INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The mechanisms by which hosts kill parasites are complex, and the role of eosinophils in this process is contentious. To study this question, transgenic mice which constitutively overexpress interleukin-5 (IL-5) and become eosinophilic (6), IL-5 gene knockout mice (24), and mice treated with monoclonal antibodies to IL-5 (4) are interesting tools of investigation. These protocols were used in diverse experimental helminthiases (24).

In filariases, previous works, including two studies with antibodies to IL-5, argue for a role of eosinophils in the killing of filariae at their different stages: the microfilariae in vitro (11, 14) and in vivo (9), third- and fourth-stage larvae in vivo (3, 17, 38), and adult worms in vivo (23). However, no study has so far been performed with IL-5 transgenic mice.

We report the study of a filarial infection in IL-5 transgenic CBA/Ca mice. None of the diverse species which parasitize humans develops in laboratory mice. We used the species Litomosoides sigmodontis, a rodent filaria which, as a member of the Onchocercidae family, presents fundamental biological features similar to those of human filariae. It is the sole filaria to undergo full development in immunocompetent BALB/c mice (13, 27). It has also been shown to mature in CBA/Ca mice; however, these mice do not become microfilaraemic even though microfilariae are present in the uteri of the female worms, sometimes in high density (27).

The filarial infection in IL-5 transgenic mice was successively analyzed at three postinoculation (p.i.) time points: late (day 60 p.i.), early (day 10 p.i.), and intermediate (day 30 p.i.). This schedule was chosen according to our knowledge of the biology of L. sigmodontis in susceptible BALB/c and resistant B10.D2 mice (21). As in other filaria-host pairs (2), the life cycle begins with a phase of intense destruction of larvae within the first 2 days p.i. Only the larvae that penetrate the lymphatic vessels escape the inflammatory process induced at the site of inoculation. Thereafter, the filarial recovery rate remains stable for a long period, 60 days in BALB/c mice and 40 days in B10.D2 mice, and its level characterizes each host. Third and fourth moultings (on days 8 to 10 and 26 to 28 p.i., respectively) and adult maturation occur in the coelomic cavities, mainly the pleural. The patent phase begins in BALB/c mice on day 55 p.i. Antibody responses, which are of the Th2 type all during the course of infection in wild-type mice (20, 22), were analyzed on day 60 p.i. by measuring serum immunoglobulin G1 (IgG1) and IgG2a levels to investigate a possible alteration in the T helper balance in transgenic mice. IgE has not been studied because murine eosinophils do not express cell surface receptors that bind IgE (5) and a previous study showed that the course of L. sigmodontis infection was similar in transgenic mice overexpressing IgE and in wild-type mice (P. Maréchal, L. Le Goff, and O. Bain, unpublished data).

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Mice and infection protocol. The CBA/Ca wild-type mice were supplied by Harlan Olac. The IL-5 transgenic mice were a kind gift from B. B. Vargaftig (Pasteur Institute). These CBA/Ca IL-5 transgenic mice were generated using a transgene which imparts constitutive expression of IL-5 to CD2⁺ cells. In such transgenic mice, blood eosinophil counts reach 44 or 68%, according to the transgenic line studied, versus 2 to 5% in wild-type mice (6). Three experiments were performed. Female mice were used except for the histopathological analysis, which comprised five male and one female transgenic mouse. A prior experiment with male wild-type CBA/Ca mice (27) showed that the L. sigmodontis location, fertility, and recovery rate were similar to those in the females. Each mouse was inoculated subcutaneously in the right lumbar area with a single dose of 40 L. sigmodontis infective larvae, which were harvested by dissection of the mite vector, Ornithonyssus bacoti.

White blood cell identification. Smears of tail blood, made before inoculation, on days 10, 20, 30, and 60 p.i. were stained with May-Grünwald-Giemsa. The percentages of eosinophils, neutrophils, lymphocytes, and monocytes were determined (200 leukocytes were counted per blood sample). When a spherical aggregate of host cells was attached on a freshly recovered larva at necropsy, it was similarly stained with May-Grünwald-Giemsa.

Dissection of mice. Necropsies were performed 10 and 60 days p.i. in RPMI 1640 supplemented with 20% calf serum, in which larvae survive well, according to the procedure described previously (2, 21). Despite the fine dilaceration of tissues, some larvae escape observation at early time points, and the recovery rate is slightly lower than later on (18, 21). The location, motility, and aspects of the filariae were noted. Filariae were harvested, counted, and fixed in hot 70% ethanol for morphological analysis. Filarial development and protection were evaluated by means of the following parameters: (i) percentage of mice with filarial worms; (ii) recovery rate of filariae: [(number of worms recovered)/(number of larvae inoculated)] × 100 (F/L3); (iii) number of live worms partially surrounded by inflammatory cells (these worms were used to calculate the recovery rate); (iv) number of dead worms or pieces of worms in granulomas, called cysts in previous papers (20, 22, 27) (these were not included in the recovery rate); (v) size of worms; (vi) stage of worms; (vii) sex ratio of recovered worms (number of females/total worm burden); (viii) blood microfilaremia (number of microfilariae/10 mm³) determined at day 60 p.i. on a 10-mm³-thick blood smear stained with Giemsa; and (ix) percentage of mice with blood microfilariae. Percent protection is evaluated as [F/L3 (wild type  transgenic) × 100]/F/L3 wild type. Clinical particularities of the mice were also noted: enlarged lymph nodes, splenomegaly, ascites, etc.

Histopathological technique. On day 30 p.i., parasitological analysis was done on histological sections. IL-5 transgenic mice were fixed in toto in 10% formalin. Viscera were taken from the pleural cavity and prepared according to the histopathological procedure described previously (36). A few serial 5-µm-thick sections were cut and stained with hematoxylin-eosin-safran. Both the parasites and the lesions were located and identified.

Analysis of IgG1 and IgG2a production in serum by ELISA. Levels of parasite-specific IgG1 and IgG2a were measured by enzyme-linked immunosorbent assay (ELISA) in blood samples collected on day 60 p.i. Microtiter plates (Nunc, Roskilde, Denmark) were coated with L. sigmodontis adult extracts in 0.01 M bicarbonate buffer for 4 h at 4°C. The plates were blocked overnight at 4°C with casein (25 mg/ml) in phosphate-buffered saline (PBS)-0.05% Tween 20. The mouse serum was diluted 1:100 in PBS-2% bovine serum albumin (BSA)-0.05% Tween 20, and 100 µl was added to each well. The plates were incubated for 2 h at 37°C. Alkaline phosphatase-conjugated anti-mouse immunoglobulins (anti-IgG1 and anti-IgG2a; Sigma; 1:1,000 in PBS-2% BSA-0.05% Tween 20) were added, and the plates were incubated for 2 h at 37°C. Between each incubation step, the plate was washed three times in PBS-2% BSA-0.05% Tween 20. After the addition of p-nitrophenylphosphate (1 mg/ml in diethanolamine buffer), the optical density was determined photometrically in an ELISA reader at 405 nm (Labsystems Multiskan MS). A kinetic analysis of the enzyme reaction was performed, and the slope was calculated. The relative immunoglobulin concentration linked to the slope of the linear part of the curve was expressed in arbitrary units.

Statistical analysis. Nonparametric tests were used to assess non-normally distributed parameters: (i) the Mann-Whitney U test to compare filarial recovery rates, filarial sizes, percentages of fourth-stage larvae, and IgG1 and IgG2a levels and (ii) the Wilcoxon test to compare eosinophil percentages in paired series. P < 0.05 was considered significant.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Evolution of blood leukocyte counts. Wild-type CBA/Ca mice presented with 5.5% blood eosinophils, 18.25% neutrophils, 74.5% lymphocytes, and 1.25% monocytes before infection (n = 6). On day 10 p.i., the eosinophil and neutrophil levels doubled (Wilcoxon, P = 0.028 and 0.046, respectively) and the lymphocytes decreased (Wilcoxon, P = 0.046).

IL-5 transgenic mice show a high reduction in the filarial recovery rate 60 days p.i. The filarial recovery rate was 7.5 in IL-5 transgenic mice, compared to 51.25 in wild-type mice (Mann-Whitney U test, P = 0.02). Thus, protection reaches 65%. Granulomas were present in three of four IL-5 transgenic mice but absent in wild-type mice. The other parasitological parameters (size, absence of blood microfilariae, and percentage of mice presenting uterine microfilariae) were similar in both groups of mice (Table 1).

This reduction occurs in the pleural cavity of the IL-5 transgenic mice between 10 and 30 days p.i. On day 10 p.i. (Table 2), the filarial recovery rate was similar in wild-type and IL-5 transgenic mice (26.25% versus 25%), regardless of whether the transgenic mice were eosinophilic (27.5%) or not (22.5%). Growth is not important during the third stage; however, in IL-5 transgenic mice, both male and female larvae were longer than in wild-type mice (Mann-Whitney U test, P = 0.007 and 0.01, respectively), and nearly all of them were already at the fourth stage. In both groups of mice, larvae did not present any sign of damage, and all of them were very motile. However, during the necropsies, a few larvae were found with a spherical aggregate of host cells: six larvae from two IL-5 transgenic mice and four from four wild-type mice (Fig. 1A). These aggregates were composed of mainly degranulated eosinophils and macrophages in similar proportions; most eosinophils were juvenile (band cells); no neutrophils were identified (Fig. 1B). These cellular aggregates were localized at the end of the anterior third of the larva body (Fig. 1A). This place does not correspond to any specific structure of the worm (the excretory pore is much more anterior), and it might result from mechanical factors such as streams created by the active undulations of the larvae. They were no longer identified on larvae 1 h after incubation in dissection medium.

View larger version (126K): [in this window] [in a new window]   FIG. 1.   (A) Host cell aggregate on a live and motile larva recovered 10 days p.i. from the pleural cavity of a wild-type CBA/Ca mouse (h, head of the larva). (B) Photograph of a host cell aggregate showing degranulated eosinophils (the arrow indicates eosinophilic granules) from a transgenic mouse 10 days p.i. (C) Young adult filaria (F) severely damaged in a pleural granuloma and surrounded with Splendore-Hoeppli deposits (arrows), 30 days p.i., in a transgenic mouse. (D) Peribronchial proliferation of malignant lymphoma (arrows) in a transgenic mouse. 1 cm = 60 µm (A), 6.2 µm (B), or 85 µm (C and D).

Serum IgG1 and IgG2a levels. On day 60 p.i., the IgG2a levels were similar in IL-5 transgenic and wild-type mice (medians, 1.46 versus 1.14; Mann-Whitney U test, P = 0.55). The IgG1 levels were not significantly different (medians, 40.8 versus 17.25; Mann-Whitney U test, P = 0.67). However, even though dispersion is great in both groups, the respective medians indicate a tendency in IL-5 transgenic mice to produce more IgG1.

IL-5 transgenic mouse pathology. All IL-5 transgenic CBA/Ca mice presented ascites, pleural effusions, and splenomegaly. On sections taken on day 30 p.i., malignant lymphomas were identified in the spleen, the mesentery, and the lungs (Fig. 1D) but not in the liver; these neoplastic processes were often very extended and destroyed the tissue structure. Previous histological observations of wild-type CBA/Ca mice similarly infected with L. sigmodontis did not show any lymphoma.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Our study demonstrates protection 60 days p.i. in IL-5 transgenic mice and argues for a role of IL-5 and eosinophils in the establishment of this protection (Table 1). The filarial recovery rate on day 60 p.i. has fallen strongly in IL-5 transgenic mice compared with wild-type mice (about sevenfold decrease). In other filarial infections where mice were treated with monoclonal antibodies to IL-5, results also argue for an effector role of eosinophils in killing filariae (Table 3): the IL-5 level diminution leads to a longer survival of Onchocerca lienalis microfilariae (9) and to a lower protection in an Onchocerca volvulus vaccination protocol with irradiated larvae (17).

With the other helminths, results are heterogeneous depending on the parasite species and its biology (Table 3). Furthermore, the data for the use of monoclonal antibodies to IL-5 seem to contradict those from transgenic mice (4, 12, 29, 35) and gene knockout mice (24, 25). In these cases, an allogenic response may have been induced by the monoclonal antibody treatment, explaining the inhibition of bioactivity (24). Among all these studies, evidence for a protective role of IL-5 and eosinophils is only provided with intestinal nematodes of the Secernentea group (Table 3): Nippostrongylus brasiliensis infection decreases in IL-5 transgenic mice (8), and Heligmosomoides polygyrus infection increases in IL-5 knockout mice (24). However, no proportional correlation appears between IL-5, eosinophils, and parasitism, as observed with N. brasiliensis; the reduction in worm load and fecundity is similar whatever the IL-5 level in the different strains of transgenic mice (8). These nematodes have a very fast cycle, and, as emphasized by Dent et al. (8), "this makes unlikely any contribution from the humoral arm of the adaptive immune response."

On the contrary, with the tissue-dwelling nematode L. sigmodontis, no such rapid efficiency has been observed: the filarial recovery rate was still not reduced 10 days after larva inoculation (Table 2). We even noticed in the IL-5 transgenic mice a slight acceleration of larval growth at day 10 p.i. (increase in larval size and higher proportion of fourth-stage larvae; Table 2). This effect may be due to IL-5, as in recent experiments in which we used an IL-5 monoclonal antibody, the growth of parasites in treated animals was conversely retarded, as observed 28 days p.i. (C. Martin, K. Al Qaoud, O. Bain, M. N. Ungeheuer, P. N. Vuong, K. Paehle, B. Fleischer, and A. Hoerauf, submitted for publication). Such a relation between a cytokine and parasite growth was recently observed in an IL-7-deficient mouse schistosomiasis model, perhaps involving thyroid hormones (37). These data show that the processes that regulate parasite development are complex and may vary with the parasite stage. Incidentally, it was also noted that the lesions in transgenic mice, such as lymphomas (Fig. 1D), do not affect larval migration.

In IL-5 transgenic mice, the protection seems to be established 1 month p.i. according to the proportion of damaged filariae (41%). However, host cells (eosinophils and macrophages) are recruited earlier and adhere to some larvae in IL-5 transgenic and wild-type CBA/Ca mice (Fig. 1A). This adhering process is reversible in vitro on day 10 p.i. Dead worms in granulomas are found only in transgenic mice on day 60 p.i., showing that this process becomes intense enough in these mice because of the large numbers of eosinophils. The protection delay in IL-5 transgenic mice seems to correspond to the time necessary to produce host antibodies that will permit eosinophil degranulation (32). Indeed, previous works on this filarial model showed that the different classes of serum immunoglobulins were first detected 3 weeks after the challenge inoculation whether the mice were resistant or susceptible (1, 22), so a similar lapse of time is expected to be required in CBA/Ca mice before protection occurs. Release of the cytotoxic eosinophil substances in response to L. sigmodontis is illustrated by the presence, around young adult filarial worms, of Splendore-Hoeppli deposits (Fig. 1C), which consist in part of eosinophil major basic protein granules (10, 34).

The high protection obtained in IL-5 transgenic CBA/Ca mice, which still remains partial, is similar to that observed following vaccination with irradiated larvae in BALB/c mice. However, in vaccinated mice, vaccination is carried out within the first 2 days following the challenge inoculation. Thus, it leads to the destruction of infective larvae in the subcutaneous tissue (18). Indeed, during vaccination, specific antibodies are produced, IL-5 secretion is increased, and consequently eosinophils are present at a high density in the subcutaneous tissue. These eosinophils degranulate within the 6 h following the challenge inoculation, and they magnify the nonspecific inflammatory process normally induced after a single inoculation of larvae (19, 20).

Although the kinetics of protection in IL-5 transgenic mice differ from that previously obtained by vaccination, the mechanisms for this protection are likely to be similar: IL-5 production and eosinophilia are similarly increased in both situations. This strongly suggests that the same mechanisms are acting by the means of degranulation of the antibody-coated eosinophils. The antigen-specific IgG1, which tends to reach higher levels in transgenic mice than in wild-type mice, could be responsible for the degranulation of these leukocytes, as was evidenced in humans (15).