first and more rapidly subsequently (Fig. 1). At low infection intensities virtually all worms had disappeared by week 35 but, surprisingly, more heavily infected mice still had a MPWR of 195±4-5 at this time. This was the only occasion on which a clear significant difference was obtained in C 57 B1 1O mice, added to which the maximum likelihood analysis also gave a significant difference between the data sets for mice with light and heavy infections. Whilst we cannot eliminate the possibility that the observation may reflect density-dependent suppression of worm loss and hence the involvement of host-protective immunity, the duration of both low and high-intensity infections and the relatively small difference between the groups led us to conclude that the effects of host immunity were minimal in this strain. It is more probable, that a relatively small effect towards the tail end of the survivorship curve simply reflected the greater probability of there being some, more long-lived (possibly robust) worms in the population administered in the heavier inoculum. We interpret the survivorship curves of H. polygyrus in C 67 B1 1O mice as reflecting essentially senility, with host immunity possibly intervening in the terminal stages, to accelerate worm loss in mice with residual worm burdens.
Survival of H. polygyrus in NIH and BALB/c mice showed quite different characteristics. At low infection intensities the maximum life-span was calculated as 27-5 weeks in NIH mice and 10-4-154 weeks in BALB/c mice ( Table 2). The shape of the curve which was fitted to the function defining survivorship and the constants which were calculated for NIH mice, showed broad similarity to C 57 B1 1O mice except that parasite longevity was severely curtailed. In contrast, BALB/c mice gave quite different constants and some of the curves which we calculated were almost linear in outline ( Fig. 6 and Table 2) whilst others were clearly convex in shape (Fig. 2). Other functions were not evaluated in the present study but it is apparent that the longevity of H. polygyrus in NIH and BALB/c mice was markedly affected by factors which accelerated the expected rate of worm loss from senility alone. Itis likely that host-protective anti-adult worm responses intervened in these strains to curtail parasite survival prematurely. An immunological analysis was undertaken as part of the present investigation and our results will be published elsewhere, but two features of the data presented hereinsupport the conclusion that immunity played a significant role in abbreviating parasite survival in BALB/c mice.
Firstly, it was apparent that in BALB/c mice heavier infections persisted considerably longer than low-intensity infections. Even at low infection intensities a correlation can be seen between initial parasite establishment (Table 1) and parasite survival (Table 2 and Fig. 6). The lightest infection was established in group 9 a (Table 1) and the maximum survival was calculated as only 10-4 weeks. These observations confirm earlier reports of dosedependent suppression of worm expulsion (Dobson et al. 1985;Behnke et al. 1987) and, as proposed by the previous authors, the phenomenon probably reflects the quantitative effect of immunomodulatory factors produced by adult worms as part of the mechanism used in evading host immunity. It fol- lows that BALB/c mice, whilst initially affected by immunomodulation, nevertheless had the capacity to overcome the parasite and, depending on the intensity of infection, eventually developed insensitivity to parasite immunomodulatory factors and expelled the worms. Broadly similar results have been reported for Trichinella spiralis in rats and mice, but on a shorter time-scale (Bell, McGregor, Woan & Adams, 1983 ;Bell, Adams & Ogden, 1984;Wassom, Dougherty, Krco & David, 1984;Wakelin, Donachie & Grencis, 1985). It has been proposed that a population of T lymphocytes, controlled by genes located both within and outside the MHC, is particularly sensitive to the suppressive influence of heavy parasite burdens (Wakelin et al. 1985) and hence heavier infections are expelled more slowly than low-intensity infections.
Secondly, infections in male mice lasted longer than those in female mice, and the possibility arises of a direct detrimental effect of sex hormones on parasites residing in female mice. However, the modulatory effects of sex hormones on host immunity are well recognized in the literature (Dobson, 1961(Dobson, , 1982Solomon, 1969;Blazkovec, Orsini & Maginn, 1973). It is likely that the longer lasting infections in male BALB/c mice reflected slower development of immunity.
Two hybrid strains were investigated during this study. When C 57 B1 1O (H-2 b ) and NIH (H-2i) mice were crossed, their F t progeny (H-2 b( i) showed a considerable advantage over both parental strains. Worm expulsion commenced within 10 weeks of infection in Fl mice and in some experiments significantly accelerated rejection was identified by weeks 4-6 (Fig. 5). Clearly, gene complementation was involved, the unique combination of genes resulting from the cross, endowing Fl mice with the capacity to overcome H. polygyrus in a comparatively short period of time. In a single experiment we examined (B10G x NIH)F1 mice, a hybrid combination which would possess the H-2 q haplotype of both parental strains. Although the 2 hybrid strains were not compared directly, (BlOGxNIH)Fl mice behaved much like (C 57 B1 1O x NIH)F1, no obvious advantage of H-2<i over H-2 b( i being apparent. Both hybrid strains were affected by the intensity of infection and, as in BALB/c mice, worm expulsion began later in heavily infected animals. Male (C 57 B1 1O x NIH)F1 mice expelled worms more slowly than female mice (Fig. 2), a result which is also consistent with an immunological explanation for the short duration of primary infections in these animals.
The demonstration that the Fl hybrid mice exhibited considerably enhanced resistance to infection compared to either parental strain deserves further comment. NIH mice are recognized as being capable of generating rapid inflammatory responses in the intestine whereas C 57 B1 1O mice are slower in this respect (Alizadeh & Wakelin, 1982;Wakelin, 1980). However, C 57 B1 1O mice develop better host-protective antibody responses to H. polygyrus after repeated larval challenge (Williams & Behnke, 1983). These mechanisms may interact synergistically in Fl mice to bring about better overall protection but, as yet, this has not been confirmed experimentally. Hybrid vigour is recognized in a variety of different contexts as endowing increased fitness in competition and survival and gene complementation is a well-established phenomenon among micro and macro-parasites (reviewed by Wakelin & Blackwell (1988)). In our system, NIH and C 57 B1 1O mice arose independently through inbreeding programmes for which the selection criteria are not clear, and the two strains probably carry many contrasting alleles. If we consider the two strains as analogues of different genotypes available in wild populations, then our results may have broader implications. The reasons for the existence of susceptible individuals in wild host populations are a subject for contemporary debate (Schad & Anderson, 1985;Anderson, 19866;Bundy, Cooper, Thompson, Didier, Anderson & Simmons, 1987) but one possibility may be that other advantages balance or exceed the net disadvantage of susceptibility to a particular disease (e.g. reproductive potential or possibly, as in our case, an antibody response). Furthermore, heterozygous animals may be at an overall advantage in carrying both the resistant genes and the genes for high reproductive potential or other advantageous traits. Thus, the maintenance of genetic polymorphism at loci governing responsiveness, including the traits for susceptibility, is assured and the recurrence of susceptibles among future progeny, guaranteed. The mathematical basis for this hypothesis has been reviewed by Anderson (1986a).
The involvement of immune responses in limiting primary infections with H. polygyrus is still a controversial subject, there being little evidence for the participation of acute inflammatory reactions such as those initiated by T. spiralis and considered to play a crucial role in terminating parasitic infections in the intestine (Miller, 1984;Wakelin, 1985). Indeed, H. polygyrus is known to suppress mastocytosis, a key component of the intestinal inflammatory response, even in mouse strains such as SJL in which primary infections seldom last beyond week 10 (Dehlawi, Wakelin & Behnke, 1987;Dehlawi & Wakelin, 1988). However, our experiments provide support for the role of immunity. The variations in the survival characteristics of H. polygyrus which we have observed between mouse strains, in relation to the intensity of infection and sex of the host, cannot be explained easily, other than through the complex interaction between host protective responses against adult worms, parasite immunomodulation and host genes controlling the ability of mice to generate the former and/or resist the latter. The next step will be to define the component processes involved, but this will not be an easy task as attempts to correlate worm loss with cellular and antibody activity during primary infection have been largely unsuccessful (Behnke, 1987;Behnke etal. 1987;Williams & Behnke, 1983). Nevertheless, the information which we have provided on survivorship curves of H. polygyrus in C 57 B1 1O (BIO) and BALB/c mice should enable the BIO and BALB congenic mouse series to be exploited in identifying the host genes governing susceptibility and resistance to primary infection. The uniformity of background genes within each of these mouse series should allow M H C effects on parasite survival to be isolated and the accompanying MHC-linked immune responses to be recognized.