Gnotobiotic IL-2-Deficient Mice Lymphocyte Development in Colitis ...

Lymphoid Hyperplasia, Autoimmunity, and Compromised Intestinal Intraepithelial Lymphocyte Development in Colitis-Free Gnotobiotic IL-2-Deficient Mice1 Nikhat V. Contractor,2* Hamid Bassiri,2* Tannishtha Reya,3* Audrey Y. Park,* Daniel C. Baumgart,*† Mariusz A. Wasik,‡ Stephen G. Emerson,§ and Simon R. Carding4 IL-2-deficient (IL-22/2) mice develop disorders of the hemopoietic and immune systems characterized by anemia, lymphocytic hyperplasia, and colitis. The mechanisms responsible for these abnormalities remain unclear. To investigate the underlying basis of autoimmunity, the particular role of commensal gut flora in the initiation of colitis, and the role of IL-2 in the development of intestinal intraepithelial lymphocytes (iIEL), we evaluated IL-22/2 mice reared and maintained under gnotobiotic (germfree) conditions. By 8 wk of age, 80% (20 of 25) of germfree IL-22/2 mice show signs of disease, including anemia, disturbances in bone marrow hemopoietic cells, lymphocytic hyperplasia, and generalized autoimmunity, similar to those seen in specific pathogen-free (SPF) IL-22/2 mice. In striking contrast to SPF IL-22/2 mice, germfree IL-22/2 mice do not develop colitis. However, the numbers of gd1 and TCRab1CD8aa1 iIELs are reduced, and in lethally irradiated SPF IL-21/1 mice, reconstituted with IL-22/2 bone marrow TCRgd1 iIELs fail to develop, consistent with an important role of IL-2/IL-2R signaling in the development of gd iIELs. Consequently, our findings demonstrate that the colitis seen in SPF IL-22/2 mice depends upon the presence of intestinal bacterial flora and that environmental Ags are not responsible for the anemia and extraintestinal lymphoid hyperplasia that occur in IL-22/2 mice. Thus, germfree IL-22/2 mice represent a unique system in which the role of IL-2 deficiency in hemopoietic and immune system disorders can be investigated in dissociation from complications that may arise due to colitis. The Journal of Immunology, 1998, 160: 385–394.

C

onventionally reared and maintained IL-2-deficient (IL22/2)5 mice develop a disorder of the hemopoietic and immune systems characterized by anemia, lymphocytic hyperplasia, progressive loss of B cells, disturbances in bone marrow hemopoietic cells, and an inflammatory bowel disease resembling human ulcerative colitis (1–3). A similar lymphocytic hyperplasia and autoimmunity has been described in mice lacking the b-chain of the IL-2R (4), supporting the hypothesis that IL-2/ IL-2R signaling is required for the regulation of T cell responses in vivo. The mechanisms that initiate disease in IL-22/2 mice are not yet known, but like other autoimmune disorders, it is hypothesized that disease may result from abnormal T cell responses to environmental Ags (5). Alternatively, the disease may occur due to compromised thymic and peripheral deletion of self-reactive T cells.

Departments of *Microbiology, †Surgery, ‡Pathology and Laboratory Medicine, and §Hematology and Oncology, University of Pennsylvania School of Medicine, Philadelphia, PA 19104 1 This work was supported by grants from the University of Pennsylvania Research Foundation, the American Cancer Society (RPG-97-027), and the National Institutes of Health; a Dolores Zohrab Liebmann Fellowship (to N.V.C.); a fellowship from Deutscher Akademischer Austauschdienst (to D.C.B.), Germany; and a University of Pennsylvania M.D./Ph.D. Benjamin Franklin Scholarship (to A.P.). 2

The first two authors contributed equally to this work.

3

Current address: Department of Microbiology and Immunology, Howard Hughes Medical Institute, University of California San Francisco, San Francisco, CA 94143. 4 Address correspondence and reprint requests to Dr. Simon R. Carding, Department of Microbiology, University of Pennsylvania, 303A Johnson Pavilion, Philadelphia, PA 19104-6076. E-mail address: [email protected] 5 Abbreviations used in this paper: IL-22/2, interleukin-2-deficient mice; SPF, specific pathogen free; iIEL, intestinal intraepithelial lymphocytes; int, intermediate level.

Copyright © 1998 by The American Association of Immunologists

The hypothesis that disease induction is caused by a breakdown in the regulation of peripheral T cells is supported by several observations. First, IL-2 has been demonstrated to mediate the termination of clonal expansion by inducing apoptosis (6). Second, superantigen-stimulated T cells from IL-22/2 (7) and IL-2R achain-deficient (8) mice are more resistant to Fas-induced apoptosis than those from wild-type mice. Finally, the onset of the disease in IL-22/2 mice reared and maintained in a specific pathogen-free (SPF) environment is delayed (1). Together, these observations suggest that disease is caused by dysregulated peripheral T cell responses to environmental Ags. Despite the initial analysis of IL-22/2 mice indicating normal T cell development (9), the recent finding that IL-22/2 athymic mice fail to develop autoimmune or lymphoproliferative disorders demonstrates that disease initiation depends upon the intrathymic differentiation of IL-22/2 T cells (10). Attempts to identify abnormalities in intrathymic T cell development and negative selection in IL-22/2 mice, however, have produced contradictory results. The marked skewing toward the CD4281 phenotype in the thymi of nonmanipulated IL-22/2 mice (11) and the decreased cortical apoptosis and loss of immature CD4181 (double-positive) and accumulation of CD41 and CD81 (single-positive) thymocytes in IL-22/2 mice following antigenic stimulation (12) are consistent with abnormal intrathymic development. However, the normal selection and deletion of thymocyte subsets by endogenous superantigens or following chronic administration of Ag to MHC class I-restricted TCR-transgenic mice in the absence of IL-2 (13) argue for T cell selection mechanisms being intact in these mice. An obligatory role for IL-2 in the normal development of at least some T cell populations is supported by the recent finding that the composition of the intestinal intraepithelial lymphocytes (iIEL) compartment is greatly altered in IL-2R b-chain-deficient mice 0022-1767/98/$02.00

386 (14). In particular, there is a marked decrease in TCRab1CD8aa1 and TCRgd1 iIEL subsets. However, since both IL-2 and IL-15 signal through the IL-2R b-chain (15, 16), it is difficult to distinguish the cytokine that is responsible for the iIEL developmental defect in these mice. To investigate further the mechanism of disease in IL-22/2 mice, gnotobiotic (germfree) IL-22/2 mice were rederived and evaluated for signs of hemopoietic and immune disorders. The composition of the iIEL repertoire was also evaluated in these animals. The results show that germfree IL-22/2 mice develop hemopoietic disorders similar to those in SPF IL-22/2 mice, consistent with the hypothesis that defective central and/or peripheral tolerance to self Ags results in autoimmunity. The absence of colitis and abnormalities in the development of TCRgd1 iIEL subsets in germfree IL-22/2 mice and IL-22/2 bone marrow chimeras also demonstrate that intestinal lesions in SPF IL-22/2 mice may arise as a consequence of an abnormal mucosal immune response to enteric Ags.

Materials and Methods Mice Gnotobiotic (germfree) IL-22/2 mice (Taconic Farms, Germantown, NY), backcrossed eight generations onto the C57BL.6 background, were housed in isolator cages within a flexible film isolator in the gnotobiotic facility of the Biology Department, University of Pennsylvania. The germfree status of IL-21/1, IL-21/2, and IL-22/2 animals was verified by bacteriologic, histologic, and serologic analysis of tissues and fluids at autopsy. IL-22/2 mice obtained from heterozygous matings were identified from samples of tail DNA using a previously described PCR-based method (9). Heterozygote and/or wild-type littermates were used as controls. B6.SJLPtprcaPep3b/BoyJ(Ly5a ) (B6(CD45.1)) mice were a gift from Dr. Edward Scott (University of Pennsylvania, Philadelphia, PA).

Histopathology Tissues removed at autopsy were fixed in formalin and embedded in paraffin. Sections were stained according to standard protocols with hematoxylin-eosin or stains for the detection of microorganisms that included Gram, acid-fast, Warthin-Starry silver, and Gomori methanamine-silver nitrate stains.

Cell preparation Blood was obtained by tail bleeding or cardiac puncture and was collected in polypropylene tubes containing heparin. Blood counts were obtained using a Baker 1000 automated cell counter (Coulter, Hialeah, FL) and leukocyte differential counts were performed manually on Wright-stained blood smears. Mononuclear cells from thymus and spleen were prepared by disrupting intact tissues into ice-cold PBS. Bone marrow cells were obtained from femurs and tibias by flushing the bones with PBS and scraping the bone fragments. To prepare iIEL, the small intestine was opened longitudinally in HBSS, Peyer’s patches were removed, and the intestine was cut into small fragments that were incubated in HBSS containing 0.1 M EDTA and 5 mM DTT in an Erlenmeyer flask in a shaking water bath at 37°C for 30 min. Filtered cells were enriched for lymphocytes by Percoll gradient centrifugation, removing cells between the 65 and 45% interface. The purity of the iIEL was assessed by the degree of contamination by lamina propria-derived B2201, CD32 B cells. Intrahepatic lymphocytes were isolated by first irrigating the intact tissue with PBS to remove circulating cells. The tissue was then disrupted by passing it through a steel mesh, and the lymphocytes were enriched by Percoll gradient centrifugation as described above.

Bone marrow chimeras B6(CD45.1) mice (n 5 4 –5/group) were lethally irradiated (1.1 Gy) and injected with 1 3 106 unmanipulated bone marrow cells from B6(CD45.1) syngeneic donors or IL-22/2 (CD45.2) congenic donors. Control mice were lethally irradiated and injected with an equal volume of PBS. Control mice died 10 to 14 days postirradiation. Experimental mice were euthanized at 90 days postirradiation, and iIELs were isolated as described.

Abs and flow cytometry The following mouse and rat mAbs were used to stain and analyze thymocytes isolated from germfree IL-22/2 and IL-21/1 mice: CD90/Thy 1.2

AUTOIMMUNITY IN GERMFREE IL-2-DEFICIENT MICE (30H.12; Life Technologies, Gaithersburg, MD), CD3 (29B; Life Technologies), CD4 (CT-CD4; Caltag Laboratories, San Francisco, CA), CD8a (53-6.7; Life Technologies), CD8b (CT-CD8b; Caltag Laboratories), CD11b/Mac-1 (M1/70.15; Caltag Laboratories), CD24/heat stable Ag (M1/69; PharMingen, San Diego, CA), CD25/IL-2Ra (PC61.5.3, American Type Culture Collection, Rockville, MD), CD45R/B220 (RA3-6B2; Life Technologies), CD69 (H1.2F3; PharMingen), TCRab (H57-597; PharMingen), TCRgd (GL3; PharMingen), and Gr-1 (RB6-8C5; PharMingen). Donor-derived hemopoietic cells in bone marrow chimeras were detected using mouse-anti-CD45.2 (104; PharMingen). Streptavidin-phycoerythrin, streptavidin-Red 670 (Life Technologies), and/or streptavidinallophycocyanin (Caltag Laboratories) were also used to detect the reactivity of biotinylated primary Abs. Ab staining and three- or four-color flow cytometric analyses were conducted using a FACScan and CellQuest software (Becton Dickinson, San Jose, CA) as previously described (17).

Statistical analysis Data are presented as the mean 6 SD. Statistical significance was assessed by Student’s t test.

Results

Onset of disease in germfree IL-22/2 mice To test the hypothesis that disorders of the hemopoietic and immune systems and colitis that develops in IL-22/2 mice are initiated by an uncontrolled peripheral activation of T cells by environmental Ags (1, 10), IL-22/2 mice were rederived and maintained in a germfree environment. Histopathologic and microbial analyses of germfree IL-22/2 mice confirmed the absence of contagious pathogens (e.g., Gram-positive and Gram-negative (including Helicobacter sp) bacteria, fungi, and parasites). In addition, serologic tests for Abs against the most common murine pathogens (mouse hepatitis virus, polyoma virus, minute virus of mice, mouse adenovirus, mouse CMV, reovirus type 3, mouse pneumonia virus, Theilers’ virus, Sendai virus, and lymphocytic choriomeningitis virus) were negative. Up to approximately 8 wk of age, IL-22/2 mice exhibited no clinical signs of disease, although most were distinguishable from heterozygote or wild-type littermates by their smaller size and weight. Between 8 and 12 wk of age, 80% (20 of 25) of IL-22/2 mice developed anemia (Table I), the onset and severity of which varied with age among individual homozygous mice within the same and different litters. All the germfree heterozygous and wildtype mice have, to date, been free of any clinical or histopathologic signs of disease. In addition, a subset (5 of 25) of germfree IL22/2 mice remained clinically healthy beyond 6 mo of age. The reason for this heterogeneity in disease progression is currently not known, but the involvement of any murine commensal or pathogenic micro-organisms has been excluded. Visual inspection at autopsy of 8- to 12-wk-old IL-22/2 mice revealed a moderate to severe splenomegaly and lymphadenopathy, similar to those seen in SPF IL-22/2 mice. In contrast, the gastrointestinal tract, with the exception of an enlarged cecum ordinarily seen in germfree mice, was normal by visual inspection and histopathology in all germfree IL-22/2 mice analyzed. Thus, germfree IL-22/2 mice do not develop colitis. However, they still exhibit signs of immune system and hemopoietic disorders. To investigate this in further detail, we evaluated nonlymphoid, lymphoid, and hemopoietic organs for signs of disease.

Lymphoid hyperplasia The most consistent histopathologic finding of diseased germfree IL-22/2 mice was a moderate to severe lymphocytic hyperplasia (Fig. 1) affecting hemopoietic, lymphoid, and other organs of all (n 5 20) animals analyzed. Among nonlymphoid tissues, lymphocytic infiltrates were seen in the lungs, liver, pancreas, and kidneys

The Journal of Immunology

387

Table I. Peripheral blood analysis of germfree IL22/2 micea Differential (3109/L) Age (weeks)

IL-2 Genotype

3–5

1/1 1/2 2/2 1/1 1/2 2/2 1/2 2/2

8 12

n

RBC (106/ml)

HB (mg/dl)

5 3 4 3 6 4 4 7

7.9 6 0.9 7.2 6 1.2 7.4 6 0.8 8.6 6 0.2 8.9 6 0.9 3.8 6 0.6 9.9 6 1.9 2.6 6 2.5

15.0 6 0.5 14.8 6 0.3 13.1 6 1.1 14.2 6 0.7 15.6 6 1.0 10.0 6 0.5 14.9 6 1.8 3.7 6 2.3

HT

WBC (103/ml)

Neut

Lymph

Mono

Eosin

33 6 1 32 6 2 28 6 5 31 6 1 31 6 2 19 6 6 33 6 4 12 6 3

7.6 6 1.5 6.8 6 2.0 8.0 6 2.6 11.0 6 1.9 10.1 6 2.1 9.8 6 2.5 13.1 6 2.1 14.0 6 2.2

0.6 6 0.2 0.7 6 0.2 0.3 6 0.2b 1.7 6 0.3 1.6 6 0.3 0.6 6 0.6c 1.8 6 0.3 0.2 6 0.2d

6.2 6 1.1 6.4 6 1.3 7.0 6 1.4 9.7 6 2.5 8.6 6 2.6 6.3 6 2.2* 11.0 6 3.0 5.9 6 2.0*

0.1 6 0.1 0.2 6 0.1 0.1 6 0.1 0.2 6 0.2 0.2 6 0.1 0.6 6 0.2 0.2 6 0.1 0.6 6 0.2

0.08 6 0.02 0.09 6 0.03 0.1 6 0.04 0.1 6 0.06 0.1 6 0.02 0.2 6 0.05 0.2 6 0.1 0.3 6 0.2

a HB, hemaglobin; HT, hematocrit; WBC, white blood cells. Data represent the mean 6 SD. Manual 300-count leukocyte differentials were performed on blood smears from sex-matched littermates. Data represent the mean 6 SD. b p , 0.01 compared with 1/1 mice. c p , 0.005 compared with 1/1 mice. d p , 0.0005 compared with 1/1 mice. * Cells consisted of atypical blast-like cells that were tartrate sensitive, acid phosphatase-positive, and myeloperoxidase-negative.

(Fig. 1, a– d). Chronic pancreatitis was associated with acinar cell necrosis and atrophy as well as ductal hyperplasia, although the islets of Langerhans were unaffected by the inflammatory process. Histochemical analysis of cytocentrifuge preparations of the cellular infiltrate in the lungs showed that they comprised atypical lymphocytes, as defined by their large blast-like appearance, lack of myeloperoxidase, and presence of tartrate-resistant acid phosphatase (data not shown). Additionally, occasional foci of cell necrosis in the heart and trachea were noted in some diseased IL22/2 mice. The degree of lymphocytic hyperplasia was most marked in the lymphoid tissues. Lymph nodes of sick germfree IL-22/2 mice displayed an atypical follicular hyperplasia, with germinal centers showing poorly defined borders (data not shown). The spleens of IL-22/2 mice showed a progressive effacement in architecture due to an expansion of the white pulp areas and an increase in hemopoietic (erythroid and myleoid) cells within the red pulp (Fig. 1 and Table II). Compared with the spleen of IL21/1 mice, the red pulp of IL-22/2 mice was expanded (Fig. 1g). In younger animals (,8 wk of age), only a mild reactive white pulp hyperplasia was noted. Thus, these animals show signs of lymphocytic infiltrates in multiple organs. The exact nature of these lymphocytic infiltrates was determined by examining the lymphoid, myeloid, and erythroid cells present in the spleens and the cellular infiltrate of the livers in germfree IL-22/2 mice.

22/2 mice (Fig. 2), a phenotype normally expressed by activated T cells that accumulate and undergo apoptosis in the liver (18).

Changes in splenic lymphocyte phenotype By 6 wk of age the number of mononuclear cells recovered from the spleens of IL-22/2 mice increased by 1.5- to 2-fold compared with that in IL-21/1 mice. This change was attributable to increased numbers of both CD41 and CD81 T cells (Fig. 3) and of mature myeloid (Mac-1high) cells similar to those seen in SPF IL22/2 mice (Figs. 2 and 3). The CD41 T cells exhibited an activated phenotype, as shown by expression of CD69 and CD25 (Fig. 3A). Unlike the T cells, the proportion of B cells, as determined by anti-B220 Ab staining, was decreased in the spleens of 6-wk-old IL-22/2 mice (Fig. 3B). However, when the increase in spleen cellularity was taken into consideration, there was no significant difference in the number of B lymphocytes present in the spleens of these animals compared with that in IL-21/1 mice (Fig. 3B and Table II). However, by 8 wk of age the number of B cells in the spleens of IL-22/2 mice was reduced by approximately 2-fold compared with that in IL-21/1 littermates (Table II). Thus, germfree IL-22/2 mice develop an age-dependent disorder of the immune system similar to that described in SPF IL-22/2 mice. We next analyzed the bone marrow of germfree IL-22/2 mice to assess whether these disorders affect lymphoid and myeloid development.

Intrahepatic lymphocytes in germfree IL-22/2 mice Kramer and colleagues (10) have previously described the accumulation of CD42CD81 T cells that express intermediate levels of the TCRab in the livers of SPF IL-22/2 mice. Compared with the livers of 5-wk-old IL-21/1 littermates, approximately 10-fold more mononuclear cells were recovered from the livers of germfree IL-22/2 mice (Fig. 2). Flow cytometric analysis of these cellular infiltrates showed that the increase in cellularity in IL-22/2 mice was attributable to increased numbers of CD42CD81 and, to a lesser extent, CD41CD82 T cells (Fig. 2). Whereas the number of CD41CD82 T cells increased by approximately 13-fold on the average, there was a .30-fold increase in the number of CD42CD81 T cells. Although the IL-22/2 liver contained proportionately fewer B2201 (TCR2) B cells than those in littermate control mice (Fig. 2), when the difference in cellularity of IL-22/2 and IL-21/1 liver mononuclear cells was taken into consideration the numbers of B cells present in germfree IL-22/2 and IL-21/1 mice were similar. Also of note was the .30-fold increase in the number of B2201TCRablow/int cells in the livers of germfree IL-

Hemopoietic disorder Visual inspection of germfree IL-22/2 mice revealed brittle bones that upon a more comprehensive analysis identified striking changes in bone marrow composition that were evident as early as 4 wk of age (Fig. 4 and Table II). These changes were primarily characterized by a progressive loss of mature myeloid cells that corresponded to a 20-fold reduction in mature, Mac-1high and Gr1high, polymorphonuclear cells (Table II). The majority of residual cells expressed low levels of Mac-1/Gr-1 (Fig. 4), characteristic of immature myeloid cells (19, 20). B cell development was similarly affected in germfree IL-22/2 mice, as evidenced by the .30-fold reduction in immature, B220low, and the 10-fold reduction in more mature, B220high, B cells in the marrow (Fig. 4). Accompanying the breakdown in B cell development was the infiltration of large numbers of T lymphocytes that were primarily CD41 T cells, of which a large proportion was CD691CD251 (Table II). The T cell infiltration occurred at approximately 6 wk of age, and since the changes in myeloid cells were seen before this, it is likely that the

388

AUTOIMMUNITY IN GERMFREE IL-2-DEFICIENT MICE

FIGURE 1. Pathologic lesions in tissues of 8-wk-old germfree IL-22/2 mice. The lungs ( a ), liver ( b ), pancreas ( c ), kidney ( d ), small intestine ( e ), colon ( f ), and spleen ( g ) of IL-22/2 mice were fixed in formalin and embedded in paraffin, and 5-mm sections were counterstained with hematoxylin-eosin before photomicroscopy. Note the extensive lymphocyte hyperplasia in all tissues with the exception of the intestine. Magnification: a through e and g, 3200; f, 3300.

disruption in bone marrow hemopoiesis occurs before or coincident with the lymphoproliferative disorder. The progressive neutropenia seen in the blood of germfree IL-22/2 mice (Table I) is also consistent with bone marrow failure. The loss of mature neutrophils was accompanied by a progressive anemia and the presence of large numbers of atypical blast-like lymphocytes that were tartrate resistant, acid phosphatase positive, and myeloperoxidase negative.

Absence of colitis in germfree IL-22/2 mice The incidence of colitis in our SPF IL-2-deficient mouse colony was comparable to that described in other colonies (2). Of the animals that survived early disease (50%), characterized by splenomegaly, lymphadenopathy, and anemia, all subsequently developed colitis, with disease symptoms present between 6 and 10 wk of age. In contrast to the colitis seen in SPF IL-22/2 mice, none (n 5 25) of the germfree IL-22/2 mice developed any intestinal pathology. The small and large intestines of germfree IL-22/2 an-

imals displayed normal architecture, with no detectable lymphocyte infiltrates, epithelial cell damage, crypt hyperplasia, or signs of regeneration (Fig. 1, e and f ).

Changes in the distribution of iIEL populations To investigate whether the absence of colitis and gut flora was accompanied by changes in the intestinal T cell repertoire, we examined iIELs from both IL-22/2 and IL-21/1 mice. The iIEL repertoire of the small intestine in IL-22/2 germfree mice was distinct from that in littermate control mice. Although the total numbers of iIEL recovered from IL-22/2 and IL-21/1 mice were equivalent, those from IL-22/2 mice contained fewer CD8aa-expressing gd1 and ab1 T cells (Fig. 5B). Whereas gd1 cells represented the major iIEL population (63%) in wild-type germfree mice, they comprised only 34% of iIEL in their IL-22/2 counterparts (Fig. 5A). Similarly, the number of TCRab1CD8aa1 iIELs was reduced in germfree IL-22/2 mice, but due to an increase in CD8ab1 iIEL, the overall number of TCRab1 iIEL was higher in


389

Table II. Cellular composition of hematopoietic organs in germfree IL22/2 mice Bone Marrowa Spleena

Myeloid Cells Age (weeks)

IL-2 Genotype

n

Splenomegalyb

4

1/1 2/2 1/1 2/2 1/1 2/2

3 4 4 4 3 5

2 2 2 1/11 2 11

6 8

Colitisc

Total (3106)

% B2201

% CD31

Immatured

Maturee

Total (3106)

% B2201

% CD31

2 2 2 2 2 2

51 6 7 33 6 8 60 6 10 24 6 6 59 6 5 19 6 7

30 6 5 33 6 6 33 6 4 21 6 9 29 6 7 16 6 4

361 262 862 10 6 3 761 18 6 4f

31 6 6 56 6 5 66 6 4 84 6 10 66 6 10 83 6 12

67 6 8 40 6 7 28 6 8 13 6 5 31 6 9 18 6 6

25 6 10 28 6 7 35 6 8 58 6 15 47 6 17 73 6 21

43 6 4 40 6 10 49 6 5 35 6 9 47 6 6 16 6 4

22 6 5 25 6 6 34 6 3 23 6 6 37 6 6 21 6 5

Phenotype and composition of bone marrow and spleen cells determined by flow cytometry. Data represent the mean 6 SD. Size of IL22/2 spleen relative to that of IL21/1 mice; 2, no change; 1, 1 to 2 times larger; 11, .2 times larger. c Based upon histologic analysis. d Gr-1low and/or Mac-1low cells. e Gr-1high cells. f 60 –75% TCRab1CD41CD691 and/or CD251. a

b

FIGURE 2. Accumulation of CD42CD81 and B2201TCRlow/int intrahepatic lymphocytes in germfree IL-22/2 mice. Representative FACS profile of CD4, CD8, TCRab, and B220 expression by cells recovered from the livers of 5-wk-old germfree (GF) IL-22/2 and IL-21/1 mice. The total number of cells recovered from the liver of each animal is shown above the dot plots. The boxed region in the lower dot plots identifies the B2201TCRablow/int population of intrahepatic lymphocytes. The results shown are representative of those obtained from two pairs of mice.

germfree IL-22/2 (;1.5 3 106) than that in IL-21/1 (;1.0 3 106) mice (Fig. 5B). By contrast, the TCRab1CD4282 and TCRab1CD41 subsets of iIELs did not appear to be significantly

affected by the absence of IL-2. Thus, the absence of IL-2 results in a decrease in the numbers of CD8aa1, particularly gd1 T cells, in the iIEL compartment.

390


FIGURE 3. Phenotypic analysis of spleen cells from 8-wk-old germfree IL-22/2 mice. A, Representative FACS profile of splenic T cells and histogram profiles of CD69 and CD25 expression by CD41 T cells. The total number of cells recovered from the spleen of each animal is shown above the dot plots. B, Expression of B220 among spleen cells from wild-type (IL-21/1) and IL-2-deficient (IL22/2) germfree mice. The results shown are representative of those obtained from four pairs of mice.

Failure of gd iIELs to develop in bone marrow radiation chimeras To determine whether the decreased numbers of gd1 iIELs in germfree IL-22/2 mice was due to a defect in the ability of these

cells to differentiate, lethally irradiated CD45.11 (BL/6 congenic) mice were reconstituted with IL-22/2 (CD45.21) bone marrow. The majority of mice (four of five) reconstituted with IL-22/2 bone marrow survived through 90 days postreconstitution. At necropsy, tissues were grossly normal, and there were no signs of


391

FIGURE 4. Cellular composition of bone marrow from 8-wk-old germfree IL-22/2 mice. Representative FACS profiles of bone marrow mononuclear cells from IL-22/2 mice and wild-type (IL-21/1) littermates using lineage-specific Abs to identify myeloid (Gr-1), monocytic/macrophage (Mac-1), and B (B220) cells. The values shown for Gr-1 profiles represent the frequency of mature, Gr-1high, granulocytes. Those shown for Mac-1 and B220 represent the relative proportions of immature (Mac-1low; B220low) and mature (Mac-1high; B220high) cells present.

colitis in these animals maintained under SPF conditions (for the duration of the experiment). Analysis of the iIELs in mice reconstituted with syngeneic (host) bone marrow showed that these cells were able to fully reconstitute the iIEL compartment (Fig. 6A), such that ab1 and gd1 iIELs were present in the same proportions as those seen in unmanipulated wild-type mice (Fig. 5). On the other hand, while IL-22/2 bone marrow was able to reconstitute the majority (71%) of the iIEL compartment (Fig. 6B), they were unable to fully reconstitute TCRgd1 cells. The iIELs present consisted almost exclusively of ab1 (91%), with very few gd1 iIELs (10% or less). Additionally, the endogenously derived (host) iIELs

in IL-22/2 bone marrow-reconstituted mice were not comprised of ab1 and gd1 T cells in the same proportions as those in the mice reconstituted with syngeneic cells (Fig. 6B). Since these cells most likely arose from endogenous progenitors that survived radiation, one possible explanation for this disparity is that gd1 iIEL progenitors are more radiosensitive than ab1 iIEL progenitors. The presence of TCRgd1 iIELs in animals reconstituted with syngeneic bone marrow (Fig. 6A) argues against the possibility that the absence of these cells in mice reconstituted with IL-22/2 bone marrow is due to differences in the kinetics of gd vs ab iIEL generation and reconstitution. Instead, our results of iIEL analysis

392


FIGURE 5. Flow cytometric analyses of iIEL from 8-wk-old germfree IL-21/1 and IL-22/2 mice. The iIEL were stained with the following combinations of Abs; anti-Thy 1.2, anti-TCRab, and anti-TCRgd Abs for three-color analysis and anti-TCRgd or anti-TCRab combined with anti-CD4, anti-CD8a, and anti-CD8b Abs for four-color analysis. Viable cells were analyzed by flow cytometry as described in Materials and Methods. The numbers on each dot plot and histogram represent the percentage of total cells present in each quadrant and in the gated areas, respectively.

in germfree IL-22/2 mice and bone marrow chimeras are consistent with a requirement for IL-2 in the generation of gd1 iIELs.

Discussion Splenomegaly, lymphadenopathy, anemia, and inflammatory bowel disease are all characteristic disorders displayed by IL-22/2 mice. However, the precise nature of the role of IL-2 and the nature of the factors that initiate these disorders remain unclear. It has been proposed that the disease may be a result of aberrant responses to environmental Ags. Conversely, the disease in IL-22/2 mice may result from a breakdown in central and/or peripheral tolerance. To distinguish between these possibilities, we have rederived germfree IL-22/2 mice and evaluated them over a period of 10 mo for signs of disease. Our observations that germfree IL-22/2 mice develop hemopoietic and lymphoid disorders similar to those seen in SPF IL-22/2 mice, with the exception of colitis, demonstrate that environmental Ags are not responsible for the extraintestinal lymphoproliferative disorder and autoimmunity. The ability to distinguish between different mechanisms of disease in IL-22/2 mice is analogous to the distinction between colitis and extraintestinal disease in HLA-B27 transgenic rats (21, 22). In contrast to the inflammatory disorders that affect the in-

testine, skin, and joints of SPF transgenic rats, the intestine and joints remain histologically normal in germfree animals (22). The possibility that food Ags could initiate disease in IL-22/2 mice seems unlikely, since the intestines of germfree animals remain histologically normal, and abnormal mucosal immune responses to food Ags would be expected to result in lymphocyte infiltration and tissue damage, as seen, for example, in celiac’s disease (23, 24). Instead, our findings are consistent with defects in intrathymic T cell development and a breakdown in central and/or peripheral tolerance as the causes of the hemopoietic disorders in IL-22/2 mice. The accumulation of B2201TCRlow/int T cells that normally undergo apoptosis (18) in the livers of germfree IL-22/2 mice and the defective elimination of superantigen-stimulated peripheral T cells in IL-22/2 (7) and IL-2Ra2/2 mice (8) strongly suggest that a breakdown in peripheral tolerance is at least partially responsible for the disorders seen in IL-22/2 mice. Additionally, our findings that IL-2 is produced by dying thymocytes and that activation-induced cell death in the thymi of IL-22/2 mice is defective (H. Bassiri, P. J. Egan, E. Samoilova, Y. Chen, and S. R. Carding, unpublished observations) suggest that a breakdown in thymic tolerance may also be ultimately responsible for the splenomegaly, lymphadenopathy, and anemia seen in IL-22/2 mice.


393

FIGURE 6. Flow cytometric analyses of iIELs from IL-22/2 bone marrow chimeras. The iIELs recovered from chimeras 90 days after reconstitution with either host (IL-21/1) bone marrow ( A ) or congenic IL-22/2 bone marrow ( B ) were stained with anti-TCRab and anti-TCRgd Abs in conjunction with anti-CD45.1 or CD45.2 Abs and analysed by flow cytometry as described in Materials and Methods. The proportions (percentage) of donor-derived (CD45.21) or host-derived (CD45.11/CD45.22) cells present are shown in the upper histograms in A and B. The proportions (percentage) of donor-derived and host-derived TCRab1 and TCRgd1 iIELs shown in the lower histograms in A and B were determined by electronically gating on CD45.21 or CD45.22 cells. The results shown are representative of those obtained from four animals reconstituted with host bone marrow and five animals reconstituted with IL-22/2 bone marrow.

Comparison of the iIEL repertoire in wild-type mice maintained in an SPF or germfree environment clearly demonstrates the influence of the gut flora on shaping the iIEL repertoire. The presence of TCRab1CD8ab1 iIELs appears to be largely dependent upon bacterial colonization, whereas the CD8aa1 subset, which is thought to arise in the gut mucosa (25), is generated independently of microbial Ags. Among these cells, the generation or maintenance of TCRgd1 iIELs appears to be dependent upon IL-2. The inability of IL-22/2 bone marrow cells to generate gd1 iIELs in lethally irradiated hosts is consistent with an intrinsic defect within hemopoietic or lymphoid progenitor cells. The absence of TCRgd1 iIELs in SPF IL-22/2 bone marrow chimeras demonstrates that the reduced number of these cells in germfree IL-22/2 mice is not due to autoimmunity, the absence of gut flora, or differences in food Ags or intestinal microenvironments that may exist between SPF and germfree mice. The increased numbers of CDab1 iIELs in germfree IL-22/2 mice may simply reflect their infiltration into the epithelia and/or their expansion to fill a niche created by the failed generation of TCRgd1 iIELs. Although we have not been able to identify the defective progenitor cell(s) in IL-22/2 bone marrow, the presence of donorderived TCRab1 iIELs (Fig. 6) and ab T cells and B cells in the

blood6 of IL-22/2 bone marrow chimeras suggests that the defective cell resides in a gd progenitor population rather than in a common lymphoid progenitor or hemopoietic stem cell population. Furthermore, the normal number of gd T cells in extraintestinal tissues of SPF (26) and germfree (T. Reya, H. Bassiri, and S. R. Carding, unpublished observations) IL-22/2 mice suggests that the defect is restricted to a gd iIEL progenitor population. The reduced number of TCRgd1 iIELs in IL-2Rb2/2 mice (14) is consistent with our findings in IL-22/2 mice and suggests that IL-2/IL-2R signaling may directly influence gd iIEL progenitor cell development. Expression of IL-2Rs by a subset of T lymphoid progenitors within cryptopatches (27) of the gut is consistent with this interpretation. The source and utilization of IL-2 within the intestinal mucosa could be autocrine or paracrine in nature (28, 29), with IL-2 being produced in response to endogenous and/or food Ags. We cannot, however, exclude the possibility that IL-2 may influence TCRgd1 6 T. Reya, N. V. Contractor, M. S. Couzens, M. A. Wasik, S. G. Emerson, and S. R. Carding. Abnormal myelocytic cell development in interleukin-2 (IL-2)deficient mice: evidence for the involvement of IL-2 in myelopoiesis. Submitted for publication.

394 iIELs in other ways. For example, IL-2 may serve as a survival factor or a competence factor, enabling gd iIEL progenitors to acquire responsiveness to other growth and differentiation factors. These possibilities are not mutually exclusive, since IL-2 may have multiple effects during gd iIEL development, with its effect being determined by the developmental stage of the IL-2R-bearing cell. Finally, since the block in gd iIEL generation is incomplete in germfree IL-22/2 mice or IL-22/2 bone marrow chimeras, it is possible that there are compensatory roles for other factors (30, 31) and/or IL-2-independent pathways of gd iIEL development. Regardless of the role of these other factors, our findings clearly demonstrate the heterogeneity in the requirement for IL-2 for the development of iIELs and the existence of IL-2-dependent (TCRgd1) and independent pathways of iIEL development. The absence of colitis in germfree IL-22/2 mice is in agreement with the previous observations of Sadlack and co-workers (1) and strongly suggests that enterocolitis is a direct result of an abnormal immune response in the mucosa to the intestinal bacterial flora (32–34). It is possible that TCRgd iIEL populations, which are reduced in the absence of IL-2, possess immunoregulatory functions that contribute to the regulation of mucosal immune responses and to the maintenance of tolerance to endogenous Ags (35–37). In summary, the results of our analysis of germfree IL-22/2 mice show that the anemia and extraintestinal lymphocytic infiltration seen in SPF IL-22/2 mice are most likely the result of uncontrolled responses to endogenous Ags and not to environmental ones. In contrast, the colitis that develops in SPF IL-22/2 mice occurs in response to intestinal bacterial flora. The germfree IL22/2 mice represent, therefore, a system in which the role of IL-2 in development and regulation of the hemopoietic and immune systems can be assessed in detail in dissociation from the complications of colitis.

References 1. Sadlack, B., H. Merz, H. Schorle, A. Schimpl, A. C. Feller, and I. Horak. 1993. Ulcerative colitis-like disease in mice with a disrupted interleukin-2 gene. Cell 75:253. 2. Sadlack, B., J. Lo¨hler, H. Schorle, G. Klebb, H. Haber, E. Sickel, R. J. Noelle, and I. Horak. 1995. Generalized autoimmune disease in interleukin-2-deficient mice is triggered by an uncontrolled activation and proliferation of CD41 T cells. Eur. J. Immunol. 25:3053. 3. Horak, I., J. Lo¨hler, A. Ma, and K. A. Smith. 1995. Interleukin-2 deficient mice: a new model to study autoimmunity and self-tolerance. Immunol. Rev. 148:35. 4. Suzuki, H., T. M. Kundig, C. Furlonger, A. Wakeman, E. Timms, T. Matsuyama, R. Schmits, J. J. L. Simard, P. Ohashi, H. Griesser, T. Taniguchi, C. J. Paige, and T. W. Mak. 1995. Deregulated T cell activation and autoimmunity in mice lacking interleukin-2 receptor b. Science 268:1472. 5. Oldstone, M. B. A. 1987. Molecular mimicry and autoimmune disease. Cell 50:819. 6. Lenardo, M. J. 1991. Interleukin-2 programs mouse ab T lymphocytes for apoptosis. Nature 353:858. 7. Kneitz, B., T. Herrmann, S. Yonehara, and A. Schimpl. 1995. Normal clonal expansion but impaired Fas-mediated cell death and anergy induction in interleukin-2-deficient mice. Eur. J. Immunol. 25:2572. 8. Willerford, D. M., J. Chen, J. A. Ferry, L. Davidson, A. Ma, and F. W. Alt. 1995. Interleukin-2 receptor a chain regulates the size and content of the peripheral lymphoid compartment. Immunity 3:521. 9. Schorle, H., T. Holtschke, T. Hunig, A. Schimpl, and I. Horak. 1991. Development and function of T cells in mice rendered interleukin-2 deficient by gene targeting. Nature 352:621. 10. Kra¨mer, S., A. Schimpl, and T. Hunig. 1995. Immunopathology of interleukin (IL) 2-deficient mice: thymus dependence and suppression by thymus-dependent cells with an intact IL-2 gene. J. Exp. Med. 182:1769. 11. Hanke, T., R. Mitnacht, R. Boyd, and T. Hu¨nig. 1994. Induction of interleukin-2 receptor b-chain expression by self-recognition in the thymus. J. Exp. Med. 180: 1629. 12. Lu´dvı´ksson, B. R., B. Gray, W. Strober, and R. O. Ehrhardt. 1997. Dysregulated intrathymic development in the IL-2-deficient mouse leads to colitis-inducing thymocytes. J. Immunol. 158:104.

AUTOIMMUNITY IN GERMFREE IL-2-DEFICIENT MICE 13. Kra¨mer, S., C. Mamalaki, I. Horak, A. Schimpl, D. Kioussis, and T. Hu¨nig. 1994. Thymic selection and peptide-induced activation of T cell receptor-transgenic CD8 T cells in interleukin-2-deficient mice. Eur. J. Immunol. 24:2317. 14. Suzuki, H., G. Duncan, S., H. Takimoto, and T. Mak, W. 1997. Abnormal development of intestinal intraepithelial lymphocytes and peripheral natural killer cells in mice lacking the IL-2-receptor b chain. J. Exp. Med. 185:499. 15. Giri, J. G., M. Ahdieh, J. Eisenman, K. Shanebeck, K. Grabstein, S. Kumaki, A. Namen, L. S. Park, D. Cosman, and D. Anderson. 1994. Utilization of the b and g chains of the IL-2 receptor by the novel cytokine IL-15. EMBO J. 13:2822. 16. Grabstein, K. H., J. Eisenman, K. Shanebeck, C. Rauch, S. Srinivasan, V. Fung, C. Beers, J. Richardson, M. A. Schoenborn, M. Ahdieh, L. Johnson, M. R. Alderson, J. D. Watson, D. M. Anderson, and J. G. Giri. 1994. Cloning of a T cell growth factor that interacts with the b chain of the interleukin-2 receptor. Science 264:965. 17. Reya, T., J. A. Yang-Snyder, E. V. Rothenberg, and S. R. Carding. 1996. Regulated expression and function of CD122 (interleukin-2/interleukin-15R-b) during lymphoid development. Blood 87:190. 18. Huang, L., G. Soldevila, M. Leeker, R. A. Flavell, and I. N. Crispe. 1994. The liver eliminates T cells undergoing antigen-triggered apoptosis in vivo. Immunity 1:741. 19. Hestdal, K., F. W. Ruscetti, J. N. Ihle, S. E. Jacobsen, C. M. Dubois, W. C. Kopp, D. L. Longo, and J. R. Keller. 1991. Characterization and regulation of RB6-8C5 antigen expression on murine bone marrow cells. J. Immunol. 147:22. 20. Ralph, P., M.-K. Ho, P. B. Litcofsky, and T. A. Springer. 1983. Expression and induction in vitro of macrophage differentiation antigens on murine cell lines. J. Immunol. 130:108. 21. Hammer, R. E., S. D. Maika, J. A. Richardson, J. Tang, and J. D. Taurog. 1990. Spontaneous inflammatory disease in transgenic rats expressing HLA-B27 and human b2-microglobulin:An animal model of HLA-B27-associated human disorders. Cell 63:1099. 22. Taurog, J. D., J. A. Richardson, J. T. Croft, W. A. Simmons, M. Zhou, J. L. Ferna´ndez-Sueiro, E. Balish, and R. E. Hammer. 1994. The germfree state prevents development of gut and joint inflammatory disease in HLA-B27 transgenic rats. J. Exp. Med. 180:2359. 23. Ferguson, A. 1977. Intraepithelial lymphocytes of the small intestine. Gut 18: 921. 24. Niazi, N. M., R. Leigh, P. Crowe, and M. N. Marsh. 1984. Morphometric analysis of small intestinal mucosa. I. Methodology, epithelial volume, compartments and enumeration of intra-epithelial space lymphocytes. Virchows Arch. A Pathol. Anat. Histopathol. 404:49. 25. Rocha, B., D. Guy-Grand, and P. Vassalli. 1995. Extrathymic T cell differentiation. Curr. Opin. Immunol. 7:235. 26. Reya, T., H. Bassiri, R. Biancaniello, and S. R. Carding. 1997. Thymic stromal cell abnormalities and dysregulated T cell development in IL-2-deficient mice. Dev. Immunol. In press. 27. Kanamori, Y., K. Ishimaru, M. Nanno, K. Maki, K. Ikuta, H. Nariuchi, and H. Ishikawa. 1996. Identification of novel lymphoid tissues in murine intestinal mucosa where clusters of c-kit1 IL-7R1 Thy11 lympho-hematopoietic progenitors develop. J. Exp. Med. 184:1449. 28. Barrett, T. R., T. F. Gajewski, D. Danielpour, E. B. Chang, K. W. Beagley, and J. A. Bluestone. 1992. Differential function of intestinal intraepithelial lymphocyte subsets. J. Immunol. 149:1124. 29. Camerini, V., C. Panwala, and M. Kronenberg. 1993. Regional specialization of the mucosal immune system. J. Immunol. 151:1765. 30. Maki, K. S., S. Sunaga, Y. Komagata, Y. Kodaira, A. Mabuchi, H. Karasuyama, K. Yokomuro, J. Miyazaki, and K. Ikuta. 1996. Interleukin-7 receptor-deficient mice lack gd T cells. Proc. Natl. Acad. Sci. USA 93:7172. 31. Moore, T. A., U. von Freeden-Jeffry, R. Murray, and A. Zlotnik. 1996. Inhibition of gd T cell development and early thymocyte maturation in IL-72/2 mice. J. Immunol. 157:2366. 32. Sartor, R. B. 1992. Role of the intestinal microflora in pathogenesis and complications. In Inflammatory Bowel Diseases: Pathophysiology as Basis for Treatment. J. Scholmerich, W. Kruis, H. Goebell, W. Hohenberger, and V. Gross, eds. Kluwer Academic, Lancaster, p. 175. 33. Duchmann, R., I. Kaiser, E. Hermann, W. Mayet, K. Ewe, and K.-H. Meyer Zum Bu¨schenfelde. 1995. Tolerance exists towards resident intestinal flora but is broken in active inflammatory bowel disease (IBD). Clin. Exp. Med. 102:448. 34. Dianda, L., A. M. Hanby, N. A. Wright, A. Sebesteny, A. C. Hayday, and M. J. Owen. 1997. T cell receptor-ab-deficient mice fail to develop colitis in the absence of a microbial environment. Am. J. Pathol. 150:91. 35. Fujihashi, K., T. Taguchi, W. K. Aicher, J. R. McGhee, J. A. Bluestone, J. H. Eldridge, and H. Kiyono. 1992. Immunoregulatory functions for murine intraepithelial lymphocytes: gd T cell receptor positive (TCR1) T cells abrogate oral tolerance, while ab TCR1 T cells provide B cell help. J. Exp. Med. 175:695. 36. Komano, H., Y. Fujiura, M. Kawaguchi, S. Matsumoto, Y. Hashimoto, S. Obana, P. Mombaerts, S. Tonegawa, H. Yamamoto, S. Itohara, M. Nanno, and H. Ishikawa. 1995. Homeostatic regulation of intestinal epithelia by intraepithelial gd T cells. Proc. Natl. Acad. Sci. USA 92:6147. 37. Mengel, J., F. Cardillo, L. S. Aroeira, O. Williams, M. Russo, and N. M. Vaz. 1995. Anti-gd T cell antibody blocks the induction of oral tolerance to ovalbumin in mice. Immunol. Lett. 48:97.