J. Endocrinol. Invest. 13: 951-968, 1990
REVIEW ARTICLE
Antibodies directed to th~ insulin receptor. Clinical aspects and applications to the study of insulin action R. De Pirro, P Borboni, M.A. Marini, A. Montemurro, G. Sesti, and R. Lauro Cattedra di Endocrinologia, Universita di Ancona and Cattedra di Endocrinologia a Medicina Costituzionale, 11 Universita di Roma "Tor Vergata", Roma, Ifaly
INTRODUCTION
understanding; furthermore, they are present in some peculiar cases of altered glucose metabolism in humans.
The insulin receptor is a tetrameric glycoprotein composed by two 0:- and two ß-subunits joined together via disulfide bonds (1). The former are located extracellularly whereas the latter are exposed both to the extracellular and to the intracellular environment (Fig. 1, inset). The aminoacid receptor sequence 155-312 shows the peculiar presence of 21 cysteine residues thus suggesting that it is the insulin binding domain (1). The ß-subunit contains various tyrosine residues whic.h are candidates as primary regulatory site(s) of autophosphorylation (Fig. 1, inset). There are two insulin receptor precursors, one of 1382 and the other of 1370 aminoacids (2, 3). Insulin exerts effects taking place within seconds to minutes (e.g. cell membrane stimulation of transport), within minutes to hours (e.g. activation/inhibition of intracellular enzymes) or within hours to days (e.g. RNA/DNA synthesis). Next, soon after interaction with ,insulin, receptor molecules cluster on plasma membrane, are internalized in the cell and, finally, they are recycled back to the cell surface (Fig. 1). At the present time it is not clear whether the receptor would in toto participate in exerting these effects or peculiar receptor domains are involved in exerting each/some of those; furthermore, biological phenomena coupling receptor activity to final insulin effects are not weil defined (Fig. 1). Antibodies directed to the insulin receptor led, and still are leading, to important improvements in our
PRODUCTION OF ANTI-INSULIN RECEPTOR ANTIBODIES In 1975 Flier et al. (4) described the presence of circulating autoantibodies to the insulin receptor in three female patients; those autoantibodies were able to inhibit the binding of labeled insulin to the receptor (insulin binding), From that report we suggested that anti-insulin receptor antibodies would be a new, important tool to study insulin action. The first goal was to obtain an amount of purified insulin receptors sufficient to induce an immune response in animals. It is noteworthy that before 1979 receptor characteristics were largely unknown; there was only evidence that insulin binds to a large glycoprotein (1). Several rabbits were injected with receptor partially purified from human placenta (5, 6) but animals did not produce antibodies able to inhibit insulin binding like those described in patients (4). Thereafter, the hypothesis was raised that the insulin binding site has minor immunogenicity so that animals were indeed producing antibodies against other receptor domains. At that time inhibition of insulin binding was the only reported method to show the presence of ahtibodies. By using columns of agarose-coupled IgGs from rabbits it was possible to demonstrate that one rabbit produced antibodies directed to receptor region(s) not involved in insulin binding (Fig. 2) (6). In the meantime another research group reported on one animal producing antibodies with insulin-like effect but unable to inhibit insulin binding (7). In that period of time some studies (8, 9) raised the possibility that animals did not produce antagonistic antibodies since the insulin binding region is iden-
Key-words: Autoantibodies, insulin receptor, insulin action, extreme insulin resistance, anti insulin-receptor antibodies. Correspondence: Roberto Oe Pirro, MD., Clinica di Endocrinologia, Ospedale Torrette, 60100 Ancona, Italy.
951
R. Oe Pirro, P. Borboni, M.A Marini, et aJ.
tieal among various speeies. Aetually, it is elear that the eritieal problem is that animals were injeeted with not suffieiently purified insulin reeeptors (6). In 1979 Pileh and Czeeh (10) elueidated that the reeeptor is eomposed by a- and ß-subunits and Jaeobs et al. (11) produeed a polyelonal antibody by preparing from rat liver suffieient amounts of pure reeeptor a-subunit. Like polyelonal autoantibodies present in patients (see after) that antibody inhibited insulin binding but exerted also insulin-like effeets thus pointing out that polyclonals are not suitable to study insulin action. In 1981 Roth et al. (12) produeed the first monoelonal antibody (MA-51) whieh showed the peeuliar effeet to inhibit both insulin binding and insulin action (13). Subsequently, MA-51 was used by Montemurro et al. to prepare affinity eolumns with whieh large amounts of insulin reeeptors were purified to homogeneity and used to produee three new monoelonal antibodies (MA-5, MA-10, MA-20) (14). Few other attempts to produee monoelonal antibodies reaehed positive results (15-17). By using the notion that ehieken stores gammaglobulins inside the egg YOlk, Song et al. (18) injeeted hens with purified rat insulin a-subunit and sueeeeded in produeing polyelonal antibodies. One important point is that heqs and rabbits were injeet-
• tl
ed with the same material but no rabbit produeed antibodies (18). Unfortunately, this method eannot be used to produee monoelonal antibodies sinee at the present time ehieken myeloma eell lines are not available.
THE AUTOANTIBODY-INDUCED DISEASE STATE The speeifie binding of [125-1] Insulin to monoeytes from three women suffering from aeanthosis nigrieans was only 5-30% of normal due to autoantibodies able to inhibit the binding of insulin to receptors; two of them were diabetic and severely resistant to insulin (6,000 and 24,000 U / day, respectiveIy), the third showed impaired glucose tolerance and was resistant to insulin (0.1 U/Kg) (4). Later, the same authors (19), as weil as others (20-28), reported on a few other diabetie patients with extreme insulin resistance due to receptor autoantibodies. Several evidences indicate that insulin resistanee is provoked by autoantibodies (29-32). Ten years ago, autoantibodies to the insulin receptor were identified in a 46-year old female patient (1-1; Italy-1) with recurrent hypoglycemia and lupus nephritis. In fact the patient cells were unable to bind insulin (Fig. 3) on account of gammaglobulins of IgG class (Table 1) inhibiting the binding of insulin to receptors (Fig. 4) (34). As al ready mentioned,
INSULIN
s-sfi-+-BINDING SITE
~
r---
=
1~~ ß 1316
1322
rD.;). D
s-s
Deaetivation'lr
s-s
r---
ß
1018 Lys
1 P .... Pospho-dephosphorylation , reaetlons \ Aetivation ? ' - . . Metabolie )\.. end-points Mediator(s) ? )
9b Mierosomes
=
_
~
• ~ .../ Glucose Translocation ransporter
~
~
I t
Fig. 1 - Schematic representation of both Insulin
Protein i Regulatory KInases and enzymes Phosphatases ~
952
t J.
receptor structure and post-receptor events. There are two insulin receptor precursors. one has 1382 aminoacids (2) whereas the other has 1370 aminoacids (3), they both have the ß-subunit composed by 620 aminoacids (2, 3).
Antibodies to the insulin receptor
treatment a prompt amelioration of both insulin binding to cells and glycemia was observed (34); therefore, it was concluded that hypoglycemia was induced by anti-insulin receptor autoantibodies. The patient was healthy for aboutsix months; thereafter, she displayed recurrent hypoglycemia in the presence of autoantibodies. A new plasma exchange treatment impraved the clinical situation for some months but few days after a third plasma exchange treatment she was found comatose in bed and died. Autopsy was totally negative out of signs of lupus nephritis. At the present time, literature re ports some cases of hypoglycemia in the presence of autoantibodies to the insulin receptor (35-39) so that these autoantibodies are now classified as one of the ethiologic factors of hypoglycemia. In compari90n to patients with autoantibodies and diabetes mellitus, these patients do not have clinical peculiarities out of different glycemia; one important point to be emphasized is that hypoglycemia is a really dangeraus event since mortality rate is tremendously high. Paradigmatic is the case of two patients who died when their metabolism varied fram hyper- to hypoglycemia (25, 40).
such autoantibodies were described before in diabetic patients but not in the presence of hypoglycemia. 80th oral glucose tolerance and tolbutamide tests offered results suggesting the presence of insulin resistance since an exaggerated insulin secretion in the absence of any appreciable variation in blood glucose was present (Fig. 5). The patient insu'iin resistance was finally del'l1onstrated thraugh the intravenous infusion of insulin (0.4 U/kg in.1 h) 3 h after the patient's last meal (Fig. 6). The patient received one plasma exchange. Four days after the
20
<::
.,
.o
~
t
u..
10
><
E a.
~
8
o
5
15
25
Fractions
Fig. 2 - IgGs from a rabbit immunized against partially purified human insulin receptors were bound to aga rose and used to prepare an affinity column. An aliquot of solubilized placenta membranes was loaded onto the column at 4 C. After 14 h the column was extensively washed; fractions were collected and tested for the presence of insulin binding activity by PEG precipitation assay. No binding activity was recovered. Thereafter, 3 ng of labeled insulin (130 j.lCi/j.lg) were introduced into the column and f/ow was stopped for 24 h at 4 C. The column was washed and eluate fractions (0-0) were collected and counted. Then, 0.3 mg of porcine insulin were added to the column for 24 h at 4 C and the eluted radioactivity was counted (e-e). Finally, the column was washed with 1M KCNS to remove nonspecifically bound residual radioactivity (*-*). It is noteworthy that porcine insulin displaced most of the radioactivity bound for assumption to the receptor. Reproduced from (6) with permission.
o 0.2
10
100
Insulin (nglml)
Fig. 3 - Red blood cells trom atemale patient with hypoinsulinemia, hypoglycemia and lupus nephritis were isolated and incuba ted tor 210 min at 15·C with 251]-insulin (0.2 mg/mI) in the presence or absence ot increasing amounts ot porcine insulin. After incubation, cell-bound radioactivity was counted (33). The shaded area represents insulin binding to cells obtained trom a group ot 20 normal subjects. It is noteworthy that the patient's red blood cells totally unable to bind the hormone (34).
t
953
·R. Oe Pirro, P. Borboni, M.A. Marini, et al.
Table 1 - Insulin receptor binding inhibition.
after aperiod of hyperglycemia (26, 40), spontaneous complete remission with disappearance of autoantibodies (24, 40), gradual tendency to improve with therapy. However, none of these patients developed hyperglycemia after hypoglycemia; further, none died for hyperglycemia but several patients died for hypoglycemia. We may conclude that the incidence of this disease is actually low. The insulin binding domain, however, is a small part of the receptor moleeule (Fig. 1, inset) and in animals it is possible to induce antibodies directed to insulin receptor domain(s) other than the insulin binding site; in vitra, furthermore, monoclonal antibodies interacting with receptor regions distant from the insulin binding domain may affect insulin action without affecting insulin binding (see after). Therefore, some authors suspect that several patients have autoantibodies directed to sites different from the insulin binding domain. Various new methods of investigation have been praposed in order to reveal the presence of those antibodies. Jarrett et al. (42) labeled IgGs purified fram the serum of a patient with antireceptor antibodies and the antireceptor fraction was enriched by absorption and subsequent elution from IM-9 Iymphoblastoid cells in order to obtain a tracer directed to various domains of the insulin receptor. Other authors reported that autoantibodies may inhibit the insulin-receptor complex internalization and proposed this method to reveal immunoglobulins unable to inhibit insulin binding (43). To our knowledge these two methods did not allow identification of further affected subjects. A study group suggested that anti-insulin receptor antibodies are common in Type 1 diabetes mellitus since autoantibodies of the IgM class were found
% Sound
100
1. Control
2. Ig precipitated in ammonium-sulphate
52
3. Serum diluition 1/40
50
4. Serum treated with anti-lgG
99
Human placenta membranes (150 I'g/ml) (34) were incubated with [ '25 1]_ insulin (0.2 ng/ml) in the presence or absence 01 the indicated substances. The amount 01 IgG used in point 2 is similar to the concentration 01 IgG present in the serum diluted 1 /40 (point 3). It is noteworthy that binding inhibition in points 2 and 3 was similar. The absence 01 IgGs in point 4 was determined by immunoelectrophoresis (34).
The incidence of this pathology is higher in females than in males and it may appear at any age. Various patients have primarily an autoimmune disease; acanthosis nigricans may be present. Insulin resistance or impaired glucose tolerance is common but few patients showed normal fasting and glucosestimulated glycemia. It is noteworthy that diabetic patients are not particularly: ketosis prone. Fasting insulinemia is highly variable, being reported undetectable as weil as tremendously elevated (over 1,000 1LU/ml). Generally, exaggerated insulin responses to stimuli like glucose, tolbutamide or leuci ne may be observed; the patient 1-1 is an example. In so me cases elevated insulin levels proposed the primary diagnosis of islet cell tumor. Usually, we suspect the presence of antireceptor antibodies in patients presenting unusual alterations of glucose metabolism, epecially in the presence of acanthosis nigricans and / or autoimmune disorders. Up to 1990 literature reports less than 50.patients with anti-insulin receptor antibodies (19-28, 34-39, 41). The clinical course is variable: hypoglycemia
87c
6-
.8
5-
"0
::>
.S
-s
.S I
;;;
3-
t. 2-
-
t
r--
0
<:'!
Ci;
:t= ::>
.0
E
2
Cl)
0;
E 0
Fig. 4 - Example ot serum antagonism on 25 1Jinsulin bin ding. Red blood cells (3-4 x 1d' / ml) trom anormal subject were incubated tor 1 h at 24 C with buffer (33), the serum trom anormal subject (diluted 120), or the serum trom the 1-1 patient at the indicated dilutions. Thereatter, cells were washed twice at 24 C to remove endogenous insulin bound 10 Ihe receplor and incubaled with 25 lJ-insulin (0.2 ng/ml) tor 3 h at 15 C. Following incubation, cells were washed at 4 C and bound radioactivity was counted. As shown, it is evident the dose-dependent binding inhibition due to serum. Reproduced trom (34) with permission.
t
c
1-
0 10
102
104
Serum dilution
954
Antibodies to the insulin receptor
in 10 out of 22 insulin-dependent, untreated, diabetic patients (44). Authors, on the other hand, defined antibodies by their ability to stimulate lipogenesis in rat adipocytes; since many substances other than antireceptor antibodies can evoke insulin-like phenomena, the presence in sera of immunoglobulins to the insulin receptor it is not certain (44). Autoantibodies may immunoprecipitate the [125 1]insulin-receptor complex (46, 46). Boden et al. (47) prepared highly purified receptors labeled in the tyrosine residues and tested sera from diabetic patients for their ability to immunoprecipitate receptors.
By means of this method, they described that 10 out of 104 diabetic patients have autoantibodies; moreover, they pointed out that 9 out of the 10 positive sera did not inhibit insulin binding and did not stimulate glucose oxidation in rat adipocytes. Thus, they suggested that about 10% diabetic patients have antibodies directed to vahous domains of the insulin receptor. More recently, two re ports suggested the presence of low-titer autoantibodies in a large population of diabetic patients (48, 49) but conclusions are not certain due to methodological problems (47).
8
10 8
CI>
6
(/)
0
u~
~
CI>
(/)
0
~~ Cl E
"ES
6
äi
"O~
0 0
äi
4
0
4
2
...-...-.
•
•
•
0
2
300 0 200 c:=
=E ::>(/)::J
E~
100 700 0 2
500
= :g~
1.5
CI>
300
~~ JE
0.0 (.)0.
0.5
100
o
A
0
o
120
60
180
Minutes
B
0
40
20
60
Minutes
Fig. 5 - A: Plasma glucose and serum insulin variations during oral glucose tolerance test (75 g). The test was performed at 8.00 a.m. after 7 h fasting. B: Plasma glucose, serum insulin and serum C-peptide variations following tolbutamide infusion (1 g/2 min iv). Reproduced from (34) with permission.
955
R. Oe Pirro, P. Bo(boni, M.A. Marini, et al.
sponse to submaximal insulin concentrations, thus suggesting that chronic exposure to autoantibodies leads to decreased sensitivity of cells to insulin. The presence of a constant number of receptors on the plasma membrane is due to the so-called receptor "Iife-cycle"; it includes the receptor internalization and degradation, the receptor recycling back to the plasma membrane and the receptor synthesis. Some Authors suggested that the autoantibodies-induced cell desensitization may be due to receptor "down-regulation" produced by the accelerated rate at which receptors are degraded (57, 58). Even if the possibility that autoantibodies may provoke the receptor down-regulation exists, it is noteworthy that this may be not relevant to insulin action due to the presence of "spare-receptors". In
Pathogenesis of autoantibody-induced glycemic alterations In vitra autoantibodies show both the ability to inhibit insulin binding and to mimic most of the actions of insulin (Tab. 2) either in the case of hypoglycemic (50) or in the case of hyperglycemic patients (25, 51-55). It is worth noting that these in vitra biological effects may explain the in vivo appearance of hypoglycemia, but not the in vivo appearance of severe insulin resistance and hyperglycemia. However, it is important to point out that in vitra data were obtained in short-term experiments thus suggesting that they did not reveal the effects produced by the chronic in vivo exposure of target cells to autoantibodies. Karlsson et al. (56) reported that autoantibodies from one patient produced the expected insulin binding inhibition plus insulin-like effects in shortterm culture (2 h), but insulin binding inhibition without insulin-like effect in long-term culture (over 6 h). In conclusion, they suggested that chronic exposure to autoantibodies provokes cell desensitization in some step of insulin action. In order to address the same question, Pedersen ·et al. (24) studied the effect of insulin on adipocytes isolated from a hyperglycemic patient with autoantibodies. They reported that cells showed normal response to maximal insulin stimulation but very poor re-
Table 2 - In vitra Insulin-like effect ot anti-insulin receptor antibodies. ADIPOCYTES Insulin receptor internalization Glucose transport Glucose oxidation to CO 2 Lipogenesis Glycogen synthesis Aminoacid incorporation into prateins Activation of glycogen-synthase Inhibition of phosphorylase Activation of pyruvate dehydrogenase and acetyl-CoA carboxylase Inhibition of lipolysis
Insulin test
3T3-L 1 CELLS
0.4 U/Kg
Glucose transport Glucose oxidation to CO 2 Activation of lipoprotein lipase Glycogen synthesis Insulin receptor down-regulation
10
~
.s ~
8
8
6
MUSCLE
4
Glucose transport Stimulation of glycogen-synthase Glucose incorporation into glycogen
.2
~
8
co
2
o
LlVER
I
-10
I
30
0
(AlB) Aminoacid transport Activation of glycogen synthase Glycogen synthesis
I
60
Minutes
FIBROBLAST
Fig. 6 - (e) The effect of human insulin (0.4 U!kgl1 h) on plasma glucose; the study was pertormed 3 h after the ingestion ot breaktast due to the very low plasma glucose values in the resting state [!]. Plasma glucose va lues 3 and 4 h after breaktast the day betore and after the test; these values have been used as contrals. Repraduced tram (34) with permission.
(AlB) Aminoacid uptake IM 9 CELLS Insulin receptor internalization Insulin receptor down-regulation
956
Antibodies to the insulin receptor
showed decreased insulin, but not concanavalin A, sensitivity and responsiveness when tested in the lipogenic response assay (i.e. resistance specific for insulin). These data are in agreement with results obtained by Pedersen et al. (24) in adipocytes from an affected patient (see before). Revision of all data published in the field of antiinsulin receptor antibodies led us to some interesting observations. Studies in vitro demonstrated that sera from patients with hyper- or hypoglycemia inhibit insulin binding and, in addition, mimic insulin action in several cell types (25, 50-55). Some sera have a marked ability to inhibit insulin binding and a relatively weaker ability to mimic insulin effects; conversely, other sera show opposite actions. The same serum shows marked or weak ability to inhibit insulin binding or to stimulate insulin actions depending upon the cell type evaluated (8). The monoclonal antibody MA-51 inhibits both insulin binding and insulin effects (13). Some patients with hyperglycemia changed suddenly to hypoglycemia without appreciable variations in antibody activities (26, 40). Therefore, we hypothesized that patient sera might contain several antibodies directed to different domains of the insulin receptor involved in specific biological activities (50). The observed hyper- or
this respect Shimoyama et al. (25) reported that autoantibodies from one diabetic patient still induced glycogen synthesis in cultured rat hepatocytes after 24 hexposure. Dons et al. (59), on their part, addressed the question by simulating the in vivo situation through passive transfer of IgGs from affected patients to rats. IgGs from both hyperglycemic and hypoglycemic subjects produced hypoglycemia when acutely injected into fasting rats; injection to fed rats for several days, conversely, induced persistent hyperglycemia that declined immediately to hypoglycemic levels if rats were obliged to fast. This study would suggest that the patient metabolie situation might modulate autoantibody effects, In this respect, Shechter et al. (60, 61) injected mice with insulin in order to produce anti-insulin antibodies and, subsequently, anti-insulin receptor antibodies developing as anti-idiotypes; the laUer showed in vitro both inhibitory activity on insulin binding and insulin mimicking effects. The Authors observed that at the appearance of receptor antibodies animals had a 20 day-period of hypoglycemia followed by aperiod of hyperglycemia after 8 h fasting. Moreover, in the presence of receptor antibodies mice showed altered glucose tolerance and their adipocytes
100
50
A
Patienrs IgG
0 10.8
"c0::>
100
.0
>-
"8~
,gE
B
-::>
~.~
n;~
50
§E
g~ c
pH 3.3
Fig. 7 - The effect ot IgG subpopulations puritied
~
;;;
~
trom the serum ot I-I patient on the binding ot 25 /}-insulin (A) and 25 Ij-MA-51 monoclonal antibody (8) to IM-9 cel/s. IgG subpopulations were puritied by means ot elution trom a Protein A column through a pH gradient. Reproduced with the permission trom (50).
t
0 10.7
10.6
10.5
Protein added (M)
957
t
R. Oe Pirro, P. Borboni, MA. Marini, et al.
hypoglycemia should derive from relative potencies of these different populations of autoantibodies plus, probably, the basic metabolie situation of the patient. Differently from autoantibodies in patients (see before), the monoclonal antibody (MA-51), produced by R. Roth and 1.0. Goldfine (San Francisco, USA) (12), showed the peculiar effect to block both insulin binding and action (13), thus suggesting that it was directed to a small segment of the receptor moleeule and, probably, to apart of the insulin binding domain. Therefore, we made the hypothesis that some immunoglobulins from patients would compete with insulin, but not with MA-51 , for the binding to the receptor (50). According to this hypothesis we tried to demonstrate the presence of various antibody subpopulations in the serum of the 1-1 patient by using her IgGs, [125-1] insulin and [125-1] MA-51 in typical competition- inhibition experiments. Initial results pointed out that IgGs were more active towards insulin than MA-51; therefore, IgG subpopulations were prepared and it was clear that s~me IgG fractions were able to antagonize insulin, but not MA-51, binding (Figs. 7a and b) (62). This was the clear-cut demonstration that the insulin receptor has several antigenic sites and that in the patient serum various antireceptor antibody populations were present (50). Next, insulin-like effect of IgG subpopulations from the 1-1 patient was tested, and it was demonstrated that they showed different potencies (50). At this point, the potency to inhibit insulin binding was compared with the potency to mimic insulin action of each IgG subpopulation and it was clear that in the patient serum coexisted immunoglobulins showing insulin-like activity (agonist) and immunoglobu-
10-7
10-6
10-8
10-6
lins showing antagonist effect (Figs. 8a and b) (50). As a consequence, the hypothesis that some antibodies have only agonistic activity whereas others have only antagonistic activity was put forward (50). By studying the effect of five monoclonal antibodies in human adipocytes, Taylor et al. (63) concluded that antibodies may affect receptor function by acting on different receptor domains, even those distant from the insulin binding site. In fact, two antibodies to the a-subunit inhibited insulin binding and mimicked insulin action. Two antibodies had slight or no ability to inhibit insulin binding but they mimicked insulin action; one of these antibodies was directed to the ß-subunit. The fifth antibody was directed to the a-subunit, it inhibited insulin binding and it did not mimic insulin action; on the other hand, it inhibited insulin-induced lipogenesis or catecholamineinduced lipolysis. The Authors, furthermore, suggested that some effects may be non specific but due to the immunoglobulin property to cross-link and to aggregate receptor molecules. Two papers reported the in viva glucose metabolism in two diabetic patients with autoantibodies. One (26) reported hyperglycemia due to overpropuction of glucose through increased glycogenolysis and gluconeogenesis rather than decreased glucose utilization. Moreover, Authors suggested the presence of increased rates of glucose utilization probably due to increased glycogen synthesis and / or fatty acid esterification (i.e. dichotomy in autoantibody effects on hepatic and extraepatic insulin receptors and/or post-receptor events). The patient was treated with the oral hypoglycemic agent methyl-2-tetradecylglycidate (methyl palmoxirate), an inhibitor of free fatty acid oxidation which increases museie glucose oxidation and decreases hepatic gluconeogenesis. He changed suddenly from hyper- to hypoglycemia after 2-months treatment, and died, thus leading to the hypothesis that the drug blocked the antibody antagonistic activity on the liver and enhanced the antibody agonistic effect on the museie. The second paper (25) evaluated the in viva rates of glucose disappearance and appearance both in the absence and in the presence of insulin infusion, and the in vitra effect of both the patient's serum and the patient's 33% ammonium sulfate precipitate gammaglobulins on glycogen synthesis in cultured rat hepatocytes. By comparison of data Authors found the presence of a dichotomy between in viva and in vitra results since antibody effect on glucose disappearance was about 10% of that obtainable at
10-4
Pro tein concentration (M)
t
Fig. 8 - Inhibitory effect on 25 lj-insulin binding (.-.) and insulinlike activity (glucose transport) (*-*) of two different subpopulations of autoantibodies to the insulin receptors. A: The inhibitory effect of this subpopulation is greater than insulin-like activity. B: On the contrary, in this subpopulation the insulin-like activity is predominant as deduced by the fact that IgGs at 10-6 M stimulate glucose transport but do not inhibit insulin binding.
958
Antibodies to the insulin receptor
The problem in identifying and isolating the receptor molecule among the host of molecules present in the cell was overcome by using antibodies to immunoprecipitate specifically the receptor. Thanks to that method, for example, it was possible to demonstrate the presence of carbohydrates (64), the molecular weight and th'e intracellular life-cycle of the receptor (65, 66) as weil as receptor clustering on plasma membrane to produce insulin effects (67). In the last part of th'is review article, studies carried out by using antibodies as a tool to investigate insulin receptor structure and function will be reviewed.
maximal insulin concentrations, whereas antibody effect on glycogen synthesis was comparable to maximal insulin activity. As glucose disappearance depends mainly on muscle glucose utilization, the Authors suggested that antibodies were more effective on hepatocytes than on muscle (tissue-specific antibodies). This suggestion was in agreement with data previously reported (54) demonstrating that antibodiesfrom one patient in vitra stimulated glycogen synthesis in muscle about 3 times less than insulin whereas in other cells exerted effects like maximal insulin concentrations. The body of data reported str'ongly suggests that the relative potency of immunoglobulins acting on different sites is the ethiology of hyper- and hypoglycemia; particularly, the antibody specificity for muscle, liver and adipose tissue may playa major key role.
Species- and tissue-specificities of the insulin receptor Evaluation of 'the IgG fraction from the serum obtained from 1-1 patient indicated that immunoglobulins were more effective in inhibiting insulin binding to human than to rat cells/tissues. Since this serum contained various antibody subpopulations directed to different receptor regions (see before), the presence of IgG directed toward the human, but not the rat, insulin receptor was suggested. By testing various IgG subpopulations, finally we isolated immunoglobulins able to inhibit the binding of insulin to human cells/tissues only (Fig. 9), thus pointing out that the human insulin receptor has speciesspecific antigenic determinants (conformations) (68)
INSULIN RECEPTOR STRUCTURE AND FUNCTION It has already been mentioned that the receptoris a tetrameric glycoprotein which is composed of two subunits coming trom a receptor precursor; the receptor, after the interaction with insulin, is subjected to processes of internalization, degradation and recycling back to plasma membrane. RAT Adipocytes { Hepatocytes ~ Liver membranes Brain particles
100
o 100
I
Hepatocytes Adipocytes Monocytes Placenta RBC IM-9
'0 C
5,OE c::l
=E
~.~
.!f E ~~
50 HUMAN
• • • ...
RBC Placenta Adipocytes IM-9 cells and monocytes
o 10.8
0--'---,------.-----, 10-9
10-8
10-7
10-7
10.6
10-5
pH 4.3 subpopulation (M)
. pH 5.5 subpopulation (M)
Fig. 10 - Effect ot an IgG subpopulation, puritied trom the 1-1 serum through pH gradient on Protein A column, on 25 lj-insulin binding to various human cellsltissues. It is noteworthy that IgGs inhibit binding to tibroblast but they do not affect the binding to other tissues. Reproduced with the permission trom (62).
t
Fig. 9 - Effect ot an IgG subpopulation purified trom the 1-1 serum on 25lj-insulin binding to various human and rat tissues. It is noteworthy that IgGs do not inhibit insulin binding to rat tissues. Reproduced with the permission trom (68).
t
959
R. Oe Pirro, P. Borboni, M.A. Marini, et al.
in agreement with data that some monoclonal antibodies may recognize the human, but not the rat, insulin receptor (13, 16: 69). Next, antibodies revealed the presence of speciesspecificities even in other animals; in fact, Soos et al. (15) tested the ability of antibodies to inhibit insulin binding and to co-precipitate the insulin receptor from several species and found: i) antibodies cross-reacting weil with bovine receptors but rather less weil with rabbit receptors; ii) one antibody cross-reacting weil with rabbit receptors but showing no reaction with bovine receptors; iii) antibodies cross-reacting with human, but not with rabbit, receptors; iv) antibodies cross-reacting weil with human receptors but rather less weil with bovine receptors. As far as tissue-specificities of the insulin receptor are concerned, Shimoyama et al. (25) repported indirect evidence onthe presence of liver-specific autoantibodies in one hyperglycemic patient. In our laboratory, we tested several IgG subpopulations obtained from 1-1 patient on various cells/tissues;. finally, one IgG subpopulation able to inhibit insulin binding to human fibroblast but unable to inhibit insulin binding to a wide variety of other tissues was found (Fig. 10). Furthermore, the same subpopulation activated aminoacid' (AlB) uptake in human fibroblasts but did not mimic the action of insulin on adipocytes (62). In order to support the idea that fibroblasts have highly tissue-specific sites, experiments were carried out with the monoclonal antibody MA-51 which neither is tissue-specific nor blocks insulin-induced AlB uptake in human fibroblasts (13). 1-1 patient antibody subpopulation stimulated AlB uptake in fibroblasts in the absence as weil as in the presence, of MA-51 , thus confirming the presence of fibroblast-specific receptor site(s) (62). The demonstration of autoantibodies specifically directed against the fibroblast insulin receptor may have important consequences. In fact, it is possible that patients presenting only autoantibodies to the fibroblast receptor do exist; they may be not clinically suspected since they would not manifest an alteration in glucose metabolism. Furthermore, those autoantibodies may be the ethiological factor producing some unexplained disease phenomena in patients. Next, Roth et al. (70) purified at homogeneity the receptor of human cerebral cortex by affinity chromatography on columns of agarose-coupled mo-
noclonal antibodies and wheat germ agglutinin. Both the molecular weight and the insulin binding activity of the cerebral cortex receptor was similar to the placenta receptor. The monoclonal antibody MA51, on the other hand, bound to the cerebral cortex receptor with an affinity 300-times less than that obtained in the presence of the placenta receptor, thus suggesting that a-subunits from those two tissues have different conformations at, or close to, the insulin binding domain. Finally, Kuli et al. (16) produced the monoclonal antibodyalR-1 which does not inhibit insulin binding but immunoprecipitates the insulin receptor. This antibody shows various species- and tissue-specificities, thus pointing out that specificities are present even in receptor regions not involved in insulinreceptor interaction. The presence of tissue-specificities was suggested some time aga by Pottick et al. (71) but no evidence was so far offered. Results obtained with antibodies, on the other hand, have been recently confirmed by using other methodological approaches (72, 73), as weil as by using together inhibition-competition studies with monoclonal antibodies and SDS electrophoresis of purified receptors in acrylamide gel (74).
Insulin receptor down-regulation One important point is the characterization of the receptor domain involved in the down-regulation process; in fact, it might be responsible for some pathological phenomena and it is probably involved in the insulin-receptor complex internalization. Both polyclonal (57, 58) and monoclonal antibodies (75-78) induce receptor down-regulation, like insulin; moreover, they act by increasing receptor degradation (57, 58, 76, 78). Monoclonal antibodies MA-5 and MA-20 induce down-regulation even if they interact with receptor regions' distant from the insulin binding site (75). The monoclonal antibody MA-51 , but not insulin, induces this process in the Chinese hamster ovary cell line CHO.T -1- which expresses a receptor with a degraded intracellular domain due to the removal of the C-terminal 112 aminoacids, as weil as in the CHO.YF1 and CHO.YF3 cell lines expressing receptors in which tyrosines-1162 or both tyrosine-1162 and 1163, respectively, are replaced by phenylalanines (78). In conclusion insulin and aQtibodies would induce down-regulation via different mechanisms; moreover, the process may be activated in the absence of important portions of the intracellular receptor.
960
Antibodies to the insulin receptor
Initially, we looked at this question by using IgGs from the 1-1 patient which in vitra exerted an insulinlike effect (50). IgGs inhibited receptor phosphorylation both in the absence and in the presence of insulin (83); the phenomenon, however, was not induced by autoantibodies to the insulin binding site since it was provoked at antibody concentrations unable to affect insulin binding as weil as in a receptor preparation in which binding sites were totally occupied by insulin (i.e. antibodies were absOlutely unable to interact with insulin binding sites) (83). Moreover, hormone-receptor interaction leads to phosphorylation of Insulin Growth Factor I (IGFI) and Epidermal Growth Factor (EGF) receptors. IgGs from 1-1 patient inhibited receptor phosphorylation even in the case of both EGF receptor (83) and IGF-I receptor (84); the effect, however, was more pronounced on IGF-I receptors compared with insulin receptors (84) and it was more pronounced on insulin receptors compared with EGF receptors (83, 84). Finally, previously bound IGF-I, but not previously bound insulin, was dissociated by IgGs from receptors (83). The body of these data is in agreement with the above mentioned thesis that patients have several autoantibodies with different activities and specificities. Moreover, the monoclonal antibody MA-1 0 was able to inhibit insulin binding to human, but not rat, tissue/ cells while inhibited insulin-induced receptor phosphorylation in both species (85) (Fig. 11). It was difficult to ofter a reasonable interpretation of the result that one receptor region (i.e. the region to which MA-1 0 is directed) may affect both insulin binding and receptor phosphorylation in human, but only receptor phosphorylation in rat cells. The issue was clarified through the demonstration that MA-10 binds both the a- and the ß-subunits of the human receptor, but only the ß-subunit of the rat receptor (77); another important point was that MA-10 binds spe-
Finally, the monoclonal antibody MA-10 induces receptor down-regulation in human, but not rat, cells (75, 77); in the presence of insulin, on the other hand, it inhibits the hormone-induced down-regulation (77). Thismonoclonal binds both to the a- and to the ß-subunit of the human receptor but only to the extracellular part of the rat receptor ß-subunit (77). Therefore, all these data would suggest that insulin-receptor interaction produces ci receptor conformation change, or an intra-receptor message, which increases receptor degradation rate (i.e. receptor down-regulation) by activating/ deactivating one or more receptor domains on the a-subunit which are modulated by the extracellular region of the ß-subunit. Insulin receptor phaspharylatian and insulin action Insulin-receptor interaction provokes immediate phosphorylation of tyrosines located in the intracellular part of the ß-subunit (Fig. 1, inset). The first report was from Kasuga et al. (79) who described that insulin induces the phosphorylation of the insulin receptor; subsequently, the phenomenon was confirmed and studied with various techniques. In an elegant experiment Roth and Cassel (80) purified at homogeneity the insulin receptor by using to an affinity column of agarose-coupled MA-51; thereafter, they showed that the receptor ß-subunit phosphorylated in the presence of labeled ATP and that it has ATP binding sites. The Authors thus demonstrated that the insulin receptor is itself a protein kinase since in the reacting material only insulin receptors were present. In vitra, polyclonal autoantibodies mimic insulin action (see before); as a consequence the Authors suggested that they mimic insulin-induced receptor phosphorylation. On the contrary, in 1984 two papers reported that autoantibodies may both induce or block receptor phosphorylation (81, 82).
100
A
100 0= -0 .~b CC
"0
c
:J
'5c'"E 50
g8
"'
8'" c-O
o~
.cx
~~
80 60
o~
e--g
.,~~
40
a:~ 20
N
o
lll'" a> 1 2
3 4
0
~"-r(1--'1--'---:=---"""1
o
10-9
10.8 M Ä-1 0 (Ml
961
10-6
Fig. 11 - A: MA-1 0 inhibition ot the insulin binding to insulin receptors puritied trom: human placenta (1); human liver (2); rat liver (3); rat adipose tissue (4). It is noteworthy that MA-10 does not attect insulin bin ding to rat tissues. B: MA-1 0 inhibition ot insulin-induced autophosphorylation ot receptors puritied trom: human liver (+-T); human placenta (e-.); rat adipose tissue ( . - . ) and rat liver (1..-1..). Reproduced with the permission trom (85).
R. Oe Pirro, P. Borboni, M.A. Marini, et al.
eifieally to rat intaet eells (77) thus demonstrating that it interaets with the extraeellular portion of the ß-subunit. MA-1 0 is direeted to an epitope of the asubunit wh ich is at, or elose to, the insulin binding domain (75) and, therefore, distant from tlie ß-subunit in the aminoacid sequenee (1) (Fig. 1); eonsequently, the reeeptor must have the insulin binding site spaeely elose to the ß-subunit in its natural environment. Finally, this nation fits very weil with, and it explains, all those apparently contradictory results reported on antibody-affeeted reeepto'r phosphorylation (81, 82). The reeeptor tyrosine kinase may be aetivated adding to the reaetion either insulin or ATP (1). There is evidenee that a eritieal step in the phosphorylation proeess is the 1018 aminoacid lysine whieh direeting the third phosphate of the ATP to tyrosine residues (86) is peeuliar in ATP-indueed reeeptor phosphorylation. By studying IgGs from 1-1 patient it was found that they inhibited the insulin-indueed, but not the ATP-indueed, reeeptor phosphorylation (87) thus suggesting that these processes are different; the former neeessitating of the aetivity of the latter, but not the reverse. Autoantibodies, probably, do not interaet with the 1018 lysine region as weil as with intraeellular phenomena oeeurring after ATP aetivation of the phosphorylation proeess; it is reasonable to suggest that they affeet phosphorylation aeting on the extraeellular environment. Both sodium vanadate and the anti-ß-subunit monoelonal antibody 18-44 (15) induee phosphorylation through pathways indipendent of insulin binding site oeeupaney (88). MA-10 inhibited both sodium vanadate- and (18-44)-indueed reeeptor phosphorylation; the same result was obtained in the presenee of rat liver (88). Sinee MA-10 binds to the extraeellular portion of the rat ß-subunit (see before), the reeeptor phosphorylation indueed by-either insulin or otheragents aeting outside the insulin binding domain must be modulated by this reeeptor portion. MA-10 inhibits aminoacid uptake in both human and rat eells (77) and Morgan et al. (89) showed that the insulin-indueed ooeyte maturation is bloeked by mieroinjeeting Xenopus ooeytes with a monoelonal antibody direeted to the ß-subunit and able to inhibit reeeptor phosphorylation. In eonelusion, the phosphorylation proeess is important for long-term insulin effeets. As far as rapid insulin effeets are eoneerned, data are not in agreement. Both monoelonal antibodies MA-5 and MA-20 in-
duee glucose transport in human adipoeytes but do not provoke reeeptor phosphorylation (90). MA-10 whieh interaets with the ß- but not with the asubunit in rat tissues (77) does not affeet glucose transport in rat adipoeytes both in the absence and in the presenee of insulin (unpublished data).. These data would suggest that glucose transport is aetivated through a message involving only the a-subunit whereas aminoacid uptake or RNA/DNA synthesis are aetivated throljgh reeeptor phosphorylation. In this respeet, Chou et al. (86) reported that CHO eells with a ß-subunit mutated in aminoacid 1018 (alanine substituting lysine) lose the ability to transport glucose both in the absence and in the presenee of insulin. The same data may be obtained mutating the aminoacid in position 1150 (ph811ylalanine substituting tyrosine) (91). Furthermore, Morgan and Roth (92) reported that CHO eells mieroinjeeted with monoelonal antibodies to the intraeellular portion of the reeeptor deerease their ability to transport 2-deoxyglueose. These data would suggest, therefore, that reeeptor phosphorylation is important in the glucose transport proeess. Main differenee between experiments in agreement and those in disagreement with an involvement of the phosphorylation proeess is that the former were earried out aeting on the intraeellular portion of the reeeptor, whereas the latter were performed earried out aeting on the extraeellular environment. DIFFERENCES BETWEEN HUMAN AND PORCINE INSULIN Human and poreine insulin differ by the B-30 aminoaeid residue whieh is loeated elose to the part of insulin moleeule interaeting with the insulin reeeptor. It was suggested that this variation may produee some differenees, even if minor, in insulin-insulin reeeptor interaction. Some studies, on the other hand, did not reveal any differenee betwesn human and poreine insulin in both the binding to tissues and biologieal effeets. Monoelonal antibodies are direeted to various epitopes of the insulin reeeptor (see before) whieh are at, elose to or distant fram the insulin binding site. Therefore, it was suggested that monoelonal antibodies might reveal differenees in human and poreine insulin interaction with the human insulin reeeptor: In order to verify this hypothesis eompetition-inhibition binding studies were earried out using three monoelonal antibodies (MA-
962
Antibodies to the insulin receptor
ACKNOWLEDGMENTS
5, MA-10 and MA-20) labeled with [125-1] (93). Previous studies demonstrated that antibodies iodinated by mean of the IODOGEN method had unaltered .biological activity (93) and were therefore suitable. Competition-inhibition binding studies revealed that human and porcine insulin inhibited MA-10 and MA-20 binding to various human cells /tissue to the same extent; on the contrary, human insulin was more effective than porcine insulin in inhibiting MA-5 binding (94). MA-5 interacts with a region of the receptor a-subunit which is distant from the insulin binding domain but involved in the intra-receptor signaling transmission (75). Therefore, it is possible that human insulin is more effective than porcine insulin in inducing a species-specific intra-receptor conformational change subsequent to insulin- receptor interaction.
This work was supported in part by grants from Novo Farma.ceutici Italia, Ministero PubbHca Istruzione, Consiglio Nazionale Ricerche n. 87.00336.56, Fidia SpA.
REFERENCES 1. Goldfine 1.0. The insulin receptor: Molecular biology and transmembrane signaling. Endocr. Rev. 8:, 235, 1987. 2. Ebina Y., Ellis L., Janargin K., Edery M., Graf L., Clauser E., Ou J.H., Masiarz F., Kan YW., Goldfine 1.0., Roth RA, Rutter W.J. The human insulin receptor cDNA: The structural basis for hormone-activated transmembrane signaling. Cell. 40: 747, 1985.
CONCLUSIONS
3. Ulrich A., Bell J.R, Chen E.Y., Herrera R, Petruzzelli L.M., Dull T.J., Gray A., Coussens L., Liao Y.C., Tsubokawa M., Mason A, Seeburg P.H., Grunfeld C., Rosen O.M., Ramachandran J. Human insulin receptor and its relationship to be tyrosine kinase family for oncogenes. Nature. 313: 756, 1985.
Autoantibodies to insulin receptor may affect both insulin and insulin receptor function; therefore, they seem a good tool to understand the role of receptors in insulin action as weil as the functional domains of the receptor. In fact studies carried out with immunoglobulins have weil clarified some biological phenomena. On the other hand, antibodies are synthesized. against precise domains of the receptor molecule; therefore, in the case that receptor conformation is peculiar to provoke the biological message, immunoglobulins will reveal themselves unique in clarifying peculiar biological phenomena. For example, Herrera and Rosen (95) produced an antibody against the aminoacid sequence 11431162 of the receptor containing tyrosines 1150 and 1151; this antibody interacts with the phosphorylated, but not with the unphosphorylated, receptor thus suggesting that a conformational change is important for the phosphorylation process. Theactual gap is that we do not have techniques able to characterize the site of the receptor to which monoclonal antibodies are directed. In the evaluation of antibodies in human disease we are probably looking only at the tip of one iceberg. Some preliminary evidences would suggest that many patients could have autoantibodies; some antibodies, furthermore, would cross-react with receptors showing high-degree homology with the insulin receptor (96). It would be necessary, in the future, to clarify the real incidence of this autoimmune disease as weil as the biology of antibody-induced phenomena.
4. Flier J.S., Kahn C.R, Roth J., Bar RS. Antibodies that impair insulin receptor binding in an unusual diabetic syndrome with severe insulin resistance. Science. 190: 63, 1975. 5. Cuatrecasas P. Affinity chromatography and purification of the insulin receptor of liver cell membranes. Proc. Natl. Acad. Sci. (USA). 69: 1277, 1972. 6. Oe Pirro R., Lauro R., Gelli A.S., Bertoli A., Musiani P. Specificities of rabbit anti-human insulin receptor antibodies. J. Clin. Lab. Immunol. 2: 27, 1979. 7. Jacobs S., Chang K.J., Cuatrecasas P. Antibodies to purified insulin receptor have insulinlike activity. Science. 200: 1238, 1978. 8. Muggeo M., Ginsberg B.H., Roth J., Kahn C.R, Oe Meyts P., Neville D.M. Jr. The insulin receptor in vertebrates is functionaily more conserved during evolution than insulin itself. Endocrinology, 104: 1393, 1979. 9. Muggeo M., Van Obberghen E., Kahn C.R, Roth J., Ginsberg B.H., Oe Meyts P., Emdin S.O., Falkmer S. The insulin receptor and insulin of the atlantic hag-
963
R. Oe Pirro, P. Borboni, M.A. Marini, et al.
fish. Extraordinary conservation of binding specificity and negative cooperativity in the most primitive vertebrate. Diabetes. 28: 175, 1979.
Syndrome associated with extreme insulin-resistant diabetes. Proc. Symp. Chem. Physiol. Pathol. (Ikagaku Shinpojumu) 15: 58, 1975.
10. Pulch P., Czech M.P. Interaction of cross-linking agents with the insulin effector systems of isolated fat cells. J. Biol. Chem. 254: 3375, 1979.
21. Blackard W.G., Anderson J., Mullinax F Anti-insulin receptor antibodies and diabetes. Ann. Intern. Med., 861: 584, 1977. 22. Baldwin D. Jr., Winston F., Hashizaki R.J., Baldwin D. Sr., Holcomb H.H., Flier J.S., Rubenstein A. Insulin resistant diabetes with insulin receptor autoantibodies in a male patient without acanthosis nigricans. Diabetes Care 2: 275, 1979.
11. Jacobs S., Hazum E., Cuatrecasas P. The subunit structure of rat liver insulin receptor. Antibodies directed against the insulin-binding subunit. J. Biol. Chem. 14: 6937, 1980. 12. Roth RA, Wong K.Y., Maddux FA, Goldfine 1.0. Production of antibodies that inhibit the binding of insulin to its receptor. Biochem. Biophys. Res. Commun. 101: 979, 1981.
23. Weinstein P.S., High KA, D'Ercole A.J., Jennette J.C. Insulin resistance due to receptor antibodies: a complication of progressive systemic sclerosis. Arthritis Rheum. 23: 101, 1980.
13. Roth RA, Cassell D.J., Wong K.Y., Maddux BA, Goldfine 1.0. Monoclonal antibody to the human insulin receptor blocks insulin binding and inhibits insulin action. Proc. Natl. Acad. Sci. (USA) 79: 7312, 1982.
24. Pedersen 0., Hjollund E., Beck-Nielsen H., Kroman H. Diabetes mellitus caused by insulin-receptor blockade and impaired sensitivity to insulin. N. Engl. J. Med. 304: 1085, 1981.
14. Montemurro A., Borboni P., Lauro R., Oe Pirro R., Goldfine 1.0., Moneta E. Nuove strategie nella produzione di anticorpi monoclonali diretti contro i recettori per I'insulina e per il fattore di crescita insulinico I. J. Lab, Med. 13: 199, 1986.
25. Shimoyama R., Ray J.K., Savage C.R. Jr., Owen O.E., Boden G. In vivo and in vitra effects of antiinsulin receptor antibodies. J. Clin. Endocrinol. Metab. 59: 916, 1984.
15. So os MA, Siddle K., Baron M.D., Heward J.M., Luzio J.P., Bellatin J., Lennox E.S. Monoclonal antibodies reacting with multiple epitopes or the human insulin receptor. Biochem. J. 235: 199, 1986.
26. Mandarino L., Tsalikian E., Bartold S., Marsh H., Corney A., Buerklin E., Tutwiler G., Haymond M., Handwerger G., Rizza R. Mechanism of hyperglycaemia and response to treatment with inhibitor of fatty acid oxidation in a patient with insulin resistance due to insulin receptor antibodies. J. Clin. Endocrinol. Metab. 59: 658, 1984.
16. Kuli FC., Jacobs S., Su Y.F., Cuatrecasas P. A monoclonal antibody to human insulin receptor. Biochem. Biophys. Res. Commun. 106: 10196, 1982.
27. Fonseca V., Kholer M.A., Dandona P. Insulin receptor antibodies causing steroid responsive diabetes mellitus in a patient with myositis. Br. Med. J. 288: 1578, 1984.
17. Morgan 0.0., Roth RA Mapping surface structures of the human insulin receptor with monoclonal antibodies: Localization of main immunogenic regions to the receptor kinase oomain. Biochemistry. 25: 1364, 1986.
28. Shoelson S.R., Marshall S., Horikoshi P., Kolterman O.G., Rubenstein A.H., Olefsky J. Antiinsulin receptor antibodies in an insulin dependent diabetic may arise as autoantiidiotypes. J. Clin. Endocrinol. Metab. 63: 56, 1986.
18. Song C.S., Yu H.J., Bai D.H., Hester P.Y., Kim K.H. Antibodies to the alfa-subunit of insulin receptor from eggs of immunized hens. J. Immunol. 135: 5, 1985. 19. Van Obberghen E., Kahn C.R. Antibodies to insulin receptor. Mol. Cell. Endocrinol. 22: 277, 1981.
29. Flier J.S., Kahn C.R., Jarrett D.B., Roth J. Characterization of antibodies to the insulin receptor. A cause of insulin-resistant diabetes in man. J. Clin. Invest. 58: 1442, 1976.
20. Kibata M., Hiramatso K., Shimizo J., Fuchimoto T., Sazaki M., Shimono N., Majake K., Flier J.S., Kahn C.R. Insulin receptor antibody found in a case of Sjogren
30. Muggeo M., Kahn C.R., Bar R.S., Rechler M., Flier J.S., Roth J. The underlying insulin receptor in patients with antireceptor autoantibodies: Demonstration of normal
964
Antibodies to the insulin receptor
receptor autoantibodies. Spontaneous remission on receptor proliferation with hypoglycaemia. J. Clin. Endocrinol. Metab. 47: 985, 1978.
binding and immunological properties. J. Clin. Endocrinol. Metab. 49: 110, 1979. 31.
Muggeo M., Flier S., Ross AA, Harrison L.C., Deisserroth AB., Kahn C.R. Treatment by plasma exchange of a patient with autoantibodies to the insulin receptor. N. Engl. J. Med. 300: 477, 1979.
41.
32. Kawanishi K., Kawamura K., Nishima Y., Goto A, Okada S., Ishida T., Ofuji T., Kahn C.R., Flier J.S. Successful immunosuppressive therapy in insulinresistant diabetes caused by anti-insulin receptor autoantibodies. J. Clin. Endocrinol. Metab. 44: 12, 1977. 33.
42. Jarrett D.B., Roth J., Kahn C.R., Flier J.S. Direct method for detection and characterization of cell surface receptors ·for insulin by means of 125 1_ labeled autoantibodies against the insulin receptor. Proc. Natl. Acad. Sci. (USA) 73: 4115, 1976.
Oe Pirro R., Fusco A, Lauro R., Testa 1., Ferretti F., Oe Martinis C. Erythrocyte insulin receptors in non-insulin dependent diabetes mellitus. Diabetes. 29: 96, 1980.
43. Yoshimasa Y., Namba Y., Hanaoka M., Kohno M., Okamoto M., Hattori M., Yamada K., Kuzuya H., Imura H. A new approach to the detectior:J of autoantibodies against insulin receptors that inhibit the internalization of insulin into human cells. Diabetes 33: 1051, 1984.
34. Tardella L., Rossetti L., Oe Pirro R., Camagna A., Leonetti F., Tamburrano G., Merli M., Rossi Fanelli F., Lauro R. Circulating anti-insulin receptor antibodies in a patient suffering from lupus nephritis and hypoinsulinemic hypoglicaemia. J. Clin. Lab.lmmunol. 12: 159, 1983.
44. Maron R., Elias 0., Oe Jangh B.M., Bruining G.I., Van Rood J.J., Schechter Y., Cohen I.R. Antibodies to the insulin receptor in juvenile onset insulin-dependent diabetes. Nature 303: 817, 1983. 45.
35. Taylor S.I., Grunberger G., Marcus-Samuels B., UnderhilI L.H., Dons R.F., Ryan J., Roddan R.F., Rupe C.E., Gorden P. Hypoglycaemia associated with antibodies to the insulin receptor. N. Engl. J. Med. 307: 1421, 1982. 36.
Immunoprecipitation of the insulin receptor: A sensitive assay for receptor antibodies and a specific technique for receptor purification. J. Clin. Endocrin. Metab. 48: 59, 1979. 46. Tsushima T., Ohmori Y., Murakami H., Shizume K., Hirata Y. Immunoprecipitation of 1251-insulin crosslinked receptor. Diabetologia 74: 187, 1983.
Brand W.J., Maylor BA, Williamson D.H., Buley 1.0., Clark A, Chapel H.M., Turner H.M. Autoimmunity to insulin receptor and hypoglycaemia in patient with Hodgkin's disease. Lancet1:237,1987.
47.
Boden G., Fujita-Yamaguchi Y., Shimoyama R., Shelmet J.J., Tappy L., Rezvani 1., Owen O.E. Nonbinding inhibitory anti insulin receptor antibodies. A new type of autoantibodies in human diabetes. J. Clin. Invest. 81: 1971, 1988.
48.
Ludwig S.M., Faiman C., Dean H.J. Insulin and insulin-receptor autoantibodies in children with newly diagnosed 100M before insulin therapy. Diabetes 36: 420, 1987.
Elias 0., Cohen I.R., Shechter Y., Spirer Z., Golander
A
Antibodies to insulin receptor followed by anti-idiotype antibodies to insulin in child with hypoglycaemia. Diabetes. 36: 348, 1987. 39.
40.
Harrison L.C., Flier J.S., Roth J., Karlsson FA, Kahn
C.R.
37. Walters E.G., Tavare J.M., Denton R.M., Walters G. Hypoglycaemia due to an insulin receptor antibody in Hodgkin's disease. Lancet ii: 241, 1987. 38.
Bar R.S., Lews W.R., Rechler M.M., Harrison L.C., Svebert C., Podskalmy J., Roth J., Muggeo M. Extreme insulin resistance in ataxia telengiectasica. Defect in affinity of insulin receptors. N. Engl. J. Med. 298: 1164, 1978.
Moller D.H., Ratner R.E., Borenstein D.G., Taylor S.I. Autoantibodies to the insulin receptor as a cause of autoimmune hypoglycaemia in Systemic Lupus Erythematosus. Am. J. Med. 84: 334, 1988.
49. Tarseh BA, Thompson RA, Odugbesan 0., Barnett AH. Insulin receptor antibodies in diabetes mellitus. Clin. Exp. Immunol. 71: 85,1988.
Flier J.S., Bar R.S., Muggeo M., Kahn C.R., Roth J., Gorden P. The evolving clinical course of patients with insulin
50.
965
Oe Pirro R., Rossetti L., Roth RA, Goldfine 1.0. Characterization of the serum from a patient with insulin resistance and hypoglycaemia. Evidence for
R. Oe Pirro, P. Borboni, M.A. Marini, et al.
multiple populations of insulin receptor antibodies with different receptor binding and insulin-mimicking activities. Diabetes 33: 301, 1984. 51. Baldwin 0; Jr., Terris S., Steiner D.F. Characterization of insulin-like actions of anti-insulin receptor antibodies. Effects on insulin binding, insulin degradation and glycogen synthesis in isolated rat hepatocytes. J. Biol. Chem. 255: 4028, 1980.
J. Clin. Invest. 72: 1072, 1983. 60. Schechter Y., Elias 0., Maron R., Cohen I.R. Mouse antibodies to the insulin receptor developing spontaneously as anti-idiotypes. I. Characterization öf the antibodies. J. Biol. Chem. 259: 6411,1984. 61. Elias 0., Maron R., Cohen I.R., Schechter Y. Mouse antibodies to the insulin receptor developing spontaneously as anti-idiotypes. 11. Effects on glucose homeostasis and the insulin receptor. J. Biol. Chem. 259: 6416, 1984.
52. Belsham G.J., Brownsey R.w., Hughes W.A., Denton R.M. Antiinsulin receptor antibodies mimic the effects of insulin on the activities of pyruvate dehydrogenase and AcetylCoAcarboxilase and on specific protein phosphorylation in rat epididymal fat cells. Diabetologia 18: 307, 1980.
62. Oe PirroR., Borboni P., Lauro R., Testa S., Festa A., Oe Martinis C., Maddux BA, Goldfine 1.0. Tissue-specific antibodies against the fibroblast insulin receptorin a patient with Lupus nephritis and hypoglycaemia. Diabetes 34: 1088, 1985.
53. Kasuga M., Akamuma Y., Tsushima T., Iwamoto Y., Kosaka K., Kibata M., Kawaniski K. Effects of anti-insulin receptor autoantibodies on the metabolism of human adipocytes. Diabetes 27: 938, 1978.
63. Taylor R., Soos M.A., Wells A., Argyrakf M., Siddle K. Insulin-like and insulin-inhibitory effects of monoclonal antibodies for different epitopes on the human insulin receptor. Biochem. J. 242: 123, 1987.
54. Le Marchand-Brustel Y., Gorden P., Flier J.S., Kahn C.R., Freychet P. Anti-insulin receptor antibodies inhibit insulin binding and stimulate glucose metabolism in skeletal muscle. Diabetologia 14: 311,1978.
64. Hedo JA, Kasuga M., Van Obberghen E., Roth J., Kahn C.R. Direct demonstration of glycosylation of insulin receptor subunits by biosynthetic and externallabeling: Evidence for heterogeneity. Proc. Natl. Acad. Sci. (USA) 78: 4791,1981.
55. Kahn C.R., Baird K., Flier J.S., Jarrett D.B. Effects of autoantibodies to the insulin receptor on isolated adipocytes. J. Clin. Invest. 60, 1094, 1987.
65. Van Obberghen E., Kasuga E., Le Cam M., Hedo JA, Itin A., Harrison L.C. Biosinthetic labeling of insulin receptor. Studies of subunits in cultured human IM-9 Iymphocytes. Proc. N"atl. Acad. Sci. (USA) 78: 1052,1981.
56. Karlsson FA, Van Obberghen E., Grunfeld C., Kahn C.R. Desensitization of the insulin receptor at an early postreceptor step by prolonged exposure to antireceptor antibody. Proc. Natl. Acad. Sci. (USA) 76: 809, 1979.
66. Jacobs S., Cuatrecasas P. Insulin receptor structure and function. Endocr. Rev. 2: 251, 1981. 67. Kahn C.R., Baird K.L., Jarrett D.B., Flier J.S. Direct demonstration that receptor crosslinking or aggregation is important in insulin action. Proc. Natl. Acad. Sci. (USA) 75: 4209, 1978.
57. Grunfeld C. Antibody against the insulin receptor caus~s disappearance of insulin receptor in 3T3-L 1 cells: a possible explanation of antibody-induced insulin resistan ce. Proc. Natl. Acad. Sci. (USA) 81: 2508, 1984.
68. Oe Pirro R., Rossetti L., Montemurro A., Lauro R., Gammeltoft S., Maddux BA, Goldfine 1.0. Human autoantibodies directed against the human, but not the rat, insulin receptor. J. Clin. Endocrinol. Metab. 61: 986, 1985.
58. Taylor S.I., Marcus-Samuels B. Antireceptor antibodies mimic the effect of insulin to down-regulate insulin receptors in cultured human Iymphoblastoid (IM-9) cells. J. Clin. Endocrinol. Metabol. 58: 182, 1984.
69. Morgan 0.0., Ho L., Korn L.J., Roth RA Insulin action is blocked by a monoclonal antibody that inhibits the receptor kinase. Proc. Natl. Acad. Sci. (USA) 83: 328, 1986. 70. Roth RA, Morgan 0.0., Beaudoin J., Sara V. Purification and characterization of the human brain insulin receptor. J. Biol. Chem. 261: 3753, 1986.
59. Dons R.F., Havtik R., Taylor S.I., Baird K.L., Chernick S.S., Gorden P. Clinical disorders associated with autoantibodies to the insulin receptor. Simulation by passive tranfer of immunoglobulins to rats.
966
Antibodies to the insulin receptor
71. Pottick L.A., Moxley R.T. 111, Livingston J.N. Tissue differences in insulin receptors. Acute changes in insulin binding characteristics induced by Wheat Germ Agglutinin. Diabetes 30: 196, 1981.
The role of antireceptor antibodies in stimulating phosphorylationof the insulin receptor. . J. Biol. Chem. 259: 4396, 1984. 83. Cordera R., Gherzi R., Oe Pirro R., Rossetti L., Freidenberg G.R., Andraghetti G., Lauro R., Adezati L. Inhibition of insulin and epidermal growt~ factor (EGF) receptor autophosphorylation by a human poIyclonal IgG. Biochem. Biophys. Res. Commun. 132: 991, 1985. 84. Cordera R., Gherzi R.,.oe Pirro R., Andraghetti G., Freidenberg G.R., Minuto F., Lauro R., Giordano G., Adezati L. Insulin-like growth factor I (IGF I) receptor autophosphorylation and kinase activity. Effect of a human polyclonal antibody (plgG). Biochem. Biophys. Res. Commun. 138: 1023, 1986.
72. Burant C.F., Trentelaar M.K., Blocks M.E., BU!3e M.G. Structural differences between liver- and musclederived insulin receptors in rats. J. Biol. Chem. 261: 14361,1986. 73. Burant C.F., Trentelaar MK, Buse M.G. Tissue specific differences in the insulin receptor kin ase activated in vitro and in vivo. Endocrinology 122: 427, 1988. 74. Caro J.F., Rajn S.M., Simha M.K., Goldfine 1.0., Oohm G.L. Heteterogeneity of human liver, muscle and adipose tissue insulin receptor. Biochem. Biophys. Res. Commun. 151: 123, 1988.
85. Corde ra R., Andraghetti G., Gherzi R., Adezati L., Montemurro A., Lauro R., Goldfine 1.0., Oe Pirro R. Species specificity of insulin binding and insulin receptor protein tyrosine kinase activity. Endocrinology 121: 2007, 1987.
75. Forsayeth J.R., Monteinurro A., Maddux BA, Oe Pirro R., Goldfine 1.0. Effect of monoclon.al antibodies on human insulin receptor autophosphorylation, negative cooperativity, and down regulation. J. Biol. Chem. 262: 41340, 1987.
86. Chou C.K., Oull T.J., Russell O.S., Gherzi R., Lebwohl 0., Ullrich A., Rosen O.M. Human insulin receptors mutated at the ATP-binding site lack protein tyrosine kinase activity and fail to mediate postreceptor effects of insulin. J. Biol. Chem. 262: 1842, 1987.
76. Roth RA, Maddux BA, Cassell O.J., Goldfine 1.0. Regulation of the insulin receptor by a monoclonal antireceptor antibody. Evidence that receptor down regulation can be independent of insulin action. J. Biol. Chem. 258: 12091, 1983.
87. Gherzi R., Cordera R., Andraghetti G., Oe Pirro R., Freidenberg G.R., Lauro R., Adezati L. Regulation of insulin receptor-associated tyrosine kinase by a polyclonal IgG. Mol. Cell. Endocrinol. 53: 9, 1987.
77. Gherzi R., Sesti G., Andraghetti G.; Oe Pirro R., Lauro R., Adezati L., Cordera R. An extracellular district of the insulin receptor beta subunit with regulatory function on protein tyrosine kinase. J. Biol. Chern. 264: 8627, 1989.
88. Gherzi R., Caratti C., Andraghetti G., Bertolini S., Montemurro A., Sesti G., Cordera R. Oirect modulation of insulin receptor protein tyrosine kinase by vanadate and anti-insulin receptor monoclonal antibodies. Biochem. Biophys. Res. Commun. 152: 1474, 1988.
78. Morgan 0.0., Ellis L., Rutter W.J., Roth RA Antibody-induced down-regulation of a mutated insulin receptor lacking an intact cytoplasmic domain. Biochemistry 26: 2959, 1987.
89. Morgan 0.0., Ho L., Korn L.J., Roth RA Insulin action is blocked by a monoclonal antibody that inhibits the insulin receptor kinase. Proc. Natl. Acad. Sci. (USA) 83: 328, 1986.
79. Kasuga M., Karlsson FA, Kahn C.R. Insulin stimulates the phosphorylation of the 95,000dalton subunit of its own receptor. Science 215: 185, 1982.
90. Forsayeth J.R., Caro J.F., Sinha M.K., Maddux BA, Goldfine 1.0. Monoclonal antibodies to the human insulin receptor that activate glucose transport but not insulin receptor kinase activity. Proc. Natl. Acad. Sci. (USA) 84: 3444, 1987.
80. Roth RA, Cassell O.J. Insulin receptor: Evidence that it is a protein kinase. Science 219: 299, 1988. 81. Simpson IA, Hedo JA Insulin receptor phosphorylation may not be a prerequisite for acute insulin action. Science 223: 1301 , 1984.
91. Ellis L., Clauser E, Morgan 0.0., Edery M., Roth RA, Rutter W.J. Replacement of insulin receptor tyrosine residues 1162 and 1163 compromises insulin-stimulated kinase activity and uptake of 2-deoxyglucose.
82. Zick Y., Rees-Jones R.w., Taylor S.I., Gorden P., Roth J.
967
R. Oe Pirro, P. Barbani, M.A. Marini, et al.
94. Sesti G., Marini MA, Bertoli A., Oe Pirro R., Lauro R., Montemurro A. Monoclonal anti-insulin receptor antibodies reveal differences on human and porcine insulin binding to human tissues. Diabetes Res. Glin. Pract. 5 (Suppl. 1 ): 542, 1988.
Gell 45: 721, 1986. 92. Morgan 0.0., Roth RA Acute insulin action requires insulin receptor kinase activity: Introduction of an inhibitory monoclonal antibody into mammalian cells blocks the rapid effects of insulin. Proc. Natl. Acad. Sci. (USA) 84: 41, 1987.
95.
93. Sesti G., Marini MA, Montemurro A., Di Daniele M., Bertoli A., Gordera R., Andraghetti G., Oe Pirro R., Lauro R., Monaco F., Roche J. Preparation of monoclonal anti-insulin receptor antibodies labelled by 1251. G.R. Soc. Biol. 182: 167, 1988.
Herrera R., Rosen O.M. Autophosphorylation of the insulin receptor in vitra. J. Biol. Ghem. 261: 11980, 1986.
96. Tappy L., Fujita-Yamaguchi Y., Lebon T.R., Boden G. Antibodies to insulin-like growth factor I receptors in diabetes and other disorders. Diabetes 37: 1708, 1988.
968