1992;52:3924-3930. Published online July 1, 1992.Cancer Res Giovanni Milazzo, Francesco Giorgino, Giuseppe Damante, et al. Cancer Cell LinesInsulin Receptor Expression and Function in Human Breast
Updated Version http://cancerres.aacrjournals.org/content/52/14/3924
Access the most recent version of this article at:
Insulin Receptor Expression and Function in Human Breast Cancer Cell Lines1
Giovanni Milazzo, Francesco Giorgine, Giuseppe Damante, Chin Sung, Martha R. Stampfer, Riccardo Vigneri,Ira D. Goldfine,2 and Antonino BelfioreCattedra di Endocrinologia dell'Università di Catania, Ospedale Garibaldi, 95123 Catania, Italy {G. M., F. G., G. D., R. V., A. B.J; Division of Diabetes and
Endocrine Research, Mount Zion Medical Center of the University of California, San Francisco, California 94120 ¡G.M., C. S., I. D. G.]; Departments of Medicineand Physiology, University of California, San Francisco, California 94143 ¡I.D. G.J; Lawrence Berkeley Laboratory, University of California, Berkeley, California94720 [M. R.SJ
ABSTRACT
We have previously reported Chat insulin receptor expression is increased in human breast cancer specimens (V. Papa et al., J. Clin.Invest., 85:1503-1510, 1990). In the present study, in order to furtherunderstand the role of the insulin receptor in breast cancer, insulinreceptor expression and function were characterized in three humanbreast cancer cell lines, MCF-7, ZR-75-1, and T-47D, and compared toa nonmalignant human breast epithelial cell line, 184B5. Insulin receptor content, measured by radioimmunoassay, was elevated 5- and 3-foldin MCF-7 and ZR-75-1 breast cancer cell lines, respectively, whencompared to the nonmalignant cell line 184B5. In contrast, the insulinreceptor content of T-47D cells was not increased. The increase ininsulin receptor content in MCF-7 and ZR-75-1 cells was not due toamplification of the insulin receptor gene. Also, total insulin receptormRNA content was not increased in breast cancer cells in respect tononmalignantly transformed IK-IBS breast epithelial cells. However,significant differences in the content of receptor mRNA species wereobserved.
The insulin receptors in the breast cancer cell lines were functional:(a) In all 4 cell lines, high-affinity insulin-binding sites were detected,and, in concert with the insulin receptor radioimmunoassay data, binding capacity was highest in MCF-7 and then in ZR-75-1 cells, (b) In allcell lines, insulin stimulated insulin receptor tyrosine kinase activity.However, the effect of insulin was greater in breast cancer cell lines thanin nonmalignant breast cells, (c) In all cell lines, insulin at concentrations of l UMor less stimulated [ 'I I |th> midinc incorporation. This effect
of insulin was inhibited by 50% in MCF-7 cells and by 60% in 184B5cells when a-IR3, a monoclonal antibody to the insulin-like growthfactor I receptor, was present. In these cells, therefore, insulin wasactive via both its own receptor and the IGF-I receptor. In contrast,0-IR3 antibody was without effect in T-47D and ZR-75-1 cells, suggesting that in these cell lines insulin acted only via its receptor. In thebreast cancer cells, MA-5, an agonist monoclonal antibody to the insulinreceptor, stimulated |3Hlthymidine incorporation. This present studyindicates therefore that in breast cancer cell lines there are functionalinsulin receptors that regulate breast cancer cell growth.
INTRODUCTION
Growth factor receptors of the tyrosine kinase family play akey role in both normal and neoplastic cell growth. The insulinreceptor belongs to the tyrosine-kinase growth factor receptorfamily (1-3), and insulin mediates proliferative responses in a
Received 4/22/91; accepted 5/6/92.The costs of publication of this article were defrayed in part by the payment of
page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1Supported by the Associazione Italiana per la Ricerca sul Cancro (AIRC), NIHGrant CA24844. The Office of Energy Research, Office of Health and Environmental Research of the United States Department of Energy under Contract DE-AC03-765500098. The John A. Kerner Foundation. The Jay Gershow CancerFund, and the Ladies Auxiliary of the Veterans of Foreign Wars. G. M. is arecipient of an AIRC fellowship.
2 To whorn requests for reprints should be addressed, at Division of Diabetes andEndocrine Research. Mount Zion Medical Center, P. O. Box 7921, San Francisco,CA94120.
variety of both normal and transformed cells (4, 5). However,the role of the insulin receptor molecule in human neoplasia hasnot yet been established. Recently, we have reported that over-expression of the insulin receptor is a characteristic feature ofmany human breast cancer specimens (6). By using a specificinsulin receptor radioimmunoassay, we found that the averageinsulin receptor content of human breast cancer specimens wasapproximately 6-fold higher than that of normal breast tissue.With immunohistochemical analysis we localized the increasedexpression of the insulin receptor to the malignant epithelialcells. The insulin receptor content of the breast cancer specimens was positively correlated with tumor grade and tumorsize. Moreover, we have recently demonstrated that with fibro-blasts and ovary cells transfected with and overexpressing insulin receptors, the addition of insulin induced a ligand-depend-
ent transformed phenotype (7). These observations suggested,therefore, a possible role for insulin receptor overexpression inhuman cancer initiation and/or progression.
Human breast cancer cells in tissue culture have been important In vitro models for studying the regulation of breast cancertissue by hormones and growth factors. Prior studies have identified the presence of insulin receptors in breast cancer cell lines(8). In certain human breast cancer cell lines insulin stimulatesseveral cellular functions including cell growth (9, 10), andrecently we have reported that progestins enhance the mitoge-nic effects of insulin (11). However, it is unknown whetherinsulin receptors are overexpressed in cultured human breastcancer cells and if the insulin receptor plays a role in the growthregulation of these cells. In the present study we have quanti-tated insulin receptor content in 3 lines of human breast cancercells in tissue culture. In addition, in these cell lines, we havestudied insulin receptor tyrosine kinase activity and the abilityof insulin to stimulate cell growth via its own receptor.
MATERIALS AND METHODS
Materials
The following materials were purchased: BSA' (radioimmunoassaygrade); bacitracin; phenylmethylsulfonylfluoride; Triton X-100; poly-(Glu-Tyr); hybridization solution; sonicated salmon sperm DNA; dex-tran sulfate; formamide; A/-acetyl-n-glucosamine; wheat germ aggluti-nin-agarose; and porcine insulin were from Sigma Chemical Company(St. Louis, MO). PVC plates were from Becton Dickinson Labware(Oxnard, CA). 125I-Labeled Bolton-Hunter reagent (2200 Ci/mmol),[7-32P]ATP (3000 Ci/mmol), and |«-"P]dCTP (3000 Ci/mmol) wereobtained from Amersham International (Amersham, England). I25I-
Insulm (specific activity, 2200 Ci/mmol) and pHlthymidine (82.3Ci/mmol) were from New England Nuclear (Boston, MA).
MCF-7, T-47D, and ZR-75-1 human breast cancer cell lines (fromDr. I. Perroteau, Turin, Italy) were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, glutamine, non-
essential amino acids, penicillin, and streptomycin (Cell Culture Facility, University of California, San Francisco, CA). 184 nonimmortalizedbreast epithelial cells and 184B5 breast epithelial cells, which are animmortalized but not malignantly transformed human breast cell line(12), were cultured in MCDB-170 media supplemented with bovinepituitary extracts (Clonetic Corporation, San Diego, CA) and otherfactors as previously indicated (12). Insulin receptor monoclonal antibodies were prepared as previously described (13). «-IR3was fromOncogene Sciences, Inc. (Cambridge, MA).
Insulin Receptor Content in Cancer Cell Lines
Evaluation by Radioimmunoassay. To prepare extracts for insulinreceptor radioimmunoassay, cell monolayers were preincubated for 24h in serum-free medium containing 0.1% BSA and 10 Mg/ml transferrinand then harvested with phosphate-buffered saline and 0.2% EDTA,and the cells were counted in a Neubauer chamber. Cells (3 x IO6)were
solubilized in 1 ml of 50 ITIMHEPES buffer, pH 7.4, containing 2mg/ml bacitracin, 1 ITIMphenylmethylsulfonylfluoride. and 1% TritonX-100 for 60 min at 4°Cunder continuous shaking. The solubilized
material was then centrifuged at 10,000 x £,and the supernatant wasfrozen at -80°Cuntil assayed. The DNA content in the cellular extracts
was measured by the method of Labarca and Paigen (14).The insulin receptor radioimmunoassay was performed as previously
described (15) using pure placenta! insulin receptors for a standard (16),labeling them with 125I-Bolton Hunter reagent, and using an anti-insulin receptor polyclonal antiserum (final dilution, 1:100,000). This an-
tiserum reacts with epitopes in both the a and ßsubunits of the receptor( 15). Bound/free radioactivity separation was obtained by precipitationwith goat anti-rabbit antiserum. In this assay cross-reactivity with therelated IGF-I receptor was less than 1%.
Evaluation by I25I-Insulin Binding to Intact Cells. '' I Insulin bind
ing studies were carried out in intact cells grown to confluent mono-layers in tissue culture flasks. Cells were preincubated for 24 h inserum-free medium containing 0.1% BSA and 10 Mg/ml transferrin andthen harvested as described previouly. washed twice with phosphate-buffered saline, and then resuspended in the binding buffer (50 ITIMHEPES, pH 7.8, 120 ITIMNaCl, 1.2 HIMMgSO4, 5 HIMKC1, 15 ITIMsodium acetate, 10 ITIMglucose, 1 ITIMEDTA, 10 mg/ml BSA, and Img/ml bacitracin) at a final concentration of 3 x IO6 cells/ml. Bindingassays were carried out in 12 x 75 mm borosilicate tubes for 16 h at 4°C
(0.5 ml final volume with 40 PM labeled insulin) without or with increasing concentrations of unlabeled insulin. At the end of the incubation, 1 ml of binding buffer at 4°Cwas added, and the tubes were
centrifuged for 10 min at 2400 rpm. Cell pellets were then washed twiceand lysed with 0.03% SDS, and the radioactivity was counted in agamma counter. Nonspecific binding was 0.4-0.6% of total radioactivity and was subtracted. Scatchard plots were resolved into two orders ofbinding sites using the computer software ENZ-FITTER.
Insulin Receptor Kinase Studies
Confluent monolayers of cell cultures were harvested, solubilized,and centrifuged as described above. The soluble material was immediately applied to a 1-ml wheat germ agglutinin agarose column preequil-
ibrated with column buffer containing 150 ITIMNaCl, 0.1% TritonX-100, 1 mM phenylmethylsulfonylfluoride, 2 mg/ml bacitracin, and 50HIMHEPES, pH 7.6. Glycoproteins were eluted with the same buffercontaining 0.3 MA'-acetyl-o-glucosamine. The concentration of the in
sulin receptor in each fraction was assessed by radioimmunoassay (15).The same amount of insulin receptors from each cell line (10 ng each)
was assayed for tyrosine kinase activity toward the exogenous substratepoly(Glu-Tyr) using a specific PVC plate assay (17). The wells of PVCplates were first coated with 50 M!of rabbit anti-mouse IgG
at 4T for 16 h. After 4 washes with washing buffer (50 mMHEPES, pH7.6, 150 HIMNaCl, 0.1 % Triton X-100, 0.1% Tween 20,0.1 % BSA), theplates were incubated at 4°Cfor 16 h with 10 Mg/ml of a specific
monoclonal antibody to the insulin receptor MA-20 (13). (This antibody does not have insulin agonist effects in breast cancer cells.) Thisstep was followed by further incubation with normal mouse IgG (150Mg/ml) at 4°Cfor 16 h. Ten ng of insulin receptors, prepared fromvarious cell lines, were then added to the plates and incubated at 4°Cfor
16 h. After 4 washes with washing buffer, 24 M'of a reaction mixturecontaining 50 mM HEPES, pH 7.6, 150 m\i NaCl, 0.1% Triton X-100,2 mM MnCl2, 10 mM MgClj, and various concentrations of insulinwere added into each well and incubated at 22°Cfor 1 h. The kinasereaction was started by the addition of 3 n\ of poly(Glu-Tyr)(2 mg/ml) and IO M!of [T-12P]ATP (1 MCi/well).
After 1 h at 22°C,10 M' of the reaction mixture were spotted on 3
MM filter paper and precipitated in cold 10% trichloroacetic acid followed by scintillation counting.
Cell Growth Experiments in Monolayer Cultures
I'll] I liMiiiiliiit- Incorporation into DNA. Cells (40 x in i were
plated in 24-well tissue culture plates in their regular growth medium.After 48 h, the medium was removed and replaced by fresh mediumcontaining 0.1% BSA and 10 Mg/ml transferrin, except in T-47D cells,where Dulbecco's modified Eagle's medium supplemented with 5%
charcoal-stripped fetal calf serum and 10 RM progesterone was used(11). Seventy-two h later, variable concentrations of either insulin,MA-5, ora-IR3 were added in fresh medium. After 24 h 0.5 MCi/wellof[*H]thymidine was added for 2 h. Cells were then harvested, and the
rate of DNA synthesis was measured as described previously (18).We also measured DNA synthesis by direct assay of DNA. For this
assay, 5 x IO5cells were seeded in 35-mm multiwell plates and cultured
as previously described. After 48 h, cells were washed twice with 1 ml ofserum-free medium, and 2 ml of fresh serum-free medium were added.
After 24 h insulin was added at the indicated concentrations. Insulinwas then added every other day. Medium was changed every 2 days. Atthe end of a 5-day period the cells were solubilized in 0.03% SDS, andDNA content was determined (14).
DNA and RNA Analysis
Poly(A)+ RNA was extracted from cell monolayers (typically 10"cells) using a new, one-step method as previously described (19).Briefly, adherent cells were released with proteinase K (final concentration, 0.3 mg/ml) and solubilized in 1% sodium dodecyl sulfate.Oligo(deoxythymidine)-cellulose was directly added to the lysate andincubated overnight at 22°C.Poly(A)+ RNA was eluted from oligo-
(deoxythymidine)-cellulose by adding 3 ml of 10 mM Tris with 0.1 mMEDTA and 0.2% SDS. Poly(A)+ RNA (8 Mg)was then electrophoresed
on 1% agarose gel containing 2.2 Mformaldehyde and then transferredto nitrocellulose filters. Poly(A)+ RNA content was normalized by using an oligo(deoxythymidine) probe that was end-labeled using theenzyme T4 polynucleotide kinase (20).
High-molecular-weight DNA from cell pellets was extracted by thephenol-chloroform method, resuspended in a buffer containing 100 mMNaCl, 10 mM Tris, and 10 mM EDTA at pH 8.0, and dissolved by theaddition of 0.5% SDS in the presence of 100 Mg/ml of proteinase K at37°C.DNA was digested with the restriction endonuclease EcoRl under
standard conditions. Twenty Mgof the digested DNA were subjected toelectrophoresis on 0.8% agarose gel, followed by denaturation in NaOHbuffer and immobilization to a nitrocellulose filter as described bySouthern (21). For slot blot experiments nucleic acids were applied tonitrocellulose paper using a Slot Blot Minifold apparatus (Schleicherand Schuell, Keene, NH).
Northern and Southern blot hybridizations were carried out usingtwo human insulin receptor cDNA probes, 18.2 and 13.2 (1 and 4.2kilobases, respectively), a kind gift of Dr. G. I. Bell (University ofChicago). These were labeled with 100 MCi [T-12P]CTP by randomprimers (20) to a specific activity of IO11cpm/Mg. The nitrocellulose
filters from both Northern and Southern blots were prehybridized,hybridized, and washed as previously described (21).
Kb
RESULTS
Insulin Receptor Expression
Insulin Receptor Radioimmunoassay. In order to determinethe insulin receptor content in the breast cancer lines studied,cells were solubilized and the insulin receptor content measuredby a specific insulin receptor radioimmunoassay. Increasingamounts of cellular extracts produced competition-inhibitioncurves that were parallel to the highly purified insulin receptorstandard (Fig. 1). The content of receptors varied from 28.5ng/106 cells in the MCF-7 cells to 4.8 ng/106 cells in the T-47Dcells (Table 1). In the MCF-7 and ZR-75-1 cell lines, the insulinreceptor content was considerably higher than in 184B5 cells[5.3 ±1.7 (SEM) ng/106 cells], a nonmalignantly transformed
human breast epithelial cell line. The insulin receptor contentin T-47D cells (4.8 ±1.2 ng/106 cells) was similar that in 184B5
cells. The nonimmortalized parent cell of 184B5 cells, 184 cells,was also analyzed and had a similar insulin receptor content(4.1 ±0.22 ng/106 cells).
Insulin Receptor Gene Content. To ascertain whether geneamplification may have occurred in breast cancer cell lines,Southern blot analyses were carried out (Fig. 2). DNA wasisolated and digested with the restriction endonuclease EcoRland hybridized with labeled insulin receptor cDNA. No evidence of gene amplification was seen with DNA from MCF-7,ZR-75-1, and T-47D cells when compared with DNA fromnormal breast tissue. Slot blot hybridization was also carriedout with both insulin receptor and /3-actin probes (Fig. 3). Den-sitometric analyses were carried out, and the ratio of insulinreceptor DNA to actin DNA was the same.
100-
° ñCL '~
8 £0)oc ¿s
40
20-
O
MCF-7 1 84 i) !..
Human placestandard
T-47D
10 100
Insulin receptor (ng)
1000
1 O 1 00
Cell number (xl04)
Fig. 1. Radioimmunoassay of multiple dilutions of extracts of nonmalignantbreast epithelial cells (184B5) and breast cancer cells (MCF-7, ZR-75-1, andT-47D).
Table 1 Insulin receptor content in cultured breast cell lines
±0.22" Data are expressed as mean ±SEM of three separate assays.
23.1-
9.4-6.7-
4.4-
2.3-2.0-
W U
M
-IM
-^ I
ocnI
O•¿�nI
CDTO
Fig. 2. Southern blot analysis of insulin receptor DNA. DNA was preparedfrom the three breast cancer cell lines (MCF-7, T-47D, ZR-75-1) and normalbreast tissue. Twenty jig of fcoRI-digested DNA were subjected to agarose gelelectrophoresis followed by transfer to nitrocellulose filters and hybridizationwith labeled insulin receptor cDNA.
Insulin Receptor mRNA Content. In order to assess whetherthe differences in the insulin receptor protein content wereassociated with differences in insulin receptor mRNA levels,poly(A)"1"RNA from the breast cancer cell lines and the non-
malignant breast epithelial cell line were subjected to Northernblot analysis (Fig. 4). In all cell lines except T-47D cells, prominent bands of insulin receptor mRNA were seen at 11.0 and 8.5kilobases, although the ratio between the two bands varied inthe different cell lines (Fig. 5). In T-47D cells only the 8.5-kilobase band was prominent. Except for T-47D cells, whichhad reduced insulin receptor mRNA, there was no major quantitative difference in insulin receptor mRNA content.
IODO Insulin Receptor Functional Studies
Ligand Binding. Studies were next carried out to determinethe functional properties of the insulin receptors in the breastcancer cell lines and the nontransformed breast epithelial cells.First, in intact cells the a subunit function of the insulin receptor was investigated with ligand binding studies (Fig. 6) andanalyzed by Scatchard analysis (22). In all cell lines, bindingwas a curvilinear function, compatible with the presence of twoorders of binding sites (23), a low-affinity-high-capacity siteand a high-affinity-low-capacity site (Table 2). Each cell linehad different binding characteristics (Table 2). Major differences were seen in binding capacity. MCF-7 cells had the highest total binding capacity (high- and low-affinity sites) with avalue of 433 fmol/3 x IO6 cells, whereas T-47D cells had the
Fig. 3. Slot blot analysis of genomic DNA. Five ng of genomic DNA from thethree breast cancer cell lines (MCF-7, T-47D. ZR-75-1) and normal breast tissuewere immobilized on nitrocellulose filter and hybridized with labeled insulinreceptor cDNA or fi-actin cDNA.
Kb
Insulin induced a dose-dependent increase of [3H]thymidine
incorporation in all the cell lines (Fig. 8). In these cell linesinsulin was effective at concentrations of 0.1-1.0 HM.Maximalresponsiveness, however, was over 2-fold greater in MCF-7cells than in other cells. Other studies revealed similar resultswhen DNA content was measured after 5 days of incubationwith insulin (Table 3).
To evaluate potential insulin signaling via the IGF-I receptor,we next investigated the effect of a-IR3 (24), a monoclonalantibody that inhibits binding to the IGF-I receptor, on insulin-stimulated pHJthymidine incorporation (Fig. 9). a-IR3 inhibited 50-60% of the insulin effect in MCF-7 and 184B5cells. In contrast, this antibody did not inhibit [3H]thymidineincorporation in ZR-75-1 and T-47D cells. Normal mouse IgG
(100 n\i) produced only a slight inhibition (10%) of the insulineffect (not shown).
Next we stimulated these cells with MA-5, an insulin agonistmonoclonal antibody to the insulin receptor (25). MA-5 stimulated pHJthymidine incorporation in all cell lines studied. The
1 1.0-8.5-
(0
OT
I
00 I
CDcn
M70I
cnI
Fig. 4. Northern blot analysis of insulin receptor mRNA. Poly(A)+ RNA was
prepared from each cell line (8 vu.) and subjected to agarose gel electrophoresisfollowed by transfer to nitrocellulose filters and hybridization with labeled insulinreceptor cDNA.
lowest capacity of 58 fmol/3 X IO6 cells. The affinities of bothclasses of binding sites were relatively higher in ZR-75-1 andT-47D cells than in MCF-7 and 184B5 cells.
Insulin Receptor Kinase. We next measured the tyrosine ki-
nase activity of the insulin receptors from the cell lines. For thispurpose we used a sensitive and specific plate assay that captures the insulin receptor with an insulin receptor monoclonalantibody (17). The assay measures insulin receptor tyrosinekinase activity and not other tyrosine kinases (17). Insulin receptors were extracted from all cell lines and normalized to 10ng of insulin receptor by radioimmunoassay. In all cells insulinvia the insulin receptor stimulated the phosphorylation of poly-(Glu-Tyr). Interestingly, the insulin-stimulated receptor tyrosine kinase activity in all breast cancer cells was 5-10-foldhigher than in nonmalignant cells. The greatest increase ininsulin receptor tyrosine kinase activity was observed in ZR-75-1 cells (Fig. 7).
Insulin Effect on Cell Growth
In order to investigate the mitogenic effect of insulin on thesecell lines we studied [3H]thymidine incorporation into DNA.
Fig. 5. Bar graphs of Northern blot densitometry from Fig. 4.
0.15
•¿�= 0.05
1 50
Insulin bound (fmole/3x106 cells)
200
Fig. 6. Scatchard plots of insulin binding to human breast epithelial cells (184B5) and human breast cancer cells (MCF-7, ZR-75-1. T-47D). Data are correctedfor nonspecific binding (0.4-0.6% of total radioactvity for all cell lines).
greasiest effect of MA-5 was observed in the breast cancer celllines. 184B5 cells showed the least stimulation (Fig. 8). Normalmouse IgG (100 HM)produced little or no stimulation of [3H]-
thymidine incorporation (not shown).
DISCUSSION
Previously we studied insulin receptors in 159 human breastcancer specimens (6). The expression of these receptors washeterogeneous. The insulin receptor content in these specimensranged from 1.16 to 23.77 ng/0.1 mg protein, but the meanvalue, 6.15 ng/0.1 mg protein, was 6-fold higher than the meanvalue found in normal breast tissues, 0.96 ng/0.1 mg protein.The insulin receptor mean content in human breast cancer specimens was also higher than that observed in any normal humantissue, including liver. Immunostaining analysis revealed thatthe insulin receptor content of breast cancer specimens was dueto an increased receptor content of the malignant breast epithelial cells. We also documented that the insulin receptors inbreast cancer specimens were functional: insulin binding wasincreased in proportion to insulin receptor content as measuredby radioimmunoassay, and insulin receptor kinase activity waspresent and responsive to insulin.
In the present study, we found that insulin receptor expression in human cultured breast cancer cells in vitro was variable,as it was in the breast cancer specimens in vivo. In two of thethree cancer cell lines studied (MCF-7 and ZR-75-1), insulinreceptor expression was increased 6- and 3-fold relative to both
the nonmalignantly transformed human breast cells, 184B5,and their nonimmortalized parent 184 cells. Analysis of theinsulin receptor gene indicated that the increased insulin receptor content in MCF-7 and ZR-75-1 cells was not due to geneamplification, since no differences in insulin receptor gene copynumber were observed among the different breast cell linesexamined.
When insulin receptor mRNA was studied, no major difference was observed in the specific mRNA content of MCF-7 andZR-75-1 cells with respect to the nonmalignantly transformedbreast cell line 184B5. In the malignant (MCF-7 and ZR-75-1)and nonmalignant (184B5) cells, both the 8.5- and 11-kilobaseinsulin receptor mRNA bands were prominent even if the ratioof the two bands varied. As previously reported, in T47-D cellsonly the 8.5-kilobase band was prominent; the 11-kilobase bandwas observed only after progesterone treatment (11). In mosthuman and animal cells that have been studied, several insulinreceptor mRNA species are observed, with a size ranging widelyfrom 5.2 to 11 kilobases (3). Studies of these multiple transcripts suggest that they are due to variable splicing at the 3' endof the insulin receptor RNA (3, 26). There are also multiple 5'
start sites, but they differ by only several hundred bases (27) andthus cannot account for such size heterogeneity. Most likely, allof these multiple mRNA species are involved in insulin receptorprotein synthesis, since all are found to be associated with ri-
bosomes (28). However, the biological role for these multipletranscriptions is unknown, and, therefore, their possible role inthe enhanced insulin receptor expression in human breast cancer cells cannot be excluded. In addition to the observedchanges in mRNA quality, an enhanced translation of the insulin receptor mRNA or an enhanced stability of the insulinreceptor protein may also be responsible for the relatively highinsulin receptor content in human breast cancer cells.
Insulin receptors were functional in the different cell lines. Inagreement with insulin receptor content as measured by radioimmunoassay, insulin binding capacity was highest in MCF-7and ZR-75-1 cells and lowest in T-47D and 184B5 cells. Thehigher receptor affinity observed in ZR-75-1 and T-47D cellscould be due to changes in either posttranslational modificationof insulin receptor or the plasma membrane environment (3).
We also studied the function of the insulin receptor ßsubunitwith a tyrosine kinase plate assay (17) which separates insulinreceptor kinase activity from the kinase activities of other receptors, including the closely related IGF-I receptor. Using thisassay, with insulin receptors from all cell lines, we found thatinsulin stimulated tyrosine phosphorylation of the substratepoly(Glu-Tyr). Insulin-stimulated insulin receptor tyrosine kinase activity was higher in MCF-7 and ZR-75-1 cells than inT-47D and 184 B5 cells. This observation suggests that heterogeneity exists in insulin receptors in the different cell lines.This observation is not surprising, since heterogenity of insulinreceptor a and ßsubunits exists between various human tissues(29). One explanation for this heterogeneity is differences in theglycosylation pattern of the insulin receptors.
A mitogenic effect of insulin was observed in all breast cancercell lines, insulin being effective at physiological concentrationsof 1 nM or less. When we used «-IR3,an IGF-I receptor-blocking antibody, it was without effect on insulin-stimulated [3H]-thymidine incorporation in T-47D and ZR-75-1 cells. In contrast, it partially inhibited the insulin effect on this function inMCF-7 and 184B5 cells. This study suggested therefore that in
•¿�ov4->"
ss•¿�H
S*"•Oi
" ö>
16 -
12 -
8-
4-
0J
ZR-75-1
Insulin (M)Fig. 7. Phosphotransferase activities of the semipure insulin receptors in vitro.
Semipure insulin receptors from ZR-75-1, MCF-7, T47D, and 184B5 cells wereprepared by wheat germ agglutin chromatography. Insulin receptors (10 ng basedon radioimmunoassay) were then preincubated with various concentrations ofinsulin in PVC plates preabsorbed with the specific anti-insulin receptor monoclonal antibody MA-20. The reaction was started by the addition of an exogenoussubstrate, poly(Glu, Tyr). and I >iCi [-y-32P]ATP. The radioactivity incorporatedinto poly(Glu. Tyr) was then measured by liquid scintillation. The data are expressed as pmol of ATP incorporated into 1 mg of poly(Glu, Tyr). A representative of two separate experiments is shown. Data are mean ±SD of quadruplicatedeterminations.
Fig. 8. Stimulation of [3H]thymidine incorporation by insulin and insulin receptor agonist monoclonal antibody MA-5 in human breast epithelial cells and humanbreast cancer cells. Cells were plated in tissue culture plates in their regular growth medium. After 48 h medium was replaced with serum-free medium except for T-47D,for which 5% charcoal-stripped fetal calf serum and 10"* Mprogesterone-containing medium was used. After a further 72 h the cells were stimulated with either insulinor MA-5 for 24 h. Thymidine incorporation was performed during the last 2 h of stimulation by adding 0.S nCi ['Hlthymidine. Cells were then harvested, and the raleof |'H|thymidinc incorporation was determined as previously described (12). Normal mouse IgG at 100 n%iproduced slight stimulation (5-10%) of |*H)thymidine
incorporation. Each value is the mean ±SD of two separate experiments performed in triplicate.
i-H.2
¿0)JJCC«JO•HM U
•¿�O0Thymi incorp»over•r140
-120
-100
•¿�80
-60
-40
-20
-0
-
-75-1
184 B5
10
alpha IR3 (M)
1 0
Fig. 9. Effect of IGF-I receptor antagonist monoclonal antibody u-IR3 on| '! I |iluniiilim- incorporation stimulated by insulin in human breast epithelial and
human cancer cells. Cells were cultured in tissue culture plates as described aboveand stimulated for 24 h with I mi insulin in the absence and presence of differentconcentrations of «-IR3. ['HIThymidine incorporation was performed as described above. The amount of |'H]thymidine incorporated in the presence ofinsulin or MA-5 is represented as a percentage of the amount of ['H|thymidine
incorporated in the absence of insulin. Normal mouse IgG at 100 mi produced aslight inhibition (10%) of ['Hlthymidine incorporation. Each value is the mean ±SD of two separate experiments performed in triplicate.
MCF-7 and 184B5 cells, the insulin mitogenic effect was mediated partially via the insulin receptors and partially via theIGF-I receptors, whereas in T-47D and ZR-75-1 cells, insulinacted only via its own receptors. Moreover, MA-5, an anti-insulin receptor-specific monoclonal antibody and an insulinagonist, stimulated DNA synthesis in all cells. The mitogenicresponse to insulin receptor stimulation by MA-5 was greater inthe breast cancer cells than in the nonmalignant breast cells.These studies strongly suggested a role for insulin receptor in
Table 3 Effect of insulin on DNA synthesis in human breast cancer cell linesand in a nonmalignantly transformed human areas! epithelial cell line
Data are expressed as ng DNA ±SD for triplicate determinations of a representative experiment.
CellsInsulin
(mil0
0.11.010
100MCF-728.2±3.7
33.8±241.7±1.569.0±582.3±12T-47D31.5±6.5
34.6±441.5±546.6±759.8±8.2ZR-75-130.0±2
36.5±1.539.6±445.0±260.0±10184B534.0±1.8
34.6±0.246.2±850.0±363.0±6
the growth regulation of breast cancer cells. Moreover, thegreater mitogenic effect of insulin via the insulin receptor inbreast cancer cells when compared to nonmalignant breast cellsis in concert with the observation that breast cancer cells havegreater insulin-stimulated receptor tyrosine kinase activity. Interestingly, in MCF-7 cells Cullen et al. (30) did not observeinhibition by a-IR3 of the insulin mitogenic effect.
The biological significance of increased insulin receptor expression in breast cancer epithelial cells is unknown. Humanand animal breast carcinomas have receptors for steroid andpeptide hormones, and both in vitro and in vivo studies indicatethat their growth is at least partially hormone dependent (31-35). Various studies have demonstrated that breast carcinomascontaining high concentrations of estrogen and progesteronereceptors have a better prognosis. However, some patients withbreast carcinomas containing high concentrations of estrogenand progesterone receptors do not show a beneficial response to
hormonal therapy (36). One possible explanation for this discrepancy is that other classes of hormones such as polypeptidehormones and growth factors, either by themselves or in combination with steroid hormones, may stimulate tumor growth(31, 37). Insulin regulates the growth and metabolism of animalbreast cancer cells both in vivo and in vitro (38-40) and humanbreast cancer cells in vitro (9, 10). It is likely, therefore, that insome breast cancers insulin, either alone or in combination withsteroid and other related hormones and growth factors (41-42),plays a role in promoting their growth.
ACKNOWLEDGMENT
We thank Kirsten H. Scriven for technical assistance.
REFERENCES
1. Hanks. S. K., Quinn, A. K., and Hunter. T. The protein kinase family:conserved features and deduced phylogeny of the catalytic domains. Science(Washington DC). 241: 42-53, 1988.
2. Ullrich. 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., and Ramachandran, J. Human insulinreceptor and its relationship to the tyrosine kinase family of oncogenes.Nature (Lond.), 313: 756-761. 1985.
3. Goldfine, I. D. The insulin receptor: molecular biology and transmembranesignaling. Endocr. Rev., 8: 235-255, 1987.
4. Strauss, D. S. Growth-stimulatory actions of insulin in vitro and in vivo.Endocr. Rev., 5: 356-369, 1984.
5. Whittaker, J., Okamoto, A. K., Thys. R., Bell, G. I., Steiner, D. F., andHofmann, C. A. High-level expression of human insulin receptor cDNA inmouse NIH 3T3 cells. Proc. Nati. Acad. Sci. USA, 84: 5237-5241, 1987.
6. Papa, V., Pezzino, V., Costantino, A., Belfiore, A., Giuffrida, D., Frittitta. L.,Vannelli. G. It,, Brand, R., Goldfine, I. D.. and Vigneri, R. Elevated insulinreceptor content in human breast cancer. J. Clin. Invest., 85: 1503-1510,1990.
7. Giorgine. F., Belfiore, A., Milazzo, G., Costantino, A., Maddux, B., Whit-taker, J., Goldfine, I. D., and Vigneri, R. Overexpression of insulin receptorin fibroblasts and ovary cells induces a ligand-mediated transformed pheno-type. Mol. Endocrinol. 5: 452-459, 1991.
8. Osborne, C. K.. Bolán.G., Monaco, M. E., and Lippman, M. E. Hormoneresponsive human breast cancer in long-term tissue culture: effect of insulin.Proc. Nati. Acad. Sci. USA, 73: 4536-4540. 1976.
9. Osborne, C. K., Monaco, M. E., Lippman, M. E., and Kahn, C. R. Correlation among insulin binding, degradation, and biological activity in humanbreast cancer cells in long-term tissue culture. Cancer Res.. 38:94-101, 1978.
10. Van der Burg, B., Rutterman. G. R., Blankenstein, M. A., De Laat, S. W.,and van Zoelen, E. J. J. Mitogenic stimulation of human breast cancer cellsin a growth factor-defined medium: synergistic action of insulin and estrogen.J. Cell. Physiol.. 134: 101-108. 1988.
11. Papa, V., Reese. C. C, Brunetti, A., Vigneri, R., Siiteri, P. K.. and Goldfine,I. D. Progestins increase insulin receptor content and insulin stimulation ofgrowth in human breast cancinoma cells. Cancer Res., 50: 7858-7862, 1990.
12. Stampfer, M. R., and Bartley, J. C. Induction of transformation and continuous cell lines from normal human mammary epithelial cells after exposureto benzo(a]pyrene. Proc. Nati. Acad. Sci. USA. 82: 2394-2398, 1985.
13. Forsayeth, J. R., Montemurro. A., Maddux, B. A., DePirro, R., and Gold-fine, I. D. Effect of monoclonal antibodies on human insulin receptor auto-phosphorylation, negative cooperativi!) and down regulation. J. Biol. Chem.,262: 4134-4140, 1987.
14. Labarca, C., and Paigen. K. A simple, rapid, and sensitive DNA assay procedure. Anal. Biochem.. 102: 344-352. 1980.
15. Pezzino, V., Papa, V., Trischitta. V., Brunetti, A., Goodman, P. A., Treute-laar, M. K., Williams, J. A., Maddux, B. A., Vigneri, R., and Goldfine, I. D.Human insulin receptor radioimmunoassay: applicability to insulin-resistantstates. Am. J. Physiol.. 257: E451-E457, 1989.
16. Forsayeth, J., Maddux, B. A., and Goldfine, I. D. Biosynthesis and processing of the human insulin receptor. Diabetes, 35: 837-846. 1986.
17. Steele-Perkins, G.. and Roth, R. A. Insulin-mimetic anti-insulin receptor
18. Furlanetto, R. W., and Di Carlo, J. N. Somatomedin-C receptors and growtheffects in human breast cancer cells maintained in long term tissue culture.Cancer Res., 44: 2122-2128, 1984.
19. Hartmann, K. K. P.. Papa. V., Brown, E. J.. Doerries, U., Rosenthal. S. M.,and Goldfine, I. D. A rapid simple one-step method for isolation of poly(A)+RNA from cells in monolayer. Endocrinology, 127: 2038-2040, 1990.
20. Sambrook, J., Fritsch, E. F., and Maniatis. T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory,1989.
21. Southern, E. M. Detection of specific sequences among DNA fragmentsseparated by gel electrophoresis. J. Mol. Biol., 98: 503-517, 1985.
22. Scatchard, G. The attraction of proteins for small molecules and ions. Ann.N.Y. Acad. Sci., 51: 660-672, 1949.
23. Kahn, C. R. Membrane receptors for polypeptide hormones. Methods Mem-br. Biol., 3: 81-146, 1975.
24. Kull. F. C, Jacobs, S., Su, Y. F., Svoboda, M. E., Van WyK, J. J., andCuatrecasas, P. Monoclonal antibodies to receptors for insulin and so-matomedin C. J. Biol. Chem., 25«:6561-6566, 1983.
25. Brunetti. A., Maddux. B. A.. Wong, K. Y.. Hoffmann, C., Whittaker, J.,Sung. C.. and Goldfine, I. D. Monoclonal antibodies to the human insulinreceptor mimics a spectrum of biological effects in transfected 3T3/HIRfibroblasts without activating receptor kinase. Biochem. Biophys. Res. Commun., 165: 212-218, 1989.
26. Tewari, D. S., Cook, D. M.. and Taub, R. Characterization of the promoterregion and 3' end of the human insulin receptor gene. J. Biol. Chem., 264:16238-16245, 1989.
27. Mamula, P. W., Wong, K. Y., Maddux, B. A., McDonald, A. R., and Gold-fine, I. D. Sequence and analysis of the promoter region of the human insulinreceptor (HINSR). Diabetes, 37: 1241-1246, 1988.
28. McDonald, A. R., and Goldfine. I. D. Glucocorticoid regulation of insulinreceptor gene transcription in IM-9 cultured lymphocytes. J. Clin. Invest., 81:499-504, 1988.
29. Caro, J. F., Raju, S. M., Sinha, M. K.. Goldfine, I. D., and Dohm, G. L.Heterogeneity of human liver, muscle and adipose tissue insulin receptor.Biochem. Biophys. Res. Commun., 151: 123-129. 1988.
30. Cullen, K. J., Yee, D., Sey, W. S., Perdue, J., Hampton, B., Lippman, M. E.,and Rosen, N. Insulin-like growth factor receptor expression and function inhuman breast cancer. Cancer Res., 50: 48-53, 1990.
31. Dickson. R. B.. and Lippman, M. E. Estrogenic regulation of growth factorsecretion in human breast carcinoma. Endocr. Rev., 8: 29-43, 1987.
32. Engel, L. W.. and Young, N. A. Human breast carcinoma cells in continuousculture: a review. Cancer Res., 38: 4327-4339, 1989.
33. Benson, E. A., and Holdaway, I. M. Regulation of insulin binding to humanmammary carcinoma. Cancer Res., 42: 1137-1141. 1982.
34. Harmon, J. T.. and Hilf. R. Identification and characterization of the insulinreceptor in the R3230AC mammary adenocarcinoma of the rat. Cancer Res.,36: 3993-4000, 1976.
35. Wittliff. J. L. Steroid hormone receptors in breast cancer. Cancer (Phila.), 53:630-643, 1984.
36. Maas, H., Jonat, W., Stolzenbach, G., and Trams, G. The problem of non-responding estrogen receptor-positive patients with advanced breast cancer.Cancer (Phila.), 46: 2835-2837. 1980.
37. Sporn. M. B.. Roberts, A. B., and Driscoll, J. S. Principles of cancer biology:growth factor and differentiation. In: V. T. De Vita, S. Hellman, and S. A.Rosenberg (eds.). Cancer Principles and Practice of Oncology, pp. 49-65.Philadelphia: J. B. Lippincott Co.. 1985.
38. Heuson, J. C., Coune, A., and Heiman, R. Cell proliferation induced byinsulin in organ culture of rat mammary carcinoma. Exp. Cell Res., 45:351-360, 1967.
39. Shafie, S. M., and Hilf, R. Relationship between insulin and estrogen bindingto growth response in 7-12 dimethylbenz(a)-anthracene-induced rat mammary tumors. Cancer Res., 38: 759-764, 1978.
40. Cohen, N. D., and Hilf, R. Influence of insulin on growth and metabolism of7, 2-dimethylbenz(a)anthracene-induced mammary tumors. Cancer Res., 34:3245-3252, 1974.
41. Dickson, R. B., and Lippman, M. E. Estrogenic regulation of growth andpolypeptide growth factor secretion in human breast carcinoma. Endocr.Rev., 8: 29-43, 1987.
42. Clarke. R., Brunner, N., Katz. D., Glanz, P.. Dickson, R. B., Lippman, M. E..and Kern, F. G. The effects of a constitutive expression of transforminggrowth factor-a on the growth of MCF-7 human breast cancer cells in vitroand in vivo. Mol. Endocrinol., 3: 372-380, 1989.
