INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The gastrointestinal tract is continuously exposed to environmental antigens, which may include potential pathogens and other noxious stimuli (7). Usually the mucosal immune system can differentiate between useful and harmful antigens encountered in the lumen (32) and initiate the appropriate immune response. In this regard, most studies of enteric infections have focused on the direct effector actions taken by the immune system against invading pathogens. However, there is growing evidence that during such infections, the host develops a complex and integrated response involving the coordinated actions of all the tissues in the gastrointestinal tract. Normally passive physiological tissues are recruited by the immune system, which alters their function so that they can actively aid in host defense (7, 33). This immune regulation of physiological function occurs following infection with gastrointestinal nematode parasites such as Trichinella spiralis and results in increased fluid secretion into the lumen of the small bowel (3, 33) as well as increased intestinal propulsive activity and more rapid intestinal transit (1, 4, 7, 43). These changes in motility occur in association with a hyperresponsiveness of jejunal longitudinal muscle (11, 46, 48) and have been hypothesized to play a role in the expulsion of the parasites (7, 45). This is supported by the observation that both processes are attenuated during infection of athymic rats (19, 40, 47).

While T lymphocytes are undoubtedly major contributors to these responses, it is unclear how close a relationship exists between worm expulsion and enteric muscle function during a primary T. spiralis infection. Both CD4⁺ and CD8⁺ lymphocytes infiltrate the jejunal muscle layers during the early stages of T. spiralis infection in rats (unpublished observation), and both T-cell subsets are capable of mediating tissue damage, either directly or indirectly (15, 25). Thus, it is unclear which subset mediates the changes in muscle function. Several studies have shown that CD4 T cells taken from infected rodents can adoptively transfer protection against primary infection to naive animals (16, 24, 37). This protective effect from memory CD4 T cells is potentially quite different from the role that newly recruited CD4 T cells would play during a primary infection. Therefore, the respective roles of newly recruited CD4 and CD8 T cells in both worm expulsion and muscle function during primary infection need to be further characterized.

Equally important, CD4 T-cell activation and differentiation require their interaction with antigen-presenting cells (APC), and the role and identity of various APC within the intestine during nematode infections is not clearly understood. Major histocompatibility complex class II (MHC II) expression within the noninflamed intestine is generally limited to professional APC found within the organized lymphoid aggregates, within the Peyer's patches, or scattered within the lamina propria (10, 14). Some limited expression by intestinal epithelial cells also occurs (6, 31). However, during nematode infection the pattern of MHC II expression within the intestine is altered (28). More professional APC are recruited to the gut, and nonimmune cells such as smooth muscle cells (21, 22) and glial cells (2) may be stimulated by inflammatory cytokines to express MHC II and possibly play an active role in antigen presentation (2, 21, 22).

Our previous work primarily examined muscle function and worm expulsion in the infected rat (47, 48). However, to examine the questions concerning T-cell subsets, we switched experimental species to the mouse, in order to take advantage of the gene-targeted immunodeficient mouse strains currently available. CD8-chain-deficient (13), CD4-deficient (35), and MHC II-deficient (C2d) mice that lack CD4 T cells and class II antigen presentation (18) were infected with T. spiralis as a first approach. This allowed us to identify the critical role of CD4 T cells and MHC II expression in the physiologic responses of increased muscle tension and in parasite expulsion. These results may have implications not only for immunophysiological interactions during enteric infections but also for the maladaptive changes in physiologic function found in disease states involving the gastrointestinal tract.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Mice. Euthymic and athymic (nu/nu) C57BL/6 mice were purchased from Taconic (Germantown, N.Y.), as were mice lacking MHC II (class II^/), which were originally produced by targeted gene mutation as described by Grusby et al. (18). These mice were later backcrossed onto the C57BL/6 background five times (and thereafter designated C2d mice). A colony of C2d mice has been continuously bred in the central animal facility at McMaster University. CD4-deficient (35) and CD8-deficient (13) mice (also backcrossed onto a C57BL/6 background) were obtained from Jackson Laboratory and bred at McMaster University. Only male mice (aged 6 to 10 weeks) were used in this study, and all animals were kept under specific-pathogen-free conditions. Since the original mixed-background MHC II-deficient mice were described to spontaneously develop colitis by 4 to 6 months of age (30), we examined our C2d mice for any signs of intestinal inflammation or other pathologies. Colitis was not observed in these mice until at least 6 months of age and was observed only in a small percentage of the animals, indicating that spontaneous intestinal inflammation should not be a factor in our studies.

CD4 T-lymphocyte purification. C57BL/6 mice were euthanized, and their mesenteric, brachial, and inguinal lymph nodes were removed, placed in RPMI medium containing 10% fetal bovine serum (RPMI-10), and crushed between two sterile glass slides. The resulting single-cell suspension was placed in covered culture dishes in a 37°C incubator for 2 h. Nonadherent cells were then removed and centrifuged over Ficoll Hypaque (Pharmacia, Uppsala, Sweden) for 20 min at 500 × g. The mononuclear cells were then collected from the interface, washed twice with medium, and resuspended in a small volume of Hanks' balanced salt solution. Cells were then counted and incubated for 20 min at 4°C with magnetic beads (Miltenyi Biotec Inc., Sunnyvale, Calif.) labeled with rat anti-mouse CD4 monoclonal antibody (clone GK1.5). The cells were then passed through a VS-type Minimacs column (Miltenyi) previously washed with column buffer (phosphate-buffered saline [PBS], 2 mM EDTA, 0.5% bovine serum albumin [BSA] [pH 7.2]) and chilled to 4°C. The unbound cells were allowed to pass through the column, which was then washed three times with 3 ml of column buffer. The column was then removed from the magnet, and the purified CD4-positive cells were eluted with 5 ml of column buffer. The isolated cells were then counted and analyzed for viability, which was found to be >95%, by trypan blue exclusion. Flow cytometric analysis confirmed previous studies (39) showing >90% purity of CD4 T cells following this procedure.

Production of bone marrow chimeras. The method for production of bone marrow chimeras was adapted from that previously published (27). In brief, recipient C2d MHC II-deficient mice received 9.50 Gy of gamma irradiation delivered from a ¹³⁷Cs source. Donor C57BL/6 mice were euthanized by cervical dislocation, and their femurs were removed and flushed with Hanks' balanced salt solution. The bone marrow cells were then pooled, washed and counted by using trypan blue exclusion. At 2 to 5 h after irradiation, recipient mice were reconstituted by tail vein injection of 20 million viable bone marrow cells in sterile PBS and then given Septra in their drinking water ad libitum, for 2 weeks, until the bone marrow graft had taken. There was a 90% survival rate for mice undergoing this procedure. Mice given both CD4 T cells and bone marrow underwent the same procedure, except that 3 million to 5 million positively selected CD4 T cells were mixed with the bone marrow cells and given to recipient mice in one injection. The mice were subsequently infected, and used for experiments, 6 to 8 weeks following reconstitution.

Flow cytometry analysis. Flow cytometry staining for MHC II expression and CD4-positive T cells was performed on the peripheral blood of uninfected C2d, C57BL/6, or chimeric mice (4 to 6 weeks after reconstitution), as well as on a mixed population of spleen and mesenteric lymph node (MLN) cells isolated from various mice infected 8 days previously with T. spiralis. Blood lymphocytes were isolated from 200 µl of EDTA-treated blood centrifuged on Ficoll. Spleen and MLN cells were isolated as described above for CD4 T-lymphocyte isolation. The isolated blood cells were incubated with fluorescein isothiocyanate-labeled anti-IA^b monoclonal antibody (clone M5) prepared in our laboratory and phycoerythrin (PE)-labeled anti-CD4 (clone GK1.5 [Pharmingen]) or biotin-labeled anti-CD4 (Pharmingen) plus R-PE-labeled streptavidin (Molecular Probes, Eugene, Oreg.). Isolated spleen plus MLN cells were stained with fluorescein isothiocyanate-labeled anti-CD3 (clone 145-2C11 [Pharmingen]), PE-labeled anti-CD4, and biotin-labeled anti-IA^b (our laboratory) plus Cychrome-labeled streptavidin (Pharmingen). The cells were then fixed in 1.0% paraformaldehyde and analyzed by flow cytometry with a FACScan instrument (Becton Dickinson, San Jose, Calif.). Data analysis for CD4⁺ T cells and MHC II-expressing cells was performed with the PC-lysys software (Becton Dickinson), by gating on lymphocytes with forward- and side-scatter parameters. A minimum of 10,000 events were collected within the lymphocyte scatter gate.

Parasites and infection of mice. The T. spiralis nematode parasites used in this study originated in the laboratory of S. Desser, Department of Zoology, University of Toronto, and the colony was maintained through serial infections alternating between male Sprague-Dawley rats and male CD1 mice at McMaster University. Muscle larvae were isolated 30 to 90 days postinfection by a modification of the technique described by Castro and Fairbairn (5). Mice were infected by administration of 0.1 ml of PBS containing 350 to 400 T. spiralis larvae by gastric gavage. To minimize differences between infections, control and gene-deficient or reconstituted mice were infected from the same parasite preparation whenever possible.

Worm counts. The entire length of the small intestine was removed and, for ease of counting, divided into four equal sections. All the adult worms within a given section were then counted by a modification of a previously described method (5). Briefly, the mucosa was separated from the underlying muscularis by scraping with a glass microscope slide and mixed with 1 ml of PBS. The worms were then counted by using a scored petri dish and an inverted microscope. The worm counts in the four sections were added and expressed as the total number of worms per mouse. In accordance with published norms (49), worm rejection was considered complete when at least 98% of the maximum worm burden had been expelled from the gut.

Immunohistochemistry. Tissues were rinsed with ice-cold PBS, embedded in OCT compound, frozen with isopentane and liquid nitrogen, and stored at 70°C. Serial sections were cut at a thickness of 6 µm and fixed in ice-cold acetone for 10 min. To detect MHC II expression, the biotinylated monoclonal rat anti-mouse MHC II antibody (clone 25-9-17) (18) (Gibco) was used. Immunostaining for macrophages was performed with a rat anti-mouse monoclonal antibody (clone Cl:A3-1) specific for the macrophage marker F4/80. Immunohistochemical staining was performed by the streptavidin-biotin-peroxidase complex method, and the antibodies were diluted in 1% BSA in Tris-buffered saline (pH 7.2) (TBS). In brief, immunostaining was performed as follows. Tissue sections were washed twice in TBS, and endogenous peroxidases were blocked by submerging the slides in 1% H₂O₂ (in TBS) for 30 min. Sections were again washed twice with TBS and blocked with 1% BSA in TBS for 30 min. BSA was then washed off with another two TBS washes. Biotinylated anti-MHC II was then added to the slides at a 1:300 dilution, or anti-F4/80 was added at a dilution of 1:50 and left overnight at room temperature. The next morning, the antibody was washed off and a secondary goat anti-rat antibody (Cedarlane) was added for F4/80 staining. After 1 h, streptavidin-horseradish peroxidase conjugate (1:300 dilution), also purchased from Gibco, was added to the slides (for both MHC II and F4/80 staining), which were then incubated for 30 min. Again, after two washes, the antibodies were visualized with 3-amino-9-ethylcarbazole. Red color development was quenched by immersion of the slides in tap water followed by counterstaining with hematoxylin. Photomicrographs were taken with a Zeiss camera.

Measurement of muscle contraction. The preparation of the jejunal longitudinal muscle sections for muscle contractility analysis and the analysis of the length-tension relationships have been described previously (46). In brief, the jejunum was removed and placed in oxygenated (95% O₂, 5% CO₂) Kreb's solution and 1-cm sections of whole gut were cut from the jejunum, beginning at the ligament of Treitz and proceeding distally. The lumen of each segment was flushed with Krebs buffer prior to the insertion of short (2- to 3-mm) lengths of Silastic tubing (outer diameter, 0.065 in.; inner diameter, 0.030 in.) (Dow Corning, Midland, Mich.) into the open ends of the gut segments. The tubing was then tied in place with surgical silk. Insertion of the tubing was found to maintain the patency of the gut segments over the course of the experiments. Segments were then hung in the longitudinal axis and attached at one end to an FT03C force transducer (Grass, Quincy, Mass.), and responses were recorded on a Grass 7D polygraph. Tissues were equilibrated for 30 min at 37°C in Kreb's solution oxygenated with 95% O₂-5% CO₂ before the experiment was started. Experiments were then conducted to examine the length-tension characteristics of the muscle before and after infection. The gut segments were stretched until they reached a point where pen movement was just detectable on the polygraph. This stretch was defined as zero tension. Tissues were tested beginning at this point, and the tension was increased in 250-mg/mm² stages up to a maximum of 1,250 mg/mm². Contraction was then assessed following stimulation with 1 µM carbachol (Sigma Chemical, St. Louis, Mo.). Initial experiments indicated that this tension range was sufficient to determine the maximal responsiveness of both control and inflamed tissues. After each application of tension, the length of the tissue and the contractile response were recorded. The tissues were then rinsed twice and equilibrated at the next tension levels in fresh Kreb's buffer for 15 min, prior to the next addition of carbachol. At the end of each experiment, tissue segments were removed, blotted, and weighed, and the optimal tension (TO) and the tissue length giving the maximum contractile response were used to calculate the cross-sectional area of the tissue.

Data presentation and statistical analysis. Responses to carbachol were recorded from tracings, and this was followed by the calculation of contractile activity, expressed as milligrams of tension per unit of cross-sectional area as described previously (46). For each mouse, the mean tension was calculated from at least three segments. All the results are expressed as the means ± 1 standard error of the mean (SEM). Statistical significance was calculated by Student's t test for comparison of two means or a one-way of analysis of variance for the comparison of three or more means. Multiple comparisons were performed by the Newman-Keuls multiple comparison test. P < 0.05 was considered significant.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Euthymic C57BL/6 mice develop increased muscle contraction during T. spiralis infection. We infected euthymic C57BL/6 mice and studied the kinetics of jejunal longitudinal muscle contraction in response to 1 µM carbachol. The muscle response significantly increased by day 6 postinfection (p.i.) and peaked by day 8 p.i. at levels three times higher than those of tissues from uninfected control mice. A significant increase in muscle contraction was maintained until at least day 21 p.i. (Fig. 1).

View larger version (12K): [in this window] [in a new window]   FIG. 1.   Time course of the contractile response to 1 µM carbachol (the y axis shows the maximum tension generated) by intestinal muscle taken from infected C57BL/6 mice. Data shown are the mean ± 1 SEM of groups of four to six animals. The asterisk denotes a significant increase in tension compared to tissues from uninfected mice (time = 0 days).

Athymic, CD4-deficient, and C2d mice have attenuated muscle function during T. spiralis infection. Euthymic, athymic, CD4-deficient, CD8-deficient, and C2d mice were infected with T. spiralis to determine the impact of their respective immunodeficiencies on muscle contractility. Intestinal tissues were tested on day 8 p.i., the time of maximum increase in contraction by intestinal muscle in euthymic C57BL/6 mice during T. spiralis infection. All five strains of immunodeficient mice showed similar carbachol-induced muscle contractions (mean range, 800 to 1,000 mg/mm²) prior to infection (Fig. 2A). However, while tissues from normal C57BL/6 mice again demonstrated a large increase in contraction on day 8 p.i., tissues from athymic mice had very little increase in muscle contraction (Fig. 2B), which was not significantly greater than that generated by uninfected mice. This result clearly indicated a major role for T cells in the increased intestinal muscle contraction and confirms the previous findings in the athymic rat (47). CD8-deficient mice developed a high level of muscle contraction, similar to that seen in the C57BL/6 mice. CD4-deficient mice showed an intermediate increase in muscle function that on average was somewhat less than the increase in C57BL/6 muscle function, but there was no statistically significant difference between the muscle responses of the two strains of mice.

View larger version (21K): [in this window] [in a new window]   FIG. 2.   Maximum tension generated by muscle from uninfected mice (A) or day 8 p.i. mice (B) in response to 1 µM carbachol. Mouse strains shown are euthymic C57BL/6 (solid bar), athymic C57BL/6 (open bar), CD8-deficient (hatched bar), CD4-deficient (cross-hatched bar), and C2d (finely cross-hatched bar). Data shown are the mean ± 1 SEM from groups of four to six animals. The asterisk denotes significantly lower tension generation by muscle compared to the infected C57BL/6 mice.

Athymic, CD4-deficient, and C2d mice but not CD8-deficient mice fail to expel T. spiralis. To confirm previous findings (40, 47) of a major contribution by T lymphocytes in T. spiralis expulsion, euthymic and athymic C57BL/6 mice were infected with T. spiralis larvae and sacrificed 21 days later. The time course of their worm expulsion indicated that normal C57BL/6 mice expelled more than 90% of their parasites by day 16 p.i. and were worm free by day 21 whereas athymic mice had a chronic infection with a substantial load of worms up to day 21 (Table 1). Mice with CD4 or CD8 deficiencies were also infected to determine if their immunodeficiencies would affect worm expulsion. CD8-deficient mice were devoid of adult worms by day 21, similar to normal euthymic C57BL/6 mice, but CD4-deficient mice had substantial numbers of worms on day 21 (28 ± 15 worms per mouse; P < 0.01 with respect to the value for C57BL/6 mice).

Bone marrow and peripheral CD4 T-cell reconstitution of C2d mice. Questions about the role of CD4 T cells and MHC II expression can best be answered by reconstitution of C2d mice, using bone marrow and CD4 T-cell transfer from congenic C57BL/6 mice. We generated these congenic bone marrow chimeras and validated the success of our reconstitutions by flow cytometric analysis of recipient peripheral blood and lymphoid tissues. As expected, C57BL/6 mice had many circulating MHC II-positive cells, and approximately 27% of their lymphocytes were CD4 positive (Table 2). In contrast, C2d mice had no detectable MHC II expression and were lacking almost all CD4 T cells (which made up only 1.5% of all blood lymphocytes). Reconstitution of C2d mice with C57BL/6 bone marrow led to the presence of a large population of MHC II-positive cells within the blood (predominantly B cells and monocytes); however, there was no significant increase in the proportion of CD4⁺ cells in their peripheral blood. This was predictable because the C2d thymus cannot generate CD4 T cells due to lack of MHC II expression on the thymic epithelium and consequent lack of positive selection (18, 23, 27). Following the injection of C57BL/6 bone marrow, along with purified CD4⁺ T cells, there was a partial restoration of the CD4 T-cell population within the blood, with their frequency reaching almost 4% of all lymphocytes. However, bone marrow reconstitution of C57BL/6 mice with syngeneic C57BL/6 bone marrow alone regenerated MHC II-positive cells in blood and restored the normal proportions of CD4-positive T cells (25.1% of all blood lymphocytes).

Immunohistochemistry of intestinal MHC II and F4/80 expression. Staining for MHC II on frozen sections of the small intestines of uninfected C57BL/6 mice revealed MHC II staining predominantly on cells within lymphoid follicles and by intestinal epithelial cells (Fig. 3A). No staining could be detected within the muscle layers. Athymic mice also showed positive staining within lymphoid follicles (observations not shown), but staining on epithelial cells was decreased or absent compared to that in euthymic mice. As expected, no staining was observed on tissues removed from C2d mice (Fig. 3B). We also examined tissues from C57BL/6 mice during the height of infection (day 8 p.i.) and found reduced MHC II expression by epithelial cells but increased expression in the crypt regions and within the muscle layers (Fig. 3C). Stained cells were also seen within the crypt and muscle regions of infected athymic mice; however, fewer positive cells were seen (observations not shown). Staining for the macrophage surface marker F4/80 on serial sections revealed that many of the MHC II-positive cells within the muscle layers were F4/80-positive macrophages (Fig. 3D). As expected, no MHC II-positive cells were seen in infected C2d mice (Fig. 3E). However, following bone marrow reconstitution of C2d mice with C57BL/6 mice as donors, infection was accompanied by the presence of numerous MHC II-positive cells, again predominantly within the crypt regions and muscle layers (Fig. 3F).

View larger version (106K): [in this window] [in a new window]   FIG. 3.   Immunohistochemistry for detection of cells expressing MHC II and F4/80 within the small intestinal tissues of normal uninfected mice (A and B) or mice at 8 days p.i. (C to F). (A) Jejunal cross-section from an uninfected euthymic C57BL/6 mouse. Dark red staining is localized to cells within a lymphoid follicle (arrow, bottom right) as well as more diffuse staining of some epithelial cells (upper arrow). Magnification, ×40. (B) Jejunal cross-section from an uninfected C2d mouse, revealing absence of staining, even in a lymphoid follicle. Magnification, ×40. (C) Numerous stained cells localized to the external muscle layers of jejunal tissue (arrows) from a euthymic C57BL/6 mouse 8 days p.i. Magnification, ×400. (D) Serial section from panel C, with staining for F4/80 (for mature macrophages), showing that F4/80 staining colocalizes with many of the MHC II-positive cells found in the external muscle layers (arrows) of euthymic mice on day 8 p.i. Magnification, ×400. (E) Absence of MHC II staining in the external muscle layers and crypts of a C2d mouse on day 8. Magnification, ×400. (F) Many cells stain for MHC II within the muscle layers, myenteric plexus, and crypts of a C2d mouse (arrows), reconstituted with C57BL/6 bone marrow, on day 8 p.i. Magnification, ×400.

Muscle contraction and worm expulsion in bone marrow- and CD4 T-cell-reconstituted C2d mice. To examine the effects of the reconstitution protocols on muscle function, experiments were conducted on day 8 p.i. As shown in Fig. 4, bone marrow reconstitution of C2d mice had no effect on muscle function, with a tension generation similar to that seen in nonreconstituted C2d mice. In both cases, tension generation by muscle from C2d mice was significantly lower than that generated by muscle from infected C57BL/6 mice. The combined reconstitution of MHC II-positive bone marrow and CD4⁺ T cells to C2d mice resulted in a significant increase in muscle tension, greater than that for both bone marrow-reconstituted and unmanipulated C2d mice and not significantly different from that for infected C57BL/6 mice. C57BL/6 mice which had undergone syngeneic bone marrow grafting also developed a level of muscle contraction during infection similar to that generated by tissues from infected unmanipulated C57BL/6 mice (data not shown).

View larger version (22K): [in this window] [in a new window]   FIG. 4.   Maximum tension generated, in response to 1 µM carbachol, by intestinal muscle from infected euthymic C57BL/6 mice (solid bar) and C2d mice without manipulation (open bar) or reconstituted with bone marrow from C57BL/6 mice (hatched bar) or with bone marrow plus CD4+ T cells (cross-hatched bar) 8 days p.i. Results shown are the mean ± 1 SEM of groups of four animals. The asterisk denotes a significantly lower tension generation by muscle compared to the infected C57BL/6 mice.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Our results clearly point to a dynamic interaction between T-cell functions and intestinal muscle hyperresponsiveness during T. spiralis infection. The increased muscle tension we observed in C57BL/6 mice had a time course similar to that seen in other immunocompetent mouse strains (46). Muscle tension generation increased early, peaking on day 8 p.i., and remained significantly elevated over control levels until at least day 21 p.i. In contrast, the muscle response was significantly reduced in athymic mice, indicating a pivotal role for T cells in the development of increased muscle contraction. The CD8 knockout mice exhibited the same pattern of altered muscle function as normal C57BL/6 mice did, indicating that CD8 expression and the altered immune function of CD8 T cells in these mice (13) were not essential to the muscle hyperresponsiveness. However, a prominent role for CD4 T cells as well as MHC II expression was identified following infection of C2d mice, because tissues from these mice demonstrated attenuated muscle contraction during infection, to a degree similar to that for athymic mice. In contrast, infected CD4-deficient mice gave intermediate muscle function, not significantly different from that for normal C57BL/6 mice. Thus, CD4 expression per se is not essential. However, because these mice are known to retain some MHC II-restricted T-helper-cell function (36), it is the expression of that function which must be most important to the role of CD4 T-cell subsets in the muscle hyperresponsiveness.

The reconstitution studies of the C2d mice showed that even a small population of CD4 T cells, in the presence of MHC II-positive myeloid cells, could confer increased muscle contraction during infection. These results demonstrate clearly that class II-restricted CD4 T cells are the major T-cell subset involved, and are essential for the increased intestinal muscle contractility. They also clearly indicate that MHC II expression on the nonmyeloid tissues is not essential to the muscle response, because smooth muscle cells and others in the C2d recipients cannot express MHC II. The minimal increase in muscle function seen in unreconstituted C2d mice could also indicate a minor or compensatory role for CD8-positive T cells. Their ability to produce cytokines and mimic some of the functions of CD4 T cells under certain conditions could be at play during nematode infections as in other diseases or infections (8). Similarly, the minimal response seen in athymic mice could also reflect the actions of the small population of T cells that are extrathymically derived and frequently localize to the gut mucosa (38).

The worm expulsion that occurred by day 21 p.i. in the immunocompetent C57BL/6 mice was found to be mediated, at least in part, by CD4⁺ T cells. CD4-deficient mice, as well as athymic and C2d mice, all had significant numbers of worms remaining in their small bowels as of day 21 p.i., while expulsion was complete in CD8 knockout mice. Taken together, these results suggest that CD8 T cells play no significant role in worm expulsion but that CD4 T cells may make a significant contribution. It should be noted that while the athymic, C2d, and CD4-deficient mice were still infected on day 21, they had still lost approximately 85% of their initial worm burdens. The substantial worm loss that occurred, even in the absence of a normally functioning immune system, probably reflects factors intrinsic to the parasite. A portion of the adult Trichinella worms may spontaneously exit the bowel, limiting the duration of their own infection as a way to promote host survival until the larval worms are completely encysted. However, a properly functioning immune response involving CD4 T cells and MHC II expression considerably accelerates this worm loss, particularly between days 12 and 21 p.i.

It should also be noted that the athymic, MHC II-deficient, and CD4-deficient mice were still infected by day 28 p.i. but that the intestinal infection had been cleared by day 35 (data not shown). The ability to clear the Trichinella infection, albeit slowly, is at odds with a previously published study by Ruitenberg and Steerenberg (40) in which athymic B10 mice suffered a chronic, low-level T. spiralis intestinal infection lasting at least several months. Adult worms still present in the small bowel at these late stages of the infection may well be damaged or less fecund than earlier in the infection; however, their continued presence is probably still detrimental to the host. The differences between our results and this previous study may simply be due to intrastrain differences in T. spiralis or to the genetic background of the mice, with the C57BL/6 genetic background being capable of clearing the intestine of adult worms at the chronic stage of the infection but B10 mice not being capable of doing so. Although the baseline expulsion may be different between strains, the importance of CD4 T cells and MHC II expression in accelerating such worm expulsion is probably true for all strains.

The inability of CD4 T-cell plus MHC II-positive bone marrow reconstitution to restore normal worm expulsion to C2d mice was surprising. However, we were able to only partially reconstitute the CD4 T-cell population in these mice. Thus, the inability to accelerate expulsion probably reflects a requirement for greater numbers of available mature CD4 T cells to provide sufficient clonal development of a T. spiralis-specific immune response, allowing worm expulsion. This would agree with several previous studies demonstrating that various strains of mice differ in their degree of T-cell proliferation during T. spiralis infection and that this was correlated with the speed of expulsion (49).

Our studies indicate an important role not only for CD4 T cells but also for MHC II expression during T. spiralis infection. Previous studies with panels of congenic mouse strains found that MHC as well as non-MHC genes influence the susceptibility or resistance to T. spiralis infection (49); however, the important cell types expressing MHC II were not identified. Under uninfected, specific-pathogen-free conditions, MHC II expression within the mouse was localized only to cells within lymphoid follicles, as well as some intestinal epithelial cells. There was little or no constitutive expression within the muscle layers. However, during infection, MHC II expression on epithelial cells was reduced and numerous MHC II-positive cells appeared within the muscularis externa and crypt regions of both euthymic and athymic mice. The identity of these MHC II-positive cells was also unclear from initial immunohistochemical studies, since they could represent infiltrating inflammatory cells or resident cells such as smooth muscle and glial cells responding to proinflammatory cytokines within their microenvironment. These resident cells express MHC II ectopically in culture (2, 21, 22) and in human disease (31). However, immunostaining for the macrophage marker F4/80 indicated that many of the MHC II-positive cells found in the muscle layers of euthymic mice during infection were infiltrating macrophages. This agrees with our results from the bone marrow transplantation experiments with recipient C2d mice, which clearly showed that cells expressing MHC II (and therefore most definitely of myeloid origin) localize in the jejunal muscularis externa and crypt regions during T. spiralis infection of those chimeric mice. Thus, at least a portion of the MHC II-positive cells found in the jejunal muscle layers during T. spiralis infection are infiltrating cells, including macrophages and perhaps dendritic cells.

As shown in this and other studies, the host responds to enteric nematode infections with changes in intestinal physiology (3, 7, 33) that are moderated by the immune system. We believe that these changes play an important role in host defense against such infections and that changes in gut muscle function, along with other factors, both contribute to the expulsion of the adult worms and act to limit parasite fecundity. Although this study and others indicate a prominent role for CD4 T lymphocytes in altering the function of gut muscle, it is still not clear whether they act directly on the intestinal muscle cells or act through an intermediary cell type (MHC II-positive myeloid cells) or at a relative distance by paracrine cytokine production. Few studies have examined the direct effects of T cells on enteric muscle, but preliminary studies in culture and in vivo suggest that they can occur, with interleukin-4 as the primary candidate effector cytokine (unpublished observations). Indirect interactions are also possible, for example, through the recruitment and hyperplasia of eosinophils (17) and mast cells (44), which are both dependent upon CD4 T cells of the Th2 class. We have preliminary data indicating that interleukin-5 and eosinophils, as well as c-Kit-dependent cells, which include mast cells and interstitial cells of Cajal, are involved in the increased muscle function following a Trichinella infection (unpublished observations). As for their role in worm expulsion, current literature supports a stronger case for mast cells than eosinophils; however, there is still some controversy in the field (20, 26).

These results not only have implications for host defense against nematodes but also may have broader implications for clinical gastroenterology. CD4⁺ T cells have already been implicated in the pathogenesis of inflammatory bowel diseases and are thought to initiate much of the tissue damage seen in these diseases (34, 42). As well, both MHC II expression and T lymphocytes have been found within the external intestinal muscle layers (9, 12) of patients with Crohn's disease, and recent studies indicate that these infiltrating T cells are both activated and dividing, suggesting that they are responding to antigen and antigen presentation within the muscle layers (12). There are also cases of intestinal pseudo-obstruction where immune cells have been identified within the neuromuscular layers of the gut (29, 41). This study thus raises the possibility that CD4⁺ T cells are a putative cause of the motility disturbances associated with these diseases, as well as with the functional bowel disorders which can arise during enteric infections and persist long after the resolution of mucosal inflammation and immune system activation.