CIRCULATING IMMUNE COMPLEXES IN DERMATITIS HERPETIFORMIS

400 of the clinically unaffected skin in and fibrin 12 and complement 9,13 are present D.H.,9-11 in the involved skin. On the basis of these observations, it seems possible that immune complexes may be formed in the intestine in response to gluten and circulate to the skin in D.H. to cause damage. We now report the presence of complexes in the sera of most D.H. patients taking a normal diet; these complexes are probably IgG, 8-lOS, and complement fixing. We have compared the frequency of the complexes in D.H. with that in two other intestinal diseases-Crohn’s disease and ulcerative colitis-in which skin lesions may also occur and in which immune mechanisms have also been suggested to play a role.

dermal

CIRCULATING IMMUNE COMPLEXES IN DERMATITIS HERPETIFORMIS A. V. HOFFBRAND J. F. MOWBRAY P. P. SEAH E. T. HOLBOROW LIONEL LIONEL FRY FRY

Departments of Experimental Pathology and Dermatology, St. Mary’s Hospital, London W2; Department of Hœmatology, Royal Postgraduate Medical School, London W12; and Medical Research Council Rheumatism Unit, Canadian Red Cross Memorial Taplow, Berkshire

Hospital,

complexes have been detected anti-complementary method in by the sera of 12 (80%) of 15 patients with dermatitis herpetiformis (D.H.) taking a normal diet, and in 4 (36%) out of 11 taking a gluten-free diet. The size of the immune complexes was between 8 and 10S. Immune complexes were detected in the sera of only 3 (12%) of 25 D.H. patients by the C’lq method, but this method detected complexes in 9 (28%) of 32 patients with Crohn’s disease, 2 of 6 patients with cœliac disease, and 2 of 6 patients with ulcerative Summary

Immune

Patients and Methods Patients The sera from 26 patients with D.H., 32 with Crohn’s disease, 6 with coeliac disease, and 6 with ulcerative colitis were examined. Of the 26 patients with D.H., 15 were taking a normal diet and 11 a gluten-free diet (G.F.D.). The tests were also carried out on the sera of 20 male and 10 female healthy adult volunteers.

an

colitis. Introduction

GLUTEN plays a central role in the pathogenesis of coeliac disease (c.D.) and dermatitis herpetiformis (D.H.), but the precise mechanism by which gluten damages the small intestine in C.D. and D.H. and causes the skin lesions in D.H. is unknown. There is accumulating evidence, however, that immunological processes are involved. Low IgM levels 1,2 and antireticulin antibody3 have been found in the sera of about a third of the patients. There is a proportional increase in IgM and IgG producing plasma-cells compared with IgA cells in the small intestinal mucosa in C.D.4-6In addition, antibodies reactive with the basement membrane of the intestinal epithelium have been reported,’ and Shiner and Ballardhave demonstrated the presence of IgA and complement in the jejunal mucosa after gluten challenge. Although gluten could damage the small intestine by direct contact, it could damage the skin only if a toxic fraction entered the bloodstream. However, IgA deposits occur in the DR MANNINEN AND OTHERS:

4. 5. 6. 7.

REFERENCES—continued

Ojala, K., Karjalainen, J., Reissell, P. Lancet, 1972, i, 150. Greenberger, N. J., MacDermott, R. P., Martin, J. F., Dutta, S. J. Pharmac. exp. Ther. 1969, 167, 265. Caldwell, J. H., Martin, J. F., Dutta, S., Greenberger, N. J. Am. J. Physiol. 1969, 217, 1747. Beerman, B., Hellström, K., Rosén, A. Circulation, 1971, suppl. II, p. 139.

8. Huffman, D. H., Azarnoff, D. L. J. Am. med. Ass. 9. Eisner, M. Br. med. J. 1968, iv, 679. 10.

11. 12. 13. 14. 15. 16.

papillse

1972, 222, 957.

Justin-Besançon, L., Cornet, A., Grivaux, M., Guerre, J., Wattez, E., Castera, R., Ecoiffier, J. Presse méd. 1965, 73, 763. Margieson, G. R., Sorby, W. A., Williams, H. B. L. Med. J. Aust. ,

.

1966, ii, 1272. James, W. B., Hume, R. Gut, 1968, 9, 203. Schwartz, I., Lehman, E., Ostrove, R., Seibel, J. M. Gastroenterology, 1953, 25, 416. Goodman, L. S., Gilman, A. The Pharmacological Basis of Therapeutics; p. 524. New York, 1970. Manninen, V., Melin, J., Reissell, P. Lancet, 1972, i, 490. Reissell, P., Manninen, V., Ojala, K., Karjalainen, J. Ann. clin. Res. (in the press).

’

Methods for Detection of Immune Complexes Anti-complementary method.-The method is based on the binding of C’lto circulating complexes after heat inactivation at 56 °C of the C’l which may have been bound in the blood, with a subsequent back titration of added guineapig complement. Fresh sera were stored at - 70°C until assayed. They were thawed and heated at 56°C for sixty minutes in 0-1 ml. volumes. Guineapig complement (2-5 units) was added and the samples maintained at 4°C for thirty minutes. 0-1 ml. of a 3% suspension of optimally sensitised sheep red blood-cells were added and incubated at 37 °C for a further fifteen minutes. The samples were then diluted with 1 0 ml. of barbitonebuffered saline (pH 7-2) and after centrifugation at 3000g for five minutes the lysis was calculated from the released hxmoglobin in the supernatant using a Gilford 300N spectrophotometer. The curve for the complement titration was constructed using known standards, and the loss of hasmolytic complement activity was determined. The results are expressed as the units of complement left in the assay which had a lower 95% confidence limit of 1-1units remaining. C’lq method.-The method of Agnello et a1.14 was used with the gel buffer of molarity 10 mM. Size of Circulating Immune Complexes Using the anti-complementary technique described above, complexes in the fractions obtained by gel filtration of serum samples on columns of ’Sephadex G-200’ could be assayed. The position of the activity was determined as were the positions of IgG, IgM, and IgA by radial immunodiffusion assay on the same samples. The isolated peak of activity was collected and applied to the top of a 5-44% sucrose density gradient in phosphatebuffered 0-15M saline, pH 7. The gradient was centrifuged at 410,000 g for 7 hours and the anti-complementary activity of the fractions from the gradient tubes determined and the density measured. Dialysis before assay was not required because sucrose, at concentrations below 3000’ did not interfere with the assay. The molecular size of the complexes was determined by the gel-filtration technique using markers of known molecular weights, and their density was determined by the ultracentrifugation method.

-

Results

Dermatitis

Herpetiformis Anti-complementary method.-Circulating

immune

401

complexes were detected in the sera of 12 (80 %) of the 15 patients taking a normal diet, but in only 4 (36%) of the 11 sera from patients taking a G.F.D. C’lq method.-3 (12%) of the 25 patients showed immune complexes in the sera by this test. All 3 positive results occurred in the sera of the 15 patients taking a normal diet. Controls None of the 30 controls gave positive tests by either the anti-complementary or the C’lq techniques. Other Conditions 9 (28%) of 32 sera from patients with Crohn’s disease showed the presence of immune complexes by the C’lq method, 2 of 6 sera from patients with coeliac disease, and 2 of 6 sera from patients with ulcerative colitis also showed the presence of immune complexes by the C’lq method. These sera were not tested by the anti-complementary method. Size of Circulating Immune Complexes in D.H. The size of the immune complexes detected by the anti-complementary method in 4 patients tested was found to be between 8 and 10S (see figure). This is the size of IgG complexes. In the four sera studied the densities of the complexes were found to be 1-22, 1-27, 1-41, and 1-37. All four figures are less than the density of nucleic acids and implies that these complexes are unlikely to be virus complexes.

in the exact type of complex which each technique detects. Whether the complexes are involved in the pathogenesis of the skin and gut lesions or are secondary to this is at the present uncertain. The significantly lower frequency of circulating complexes in patients taking a G.F.D. could imply that these complexes contain part of the gluten molecule. The fact that these were present at all in four patients taking a G.F.D. may then be that the diet was not strictly gluten-free. Alternatively, it may be that the complexes do not contain part of the gluten molecule and it is possible that they are related to part of the reticulin molecule. Anti-reticulin antibody occurs in D.H. and C.D.3 The titre of anti-reticulin antibody falls slowly when the patient takes a G.F.D., and this shows a good correlation to the clinical results of the healing of the small bowel and skin. In addition, cross-reactivity has been demonstrated between fraction ill of gluten and reticulin.111 It may be, therefore, that if connective tissue is damaged by gluten, complexes might be formed for some considerable time later from this damaged connective tissue if this takes a long time to heal. The finding of circulating immune complexes and the supposition that they migrate to the skin and cause lesions there would be compatible with the presence of complement 9,13and fibrin 12 in the skin lesions. However, the fact that the upper part of the small intestine is more severely affected than the ileum would suggest that, rather than the damage to the gut being caused solely by circulating complexes, some form of direct effect of the gluten on the intestine might be operative. Circulating immune complexes in Crohn’s disease and ulcerative colitis were mentioned in an abstract. 155 Their incidence, however, was not recorded. We have found that almost a third of patients with each of these diseases have complexes in their sera by the C’lq method, and this is similar to that found by the anticomplementary method (unpublished). However, before attributing a direct causative role to immune complexes in D.H., Crohn’s disease, and ulcerative colitis, the possibility that these complexes are secondary to the disease process must be considered. Further studies are needed to define the characteristics of these complexes and to elucidate their role, if any, in the pathogenesis of the gut and skin lesions in these diseases. We thank Mrs Theresa Rufus and Mr K. Ganeshaguru for expert technical assistance and Dr D. L. Brown for help with the C’lq method. L. F. and P. P. S. are in receipt of grants from the Medical Research Council and the Wellcome Trust.

Requests for reprints should be addressed to L. F., Departof Dermatology, St. Mary’s Hospital, London W2 1NY.

ment

Discussion

These results show that immune complexes

REFERENCES

are

present in the sera of most patients with glutensensitive enteropathy and the skin lesions of D.H. The size of the complexes are between 8 and 10S, which implies that they are IgG complexes. We are uncertain why we detected complexes more often by the anti-complementary method compared with the C’lq method, but it is presumably related to differences in the sensitivity of the two methods or to differences

1. 2. 3. 4.

5. 6. 7.

Fry, L., Keir, P., McMinn, R. M. H., Cowan, J. D., Hoffbrand, A. V. Lancet, 1967, ii, 729. Hobbs, J. R., Hepner, G. W. ibid. 1968, i, 217. Seah, P. P., Fry, L., Hoffbrand, A. V., Holborow, E. J. ibid. 1971, i, 834. Crabbé, P. A. Signification du tissu lymphoide de mugueses digestive; pp. 52, 117. Brussels, 1967. Douglas, A. P., Crabbé, P. A., Hobbs, J. R. Gut, 1969, 10, 413. Soltoft, J. Clin. exp. Immun. 1970, 6, 413. Dick, H. D., Fraser, N. G., Murray, D. Br. J. Derm. 1969, 81, 692. References continued overleaf

402 Patients Studied

Preliminary

Communications

EVIDENCE FOR COMPLEMENT-BINDING IMMUNE COMPLEXES IN ADULT CŒLIAC DISEASE, CROHN’S DISEASE, AND ULCERATIVE COLITIS C. C. BOOTH

WILLIAM F. DOE

D. L. BROWN

Sera were collected from 91 patients with adult coeliac disease, all of whom had an abnormal jejunal biopsy which improved following treatment with a gluten-free diet. 10 patients showing the characteristic clinical, radiological, and biopsy findings of ulcerative colitis and 21 patients with Crohn’s disease in whom the diagnosis was based on typical clinical and radiological findings, and in some instances confirmed by histology, were also studied. Sera from 31 healthy laboratory workers were used as controls. All

sera were

stored

M.R.C. Intestinal

Malabsorption Group, Department of Medicine, and Department of Immunology, Royal Postgraduate Medical School, Ducane Road, London W12

Precipitin reactions to the C1q comSum ary ponent of complement were found in the sera of a high proportion of patients with adult cœliac disease, Crohn’s disease, and ulcerative colitis, providing evidence for circulating immune complexes in these diseases. Adult cœliac patients taking a normal diet had a significantly higher frequency of positive reactions than those taking a gluten-free diet, suggesting that a dietary antigen may be a component of the complex.

at

-20°C until tested. RESULTS

Results obtained in the three groups of patients are in table i. In 30% of sera from adult coeliac

set out

TABLE I-PREVALENCE OF

TABLE SENCE

Clq

PRECIPITATION

II-RELATIONSHIP BETWEEN GLUTEN-FREE DIET AND PREOF

Clq

PRECIPITATION IN ADULT CCF-LIAC DISEASE SERA

INTRODUCTION

EVIDENCE suggests that circulating antigen/antibody complexes are present in the sera of some patients with adult coeliac disease, Crohn’s disease, and ulcerative We have previously observed mixed colitis. 1,2 cryoglobulinaemia and vasculitis in association with adult coeliac diseaseand others have reported the presence of IgG immune complexes in the sera of some patients with Crohn’s disease and ulcerative colitis.2 Recently Agnello et al.3 have described a technique for detecting certain circulating immune complexes using the ability of the Clqcomponent of complement to precipitate soluble complexes in agar gel. We have used this technique to study the prevalence of Clqbinding immune complexes in these diseases. MATERIALS AND METHODS

ClPrecipitation

Test

The Clq component of complement was prepared from fresh human serum by D.N.A. precipitation and ’Sephadex G-200’ gel filtration as described by Agnello et al.3 Clq precipitation was tested in agarose gel, and the plates were read after 48 hours at room temperature and again after 72 hours at 4°C. All positive reactions were confirmed by repeat testing. Anti-Clq antibody gave a line of identity with precipitin lines of positive sera. Heat-aggregated IgG was used as a positive control.

DR MOWBRAY AND OTHERS: 8. 9. 10.

11. 12. 13. 14.

15.

REFERENCES—continued

Shiner, M., Ballard, J. Lancet, 1972, i, 1202. van der Meer, J. B. Br. J. Derm. 1969, 81, 493. Chorzelski, T. P., Beutner, E. H., Jablonska, S., Blaszcyk, M., Triffshauser, C. J. invest. Derm. 1971, 56, 373. Seah, P. P., Fry, L., Stewart, J. S., Chapman, B. L., Hoffbrand, A. V., Holborow, E. J. Lancet, 1972, i, 611. Mustakallio, K. K., Blomqvist, K., Laiho, K. Ann. clin. Res. 1970, 2, 13. Holubar, K., Doralt, M., Eggerth, G. Br. J. Derm. 1971, 85, 505. Agnello, V., Winchester, R. J., Kunkel, H. G. Immunology, 1970, 19, 909. Jewell, D. P., MacLennan, I. C. M., Truelove, S. C. Gut, 1972, 13, 839

(abstr.).

*p< 0-001

(xi= 13°79).

patients there was a precipitin reaction with the purified Clq component of complement. Similarly, 57% of Crohn’s patients’ sera and 20% of ulcerative-colitis sera gave positive reactions. By contrast, only 1 of 31 sera from a healthy control group was positive for Clq precipitation. Effect of Gluten-free Diet

in Adult Cceliac Disease

When sera from coeliac patients taking a normal diet were compared with those from patients on a glutenfree diet, there was a significant difference in the prevalence of Clq precipitin reactions (table 11). 13 adult caeliac patients whose sera were Clq-positive while taking a normal diet were re-studied after 6 months on a gluten-free diet. In 9 of these patients Clq reactivity disappeared from the sera following treatment with a

gluten-free diet. DISCUSSION

The Clqprecipitation test is a recognised technique for detecting certain circulating antigen/antibody complexes.3 These are apparently small, soluble immune complexes capable of binding the Clq component of complement. 3,44 In an established " immune-complex

disease " such

systemic lupus erythematosus, Clqreacting complexes may be detected in the circulation during disease activity,and may be an essential feature in the pathogenesis of the disease. Our observations on the sera of patients with adult coeliac disease taking a normal diet show that a high proportion contain Clq-reactive material, but that the as