0021-972X/89/6804-0766$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1989 by The Endocrine Society

Vol. 68, No. 4 Printed in U.S.A.

Thyroid Peroxidase and Thyroid Microsomal Autoantibodies* NAOKATA YOKOYAMA, ALVIN TAUROGt, AND GEORGE G. KLEE Department of Pharmacology, University of Texas Southwestern Medical Center, Dallas, Texas 75235; and the Division of Laboratory Medicine, Mayo Clinic, Rochester, Minnesota 55905

ABSTRACT. The antithyroid microsomal antibodies found in the serum of patients with autoimmune thyroid disease are

directed largely, if not entirely, against thyroid peroxidase (TPO). In this study we used a highly purified, well characterized, large tryptic fragment of porcine TPO (hereafter referred to as purified porcine TPO) to examine possible differences among microsomal antibodies in patients with autoimmune thyroid disease. Antibodies against this TPO preparation and also against a synthetic peptide corresponding to residues 780-793 of the deduced sequence of the native enzyme were compared with microsomal antibodies from patients in immunoblot experiments. The antiporcine TPO and antisynthetic peptide antibodies reacted with crude preparations of human TPO. Binding of serum microsomal antibodies to purified porcine TPO was also found. Purified porcine TPO shows two fragments after gel

S

electrophoresis under reducing conditions: a 59K fragment corresponding to the amino end of the molecule, and two approximately 30K fragments corresponding to the carboxyl end. Using an immunoblot procedure with purified porcine TPO as the antigen, we found that at least two epitopes were involved in microsomal antibody production: one associated with the 59K fragment and the other with the approximately 30K fragment(s). The distribution of serum antibodies against these epitopes differed among the patients, indicating that these antibodies comprise a heterogeneous group. Serum from patients with autoimmune thyroid disease significantly inhibited human TPO activity, raising the possibility that microsomal antibodies may contribute to the impaired thyroid function that occurs in some patients with autoimmune thyroid disease. (J Clin Endocrinol Metab 68: 766, 1989)

antibodies against purified TPO and against residues 780-793 to study human thyroid microsomal antigen and serum antibodies in patients with AITD. We found that the antibodies bound readily to human thyroid microsomal antigen and purified porcine TPO, and that considerable heterogeneity exists among microsomal antibodies from different patients. We also provide further evidence that thyroid microsomal autoantibodies can significantly inhibit the enzymatic activity of TPO.

ERUM from patients with autoimmune thyroid disease (AITD) usually contains antibodies against a thyroid microsomal antigen (1). The nature of this antigen remained unknown for more than 20 yr after its discovery, but recently it has become clear that the thyroid microsomal antigen is very closely related to, if not identical with, thyroid peroxidase (TPO) (2-4). In a previous report (5) we described the isolation of highly purified and extremely active porcine TPO from the particulate fraction of porcine thyroids after initial treatment with trypsin plus detergent. The cDNA of porcine TPO has now been cloned, and the full-length amino acid sequence has been reported (6). Very recently (7), we were able to characterize our purified trypsin fragment of TPO in relation to the known sequence of the native enzyme. For this purpose we used antibodies prepared against the purified enzyme and against synthetic peptides representing known regions of the native enzyme.

Materials and Methods Highly purified tryptic fragment of porcine TPO (hereafter referred to as purified porcine TPO) This was described in detail (preparation XII) in a previous report (7). Native human TPO For the preparation of crude native human TPO, samples of freshly thawed thyroid tissue (stored at - 7 0 C) that had been surgically removed from patients with Graves' disease were homogenized, centrifuged, and solubilized with 0.1% deoxycholate, as previously described (8). In some instances, trypsin was included in the solubilization mixture.

In this study we used highly purified porcine TPO and Received October 13,1988. Address all correspondence and requests for reprints to: Dr. Alvin Taurog, Department of Pharmacology, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, Texas 75235. • This work was supported by USPHS Grant 5-RO1-DK-03612. t USPHS career research awardee.

Antibody to purified porcine TPO The preparation of this antibody was described previously (7). 766

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 January 2017. at 08:05 For personal use only. No other uses without permission. . All rights reserved.

TPO AND THYROID MICROSOMAL AUTOANTIBODIES Antibody to synthetic peptide This antibody was prepared against a synthetic peptide representing residues 780-793 of the deduced sequence of porcine TPO, as previously described (7). Gel electrophoresis and immunoblotting

TSH. Serum total T4 was measured by RIA [normal range, 512.5 Mg/dL (64-161 nmol/L)], and TSH was measured by a modification of the Corning chemiluminescent assay (8). Effect of incubation with antibodies on catalytic activity of TPO Experiments were performed with both crude human TPO and purified porcine TPO. Catalytic activity usually was measured by the guaiacol assay, but in one experiment with human TPO the iodide oxidation assay also was employed.

These procedures were described previously (7). Patient serum samples Serum samples were obtained from 12 patients with AITD, selected because they had high titers of antithyroid microsomal autoantibodies and low titers of antithyroglobulin autoantibodies. This selection was made to increase the probability of immunoreaction with purified porcine TPO and avoid possible interference from the presence of antibodies against thyroglobulin. Only limited information was available regarding the patients' clinical status (Table 1). Eight of the 12 patients had Hashimoto's thyroiditis, 1 had Graves' disease, 1 had Hashimoto's thyroiditis after childbirth and multiple endocrine problems (pheochromocytoma, metastatic islet cell carcinoma, and Von Kippel-Lindau disease), and 2 others had undisclosed thyroid disorders. All 12 patients had antithyroid microsomal autoantibody titers of at least 1:100,000, and all but 1 had antithyroglobulin autoantibody titers less than 1:100. Autoantibody titers were measured using Ames Sera-Tek kits (Elkhardt, IN). Also shown in Table 1 are values for serum T4 and TABLE

767

Human TPO. Patient serum samples were used as the antibody source. The globulin fraction was isolated by precipitation with ammonium sulfate. Ten microliters of deoxycholate-solubilized human TPO (3.7-5.8 mg protein/mL) were added to 50 nL globulin, derived from 50 fxL serum, in 67 mmol/L phosphate, pH 7.0, and the mixtures were allowed to stand at 4 C for 24 h. After the incubation period, 10 /xL were added to 900 nL of a guaiacol assay solution containing 33 mmol/L guaiacol and 0.5 g/L BSA in 67 mmol/L phosphate, pH 7.0. The reaction was started by the addition of 10 nL 33 mmol/L H2O2, and the change in absorbance at 470 nm per min was measured in a Cary 219 spectrophotometer (Varian Instrument Division, Palo Alto, CA). For the iodide oxidation assay, 10 /xL were assayed in the Cary spectrophotometer, as described previously (9). Porcine TPO. Two antibodies were used, rabbit antiporcine TPO serum and rabbit antipeptide serum. Varying amounts of antiserum (10-100 nh) were added to highly purified porcine

1. Laboratory and clinical findings in the patients with AITD

Patient no.

Age (yr)

Sex

1

45

M

T4

TSH (mlU/L)

1:102,400

11.1

1:102,400

Hashimoto's thyroiditis

39.8

1:102,400

Multiple endocrine problems; history of Hashimoto's thyroiditis following pregnancy Slightly tender thyroid, no prior history of endocrine disorders Moderate goiter, exophthalmos; probable Hashimoto's thyroiditis Hypothyroidism due to Hashimoto's thyroiditis, T4 replacement Hashimoto's thyroiditis with smooth enlarged thyroid Hashimoto's thyroiditis with borderline hypothyroidism; multinodular goiter No clinical information

11

F

3

70

F

4

29

F

5

61

F

3.5 29.2

1:102,400

40.0

1:102,400

3.4

1:409,600

6.9

1:102,400

14.7

1:409,600

6

66

F

7

56

F

3.1 (40) 5.7 (73) 4.2 (54)

8

47

F

9

56

F

10

51

F

3.2 (41)

11

13

F

6.4

1:102,400

0.01

1:409,600

(82)

12

53

F

1.0 (13)

Clinical findings

7.8

1:102,400

2

(45) 3.2 (41)

Thyroglobulin antibody titer

Hashimoto's thyroiditis with large firm goiter and hypothyroidism Hashimoto's thyroiditis

1.9 (24)" 3.5 (45) 4.2 (54) 4.6 (59)

Microsomal

antibody titer

1:409,600

1:102,400

Graves' disease with diffusely enlarged thyroid Hashimoto's thyroiditis, T4 replacement

' The T4 values in parentheses represent SI units (nanomoles per L).

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 January 2017. at 08:05 For personal use only. No other uses without permission. . All rights reserved.

JCE & M • 1989 Vol68«No4

YOKOYAMA, TAUROG, AND KLEE

768

TPO (2.4 ng in a final volume of 1 mL). Preimmune serum was used in control samples. The mixtures were allowed to stand at 4 C for 24 h, and 10 nh were removed for assay of guaiacol activity.

Results Binding of porcine TPO antiserum to human TPO Figure 1 shows the results of experiments in which crude preparations of solubilized TPO from Graves' thyroid tissue were subjected to gel electrophoresis, transferred to nitrocellulose sheets, and immunoblotted with antiserum to purified porcine TPO. With crude human TPO solubilized with deoxycholate (Fig. 1, lanes a and c), two major bands were observed at 100K and 110K. These bands were largely eliminated when the antiserum to porcine TPO was pretreated with purified porcine TPO (results not shown), indicating that they did not arise through nonspecific binding. The results in Fig. 1 with antiserum to porcine TPO closely resemble those reported by other investigators using monoclonal antibodies against human TPO (10, 11). Lanes b and d in Fig. 1 show the results obtained with crude human TPO solubilized with trypsin plus deoxycholate. Two bands were found at 86K and 95K, indicating that the trypsin treatment removed about 15K from each of the components.

Binding of microsomal antibodies to human and porcine TPO Figure 2 shows representative results obtained when serum from patients with AITD was reacted with crude human TPO after the latter had been electrophoresed and transferred to nitrocellulose sheets. The results obtained with serum from the other nine patients were very similar to those shown in Fig. 2, and they mimicked the results obtained with the rabbit antiserum to purified porcine TPO. These observations confirm the conclusion drawn by previous investigators that serum microsomal antibodies in patients with AITD are directed against thyroid peroxidase. Lane b in Fig. 2 shows that no bond was found when crude porcine TPO was substituted for crude human TPO. However, as shown in Fig. 4, patient serum samples gave positive immunoblots when the amount of porcine TPO applied to the gel was greatly increased by using highly purified porcine TPO. Binding of antisynthetic peptide antibody of porcine and human TPO In a previous report (7) we described the preparation of an antiserum to a synthetic peptide representing res-

a b c

d e

f

a b e d

FiG. 1. Immunoblots with rabbit antiporcine TPO after SDS-PAGE under reducing conditions of the following antigens: a, deoxycholatesolubilized particulate fraction from Graves' thyroid A; b, deoxycholateplus trypsin-solubilized particulate fraction from Graves' thyroid A; c, deoxycholate-solubilized particulate fraction from Graves' thyroid B; and d, deoxycholate- plus trypsin-solubilized particulate fraction from Graves' thyroid B.

FlG. 2. Immunoblots with serum from patients 2, 10, and 11 with AITD after SDS-PAGE under reducing conditions of the following antigens: a, deoxycholate-solubilized particulate fraction from Graves' thyroid C; b, deoxycholate-solubilized particulate fraction from porcine thyroid; c, deoxycholate-solubilized particulate fraction from Graves' thyroid C; d, deoxycholate- plus trypsin-solubilized particulate fraction from Graves' thyroid C; e, deoxycholate-solubilized particulate fraction from Graves' thyroid C; and f, deoxycholate- plus trypsin-solubilized particulate fraction from Graves' thyroid C.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 January 2017. at 08:05 For personal use only. No other uses without permission. . All rights reserved.

TPO AND THYROID MICROSOMAL AUTOANTIBODIES idues 780-793 of native porcine TPO. This region of the deduced amino acid sequence of porcine TPO was chosen because, on hydropathic analysis (6), it represented a hydrophilic region likely to interact with an antibody to the peptide. Moreover, it contained an amino-terminal cysteine for coupling to hemocyanin. It was also located on the amino-terminal side of the putative membrane spanning region (6). The antibody against this synthetic peptide proved very useful as a probe in elucidating the structure of purified TPO, as previously described (7). Figure 3 shows the results of immunoblot experiments with the antibody against the synthetic peptide, using crude porcine TPO and crude human TPO as antigens. The results with crude porcine TPO are similar to those previously reported (7). The deoxycholate-solubilized enzyme showed a band at 106K, while the trypsin-solubilized TPO showed bands at 88K and 29K. The 29K band corresponds to a carboxy-terminal fragment. Crude human TPO solubilized with deoxycholate also contained bands at 100K and 110K, presumably the same proteins reacting with the microsomal antibodies (Fig. 2). Since the antisynthetic peptide antibody is directed against a very restricted region of TPO, these results provide further evidence that TPO is the antigen that elicits the production of serum microsomal antibodies. Surprisingly, the major band on the immunoblot with the antisynthetic peptide antibody appeared at 28K. Similar results were obtained in two other experiments with crude TPO prepared from the thyroids of other

a b

769

Graves' patients. Presumably, therefore, even when trypsin was not added during the solubilization procedure, there was sufficient endogenous proteolysis to release a 28K carboxy-terminal fragment from human TPO. Such a band was not found with similarly prepared crude porcine TPO. Heterogeneity among serum microsomal antibodies in patients with AITD Figure 4 shows immunoblots obtained with purified porcine TPO as the antigen and with patient serum samples as the antibody sources (lanes a-d). For comparison the figure also shows results using the same antigen and rabbit antiporcine TPO serum (lane f) and rabbit antisynthetic peptide serum (lane e). As previously reported (7), studies with purified porcine TPO indicated that the initial trypsin-detergent solubilization procedure resulted in peptide bond cleavage within a disulflde loop (after Arg residue 561) and also at alternate sites near the carboxy-terminus. This gave rise to fragments at 59K, 32K, and 29K, which were visualized on immunoblots prepared with antiporcine TPO serum (Fig. 4, lane f). Immunoblots prepared with the antisynthetic peptide antibody displayed only the 29K and 32K fragments (Fig. 4, lane e). The band at 88K represents intact TPO that was not cleaved within the disulflde loop by the initial trypsinization or was not reduced by the

a

c

-28K 29K — FlG. 3. Immunoblots with rabbit antipeptide serum after SDS-PAGE under reducing conditions of the following antigens: a, deoxycholatesolubilized particulate fraction from porcine thyroid; b, deoxycholateplus trypsin-solubilized particulate fraction from porcine thyroid; and c, deoxycholate-solubilized particulate fraction from Graves' thyroid A.

FlG. 4. Immunoblots with purified porcine TPO as antigen, comparing patient serum, rabbit antipeptide serum, and rabbit antiporcine TPO serum as the antibody sources. Purified porcine TPO was subjected to SDS-PAGE under reducing conditions, transferred to a nitrocellulose sheet, and blotted with a dilution of patient or rabbit serum. Lanes ad show results with a 1:400 dilution of serum from patients 1, 9, 11, and 10, respectively. Lane e shows results with a 1:500 dilution of rabbit antipeptide serum, and lane f with a 1:1000 dilution of rabbit antiporcine TPO serum. In lanes a-d, 10 ng purified TPO were applied to the gels, and in lanes e and f, 5 /*g were used.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 January 2017. at 08:05 For personal use only. No other uses without permission. . All rights reserved.

YOKOYAMA, TAUROG, AND KLEE

770

JCE & M • 1989 Vol68«No4

TPO after incubation for 24 h at 4 C with globulin derived from 50 /uL patient or normal serum. There was considerable variability among both control and patient groups, but the difference between the means of the two groups, expressed as a percentage of the TPO value in the absence of any globulin, was highly significant (mean ± SD for patient group, 60 ± 19% vs. 102 ± 17% for the normal group; P < 0.001). The results of a second experiment comparing inhibition based on guaiacol and iodide oxidation assays are shown in Table 2. Although the control globulin fractions alone had an appreciable inhibitory effect in the iodide oxidation assay, inhibition in the presence of patient globulin fractions was greater. The magnitude of the mean inhibitory effect of the patients' globulin fraction on the iodide oxidation assay was quantitatively similar to that based on the guaiacol assay, and the correlation coefficient was 0.80 when the inhibitory effects of individual globulin fractions in the two assays were compared. Some patient serum globulin fractions had greater inhibitory effects than others, and there is little doubt that binding of the antibody to TPO in these serum samples affected the catalytic site of the enzyme, either directly or indirectly. All patient serum fractions inhibited TPO activity by at least 15%, based on the mean of three different guaiacol assays, and even though there was considerable variability in individual assay results, no patient serum was totally devoid of inhibitory activity. There was no correlation between the degree of enzyme inhibition and the type of immunoblot pattern shown in Fig. 4. Figure 5B shows the results of an experiment performed with purified porcine TPO and rabbit antiporcine TPO. In this experiment serum rather than the globulin

mercaptoethanol treatment. At least four different immunoblot patterns were observed with patient serum samples (Fig. 4). The results in lane a closely resembled those with the antisynthetic peptide serum (lane e). Comparable results were obtained with serum from two other patients (no. 3 and 6). The microsomal antibodies in these patients, therefore, were directed against an epitope(s) near the carboxyl-terminus, but not any epitope in the 59K fragment. The results in lane b also showed antibodies directed primarily against the carboxyl end of the protein, although in these serum samples a faint band was also observed at 59K. The results were similar with serum from three other patients (no. 2, 4, and 12). The immunoblot in lane c of Fig. 4 showed only faint bands at 29K and 32K and a much more intense band at 59K. The microsomal antibodies in this serum, therefore, were directed primarily against an epitope(s) in the larger amino-terminal fragment of the purified porcine TPO. Similar results were observed in serum from three other patients (no. 5, 7, and 8). The immunoblot pattern in lane d of Fig. 4 was the most complex and appeared in the serum from only one patient. Bands were observed at 29K, 32K, 40K, and 59K. The serum antibodies in this patient, therefore, were directed against sites within both the larger and the smaller fragments of purified porcine TPO. Inhibitory effect of antibodies on TPO enzyme activity To demonstrate significant inhibition of catalytic activity by microsomal autoantibodies we found it necessary to use a limiting amount of enzyme. Figure 5 A shows the guaiacol activity of deoxycholate-solubilized human

A. Human FIG. 5. Inhibitory effect of antibodies on enzyme activity of TPO. A, Crude human TPO was incubated for 24 h at 4 C with the globulin fraction derived from 50 /xL patient (•) or normal (O) serum, and enzyme activity was measured by guaiacol assay. The results are expressed as a percentage of the activity of crude TPO incubated with buffer. B, Highly purified porcine TPO was incubated for 24 h at 4 C with varying volumes of rabbit antiporcine TPO serum (•) or rabbit preimmune serum (A). The results are expressed as a percentage of the activity of purified porcine TPO incubated with buffer.

O

o . o o

100 _

o • o o

80 60 40 20

t

Patient

Normal

20

40

60

80

f t I Of Rabbit Serum

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 January 2017. at 08:05 For personal use only. No other uses without permission. . All rights reserved.

100

TPO AND THYROID MICROSOMAL AUTOANTIBODIES TABLE 2. Inhibition of TPO catalytic activity by globulin fractions of serum containing microsomal autoantibodies and normal serum: comparison of guaiacol and iodide oxidation assays Guaiacol assay (% of control activity)

Iodide oxidation assay (% of control activity)

Normal subjects (n = 8)

Patients (n = 11)

Normal subjects (n = 8)

Patients (n= 11)

93.6 ± 10.5

54.0 ± 24.0°

73.8 ± 7.9

41.7 ± 15.9°

° P < 0.001 us. normal subjects. TABLE 3. Comparison of inhibitory effect of antiporcine TPO and antipeptide serum on guaiacol activity of porcine TPO % of preimmune serum value Serum added 10 25 50 100

Antiporcine TPO serum

Antipeptide serum

11 33 27 19

98 101 105 95

fraction was used. Twenty-five microliters of serum markedly inhibited guaiacol activity. The inhibitory effect was slightly greater than previously reported by Cooper et al. (12). The results in Fig. 5B provide further evidence that antibodies to TPO can inhibit the catalytic activity of the enzyme. Table 3 shows the results of an experiment comparing the inhibitory effect of antiporcine TPO with that of antisynthetic peptide serum on the guaiacol activity of purified porcine TPO. The antiporcine TPO serum again had a marked inhibitory effect, but the antisynthetic peptide serum was completely inactive. These results indicate that residues 780-793 in the porcine TPO molecule lie outside the catalytic site for guaiacol activity.

Discussion Reports from several laboratories (2-4, 10, 11, 13-20) have established that TPO is very closely related to the thyroid microsomal autoantigen that elicits the production of the serum microsomal antibodies associated with AITD. Recently, the full-length primary amino acid sequence of both porcine (6) and human (21, 22) TPO was deduced from cDNA cloning experiments, and a 72% homology was found. The highly purified porcine TPO used in this study was a large tryptic fragment of the native enzyme, and we recently (7) characterized its structure in relation to that of the native enzyme. The high degree of homology between porcine and human TPO made it likely that our porcine preparation might be useful for further immunological characterization of human serum microsomal antibodies. In immunoblot experiments with crude native human

771

TPO as antigen and antiporcine TPO serum as antibody, two major bands were found, at 100K and 110K, when gel electrophoresis was performed under reducing conditions. Similar results were obtained by previous investigators (10, 11) using monoclonal antibodies against human TPO. Kimura et al. (22) reported the existence of two alternatively spliced mRNAs for human TPO, suggesting that the two bands on the immunoblots represent proteins that arise from separate mRNAs. However, Hamada et al. (23) presented evidence that the smaller protein was derived from the larger one by proteolytic activity. In immunoblot experiments similar to those described above, but with crude native porcine TPO instead of crude human TPO, we did not find two separate bands for TPO. In immunoblot experiments with purified porcine TPO as antigen and patient serum as the antibody source, several different patterns were observed when serum from different patients was used. Comparison with the patterns found with the antiporcine TPO serum and the antisynthetic peptide serum provided information concerning the location of the epitopes against which the patient's serum antibodies were directed. Our previous characterization of the purified porcine TPO indicated that the initial trypsin treatment cleaved a peptide bond after Arg residue 561 within a disulfide loop. When purified porcine TPO was subjected to sodium dodecyl sulfate-poly acrylamide gel electrophoresis (SDS-PAGE) under reducing conditions, we found fragments at 59K, 32K, and 29K. The larger fragment contained the aminoterminus, while the two smaller fragments corresponded to alternatively cleaved fragments near the carboxylterminus. The antisynthetic peptide antibody, which was directed against a defined region near the carboxylterminus, reacted only with the 29K and 32K fragments and not with the 59K fragment. In 7 of the 12 serum samples examined, the antibodies reacted primarily with the 29K and 32K fragments, indicating that they were directed against an epitope(s) in this portion of the molecule. In 4 serum samples the antibodies reacted primarily with the 59K fragment, indicating that the epitope(s) was in this part of the molecule. The serum of one patient reacted strongly with both portions of the molecule, indicating the presence of at least two different antibodies in the same serum. These results indicate that considerable heterogeneity exists among serum microsomal antibodies in different patients. There was no obvious correlation between the immunoblot pattern and the clinical diagnosis or history. Libert et al. (18) isolated 10 microsomal antigen clones from a Xgtll cDNA library, using serum from a patient with Hashimoto's thyroiditis as an antibody source. One of these clones reacted with serum samples from 17 of 18 patients with Hashimoto's thyroiditis, indicating that

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 January 2017. at 08:05 For personal use only. No other uses without permission. . All rights reserved.

772

YOKOYAMA, TAUROG, AND KLEE

it harbored a major epitope involved in thyroid autoimmunity. The nucleotide sequence of this clone was found to encode a segment corresponding to residues 590-675 of human TPO, indicating that the epitope lies within this region. It is of interest, however, that this region is not included in the 59K fragment. Thus, in serum from 5 of the patients that we studied, the major epitope did not correspond to that which appeared to be common to almost all the Hashimoto's patients studied by Libert et al. (18). Heterogeneity among microsomal antibodies in serum from patients with AITD previously was reported by several groups of investigators (13, 20, 24). Hamada et al. (24), using immunoblot analysis, found differences in immunoreactivity among patient serum samples toward native, SDS-denatured, and SDS-denatured and reduced thyroid microsomal antigen. They suggested that antibodies against denatured or denatured and reduced microsomal antigen may be related to destruction of the thyroid gland. Doble et al. (20) used a combination of immunochemical and enzymatic techniques to examine different epitopes recognized by autoantibodies to the thyroid microsomal antigen. They concluded that there is a minimum of six autoantigenic epitopes on the thyroid microsomal antigen that can provoke immune reactivity. Kohno et al. (13) used trypsin-solubilized purified human TPO for their studies and concluded that at least three epitopes of TPO are recognized by serum autoantibodies. Both Hamada et al. (24) and Doble et al. (20) reported that the microsomal antibodies in the majority of their patients with AITD did not react with SDS-denatured TPO, as determined in immunoblot experiments using reducing conditions. These results contrast with the results of our study, in which positive immunoblots were found in all 12 patients with both purified porcine TPO and deoxycholate-solubilized human thyroid microsomes as antigens. A possible reason for this difference is that we used only serum containing very high titers of microsomal antibodies (>l:100,000). There is some disagreement regarding the inhibitory effect of microsomal autoantibodies on the catalytic activity of TPO. Portmann et al. (2) concluded that when detergent-solubilized TPO was bound to antibody it retained its guaiacol activity; loss of activity required removal of the immune complex. Kohno et al. (13), on the other hand, used human TPO that had been purified after solubilization with trypsin plus cholate, and they reported that guaiacol activity was inhibited about 20% after incubation with immunoglobulin G fractions of serum from patients with AITD. However, there was considerable variation in the extent of inhibition, and they concluded that serum from some of the patients contained no inhibitory activity. More recently, Doble et al. (20), using crude detergent-solubilized human TPO,

JCE & M • 1989 Vol68«No4

reported that autoantibodies from the majority of patients with AITD significantly inhibited guaiacol activity. Again, there was extensive variability, and some serum samples showed very little inhibitory effect. Even though we used only high titer serum samples, we also found considerable variation in inhibitory activity, and the globulin fraction from some serum samples showed only a small inhibitory effect. However, based on the results of repeated assays even this small inhibitory effect (~15%) appeared to be significant. The globulin fraction from most serum samples, however, showed a highly significant inhibitory effect on TPO catalytic activity. Khoury et al. (25) have provided evidence that in AITD microsomal autoantibodies are present in the apical portion of follicular cells, the probable site of TPO-catalyzed iodination. The inhibitory effect on enzyme activity raises the possibility that microsomal antibodies contribute to the impaired thyroid function that occurs in some patients with AITD. References 1. Doniach D, Bottazzo F, Drexhage HA. The autoimmune endocrinopathies. In, Lachmann PJ, Peters K, ed. Clinical aspects of immunology. Oxford: Blackwell Scientific Publications; 1982;2: 903-37. 2. Portmann L, Hamada N, Heinrich G, DeGroot LJ. Anti-thyroid peroxidase antibody in patients with autoimmune thyroid disease: possible identity with anti-microsomal antibody. J Clin Endocrinol Metab. 1985;61:1001-3. 3. Czarnocka B, Ruf J, Ferrand M, Carayon P, Lissitzky S. Purification of the human thyroid peroxidase and its identification as the microsomal antigen involved in autoimmune thyroid disease. FEBS Lett. 1985;190:147-51. 4. Kotani T, Umeki K, Matsunaga S, Kato E, Ohtaki S. Detection of autoantibodies to thyroid peroxidase in autoimmune thyroid disease by micro-ELISA and immunoblotting. J Clin Endocrinol Metab. 1986;61:928-33. 5. Rawitch AB, Taurog A, Chernoff SB, Dorris ML. Hog thyroid peroxidase: physical, chemical, and catalytic properties of the highly purified enzyme. Arch Biochem Biophys. 1977;194:244-57. 6. Magnusson RP, Gestautas J, Taurog A, Rapoport B. Molecular cloning of the structural gene for porcine thyroid peroxidase. J Biol Chem. 1987;262:13885-8. 7. Yokoyama N, Taurog A. Porcine thyroid peroxidase: relationship between the native enzyme and an active, highly purified tryptic fragment. Mol Endocrinol. 1988;2:838-44. 8. Klee GG, Hay ID. Sensitive thyrotropin assays: analytic and clinical performance criteria. Mayo Clin Proc. 1988;63:1123-32. 9. Nakashima T, Taurog A. Improved assay procedures for thyroid peroxidase; application to normal and adenomatous thyroid tissue. Clin Chim Acta. 1978;83:129-40. 10. Ohtaki S, Kotani T, Nakamura Y. Characterization of human thyroid peroxidase purified by monoclonal antibody-assisted chromatography. J Clin Endocrinol Metab. 1986;63:570-6. 11. Portmann L, Fitch FW, Havran W, Hamada N, Franklin WA, DeGroot LJ. Characterization of the thyroid microsomal antigen, and its relationship to thyroid peroxidase, using monoclonal antibodies. J Clin Invest. 1988;81:1217-24. 12. Cooper DS, Maloof F, Ridgway EC. Rat thyroid peroxidase (TPO) biosynthesis in vitro: studies using antiserum to porcine TPO. Endocr Res. 1987; 13:15-29. 13. Kohno Y, Hiyama Y, Shimojo N, Niimi H, Nakajima H, Hosoya T. Autoantibodies to thyroid peroxidase in patients with chronic

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 January 2017. at 08:05 For personal use only. No other uses without permission. . All rights reserved.

TPO AND THYROID MICROSOMAL AUTOANTIBODIES

14.

15.

16. 17. 18. 19.

thyroiditis: effect of antibody binding on enzyme activities. Clin Exp Immunol. 1986;65:534-41. Czarnocka B, Ruf J, Ferrand M, Lissitzky S, Carayon P. Interaction of highly purified thyroid peroxidase with anti-microsomal antibodies in autoimmune thyroid diseases. J Endocrinol Invest. 1986;9:135-8. Ruf J, Czarnocka B, DeMicco C, Dutoit C, Ferrand M, Carayon P. Thyroid peroxidase is the organ-specific microsomal autoantigen involved in thyroid autoimmunity. Acta Endocrinol [Suppl] (Copenh). 1987;281:49-56. Kotani T, Ohtaki S. Characterization of thyroid follicular cell apical plasma membrane peroxidase using monoclonal antibody. Endocrinol Jpn. 1987;34:407-13. Nakajima Y, Howells RD, Pegg C, Davies Jones E, Rees Smith B. Structure-activity analysis of microsomal antigen/thyroid peroxidase. Mol Cell Endocrinol. 1987;53:15-23. Libert F, Ruel J, Ludgate M, et al. Thyroperoxidase, an autoantigen with a mosaic structure made of nuclear and mitochondrial gene modules. EMBO J. 1987;6:4193-6. Mariotti S, Anelli S, Ruf J, et al. Comparison of serum thyroid microsomal and thyroid peroxidase autoantibodies in thyroid dis-

773

ease. J Clin Endocrinol Metab. 1987;65:987-93. 20. Doble ND, Banga JP, Pope R, Lalor E, Kilduff P, McGregor AM. Autoantibodies to the thyroid microsomal/thyroid peroxidase antigen are polyclonal and directed to several distinct antigenic sites. Immunol. 1988;64:23-9. 21. Magnusson RP, Chazenbalk GD, Gestautas J, Seto P, Filetti S, DeGroot LJ, Rapoport B. Molecular cloning of the complementary deoxyribonucleic acid for human thyroid peroxidase. Mol Endocrinol. 1987;1:856-61. 22. Kimura S, Kotani T, McBride OW, et al. Human thyroid peroxidase: complete cDNA and protein sequence, chromosome mapping, and identification of two alternately spliced mRNAs. Proc Natl Acad Sci USA. 1987;84:5555-9. 23. Hamada N, Portmann L, DeGroot LJ. Characterization and isolation of thyroid microsomal antigen. J Clin Invest. 1987;79:81925. 24. Hamada N, Jaeduck N, Portmann L, Ito K, DeGroot LJ. Antibodies against denatured and reduced thyroid microsomal antigen in autoimmune thyroid disease. J Clin Endocrinol Metab. 1987;64:230-8. 25. Khoury EL, Bottazzo GF, Roitt IM. The thyroid "microsomal" antibody revisited. J Exp Med. 1984;159:577-91.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 January 2017. at 08:05 For personal use only. No other uses without permission. . All rights reserved.