Clin. exp. Immunol. (1977) 29, 256-260.
Experimental autoimmune thyroiditis in the vervet monkey D. J. PUDIFIN, JUNE DUURSMA & P. BRAIN Department of Medicine, University of Natal, and Natal Institute of Immunology
(Received 30 November 1976) SUMMARY
Autoimmune thyroiditis was induced in two vervet monkeys by immunizing them with human thyroid extract. As expected, both animals developed cytotoxic antibodies which were active against human thyroid and other human cells in tissue culture. In addition, a second serum factor was detected. This was capable of sensitizing human thyroid (and other) cells and rendering them susceptible to killing by normal human lymphocytes. This finding implicates antibody-dependent cell-mediated cytotoxicity in the pathogenesis of autoimmune thyroiditis.
INTRODUCTION Experimental autoimmune thyroiditis has been induced and studied in various animal species (Weigle, 1973) using either homologous or heterologous thyroid extract as the provoking antigen. Despite these experiments and the many observations made on autoimmune thyroiditis in humans, the immunopathology of the disorder is not yet clearly understood. We are particularly interested in the possible role of antibody-dependent cell-mediated cytotoxicity (ADCMC) in autoimmune disease. Using a subhuman primate immunized with human thyroid extract as our experimental model, we have attempted to assess this, as well as other possible pathogenetic mechanisms. MATERIALS AND METHODS Two male vervet monkeys (Cercopithecus aethiops pygerythrus) (number 369 and 379) aged 3 years and about 4 years and weighing 3-2 and 4-2 kg were used. Venous blood samples were taken on two occasions during the 3 weeks prior to immunization. Observations were made on these and on post-immunization specimens which were taken over 9 months. All tests included controls using blood drawn from three healthy vervet monkeys matched for age and weight. Partial thyroidectomy was carried out on the two test animals at 6 and 9 months respectively and they were killed at 9 5 months. Thyroid antigen. Portions of several human thyroidectomy specimens weighing 22 5 g in all were mixed in an equal volume of phosphate-buffered saline (PBS), homogenized in a blender, and then treated in a sonicator. After centrifugation the clear supernatant, the protein content of which was 2-8 g/100 ml, was bottled in 1 ml aliquots and stored at - 20'C. Immunization procedure. Each test animal received 1 ml of the antigen with 1 ml of Freund's complete adjuvant by injection into the pads of one hand and one foot. Further injections of 1 ml of antigen were given i.m. without adjuvant on days 20, 50 and 110. Thyroid antibodies. Thyroglobulin antibodies were sought using the tanned red-cell agglutination test (Wellcome). Indirect immunofluorescence with both human and monkey thyroid substrates was also employed. Sera were diluted 1:10 for testing, and conjugated rabbit anti-human IgG (Wellcome) was used to detect fixed antibody. Cytotoxicity tests. Target cells were monolayer cultures grown from monkey thyroid, human thyroid and human skin (fibroblasts). Trypsin was used to remove them from their culture flasks. For the tests, the cells were transferred, after washing and counting, to the wells of 'Microtest' plates (Falcon Plastics). After overnight culture, a count of viable, adherent cells was made. The medium in the wells was then removed by gentle suction, and suspensions of the test lymphocytes (prepared by gradient centrifugation over Ficoll-Hypaque) were substituted. The ratio of lymphocytes to target cells was 100:1. After a further 24 hr of incubation, unattached cells were washed out of the plates and surviving adherent cells were fixed with methanol and stained (Giemsa). The attached cells were then counted and comparison was made with the numbers present in wells to which no lymphocytes had been added. Correspondence: Dr D. J. Pudifin, Department of Medicine, P.O. Box 17039, Congella 4013, Natal, Republic of South Africa.
256
Experimental monkey thyroiditis
257
The effects of serum on the target cells were similarly assessed on monolayer cultures in Microtest plates. Medium in the wells was first replaced by serum diluted 1:100 or 1:1000 in tissue-culture medium; the preparation was then incubated for 6 hr. At this stage the wells were again washed and a suspension of normal lymphocytes was added, followed by a further 24 hr incubation before the final washing, fixing and staining. This procedure detects direct serum effects and ADCMC. Sera were used fresh and after heating at 560C for 30 min to destroy complement. In all these experiments, results are expressed simply according to the formula: Number of cells surviving in test wells Number of cells surviving in control wells Each figure represents the mean of the counts in ten to twenty wells. T- and B-lymphocyte counts. Our method of lymphocyte typing has been described in detail elsewhere (Brain et al., 1976). Basically it involves identification of E rosettes and of immunoglobulin-bearing lymphocytes in a combined procedure. Thus T cells, B cells, null cells and cells carrying both markers can be accurately enumerated. Regular full blood counts were performed and absolute numbers of lymphocyte types calculated. It has previously been shown (Brain, Gordon & Willets, 1970) that a large proportion of the lymphocytes of both the vervet monkey and of the chacma baboon make rosettes with red cells, and unpublished studies here have shown that cells making such rosettes seldom have detectable surface immunoglobulin; just as in man, therefore, the rosette-forming lymphocytes in the vervet are T cells. There appears to be complete cross-reactivity between vervet and human immunoglobulin, so that an anti-human immunoglobulin serum labelled with fluorescein can be used to detect B cells in this species.
RESULTS
Thyroid histology Both glands showed similar changes (Fig. 1). There were focal areas of thyroiditis with infiltration by lymphoid cells. No such abnormality was observed in a thyroid from a control monkey. At the general autopsies, no macroscopic abnormalities (apart from the thyroidectomy scars) were noted.
Thyroid antibodies Serum taken before immunization gave negative results with both tests for thyroid antibody. The haemagglutination test detected thyroglobulin antibodies in both test monkeys but in none
FIG. 1. Thyroid histology.
258
D.
J.
Pudifin, June Duursma & P. Brain
of the controls. In monkey 369, they were first detected on the 8th day after immunization. The titre rose to 1/1280 over 2 months, then fell again, but the antibodies were still detectable when the experiment was terminated. In monkey 379, the antibodies were first detected on day 30. A rise in titre, with a subsequent fall was again noticed (Fig. 2). Antibodies to thyroid colloid, detected by immunofluorescence, were found in parallel to those above. In addition, on the human thyroid substrate, fluorescence was seen on the interstitial tissue of the gland. This was not seen on the monkey thyroid substrate. Serum from monkey 379 did, however, appear to contain an antibody to human thyroid cell cytoplasm.
Cytotoxicity tests (see Table 1) Lymphocytes from immunized and control animals were cytotoxic to both types of human target cell. None of the lymphocytes had a cytotoxic effect on the target cells grown from monkey thyroid. Sera at 1:100 from the test animals exerted a strong, direct toxic effect on human cells. This effect was abolished by preheating the serum to inactivate complement. The sera did not damage the target cells grown from monkey thyroid. administered
Arntigen
25601
t
t
t
1280 _ 640 320 00 160 80 _*-* .1 40 I> a20
0'379
369
0
,0--
-0
0~~~~ I
1. ,,
0
20
40
140 160 Days FIG. 2. Thyroid antibody titres.
60
80
100
120
180
200 220 240
TABLE 1. Results of cytotoxicity tests Percentage survival of each type of target cell Effector agent(s)
Human
Medium only Control serum 1:100 serum 1:100 369 serum 1:100 379 serum 1:100 (preheated) 369 Control serum 1:1000 Control serum 1 :1000+normal human lymphocytes serum 1:1000 369 serum 1 :1000+normal human lymphocytes 369 serum 1:1000 (preheated)+normal human lymphocytes 369 serum 1:1000 379 serum 1 :1000+normal human lymphocytes 379 serum 1 :1000+normal monkey lymphocytes 369 serum 1 :1000+normal monkey lymphocytes 379
Monkey thyroid culture
culture
Human skin fibroblast
100 100 0 46 100 100 88 100 26 61 100 38
100 98 0 5 100 100 84 100 14 37 100 19
100 100 100 100
-
100 100
thyroid
-
-
100
Experimental monkey thyroiditis
259
The same sera at a dilution of 1:1000 had no direct cytotoxic effect on any ofthe target cells. However, the addition of normal human lymphocytes after removal of the serum resulted in damage to the human target cells (ADCMC). This was so whether fresh or preheated serum was used. When the target cells were from a culture of monkey thyroid, and normal monkey lymphocytes were used as aggressors with the same sera, no ADCMC was observed.
Circulating lymphocytes There were fairly wide fluctuations in the total numbers of circulating lymphocytes in the test monkeys (1125-3984/mm3), but the ratio of T cells to B cells remained relatively constant. The fluctuations in total lymphocyte numbers did not follow a recognizable pattern or have any relationship to the times of the immunizing injections. Similar fluctuations were seen before immunization and in the control animals. DISCUSSION The experimental model which we have used appears satisfactory for the study of autoimmune thyroiditis because the two main criteria of the disease, viz demonstrable thyroid antibodies and histological lesions in the thyroid gland, are adequately fulfilled. The condition has been studied in numerous other animals such as the rat (Jones & Roitt, 1961), rabbit, guinea-pig, dog (Terplan et al., 1960) and Rhesus monkey (Kite, Argue & Rose, 1966). The present model, in which a subhuman primate is immunized with human thyroid antigen, is perhaps the closest approximation to the human disease. Antibody to thyroglobulin was readily detected by two techniques. The rise to high titres soon after its appearance, with subsequent decline, is similar to the pattern seen during the induction of thyroiditis in the rabbit. The initial rise in the rabbit coincides with the development of lesions in the thyroid, while the decline has been shown to be due to combination with thyroglobulin released from damaged thyroid cells (Clinton & Weigler, 1972). Rabbit thyroiditis can be transferred by serum alone, and the similar behaviour pattern of the antibody in the monkey suggests a similar pathogenetic role. Antibody to thyroid cell cytoplasm is rarely found in experimental thyroiditis although it is a regular feature of the human disease (Kite, Argue & Rose, 1966). Monkey 369 developed this antibody, which was demonstrable by immunofluorescence on monkey and human substrate, over the 2 months during which the thyroglobulin-antibody titre was at its highest. Cross-reactivity of this sort has been shown elsewhere between human and Rhesus monkey thyroid-cell cytoplasm (Beutner & Witebsky, 1963). Cytotoxicity. The indiscriminate damaging effect exerted by lymphocytes from immunized and control monkeys on target cells of human origin can be regarded only as a non-specific phenomenon for which there is no clear explanation. The fact that the same lymphocytes did no damage to target cells of monkey origin implicates some form of non-specific or xenogeneic effect. We were thus not able to draw any conclusions about the role of direct lymphocyte cytotoxicity in this experimental model. The severe damage to human thyroid and skin fibroblast cultures brought about by serum at 1:100 from both test monkeys suggests the presence of complement-dependent cytotoxic antibody. This was borne out by repeating the experiment using the serum after heating it to 560C for 30 min, and finding that its cytotoxic effect was abolished. In the same experiment, no cytotoxic effect was seen when the sera were used at a dilution of 1:1000. The addition of normal human lymphocytes, however, resulted in a clear cytotoxic effect on the target cells in the wells which contained serum from the immunized monkeys. The effect is not complementdependent. From this, it can be concluded that another antibody is present and that this antibody, which is active in high dilution, sensitizes the target cells. These sensitized cells are then susceptible to attack by normal lymphocytes. Such antibody-dependent cytotoxic effects have previously been shown by more indirect means (Jones & Roitt, 1961; Kite, Argue & Rose, 1966) and the evidence presented here strongly supports the possibility of a role for ADCMC in autoimmune thyroiditis. The sera were used in an identical way in another experiment, in which the target cells came from a
260
D. J. Pudifin, June Duursma & P. Brain
culture of monkey thyroid and the lymphocytes from a normal monkey. The absence of demonstrable ADCMC here was disappointing, but it is quite possible that the target cells were connective tissue cells and not thyroid cells. This would explain the negative result since no effect on non-thyroid monkey cells would be expected. The activity of both antibody types against human thyroid and skin cells is hardly surprising as the immunizing antigen was a crude extract of thyroid tissue from a number of donors and must have contained a variety of non-thyroid human antigens, The direct cytotoxic effect on thyroid cells of sera from animals or patients with autoimmune thyroiditis is well known (Irvine, 1960, 1962; Pulvertaft et al., 1961; Kite et al., 1966). ADCMC is, however, a more recently described mechanism (MacLennan & Loewi, 1968) and its involvement in disease production is not clearly established. Calder et a!. (1973) first showed that serum from patients with Hashimoto's thyroiditis, in combination with normal human lymphocytes, brought about lysis of thyroglobulin-coated chicken red blood cells. We feel that the results reported here support the possibility that ADCMC participates in producing autoimmune thyroiditis. They also indicate as a corollary its implication in other reactions such as graft rejection. The support of the Medical Research Council is gratefully acknowledged. REFERENCES BEUTNER, E.H. & WITEBSKY, E. (1963) Studies on organ
specificity. XV. Immunohistologic evaluation of reactions produced by thyroid antibodies. J. Immunol. 91, 204. BRAIN, P., Cox, J., DuuRsmA, J. & PUDIFIN, D.J. (1976) T and B lymphocytes in three population groups. Clin. exp. Immunol. 23, 248. BRAIN, P., GORDON, J. & WILLETS, W.A. (1970) Rosette formation by peripheral lymphocytes. Clin. exp. Immunol. 6, 681-688. CALDER, E.A., PENHALE, W.J., McLENNAN, D., BARNEs, E.W. & IRVINE, W.J. (1973) Lymphocyte-dependent antibody-mediated cytotoxicity in Hashimoto's thyroiditis. Clin. exp. Immunol. 14, 153. CLINTON, B.A. & WEIGLER, W.O. (1972) Cellular events during the induction of experimental thyroiditis in the rabbit. J. exp. Med. 136, 1605. IRVINE, W.J. (1960) The cytotoxic factor in thyroid disease. Scot. med.J. 5, 511. IRVINE, W.J. (1962) Studies in the cytotoxic factor in thyroid disease. Brit. med. 7. i, 1444.
JONES, H.E.H. & RoITT, I.M. (1961) Experimental autoimmune thyroiditis in the rat. Brit. J. exp. Path. 42, 546. KITE, J.H., ARGUE, H. & ROSE, N.R. (1966) Experimental thyroiditisintheRhesus monkey. Clin. exp.Immunol. 1, 139. MAcLENNAN, I.C.M. & LOEWI, G. (1968) Effect of specific antibody to target cells on their specific and non-specific interactions with lymphocytes. Nature (Lond.), 219, 1069. PULVERTAFT, R.V., DONIACH, D., RoiTT, I.M. & HUDSON, R.V. (1961) The cytotoxic factor in Hashimoto's disease and the incidence in the other thyroid diseases. Brit. 5. exp. Path. 42, 496. TERPLAN, K.L., WITEBSKY, E., RosE, N.R., PAINE, J.R. & EGAN, R.W. (1960) Experimental thyroiditis in rabbits, guinea-pigs and dogs, following immunization with thyroid extracts of their own and heterologous species. Amer.]. Path. 36, 213. WEIGLE, W.O. (1973) Experimental autoimmune thyroiditis. Path. Annual, 8, 329.
Experimental autoimmune thyroiditis in the vervet monkey D. J. PUDIFIN, JUNE DUURSMA & P. BRAIN Department of Medicine, University of Natal, and Natal Institute of Immunology
(Received 30 November 1976) SUMMARY
Autoimmune thyroiditis was induced in two vervet monkeys by immunizing them with human thyroid extract. As expected, both animals developed cytotoxic antibodies which were active against human thyroid and other human cells in tissue culture. In addition, a second serum factor was detected. This was capable of sensitizing human thyroid (and other) cells and rendering them susceptible to killing by normal human lymphocytes. This finding implicates antibody-dependent cell-mediated cytotoxicity in the pathogenesis of autoimmune thyroiditis.
INTRODUCTION Experimental autoimmune thyroiditis has been induced and studied in various animal species (Weigle, 1973) using either homologous or heterologous thyroid extract as the provoking antigen. Despite these experiments and the many observations made on autoimmune thyroiditis in humans, the immunopathology of the disorder is not yet clearly understood. We are particularly interested in the possible role of antibody-dependent cell-mediated cytotoxicity (ADCMC) in autoimmune disease. Using a subhuman primate immunized with human thyroid extract as our experimental model, we have attempted to assess this, as well as other possible pathogenetic mechanisms. MATERIALS AND METHODS Two male vervet monkeys (Cercopithecus aethiops pygerythrus) (number 369 and 379) aged 3 years and about 4 years and weighing 3-2 and 4-2 kg were used. Venous blood samples were taken on two occasions during the 3 weeks prior to immunization. Observations were made on these and on post-immunization specimens which were taken over 9 months. All tests included controls using blood drawn from three healthy vervet monkeys matched for age and weight. Partial thyroidectomy was carried out on the two test animals at 6 and 9 months respectively and they were killed at 9 5 months. Thyroid antigen. Portions of several human thyroidectomy specimens weighing 22 5 g in all were mixed in an equal volume of phosphate-buffered saline (PBS), homogenized in a blender, and then treated in a sonicator. After centrifugation the clear supernatant, the protein content of which was 2-8 g/100 ml, was bottled in 1 ml aliquots and stored at - 20'C. Immunization procedure. Each test animal received 1 ml of the antigen with 1 ml of Freund's complete adjuvant by injection into the pads of one hand and one foot. Further injections of 1 ml of antigen were given i.m. without adjuvant on days 20, 50 and 110. Thyroid antibodies. Thyroglobulin antibodies were sought using the tanned red-cell agglutination test (Wellcome). Indirect immunofluorescence with both human and monkey thyroid substrates was also employed. Sera were diluted 1:10 for testing, and conjugated rabbit anti-human IgG (Wellcome) was used to detect fixed antibody. Cytotoxicity tests. Target cells were monolayer cultures grown from monkey thyroid, human thyroid and human skin (fibroblasts). Trypsin was used to remove them from their culture flasks. For the tests, the cells were transferred, after washing and counting, to the wells of 'Microtest' plates (Falcon Plastics). After overnight culture, a count of viable, adherent cells was made. The medium in the wells was then removed by gentle suction, and suspensions of the test lymphocytes (prepared by gradient centrifugation over Ficoll-Hypaque) were substituted. The ratio of lymphocytes to target cells was 100:1. After a further 24 hr of incubation, unattached cells were washed out of the plates and surviving adherent cells were fixed with methanol and stained (Giemsa). The attached cells were then counted and comparison was made with the numbers present in wells to which no lymphocytes had been added. Correspondence: Dr D. J. Pudifin, Department of Medicine, P.O. Box 17039, Congella 4013, Natal, Republic of South Africa.
256
Experimental monkey thyroiditis
257
The effects of serum on the target cells were similarly assessed on monolayer cultures in Microtest plates. Medium in the wells was first replaced by serum diluted 1:100 or 1:1000 in tissue-culture medium; the preparation was then incubated for 6 hr. At this stage the wells were again washed and a suspension of normal lymphocytes was added, followed by a further 24 hr incubation before the final washing, fixing and staining. This procedure detects direct serum effects and ADCMC. Sera were used fresh and after heating at 560C for 30 min to destroy complement. In all these experiments, results are expressed simply according to the formula: Number of cells surviving in test wells Number of cells surviving in control wells Each figure represents the mean of the counts in ten to twenty wells. T- and B-lymphocyte counts. Our method of lymphocyte typing has been described in detail elsewhere (Brain et al., 1976). Basically it involves identification of E rosettes and of immunoglobulin-bearing lymphocytes in a combined procedure. Thus T cells, B cells, null cells and cells carrying both markers can be accurately enumerated. Regular full blood counts were performed and absolute numbers of lymphocyte types calculated. It has previously been shown (Brain, Gordon & Willets, 1970) that a large proportion of the lymphocytes of both the vervet monkey and of the chacma baboon make rosettes with red cells, and unpublished studies here have shown that cells making such rosettes seldom have detectable surface immunoglobulin; just as in man, therefore, the rosette-forming lymphocytes in the vervet are T cells. There appears to be complete cross-reactivity between vervet and human immunoglobulin, so that an anti-human immunoglobulin serum labelled with fluorescein can be used to detect B cells in this species.
RESULTS
Thyroid histology Both glands showed similar changes (Fig. 1). There were focal areas of thyroiditis with infiltration by lymphoid cells. No such abnormality was observed in a thyroid from a control monkey. At the general autopsies, no macroscopic abnormalities (apart from the thyroidectomy scars) were noted.
Thyroid antibodies Serum taken before immunization gave negative results with both tests for thyroid antibody. The haemagglutination test detected thyroglobulin antibodies in both test monkeys but in none
FIG. 1. Thyroid histology.
258
D.
J.
Pudifin, June Duursma & P. Brain
of the controls. In monkey 369, they were first detected on the 8th day after immunization. The titre rose to 1/1280 over 2 months, then fell again, but the antibodies were still detectable when the experiment was terminated. In monkey 379, the antibodies were first detected on day 30. A rise in titre, with a subsequent fall was again noticed (Fig. 2). Antibodies to thyroid colloid, detected by immunofluorescence, were found in parallel to those above. In addition, on the human thyroid substrate, fluorescence was seen on the interstitial tissue of the gland. This was not seen on the monkey thyroid substrate. Serum from monkey 379 did, however, appear to contain an antibody to human thyroid cell cytoplasm.
Cytotoxicity tests (see Table 1) Lymphocytes from immunized and control animals were cytotoxic to both types of human target cell. None of the lymphocytes had a cytotoxic effect on the target cells grown from monkey thyroid. Sera at 1:100 from the test animals exerted a strong, direct toxic effect on human cells. This effect was abolished by preheating the serum to inactivate complement. The sera did not damage the target cells grown from monkey thyroid. administered
Arntigen
25601
t
t
t
1280 _ 640 320 00 160 80 _*-* .1 40 I> a20
0'379
369
0
,0--
-0
0~~~~ I
1. ,,
0
20
40
140 160 Days FIG. 2. Thyroid antibody titres.
60
80
100
120
180
200 220 240
TABLE 1. Results of cytotoxicity tests Percentage survival of each type of target cell Effector agent(s)
Human
Medium only Control serum 1:100 serum 1:100 369 serum 1:100 379 serum 1:100 (preheated) 369 Control serum 1:1000 Control serum 1 :1000+normal human lymphocytes serum 1:1000 369 serum 1 :1000+normal human lymphocytes 369 serum 1:1000 (preheated)+normal human lymphocytes 369 serum 1:1000 379 serum 1 :1000+normal human lymphocytes 379 serum 1 :1000+normal monkey lymphocytes 369 serum 1 :1000+normal monkey lymphocytes 379
Monkey thyroid culture
culture
Human skin fibroblast
100 100 0 46 100 100 88 100 26 61 100 38
100 98 0 5 100 100 84 100 14 37 100 19
100 100 100 100
-
100 100
thyroid
-
-
100
Experimental monkey thyroiditis
259
The same sera at a dilution of 1:1000 had no direct cytotoxic effect on any ofthe target cells. However, the addition of normal human lymphocytes after removal of the serum resulted in damage to the human target cells (ADCMC). This was so whether fresh or preheated serum was used. When the target cells were from a culture of monkey thyroid, and normal monkey lymphocytes were used as aggressors with the same sera, no ADCMC was observed.
Circulating lymphocytes There were fairly wide fluctuations in the total numbers of circulating lymphocytes in the test monkeys (1125-3984/mm3), but the ratio of T cells to B cells remained relatively constant. The fluctuations in total lymphocyte numbers did not follow a recognizable pattern or have any relationship to the times of the immunizing injections. Similar fluctuations were seen before immunization and in the control animals. DISCUSSION The experimental model which we have used appears satisfactory for the study of autoimmune thyroiditis because the two main criteria of the disease, viz demonstrable thyroid antibodies and histological lesions in the thyroid gland, are adequately fulfilled. The condition has been studied in numerous other animals such as the rat (Jones & Roitt, 1961), rabbit, guinea-pig, dog (Terplan et al., 1960) and Rhesus monkey (Kite, Argue & Rose, 1966). The present model, in which a subhuman primate is immunized with human thyroid antigen, is perhaps the closest approximation to the human disease. Antibody to thyroglobulin was readily detected by two techniques. The rise to high titres soon after its appearance, with subsequent decline, is similar to the pattern seen during the induction of thyroiditis in the rabbit. The initial rise in the rabbit coincides with the development of lesions in the thyroid, while the decline has been shown to be due to combination with thyroglobulin released from damaged thyroid cells (Clinton & Weigler, 1972). Rabbit thyroiditis can be transferred by serum alone, and the similar behaviour pattern of the antibody in the monkey suggests a similar pathogenetic role. Antibody to thyroid cell cytoplasm is rarely found in experimental thyroiditis although it is a regular feature of the human disease (Kite, Argue & Rose, 1966). Monkey 369 developed this antibody, which was demonstrable by immunofluorescence on monkey and human substrate, over the 2 months during which the thyroglobulin-antibody titre was at its highest. Cross-reactivity of this sort has been shown elsewhere between human and Rhesus monkey thyroid-cell cytoplasm (Beutner & Witebsky, 1963). Cytotoxicity. The indiscriminate damaging effect exerted by lymphocytes from immunized and control monkeys on target cells of human origin can be regarded only as a non-specific phenomenon for which there is no clear explanation. The fact that the same lymphocytes did no damage to target cells of monkey origin implicates some form of non-specific or xenogeneic effect. We were thus not able to draw any conclusions about the role of direct lymphocyte cytotoxicity in this experimental model. The severe damage to human thyroid and skin fibroblast cultures brought about by serum at 1:100 from both test monkeys suggests the presence of complement-dependent cytotoxic antibody. This was borne out by repeating the experiment using the serum after heating it to 560C for 30 min, and finding that its cytotoxic effect was abolished. In the same experiment, no cytotoxic effect was seen when the sera were used at a dilution of 1:1000. The addition of normal human lymphocytes, however, resulted in a clear cytotoxic effect on the target cells in the wells which contained serum from the immunized monkeys. The effect is not complementdependent. From this, it can be concluded that another antibody is present and that this antibody, which is active in high dilution, sensitizes the target cells. These sensitized cells are then susceptible to attack by normal lymphocytes. Such antibody-dependent cytotoxic effects have previously been shown by more indirect means (Jones & Roitt, 1961; Kite, Argue & Rose, 1966) and the evidence presented here strongly supports the possibility of a role for ADCMC in autoimmune thyroiditis. The sera were used in an identical way in another experiment, in which the target cells came from a
260
D. J. Pudifin, June Duursma & P. Brain
culture of monkey thyroid and the lymphocytes from a normal monkey. The absence of demonstrable ADCMC here was disappointing, but it is quite possible that the target cells were connective tissue cells and not thyroid cells. This would explain the negative result since no effect on non-thyroid monkey cells would be expected. The activity of both antibody types against human thyroid and skin cells is hardly surprising as the immunizing antigen was a crude extract of thyroid tissue from a number of donors and must have contained a variety of non-thyroid human antigens, The direct cytotoxic effect on thyroid cells of sera from animals or patients with autoimmune thyroiditis is well known (Irvine, 1960, 1962; Pulvertaft et al., 1961; Kite et al., 1966). ADCMC is, however, a more recently described mechanism (MacLennan & Loewi, 1968) and its involvement in disease production is not clearly established. Calder et a!. (1973) first showed that serum from patients with Hashimoto's thyroiditis, in combination with normal human lymphocytes, brought about lysis of thyroglobulin-coated chicken red blood cells. We feel that the results reported here support the possibility that ADCMC participates in producing autoimmune thyroiditis. They also indicate as a corollary its implication in other reactions such as graft rejection. The support of the Medical Research Council is gratefully acknowledged. REFERENCES BEUTNER, E.H. & WITEBSKY, E. (1963) Studies on organ
specificity. XV. Immunohistologic evaluation of reactions produced by thyroid antibodies. J. Immunol. 91, 204. BRAIN, P., Cox, J., DuuRsmA, J. & PUDIFIN, D.J. (1976) T and B lymphocytes in three population groups. Clin. exp. Immunol. 23, 248. BRAIN, P., GORDON, J. & WILLETS, W.A. (1970) Rosette formation by peripheral lymphocytes. Clin. exp. Immunol. 6, 681-688. CALDER, E.A., PENHALE, W.J., McLENNAN, D., BARNEs, E.W. & IRVINE, W.J. (1973) Lymphocyte-dependent antibody-mediated cytotoxicity in Hashimoto's thyroiditis. Clin. exp. Immunol. 14, 153. CLINTON, B.A. & WEIGLER, W.O. (1972) Cellular events during the induction of experimental thyroiditis in the rabbit. J. exp. Med. 136, 1605. IRVINE, W.J. (1960) The cytotoxic factor in thyroid disease. Scot. med.J. 5, 511. IRVINE, W.J. (1962) Studies in the cytotoxic factor in thyroid disease. Brit. med. 7. i, 1444.
JONES, H.E.H. & RoITT, I.M. (1961) Experimental autoimmune thyroiditis in the rat. Brit. J. exp. Path. 42, 546. KITE, J.H., ARGUE, H. & ROSE, N.R. (1966) Experimental thyroiditisintheRhesus monkey. Clin. exp.Immunol. 1, 139. MAcLENNAN, I.C.M. & LOEWI, G. (1968) Effect of specific antibody to target cells on their specific and non-specific interactions with lymphocytes. Nature (Lond.), 219, 1069. PULVERTAFT, R.V., DONIACH, D., RoiTT, I.M. & HUDSON, R.V. (1961) The cytotoxic factor in Hashimoto's disease and the incidence in the other thyroid diseases. Brit. 5. exp. Path. 42, 496. TERPLAN, K.L., WITEBSKY, E., RosE, N.R., PAINE, J.R. & EGAN, R.W. (1960) Experimental thyroiditis in rabbits, guinea-pigs and dogs, following immunization with thyroid extracts of their own and heterologous species. Amer.]. Path. 36, 213. WEIGLE, W.O. (1973) Experimental autoimmune thyroiditis. Path. Annual, 8, 329.
