Human monoclonal antibodies to thyroid antigens derived by hybridization of lymphocytes from a diabetic patient

Human Monoclonal

Antibodies to Thyroid Antigens Derived by Hybridization of Lymphocytes From a Diabetic Patient

K.S. Tan, C.S. Foster,

M. DeSilva,

P.G.H.

Byfield, A.R.

Medlen,

J.M.

Wright,

and V. Marks

Human monoclonal antibodies to human endocrine cells have been obtained following the generation of immunoglobulinsecreting interspecies lymphocyte hybridomas. Peripheral blood lymphocytes from an adult patient presenting with acute onset, Type I. diabetes mellitus were fused in vitro with mouse myeloma cells of the NSl cell line. Initial selection of resulting hybridomas was made by their ability to proliferate in HAT medium. Those hybridomas secreting human immunoglobulins were identified by radioimmunoassay and, thereafter, cloned at frequent intervals to ensure continued antibody production. Human monoclonal antibodies selected in this manner are being employed to identify those epitopes which are common antigenic targets during initial stages of autoimmune-mediated diabetes mellitus and associated multiple endocrinopathies. Of these antibodies, one (HML 3.22) recognizes an epitope present on the human TSH receptor and a second (HML 3.21) identifies a component of thyroglobulin. The potential value of human monoclonel antibodies as probes for analyzing autoimmune-mediated endocrine diseases is discussed. D 1987

A

by Grune

UTOIMMUNE

& Stratton,

Inc.

MECHANISMS

mediate

the patho-

genesis of many common human diseases. Some disorders, such as myasthenia gravis,’ Graves’ disease,2,3 and Hashimoto’s thyroiditis4 are characterized by a predominantly organ-specific autoimmune response. Conversely, in type I diabetes meIlitus,s.6 SjGgren’s syndrome, rheumatoid arthritis, and systemic lupus erythematosus’ the autoimmune response frequently involves autoantibodies to several different endocrine tissues. Unfortunately. because of the low titre and polyclonality of the involved antibodies, the identities of those antigenic determinants responsible for initiating tissue damage remain unknown. There is now firm evidence of a specific association between HLA subtype and the probability of developing diabetes mellitus.’ Singal et al9 demonstrated a correlation between antigens of the HLA-B locus and type I but not type II diabetes. Recently, associations have been observed between diabetes mellitus and antigens of classes DW3, DW4,” and with DR antigens.” A high degree of correlation has been shown between the heterozygous phenotype DR3/ Wh and diabetes mellitus associated with pancreatic isletcell antibodies of both IgG and complement-fixing variety.12 Nevertheless, evidence is now accumulating from rodent studies that the cell-mediated autoimmune destruction of pancreatic @-cells, which is characteristic of the early stage of diabetes mellitus,‘3.‘4 may be directed towards tissuespecific antigens which are independent of MHC determinants.15 As indicated by Lernmark,16 present knowledge of immunogens specifically associated with the onset of type I diabetes is fragmentary. There is now an urgent requirement for the involved antigens from different endocrine tissues to be identified and the pathogenetic importance of the immune response against these antigens to be determined. Autoimmune diseases affecting various combinations of endocrine tissues occur both independently and in conjunction with insulin-dependent diabetes melIitus.‘7*‘8 In primary autoimmune thyroid disease, antibodies binding to the endogenous TSH receptor and having inhibitory” or stimulatory20.2’ functions are recognized. Vafente et a12’ employed a panel of human monoclonal antibodies generated using lymphocytes from patients with Graves’ disease, and confirmed the presence of several distinct immunoreactive epitopes within the TSH receptor. Nevertheless, despite the Merabolism, Vol

36, No 4 (April). 1987: pp 327-334

occurrence of both hyperthyroid and hypothyroid disease in association with type I diabetes, no human monoclonal antibodies to the TSH receptor have yet been reported from diabetic patients. Since autoantibodies to components of thyroid tissue are frequently associated with acute onset diabetes mellitus, we have included methods for detecting human monoclonal antibodies to a variety of thyroid antigens as a routine component of our assay protocol. The long-term objective of our studies is to accumulate a panel of human monoclonal antibodies with which to test the hypothesis that, despite associations between HLA type and diabetes mellitus, cellular structures other than HLA determinants might be the initial mediators of autoimmune endocrine diseases. The approach has been to immortalize human lymphocytes activated early in type I diabetes mellitus by generating lymphocyte hybridomas secreting human monoclonal antibodies. In this manner we are attempting to identify those epitopes which become recognized during the acute phase of diabetes mellitus with associated autoantibodies to various endocrine tissues. We now report the successful generation of the first in an intended series of interspecies hybridomas secreting human monoclonal antibodies to human endocrine tissues from a patient with diabetes mellitus. We describe in detail the generation, selection, and characterization of human monoclonal antibody HML 3.22 which recognizes an epitope present on the human thyroid TSH receptor. A second human monoclonal antibody from this series (HML 3.21) identifies an epitope present on human thyroglobulin. The

From Guildhay Antisera Ltd, Departments of Biochemistry and Microbiology, University of Surrey, Guildford; Endocrinology Research Group, Clinical Research Centre. Harrow, Middlesex; and the Department of Histopathology, St Bartholomew’s Hospital, London, United Kingdom. Address reprint requests to KS. Tan, PhD, Guildhay Antisera Ltd, Department of Biochemistry. University of Surrey, Guildford. Surrey, GU2 SXH, United Kingdom; or to C.S. Foster. MD, PhD. MRCPath. Division of Biochemical Development and Molecular Diseases. Children’s Hospital of Philadelphia. 34th St & Civic Center Blvd. Rm 6157. Philadelphia, PA 19104. D I987 by Grune & Stratton, Inc. 00.?6-0495/87/3604-0005~03.00/0

327

328

TAN ET AL

potential value of this technique, relative to other methods of generating probes of autoimmune diseases, is discussed. [In the nomenclature used herein, “HM” refers to human monoclonal. The letter “L” is our reference code for the particular patient from whom the lymphocytes were obtained. The numbers refer to culture plates and wells from which the hybridoma clones were derived. Therefore, in this paper, “HML3.21” and “HML3.22” are equivalent to human monoclonal L3.21 and human monoclonal L3.22, respectively.] MATERIALS

AND METHODS

Generation of Hybridomas

Peripheral blood lymphocytes were obtained from a series of acute onset type I diabetic patients presenting with various combinations of complement-fixing islet-cell antibodies together with thyroid microsomal antibodies and antithyroglobulin antibodies. The patient from whom lymphocytes were obtained to perform the studies detailed herein was a 35year-old Caucasian male with acute onset (type I) diabetes mellitus. He had a strong family history of diabetes mellitus with associated thyroid disease. Initially, he was euthyroid, but has since developed clinical and biochemically confirmed hypothyroidism. His presenting auotimmune status is shown in Table 1. Lymphocytes were separated from peripheral blood by Ficollhypaque centrifugation. Isolated lymphocytes were either maintained in short-term (three to five days) liquid tissue culture (RPM1 medium containing 10% fresh autologous human serum at 37OC in an atmosphere of 10% CO2 in air and 100% humidity) or were directly fused with cells of murine myeloma line P3-NSl/ 1-AG4- 1 using poly-(ethylene glycol) 1500 (Sigma, St Louis) as previously described.*3 After fusion, cells were plated out into 48 x 2 mL culture wells containing Dulbecco’s MEM and 20% fetal calf serum (Gibco, Paisley, Scotland) in the presence of 1 x 10’ mL_’ murine thymocytes as feeder cells. Cultures were maintained at 37“C in an atmosphere of 10% CO2 in air and 100% humidity. Eighteen days after fusion, culture supernatants from wells containing visually identifiable hybridoma colonies were tested for the presence of human IgG and IgM. Fifty-microliter aliquots of culture supernatants were assayed by radioimmunoassay using affinity-purified sheep anti-(human immunoglobulin) obtained from Guildhay Antisera (Guildford, United Kingdom). Assays were performed at 4“C for two hours. Precipitates were formed using poly-(ethylene glycol) 6,000 at a final concentration of 2% (w/v) for two hours. Hybridomas secreting human immunoglobulins were cloned by limiting dilution. Only those cultures showing vigorous proliferation were rescreened by identical radioimmunoassay after a further period of 14 days culture. Thereafter, hybridomas continuing to secrete high titres of human immunoglobulins were recloned by limiting dilution at regular intervals. Table 1. Autoimmune

Status at Presentation

Whom Human Monoclonal

of Patient From

Antibody HML 3.22 Was Derived

AntibodyType

Titre

Anti-islet cell (complement fixing)

1:5 120 (normal < 1:40)

Anti-islet cell (IgG)

1:5 120 (normal < 1:40)

Anti-thyroid microsomal

1:640 (normal < 1:40)

Anti-mitochondrial

Negative

Anti-nuclear

Negative

Anti-gastric parietal cell

Negative

Tanned red cell anti-thyroglobulin Sheep cell agglutination test for rheumatoid factor

1:40 (normal < 1:40) Negative

Characterization of Monoclonal Antibodies Human monoclonal antibodies were characterized by the Ouchterlony precipitin technique using cellulose acetate substrates impregnated with barbitone buffer (60 mmol/L, pH 8.6) after the method of Kohn et aLz4 Duplicates of tissue-culture supernatants were set up against affinity-purified sheep anti-(human IgG) and sheep anti-(mouse IgG) obtained from Guildhay Antisera, (Guildford, United Kingdom) and left for 48 hours in a humid chamber at 4°C. After washing in 3% (w/v) NaCl containing 0.1% (v/v) Triton 405, precipitin lines were visualized by staining the cellulose acetate in 0.2% saturated Nigrosin solution in 5% (v/v) acetic acid for ten minutes. Destaining was performed in tap water. Cellulose acetate immunofixation was employed to identify the specific class of human monoclonal antibodies. One-microliter aliquots of tissue-culture supernatants were applied to cellulose acetate strips previously impregnated with barbitone buffer. Electrophoresis was performed in a humid chamber at 120 volts for one hour. The strips were then impregnated with rabbit anti-(light chain specific) antisera (DAKO, Copenhagen) prior to washing and staining as described above. Immunoglobulins secreted by human-mouse hybridoma cells were metabolically labeled by culture with 4 &i x mL_’ “C-leucine (Amersham, United Kingdom) for 12 hours in leucine-free Dulbecco’s MEM supplemented with 10% dialyzed and heat-inactivated fetal calf serum (Gibco). After labeling, culture supernatants were aspirated from the cells and saved. Cells were washed twice with ice-cold isotonic sodium phosphate-buffered saline prior to lysis in a buffer containing 10 mmol/L sodium phosphate (pH 7.4) 0.1 mol/L NaCl, 1% Triton X-100, 0.5% (w/v) sodium deoxycholate, 0.1% (w/v) SDS, and 10 mmol/L benzamidine. Cell lysates were centrifuged at 100,000 g for 60 minutes at 4OC and supernatants were saved. Antibodies were precipiated from culture supernatants and from cell lysates using affinity-purified donkey anti-(human immunoglobulin) (Guildhay) and poly-(ethylene glycol) 6,000 (final concentration 2%) for two hours. Precipitated antibodies were dissolved in reducing buffer containing 10 mmol/L Tris-HC1 (pH 6.8) 2%-SDS, and 5% 2-mercaptoethanol. Reduced samples were boiled for approximately one minute and separated on a 10% polyacrylamide gel according to the method of Laemmli.‘s Following electrophoresis, gels were fixed, impregnated with EN’HANCE (New England Nuclear; Stevenage, UK), dried, and fluorographed at -80°C for periods of between seven days and one month. Karyotyping Studies Karyotyping of hybridoma cells was performed by conventional technique. Cells were cultured to 5 x IO6mL_’ in Dulbecco’s MEM supplemented with 20% fetal calf serum (Gibco) at 37°C and in an atmosphere of 10% CO2 in air and 100% humidity. To inhibit cell division in metaphase, culture medium was replaced by Liebowitz medium L15 containing 10% fetal calf serum and 0.2% colchicine for 12 hours. Thereafter, cells were maintained in air at 37OCand an atmosphere of 100% humidity. Cells were then harvested and chromosome “spreads” were obtained by allowing drops of medium containing cells to fall onto clean glass microscope slides. After drying and “aging” for 2 weeks, spread chromosomes were banded by digestion with 0.05% trypsin for 5 to 10 seconds prior to staining with 0.3% aqueous Giemsa for four minutes. Immunohistochemistry Immunohistochemical studies were employed to establish the tissue specificities of the human monoclonal antibodies. A spectrum of fresh-frozen and formalin-fixed human tissues (both endocrine and nonendocrine) were submitted to immunohistochemical stain-

329

HUMAN MONOCLONAL ANTIBODIES TO THYROID ANTIGENS

ing. Human thyroid tissues were obtained fresh at surgical thyroidectomies performed for investigation of nodular hyperplasia. Tissues from patients with known thyrotoxicosis or thyroid malignancy were not employed. All other tissues were fresh surgical specimens. Autopsy tissues were not obtained. Tissue sections were incubated overnight at 4OCwith hybridoma culture supernatants concentrated tenfold using Amicon filters. After washing, human immunoglobulins were detected by incubation of sections with FITC-conjugated, affinity-purified, class-specific antisera (Sera Laboratories, Crawley Down, Sussex, United Kingdom) or by enzyme immunohistochemistry as previously described.26 lmmunohistochemical controls comprised two adjacent tissue sections. One of these was overlayered with 10% fetal calf serum in culture medium, instead of hybridoma supcrnatant. The other section was incubated directly with the secondary antiserum.

Inhibition

of‘25L!Jptake

Studies

TSH-stimulated iodide uptake into FRTL-5 cells was assessed in the presence of monoclonal antibody HML 3.22. FRTL-5 cells were plated in 24-well culture dishes (2.5 x IO’ cells/ml) and maintained for three days in Coon’s modified Ham’s F- 12 M medium containing a supplement of six hormones (6H) including TSH as described previously.” The medium was then replaced by 5H medium (without TSH) and culture continued for a further seven days. Aliquots (0.5 mL) of tissue-culture supernatants containing monoclonal antibody HML 3.22 diluted I:10 and 1:50 in 5H medium and containing increasing concentrations of TSH were added to the FRTL-5 cells and incubated for 48 hours. Thereafter, the medium was aspirated, replaced by fresh 5H medium and the cells incubated for a further 24 hours. Iodide uptake was then determined as described by Valente et al? 5H medium was aspirated and replaced by 0.5 mt HBSS medium containing 0.1 &i carrier free Na’*? and 10 pmol Nal/L to give a specific activity of 20 mCi/mmol. After 40 minutes at 37OC, the reaction was terminated by aspiration and the cells washed with 1 mL ice-cold HBSS. To assess the amount of ‘“‘I taken up by the cells, 0.5 mL 95% ethanol was added to each well, incubated for 20 minutes, and then transferred to scintillation vials to determine the contained radioactivity.

concentration of 15%. After mixing, tubes were centrifuged at 15,000 g for 60 min at 4OC and receptor-bound ‘*‘I-TSH in the pellets was measured. RESULTS

Approximately 200 hybridoma colonies were visualized within the cultures 18 days after fusing I x IO’ peripheral blood lyphocytes with 1 x 10s NSI hybridoma cells. Radioimmunoassay of culture supernatants from wells

containing hybridoma colonies identified the presence of secreted human IgG or IgM immunoglobulins. After serial cloning individual colonies by limiting dilution, 19 cloned hybridomas were selected for detailed investigation. Monoclonal antibodies from two of these 19 hybridomas reacted with human thyroid tissue, and were studied first. Hybrid clones HML 3.21 and 3.22 were shown by radioimmunoassay to secrete human IgG immunoglobulin. This finding was confirmed by radial diffusion immunoprecipitation (Ouchterlony) technique performed on celluloseacetate substrates. No cross-reactivity with anti-(mouse IgG) was observed. Immunofixation studies showed both human IgGs to be of K light-chain subclass. Following immunoprecipitation of proteins from cell lysates and culture supernatants of hybridoma cells metabolically labeled with “C-leucine, polyacrylamide gel electrophoresis of the immune precipitates was performed under mild reducing conditions. Autoradiography of the gels revealed three distinct protein bands precipitated from each hybridoma. The L 3.2r

L3.22

NS-1

4 200Kd 4

TSH Receptor Assay An essay for TSH receptor antibodies, based on the TRAb assay described by Southgate et al,** was employed to confirm the specific binding of antibody HML 3.22 to the TSH receptor. Antibody was first prepared from tissue culture supernatants as described by Eisenbarth et al.“9 After precipitation of antibody by addition of an equal volume of saturated (NH&SO, to the medium, the precipitate was suspended in %oth its original volume of 20 mmol/L sodium phosphate-buffered saline (pH 7.5) (PBS) and then dialyzed against PBS. To provide a control, medium from cultures of human-mouse hybridoma HML 3.1 was processed in a similar manner. Dilutions of the dialyzed antibody solutions were assayed, in duplicate, for TSH receptor binding activity. Aliquots of antibody, diluted to 50 rL with PBS, were incubated for 30 minutes at 20°C with 100 rL of detergent-solubilized porcine TSH receptor” in 10 mmol/L TrisHCI buffer (pH 7.5) containing 50 mmol/L NaCl and 0.1% (w/v) BSA. Bovine TSH (Thytropar, Armour Pharmaceuticals; Eastbourne, Sussex, UK) labeled with ‘ZsI-icdine (5,000 cpm, 50 gL) using chloramine-T and diluted in Tris-NaCl-BSA buffer was then added. The solutions were mixed and incubation continued overnight at 20°C. Incubation mixtures were then diluted to 1 mL by addition of 800 FL TSH-NaCI-BSA. Receptor-hormone and receptorantibody complexes were precipitated by addition of 1 mL 30% (w/v) poly (ethylene glycol) (4000) in I mol/L NaCl to give a final

100

4

69

4

46

4

30

1,,

Fig 1. Autoradiography of “C-leucine-labeled immunoprecipitates from hybridoma supernatants separated by SOS-lo% (w/v) polyacrylamide gel electrophoresis. lmmunoprecipiteted proteins separated with approximate relative mobilities of 158 kD, 52 kD, and 20 kD corresponding to intact IgG, and human H and L chains, respectively. No immune precipitates were obtained from the supernatant or from the cell extracts of the NS-1 cell line.

330

bands were of approximate relative mobilities (Mr) 158 kD, 52 kD, and 20 kD, corresponding to intact human IgG, and human heavy and light chains, respectively (Fig 1). No immunoprecipitated bands were observed in the NSl control lanes. Karotyping studies of the NSl cell line showed chromosomes which were predominantly acrocentric and thus characteristic of a murine origin. However, examination of the karyotypes of hybridoma cells HML3.21 and HML3.22 (Fig 2) revealed the consistent presence of several additional chromosomes which were metacentric and submetacentric. These latter are typical of a human origin and included, among others, the immunoglobulin-associated chromosomes 2, 14, and 22. Immunohistochemical studies (Table 2) indicated monoclonal antibody HML 3.22 to bind strongly to basolateral plasma membranes of frozen sections of thyroid follicular epithelial cells (Fig 3) but not to formalin-fixed, paraffin wax-embedded thyroid or to any other tissues. Monoclonal

TAN ET AL

antibody HML 3.21 bound to colloid within thyroid follicles, but to no other tissue studied. Assays to determine the uptake of ‘251-iodine by FRTL-5 cells indicated that monoclonal antibody HML 3.22 specifically inhibited the stimulatory effect of TSH (Fig 4). Following 48 hours incubation, a I:10 dilution of this monoclonal antibody blocked TSH-stimulated uptake of iodine at concentrations of up to 100 KU x mL_’ bovine TSH. Identical cultures incubated with control media showed a dosedependent response to TSH. At a higher dilution (I 50) the antibody continued to block TSH-stimulated iodine uptake, although this inhibition could be reversed by higher concentrations of TSH with significant (P < .002) increase in iodine uptake. The TRAb assay (Fig 5) showed increasing specific precipitation of lZSI-TSH with increasing amounts of dialyzed monoclonal antibody HML 3.22. Control antibody HML 3.21, which bound to thyroglobulin but which was not considered to be directed to a determinant on the TSH

Appearance of a typical chromosome “spread” from hybridoma HML 3.22. The majority of such preparations from the resultant Fig 2. stable clone contained 82 chromosomes. Karyotype analysis revealed the consistent inclusion of 15 metacentric and submetecentric chromosomes characteristic of human origin, together with a majority of typical murine chromosomes. Those human chromosomes which could be identified unequivocally have, wherever possible, been assigned numbers.

HUMAN

MONOCLONAL

ANTIBODIES

TO

THYROID

331

ANTIGENS

indirect immunofluorescence photomicrograph of paraffin-wax embedded section of human thyroid follicle stained with human Fig 3. monoclonal antibody HML 3.22. Bound immunoglobulin was visualized using FITC-conjugated rabbit anti-(human IgGj. In the tissue sections, observed staining varied in intensity between follicles. Some follicles were not stained, and frequently the staining was not bright. Nevertheless, the pattern of staining was constant in distribution to the basolateral plasma membranes of the follicular epithelial cells. Staining of nonepithelial structures was not seen. The basolateral plasma membranes of the follicular epithelial cells illustrated herein are particularly clearly depicted (x55Bj.

Table 2. lmmunohistochemical Staining of Normal Human Tissues Using Monoclonal Antibody HML 3.22

Tissue

FKTZf3l

Paraffin Wax

Section

Section

Endocrine tissues Thyroid +++

+

Colloid

Epithelial

cells

-

_

Stroma

-

_

_

_

Pancreas Islet cells Exocrine

cells

Stroma

_

_

_

_

Adrenal Cortex

_

Medulla

_

_

_

_

_

_

Testis Germ

cells

Leydig

cells

Pituitary Adenohypophysis

_

_

Neurohypophysis

_

_

_

_

Nonendocrine Resting

tissues breast

Stomach

-

_

Duodenum

_

_

Colon

_

_

receptor, showed no precipitation outside the range obtained with normal serum samples and thus regarded as representing the “background” for this method. DISCUSSION

Human monoclonal antibodies are potentially valuable reagents for identifying cellular antigens recognized during the acute stages of human autoimmune diseases and hence for determining those epitopes which are of pathogenetic importance. Herein we report the successful generation of a panel of human monoclonal antibodies following fusion of peripheral blood lymphocytes obtained from a patient with acute onset diabetes mellitus, thyroid autoantibodies, and hypothyroidism. Two of the monoclonal antibodies identify human thyroid antigens but no other endocrine tissue examined. One (HML 3.22) has been shown to identify an epitope present on the TSH receptor of thyroid epithelium. The other (HML 3.21) identifies a com,ponent of thyroglobulin. Precise events involved in the pathogenesis of insulindependent diabetes mellitus remain uncertain.r6,” Current hypotheses range from cross-reactivity between islet beta cells and an environmental antigen to a subtle primary abnormality in the immune systems of susceptible individuals which leads to autoimmune destruction of islet cells.32*33 In this disease it is likely that a restricted but highly specific and susceptible group of cell-surface determinants become recognized by the immune system and initiate subsequent pathogenetic mechanisms. However, the etiology of such recognition is presently unknown.

332

TAN ET AL

0.4

U

HML 3.22

o- ---- -0

Control

100 bTSH

()IUml-‘)

bTSH(pUml“)

Fig 4. Specific inhibition of TSH-stimulated uptake of ‘?by FRTL-5 cells in the presence of tissue-culture supernatant medium containing human monoclonal antibody HML 3.22 at dilution 190 and 150. Culture medium containing HML 3.21 was employed as the antibody control (not shown) and gave levels of “‘l-uptake by FRTL-5 cells identical to those of the medium control without antibody. In the presence of HML 3.22, the maximum uptake of “‘l-iodine obtained was less than 10% that of the control.

100

200

300

400

500

iodine

The pathogenesis of multiple autoimmune endocrinopathies associated with type I diabetes mellitus and simultaneousty affecting several different endocrine tissues is even more obscure. One possibility is that some cellular determinants, important in the pathogenesis, are expressed by a variety of endocrine organs, while other determinants appear organ-specific. However, mere exposure of structural or functional components of cells to the immune system following damage by viruses or otherwise is, of itself, an insufficient stimulus to induce a sustained autoimmune attack.34 Therefore, progressive autoimmune disease, simultaneously affecting several different endocrine organs, is not a nonspecific event. Rather, the pathogenesis appears to be precise and to involve a select group of cellular determinants. The findings of the present study now raise the extremely important question of the relevance to pathogenesis of those tissue epitopes recognized by a patient’s immune system and to which human monoclonal antibodies may be selected. Unfortunately, the frequency of autoantibodies to nonpathological tissues occurring in the normal population is presently unknown. Only since the development of techniques for generating human monoclonal antibodies has it been possible to probe the repertoire of epitopes recognized by an individual human’s immune system. Nevertheless, identification of antibodies to a particular antigen, in a single patient, is insifficient evidence to ‘ascertain that the antigen is of pathogenetic importance in a specific autoimmune endocrine disease. Satoh et at3’ showed several human monoclonal antibodies produced from patients with multiple autoimmune endocrinopathies to cross-react with antigens present on a variety of endocrine organs apparently unaffected by the primary autoimmune disease. Hawkins et al36 examined a randomly selected population of supposedly healthy subjects and identified the occurrence of thyroid microsomal antibod-

Equivalent volume of original hybridoma culture medium (pl) Fig 5. Specific immunoprecipitation of ‘asl-labeled TSH by human monoclonal antibody HML 3.22. Values for the control antibody (HML 3.21) fall within the normal range (0 to 0.1 mU x mL_’ TSH) for normal human serum. The points shown are the mean values of duplicate assays.

ies to be a specific and sensitive indicator of previously undetected autoimmune hypothyroidism. However, Riley et a13’ demonstrated the incidence of thyroglobulin antibodies in insulin-dependent diabetes mellitus to be identical to the normal population (1.6%), and thus of little immunological or clinical relevance. Conversely (and in the same group of young diabetics), Riley et al” confirmed and emphasized a 17.6% incidence of thyroid microsomal antibodies to be highly significant indicator of thyroid insufficiency, whether or not clinically overt. At present, too few data have been published to define the pathological significance of antibodies to other determinants (such as the TSH receptor) occurring in the normal population, or in patients with autoimmune-associated diabetes mellitus. Therefore, from the data reported herein, we cannot conclude that an anti-TSH receptor antibody is of pathogenetic significance in the etiology of thyroid disease associated with diabetes mellitus, despite this patient now being hypothyroid. However, should additional anti-TSH monoclonal antibodies be generated from other patients with acute onset diabetes mellitus, then the present association will become strengthened, particularly if these patients also proceed to thyroid insufficiency. Furthermore, comparison of antibody HML 3.22 with other reagents will determine whether all those epitopes recognized on the TSH receptor are identical or different. Although we have successfully produced several human monoclonal antibodies, the present technique is inefficient” and probably accounts for the absence of anti-TMA antibodies in the reported study. During initial stages of the cultures, some difficulty was experienced in maintaining immunoglobulin production by many of the hybridomas. Presumably, this was due to instability of the hybrid genome, with

333

HUMAN MONOCLONAL ANTIBODIES TO THYROID ANTIGENS

selective segregation of human chromosomes, a problem associated with interspecies hybridomas of this type. Nevertheless, following repeated single-cell cloning, the remaining human x mouse hybridomas have been cultured for more than one year and are presently stable. Since no human myeloma or plasmacytoma cell line is available which reliably forms hybridomas with human lymphocytes and produces human monoclonal antibodies in large amounts,39 we are currently

investigating

other

techniques

(including

non-

specific EB viral transformation,40 specific transfection with c-myc

oncogene,4’

immortalizing is to develop

and

activated a

protein human

synthesis lymphocytes.

inhibitors42) Our

for

intention

more efficient system for producing human

monoclonal antibodies to a greater range of cellular components recognized in diabetes mellitus and associated autoimmune endocrine diseases. ACKNOWLEDGMENT The authors

thank

Professor

Joachim

Kohn and Julian

C. RayDr Mike Butler and Professor Ray Spier allowed us free access to their excellent tissue culture facilities. We are grateful to Laurie Smith, Severalls Hospital, Colchester, United Kingdom and to Professor

mond for the immunotixation and Ouchterlony assays.

Beverley Emmanuel, Children’s Hospital, Philadelphia, for assistance with the karyotyping procedures and with their subsequent

interpretation, respectively.

REFERENCES I. Linstrom J: Autoimmune response to acetylcholine receptors in myasthenia gravis and its animal model. Adv Immunol 27:1-50, 1979 2. Doniach D: Humoral and genetic aspects of thyroid autoimmunity, in Irvine WJ (ed): Autoimmunity in Endocrine Disease. Clin Endocrinol Metab 4:267-285, 1975 3. McKenzie JM, Jakarija M: A reconsideration of a thyroidstimulating immunoglobulin as the cause of hyperthyroidism in Graves’ disease. J Clin Endocrinol Metab 42:778-78 1, 1976 4. Kidd A, Okita N, Row VV, et al: Immunologic aspects of Graves’ and Hashimoto’s diseases. Metabolism 29:80-99. 1980 5. Lendrum R, Walker G, Cudworth AC, et al: Islet-cell antibodies m diabetes mellitus. Lancet 2: l273- 1276, 1976 6. Irvine WJ, McCallum CJ, Gray RS, et al: Pancreatic islet-cell antibodies in diabetes mellitus correlated with the duration and type of diabetes, coexistent autoimmune disease, and HLA type. Diabetes 26:138-147. 1977 7. Dons RF, Havlik R. Taylor SI, et al: Clinical disorders associated with autoantibodies to insulin receptor. Simulation by passive transfer to rats. J Clin Invest 75:89-99, 1983 8. Bertrams J: The HLA association of insulin-dependent (type 1) diabetes mellitus. Behring Inst Mitt 75:89-99, 1984 9. Singal DP, Blajchman MA: Histocompatibility (HL-A) antigens, lymphocytotoxic antibodies and tissue antibodies in patients with diabetes mellitus. Diabetes 22:429-432, 1973 10. Svejgaard A, Platz P, Ryder LP, et al: HL-A and disease associations: A survey. Transplant Rev 22:3-43, 1975 11. Svejgaard A, Platz P, Ryder LP: Joint report: Insulindependent diabetes mellitus, in Terasaki PI (ed): Histocompatibility Testing. Los Angeles, University of California, 1980, pp 638-656 12. Glichmann H, Zorcher B, Greulich B, et al: Correlation of islet-cell antibodies and HLA-DR phenotypes with diabetes mellitus in adults. Diabetologia 27:90-92, 1984 13. Naji A, Silvers WK, Bellgrau D, et al: Prevention of diabetes in rats by bone marrow transplantation. Ann Surg 194:328-338, 1981 14. Like AA, Butler L, Williams RM, et al: Spontaneous autoimmune diabetes mellitus in the BB rat. Diabetes 31:7-13, 1982 (suppl 1) 15. Weringer EJ. Like AA: Immune attack on pancreatic islet transplants in the spontaneously diabetic biobreeding/Worcester (BB/W) rat is not MHC restricted. J Immunol 134:2383-2386, 1985 16. Lernmark A: Cell-mediated immunity in Type 1 (insulindependent) diabetes: Update 84, in Andreani D, DiMario U, Federlin KF. et al (eds): Immunology in Diabetes. London, Kimpton, 1984, pp 121-131 17. Gorsuch AN, Dean BM, Bottazzo GF, et al: Evidence that

type 1 diabetes and thyrogastric autoimmunity have different genetic determinants. Br Med J 280: 145-147. 1980 18. Maclaren NK, Riley WJ: Thyroid, gastric and adrenal autoimmunities associated with insulin-dependent diabetes mellitus. Diabetes Care 8:34-38. 1985 (suppl 1) 19. Drexhage HA, Bottazzo GF, Bitensky L, et al: Thyroid growth-blocking antibodies in primary myxoedema. Nature 289:594-596, I98 1 20. Rees-Smith B, Hall R: Thyroid-stimulating immunoglobulins in Graves’ disease. Lancet 2:427-43 I, 1974 21. Clague R, Mukhtar ED, Pyle GA, et al: Thyroid-stimulating immunoglobulins and the control of thyroid function. J Clin Endocrinol Metab 43:550-556, 1976 22. Valente WA, Vitti P, Yavin Z, et al: Monoclonal antibodies to the thyrotropin receptor: Stimulating and blocking antibodies derived from the lymphocytes of patients with Graves’ disease. Proc Natl Acad Sci USA 79:6680-6684, 1982 23. Foster CS: Lymphocyte hybridomas. Cancer Treat Rev 9:5984, 1982 24. Kohn J, Riches JG, Raymond JC: The use of alkalinephosphatase conjugated second antibody for the enhancement of immunoprecipitation techniques on cellulose acetate membranes. J lmmunol Methods 76:1 l-16, 1985 25. Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophase T,. Nature 227:680-685. 1970 26. Foster CS, Edwards PAW, Dinsdale EA. et al: Monoclonal antibodies to the human mammary gland: (1) Distribution of determinants in non-neoplastic mammary and extra-mammary tissues. Virchows Arch 394:279-293, 1982 27. Ambesi-Impiombato FS, Parks LAM, Coon MG: Culture of hormone-dependent functional epithelial cells from rat thyroids. Proc Natl Acad Sci USA 77:3455-3459, 1980 28. Southgate K, Creagh F, Teece M, et al: A receptor assay for the measurement of TSH receptor antibodies in unextracted serum. Clin Endocrinol 20:539-543, 1984 29. Eisenbarth GS, Linnenbach A, Jackson R. et al: Human hybridomas secreting anti-islet autoantibodies. Nature 300:264-267, 1982 30. Rees-Smith B, Hall R: Measurement of thyrotropin receptor antibodies. Methods Enzymol 74:405-420, 198 1 31. Andreani D: Immunology in Diabetes: Introduction and Overview, in Andreani D, DiMario U, Federlin KF, et al (eds): Immunology in Diabetes. London, Kimpton, 1984, pp I-20 32. Cahill GF, McDevitt HO: Insulin-dependent diabetes mellitus: The initial lesion. N Engl J Med 304:1454-1465, 1981 33. Ptak W, Rewicka M, Ptak M, et al: Deficiency of contrasuppressor lymphocytes in alloxan-diabetic mice. 38:13-21, 1986

334 34. Wall JR, Fang S-L, Ingbar SH, et al: Lymphocyte transformation in response to human thyroid extract in patients with subacute thyroidits. J Clin Endocrinol Metab 43587-590, 1976 35. Satoh J, Prabhakar BS, Haspel MV, et al: Human monoclonal autoantibodies that react with multiple endocrine organs. N Engl J Med 309:217-220, 1983 36. Hawkins BR, Dawkins RL, Burger HG, et al: Diagnostic significance of thyroid microsomal antibodies in randomly selected population. Lancet 2:1057-1059, 1980 37. Riley WJ, Maclaren NK, Lezotte DC, et al: Thyroid autoimmunity in insulin-dependent diabetes mellitus: The case for routine screening. J Pediatr 98:350-354, 1981 38. Riley W, Maclaren N, Rosembloom A: Thyroid disease in young diabetics. Lancet 2:489-490, 1982

TAN ET AL

39. Boyd JE, James K, McClelland DBL: Human monoclonal antibodies-production and potential. Trends Biotech 2:70-76, 1984 40. Katamine S, Otsu M, Tada K, et al: Epstein-Barr virus transforms precursor B cells even before immunoglobulin gene rearrangements. Nature 309:369-372, 1984 41. Mougneau E, Lemieux L, Rassoulzadegan M, et al: Biological activities of v-myc and rearranged c-myc oncogenes in rat fibroblast cells in culture. Proc Nat1 Acad Sci USA 81:5758-5762, 1984 42. Kobayashi Y, Asada M, Higuchi M, et al: Human T cell hybridomas producing lymphokines. I. Establishment and characterization of human T-cell hybridomas producing lymphotoxin and migration inhibitory factor. J Immunol 128:27 14-27 18, 1982