Severe colitis in mice with aberrant thymic selection

Immunity, Vol. 3, 27-39, July, 1995, Copyright 0 1995 by Cell Press

Severe Colitis in Mice with Aberrant Thymic Selection

Georg A. Hollllnder, l II Stephen J. Slmpson,tll Emiko Mizoguchi,S Allkl Nichogiannopoulou,§ Jlan She,? Jo&Carlo8 Gutlerrez-Ramos,§ Atul K. Bhan,* Steven J. Burakoff,’ Baoping Wang,t and Cox Terhorstt *Division of Pediatric Oncology Dana Farber Cancer institute tDivision of Immunology Beth Israel Hospital *Department of Pathology Massachusetts General Hospital §Center for Blood Research Harvard Medical School Boston, Massachusetts 02115

Summary

TgE26 mice display an arrest very early in T cell development that has a profound effect on the architecture of thymic stromal ceils. We have recently demonstrated that transplantetion of wild-type bone marrow cells restores the thymlc microenvironment of fetal but not adult Tgr26 mice. Here, we report that T cell-recorr stituted adult Tgd6 mice develop a spontaneous wasting syndrome characterized by extensive inflammation of the colon, resembling human ulcerative colitis. Colitis in these animals was marked by substantial infiltration of the colon by activated thymusderived CD4+ T cells. Importantly, bone marrow-transplanted TgE26 mice previously engrafted with a fetal Tg&26 thy- mu8 did not develop colitis. These results suggest that T cells selected in an aberrant thymic mlcroenviron- ment contain a population of cells able to induce severe colitis that can be prevented by T cells that have undergone normal thymic development.

Introduction

Inflammatory bowel disease (IBD) in humans encompasses at least two separate forms of intestinal inflammation: ui- cerative colitis and Crohn’s disease. The etiology of both of these disorders is unknown, although several observations provide evidence for an immunopathological, possibly autoimmune, origin (Sartor, 1990, 1994; Podolsky, 1991a,1991b;Mihadaetai.,1969;Wuetal.,1991;Ligum- ski et al., 1990; Brynskov et al., 1992). Distinct changes in the number of macrophages, B cells, and T cells, as well as elevated levels of cytokines have been noted in colonic tissue from patients with IBD (Podolsky, 1991 b). Consistent with these findings of an altered immune response in human IBD are observations that colitis in experimental animal models may result from acytokinedysregu- lation and the absence of regulatory T cells. Mice deficient

IlBoth authors contributed equally to this work.

for interfeukin-2 (IL-2) or IL-1 0 are prone to develop severe IBD (KQhn et al. 1993; Sadlack et al., 1993) and mice deficient in ap T cell receptor (T(X)-positive cells develop spontaneously an intestinal inflammation akin to the human ulcerative colitis (Mombaerts et al., 1993). The complex mechanisms operative in the induction and mainte- nance of colitis in each of these experimental systems are not precisely defined. However, it was recently demonstrated that the transfer of a subpopulation of peripheral T cells, CD4+ CD45RBh’, to T cell-deficient mice is sufficient to confer a lethal wasting syndrome with IBD. The elevated secretion of interferon? (IFNT) and tumor necro- sisfactor(TNF) by intestinaiTceilssuggest that thedisrup- tion in the regulation of Thl-celis may be one mechanism underlying the pathogenesis of colitis in these mice (Pow- rie et al., 1993, 1994b; Morrissey et al., 1993). This dual involvement of T cells as regulatory cells that prevent the occurrence of IBD and as effector cells able to mediate inflammation is now established (Mombaerts et al., 1993; Powrie et al., 1993, 1994b; Morrissey et al., 1993).

Within the intestinal mucosa of normal adult mice, CD3+ T lymphocytes are compartmentalized in three anatomi- cally distinct areas: the Peyer’s patches, the lamina propria, and the epithelial region as demarcated by a basement membrane. Developmentally, intestinal intraepithelial lyrn- phocytes (ilELs) are considered to comprise two distinct populations, i.e., T cells that have matured in the thymus and those that have developed independently of the thymic microenvironment (Poussier and Julius, 1994). Mech- anisms selecting the repertoire of most thymusderived T cells have been well characterized using different trans- genie and gene-targeted mouse models (von Boehmer, 1990; Pfeffer and Mak, 1994; Robey and Fowlkes, 1994). By contrast, events operative in the selection of gut- derived T cells have been less well defined. For example, self-reactive T cells are detected among IELs (Rocha et al., 1991, 1992; Barrett et al., 1992; Poussier et al., 1992, 1993), although these af3 TCR’cells are functionally unre- sponsive when stimulated via their TCR (Poussier et al., 1992, 1993; Barrett et al., 1993; Poussier and Julius, 1994).

Central and peripheral immunological tolerance are achieved by several complementary but mutually exclusive processes, namely clonal deletion, cional anergy, and active suppression (Kroemer and Martinez, 1992). During intrathymic T cell development, a close and bidirectional interaction between stromai cells and thymocytes is critical for the first two of these mechanisms to be effective. Erroneous selection of thymocytes may be responsible for the selection and expansion of self-reactive T cells or may contribute to the absence of regulatory T cells important in the control of self-reactive lymphocytes. It is, therefore, conceivable that any events that disrupt the induction of tolerance may constitute the initiating step in the pathogenesis of autoimmune disorders.

Mice transgenic for the human CD3s gene, designated Tgs26, display a very early arrest in T cell development,

Immunity 26

Figure 1. Gross Pathology of the Colon of Transplanted Tga26 Mice Comparison of large bowel size of a transplanted Tgs26 mouse with intestinal disease (A) and a normal mouse (6).

which prevents the induction of a normal thymic microenvironment (Wang et al., 1994, 1995). Transplantation of adult Tgs26 mice with normal T cell-depleted bone marrow cells results neither in the reconstitution of the thymic microenvironment nor in normal T cell ontogeny (Hol- lander et al., 1995). Here, we report that T cells selected inathymusgreatlydeficientof anormalarchitecturecause a lethal wasting syndrome with severe colitis. These animals were studied in detail to assess the immunopathol- ogy underlying this disease. The results presented here directly underscore the critical role of the thymic microenvironment in the education of thymus-derived regulatory T cells crucial for the homeostasis of the mucosal immune system. More importantly, the presence of mature T cells selected in an aberrant thymic microenvironment appear to be critical in the pathogenesis of one form of IBD.

Results

Intestinal Disease In Bone Marrow-Reconstituted Tgd6 Mice A general wasting syndrome was noted 5-6 weeks after transplantation in Tgs26 mice engrafted with T cell-depleted normal (C57BU6 x CBA/J)Fl bone marrow cells. Severe weight loss, hunched posture, and chronic diarrhea with occasional bleeding were uniform signs of this disease, whereas anorectal prolaps occurred rarely. Within 2-3 weeks after onset of the first clinical symptoms, all mice succumbed to a wasting syndrome, thus demon- strating a 100% penetrance of the disease. At necropsy, these animals revealed a general lymphadenopathy. When compared with normal mice, lymph nodes were enlarged 2- to 3-fold but they showed neither loss of their anatomical confinement nor signs of suppuration. More- over, transplanted Tgs26 mice presented with extensive pathological changes typical of IBD (Figure 1). The colon was most prominently affected, with marked dilatation and thickening of the gut wall resulting in a rigid and shortened large bowel. In contrast, the small bowel was never macro-

scopically enlarged. Consistently associated with the pathological changes of the colon were greatly enlarged mesenteric lymph nodes, while both size and number of Peyer’s patches appeared to be normal. The contents of colon and rectum were loose in transplanted and Tgs26 mice with disease in contrast with the beaded appearance of feces of wild-type mice. All macroscopic changes directly correlated with the degree of clinical disease and could neither be found in bone marrow-transplanted wild- type animals, nor in RAG-T’” mice transplanted with T cell-depleted bone marrow cells (see below; data not shown).

Hlstopathology of Colitis Microscopic analysisof transverse sections from the colon of unmanipulated and transplanted Tgs26 mice revealed an IBD involving mucosa and submucosa with profound thickening of the colon wall in the bone marrow-reconstituted animals (Figures2Aand 28). Typical features included hyperplasia, elongation, distortion, and branching of crypts as well as a loss of the normal mucin-producing goblet cells. Crypt abscesses could also frequently be found, es- pecially in animals with severe inflammation of the large bowel (Figures 28 and 2D, arrow). Mucosal ulcerations were not identified and there was no indication of the formation of granulomas, fissures, or transmural fibrosis. However, the epithelial lining and the lamina propria were infiltrated diffuselywith mononuclear cells, which regularly formed focal aggregates in the lamina propria (Figures 2C and 2D). Inflammatory mononuclear cells were also present in the muscularis mucosae. In comparison to wild- type mice (Figure 2E), the large majority of lymphocytes in the lamina propria of diseased Tgs26 mice were CD3+CD4+ T cells (Figures 2F and 2G). Similarly, most of the T cells among the population of IELs were also CD4+, a phenotype observed only for lo%-20% of IELs in wild-type mice (Ta- ble 1). CD6+ T cells were, in contrast, only rarely identified in both the epithelial compartment and the lamina propria (Table 1). Moreover, staining for the expression of MHC

Aberrant Thymic Selection and Severe Colitis 29

Figure 2. Histological Analysis of Colitis in Transplanted Tgs26 Mice

A low power view (magnification, 10x) of a colon from an untransplanted Tgs26 mouse (A) and a transplanted Tgs26 mouse (B) with IBD using hematoxylin-eosin (H-E) staining. In comparison to the untransplanted Tgs26 mouse, the section of the T cell reconstituted Tgs26 mouse displays an increased wall thickness, distortion of the crypt architecture, the formation of crypt abscess (arrow), and the loss of goblet cells. High power view (HE stain, 40 x) of the inflamed colon shows the enlargement of the mucosal layer and the diffuse infiltration with mononuclear cells (C), as well as branching of the crypts and crypt abscesses (ID], arrow). lmmunohistolqical staining of frozen sections from a normal mouse (E) and a transplanted Tgs26 mouse (F) using anti-CD%specific antibodies demonstrate the large number of strongly positive cells in both the lamina propria and the intraepithelial compartment of transplanted Tgs26 mice with colitis. The majority of these infiltrating T cells express CD4 (G).

class II molecules on the surface of large bowel epithelial an apparent decrease in the number of intestinal T cells cells was positive in diseased but not in normal mice (data in the small intestine when compared with wild-type mice, not shown; Bland, 1988). Microscopic analysis from sec- a finding further suggested by the low yield of small intesti- tions of the small bowel failed to demonstrate pathological nal T cells isolated (data not shown). This reduced cell signs of inflammation. However, immunohistology revealed numberparalleled the relativeTcelllymphopeniacommon

Immunity 30

Table I. Phenotypic Analysis of a5 TCR’ Cblonic T Cells

Lymph Node Large Bowel LPL Large Bowel IEL Percentage of Percentage of Percentage of

Mice ab TCR’ CD4" CD0’ a0 TCR+ CD4' CDW af3 TCR’ CD4’ CDV

(C57BL6 x CBA/J)Fl 62 65 35 50 35 65 49 10 62

Tgs26BMT x 1 12 61 19 90 90 4 92 95 5 2 14 03 17 90 76 22 00 96 4 3 6 70 22 96 92 0 61 70 30

Tg ~26 mice were transplanted and lymphocytes from small and large intestine were isolated 5-6 weeks after transplantation as described in Experimental Procedures. Lymphocytes were stained for a5 TCR, CD4, and CD6 expression. The values given are representative for the analysis of 20 mice. (C57BU6 x CBA/J)Fl mice were used as controls. a Defined as the relative percentage of a6 TCR’ cells.

for spleen and peripheral lymph nodes of bone marrow- transplanted Tgs26 mice (Hollander et al., 1995). Thus, the IBD in bone marrow-transplanted Tge26 mice was restricted to the large bowel and was correlated with an increased number of CD4+ af3 TCR+ T cells.

Phenotypic Analysis of Colonic and Peripheral T Cells For further phenotypic analysis, lymphocytes from small and large bowel of normal Fl mice and transplanted Tgs26 mice with colitis were isolated and evaluated by multipa- rameter flow cytometry. All intestinal T cells were of donor origin (H-2k~~; data not shown). As noted by immunohistology, the large majority of IELs in diseased mice were CD4+ a3 TCR+ cells, changing the ratio of CD4+ to CD6+ IELs greatly (Table 1). Untransplanted Tge26 mice were com- pletely devoid of T cells in all secondary and tertiary lymphoid organs (Wang et al., 1994, 1995). To determine whether T lymphocytes in the epithelial lining of sick animals were of extrathymic origin or thymus-derived T cells, both small and large bowel lymphocytes were further phe- notyped. Nearly all CD4+ IELs in transplanted Tgs26 mice with colitis expressed CD5 (Figure 3A, top), a marker used to identify thymus-dependent IELs (Croitoru et al., 1992). To distinguish further the intraepithelial T cells, IELs were also stained for the 135 kDa aE-integrin molecule recog- nized by monoclonal antibody (MAb) M290 (Kilshaw and Murant, 1991). The expression of the M290 epitope is al- most exclusively restricted to T cells of the intraepithelial compartment, since little or no reactivity is seen for other peripheral lymphocytes. Flow cytometry of IELs from transplanted Tgs26 mice with colitis revealed a much lower frequency of aE-integrin-positive cells when compared with IELs from wild-type control animals (Figure 3A, bottom). Taken together, these results suggest that the increased number of lymphocytes and in particular CD4+ CD5+ M290- a3 TCR+ cells found in the epithelial lining of transplanted Tgs26 mice are most likely thymus derived and have immigrated from asecondary lymphoid compartment to the intestinal epithelium. In accordance with these observations is our finding that the majority of the CD6+ Tcells in thiscompartment expressed simultaneously both a and 6 chains of the CD6 coreceptor (data not shown).

The expression of the CD6a chain in the absence of the CD66 chain is, however, unique to IELsand is acharacter- istic feature of the extrathymic lineage development (Poussier and Julius, 1994).

Further phenotypic analysis revealed that most of the T cells in the intraepithelial compartment and the lamina propria expressed surface markers typical of activated T cells. IELs and lamina propria lymphocytes (LPLs) from the colon of transplanted Tgs26 animals with colitis displayed a decreased expression of CD62L when compared with colonic T cells from wild-type mice (Figure 3s). The conjecture that the CD4+ cells in transplanted Tgs26 mice with colitis represent activated lymphocytes was further emphasized by positive staining for CD69, an early T cell activation marker, and for CD44, a marker for memory T cells (data not shown). T cells with an identical phenotype have been noted in lymph nodes and spleens of transplanted Tgs26 mice (Hollander et al., 1995).

Using the surface expression of CD45RB, CD4+ T cells can be differentiated into distinct subpopulations (Bot- tomly et al., 1969; Powrie and Mason, 1990a, 1990b). Functionally, CD45RBhi and CD45RB” subsets of CD4+ T cells are thought to represent naive and memory T cells, respectively, since the CD45RB”’ antigen expression is lost upon Tcell activation (Lee et al., 1990). In transplanted Tgs26 mice with colitis, CD4+ T cells from peripheral and mesenteric lymph nodes, the lamina propria, and the epithelial compartment of the colon displayed mostly a CD45RBb phenotype. In contrast, the corresponding population of CD4+ T cells in wild-type mice displayed gener- ally a CD45RBhi surface expression with a distinct but mi- nor population being CD45RBb. (Figure 3C). Thus, the great majority of the T cells found in the inflamed colon of transplanted Tgs26 mice are of thymic origin and display the phenotype of activated lymphocytes.

Cytotoxic Potential of Colonlc T Cells Previously, we had shown that peripheral T cells in bone marrow-transplanted Tge26 recipients had lost their capacity to respond to allogeneic cells but had retained their function to proliferate if stimulated with mitogens (Hol- lander et al., 1995). Since an increased number of IELs and LPLs could be found within the mucosa of the large bowel of transplanted Tgs26 mice with colitis but not wild-


IEL Transplanted Wild Type

Tg E26

. CD 5

aMzpO -Integrln

LN

B Transplanted Wild Type

4 1 Tg ~-26 1

CD62L

IEL LPL

CD45RB

Figure 3. Phenotypic Analysis of Peripheral T Cells in Tgs26 Transplanted Recipients (A) Staining of large bowel IELs from transplanted Tgs26 mice and wild-type mice for the surface expression of CD5 and the aE-integrin. The increased percentage of CD5+ T cells among IELs of transplanted Tgs26 mice suggests that these cells have developed in the thymus and immigrated to the epithelial compartment. T cells accumulating in the inflamed colon of transplanted Tgs26 mice display a marked daoreaee of MAb M290+ staining in comparison to control mice. (B) Staining for CD62L expression among CD4+CD8‘ IELs from the large bowel of transplanted Tgs26 mice and wild-type mice. In comparison to normal mica, this subpopulation of IELs from transplanted Tgs26 mice has a decreased surface expression of OWL, a marker typical of activated T cells. (C) Analysis of CD4+CD& T lymphocytes from transplanted Tgs26 mice (A) and wild-type mice (B) for the expression of CD45RB. Cells were obtained from axillary lymph nodes(LN), mesenteric lymph nodes (MesLN), lamina proprfa(LPL), andepithelium (IEL)of the colon. The s&population of CD4+CD8- IELs from transplanted Tgs26 mice with colitis display a decreased expression of CD45RB in comparison to wil&ype mice.

Immunity 32

IEL

7---

type mice, we sought to examine their functional capacity using a redirected cytotoxicity assay. As demonstrated in Figure 4, both IELs and LPLs from transplanted Tgs26 mice with inflammatory bowel disease displayed a very high degree of cell-mediated killing with up to 60% lysis of P615 target cells. In comparison, wild-type mice consistently displayed a low killing in redirected cytotoxicity assays. The increased cytotoxic capacity of T cells from Tge26 mice with colitis was, however, limited to intestinal lymphocytes, in that splenic T cells from the same animal did not display redirected cytotoxicity. In the absence of anti-CD3 antibodies, the cell lysis was reduced to background level, suggesting that IELs and LPLs did not kill spontaneously, but that their cytotoxic activity depended on TCR-CDSmediated activation. Furthermore, the nature of the target cells ruled out the possibility, that the increased cytotoxicity was mediated by natural killer cell activity.

Transfer of Peripheral T Ceils Conveys Colitis Intraepithelial T cells in the large intestineof bone marrow- transplanted Tgs26 mice displayed a phenotype typical of thymus-derived activated T cells, while the general capacity for redirected killing is a functional feature characteris- tic of IELs. It was thus conceivable that T cells from spleen

A

Figure 4. Redirected Cytotoxicity of Lympho- cytes

The cytotoxic potential of T cells from transplanted Tgs26 mice (closed circle) and wild- type mice (open circle) were compared. T cells were isolated from spleen and lamina propria and epithelial compartment of the colon and tested in the presence of anti-murine CD3e for their redirected cytotoxic capacity. T cells from Tgs26 mice were also assayed in the absence (open square) of anti-murine CD3s to assess the necessity of TCR-CDBmediated cell activation.

and lymph nodes contained a subpopulation of aberrantly selected cells, which, upon migration to the lamina propria and the intestinal epithelium, acquired the competence for cytotoxicity. To test whether thymus-derived peripheral T cells are indeed accountable for the IBD observed in transplanted Tgs26 mice, T lymphocytes isolated from peripheral (i.e., submandibular, axillary, and inguinal) and mesentery lymph nodes of wild-type mice or transplanted Tge26 mice with colitis by FACS sorting, were injected into I?AG-~“~’ mice or untransplanted ~26 transgenic mice. The RAG-T”” have previously been bred to a C57BU6 background. Over the course of 2-6 weeks, all mice injected with lo6 purified T cells from sick mice developed a distinct wasting syndrome with severe diarrhea. Sick secondary recipients revealed macroscopically enlarged mesenteric lymph nodes and an enlarged colon. Tissue sections of the colon from these mice displayed the com- parable pathological features of structural changes and lymphocyticcell infiltration astypical of the sick donor mice (Figure 5; Figure 6). (Identical results were seen in Tgs26 mice injected with T cells from bone marrow-transplanted Tgs26 mice with colitis.) In contrast, RAG-Pun’ mice that received T cells from control animals remained free of colitis and lymphadenopathy, although T cells were present in the colonic mucosa of these animals (Figure 6).

B

Figure 5. Histological Analysis of the Colon from RAG”“’ Mice Injected with Peripheral T Lymphocytes from Transplanted Tgs26 Mice with Colitis

(A) H-E staining (magnification, 10 x) of the colon from RAGMH mice injected with purified lymph node T cells from transplanted Tgc26 mice with inflammation of the large bowel displays changes alike the pathology seen in the donor mice. The section shows an exceptionally inflammed large bowel with signs of ulcerations, increased thickness of the mucosa, massive infiltration of mononuclear cells, and diminished number of crypts with loss of goblet cells. (B) lmmunohistological staining for CD3s expression in the colon of T cell reconstituted RAG”“” mice demonstrates the prominent infiltration with T cells (magnification, 40 x).


IEL LPL

CDS

Transplantation of Fetal Thymic Tissue Prevents Colitis Taken together, the above experiments suggested that the peripheral T ceils from secondary lymphoid tissue may play an important role in the development of colitis. Since these T ceils matured in the aberrant thymic microenvironment of bone marrow-transplanted Tgs26 mice, their thymic selection may have been abnormal. in particular, the lack of positive selection of regulatory T cells may have resulted in the uncontrolled expansion of T ceils, which may account for the occurrence of colitis. To test this hy- pothesis, adult Tgs26 mice were reconstituted both with normal nontransgenic T ceil precursor cells and with a syngeneic and transgenic but fetal thymic microenvironment. Mice were first grafted with day 16 TgQ6 fetal thymus and subsequently transplanted with T ceil-depleted bone marrow ceils from (C57BU6 x CBA/J)Fl donors. As previously shown, the fetal graft in the transplanted Tge26 mice developed a normal thymic microenvironment, while the orthotopic thymus remained structurally abnormal (Hoiiander et al., 1995). Thymus-transplanted and bone marrow-engrafted mice were alive and disease- free for at least 6 months after transplantation. in contrast, Tgs26 recipients transplanted with only bone marrow cells died within 9 weeks after engraftment.

Detailed analysis of the double-transplanted animals did not disclose generalized lymphadenopathy nor macroscopic features of inflammatory bowel disease. Moreover, normal intestinal histology was observed on cross-sec- tional analysis of the large bowel tissue (data not shown).

IELs from double-transplanted Tgs26 and wild-type mice were isolated from the large bowel and tested phenotypi- tally as well as functionally. Engraftment and subsequent reconstitution of thymic microenvironment resulted in a T

Figure 6. Flow Cytometric Analysis of the Co- lon from RAW Mice Injected wtth Peripheral T Lymphocytes from Transplanted Tgs26 Mice with Colitis Flow cytometry of IELs and LPLs from RAG”“’ mice transplanted with T cells from peripheral (submandibular, axillary, and inguinal) and mesentery lymph nodes of Tgs36 mice with colitis (A) and of wild-type mice (6). Cells were stained for CD4 and CD6 after electronic gating for CD3 expression.

ceil repertoire effective in preventing the expansion of CD4+ T cells in both peripheral and gut-associated lymphoid tissue (Figure 7A; Hoiilinder et al., 1995). Further- more, IELs and LPLs from double-transplanted (i.e., bone marrow and thymus) transgenic and normal mice displayed identical cytotoxicity in a redirected assay (Figure 78). Since the orthotopic thymus continued to have an aberrant T cell selection (Hollander et al., 1995), this finding suggests that a functional thymic microenvironment will select a population of regulatory T ceils competent to prevent the development of colitis.

Discussion

Here, we report that adult Tgs26 mice engrafted with T ceil-depleted bone marrow cells from (C57BU6 x CBA/ J)Fl donors develop a lethal wasting syndrome with severe colitis. Coionic T ceils from the epitheiial lining and the lamina propria exhibited unique characteristics: the majority of IELs and LPLs had a phenotype of activated thymus-derived CD4+ T ceils and displayed greatly increased cytotoxic activity. Although T ceils with an identical phenotype were prevalent in lymph nodes and the spleen of transplanted Tgs26 mice, these cells did not demonstrate increased redirected cytotoxicity. However, the capacity of peripheral T ceils from transplanted Tgs26 mice to induce severe colitis was demonstrated in transfer experiments. injection of a small number of lymph node T ceils (106) caused severe colitis in RAGnU” recipients, which was macroscopically and microscopically identical to the disease in T cell donors. In contrast, none of these striking phenotypic and pathological changes were noted in Tge26 mice transplanted with T cell-depleted (C57BU6 x CBA/J)Fl bone marrow ceils and fetal thymus

Immunity 34

A

B 40

IEL

Z-T-

LPL

ld 3 I 4 10’ za

20 30

E:T Ratio

from day 16 fetal Tgs26 mice. Disease-free long-term survival was also observed in Tgs26 mice that were reconstituted in utero with Fl bone marrow cells and thus developed a functionally normal thymus (data not shown; Holliinder et al., 1995). These bone marrow transplantation and transfer experiments provide evidence that thymic selection in the aberrant microenvironment of Tge26 mice results in a repertoire of peripheral T cells sufficient to induce colitis. Furthermore, the observations in double- transplanted Tgs26 mice underscore the significance of a normal thymic microenvironment for the possible gener- ation of regulatory T cells important in the immunological homeostasis of the mucosa-associated lymphoid tissue.

A number of unique features sets this form of colitis apart from previously described murine models of IBD and

Figure 7. Transplantation of a Normal Thymic Microenvironment Prevents colitis in Trans- planted TgsZ6 Mice: Phenotypic and Func- tional Analysis of Cotonic T Cells (A) Staining for the surface expression of CD4 and CD8 on IELs and LPLs from Tgs26 mice cotransplanted with Tgs26 fetal thymus (day 16) and T cell-depleted (C57BU6 x CBA/J)Fi bone marrow cells (A) and from wild-type mice (B). (B) T cells were isolated from the intraeptthelial compartment from normal Fl mice (open circle) and from TgQ6 mice cotransplanted with Tgs26 fetal thymus (day 16) and T cell-depleted (C57BU6 x CBA/J)Fl bone marrow cells (closed circle). The cytotoxic activity of small bowel IELs from Fl mice (open square) and Tgs26 mice transplanted with thymus and bone marrow cells (closed square) are shown for comparison. The cytotoxic activity displays no difference between Fl mice and transplanted Tgc26 mice.

argues, thus, for multiple pathogenic mechanisms operative in the induction of IBD. In comparison to the study with TCR a mutant, TCR 6 mutant, TCR 6x6 double- mutant, and class II major histocompatibility-deficient mice (Mombaerts et al., 1993), reconstituted Tgs26 mice developed spontaneously a wasting syndrome with severe colitis. Both frequency of disease and kinetics of onset were different between these two models in that bone marrow-transplanted Tgs26 mice developed severe colitis within a few weeks after T cell reconstitution, while only approximately a third of the TCR mutant mice developed colitis gradually after 16 weeks, with differences in the severity of IBD depending on theTCR mutation. A histolog- ically similar form of colitis developed also in mice deficient for IL-2 (Sadlack et al., 1993). Although the predominance


of CD4+ ab TCR+ T cells infiltrating mucosa and submucosa was not reported originally in this IBD model, CD4+CD45RBb ap TCR’ T cells are prominent both in the periphery and in the colon (unpublished data). Despite circumstantial evidence that IBD may develop due to B cell hyperresponsiveness in the absence of IL-2-secreting regulatory T cells, the present report and other studies (Powrie et al., 1993,1994a, 1994b; Morrissey et al., 1993) contrast this contention. We show here that the transfer of unseparated T cells alone is able to induce colitis in T and B cell-deficient recipients. The unique role of T cells in the prevention of colitis was underlined by experiments in which Tg&26 mice were reconstituted with both bone marrow cells and a normal thymic microenvironment. In these double-transplanted recipients, it is likely that regulatory T cells developed in the thymic environment of engrafted Tge26 fetal tissue prevented the induction of colitis by erroneously selected T cells derived from the orthotopic thymus.

Recently, Powrie and co-workers (Powrie et al., 1994b) provided evidence that Thl cells and, in particular, IFNy production are involved in the pathogenesis of a murine model of Crohn’s disease. In comparison to the model presented here, the phenotype of cells competent to transfer colitis was strikingly different. The transfer of CD45RBhi CD4+ T cells from wild-type mice induced severe colitis in syngeneic scid mice, while the injection of unseparated or CD45RBi0CD4+ T cells prevented the induction of disease. The apparent functional heterogeneity between CD45RBhi CD4+ and CD45RB” CD4+ T cells with regards to their capacity to induce colitis may reflect two distinct stages of special T cell function (Powrie and Mason, 1990a, 1990b; Sparshott et al., 1991; Powrie et al., 1994a; Morrissey et al., 1993). The CD45RBh’ CD4+ T cell subset appears to represent naive Thl cells secreting high levels of IL-2 and IFNy but only little IL-4, whereas memory T cells are CD45RBl” and produce both IL-2 and IL4 when stimulated (Birkeland et al., 1966; Bottomly et al., 1969; Lee et al., 1990; Bradley et al., 1992). Further studies in mice infected with the intracellular protozoan Leishmania major have suggested that CD45RBhi T cells may represent also a heterogeneous population of naive cells and primedTh1 cells (Powrieet al., 1994a). Whetherthe preva- lence of CD45RB” T cells in transplanted Tg&26 mice with severe colitis represents a Thl- or The-restricted population, as it does in the models mentioned above, is presently unknown. In any event, the experiments presented here differ substantially from investigations by Powrie and colleagues (Powrie et al., 1994b; Morrissey et al., 1993), in that transfer of unseparated peripheral T cells from transplanted Tge26 mice with bowel disease could cause severe colitis. It is, furthermore, of note, that although the majority of these cells were CD4+ as many as 60% expressed the CD45RB’O phenotype (Figure 3C).

The relative percentage of CD4+ T cells was greatly increased in spleen, peripheral lymph nodes (including the mesenteric lymph nodes), and the gut-associated lymphoid tissue of transplanted Tgs26 mice (Table 1; Figure 2F). However, a normal CD4:CD8 ratio was noted among single-positive mature thymocytes(HollBnder et al., 1995).

This discrepancy suggests that the increase in CD4+ cells among peripheral T cells may be the result of a secondary expansion to a yet unidentified MHC class II-restricted antigen. This explanation would be consistent with the low surface expression of CD45RB among the CD4+ T cells of transplanted Tgs26 mice reflecting a memory cell function after exposure to antigen. Interestingly, Tcells accumulating within the epithelial lining of the large bowel are of thymic origin, as demonstrated by the presence of CD5 expression and possibly by the absence of aE-integrin molecules on the cell surface. Transfer experiments with lymph node T cells suggested further that the IELs had most likely immigrated from secondary lymphoid tissue to the mucosal lining. The molecular nature of this preferen- tial homing to the ilEL compartment remains presently obscure, although it has been suggested that the traffic of T cells from secondary lymphoid tissue (e.g., Peyer’s patches) to the intestinal mucosa may represent a physio- logical event (Guy-Grand et al., 1978). These observations do not exclude, however, the possibility that the predominance of thymus-derived CD4+ T cells within the epithelium of the colon may be the result of further in situ expansion in response to antigens. Indeed, it has been proposed that enterocytes are involved in antigen presentation in as much as they express CD1 (Bleicher et al., 1990) and MHC class I plus class II antigens (Bland, 1986). Consis- tent with this concept, immunohistology of the colon from transplanted Tgs26 mice revealed an increased expression of MHC class II molecules on epithelial cells (Cerf- Bensussan et al., 1984; Mayer and Shlien, 1987; Bleicher et al., 1990; Wu et al., 1991). The likelihood, however, that the antigens are presented to the immune system exclusively within the intestinal compartment is arguable because large numbers of thymus-derived CD4+ T cells with an identical phenotype were also present in lymph nodes and spleen.

Although the different veritable functions of ilELs have yet to be defined, the physical location of these cells within the epithelial layer of the gut has suggested that they may play an important role as a first line of defense against enteric antigens. UnlikeTcells in lymph nodes and spleen, the population of ilELs from the small bowel and to a lesser degree the large bowel comprise spontaneous cytotoxic activity. Although a precise definition of the phenotype of the effector cells and the factors that influence their activation remain largely unknown, both ap TCR’ and y6 TCR’ cells have been shown to mediate redirected cytotoxicity among murine ilELs (Klein and Mosley, 1993). The cytotoxicity observed in transplanted Tgs26 mice was greatly increased among ilELs and LPLs from large bowel in comparison to the spontaneous cytotoxicity displayed by ilELs of normal mice. Importantly, the degree of cytotoxic activity was directly proportional to the overall extent of the wasting syndrome and, in particular, the severity of colitis. Intestinal IELsfrom micewithamarginallyenlarged colon and a less prominent infiltration with CD5’ CD4+ CD45RB’O CD62L- lymphocytes displayed cytotoxic activity that was intermediate between that of severely sick animals and controls (data not shown). Although not for- mally tested, two independent observations suggest that

Immunity 36

CD4+ lymphocytes account fully for this increased degree of cytotoxic activity. First, ap TCR+ CD4+ cells represented occasionally over 95% of all the ilELs recovered in sick Tgs26 mice and these preparations showed the highest degree of redirected cytotoxicity. Second, preliminary experiments with transfer of purified CD4+ T cells from sick mice to RAG”“” recipients resulted both in disease and increased cytotoxicity of infiltrating lymphocytes. It thus appears that the unique environment of the intestine induced the CD4+ T cell immigrants to their novel functional capacity for redirected killing. It is, however, not known whether subpopulations of ilELs exist in normal mice with identical phenotype and function.

The pathogenic function of infiltrating T cells in transplanted Tgs26 mice is presently not determined. It is, however, likely that the increased cytotoxicity of the infiltrating cells contributes to the severe pathological changes ob- sewed. Furthermore, colonic CD4+ T cells may also se- crete proinflammatory cytokines, which perpetuate the severe histopathological changes. An alternative but mutually not exclusive mechanism in the pathogenesis of this form of colitis may also be the lack of regulatory lympho- kines. These potent suppressors of cytokine synthesis by macrophages, natural killer cells, and T cells may regulate the normal immune response to enteric antigens. Thus, their relative or complete absence may contribute to the development of colitis. Consistent with a cytokine-mediated pathogenesis of colitis are the recent findings of increased IFNy and TNF production by CD4+ T cells in the IBD model of Powrie and colleagues (Powrie et al., 1994a, 1994b).

Taken together, this novel model of IBD defines the im- portance of a normal thymic function for the prevention of colitis. The aberrant selection of T cells in the orthotopic thymus of bone marrow-transplanted TgE26 mice has resulted in a peripheral repertoire of cells competent to induce a lethal wasting syndrome. Interestingly, the peripheral expansion of CD4+ T cells upon emigration to secondary and gut associated lymphoid tissue appears to be compatible with the notion that an abnormal thymic microenvironment permitted the selection of potentially self- reactive T cells. Upon recognition of MHC class II-restricted antigens in the periphery, these T cells expand and mediate a process of severe intestinal inflammation. Because of our ignorance of the nature of the antigen, it remains further to be determined why transplanted TgE26 mice develop only severe colonic but not other organ inflammation. Alternatively, but not mutually exclusive, the absolute increase in CD4+ T cells may have been brought about by the absence of positive thymic selection of regulatory T cells. Although this rationale remains speculative, co- transplantation experiments with engraftment of bone marrow cells and fetal thymic tissue support this latter conclusion. By further analyzing the mechanisms operative in immunoregulation, bone marrow-transplanted Tg&26 mice may serve as an invaluable model for the better un- derstanding of the pathogenesis of colitis.

Experimsntsl Procedures

Mice and Bone Marrow Transplant&Ion The Tge26 mice were generated as previously described by overex.

pression of the full-length human CD& gene (Wang et al., 1994). Transgenic mice werecrossed into CBA/J (H-27, mated lo homozygos- ity, and screened by Southern blotting. RAG-2”1 mice (H-23 and (C57BU6 x CBAN)Fl were purchased from Jackson Laboratories (Bar Harbor, Maine). Tcell-depleted (C57Bl/6 x CBAfJ)Fl bonemar- row cells were used as donor cells to avoid altorecognition of the TgQ6 and RAG2”” recipients, aged 6-12 weeks. T cell depletion was achieved by two rounds of treatment with MAb anti-Thyl.2 (clone 30. H12) followed by rabbit complement lysis. Bone marrow recipients were pretreated with Mluorouracil (5FU; 150 mg/kg) 48 hr prior to transplantation of 10 x 10T cell-depleted bone marrow cells. Identi- cal iesults were obtained in recipient animals using total body irradi- ated (INO rads).

In experiments to ascertain normal thymic function, fetal day 16 thymic lobes from homozygous Tg~26 embryos were transplanted uni- laterally under the kidney capsule of adult Tge26 mice 24 hr after injection of 5FU. The graft was followed 1 day later by transplantation of IO x 1W T cell-depleted bone marrow cells.

HIstologlcsl Anslyrla Tissue from the small and large intestine were surgically removed. For conventional histology, tissue samples were fixed in 10% buffered formalin and embedded in paraffin. Sections (4 pm) were cut and placed on gelatin-coated microscope slides for staining with hematoxylin and eosin using standard techniques.

For immunohistology, tissue samples were embedded in OCT com- pound (Ames, Elkhart, Indiana), snap-frozen in liquid nitrogen, and subsgquently stored at -80°C. Sections were stained by an avidin- biotin complex method as described previously (Cerf-Bensussan at al., 1963; Mombaerts et al., 1993). In brief, 4 pm thick tissue sections were air dried for 2 hr, fixed in acetone for 7 min. air dried again, and then incubated with various primary antibodies. For detection, biotinylated secondary antibodies were used, followed by a I:100 dilu- tion of avidin-biotinylated paroxidase complex (Dako, Santa Barbara, California). Each step was followed by three washes with phosphate- bufferadsaline(PBS). Thestaining wasdonein asolutionof39minc& ethylcarbazol (Aldrich, Milwaukee, Wisconsin), followed by postfixa- tion with 2% pareformaldehyde and counterstained with hematoxylin. The antibodies used were the following: anti-CDSa (clone YCD3.1), anti-CD4 (GK1.5). anti-CD8 (53-6.7). anti-ab TCR (H57-597), anti+ TCR (GL3). anti-MHC class II (I-Ah; 1 l-5.2) followed by either goat anti-hamster (1:20, Vector, Burlingama, California) or rabbi anti-rat (l:lOO, Vector, Burlingame. California). Incubation (30 s) with 0.3% hydrogen peroxide in PBS was used to block endogenous peroxidase activity, while sequential incubations with avidin and biotin (Vector, Burlingame, California) was used to block endogenoua biotin.

laolstlon of Inteatlnsl Lymphocytea Small and large intestines were surgically removed and washed by flushing with cold PBS containing 5% fetal calf serum until the lumen was free of detritus. The intestine was then cu1 into small pieces for separate isolation of lamina propria and intraepithelial lymphocytes using a standard protorol previously described (Lefranwia, 1994). In brief, ilELs were separated from nonlymphoid tissue by shaking twice at 37OC in PBS supplemented with 2% fetal calf serum and 2.5 x lo* M f3-mercaptoethanol. After one further incubation in 1 mM EDTA lo remove remaining epithelial cells, LPLs ware isolated using anzy- matic digestion of tissue as previously described (Sadlack et al., 1993). Both lymphocyte populations were subsequently enriched on a discon- tinuous percoll gradient (67%, 44%).

flow Cytomewlc Analyala and Call Sorting Isolated IELs and LPLs were preblocked with anti-CDW32 (clone 2.462) and then stained for three-color flow cytometric analysis using a combination of the following MAbsz anti-CD3 (145-X11); antiCD4 (RM4S); anti-CD5 (53-7.3); antiCD8o (53-8.7); antiCD6@ (63-5.8); anti-CD44 (IM7); antiCD45RB (16A); anti-CD62L (MEL-14); ant&E integrin (M290); anti-a8 TCR (H57-597); and anti76 TCR (GU) (all obtained from PharMingen, San Diego, California). Data was acquired and analyzed using a FACScan and Lysys &Ware (Becton Dickinson, San Jose, California), respectively. For cell sorting, lymph node cells were stained simultaneously for CD4 (RM4-5) and CD8a (53.8.7) and positively selected using cell sorting with a FACSVantags (Coultar, Hialeah, Florida).


Cytotoxiclty Assay Cytotoxic activity was measured in standard s’Cr-reIease assay. Freshly isolated IELs and LPLs were washed extensively in PBS and resuspended in RPM1 1640 supplemented with 5% fetal calf serum and penicillin, streptomycin, Lglutamine and 2.5 x 1 Od M 6-mercaptoethanol. These effector cells were then incubated in g&well microtiter plates with 5’Cr-labeled P815-target cells at varying effector:target ratios in thepresenceorabsenceof anti-CD3 MAb(1 ug/ml; 1452Cil). After 4 hr of incubation at 37OC in 5% CO,, supernatant (100 ul) was harvested from each well for counting. The percentage specific lysis was calculated as follows:

Experimental s’Cr-reIease - spontaneous 6’Cr-reIease x 1oo Maximum 5’Cr-reIease - spontaneous s’Cr-reIease

Acknowledgments

We thank D. Allen and K. Mobisson for expert technical assistance. This work was supported by grants from the National Institutes of Health (P30 DK43351 to C. T. and A. K. B.; 5 PO1 CA39542-09 to S. J. 6.; and ROl DK47877-01 to A. K. B.), and the Crohn’s and Colitis Foundation of America to C. T. S. J. S. is a recipient of a fellowship from the Crohn’s and Colitis Foundation of America.

Received January 17, 1995; revised April 20, 1995.

References

Barrett, T. A., Delvy, M., Kennedy, D. M., Lefrancois, L., Matis, L. A., Dent, A. L., Hendrick, S. M., and Bluestone, J. A. (1992). Mechanisms of self-tolerance of ~8 T cells in epithelial tissue. J. Exp. Med. 775, 85-70.

Barrett, T. A., Koons, A. M., Bluestone, J. A., and Hurst, S. D. (1993). Tolerance of intraepithelial lymphocytes in the intestine. In Mucosal Immunology: lntraepithelial Lymphocytes, H. Kiyono and J. Ft. McGhee, eds. (New York: Raven Press), pp. 115-130.

Birkeland, M. L., Metlay, J., Sanders, V. M., Fernandez-Botran, R., Vitetta, E. S., Steinman, R. M., and Pure, E. (1988). Epitopeson CD45R IT2001 molecules define differentiation antigens on murine B and T lymphocytes. J. Mol. Cell. Immunol. 4, 71-85.

Bland, P. (1988). MHC class II expression by the gut epithelium. Immu- not Today 9, 174-l 78.

Bleicher, P. A., Balk, S. P., Hagen. S. J.. Blumber, R. S., Flotte, T. J., and Terhorst, C. (1990). Expression of murine CD1 on gastrointestinal epithelium. Nature 250, 879-682.

Bottomly, K., Luqman, M., Greenbaum, L., Carding, S., West, J., Pas- qualini, T., and Murphy, D. B. (1989). A monoclonal antibodytoCD45R distinguishes CD4 T cell populations that produce different cytokines. Eur. J. Immunol. 16, 617-623.

Bradley, L. M., Atkins, G. G., and Swain, S. L. (1992). Longterm CD4+ memory T cells from the spleen lack MEL-14, the lymph node homing receptor. J. Immunol. 148, 324.

Brynskov, J., Nielson, 0. H., Ahnfelt-R&tne, I., and Bendtzen, K. (1992). Cytokines in inflammatory bowel disease. Stand. J. Gastroen- terol. 27, 897-906.

Cerf-Bensussan, N., Schneeberger, E. E., and Bhan, A. K. (1983). lmmunohistologic and immunoelectron microscopic characterization of the mucosal lymphocytes of human small intestine by the use of monoclonal antobodies. J. Immunol. 730, 2615-2622.

Cerf-Bensussan, N., Ouaroni, A., Kurnick, J. T., and Bhan, A. K. (1964). lntraepithelial lymphocytes modulate la expression by intestinal epithelial cells. J. Immunol. 732, 224-235.

Croitoru, K., Bienenstock, J., and Ernst, P. B. (1992). CD5 negative murine intraepithelial lymphocytes (IEL) represent a thymic- independent lineage with distinct repertoire. FASEB J. 6, 1707-1715.

Guy-Grand, D., Griscelli, C., and Vassalli, P. (1978). The mouse T lymphocyte, a novel type of T cell: nature, origin and traffic in mice in normal and graft versus host conditions. J. Exp. Med. 748, 1661- 1677.

Hollander, G. A., Wang, B., Nichogiannopoulou, A., Platenberg, P.-P., van Ewijk, W., Burakoff, S. J., Gutierrez-Ramos, J.-C., and Terhorst,

C. (1995). A developmental control point in the induction of the thymic cortex regulated by a subpopulation of prothymocytes. Nature 373, 350-353.

Kilshaw, P. J., and Murant, S. J. (1991). Expression and regulation of 61(6p) integrins on mouse lymphocytes: relevance to the mucosal immune system. Eur. J. Immunol. 27,2691-2597.

Klein, J. A. and Mosley, R. L. (1993). Phenotypicand cytotoxiccharac- teristics of intraepithelial lymphocytes. In Mucosal Immunology: Intra- epithelial Lymphocytes. H. Kiyono and J. R. McGhee, eds. (New York: Raven Press), pp. 33-60.

Kroemer. G., and Martinez, A. C. (1992). Mechanisms of self tolerance. Immunol. Today 73, 401-404.

Kiihn, R., LOhler, J., Rennick, D., Rajewski, K., and Miiller, W. (1993). Interleukin-10 deficient mice develop chronic enterocolitis. Cell 75, 263-274.

Lee, W. T., Yin, X. M., and Vitetta, E. S. (1990). Functional and ontoge- netic analysis of murine CD45R” and CD45Rb CD4+ T cells. J. Immu- nol. 744, 3288.

Lefrancois, L. (1994). Isolation of mouse small intestinal intraepithelial lymphocytes. In Current Protocols in Immunology, J. E. Coligan, A. M. Kruisbeek, D. H. Marguiles, E. M. Shevach, and W. Strober, eds. (New York: J. Wiley and Sons, Incorporated), pp. 3.19.1-3.19.7.

Ligumski, M., Simon, P. L., Karmeli, F., and Rachmilewitz, D. (1990). Role of interleukin-1 in inflammatory bowel disease: enhanced production during active disease. Gut 37, 686889.

Mayer, L., and Shlien, R. (1987). Evidence for function of la molecules on gut epithelial cells in man. J. Exp. Med. 766, 1471-1483.

Mihada, Y. Ft., Wu, K., and Jewell, D. P. (1989). Enhanced production of interleukin-16 by mononuclear cells isolated from mucosa with active ulcerative colitis or Crohn’s disease. Gut 30, 835-838.

Mombaerts, P., Mizoguchi, E., Grusby, M. J., Glimcher, L. H., Bhan, A. K., and Tonegawa, S. (1993). Spontaneous development of inflammatory bowel disease in T cell receptor mutant mice. Cell 75, 275- 282.

Morrissey, P. J., Charrier, K., Braddy, S., Liggitt, D., and Watson, J. D. (1993). CD4+ T cells that express high levels of CD45RB induce wasting diserase when transferred into congenic severe combined immunodeficient mice: disease development is prevented by co- transfer of purified CD4+ T cells. J. Exp. Med. 778, 237-244.

Pfeffer, K., and Mak, T. W. (1994). Lymphocyte ontogeny and activation in gene targeted mutant mice. Annu. Rev. Immunol. 12,367-411.

Podolsky, D. K. (1991a). Inflammatory bowel disease I. N. Engl. J. Med. 325,928-937.

Podolsky, D. K. (1991b). Inflammatory bowel disease II. N. Engl. J. Med. 325, 1008-1016.

Poussier, P., and Julius, M. (1994). Thymus independent T cell development and selection in the intestinal epithelium. Annu. Rev. Immunol. 72, 521-553.

Poussier, P., Eduard, P., Lee, C., Binnie, M., and Julius, M. (1992). Thymus-independent development and negative selection of T cells expressing T cell receptor a5 in the intestinal epithelium: evidence for distinct circulation patterns of gut and thymusderived T lymphocytes. J. Exp. Med. 776, 187-199.

Poussier, P., Teh, H. S., and Julius, M. (1993). Thymus-independent positive and negative selection of T cells expressing an MHC class I restricted transgenic T cell receptor o/6 in the intestinal epithelium. J. Exp. Med. 778, 1947-1957.

Powrie. F., and Mason, D. (1990a). Subset of rat CD4+ T cells defined by their differential expression of variants of the CD45 antigen: developmental relationships and in vitro and in vivo functions. Curr. Topics Microbial. Immunol. 159, 79-92. ’

Powrie, F., and Mason, D. (199Ob). Ox22”w CD4+ T cells induce wasting disease with multiple organ pathology: prevention by the 0X-22” subset. J. Exp. Med. 772, 1701.

Powrie, F., Leach, M. W., Mauze, S., Caddle, L. B, and Coffman, R.B. (1993). Phenotypically distinct subsets of CD4’ T cell induce or protect from chronic intestinal inflammation in C.B-17 scid mice. Int. Immunol. 5, 1461-1471.

Powrie, F., Correa-Oliveira, R., Mauze, S., and Coffman, R. L. (1994a).

Immunity 38

Regulatory interactions between CD45RBhi and CD45RBbCD4+T cells are important for the balance between protective and pathogenic cell- mediated immunity. J. Exp. Med. 179. 589-600.

Powrie, F., Leach, M. W., Mauze, S., Menon, S., Barcomb Caddle, L., and Coffman, R. L. (1994b). Inhibition of Thl responses prevents inflammatory bowel disease in scid mice reconstituted with cd45RB”’ CD4+ T cells. Immunity 7, 553-562.

Robey, E., and Fowlkes. B. J. (1994). SelectiveeventsinTcell development. Annu. Rev. Immunol. 12, 675-705.

Rocha, B., Vassalli, P., and Guy-Grand, D. (1991). The V5 repertoire of homodimeric a CD8’ intraepithelial lymphocytes reveals a major extrathymic pathway of T cell differentiation. J. Exp. Med. 173, 483- 486.

Rocha, B., von Boehmer, H., and Guy-Grand, D. (1992). Selection of intraepithelial lymphocytes with CD8 co-receptors by self-antigen in the murine gut. Proc. Natl. Acad. Sci. USA 89, 5336-5340.

Sadlack, B., Merz, H., Schorle, H., Schimpl, A., Feller, A. C., and Horak, I. (1993). Ulcerative colitis-like disease in mice with adisrupted interleukin-2 gene. Cell 75, 253-261.

Sartor, R. B. (1990). Roleof intestinal microflorain initiation and perpet- uation of inflammatory bowel disease. Canadian J. Gastroenterol. 4, 271-277.

Sartor, R. B. (1994). Cytokines in intestinal inflammation: pathophysio- logical and clinical considerations. Gastroenterology 106, 533-549.

Sparshott, S. M., Bell, E. B., and Sarawar, S. R. (1991). CD45R CD4 T cell subset-reconstituted nude mice: subset dependent survival of recipients and bi-directional isoform switching. Eur. J. Immunol. 21, 993-l 000.

von Boehmer, H. (1990). Developmental biology of T cell in T cell- receptor transgenic mice. Annu. Rev. Immunol. 8, 531-556.

Wang, B., Biron, C., She, J., Higgins, K., Sunshine, M.-C., Lacy, L., Lonberg, N., and Terhorst, C. (1994). A block in both early T lymphocyte and natural killer cell development in transgenic mice with high copynumbersoftheCD3~gene. Proc. Natl. Acad. Sci. USAQ7,9402- 9406.

Wang, B., Levelt, C., Salio, M., Zheng, D., Sancho, J., Liu, C.-P., Huang, M., Higgins, K., Sunsine, M.J., Eichman, K., Lacy, E., Lon- berg, N., and Terhorst, C. (1995). Abrogation of early T cell develop ment by excessive signal transduction through CD3-E. Int. Immunol. 7, 435-448.

Wu, M., Van Kaer, L., Itohara, S., and Tonegawa, S. (1991). Highly restricted expression of the thymus leukemia antigens on intestinal epithelial cells. J. Exp. Med. 174, 213-218.

Date post:	21-Jan-2023
Category:	Documents
Upload:	uni
View:	0 times
Download:	0 times

Severe colitis in mice with aberrant thymic selection

Documents