Abstract
Resistance to intestinal amoebiasis is mouse strain dependent. C57BL/6 (B6) mice clear Entamoeba histolytica within hours of challenge, whereas C3H and CBA strains are susceptible to infection and disease. In this study, we show using bone marrow (BM) chimeric mice that mouse strain-dependent resistance is mediated by nonhemopoietic cells; specifically, B6 BM → CBA recipients remained susceptible as measured by amoeba score and culture, whereas CBA BM → B6 recipients remained resistant. Interestingly, hemopoietic IL-10 was required for maintaining the resistance of B6 mice, in that B6 IL-10-deficient mice and IL-10−/− BM → wild-type recipients, but not IL-10+/+ BM → IL-10−/− recipients, exhibited higher amoeba scores than their wild-type controls. Additionally, C57BL/10 IL-10−/−Rag2−/− mice exhibited diminished amoeba scores and culture rates vs IL-10−/− mice, indicating that lymphocytes potentiated the susceptibility of IL-10-deficient mice. We conclude that nonhemopoietic cells mediate the natural resistance to intestinal amoebiasis of B6 mice, yet this resistance depends on hemopoietic IL-10 activity.
Entamoeba histolytica is the agent of amoebic colitis and liver abscess and second only to malaria in global mortality due to protozoa (1). In addition to invasive disease, the parasite asymptomatically colonizes the intestine of up to 11% of endemic populations (2, 3). The mechanisms of human resistance to colonization or invasive disease are poorly understood. One contributing factor is fecal IgA to the parasite, particularly to its galactose-inhibitable adherence lectin, which correlates with partial protection against infection in children and vaccinated mice (3, 4). Additionally, a protective role for IFN-γ, TNF-α, and NO on macrophages has been demonstrated in vitro (5). Yet TNF-α also contributes to amoebic-induced epithelial damage (6) and is chemotactic for amoeba (7), such that the role of proinflammatory cytokines may depend on the cellular context.
To define the host mechanisms of resistance to intestinal amoebiasis, we have used a mouse model of amoebic colitis that develops upon intracecal inoculation of trophozoites (4, 8, 9). A majority of CBA or C3H (either TLR4 mutant or wild-type (WT))3 mice develop patent infections upon challenge. Early after infection, viable amoeba are seen predominantly at sites of epithelial breakdown and a robust inflammatory response ensues that is phenotypically protective, insofar as neutrophil-depleted or dexamethasone-treated C3H or CBA mice show increased infection rates and/or disease scores (9).
In contrast, several other inbred mouse strains such as C57BL/6 (B6) are highly resistant to the infection, exhibiting rapid clearance of trophozoites within hours of challenge. As this timing suggests, their resistance is innate and persists in B6 SCID mice (9). In contrast to susceptible mouse strains, the B6 response to amoeba histologically demonstrates a normal mucosa without inflammation or epithelial breakdown. Indeed, gene expression analysis suggests the absence of a response to amoeba in B6 mice (only 1 of 12,422 genes or expressed sequence tag was dysregulated between B6 sham- and B6 amoeba-challenged mice; data published in (9) and available at 〈https://genes.med.virginia.edu/public data/index.cgi〉). Moreover, B6 resistance does not rely on an innate proinflammatory response in that IL-12p40-deficient, inducible NO synthase-deficient, phagocyte oxidase-deficient, MyD88-deficient, or neutrophil-depleted B6 mice remain resistant (8, 9).
Because an innate proinflammatory response appeared irrelevant for B6 clearance, we embarked upon the hypothesis that B6 resistance owes to the strain’s capacity to maintain intestinal homeostasis in the face of amoeba challenge. We examined the role of IL-10, because IL-10 regulates the host immune response to gut flora (10, 11, 12) and prevents diverse lymphoid and epithelial defects (13, 14). Through this work we find that B6 resistance to intestinal E. histolytica infection occurs at the nonhemopoietic level, yet that the protective capacity of this compartment is maintained via hemopoietic IL-10.
Materials and Methods
Mice
CBA/J, C57BL/6, C57BL/6 IL-10−/−, and C57BL/6 (CD45.1+) mice were purchased from The Jackson Laboratory or obtained from the research colony of E. Leiter (also at The Jackson Laboratory). C57BL/10 mice were purchased from Taconic Farms. C57BL/10 IL-10−/− and IL-10−/−/Rag2−/− mice were generated as described previously (15) and obtained from the National Institutes of Allergy and Infectious Diseases Taconic colony (T. Wynn, Bethesda, MD). Animals were maintained under specific pathogen-free conditions at the University of Virginia and were challenged at 6–12 wk of age. The Institutional Animal Care and Use Committee approved all protocols.
Parasites and intracecal inoculation
Trophozoites for intracecal injections were originally derived from laboratory strain HM1:IMSS (American Type Culture Collection) that were sequentially passaged in vivo through the mouse cecum. Cecal contents were cultured in trypsin-yeast-iron (TYI-S-33) medium supplemented with 25 U/ml penicillin and 25 mg/ml streptomycin until trophozoite growth was axenic as confirmed by the absence of bacterial growth on Trypticase soy agar with 5% sheep blood (BD Biosciences). For all intracecal inoculations, axenic trophozoites were grown to the log phase and counted with a hemacytometer, and 2 × 106 trophozoites in 150 μl were injected thrice intracecally after laparotomy as described (8).
Bone marrow (BM) chimera
Female CBA/J, C57BL/6 IL-10+/+ (CD45.2), C57BL/6 IL-10+/+ (CD45.1), and C57BL/6 IL-10−/− mice (CD45.2) were used for BM chimera mice. Recipient mice were total body irradiated (1100–1300 rad from a 137Cs source) and reconstituted i.v. within 6 h with BM cells prepared (for syngeneic chimeras 1 × 107 cells were transferred, whereas for allogeneic chimeras 2 × 107 cells were transferred) from donor femurs and tibias and then rested for 7–8 wk. For allogeneic BM chimeras, BM cells were pretreated with rabbit antimouse T cell antiserum (Thy 1, CL2005; Cedarlane Laboratories) and rabbit complement (CL3051; Cedarlane Laboratories) to prevent graft-vs-host disease. During weeks 0–3, chimeric mice were given trimethoprim/sulfamethoxazole (5.0/0.86 mM)-supplemented water. Eight groups were generated. For the mouse-strain chimera experiment: CBA/J to CBA/J, CBA/J to C57BL/6, C57BL/6 to CBA, and C57BL/6 to C57BL/6; for the IL-10−/− chimera experiment: IL-10+/+(CD45.1) to IL-10+/+(CD45.2), IL-10−/−(CD45.2) to IL-10+/+(CD45.1), IL-10+/+(CD45.1) to IL-10−/−(CD45.2), and IL-10−/−(CD45.2) to IL-10−/−(CD45.2). More than 90% engraftment of hemopoietic cells from donor origin was confirmed on peripheral blood leukocytes at 6 wk postreconstitution, and on splenocytes and mesenteric lymph node (MLN) cells upon sacrifice by FACS using conjugated Abs to H-2Kb (AF6-88.5) and H-2Kk (36-7-5) for the allogeneic chimeras and CD45.1 (A20) and CD45.2 (104) for the IL-10−/− chimera experiment (Abs obtained from BD Biosciences).
Pathology and scoring of amoebic colitis
Mice were sacrificed and each cecum longitudinally bisected. One-half of the cecum was placed in Hollande’s fixative, cut into three to five equal cross-sections, paraffin embedded, and 4-μm sections were stained with H&E. Histopathology was scored blindly for each mouse as described previously (8). Briefly, numbers of histologically visible amoeba were scored 0–5 (0, none; 1, present but difficult to locate; 2, occasional, up to 10% of the lumen occupied by ameba; 3, moderate, up to 25% of lumen occupied; 4, heavy, up to 50% of lumen occupied; 5, virtually complete occupation of the lumen by ameba). Degree of inflammation was scored 0–5 (0, normal; 1, mucosal hyperplasia; 2, spotty infiltration of inflammatory cells not involving the entire thickness of the mucosa; 3, marked increase in inflammatory cells involving full thickness of mucosa; 4, marked increase in inflammatory cells of mucosa and submucosa, with tissue architecture intact; 5, complete destruction of cecal architecture by inflammation). To determine infection rate by culture, the contents of the other half of the cecum were rinsed in 1 ml of PBS and cultured in TYI-S-33 medium.
RNA expression
Affymetrix gene chip analysis was performed according to the manufacturer’s instructions using murine MgU74Av2 arrays. Full public access to the raw data is available at 〈https://genes.med.virginia.edu/public data/index.cgi〉 under “Eric Houpt IL10KO Amebic Colitis.” Information about a microarray experiment (〈www.mged.org/miame〉) includes the following. Ceca were obtained from female 6-wk-old C57BL/6 WT and IL-10−/− mice 18 h after intracecal amoeba or sham challenge (n = 5 C57BL/6 WT amoeba, n = 3 C57BL/6 WT sham, n = 3 C57BL/6 IL-10−/− amoeba, n = 3 C57BL/6 IL-10−/− sham) for a total of 14 samples. Cecal tissue was rinsed in sterile PBS and then placed in RNAlater (Ambion), homogenized, and total RNA extracted using the RNAeasy kit (Qiagen). Ribosomal peaks were intact and showed no evidence of degradation. The BioB control cRNA was spiked in at the detection threshold of 1.5 pM and received a “present” detection call. Housekeeping genes had 3′ to 5′ detection ratios of <4. Complementary DNA synthesis, in vitro transcription to cRNA, and hybridization were performed as described (per 〈http://www.healthsystem.virginia.edu/internet/biomolec/genechipprotocols.cfm〉).
Raw data from the arrays were normalized at probe level using GC content-based algorithm (16). The detection call (Present, Marginal, Absent) for each probe set was obtained using the GeneChip Operating Software (〈http://www.affymetrix.com/products/software/specific/gcos.affx〉). Only genes with at least one present call across all the compared arrays were kept for downstream statistical analyses. Differentially expressed genes were identified using Linear Models for Microarray data algorithm (17), and false discovery rate (FDR) was used for multiple testing correction (18). Genes with fold change ≥2 and FDR <0.05 were identified as differentially expressed genes. Significantly overrepresented gene ontology terms among the list of differentially expressed genes were identified using GO::TermFinder program (19) with a cutoff value of FDR <0.05.
Depletion of CD4+CD25+ T cells
A total of 500 μg of anti-mouse CD25 (rat IgG1, TIB 222 also referred to as PC61; obtained from K. Tung, Charlottesville, VA) or control rat IgG (Lampire Biological Laboratories) was administered i.p. to C57BL/6 mice every 5 days for 5 doses on days 0–20. To examine depletion, splenocytes and MLN cells were tested by FACS using FITC-conjugated anti-CD25 7D4 mAb (which binds to CD25 noncompetitively with PC61; Ref. 20), allophycocyanin-conjugated anti-CD4 mAb, as well as the PE-anti-mouse/rat Foxp3 staining set (eBioscience). Mice were then challenged with E. histolytica trophozoites on day 24 and sacrificed 3 days postchallenge for evaluation of infection.
Statistics
Group averages were compared using Student t test, and proportions (e.g., culture-positive rates) were compared using Fisher’s exact test. Data are expressed as mean ± SE. All p values were two tailed.
Results
Mouse strain-dependent resistance to infection is mediated by nonhemopoietic cells
To elucidate which cellular compartment determines resistance to intestinal amoebiasis in mice, we constructed B6→CBA BM chimeras. Chimerism was confirmed in PBMCs, MLNs, and splenocytes (Fig. 1,A). Seven weeks posttransfer, mice were intracecally challenged with E. histolytica trophozoites. At 4 days postintracecal challenge, a time point by which the resistance to infection has occurred in B6 mice (9), CBA→B6 chimeras recapitulated the phenotype of resistant B6→B6 mice with a low amoeba score, inflammation score (Fig. 1, B and C), and culture-positive rate (1 of 15 and 0 of 9 ceca were culture positive, respectively; p = NS). Likewise, B6→CBA chimeras exhibited the susceptibility of CBA→CBA mice according to amoeba score, inflammation score, and culture-positive rate (13 of 14 and 6 of 7 ceca were culture positive, respectively; p = NS).
C57BL/6 mouse resistance to E. histolytica is mediated by nonhemopoietic cells. Recipient mice were total-body irradiated (1300 rad from a 137Cs source) and reconstituted i.v. within 6 h with cells prepared from donor femur and tibia. Engraftment of >95% donor hemopoietic cells were confirmed by FACS for H-2Kb and H-2Kk at 6 wk posttransfer on PBMCs and upon sacrifice on splenocytes (A; n = 3/group, one representative mouse per group shown) and MLN. At 7 wk posttransfer, mice were challenged intracecally with 2 × 106 E. histolytica trophozoites and sacrificed after 4 days for evaluation of infection according to histologic amoeba score (B) and inflammation (C). ∗, p < 0.05; n = 9, 14, 15, and 7 for the four groups indicated left to right on the x-axis. Data shown as mean + SE.
C57BL/6 mouse resistance to E. histolytica is mediated by nonhemopoietic cells. Recipient mice were total-body irradiated (1300 rad from a 137Cs source) and reconstituted i.v. within 6 h with cells prepared from donor femur and tibia. Engraftment of >95% donor hemopoietic cells were confirmed by FACS for H-2Kb and H-2Kk at 6 wk posttransfer on PBMCs and upon sacrifice on splenocytes (A; n = 3/group, one representative mouse per group shown) and MLN. At 7 wk posttransfer, mice were challenged intracecally with 2 × 106 E. histolytica trophozoites and sacrificed after 4 days for evaluation of infection according to histologic amoeba score (B) and inflammation (C). ∗, p < 0.05; n = 9, 14, 15, and 7 for the four groups indicated left to right on the x-axis. Data shown as mean + SE.
IL-10 is required for resistance to intestinal amoebiasis in C57BL/6 mice
These results suggested that the strain-dependent phenotype of resistance or susceptibility is governed by nonhemopoietic cells. We next examined the outcome of infection in IL-10−/− mice on the resistant C57BL/6 background, because these mice acquire abundant defects in their epithelium (13, 14, 21, 22, 23). When sacrificed at 9–10 days postchallenge, B6 IL-10-deficient mice exhibited a significantly higher amoeba score, inflammation score, and culture-positive rate than their WT counterparts (Fig. 2, A and B; 10 of 22 IL-10−/− ceca were culture positive vs 1 of 28 WT; p < 0.05). Intestinal pathology of amoeba-challenged IL-10−/− mice revealed extensive infiltrates in the lamina propria and epithelial ulceration vs relatively normal histology in amoeba-challenged IL-10+/+ mice. Naive IL-10−/− mice housed in our specific pathogen-free facility exhibited minimal histologic abnormalities in the cecum before challenge, such that the susceptibility of IL-10−/− mice did not apparently require histologically overt pre-existing disease, and the amoeba-induced inflammation seen in IL-10−/− developed acutely (Fig. 2 C). Notably, the susceptibility of B6 IL-10−/− mice appeared to be distinct from that of CBA mice in terms of a higher inflammation:amoeba score ratio (2.70 ± 0.59 vs 1.07 ± 0.18 in B6 IL-10−/− and CBA/J, respectively; n = 22 and 11; p < 0.05). Furthermore, B6 IL-10−/− mice appeared to ultimately clear the parasite (0 of 8 C57BL/6 IL-10−/− mice were infected by histology or culture after 14 days) in contrast to the chronic nonhealing infection of C3H or CBA mice (9). These results suggested that IL-10 functions to maintain C57BL/6 resistance to the early establishment of infection.
IL-10 promotes resistance to intestinal amoebiasis. WT (□) or IL-10−/− (▪) C57BL/6 or CBA mice were intracecally challenged with 2 × 106 E. histolytica trophozoites and sacrificed after 9–10 days for evaluation of infection according to histologic amoeba score (A) and inflammation (B). ∗, p < 0.05; n = 28, 22, and 11 for the three groups indicated left to right on the x-axis. Data represent three independent experiments and are shown as mean + SE. C, Representative histology from the ceca of C57BL/6 WT or IL-10−/− mice before and after amoeba challenge demonstrated normal histology in all mice except for abundant inflammation and ulceration in amoeba-challenged IL-10−/− mice. Amoeba indicated with arrowheads (H & E, ×200; inset, ×1000).
IL-10 promotes resistance to intestinal amoebiasis. WT (□) or IL-10−/− (▪) C57BL/6 or CBA mice were intracecally challenged with 2 × 106 E. histolytica trophozoites and sacrificed after 9–10 days for evaluation of infection according to histologic amoeba score (A) and inflammation (B). ∗, p < 0.05; n = 28, 22, and 11 for the three groups indicated left to right on the x-axis. Data represent three independent experiments and are shown as mean + SE. C, Representative histology from the ceca of C57BL/6 WT or IL-10−/− mice before and after amoeba challenge demonstrated normal histology in all mice except for abundant inflammation and ulceration in amoeba-challenged IL-10−/− mice. Amoeba indicated with arrowheads (H & E, ×200; inset, ×1000).
Susceptibility of IL-10−/− mice is associated with a proinflammatory response
Because of the abundance of both lymphoid and epithelial intestinal defects that have been described in IL-10-deficient mice, to begin to understand the mechanism of IL-10−/− susceptibility we characterized the intestinal environment after E. histolytica challenge by microarray. Total RNA was prepared from three IL-10−/− and five IL-10+/+ ceca 18 h postchallenge and hybridized to Affymetrix MgU74Av2 arrays. Of 7264 genes, 399 were recorded as differentially expressed as detailed in Materials and Methods. Patterns of gene ontology were queried among the differentially expressed genes and indicated that “immune response” was the most statistically overrepresented gene ontology function in IL-10−/− mice (FDR = 0.0025), whereas “regulation of cellular process” was most overrepresented in IL-10+/+ mice (FDR = 0.02). Most (209 of 399) of the differentially expressed genes were not amoeba-specific, and were common to the identical sham-challenge cecal tissue comparison (three IL-10−/− and three IL-10+/+ ceca 18 h after sham challenge; Table I). Overall, the profile of the amoeba or sham-infected IL-10−/− ceca suggested a mixed inflammatory environment with up-regulation of several Ag presentation-related genes, certain innate immune genes and chemokines, as well as the anti-inflammatory gene Pap.
Upregulated immune response genes in the ceca of C57BL/6 IL-10−/− mice 18 h after E. histolyticachallenge
| Gene | Description | Affymetrix Probe Set | IL-10−/− (signal intensity) | IL-10+/+ (signal intensity) | Fold Change | p Value |
|---|---|---|---|---|---|---|
| H2-Ab1 | Histocompatibility 2, class II | 100998_at | 992 | 126 | 7.9 | 0.010 |
| Pap | Pancreatitis-associated protein | 161890_f_at | 34958 | 4784 | 7.3 | 0.014 |
| Cdkn1a | Cyclin-dependent kinase inhibitor 1A (P21) | 98067_at | 315 | 65 | 4.9 | 0.032 |
| Reg3g | Regenerating islet-derived 3 γ | 162187_f_at | 1799 | 382 | 4.7 | 0.023 |
| Psmb9 | Proteosome subunit β type 9b | 93085_at | 569 | 124 | 4.6 | 0.000 |
| Cd74 | Ia-associated invariant chain | 101054_at | 3722 | 874 | 4.3 | 0.026 |
| PGLYRP1 | Peptidoglycan recognition protein 1 | 162475_f_at | 4608 | 1659 | 2.8 | 0.003 |
| OTUB1 | OTU domain, ubiquitin aldehyde binding 1c | 94336_at | 65 | 24 | 2.7 | 0.006 |
| C4a | Complement component 4 | 103033_at | 156 | 65 | 2.4 | 0.027 |
| Cxcl12 | Stromal cell-derived factor 1 | 160511_at | 136 | 60 | 2.3 | 0.049 |
| lrf8 | IFN regulatory factor 8 | 98002_at | 880 | 395 | 2.2 | 0.011 |
| B2m | β2 microglobulin | 93088_at | 14690 | 7496 | 2.0 | 0.002 |
| Gene | Description | Affymetrix Probe Set | IL-10−/− (signal intensity) | IL-10+/+ (signal intensity) | Fold Change | p Value |
|---|---|---|---|---|---|---|
| H2-Ab1 | Histocompatibility 2, class II | 100998_at | 992 | 126 | 7.9 | 0.010 |
| Pap | Pancreatitis-associated protein | 161890_f_at | 34958 | 4784 | 7.3 | 0.014 |
| Cdkn1a | Cyclin-dependent kinase inhibitor 1A (P21) | 98067_at | 315 | 65 | 4.9 | 0.032 |
| Reg3g | Regenerating islet-derived 3 γ | 162187_f_at | 1799 | 382 | 4.7 | 0.023 |
| Psmb9 | Proteosome subunit β type 9b | 93085_at | 569 | 124 | 4.6 | 0.000 |
| Cd74 | Ia-associated invariant chain | 101054_at | 3722 | 874 | 4.3 | 0.026 |
| PGLYRP1 | Peptidoglycan recognition protein 1 | 162475_f_at | 4608 | 1659 | 2.8 | 0.003 |
| OTUB1 | OTU domain, ubiquitin aldehyde binding 1c | 94336_at | 65 | 24 | 2.7 | 0.006 |
| C4a | Complement component 4 | 103033_at | 156 | 65 | 2.4 | 0.027 |
| Cxcl12 | Stromal cell-derived factor 1 | 160511_at | 136 | 60 | 2.3 | 0.049 |
| lrf8 | IFN regulatory factor 8 | 98002_at | 880 | 395 | 2.2 | 0.011 |
| B2m | β2 microglobulin | 93088_at | 14690 | 7496 | 2.0 | 0.002 |
Four other histocompatibility 2 class II transcripts were also up-regulated in IL-10−/− mice (98035_g_at, 94285_at, 93092_at, 97541_f_at).
Proteosome subunit β type 8 (102791_at) was also up-regulated in IL-10−/− mice.
Bold genes were those up-regulated only in amoeba-challenged IL-10−/− mice, all others were up-regulated in both amoeba- and sham-challenged IL-10−/− mice (vs WT controls).
Hemopoietic cells are a required source of protective IL-10
Although we were seeking to find in IL-10−/− ceca the loss of canonical nonhemopoietic cell transcripts, the microarray analysis did not clearly incriminate a single cellular subset in the susceptibility defect of IL-10−/− mice. We therefore broadly examined the cellular anatomy of IL-10 production via BM chimeras. IL-10−/− → IL-10+/+ BM chimeras were generated, and chimerism was confirmed using CD45 markers in PBMCs, MLN cells, and splenocytes (Fig. 3,A). Early susceptibility was examined at 2–3 days postchallenge. As expected, a lower amoeba score, inflammation score, and culture-positive rate (4 of 15 vs 6 of 7; p < 0.05) was observed in IL-10+/+ → IL-10+/+ vs IL-10−/− → IL-10−/− chimeras (Fig. 3, B and C). Notably, the culture-positive rate at this early time point is relatively high in the resistant B6 IL-10+/+ → IL-10+/+ controls, therefore we feel the quantitative amoeba and inflammation scores are more meaningful for early events. The comparison of IL-10+/+ → IL-10+/+ vs IL-10−/− → IL-10+/+ mice demonstrated that hemopoietic IL-10 was required for resistance (IL-10−/− → IL-10+/+ mice exhibited statistically higher amoeba score, inflammation score, and culture-positive rate (13 of 20 vs 4 of 15; p < 0.05)). Additionally, hemopoietic IL-10 was largely sufficient for resistance, because IL-10+/+ → IL-10−/− mice exhibited a statistically lower amoeba score and inflammation score vs IL-10−/− → IL-10−/− mice (culture-positive rate 8 of 15 vs 6 of 7, respectively; p = 0.19).
Resistance to intestinal amoebiasis in C57BL/6 mice is maintained by hemopoietic IL-10 production. Recipient B6 IL-10+/+ or IL-10−/− mice were total-body irradiated (1100 rad from a 137Cs source) and reconstituted i.v. within 6 h with 1 × 107 cells prepared from donor femur and tibia. Engraftment of >90% donor hemopoietic cells were confirmed by FACS for CD45 Ags at 6 wk posttransfer on PBMCs and upon sacrifice on splenocytes (A; n = 4/group, one representative mouse per group shown) and MLN. At 7 wk posttransfer, mice were challenged intracecally with 2 × 106 E. histolytica trophozoites and sacrificed after 2–3 days for evaluation of infection according to histologic amoeba score (B) and inflammation (C). ∗, p < 0.05; n = 15, 20, 15, and 7 for the four groups indicated left to right on the x-axis. Data represent two independent experiments and are shown as mean + SE. D, Representative histology from the ceca of BM chimeras demonstrated benign histology in IL-10WT → IL-10WT mice, mucosal hyperplasia with amoeba in IL-10KO → IL-10WT mice, mild mucosal hyperplasia in IL-10WT → IL-10KO mice, and severe inflammation and ulceration in IL-10KO → IL-10KO mice. H & E, ×200.
Resistance to intestinal amoebiasis in C57BL/6 mice is maintained by hemopoietic IL-10 production. Recipient B6 IL-10+/+ or IL-10−/− mice were total-body irradiated (1100 rad from a 137Cs source) and reconstituted i.v. within 6 h with 1 × 107 cells prepared from donor femur and tibia. Engraftment of >90% donor hemopoietic cells were confirmed by FACS for CD45 Ags at 6 wk posttransfer on PBMCs and upon sacrifice on splenocytes (A; n = 4/group, one representative mouse per group shown) and MLN. At 7 wk posttransfer, mice were challenged intracecally with 2 × 106 E. histolytica trophozoites and sacrificed after 2–3 days for evaluation of infection according to histologic amoeba score (B) and inflammation (C). ∗, p < 0.05; n = 15, 20, 15, and 7 for the four groups indicated left to right on the x-axis. Data represent two independent experiments and are shown as mean + SE. D, Representative histology from the ceca of BM chimeras demonstrated benign histology in IL-10WT → IL-10WT mice, mucosal hyperplasia with amoeba in IL-10KO → IL-10WT mice, mild mucosal hyperplasia in IL-10WT → IL-10KO mice, and severe inflammation and ulceration in IL-10KO → IL-10KO mice. H & E, ×200.
Because hemopoietic IL-10 played a role in maintaining the resistant state of C57BL/6 mice, we sought to examine the role of CD25+ T regulatory cells in this strain. We depleted the CD25+ population via administration of anti-CD25 mAb (PC61) on every 5 days × 5 doses (CD4+CD25+ proportion of CD3+ splenocytes reduced to 0.27 from 5.06%; n = 3/group), challenged mice on day +24, and sacrificed mice after 3 days. The effect of this intervention on amoeba score was unchanged vs control Ab-treated mice (n = 14 mice/group; data not shown). This absence of effect was potentially confounded by residual CD4+Foxp3+ cells (e.g., only reduced in spleen from 20.13 ± 2.46 to 12.12 ± 1.51% with anti-CD25 mAb treatment; n = 3/group) or IL-10 production by other cellular sources.
As for nonhemopoietic sources of IL-10, the reciprocal BM chimera comparisons indicated that nonhemopoietic IL-10 was not required for resistance as measured by amoeba score (IL-10+/+→ IL-10+/+ vs IL-10+/+ → IL-10−/−; p = NS) but did contribute to suppress inflammation (inflammation score lower in IL-10+/+ → IL-10+/+ vs IL-10+/+ → IL-10−/−; p < 0.05; Fig. 3 C). Moreover, a role of nonhemopoietic IL-10 became apparent in the absence of hemopoietic IL-10 (i.e., IL-10−/− → IL-10+/+ mice showed a statistically lower amoeba score and inflammation score vs IL-10−/− → IL-10−/− mice).
Lymphocytes partly contribute to susceptible phenotype of IL-10−/− mice
T lymphocytes are important targets of IL-10 activity in many models. Therefore, we sought to examine the requirement for lymphocytes in the susceptibility of IL-10−/− mice. We examined amoebic infection in Rag2−/−× IL-10−/− mice, which were available on the C57BL/10 background. C57BL/10 Rag2−/−× IL-10−/− exhibited diminished susceptibility vs C57BL/10 IL-10−/− mice according to amoeba score (Fig. 4,A) or culture-positive rate (double knockout (KO) 5 of 25 vs 7 of 12 KO; p < 0.05). The effect of lymphocyte deficiency was modest, and inflammation score was not statistically altered (Fig. 4 B), perhaps because the underlying strain susceptibility of C57BL/10 IL-10−/− was less than that of C57BL/6 IL-10−/− mice (data not shown). However, these results suggested that lymphocytes partially contributed to the susceptible state of IL-10-deficient mice.
Lymphocytes partially contribute to the susceptibility of IL-10−/− mice to intestinal amoebiasis. C57BL/10 WT (WT, □), C57BL/10 IL-10−/− (KO, ▪), and C57BL/10 IL-10−/−× Rag−/− (DKO, ▦) were intracecally challenged with 2 × 106 E. histolytica trophozoites and sacrificed after 3 days for evaluation of infection according to histologic amoeba score (A) and inflammation (B). ∗, p < 0.05; n = 11, 12, and 25 left to right on x-axis. Data represent three independent experiments and are shown as mean + SE.
Lymphocytes partially contribute to the susceptibility of IL-10−/− mice to intestinal amoebiasis. C57BL/10 WT (WT, □), C57BL/10 IL-10−/− (KO, ▪), and C57BL/10 IL-10−/−× Rag−/− (DKO, ▦) were intracecally challenged with 2 × 106 E. histolytica trophozoites and sacrificed after 3 days for evaluation of infection according to histologic amoeba score (A) and inflammation (B). ∗, p < 0.05; n = 11, 12, and 25 left to right on x-axis. Data represent three independent experiments and are shown as mean + SE.
Discussion
This work shows that the inherent resistance of C57BL/6 mice to amoebic infection is governed by nonhemopoietic cells yet can be altered by the hemopoietic compartment, specifically needing hemopoietic IL-10 for maintenance. These findings may be a clue to the variable resistance to colonization or invasive disease observed in humans. In the steady state, some individuals may exhibit a nonhemopoietic predisposition to susceptibility. Others are endowed with a resistant phenotype, which can be lost with hemopoietically derived alterations.
A protective role for IL-10 is at first counterintuitive given several systemic infectious disease models such as Listeria monocytogenes, Toxoplasma gondii, and Trypanosoma cruzi infection (24, 25, 26), where IL-10 dampens protective immunity but prevents immunopathology. Interestingly, many gastrointestinal infections behave differently, and IL-10-deficient mice exhibit worse infection and disease burdens with Trichuris muris and Helicobacter hepaticus (27, 28). Disease in the latter model owes to the failure to generate a protective T regulatory cell population (29) and is manifest after 20 days during the phase of adaptive immunity. In this amoebiasis model, by contrast, the effect of IL-10 is on the immediate phase, because B6 IL-10-competent mice resist amoebic infection within hours (9).
This timing would predict that the susceptibility of IL-10-deficient mice reflects a predisposed state of the IL-10−/− intestine. The microarray results support this hypothesis, because most dysregulated genes in IL-10−/− mice after amoeba challenge were present with the identical sham comparison. As for potential mechanisms of susceptibility in IL-10−/− mice, lymphocyte-dependent pathways are suggested by the diminished susceptibility in IL-10−/−× Rag−/− mice (albeit this comparison was performed on the C57BL/10 background). IL-10 inhibits resident T cell responses to host intestinal flora (22, 30, 31), and one could propose that resultant proinflammatory T cell responses in IL-10−/− mice thereby predispose the otherwise-resistant B6 nonhemopoietic cells to early amoebic infection upon challenge. One possibility may involve mucus production. The epithelium of IL-10−/− mice exhibits defects in synthesis of Muc2 (23), a major component of human colonic mucus that limits E. histolytica adherence to host cells (32). Finally, the epithelium of IL-10−/− mice can be predisposed toward increased apoptosis (14), and these cellular corpses could serve as fuel for the amoeba as happens in vitro (33, 34). Another intriguing possibility is that trophozoites, being the tissue phase of the parasite, benefit from an IL-10-deficient proinflammatory state; for instance, TNF-α (though TNF message was not increased in Table I) is known to promote E. histolytica chemotaxis, adherence, and trophozoite-induced epithelial damage (6, 35, 36).
The susceptibility-enhancing effect of IL-10 deficiency and its correlation with an inflammatory environment as described by the microarrays is seemingly at odds with our previous work showing inflammation was host-protective in CBA or C3H mice (namely, neutrophil depletion or dexamethasone therapy diminished the resistance rate of CBA, C3H, and C3H SCID mice; Ref. 9). We would argue that first it is important to note that the phenotype of early susceptibility in B6 IL-10−/− mice vs the chronic infection of WT CBA or C3H are distinct, such that their mechanisms of susceptibility need not overlap. At present, we are investigating the hypothesis that successful infection requires two phases: step one is survival amid nonhemopoietic cells (e.g., epithelium), and step two is survival amid the inflammatory response. Accordingly, B6 mice are naturally resistant at step 1; the inflammatory cascade is not recruited nor needs to be, hence inducible NO synthase, IL-12, phagocyte oxidase, and MyD88-deficient and neutrophil-depleted B6 mice remain resistant (9). CBA mice possess an underlying nonhemopoietic predisposition to amoebic establishment and therefore are susceptible to initial infection at step 1; however, the inflammatory response that ensues provides partial protection (diminishing susceptibility to ∼60% from near-complete levels; Ref. 9). In B6 IL-10-deficient mice, hemopoietically derived alterations lead to defects in nonhemopoietic cells that render these mice partly susceptible to initial infection, although ultimately these mice clear infection as well. If this speculative model or parts therein are upheld, inflammation in this system is not qualitatively good or bad for the host but can be protective or deleterious depending on the timing and cellular context.
Acknowledgments
We thank William A. Petri, Jr., Kenneth S. Tung, Shinji Okano, and Hisakata Yamada for advice. We acknowledge the Research Histology Core of the Center for Research in Reproduction and the Morphology Imaging Core of the Digestive Disease Research Center for histology support, all at the University of Virginia. We thank the University of Virginia Biomolecular Research Facility (Alyson Prorock, Yongde Bao) and the Virginia Bioinformatics Institute (Zhangjun Fei, Adam Jerauld, Oswald Crosta, Clive Evans) for performing the gene chip analyses.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by National Institutes of Health Grant AI052444-01, the Crohn’s and Colitis Foundation of America, a Pilot/Feasibility Award from the University of Virginia Silvio O. Conte Digestive Disease Research Center (all to E.H.), and the Uehara Memorial Foundation (to S.H.).
Abbreviations used in this paper: WT, wild type; BM, bone marrow; MLN, mesenteric lymph node; FDR, false discovery rate; KO, knockout.
