J. Endocrinol. Invest. 14: 913-918 , 1991
Evidence that human and porcine insulin differently affect the human insulin receptor: studies with monoclonal anti-insulin receptor antibodies G. Sesti*, M.A Marini*, A Montemurro*, P. Borboni*, G. Di Cola*, A Bertoli*, R. De Pirro**, and R. Lauro* *Cattedra di Endocrinologia, Dipartimento di Medicina Interna, II Universita di Roma-Tor Vergata, and ** Cattedra di Endocrinologia, Universita di Ancona, Italy ABSTRACT. Binding studies have been carried out with radioiodinated monoclonal antibodies directed to various epitopes of the insulin receptor in order to detect differences between human and porcine insulin in the interaction with the human insulin receptor. Human insulin was more effective that porcine insulin at inhibiting the binding of 1251-MAS to IM-9 cells, Hep-2 human larynx cells and human placenta membranes. On the contrary, human
and porcine insulin showed similar inhibitory effect on the binding of two other labeled anti-insulin receptor monoclonal antibodies, thus ruling out the possibility that results were due to experimental artifacts. Although several interpretations are possible, data reported suggest that human insulin and porcine insulin might differently affect the insulin receptor, even if, the biological significance of these findings remains unknown.
INTRODUCTION
bility that human and porcine insulin differently affect the human insulin receptor.
Human and porcine insulin differ by the B-30 amino acid residue which is located close to the domain of the insulin molecule interacting with the insulin receptor (1). Therefore, it has been suggested that this variation may have some effect, even if minor, on the interaction of insulin with its receptor. In vitro and in vivo studies carried out to clarify this issue did not reveal any difference in the binding properties of human insulin and porcine insulin (2-8), as well as in the bioeffect of the two insulins (9, 10). On the other hand, it has been shown that mutation of a single amino acid residue of the insulin molecule determines severe resistance due to a marked defective insulin bioeffect (11). Monoclonal antibodies to the insulin receptor are a valuable tool to study both insulin receptor function and insulin action since they are specifically directed to various epitopes of the receptor (12-20). In the present study three monoclonal antibodies directed to different domains of the insulin receptor have been used in order to investigate the possi-
MATERIALS AND METHODS 125 1(16 .5 mCi//lg) was purchased from Amersham (Buckinghamshire, England); 10DO-GEN was purchased from Pierce (Rockford , IL, USA); porcine and human insulin were kindly provided by Novo Industries (Copenhagen, Denmark). 125 1_ A 14 monoiodo-porcine .and -human insulin, prepared by the lactoperoxidase method (280-300 /lCi//lM) (21), were a gift from Prof. Navalesi (Pisa, Italy). All other chemicals were purchased from Sigma Chemical Co. (St. Louis, MO, USA).
Monoclonal antibodies iodination Three monoclonal anti - insulin receptor antibodies (MA-5, MA-1O and MA-20) (17) have been used . Antibodies, purified on a Affi-Gel Protein A column (Bio-Rad , Richmond, CA, USA), were iodinated by the 10DO-GEN method, as previously described (22). The specific activity was 3-5 /lCi//lg. Binding Studies Human lymphoblast cells (1M-g) (1x106 cells/ml) suspended in 100 mM Hepes buffer (pH 7.8) containing 120 mM NaCI, 1.2 mM Mg S04, 1 mM EDTA, 15 mM sodium acetate, 10 mM glucose, 10 mg/ml bovine serum albumin (BSA) were incubated for 100 min at
Key-words: Human insulin . porci ne insulin . insulin receptor , monoclonal anti-insulin receptor antibod ies. Correspondence: Giorgio Sesti, MD., Dipartimento di Medicina Interna. II Universita di Roma, Via O. Raimondo , 00173 Roma, Italy. Received November 5. 1990; accepted June 22, 1991.
913
G. Sesti, MA M arin i, A M ontemurro , et al.
1S C with 1251 -monoclonal antibody (MA-S, MA-10 or MA-20) at tracer concentration (70 pM) in the presence of absence or increasing amounts of porcine insulin , human insulin or unlabeled monoclonal antibody. In another set of experiments IM-9 cells were incubated with 125 1 human insulin or 125 1 porcine insulin at tracer concentration (SO pM) in the presence or absence of various amounts of human insulin , porcine insulin or unlabeled MA-S . Bound and free radioactivity were separated by centrifuging cells twice at 4 C in Hepes buffer. Human larynx carcinoma cells (Hep-2) were grown to confluence in 6-well multiwells , washed twice with 100 mM Hepes buffer (pH 7.9) containing 120 mM NaCI, 1.2 mM MgS0 4 , 15 mM sodium acetate, 10 mM glucose, 10 mg/ml BSA and incubated in the same buffer for 90 min at 15 C with 1251 -monoclonal antibody (MA-5 , MA-10 or MA-20) at tracer concentration (70 pM) in the presence or absence of increasing amounts of porcine insulin , human insulin or unlabeled monoclonal antibody. In another set of experiments HEp-2 cells were incubated with 125 1 -human or 125 1 -porcine insulin at tracer concentration (50 pM) in the presence or absence of various amounts of human insulin , porcine insulin or unlabeled MA-5. Bound radioactivity was determined by washing cells twice with 10 mM Tris-HCI (pH 7.8) containing 154 mM NaCI and scraping cells from wells with 0.03% (w/v) sodium dodecyl sulfate (SOS). After counting, the protein content of the solubilized cells was measured by means of the dye method of Bradford (23). Human placenta membranes (150 Ilg/ml), prepared as previously described (24) , were incubated for 60 min at 22 C in 50 mM Tris buffer (pH 7.8) con-
taining 10 mM MgCI2' 2mM EDT A , 5 mg/ml BSA with 1251 -monoclonal antibody (MA-S, MA-10 or MA-20) at tracer concentration (70 pM) in the presence or absence of increasing amounts of porcine insulin , human insulin or unlabeled monoclonal antibody. In another set of experiments human placenta membranes were incubated with 125 1 -human or 125 1 -porcine insulin at tracer concentration (50 pM) in the presence or absence of various amounts of human insulin , porcine insulin or unlabeled MAS. Bound and free radioactivity were separated by centrifuging placenta membranes twice at 4 C in Tris buffer. Experiments with 125 1 -monoclonal antibody (MA-5, MA-10 or MA-20) have been always carried out with the same stock and concentrations of human insulin and porcine insulin . Nonspecific binding , defined as the binding in the presence of 1IlM monoclonal antibody, human insulin or porcine insulin , was subtracted and it was less than 10% of total binding . Both 125 1 -monoclonal antibody and 125 1 -insulin degradations were determined before and at the end of each incubation period by both 10% TCA precipitation and rebinding to fresh placenta membranes or cells. Statistical analysis Statistical analysis was performed by using Student's t test for unpaired data. RESULTS
In order to evaluate differences between human and porcine insulin in their ability to interact with
20
/
/
10-10
~
~~
~~
TOTAL PROTEIN [M]
914
Fi g. 1 - The effect of huma n insulin (A- A) and porcine insulin (e_) on (1 251)-MA- 10 binding to IM9 cel/s. IM-9 cel/s (1 x 10 6 eel/simi) were incubated for 100 min at 15 C with 1251-MA- 10 (70 p M) in the presence or absence of increasing amounts of porcine insulin or human insulin. Nonspecific binding was calculated and subtracted. Values are the means ± SO of three experiments carried out in triplica te. (' p < 0.01, **p < 0.001).
Human and pork insulin binding to receptor
the human insulin receptor , competition-inhibition studies were carried out using three radioiodinated monoclonal antibodies to the human insulin receptor (MA-5, MA-10 or MA-20). Human and porcine insulin inhibited to the same extent the binding of 125 1-MA-1O to IM-9 cells (Fig. 1), Hep-2 cells and human placenta membranes (data not shown). Similar results were obtained with 125 1-MA20 (data not shown). The finding that the two insui ins did not show any difference in their ability to compete for the insulin receptor confirmed previously reported data (2-8) as well as our preliminary binding inhibition studies carried out with either 125 1_ human insulin or 1251-porcine insulin (data not shown) . On the contrary, human insulin was more effective that porcine insulin at inhibiting 125 1-MA-5
binding to IM-9 cells, 50% inhibition occurring at 5x10- 7 M with human insulin and at 9x10-7 M with porcine insuUn (Fig . 2A) . Same data were obtained in studies carried out with HEp-2 cells, 50% inhibition occurring at6x10-7 M with human insulin and at 9x10- 7 M with porcine insulin (Fig . 28). Next, isolated human placenta membranes were used in order to confirm the data in a cell-free system. Again, human insulin was more effective than porcine insulin at inhibiting 1251-MA-5 binding (Fig. 2C). As shown , differences were statistically significant in the range of 10-8 M-5x1 0-6 M insulin concentration ; in pratice, at any insulin concentration able to inhibit 125 1-MA-5 binding (Figs. 2A, B and C). Finally, unlabeled MA-5 inhibited to the same extent the binding of both 125 1-human and 125 1-porcine insulin
A 80 )(
...'"
:Ii
...o
0
£
£60
60
Q
Q
Z
Z
::l
::l
CD
CD
1/)40
I/)
o
0
'"
40
'"
2
:Ii .!. ~
"
20
20
•• 0 I 0~~/~--~~='.'-----~1~~'~-----±IV~·-------Jlcr
I
/
I
10"
I
10-1
TOTAL PROTEIN
TOTAL PROTEIN [M]
I
I
10"
10'"
(M]
c
100
80 )( ~
...260 0
.£. Q
Z ::l
40
Fig. 2 - The effect of human insulin (4- 4) and porcine insulin (e---.) on 1251-MA-5 binding to IM-9 cells (panel A), Hep-2 cells
0 II
..,
~
2 20 .!..
III
~
0
I
10-1 TOTAL PROTEIN
I
[M]
~
(panel B) and human placenta membranes (panel C).IM-9 cells (1x1Cf3 cells/ml), HEp-2 cells (grown to confluence in 6-weJI multiwells) or human placenta membranes (150 Ilg/ml) were incubated with 1251-MA-5 (70 pM) in the presence or absence of increasing amounts of porcine insulin or human insulin. Nonspecific binding was calculated and subtracted. Values are the mean ± SO of five experiments carried Qut in triplicate. ip < 0.01;**p < 0.001).
915
G. Sesti, MA Marini, A Montemurro, et al.
100
">c
80
'"' Q
. I
60
Z
40
i
~;==::::
~
:::I
., ::)
Z
! !
';"7
.:0-MA-5 [M)
to IM-9 cells (Fig. 3), HEp-2 cells and human placenta membranes (data not shown). Degradation of both 125 1 -monoclonal antibodies and 125 1 -insulins, verified at the beginning and the end of each incubation period , was negligible.
Fig. 3 - The effect of MA-S monoc lonal antibody on 1251-human insulin (A- A) and 1251-porcine insulin (e-e) binding to IM-9 cel/s. IM-9 cel/s (lxl(J3 cel/slml) were incubated for lOa min at IS C with 1251-human or 1251-porcine insulin (SO pM) in the presence or absence of increasing amounts of unlabeled MA-S. Nonspecific binding was calculated and subtracted. Values are the mean ± SO of three experiments carried out in triplicate.
Data reported raised several possible interpretations. Results may derive from a sterical inhibitory effect of the two insulins on MA-5 binding , so that inhibition could be related to the possible proximity between the 8-30 aminoacid residue (different for human and porcine insulin) and the receptor domain recognized by MA-5. This possibility is unlikely since MA-5 does not affect differently the binding of porcine and human insulin (Fig. 3) and monoclonal antibodies do not bind insulins per se (1720) ; furthermore, it has been shown that MA-5 interacts with a receptor region distant from the insulin binding site and that it inhibits insulin binding allosterically by altering the receptor conformation (17,26). Another possibility is that the two insulins may induce sligthly different conformational changes of the receptor; in this respect, it has been demonstrated that antibodies to the receptor molecule can recognize peculiar conformational changes of the receptor (26, 27) which are thought to be involved in the transmembrane signal transmission (28). According to this method of reasoning, MA-5 would recognize better the conformation induced by human insulin than that induced by porcine insulin. On the other hand , the finding that human insulin was significantly more effective than porcine insulin at hormone concentrations which are superphysiological with respect to insulin receptor binding raises the possibility that results are due to phenomena which are mediated by low affinity, high capacity insulin binding sites rather than by high affinity insulin receptors. Finally, we cannot rule out the possibility that high affinity binding of MA-5 re-
DISCUSSION
In the present investigation three monoclonal antibodies (MA-5, MA-1O and MA-20) directed to different epitopes of the human insulin receptor (17,25) were used in order to investigate differences between human and porcine insulin in the interaction with the human insulin receptor. Monoclonal antibodies, purified on Affi-Gel Protein A column to homogeneity, were iodonated by means of the 1000GEN method (22). It has been demonstrated that antibodies labeled by this iodination procedure retain their full biological activity and interact with the insulin receptor better than those labeled by other methods (22). Human insulin and porcine insulin inhibited to the same extent the binding of both MA-1 0 and MA-20 to all cells and tissue tested . On the contrary, human insulin was more effective that porcine insulin at displacing MA-5 from the insulin receptor. It is unlikely that results were due to artifacts determined by the different aggregration state of the two insulins since differences did not appear in experiments carried out with 125 1 -MA-1O or 125 1 -MA-20. Furthermore, tracer degradation did not affect results since it was negligible either at the beginning or at the end of the incubation period .
916
Human and pork insulin binding to receptor
10. Home D, Shepherd G.A.A., Noy G., Massi-Benedetti N., Hanning I., Burrin J.M., Alberti K.G.M.M. Comparison of the activity and pharmacokinetics of porcine insulin and human insulin (Novo) as assessed by the glucose clamp technique in normal and diabetic man. Diabetes Care 6: 23, 1983. 11. Tager H., Given B., Baldwin D., Mako M., Markese J., Rubenstein A, Olefsky J., Kobayashi M., Kolterman 0., Poucher R. A structurally abnormal insulin causing human diabetes. Nature 281: 122,1979. 12. Roth R., Cassel D.J., Wong K.Y., Maddux B., Goldfine 1.0. Monoclonal antibodies to the human insulin receptor block insulin binding and inhibit insulin action. Proc. Natl. Acad. Sci. USA 79: 7312,1982. 13. Kull F.C., Jacobs S., Svoboda M.E., Van Wyk J.J., Cuatrecasas P. Monoclonal antibodies to receptors for insulin and somatomedin-c. J. Bioi: Chem. 258:6561,1983. 14. Soos M.A., Siddle K., Baron M.D., Heward J.M., Luzio J.P., Bellatin J., Lennox E.S. Monoclonal antibodies reacting with multiple epitopes on the human insulin receptor. Biochem. J. 235: 199,1986. 15. Morgan D.O., Ho L., Roth R. Insulin action is blocked by a monoclonal antibody that inhibits the insulin receptors kinase. Proc. Natl. Acad. Sci. USA 83: 328, 1986. 16. Morgan D.O., Roth R. 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. 17. Forsayeth J.R., Montemurro A, Maddux B., De Pirro R., Goldfine 1.0. Effect of monoclonal antibodies on human insulin receptor autophosphorylation, negative cooperativity and down-regulation. J. BioI. Chem. 262: 4134,1987. 18. Cordera R., Andraghetti G., Gherzi R., Adezati L., Montemurro A, Lauro R., Goldfine 1.0., De Pirro R. Species specificity of insulin binding and insulin receptor protein tyrosine kinase activity. Endocrinology 121: 2007,1987.
quires a peculiar conformation of the receptor, and that the two insulins differently affect the ability of the receptor to assume that conformation. The biological significance of these findings is unknown; however, data reported suggest that human insulin and porcine insulin might differently affect the insulin receptor.
ACKNOWLEDGEMENTS This work was supported by grants from NOVO Italia and Ministero Pubblica Istruzione 40% and 60%.
REFERENCES 1. Pullen R.A., Lindsay D.G., Wood SP., Ticle I.J., Blundell T.L., Wollmer A, Krail G., Brandenburg D., Zahn H., Gliemann J., Gammeltoft S. Receptor-binding region of insulin. Nature 259: 369, 1976. 2. De Meyts P., Halban P., Hepp O.K. In vitro studies on biosynthetic human insulin: an overview. Diabetes Care 4: 144, 1981. 3. Fehlmann M., Le Marcharnd-Brustel Y., Dlais-Kitabgi J., Morin 0., Freychet P. Biological activity and receptor binding properties of biosynthetic human insulin in isolated rat hepatocytes and mouse soleus in vitro. Diabetes Care 4: 223, 1981. 4. Fussgaenger R.D., Ditschuneit H.H., Martini H., Wieauer-De Lenardis H., Etzrdt H., Thun C.H., Ditschuneit H., Pfeiffer E.F., Enzamann F. Potency of biosynthetic human insulin determined
5.
6.
7.
8.
9.
in vitro. Diabetes Care 4: 228, 1981. Gammeltoft S. Receptor binding of biosynthetic human insulin on isolated pig hepatocytes. Diabetes Care 4: 235, 1981. Olefsky J.M. Insulin binding, biologic activity and metabolism of biosynthetic human insulin. Diabetes Care 4: 244, 1981. Sonne 0., Foley J.E., Gliemann J. The biologic potency and binding affinity of biosynthetic human insulin in isolated rat adipocytes. Diabetes Care 4: 250, 1981. Zapf J., Froesch E.R. Comparison of pancreatic human and biosynthetic human insulin with respect to their action on adipocytes and chick embryo fibroblast. Diabetes Care 4: 257, 1981. Ebinara A, Kondo K., Ohashi K., Kosaka K., Kuzuya T., Matsuda A Comparative clinical pharmacology of human insulin (Novo) and porcine insulin in normal subjects. Diabetes Care 6: 17,1983.
19. Gherzi R., Caratti C., Andraghetti G., Bertolini S., Montemurro A, Sesti G., Cordera R. Direct modulation of insulin receptor protein tyrosine kinase by vanadate and anti-insulin receptor monoclonal antibodies. Biochem. Biophys. Res. Commun. 152: 1474,1988. 20. Gherzi R., Sesti G., Andraghetti G., De Pirro R., Lauro R., Adezati L., Cordera R. An extracellular domain of the insulin receptor beta-
917
G. Sesti, MA. Marini, A. Montemurro, et al.
subunit with regulatory function on protein-tyrosine kinase. J. BioI. Chem. 264: 8627,1989.
Characteristic of the human placenta insulin receptor. LAB. J. Res. Lab. Med. 7: 23, 1980. 25. Forsayeth J.R., Caro J.F., Sinha M.K., Maddux BA, Goldfine I.D. Monoclonal antibodies to the human insulin receptor that activate glucose transport but not insulin receptor kinase activity. Proc. Natl. Acad. Sci. USA 84: 3448, 1987.
21. Navalesi R., Pilo A., Ferrannini E. Kinetic analysis of plasma insulin disappearance in non ketotic diabetic patients and in normal subjects. J. Clin. Invest. 61: 197,1979. 22. Sesti G., Marini MA, Montemurro A., Di Daniele N., Bertoli A., Cordera R., Andraghetti G., De Pirro R., Lauro R., Monaco F., Roche J. Preparation of monoclonal anti-insulin receptor antibodies labelled by 125_1. C.R. Soc. BioI. 182: 167,1988.
26. Gu J.L., Goldfine I.D., Forsayeth J.R., De Meyets P. Reversal of insulin-induced negative cooperativity by monoclonal antibodies that stabilize slowly dissociating ("Ksuper) state of the insulin receptor. Biochem. Biophys. Res. Commun. 150: 694, 1988. 27. Herrera R., Rosen O.M. Autophosphorylation of the insulin receptor in vitro. J. Bioi. Chem. 261: 11980,1986.
23. Bradford M.M. A rapid and sensitive method for the quantitation of micrograms quantities of protein utilizing the principle of protein dye binding. Anal. Biochem. 72: 248,1976.
28. Kahn C.R., White M.F. The insulin receptor and the molecular mechanism of insulin action. J. Clin. Invest. 82: 1151, 1988.
24. Magnatta R., Spallone L., De Pirro R., Fusco A., Forte F., Scandurra R., Lauro R.
918
