Ectopic Class II Major Histocompatibility Antigens in Hirschsprung’s Disease and Neuronal Intestinal Dysplasia By Seiichi Hirobe, Daniel P. Doody, Daniel P. Ryan, Samuel H. Kim, and Patricia K. Donahoe Boston,
Massachusetts
@Although the etiology of Hirschsprung’s disease and neuronal intestinal dysplasia remains obscure, both have histological abnormalities involving ganglion cells and neuronal elements. Searching for a common pathway that may inhibit normal maturation of neurogenic precursors, we examined the possible role of an immune mechanism in the maldevelopment of the enteric neural network. Six patients with Hirschsprung’s disease were studied by comparing biopsy specimens from diseased colon with ones taken from proximal ganglionic colon in the same patients. These were similarly compared with colonic biopsy specimens from patients studied with chronic constipation or bowel removed at the time of operation for other disorders. Biopsies were taken from four other patients with neuronal intestinal dysplasia. Each was examined by hematoxylin 81 eosin staining, acetylcholinesterase histochemistry, and immunohistochemistry of major histocompatibility complex (MHC) class I and class II antigens. All rectal samples from Hirschsprung’s disease patients exhibited elevated acetylcholinesterase histochemistry and absent ganglia to confirm the diagnosis. These findings were correlated with marked elevation of class II MHC in the aganglionic area, whereas the proximal normal ganglionic segments showed no elevation. Rectal biopsy specimens from patients with chronic constipation exhibited no such elevation. A similar elevation of class II MHC was detected in the mucosa and submucosa of all four patients with the rare neuronal intestinal dysplasia disorder whose diagnosis was confirmed by giant ganglia in Auerbath’s plexuses, aberrant Meissner’s ganglia in the lamina propria mucosa, and giant neurofibrils in the mucosa and submucosa. The correlation of elevated class II MHC in these two neuronal dysfunction disorders may indicate an underlying autoimmune mechanism as is seen in thyroiditis and insulin dependent diabetes mellitus. These observations confirm those found in a small number of Hirschsprung’s disease patients reported earlier, and extend the observation to other neuronal dysfunction disorders. These findings provoke a reexamination of the etiology of Hirschsprung’s disease and may predict alternate therapies. Copyright o 1992 by W.B. Saunders Company INDEX WORDS:
Hirschsprung’s
disease; neuronal intestinal
dysplasia; class II major histocompatibility
complex.
H
IRSCHSPRUNG’S disease (HD) is characterized by the absence of enteric ganglia and the presence of abnormal neural elements that, more commonly, involves the distal hindgut.’ The etiology of HD remains unclear, but has been attributed to a failure of migration of neural crest cells into the distal colon,’ to neural precursor abnormalities, or to a nonpermissive intestinal microenviromnent.3 The early timing of normal neural crest migration in the fetus to the distal colon4 and the usual lack of association of HD with major hindgut anomalies make failure of Journal of Pediatric Surgery,
Voi 27.
No 3
(March), 1992: pp 357-363
migration a less satisfying explanation. Therefore, we earlier proposed that factors leading to failure of differentiation after migration could be responsible for the disease complex.5 The cellular specificity of neural destruction led us further to hypothesize that an immune mechanism could underlie HD.h Neuronal intestinal dysplasia (NID) shows a distinct morphologic picture characterized by hyperplasia of enteric ganglia, presence of isolated ganglia in the lamina propria mucosa, and increased activity of acetylcholinergic nerve fibers.’ Until 1983, the association of NID and HD was infrequently reported.‘,“ However, more current series suggest that NID may occur in 26% of children with HD and that up to 75% of cases of NID have some element of aganglionosis.‘.” The coincidence of these two diseases suggests that a common factor may disturb the normal maturation of the enteric ganglia. The cell-specific neuronal abnormalities also apparent in NID anomalies led us to examine the possible role of immune mechanisms in this maldevelopment of the enteric nervous system. Therefore, we studied the expression of the major histocompatibility complex (MHC) class I and II antigens in a series of intestinal biopsy specimens using immunohistochemical techniques and correlated elevation of class II with the diseased state. MATERIALS AND METHODS Six children with HD (mean age, 2.7 years: range, 5 days to 7 years), five with rectosigmoid aganglionosis and one with total colon aganghonosis, had six biopsies from the aganglionic area, three from the transitional area, and three from normoganglionic area. Pathological investigation showed a classical picture of HD, showing the absence of enteric ganglion cells and elevation of mucosal and submucosal acetylcholinesterase (AChE) activity. Four children had histologically diagnosed NID (mean age, 3.3 years: range, 6 months to 6 years). Of these, three had accompanying HD. All colonic specimens in the children with NID showed moderate elevation of AChE-positive nerve fibers. isolated ganglia in the lamina propria. and hyperplasia of submucosal and myen-
From the Division of Pediatric Surgery and the Pediatric Surgical Research Laboratories, Massachusetts Generai Hospital and Harvard Medical School, Boston, MA. Presented at the 22nd Annual Meeting of the American Pediattic SurgicalAssociation, Lake Buena vista, Florida, Mav 15-18. 1991. Address reprint requests to Patricia K. Donahoe, MD, Department of Pediattic Surgery, Massachusetts General Hospital, Warren Bldg, Room 1133. Boston, MA 02114. Copyright o 1992 by W.B. Saunders Company 0022-3468/92/2703-0016$03.OOiO 357
358
HIROBE ET AL
teric plexuses. Dysplastic giant ganglia in the submucosal plexuses were often associated with thickened nerve fibers.’ For comparison we studied biopsy specimens from 10 children (mean age, 6.7 years; range, 3 months to 15 years), including 3 rectal specimens from patients with severe chronic constipation, 3 colonic specimens from patients with anorectal malformations, 1 colonic specimen from one patient each with ulcerative colitis, Crohn’s disease, and volvulus, and 1 jejunal specimen from a patient with a choledocal cyst. Specimens from each of these patients failed to exhibit abnormalities in the enteric nervous system by AChE or hematoxylin and eosin stains.
Antibodies and Immunohistochemistry Mouse monoclonal antibody specific for a skeletal determinant of human HLA-A,B,C (W6/32. Seralab, Crowley Down, England)” and rat monoclonal antibody specific for a monomorphic determinant of human HLA-DR (YE 2/36 HLK, Seralab)” were used to detect MHC class I and II antigens. Specimen were fixed in Zamboni’s solution for 6 hours at 4”C, washed overnight in 0.1
mol/L phosphate buffered saline (pH 7.4) containing 7% sucrose, embedded in OCI compound (Miles Scientific, Naperville, IL), then frozen and stored in liquid nitrogen until sectioned. The frozen tissues were cut with a cryostat into serial S-km sections and mounted on gelatin-coated glass slides at -2o”C, and air-dried for 30 minutes at room temperature. They were first treated with 1% hydrogen peroxide:50% methanol for 30 minutes at 4°C and then with 5% bovine serum albumin (BSA) in 0.05 mol/L Tris-HCl buffer (pH 7.4) in saline (TBS) for 30 minutes. Monoclonal antibodies at optimal dilution (HLA-DR and HLA-A,B,C: 1 to 2,000) with TBS containing 5% BSA were added for 45 minutes at room temperature in a moist chamber. In an earlier study,“6 the first monoclonal antibody was biotinylated. After comparative studies, we elected to biotinylate the second antibody for these studies. After rinsing with TBS, the second antibodies, either appropriate biotinylated horse antimouse or rabbit antirat immuno-
giibuiin G (heavy and light chain specific) (Vector Laboratories, Inc, Burlingame, CA) diluted 1:200 and containing 5% human serum, were added for 30 minutes. Thereafter, sections were treated for 30 minutes with ABC reagent (avidin DH and biotinylated peroxidase H, Elite kit; Vector Laboratories, Inc). A mixture of 0.05% 3,3’-diaminobenzidine tetra-hydrochloride (grade II; Sigma Chemical Co, St Louis, MO) in 0.05 mol/L Tris-HCl buffer (pH 7.4), and 0.01% hydrogen peroxide and 10 mmol/L sodium azide was then added to develop the peroxidase reaction. Sections were lightly counterstained with 1% methyl green and examined with a light microscope. As a negative control, nonimmune appropriate rat or mouse serum were used in place of primary antibodies. Lymphoid tissue in Peyer’s patches was well visualized after biotinylating the second antibody and, therefore, provided positive control cells within selected specimens. Selected sections were stained with hematoxylin and eosin and for AChE histochemistry using Karnovsky and Roots’ direct coloring method.” RESULTS
MHC Class II Antigen Expression In the control bowel, MHC class II antigen was seen in the lamina propria mucosa. These class II-positive cells had either a small round cell or dendritic morphology and tended to be polarized toward the lumen (Fig 1A). In the submucosal and muscle layers, small numbers of cells with dendritic
Fig 1. Elevated class II MHC antigen expression in the mucosa of eganglionic colon. (A) Normoganglionic colon: normal levels of class II-positive cells (open arrow) with small round cell or dendritic morphology are seen in the lamina propria. Staining in Peyer’s patches (PP) provides a positive control tissue for comparison (closed arrow). (B) Aganglionic colon: marked increase of class II antigen (open arrow) staining is seen in the lamina propria where positive cells are diffusely scattered. (Original magnification x 100.)
morphology or ceils lining vessels stained for class II. However, no staining could be found within the nerve plexuses (Fig 2A). In the aganglionic colon of patients with HD, there cells was a marked increase in class II staining distributed throughout the intestinal wall in all cases (Fig 3B). The pattern of distribution of these cells correlated closely with the AChE-positive hypertrophied nerve fibers (Fig 3A). There was a decided increased distribution of class II antigens in the
359
MI-K ANTIGENS IN NEUROENTERIC DISORDERS
colon. The abnormal transitional ganglia were decorated with ectopic class II (Figs 2B and 2C). The proximal normoganghonic colon of HD showed a pattern of class II expression that was similar to that found in control specimens. Specimens taken from patients with NID had a marked increase in the distribution of class II antigen in the cellular elements of the lamina propria (Fig 5B), where moderate elevation of AChE-positive nerve fibers and ectopic isolated ganglia were also found (Fig 5A). Dysplastic giant ganglia in the submucosal plexus were often associated with thickened AChE-positive nerve fibers (Fig 5C), which stain for class II antigen (Fig 5D). In Auerbach’s plexus numerous giant ganglia (Fig 5E) with associated class II-positive cells were also seen (Fig 5F). MHC Class I Expression
Specimens taken from patients with HD, NID, and from normal controls stained identically with antibodies to MHC class I antigens in all epithelial and interstitial cells in the lamina propria, and in vesselassociated cells and cells with dendritic morphology in the submucosa and muscle layers. Class I antigen stained the myenteric plexuses of control specimens, the transitional plexuses of HD, the hyperplastic plexuses of NID, and the thickened nerve fibers in the affected colon of HD and NID. DISCUSSION
MHC class II antigen is a cell surface glycoprotein involved in the immune recognition of foreign tissue and in the regulation of the immune response. In order to induce an immune reaction, antigen must first be processed intracellularly and then be presented on the cell surface as peptides bound to the MHC molecules expressed on the antigen presenting Fig 2. Class II expression In transkion zone myenteric plexus55 of a HD patient. (A) Normoganglionic colon for comparison shows a small number of class II-positive cells (arrows) outside the myenteric plexus. Serial sections of the transition colon stained for (B) class II and (C) AChE show increased expression in a hypertrophic nervr (open arrow), and transitional ganglia (arrows) are surrounded by cells expressing class II antigen. (Original magnification x200.)
lamina propria where positive cells tended to scatter diffusely (Fig 1B). Specimens from older patients tended to express more cellular class II antigens in the lamina propria. Class II staining within the nerve fibers was seen in cells with two or three processes (Figs 4B and 4C) and occasionally in larger fibrous components (Fig 4A). These findings were found as early as the 5-day-old neonate. In the transition zones of three patients with classical HD, class II expression was elevated in the AChE-positive hypertrophied nerve fibers as had been observed in the aganglionic
Fig 3. Conspicuous increase of (A) AChE and (8) class II expmssfon in the submucose of agangffonic rectum in the case of a &day-old neonate. Class II expression correlates wfth the AChE-posftive hypertrophied nerve bundles. MM, muscularis mucosa. (Original magnification x200.)
360
HROBE ET Al.
increased tissue specific expressions of class II antigens. Moreover, the induced expression of MHC antigens on islets of Langerhans in transgenic models has resulted in failure of islet maturation and secondary diabetes.19 The present study explores the possibility that the enteric nervous system could be the cell specific target of disturbances of the immune response, and so examines the expression of major histocompatibility antigens in the dysfunctional tissue of HD and NID. The staining patterns of MHC class I and II antigens in specimens without neuroenteric disorders and in positive control Iymphoid tissue (Fig 1A) were similar to those previously reported.W,21In the case of HD and NID, class I expression was precisely the same as that observed in normal tissue, but there was marked elevation of class II expression in the aganglionic areas, especially within the hypertrophied nerve
Class
II
Fig 4. Class II expression in the neural elements of an aganglionic areas. (A) Increased class II staining is seen in gbrillar components of hypettrophied nerves (original magnification x400). (6) Cross-section through hypertrophied nerve in the intermyenteric space demonstrates class II-positive interdigitating cells (arrow) (original magnification x200). CM, circular muscle; LM, longitudinal muscle. (C) Class II-positive cells in B at higher magnification (original magnification x 1,000) have short intercalating processes (arrow).
cells. The antigen-MHC complex is then recognized by the T cell receptor on the T lymphocyte membrane. Activated T cells then mediate an array of immunologic effector mechanisms that can lead to tissue damage.14 MHC molecules, especially class II, are known to play a major role in pathological states. A considerable body of evidence has accumulated that indicates that a quantitative variation in class II expression can be associated with profound alterations in immune responsiveness and immunologically mediated disease.15 For example, diseases such as thyroiditis,16 multiple sclerosis,17 and insulin-dependent diabetes mellitus” have been associated with aberrant and
Fig 5. Expression of AChE (left column; A, C. E) and class II (right column; B, D, F) in NID. (A) In the mucosa, there are increased AChE-positive nerve fibers and the presence of an ectopic ganglion cell (arrow), whereas(B) a disseminated and increased distribution of class II bearing cells is found in the lamina propria. (C) In serial sections of the submucosa, thickened nerve fiber with a ganglion (arrow) are seen in Meissner’s submucosal plexus by AChE stain. (D) Class II expression is seen in the nerve fibers but not in the ganglia (arrow). In the intermyenteric plexus of Auerbach, (E) AChE stain shows numerous hypertrophic ganglia (arrow) and (F) an increased population of associated Class II positive cells (arrow). (Original magnifications A,B,E,F: x400; C,D: x200.)
361
MHC ANTIGENS IN NEUROENTERIC DISORDERS
fibers (Figs 3 and 4). Previous studies indicate class II antigens are not expressed in the central nervous systemI nor in the peripheral nervous systems.“.23 This deficiency of class II antigens in normal nerve tissue is thought to be responsible for the apparent lack of neural immunoreactivity. Therefore, the ectopic presence of class II expression in and about the neural elements of these disorders can be considered aberrant and abnormal. The aganglionic colon in ‘HD may represent the final stage of a major immunological insult, so we intensely examined the transition zone, searching for clues in an area of continued pathophysiologic activity. Where previous electronmicroscopic studies had shown degenerated ganglion cells,“’ these hypoganglionic plexuses were decorated with class II antigens (Fig 2B). Continuing immunological events, visualized in the transition zone as in these cases, may be responsible for the 5% to 15% of patients whose symptoms persist after prior operations for HD.‘5.‘h We also suspected that patients with NID would provide morphologic evidence of an immunologic process because of the specificity of the target tissue and the concurrence of the two diseases. All specimens examined in the children with NID showed elevated expression of class II antigens (Table 1). Therefore, it is feasible’that an immune mechanism may play a role in its etiology, and that a common mechanism may account for the etiology of both HD and NID. The intestinal mucosa is a major site of contact with the external environment and contains numerous immune effector cells that form gut-associated lymphoid tissue, a major component of the body’s immune defense.” Recently, it was recognized that there were consistent interaction between enteric nervous systems and mucosal immune systems, each of which was influenced by the other? In the lamina propria of HD and NID, AChE-positive ‘nerve fibers are accompanied by class II-positive cells whose increased distribution may represent macrophages, dendritic cells, and, possibly, B cells. It is possible that the large numbers of cholinergic nerve fibers or their released peptides may initiate an abnormal reaction Table 1. MHC Class II Expression in Bowel Specimens No Studied
Class II Elevation
Hirschsprung’s disease Aganglionic
6
6
Hypoganglionic
3
3
Normoganglionic
3
0
4
4
Control
10
0
Total samples
26
Neuronal intestinal dysplasia
of the mucosal immune system. Conversely, the antigens presented at the mucosal surface may initiate an immune response that may incite hypertrophy of the nerves or degeneration of the ganglia. It is apparent that further studies will be needed to clarify the immunopathological mechanisms that underlie HD. The correlation of class II elevation with HD is now evident; the role of this abnormal finding with the pathogenesis of HD remains to be elucidated. However, the lessons learned from other disease states’6-18lead us to conclude that bowel with elevated class II expression may be highly susceptible to abnormal responses of immune origin. The findings of class II expression in the nerve fibers of a 5-day-old neonate indicates that this abnormal expression can occur very early and, one would presume, in utero. These findings may reflect an ongoing immunologic response. An alternative consideration is that MHC expression during development may create a nonpermissive microenvironment for neurogenic differentiation. In addition, the abnormal and increased expression of MHC antigens may reflect an epiphenomenon, related to changes in the extracellular matrix or to the local, increased concentration of cholinergic neurotransmitters. Ontogeny studies of class II antigen in both ganglionic and aganglionic colon in animal models with congenital megacolon” should be helpful in facilitating a study of the immune mechanisms of aganglionosis. In vitro studies may allow a more precise definition of the mechanisms that are responsible for the neuroenteric dysplasia that occurs in HD and NID. Further, experiments can be designed to examine the role of cholinergic substrates and products and luminal antigens in stimulating MHC class II expression. It may be possible to define an inciting antigen by eluting peptides bound to an affected individual’s class II molecules. In this way, immunomodulatory therapy can be individually designed in the future. The offending antigens causing experimental allergic encephalomyelitis, myelin basic protein3’ and experimental uveitis, retinal S-antigen3’ have been identified. It should be possible to define the offending antigen from the large quantity of bowel resected in cases of NID and HD, which provide a unique opportunity to biochemically identify an etiologic factor that is specific both to the disease process and to each affected individual in order to tailor subsequent care. It is important to restudy patients with HD who do not do well and continue to have symptoms after a technically satisfactory pull-through procedure. We recommend studying remaining fixed tissue for signs of NID. If sections have been frozen for AChE
362
HIROBE ET AL
histochemistry, we recommend restudy of these sections for the characteristic morphologic signs of ectopic lamina propria mucosa ganglion cells, increased population of AChE fibers, and enlarged myenteric ganglia. In addition, we now recommend restaining for class II MHC immunohistochemistry.
We suspect that gradations of abnormalities will occur between HD and NID that will link these as functional and anatomic autoimmune abnormalities with a common etiology. An understanding of the common underlying defect will help direct novel theories for the future.
REFERENCES 1. Swenson 0, Bill AH: Resection of the rectum and rectosigmoid with preservation of the sphincter for benign spastic lesions producing megacolon. Surgery 24:212-220,1948 2. Okamoto E, Ueda T: Embryogenesis of intramural ganglia of the gut and its relation to Hirschsprung’s disease. J Pediatr Surg 2:437-443,1967 3. Molenaar JC, Tibboel D, van der Kamp AWM, et al: Diagnosis of innervation-related motility disorders of the gut and basic aspects of enteric nervous system development. Prog Pediatr Surg 24:173-185,1989 4. Ito Y, Donahoe PK, Hendren WH: Differentiation of intramural ganglia in the dissociated rectosigmoid of the rat: An organ culture study. J Pediatr Surg 12:969-975,1977 5. Kamagata S, Donahoe PK: The effect of fibronectin on cholinergic differentiation of the fetal colon. J Pediatr Surg 20:307-314, 1985 6. Kuroda T, Doody DP, Donahoe PK: Aberrant colonic expression of MHC class II antigens in Hirschsprung’s disease. Aust N Z J Surg 61:351-359,199l 7. Fadda B, Maier WA, Meier-Ruge W, et al: Neuronale intestinale Dysplasie, eine kritische lo-Jahres-Analyse klinischer und bioptischer Diagnostik. Z Kinderchir 38:305-311,1983 8. Puri P, Lake BD, Nixon HI-I, et al: Neuronal colonic dysplasia: An unusual association of Hirschsprung’s disease. J Pediatr Surg 12:681-685,1977 9. Scharli AF, Meier-Ruge W: Localized and disseminated forms of neuronal intestinal dysplasia mimicking Hirschsprung’s disease. J Pediatr Surg 16:164-170, 1981 10. Briner J, Oswald HW, Hirsig J, et al: Neuronal intestinal dysplasia-Clinical and histochemical findings and its association with Hirschsprung’s disease. Z Kinderchir 41:282-286, 1986 11. Parham P, Barnstable CJ, Bodmer WF: Use of a monoclonal antibody (W6/32) in structural studies of HLA-A,B,C, antigens. J Immunol123:342-349,1979 12. Brickell M, McConnell I, Milstein C, et al: A monoclonal antibody to the HLA-DR product recognizes a polymorphic Ia determinant in mice. Immunology 43:493-501,198l 13. Kamovsky MJ, Roots L: A “direct coloring: thiocholine” method for cholinesterase. J Histochem Cytochem 12:219-221, 1964 14. Hohlfeld R: Neurological autoimmune disease and the trimolecular complex of T-lymphocytes. Ann Neural 25:531-538, 1989 15. Janeway CA, Bottomly K, Babich J, et al: Quantitative variation in Ia antigen expression plays a central role in immune regulation. Immunol Today 5:99-105,1984 16. Bottazo GF, Pujol-Borrell R, Hanafusa T, et al: Role of aberrant HLA-DR expression and antigen presentation in induction of endocrine autoimmunity. Lancet 2:1115-1118,1983
17. Traugott U: Multiple sclerosis: Relevance of Class I and Class II MHC-expressing cells to lesion development. J Neuroimmunol 16:283-302,1987 18. Bottazo GF, Dean MB, McNally JM, et al: In situ characterization of autoimmune phenomena and expression of HLA molecules in the pancreas in diabetic insulitis. N Engl J Med 313:353360,1985 19. Markmann J, Lo D, Naji A, et al: Antigen presenting function of class II MHC expressing pancreatic beta cells. Nature 336:476-479,1988 20. Csiba A, Whitwell HL, Moore M: Distribution of histocompatibility and leukocyte differentiation antigens in normal human colon and in benign and malignant colonic neoplasms. Br J Cancer 50:699-709,1984 21. Daar AS, Fabre JW: The membrane antigens of human colorectal cancer cells: Demonstration with monoclonal antibodies of hetergeneity within and between tumours and of anomalous expression of HLA-DR. Eur J Cancer Clin Oncol19:209-222,1983 22. Olsson T, Holmdahl R, Klareskog L, et al: Dynamics of Ia-expression of cells and T lymphocytes of different subsets during experimental allergic neuritis in Lewis rats. J Neural Sci 66:141149,1984 23. Daar AS, Fuggle SV, Fabre JW, et al: The detailed distribution of MHC class II antigens in normal human organs. Transplantation 38:293-298,1984 24. Ito Y, Tatekawa I, Nishiyama F, et al: Ultrastructural localization of acetylcholinesterase activity in Hirschsprung’s disease. Arch Path01 Lab Med 111:161-165, 1987 25. Kleinhaus S, Boley SJ, Sheran M, et al: Hirschsprung’s disease. A survey of the members of the Surgical Section of the American Academy of Pediatrics. J Pediatr Surg. 14:588-597, 1979 26. Kliick P, Tibboel D, Leendertse-Jerloop K, et al: Disturbed defecation after colectomy for aganglionosis investigated with monoclonal antineurofilament antibody. J Pediatr Surg 21:845-847, 1986 27. Bienenstock J, Befin AD: Review: Mucosal immunology. Immunology 41:249-270,198O 28. Stead RH, Bienenstock J, Stanisz AM: Neuropeptide regulation of mucosal immunity. Immunol Rev 100:333-359.1987 29. Lane PW: Association of megacolon with two recessive spotting genes in the mouse. J Hered 57:29-31,1966 30. Lider 0, Santos L, Lee C, et al: Suppression of experimental autoimmune encephalomyelitis by oral administration of myelin basic protein. II. Suppression of disease and in vitro immune responses is mediated by antigen-specific CD8+ T-lymphocytes. J Immunol142:748-752,1989 31. Nussenblatt R, Caspi R, Mahdi R, et al: Inhibition of S-antigen introduced experimental autoimmune uveoretinitis by oral induction of tolerance with S-antigen. J Immunol 144:16891695.1990
MHC ANTIGENS
IN NEUROENTERIC
363
DISORDERS
Discussion P. Pun’ (Dublin, h-eland): Class II antigens are generally thought to be expressed by antigenpresenting cells like macrophages and dendritic cells. It is also well known that other cell types under pathological conditions can express class II molecules. Epithelial cells in a number of conditions such as intestine and trachea can express class II antigens. On the other hand, class I antigens are considered to be expressed on the surface of all nucleated cells. As described in the manuscript, the lamina propria of the gut can contain many cells that express class II antigen and, in a disease like Hirschsprung’s disease in which there is a high incidence of enterocolitis and, therefore, inflammatory responses, increase in the class II antigen expression in the lamina propria will reflect inflammatory responses. We have not looked at the actual expression in the lamina propria or in the epithelial cells. We have shown that patients with enterocolitis associated with Hirschsprung’s disease have marked increase in neuronal size in the colon. However, the authors normal observation of class II antigen expression in neural fibers in patients with Hirschsprung’s disease and in the disorder of neuronal intestinal dysplasia is particularly intriguing. There is speculation that this may represent some form of immunological attack with a possible autoimmune etiology is solely based given the widely described associated of class II antigen expression with autoimmune diseases. If this observation represents a meaningful autoimmune response, then a search for organ antibody with neural elements in the colon would be particularly interesting. However, the detection of inappropriate or ectopic class II expression is not always in the context of auto-immune immunity, particularly in relation to bowel. For example, class II expression has been described in the mucosa of patients affected by carcinoma, colon disease, ulcerative colitis, and also in intestine biopsies from patients with celiac diseases. In these diseases, the existence of an autoimmune mechanism is not clear, certainly not up until now. However, Dr Doody’s paper has clearly presented us with a new challenge of our understanding of the possible cause of Hirschsprung’s disease and the clinical consequences arising
from it. I look forward with interest to further developments in this area. I do have two short questions for you. When you describe your research as marked elevation of class II antigen in aganglionosis, is it more the case that you may be talking about the appearance of class II antigens in interstitial hypertrophic neural fibers. And, second, have you considered looking for the evidence of auto antibodies to the neural fibers? J. Larger (Hamilton, Ontario): We have had a few patients with pseudoobstruction-type syndromes who have normal ganglion cells on their biopsy specimens and who have a waxing and waning type of history and we’ve often wondered whether this might represent some type of autoimmune disease. Have you have had an opportunity to look at any patients with pseudoobstruction? D. Doody (response): Dr Puri’s points are well taken. To answer his questions specifically, he asked if the increased class II expression does not seem to be more in the hypertrophic nerve trunks themselves, we did describe it on the hypertrophic nerve trunks, but also we seem to notice that there is an even more increased expression on small cells that appear to be associated with the hypertrophic nerve trunks. These cells need to be defined histologically before any further comment can be made. We have not looked specifically at autoantibodies, although we have begun investigation looking at different T-lymphocyte population, as well as natural killer cells, and those studies are only preliminary. I won’t be able to really tell you too much about it, although there does appear to be an increased expression of the cytotoxic T-lymphocytes. Dr Langer, we have had an opportunity to look at one patient who falls under the classification of pseudoobstruction. This patient has what appears to be ganglion cells, but on longitudinal section, these cells appear to be somewhat hypoganglionic and perhaps a little bit dysplastic. This specimen also showed increased MHC class II expression, both in the lamina propria and around these cells, so it may be that this will be a very accurate marker for pseudoobstruction that has not really been defined before.
Massachusetts
@Although the etiology of Hirschsprung’s disease and neuronal intestinal dysplasia remains obscure, both have histological abnormalities involving ganglion cells and neuronal elements. Searching for a common pathway that may inhibit normal maturation of neurogenic precursors, we examined the possible role of an immune mechanism in the maldevelopment of the enteric neural network. Six patients with Hirschsprung’s disease were studied by comparing biopsy specimens from diseased colon with ones taken from proximal ganglionic colon in the same patients. These were similarly compared with colonic biopsy specimens from patients studied with chronic constipation or bowel removed at the time of operation for other disorders. Biopsies were taken from four other patients with neuronal intestinal dysplasia. Each was examined by hematoxylin 81 eosin staining, acetylcholinesterase histochemistry, and immunohistochemistry of major histocompatibility complex (MHC) class I and class II antigens. All rectal samples from Hirschsprung’s disease patients exhibited elevated acetylcholinesterase histochemistry and absent ganglia to confirm the diagnosis. These findings were correlated with marked elevation of class II MHC in the aganglionic area, whereas the proximal normal ganglionic segments showed no elevation. Rectal biopsy specimens from patients with chronic constipation exhibited no such elevation. A similar elevation of class II MHC was detected in the mucosa and submucosa of all four patients with the rare neuronal intestinal dysplasia disorder whose diagnosis was confirmed by giant ganglia in Auerbath’s plexuses, aberrant Meissner’s ganglia in the lamina propria mucosa, and giant neurofibrils in the mucosa and submucosa. The correlation of elevated class II MHC in these two neuronal dysfunction disorders may indicate an underlying autoimmune mechanism as is seen in thyroiditis and insulin dependent diabetes mellitus. These observations confirm those found in a small number of Hirschsprung’s disease patients reported earlier, and extend the observation to other neuronal dysfunction disorders. These findings provoke a reexamination of the etiology of Hirschsprung’s disease and may predict alternate therapies. Copyright o 1992 by W.B. Saunders Company INDEX WORDS:
Hirschsprung’s
disease; neuronal intestinal
dysplasia; class II major histocompatibility
complex.
H
IRSCHSPRUNG’S disease (HD) is characterized by the absence of enteric ganglia and the presence of abnormal neural elements that, more commonly, involves the distal hindgut.’ The etiology of HD remains unclear, but has been attributed to a failure of migration of neural crest cells into the distal colon,’ to neural precursor abnormalities, or to a nonpermissive intestinal microenviromnent.3 The early timing of normal neural crest migration in the fetus to the distal colon4 and the usual lack of association of HD with major hindgut anomalies make failure of Journal of Pediatric Surgery,
Voi 27.
No 3
(March), 1992: pp 357-363
migration a less satisfying explanation. Therefore, we earlier proposed that factors leading to failure of differentiation after migration could be responsible for the disease complex.5 The cellular specificity of neural destruction led us further to hypothesize that an immune mechanism could underlie HD.h Neuronal intestinal dysplasia (NID) shows a distinct morphologic picture characterized by hyperplasia of enteric ganglia, presence of isolated ganglia in the lamina propria mucosa, and increased activity of acetylcholinergic nerve fibers.’ Until 1983, the association of NID and HD was infrequently reported.‘,“ However, more current series suggest that NID may occur in 26% of children with HD and that up to 75% of cases of NID have some element of aganglionosis.‘.” The coincidence of these two diseases suggests that a common factor may disturb the normal maturation of the enteric ganglia. The cell-specific neuronal abnormalities also apparent in NID anomalies led us to examine the possible role of immune mechanisms in this maldevelopment of the enteric nervous system. Therefore, we studied the expression of the major histocompatibility complex (MHC) class I and II antigens in a series of intestinal biopsy specimens using immunohistochemical techniques and correlated elevation of class II with the diseased state. MATERIALS AND METHODS Six children with HD (mean age, 2.7 years: range, 5 days to 7 years), five with rectosigmoid aganglionosis and one with total colon aganghonosis, had six biopsies from the aganglionic area, three from the transitional area, and three from normoganglionic area. Pathological investigation showed a classical picture of HD, showing the absence of enteric ganglion cells and elevation of mucosal and submucosal acetylcholinesterase (AChE) activity. Four children had histologically diagnosed NID (mean age, 3.3 years: range, 6 months to 6 years). Of these, three had accompanying HD. All colonic specimens in the children with NID showed moderate elevation of AChE-positive nerve fibers. isolated ganglia in the lamina propria. and hyperplasia of submucosal and myen-
From the Division of Pediatric Surgery and the Pediatric Surgical Research Laboratories, Massachusetts Generai Hospital and Harvard Medical School, Boston, MA. Presented at the 22nd Annual Meeting of the American Pediattic SurgicalAssociation, Lake Buena vista, Florida, Mav 15-18. 1991. Address reprint requests to Patricia K. Donahoe, MD, Department of Pediattic Surgery, Massachusetts General Hospital, Warren Bldg, Room 1133. Boston, MA 02114. Copyright o 1992 by W.B. Saunders Company 0022-3468/92/2703-0016$03.OOiO 357
358
HIROBE ET AL
teric plexuses. Dysplastic giant ganglia in the submucosal plexuses were often associated with thickened nerve fibers.’ For comparison we studied biopsy specimens from 10 children (mean age, 6.7 years; range, 3 months to 15 years), including 3 rectal specimens from patients with severe chronic constipation, 3 colonic specimens from patients with anorectal malformations, 1 colonic specimen from one patient each with ulcerative colitis, Crohn’s disease, and volvulus, and 1 jejunal specimen from a patient with a choledocal cyst. Specimens from each of these patients failed to exhibit abnormalities in the enteric nervous system by AChE or hematoxylin and eosin stains.
Antibodies and Immunohistochemistry Mouse monoclonal antibody specific for a skeletal determinant of human HLA-A,B,C (W6/32. Seralab, Crowley Down, England)” and rat monoclonal antibody specific for a monomorphic determinant of human HLA-DR (YE 2/36 HLK, Seralab)” were used to detect MHC class I and II antigens. Specimen were fixed in Zamboni’s solution for 6 hours at 4”C, washed overnight in 0.1
mol/L phosphate buffered saline (pH 7.4) containing 7% sucrose, embedded in OCI compound (Miles Scientific, Naperville, IL), then frozen and stored in liquid nitrogen until sectioned. The frozen tissues were cut with a cryostat into serial S-km sections and mounted on gelatin-coated glass slides at -2o”C, and air-dried for 30 minutes at room temperature. They were first treated with 1% hydrogen peroxide:50% methanol for 30 minutes at 4°C and then with 5% bovine serum albumin (BSA) in 0.05 mol/L Tris-HCl buffer (pH 7.4) in saline (TBS) for 30 minutes. Monoclonal antibodies at optimal dilution (HLA-DR and HLA-A,B,C: 1 to 2,000) with TBS containing 5% BSA were added for 45 minutes at room temperature in a moist chamber. In an earlier study,“6 the first monoclonal antibody was biotinylated. After comparative studies, we elected to biotinylate the second antibody for these studies. After rinsing with TBS, the second antibodies, either appropriate biotinylated horse antimouse or rabbit antirat immuno-
giibuiin G (heavy and light chain specific) (Vector Laboratories, Inc, Burlingame, CA) diluted 1:200 and containing 5% human serum, were added for 30 minutes. Thereafter, sections were treated for 30 minutes with ABC reagent (avidin DH and biotinylated peroxidase H, Elite kit; Vector Laboratories, Inc). A mixture of 0.05% 3,3’-diaminobenzidine tetra-hydrochloride (grade II; Sigma Chemical Co, St Louis, MO) in 0.05 mol/L Tris-HCl buffer (pH 7.4), and 0.01% hydrogen peroxide and 10 mmol/L sodium azide was then added to develop the peroxidase reaction. Sections were lightly counterstained with 1% methyl green and examined with a light microscope. As a negative control, nonimmune appropriate rat or mouse serum were used in place of primary antibodies. Lymphoid tissue in Peyer’s patches was well visualized after biotinylating the second antibody and, therefore, provided positive control cells within selected specimens. Selected sections were stained with hematoxylin and eosin and for AChE histochemistry using Karnovsky and Roots’ direct coloring method.” RESULTS
MHC Class II Antigen Expression In the control bowel, MHC class II antigen was seen in the lamina propria mucosa. These class II-positive cells had either a small round cell or dendritic morphology and tended to be polarized toward the lumen (Fig 1A). In the submucosal and muscle layers, small numbers of cells with dendritic
Fig 1. Elevated class II MHC antigen expression in the mucosa of eganglionic colon. (A) Normoganglionic colon: normal levels of class II-positive cells (open arrow) with small round cell or dendritic morphology are seen in the lamina propria. Staining in Peyer’s patches (PP) provides a positive control tissue for comparison (closed arrow). (B) Aganglionic colon: marked increase of class II antigen (open arrow) staining is seen in the lamina propria where positive cells are diffusely scattered. (Original magnification x 100.)
morphology or ceils lining vessels stained for class II. However, no staining could be found within the nerve plexuses (Fig 2A). In the aganglionic colon of patients with HD, there cells was a marked increase in class II staining distributed throughout the intestinal wall in all cases (Fig 3B). The pattern of distribution of these cells correlated closely with the AChE-positive hypertrophied nerve fibers (Fig 3A). There was a decided increased distribution of class II antigens in the
359
MI-K ANTIGENS IN NEUROENTERIC DISORDERS
colon. The abnormal transitional ganglia were decorated with ectopic class II (Figs 2B and 2C). The proximal normoganghonic colon of HD showed a pattern of class II expression that was similar to that found in control specimens. Specimens taken from patients with NID had a marked increase in the distribution of class II antigen in the cellular elements of the lamina propria (Fig 5B), where moderate elevation of AChE-positive nerve fibers and ectopic isolated ganglia were also found (Fig 5A). Dysplastic giant ganglia in the submucosal plexus were often associated with thickened AChE-positive nerve fibers (Fig 5C), which stain for class II antigen (Fig 5D). In Auerbach’s plexus numerous giant ganglia (Fig 5E) with associated class II-positive cells were also seen (Fig 5F). MHC Class I Expression
Specimens taken from patients with HD, NID, and from normal controls stained identically with antibodies to MHC class I antigens in all epithelial and interstitial cells in the lamina propria, and in vesselassociated cells and cells with dendritic morphology in the submucosa and muscle layers. Class I antigen stained the myenteric plexuses of control specimens, the transitional plexuses of HD, the hyperplastic plexuses of NID, and the thickened nerve fibers in the affected colon of HD and NID. DISCUSSION
MHC class II antigen is a cell surface glycoprotein involved in the immune recognition of foreign tissue and in the regulation of the immune response. In order to induce an immune reaction, antigen must first be processed intracellularly and then be presented on the cell surface as peptides bound to the MHC molecules expressed on the antigen presenting Fig 2. Class II expression In transkion zone myenteric plexus55 of a HD patient. (A) Normoganglionic colon for comparison shows a small number of class II-positive cells (arrows) outside the myenteric plexus. Serial sections of the transition colon stained for (B) class II and (C) AChE show increased expression in a hypertrophic nervr (open arrow), and transitional ganglia (arrows) are surrounded by cells expressing class II antigen. (Original magnification x200.)
lamina propria where positive cells tended to scatter diffusely (Fig 1B). Specimens from older patients tended to express more cellular class II antigens in the lamina propria. Class II staining within the nerve fibers was seen in cells with two or three processes (Figs 4B and 4C) and occasionally in larger fibrous components (Fig 4A). These findings were found as early as the 5-day-old neonate. In the transition zones of three patients with classical HD, class II expression was elevated in the AChE-positive hypertrophied nerve fibers as had been observed in the aganglionic
Fig 3. Conspicuous increase of (A) AChE and (8) class II expmssfon in the submucose of agangffonic rectum in the case of a &day-old neonate. Class II expression correlates wfth the AChE-posftive hypertrophied nerve bundles. MM, muscularis mucosa. (Original magnification x200.)
360
HROBE ET Al.
increased tissue specific expressions of class II antigens. Moreover, the induced expression of MHC antigens on islets of Langerhans in transgenic models has resulted in failure of islet maturation and secondary diabetes.19 The present study explores the possibility that the enteric nervous system could be the cell specific target of disturbances of the immune response, and so examines the expression of major histocompatibility antigens in the dysfunctional tissue of HD and NID. The staining patterns of MHC class I and II antigens in specimens without neuroenteric disorders and in positive control Iymphoid tissue (Fig 1A) were similar to those previously reported.W,21In the case of HD and NID, class I expression was precisely the same as that observed in normal tissue, but there was marked elevation of class II expression in the aganglionic areas, especially within the hypertrophied nerve
Class
II
Fig 4. Class II expression in the neural elements of an aganglionic areas. (A) Increased class II staining is seen in gbrillar components of hypettrophied nerves (original magnification x400). (6) Cross-section through hypertrophied nerve in the intermyenteric space demonstrates class II-positive interdigitating cells (arrow) (original magnification x200). CM, circular muscle; LM, longitudinal muscle. (C) Class II-positive cells in B at higher magnification (original magnification x 1,000) have short intercalating processes (arrow).
cells. The antigen-MHC complex is then recognized by the T cell receptor on the T lymphocyte membrane. Activated T cells then mediate an array of immunologic effector mechanisms that can lead to tissue damage.14 MHC molecules, especially class II, are known to play a major role in pathological states. A considerable body of evidence has accumulated that indicates that a quantitative variation in class II expression can be associated with profound alterations in immune responsiveness and immunologically mediated disease.15 For example, diseases such as thyroiditis,16 multiple sclerosis,17 and insulin-dependent diabetes mellitus” have been associated with aberrant and
Fig 5. Expression of AChE (left column; A, C. E) and class II (right column; B, D, F) in NID. (A) In the mucosa, there are increased AChE-positive nerve fibers and the presence of an ectopic ganglion cell (arrow), whereas(B) a disseminated and increased distribution of class II bearing cells is found in the lamina propria. (C) In serial sections of the submucosa, thickened nerve fiber with a ganglion (arrow) are seen in Meissner’s submucosal plexus by AChE stain. (D) Class II expression is seen in the nerve fibers but not in the ganglia (arrow). In the intermyenteric plexus of Auerbach, (E) AChE stain shows numerous hypertrophic ganglia (arrow) and (F) an increased population of associated Class II positive cells (arrow). (Original magnifications A,B,E,F: x400; C,D: x200.)
361
MHC ANTIGENS IN NEUROENTERIC DISORDERS
fibers (Figs 3 and 4). Previous studies indicate class II antigens are not expressed in the central nervous systemI nor in the peripheral nervous systems.“.23 This deficiency of class II antigens in normal nerve tissue is thought to be responsible for the apparent lack of neural immunoreactivity. Therefore, the ectopic presence of class II expression in and about the neural elements of these disorders can be considered aberrant and abnormal. The aganglionic colon in ‘HD may represent the final stage of a major immunological insult, so we intensely examined the transition zone, searching for clues in an area of continued pathophysiologic activity. Where previous electronmicroscopic studies had shown degenerated ganglion cells,“’ these hypoganglionic plexuses were decorated with class II antigens (Fig 2B). Continuing immunological events, visualized in the transition zone as in these cases, may be responsible for the 5% to 15% of patients whose symptoms persist after prior operations for HD.‘5.‘h We also suspected that patients with NID would provide morphologic evidence of an immunologic process because of the specificity of the target tissue and the concurrence of the two diseases. All specimens examined in the children with NID showed elevated expression of class II antigens (Table 1). Therefore, it is feasible’that an immune mechanism may play a role in its etiology, and that a common mechanism may account for the etiology of both HD and NID. The intestinal mucosa is a major site of contact with the external environment and contains numerous immune effector cells that form gut-associated lymphoid tissue, a major component of the body’s immune defense.” Recently, it was recognized that there were consistent interaction between enteric nervous systems and mucosal immune systems, each of which was influenced by the other? In the lamina propria of HD and NID, AChE-positive ‘nerve fibers are accompanied by class II-positive cells whose increased distribution may represent macrophages, dendritic cells, and, possibly, B cells. It is possible that the large numbers of cholinergic nerve fibers or their released peptides may initiate an abnormal reaction Table 1. MHC Class II Expression in Bowel Specimens No Studied
Class II Elevation
Hirschsprung’s disease Aganglionic
6
6
Hypoganglionic
3
3
Normoganglionic
3
0
4
4
Control
10
0
Total samples
26
Neuronal intestinal dysplasia
of the mucosal immune system. Conversely, the antigens presented at the mucosal surface may initiate an immune response that may incite hypertrophy of the nerves or degeneration of the ganglia. It is apparent that further studies will be needed to clarify the immunopathological mechanisms that underlie HD. The correlation of class II elevation with HD is now evident; the role of this abnormal finding with the pathogenesis of HD remains to be elucidated. However, the lessons learned from other disease states’6-18lead us to conclude that bowel with elevated class II expression may be highly susceptible to abnormal responses of immune origin. The findings of class II expression in the nerve fibers of a 5-day-old neonate indicates that this abnormal expression can occur very early and, one would presume, in utero. These findings may reflect an ongoing immunologic response. An alternative consideration is that MHC expression during development may create a nonpermissive microenvironment for neurogenic differentiation. In addition, the abnormal and increased expression of MHC antigens may reflect an epiphenomenon, related to changes in the extracellular matrix or to the local, increased concentration of cholinergic neurotransmitters. Ontogeny studies of class II antigen in both ganglionic and aganglionic colon in animal models with congenital megacolon” should be helpful in facilitating a study of the immune mechanisms of aganglionosis. In vitro studies may allow a more precise definition of the mechanisms that are responsible for the neuroenteric dysplasia that occurs in HD and NID. Further, experiments can be designed to examine the role of cholinergic substrates and products and luminal antigens in stimulating MHC class II expression. It may be possible to define an inciting antigen by eluting peptides bound to an affected individual’s class II molecules. In this way, immunomodulatory therapy can be individually designed in the future. The offending antigens causing experimental allergic encephalomyelitis, myelin basic protein3’ and experimental uveitis, retinal S-antigen3’ have been identified. It should be possible to define the offending antigen from the large quantity of bowel resected in cases of NID and HD, which provide a unique opportunity to biochemically identify an etiologic factor that is specific both to the disease process and to each affected individual in order to tailor subsequent care. It is important to restudy patients with HD who do not do well and continue to have symptoms after a technically satisfactory pull-through procedure. We recommend studying remaining fixed tissue for signs of NID. If sections have been frozen for AChE
362
HIROBE ET AL
histochemistry, we recommend restudy of these sections for the characteristic morphologic signs of ectopic lamina propria mucosa ganglion cells, increased population of AChE fibers, and enlarged myenteric ganglia. In addition, we now recommend restaining for class II MHC immunohistochemistry.
We suspect that gradations of abnormalities will occur between HD and NID that will link these as functional and anatomic autoimmune abnormalities with a common etiology. An understanding of the common underlying defect will help direct novel theories for the future.
REFERENCES 1. Swenson 0, Bill AH: Resection of the rectum and rectosigmoid with preservation of the sphincter for benign spastic lesions producing megacolon. Surgery 24:212-220,1948 2. Okamoto E, Ueda T: Embryogenesis of intramural ganglia of the gut and its relation to Hirschsprung’s disease. J Pediatr Surg 2:437-443,1967 3. Molenaar JC, Tibboel D, van der Kamp AWM, et al: Diagnosis of innervation-related motility disorders of the gut and basic aspects of enteric nervous system development. Prog Pediatr Surg 24:173-185,1989 4. Ito Y, Donahoe PK, Hendren WH: Differentiation of intramural ganglia in the dissociated rectosigmoid of the rat: An organ culture study. J Pediatr Surg 12:969-975,1977 5. Kamagata S, Donahoe PK: The effect of fibronectin on cholinergic differentiation of the fetal colon. J Pediatr Surg 20:307-314, 1985 6. Kuroda T, Doody DP, Donahoe PK: Aberrant colonic expression of MHC class II antigens in Hirschsprung’s disease. Aust N Z J Surg 61:351-359,199l 7. Fadda B, Maier WA, Meier-Ruge W, et al: Neuronale intestinale Dysplasie, eine kritische lo-Jahres-Analyse klinischer und bioptischer Diagnostik. Z Kinderchir 38:305-311,1983 8. Puri P, Lake BD, Nixon HI-I, et al: Neuronal colonic dysplasia: An unusual association of Hirschsprung’s disease. J Pediatr Surg 12:681-685,1977 9. Scharli AF, Meier-Ruge W: Localized and disseminated forms of neuronal intestinal dysplasia mimicking Hirschsprung’s disease. J Pediatr Surg 16:164-170, 1981 10. Briner J, Oswald HW, Hirsig J, et al: Neuronal intestinal dysplasia-Clinical and histochemical findings and its association with Hirschsprung’s disease. Z Kinderchir 41:282-286, 1986 11. Parham P, Barnstable CJ, Bodmer WF: Use of a monoclonal antibody (W6/32) in structural studies of HLA-A,B,C, antigens. J Immunol123:342-349,1979 12. Brickell M, McConnell I, Milstein C, et al: A monoclonal antibody to the HLA-DR product recognizes a polymorphic Ia determinant in mice. Immunology 43:493-501,198l 13. Kamovsky MJ, Roots L: A “direct coloring: thiocholine” method for cholinesterase. J Histochem Cytochem 12:219-221, 1964 14. Hohlfeld R: Neurological autoimmune disease and the trimolecular complex of T-lymphocytes. Ann Neural 25:531-538, 1989 15. Janeway CA, Bottomly K, Babich J, et al: Quantitative variation in Ia antigen expression plays a central role in immune regulation. Immunol Today 5:99-105,1984 16. Bottazo GF, Pujol-Borrell R, Hanafusa T, et al: Role of aberrant HLA-DR expression and antigen presentation in induction of endocrine autoimmunity. Lancet 2:1115-1118,1983
17. Traugott U: Multiple sclerosis: Relevance of Class I and Class II MHC-expressing cells to lesion development. J Neuroimmunol 16:283-302,1987 18. Bottazo GF, Dean MB, McNally JM, et al: In situ characterization of autoimmune phenomena and expression of HLA molecules in the pancreas in diabetic insulitis. N Engl J Med 313:353360,1985 19. Markmann J, Lo D, Naji A, et al: Antigen presenting function of class II MHC expressing pancreatic beta cells. Nature 336:476-479,1988 20. Csiba A, Whitwell HL, Moore M: Distribution of histocompatibility and leukocyte differentiation antigens in normal human colon and in benign and malignant colonic neoplasms. Br J Cancer 50:699-709,1984 21. Daar AS, Fabre JW: The membrane antigens of human colorectal cancer cells: Demonstration with monoclonal antibodies of hetergeneity within and between tumours and of anomalous expression of HLA-DR. Eur J Cancer Clin Oncol19:209-222,1983 22. Olsson T, Holmdahl R, Klareskog L, et al: Dynamics of Ia-expression of cells and T lymphocytes of different subsets during experimental allergic neuritis in Lewis rats. J Neural Sci 66:141149,1984 23. Daar AS, Fuggle SV, Fabre JW, et al: The detailed distribution of MHC class II antigens in normal human organs. Transplantation 38:293-298,1984 24. Ito Y, Tatekawa I, Nishiyama F, et al: Ultrastructural localization of acetylcholinesterase activity in Hirschsprung’s disease. Arch Path01 Lab Med 111:161-165, 1987 25. Kleinhaus S, Boley SJ, Sheran M, et al: Hirschsprung’s disease. A survey of the members of the Surgical Section of the American Academy of Pediatrics. J Pediatr Surg. 14:588-597, 1979 26. Kliick P, Tibboel D, Leendertse-Jerloop K, et al: Disturbed defecation after colectomy for aganglionosis investigated with monoclonal antineurofilament antibody. J Pediatr Surg 21:845-847, 1986 27. Bienenstock J, Befin AD: Review: Mucosal immunology. Immunology 41:249-270,198O 28. Stead RH, Bienenstock J, Stanisz AM: Neuropeptide regulation of mucosal immunity. Immunol Rev 100:333-359.1987 29. Lane PW: Association of megacolon with two recessive spotting genes in the mouse. J Hered 57:29-31,1966 30. Lider 0, Santos L, Lee C, et al: Suppression of experimental autoimmune encephalomyelitis by oral administration of myelin basic protein. II. Suppression of disease and in vitro immune responses is mediated by antigen-specific CD8+ T-lymphocytes. J Immunol142:748-752,1989 31. Nussenblatt R, Caspi R, Mahdi R, et al: Inhibition of S-antigen introduced experimental autoimmune uveoretinitis by oral induction of tolerance with S-antigen. J Immunol 144:16891695.1990
MHC ANTIGENS
IN NEUROENTERIC
363
DISORDERS
Discussion P. Pun’ (Dublin, h-eland): Class II antigens are generally thought to be expressed by antigenpresenting cells like macrophages and dendritic cells. It is also well known that other cell types under pathological conditions can express class II molecules. Epithelial cells in a number of conditions such as intestine and trachea can express class II antigens. On the other hand, class I antigens are considered to be expressed on the surface of all nucleated cells. As described in the manuscript, the lamina propria of the gut can contain many cells that express class II antigen and, in a disease like Hirschsprung’s disease in which there is a high incidence of enterocolitis and, therefore, inflammatory responses, increase in the class II antigen expression in the lamina propria will reflect inflammatory responses. We have not looked at the actual expression in the lamina propria or in the epithelial cells. We have shown that patients with enterocolitis associated with Hirschsprung’s disease have marked increase in neuronal size in the colon. However, the authors normal observation of class II antigen expression in neural fibers in patients with Hirschsprung’s disease and in the disorder of neuronal intestinal dysplasia is particularly intriguing. There is speculation that this may represent some form of immunological attack with a possible autoimmune etiology is solely based given the widely described associated of class II antigen expression with autoimmune diseases. If this observation represents a meaningful autoimmune response, then a search for organ antibody with neural elements in the colon would be particularly interesting. However, the detection of inappropriate or ectopic class II expression is not always in the context of auto-immune immunity, particularly in relation to bowel. For example, class II expression has been described in the mucosa of patients affected by carcinoma, colon disease, ulcerative colitis, and also in intestine biopsies from patients with celiac diseases. In these diseases, the existence of an autoimmune mechanism is not clear, certainly not up until now. However, Dr Doody’s paper has clearly presented us with a new challenge of our understanding of the possible cause of Hirschsprung’s disease and the clinical consequences arising
from it. I look forward with interest to further developments in this area. I do have two short questions for you. When you describe your research as marked elevation of class II antigen in aganglionosis, is it more the case that you may be talking about the appearance of class II antigens in interstitial hypertrophic neural fibers. And, second, have you considered looking for the evidence of auto antibodies to the neural fibers? J. Larger (Hamilton, Ontario): We have had a few patients with pseudoobstruction-type syndromes who have normal ganglion cells on their biopsy specimens and who have a waxing and waning type of history and we’ve often wondered whether this might represent some type of autoimmune disease. Have you have had an opportunity to look at any patients with pseudoobstruction? D. Doody (response): Dr Puri’s points are well taken. To answer his questions specifically, he asked if the increased class II expression does not seem to be more in the hypertrophic nerve trunks themselves, we did describe it on the hypertrophic nerve trunks, but also we seem to notice that there is an even more increased expression on small cells that appear to be associated with the hypertrophic nerve trunks. These cells need to be defined histologically before any further comment can be made. We have not looked specifically at autoantibodies, although we have begun investigation looking at different T-lymphocyte population, as well as natural killer cells, and those studies are only preliminary. I won’t be able to really tell you too much about it, although there does appear to be an increased expression of the cytotoxic T-lymphocytes. Dr Langer, we have had an opportunity to look at one patient who falls under the classification of pseudoobstruction. This patient has what appears to be ganglion cells, but on longitudinal section, these cells appear to be somewhat hypoganglionic and perhaps a little bit dysplastic. This specimen also showed increased MHC class II expression, both in the lamina propria and around these cells, so it may be that this will be a very accurate marker for pseudoobstruction that has not really been defined before.
