[CANCER RESEARCH 48. 6691-6696. December 1. 1988]

Insulin-like Growth Factor II mRNA Expression in Human Breast Cancer

Douglas Yee1, Kevin J. Cullen1, Soonmyoung Paik1, James F. Perdue, Brian Hampton, Arnold Schwartz,Marc E. Lippman1, and Neal Rosen1

Medical Breast Cancer Section National Cancer Institute, Bethesda, Maryland[D. Y., K. J. C., S. P., M. E. L., N. R.]; Holland Research Labs, American Red Cross,Rockville, Maryland [J. F. P., B. H.J; and the Department of Pathology, George Washington University Medical Center, Washington, DC ¡A.SJ

ABSTRACT

Insulin-like growth factor II is a growth factor important in fetaldevelopment. Several cancer tissues and cell lines have been reported toexpress IGF-II and rat IGF-II is mitogenic for breast cancer cell lines.Using Northern analysis and ribonuclease protection assays, IGF-IImRNA was detected in normal fibroblasts and in the established breastcancer cell line, T47D. In this cell line, steady state levels of IGF-IImessage were increased by treatment with est rail in). 10 nM IGF-II,purified from human serum, was mitogenic for breast cancer cell lines.In vitro, IGF-II may act as an autocrine growth factor for some cell lines.

RNA derived from breast cancer, pathologically normal breast tissue,and benign breast disease also contained IGF-II mRNA. When pairedsamples of normal and cancer tissue were obtained from the breast of thesame patient, the level of IGF-II mRNA expression in the normal tissuewas at least that found in the cancer. This is consistent with previousobservations that show IGF-II is expressed in mesenchyme.

These findings suggest that in breast cancer IGF-II is produced bystromal tissue elements and potentially by the malignant epithelial cells.Therefore, IGF-II may function as an autocrine or a paracrine growthfactor in different breast tumors.

INTRODUCTION

Autocrine stimulation of growth has been theorized to beimportant in malignant transformation (1). Breast cancer is ahormonally responsive tumor and has been associated withseveral putative autocrine and/or paracrine growth factors,including transforming growth factor a (2), platelet-derivedgrowth factor (3), and cathepsin D (4). Insulin-like growth

factor I is mitogenic for breast cancer cell lines and mediaconditioned by these cell lines contain an immunoreactive IGF-I2-related protein (5). Therefore IGF-I related protein may bean autocrine growth factor for breast cancer. Since IGF-I andinsulin-like growth factor II are homologous proteins that canboth bind to the type I receptor (6), we have examined breastcancer cell lines and tissues to determine if IGF-II is also apotential autocrine growth factor in breast cancer.

Insulin-like growth factor II is a mitogenic polypeptideclosely related to insulin and IGF-I (7). Although the physiologic roles for insulin and IGF-I are identified, the functions ofIGF-II are not well known. Expression of IGF-II is high in thefetal rat and falls at birth (8, 9). In the human fetus, in situhybridization histochemistry has localized IGF-II mRNAexpression to the mesenchymal tissues (10). Although adultliver and kidney express IGF-II mRNA, the level of messageabundance is 10- to 40-fold less than in the fetus (11). Theseobservations have lead to the belief that IGF-II is important infetal growth and development; the function of IGF-II in the

adult is not known.IGF-II expression has also been noted in several tumors.

Wilms' tumor and hepatoblastoma express high levels of IGF-

Reeeived 5/4/88; revised 8/23/88; accepted 8/29/88.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 inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

' Current address: Lombard: Cancer Research Center, Georgetown UniversityMedical Center, 3800 Reservoir Road, N. W., Washington. DC. 20007.

2The abbreviations used are: IGF-I, insulin-like growth factor I; IGF-II,insulin-like growth factor II; IMEM, improved minimum essential medium.

II mRNA (11). In addition breast cancer cell lines, pheochro-mocytoma tissue, colon cancer tissues, and some sarcomas havebeen reported to express IGF-II mRNA (12-14). Rat multiplication-stimulating activity-Ill (rat IGF-II) has been found to bemitogenic for the breast cancer cell line T47D (15). This suggests that IGF-II could potentially act as an autocrine growthfactor in breast cancer.

To evaluate the role of IGF-II in breast cancer, we have usedNorthern analysis and the ribonuclease (RNase) protectionassay to evaluate the expression of authentic IGF-II mRNA inbreast cancer cell lines and tissues. We have also examined themitogenicity of purified human IGF-II protein in a serum-freein vitro system. We have found that the estrogen responsivebreast cancer cell line, T47D, expresses IGF-II mRNA. In thiscell line, the message is increased by estrogen treatment. 10 nMIGF-II is mitogenic for breast cancer cells in culture. Therefore,IGF-II may be an autocrine growth factor for these cells.

RNAs derived from breast cancer tissue, normal breast tissueand benign breast disease contain IGF-II message. Since pairednormal and cancer RNAs both express IGF-II mRNA at equivalent levels, this suggests that the stromal elements of the tissuemay also express IGF-II and in some tumors IGF-II may actas a paracrine growth factor in vivo.

MATERIALS AND METHODS

Cells and Cell Culture. MCF-7 cells were originally obtained fromDr. Marvin Rich, Michigan Cancer Foundation, Detroit, MI (16). Thecells used in most experiments were early passage number 30-40,however some later passage subclones were also tested. ZR75B wasinitially established at the National Cancer Institute (17) and furthersubcloned by our laboratory. Hs578T was a gift from Dr. AdelineHackett, Peralta Cancer Research Institute, Oakland, CA (18). Fibro-blast cell lines were obtained from the Coriell Research Institute,Camden, NJ. All other cell lines were obtained from the AmericanType Culture Collection, Rockville, MD.

All carcinoma cells were maintained in Improved minimum essentialmedium (Biofluids, Rockville, MD) with 10% fetal calf serum (GIBCO,Detroit, MI). Cells used for estrogen regulation experiments wereroutinely grown in 5% fetal calf serum stripped of estrogen by sulfataseand dextran-coated charcoal treatment (19). Fibroblasts were maintained in Eagle's minimum essential medium (Biofluids) with 20% fetal

calf serum (GIBCO) and twice normal concentrations of nonessentialamino acids, glutamine and vitamins (GIBCO). All media containedpenicillin (100 U/ml) and streptomycin (100 mg/ml).

Mitogenicity Studies. IGF-II was isolated from human Cohn fractionIV as previously described (20). The final product was homogeneous asevidenced by sodium dodecyl sulfate-polyacrylamide gel electrophoresisand jV-terminal amino acid sequence analysis. The ¿V-terminalaminoacid sequence analysis demonstrated that the material was greater than99% human IGF-II. Amino acid composition analysis was employedto determine the molar concentration of IGF-II used in the mitogenicity

experiments.MCF-7 and T47D cells were plated in 0.5 ml IMEM with 5%

sulfatase treated, charcoal stripped calf serum in 24 u dl culture plates.MCF-7 cells were plated at a density of 25,000 per well and T47D cellswere plated at 50,000 per well. After 24 h, the medium was removedand replaced with serum free medium consisting of IMEM with 2 mg/liter fibronectin, 2 mg/liter transferrin, 20 mM 4-(2-hydroxyethyl)-l-

6691

Research. on December 9, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

IGF-II AND BREAST CANCER

piperazineethanesulfonic acid buffer, 292 mg/liter glutamine, and traceelements (Biofluids). After an additional 24 h, this was replaced withserum free medium containing either no additives, l UMestradici, 5nM IGF-I, 1 nM IGF-II, or 10 nM IGF-II. Each additive was studied induplicate wells. Media and additives were changed at 48 h, and cellswere harvested in PBS (GIBCO) with 0.02% EDTA after the additionof growth factor containing media at 0,48, and 96 h. Cells were countedvisually with a hemocytometer.

Nucleic Acid Preparation. Fresh tissue was obtained in the operatingroom and grossly cancerous tissue was dissected away from normalappearing tissue. The specimens were divided, and one portion wassubmitted for pathological exam. The other portion was frozen on dryice and then stored at -70°Cuntil RNA was prepared.

Cells used for estrogen regulation experiments were passaged inmedia containing 5% sulfatase treated, charcoal stripped calf serum,and allowed to attach for 24 h. Then the cells were washed three timeswith PBS (GIBCO) and refed with phenol red free, serum free media.24 h later, 17/3-estradiol (Sigma, St. Louis, MO) was added to the cellmedia in the concentrations noted. In some experiments, phenol red(0.01 g/liter Spectrum, Gardenia, CA) was added to the medium. Cellswere harvested at the indicated time points.

Total tissue and cellular RNAs were prepared using the guanidinium/cesium chloride method (21). Tissue was first pulverized on dry ice andthen homogenized with the Polytron (Brinkmann, Westbury, NY) in 4M guanidinium thiocyanate solution. Cells were either harvested directly in guanidinium or pelleted for later processing. All RNA concentrations were determined spectrophotometrically. The concentrationand integrity of the RNA were confirmed on denaturing 1.1% agaroseminigels containing 7.2% formaldehyde prior to use. Poly(A)+ RNA

was selected using oligo(dT) chromatography (Boehringer-Mannheim,Indianapolis, IN) as previously described (22).

Tissue and cellular DNAs were prepared as described (23).Probes. The IGF-II probe was a gift from Graeme Bell (24). The

854-basepair Pstl-Pstl fragment was cloned in pGem4 (Promega, Madison, WI), a vector with both T7 and SP6 promotor sequences. Theplasmid was linearized with either Acci or Rsal prior to transcription.This resulted in probe sizes of 292 and 336 basepairs, respectively (Fig.1).

Ribonuclease Protection Assay. RNA probes were synthesized according to the instructions of the manufacturer (Promega). LinearpGem4-IGFII was transcribed with T7 polymerase. DNase I (Promega)was added to digest the DNA template.

Hybridization was carried out as previously described (25). 20 to 60Mgof RNA was hybridized with 5 x IO4cpm of probe in 30 ß\of buffer(80% formamide, 40 mM piperazine-W,./V"-bis(2-ethanesulfonic acid),0.4 M NaCl, 0.1 M EDTA) for 12-16 h at 50°C.Samples were then

digested with RNase A (Sigma) at a concentration of 40 fig/ml for 30min at 2S°C.15 min of Proteinase K (Sigma) and sodium dodecyl

sulfate (BRL, Bethesda, MD) treatment was used to terminate theRNase digestion. The samples were extracted once with phenol-chlo-roform-isoamyl alcohol (20:20:1) and then were precipitated with l /igtRNA (Sigma) and two volumes of absolute ethanol. The pellets wereresuspended in 5 M'of an 80% formamide loading buffer, boiled, thenelectrophoretically fractionated on a 6% polyacrylamide gel containing8 M urea. Size markers were prepared by digesting pBR322 with Afspl(New England Biolabs, Beverly, MA) and the products were end-labeledwith "P dCTP (NEN, Wilmington, DE) and Klenow fragment (Boehringer-Mannheim). The gel was dried and exposed to X-ray film in the

A. 5'

B.

C.

H3'

Ace II—

Rsa I

Pst I—¿�I

Pst IFig. 1. IGF-II probes used in RNase protection assay. A, IGF-II cDNA, boxed

area represents coding portion of the pre-pro-hormone (IH)coding portion ofmature circulating IGF-II P); B, 292-basepair Acc\-Pst\ probe; C, 336-basepairRsa\-Pst\ probe.

presence of a Quanta III (Kodak, Rochester, NY) intensifying screenfor 12 h to 1 week.

Northern and Southern Analysis. Heat denatured total cellular RNAor poly(A)'1"selected RNA was electrophoretically size fractionated on

a 1.1% agarose gel containing 7.2% formaldehyde. RNA was transferred to nitrocellulose (Schleicher and Schuell, Keene, NH) or nylon(Nytran, Schleicher and Schuell) membranes by capillary action. Probeswere labeled by random priming (Boehringer-Mannheim) to a specificactivity of 2-5 x IO8 cpm/iig DNA and hybridization was in 50%

formamide, 450 mM NaCl, 45 mM sodium citrate, 100 Mg/ml tRNA,and 10% dextran sulfate at 42°Cfor 12-16 h. Blots were washed at

room temperature in 300 mM NaCl, 30 mM sodium citrate for 1 h,followed by a final wash in 15 mM NaCl, 1.5 mM sodium citrate at62°Cfor 15-30 min. Blots were exposed to X-ray film and a QuantaIII (Kodak) intensifying screen at -70°Cfor 4-10 days.

10 jig of DNA were digested completely with ////¡dill, BamHl, or£coRI, (Boehringer-Mannheim), fractionated by electrophoresis, denatured, neutralized, and transferred to nylon membranes (Nytran,Schleicher and Schuell) by capillary action. Hybridization and washingwere carried out under the same conditions used for Northern analysis.Blots were exposed to X-ray film and a Quanta III (Kodak) intensifyingscreen at -70°Cfor 16 h.

RESULTS

Northern Blot Analysis of IGF-II mRNA. Northern blotanalysis of poly(A)+ mRNAs from several breast cancer celllines was used to detect IGF-II expression. MCF-7, MDA MB231, ZR75B, and Hs578T did not express IGF-II mRNA, onlyT47D expressed IGF-II message (Fig. 2A and data not shown).Fig. 2A shows that RNA from this cell line reveals three bandswith sizes of 6.0, 4.8 and 0.5 kilobases. IGF-II message wasalso detected in RNA isolated from breast cancer tissues (Fig.2B). The transcript sizes were similar to those seen in T47D,however in Patient C. R. the 6.0-kilobase transcript was lessabundant than in Patient J. P. HepG2 is a cell line derivedfrom a hepatocellular carcinoma and expresses high levels of

m

A.

6.0 —¿�

4.8 —¿�

Ûr«.co12ŒN

CL—¿�3

(D

BCOCL

rro

(BCL

SI

0.5 —¿�

I —¿�6.0....

—¿�0.5

Fig. 2. Northern blot analysis of IGF-II message in breast cancer cell linesand tissue. Blots were hybridized with the 854-basepair Pstl-Pstl fragment ofphigf2. A, 5 jig poly(A)+ selected mRNA from breast cancer cell lines. Radiograph was exposed for 7 days; B, 10 ¿igof total RNA from two patients andHepG2. Radiograph was a 4-day exposure.

6692

Research. on December 9, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

IGF-II AND BREAST CANCER

IGF-II mRNA (13). HepG2 RNA demonstrated the 6.0- and4.8-kilobase transcript sizes, but did not express the smaller0.5-kilobase message (Fig. 2B).

In normal tissues, IGF-II mRNA is transcribed from twodifferent 5' untranslated exons; in the fetal tissues a 6.0-kilo-

base transcript is seen, while in adult liver the IGF-II mRNAis 5.3-kilobase (26). Irminger et al. has found that expressionof the different 5' sequences are also tissue specific (27). Fur

thermore, the 4.8-kilobase band noted on Northern blot analysismay reflect hybridization of the IGF-II probe to 28S ribosomalRNA. Total cellular RNA from breast cancer cell lines andtissues generally displayed a prominent 4.8-kilobase band andpoly(A+) RNA occasionally demonstrated the 4.8-kilobase

band. This suggests that this band may represent hybridizationof the IGF-II probe to ribosomal RNA. The origin of the smaller0.5-kilobase band is uncertain.

Because Northern analysis of mRNA with the IGF-II cDNAprobe may yield several transcript sizes that could representcross hybridizing species, we turned to RNase protection assaysto probe for authentic IGF-II message.

Ribollili lease Protection Assays for IGF-II mRNA. We usedsolution hybridization of a 32P-labeled antisense IGF-II RNA

probe to cellular RNA followed by incubation of the productswith ribonuclease A to detect authentic IGF-II mRNA. Thistechnique has been used to detect single nucleotide mismatchesin mRNA (28). RNA hybrids between the probe and cross-reacting mRNA species are sensitive to digestion by the nu-clease. IGF-II cDNA probes hybridize with ribosomal RNAand IGF-II shares sequence homology with IGF-I, insulin, andrelaxin; RNase protection assays can eliminate detection ofthese messages.

Two probes were used in the RNase protection assay, bothprobes include the coding regions for the entire D and Edomains of the IGF-II protein, as well as 41 basepairs of the3'-untranslated region (Fig. IA). The Accl-Pstl probe (Fig. IB)

used in the initial studies yielded a consistent protected fragment of the exact size as the probe (Fig. 3, tRNA lane). Thisfragment arose from anomalous transcription of sense RNAwhile synthesizing the antisense probe. Prolonged exposures ofthe autoradiograph allowed detection of this larger sense probe(Data not shown). The fragment protected by hybridization ofthe probe to IGF-II mRNA was 24 basepairs smaller than the

probe due to the digestion of linker sequences transcribed fromthe pGEM vector. This authentic protected signal was easilydetected in HepG2 (Fig. 3) or liver RNA (Fig. 6/1) sample, andwas not apparent in lanes hybridized with tRNA. Later studiesused an Rsa\-Pstl probe (Fig. 1C) and the artifactual protectedfragment was eliminated (Fig. 5/1, lane 2).

Often breast cancer cell lines examined, only T47D expressedauthentic IGF-II message (Fig. 3). Some late passage subclonesof MCF-7 were found to express low levels of IGF-II message(Data not shown). Early passage MCF-7 cells growth in serumfree media, at different cell densities, or in the presence ofestrogen, dexamethasone, or growth hormone did not expressIGF-II mRNA (data not shown). Increasing the amount of

». - -17

_ —¿�- •¿�- -8.5

Fig. 4. Southern blot analysis of cellular DNA digested with ///ml 111.10 pgof DNA was loaded in each lane. /, placenta; 2, HepG2; 3, MCF-7; 4, HBL 100;5, T47D. The apparent increase in the HBL100 band intensity was due to slightoverloading of the gel. Radiograph exposure was for 16 h.

A.

IO

I°CD m

•¿�•*ODX

m

i i I

Fig. 3. RNase protection assay for IGF-II mRNA in breast cancer cell lines.30 fig of total RNA was used in each lane, the 292-basepair Accl-Pstl probe wasused. Radiograph exposure was for 24 h. All cell lines were breast cancer celllines except HepG2 (hepatocellular carcinoma) and AG 1523 (human foreskinfibroblast).

B.

Fig. 5. Effect of estrogens on IGF-II mRNA. 30 jig total RNA was used ineach lane. A, time course of IGF-II mRNA response to 1 nM 17/3-estradiol. /,probe; 2, tRNA; 3, 0 h of 170-estradiol exposure; 4, 30 min; S, 60 min; 6. 2 h; 7,4 h: .V.S h; 9, 24 h. Radiograph exposure was for 16 h. Probe was a 336-basepairRsal-Pst\ fragment. B, effect of 4-OH tamoxifen on IGF-II message, cells weretreated for 8 h prior to RNA preparation. Probe is a 292 basepair Accl-Pstlfragment. Radiograph exposure was for 24 h. /, probe; 2, HepG2 grown in serumfree media; 3. HepG2 growth in the presence of 0.1 *iM4-OH tamoxifen; 4, T47Dgrown in serum free media: 5, T47D grown in the presence of 1 nM 170-estradiol;6. T47D grown in serum free media containing phenol red; 7. T47D grown inphenol red containing media and 0.1 >iM4-OH tamoxifen; 8, tRNA.

6693

Research. on December 9, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

IGF-II AND BREAST CANCER

sample RNA in the hybridization as well as decreasing theamount of time the samples were exposed to RNase A did notresult in the detection of IGF-II mRNA in these cells (data notshown). Additionally, HBL100 also expressed IGF-II message(Fig. 3). HBL100 is a cell line that was established from normalbreast milk, but contains SV40 genetic information and hasbecome transformed (29). Therefore two transformed cell linesderived from breast epithelium were found to express IGF-IImRNA.

Digestion of DNA with Hindlll, BamHl, or £coRIshowedno gene rearrangement or amplification in any of the cell linesthat expressed IGF-II mRNA (Fig. 4 and data not shown). Thebands seen in the Mndlll digest were 17, 8.5, and 4.9 kilobases,the same sizes reported for normal human cellular DNA (30).

IGF-II Message Regulation by Estrogen. Since the cell lineT47D contains estrogen receptor protein and responds mito-genically to estrogen, we studied the response of IGF-II messageto estrogen stimulation. When cells were grown in mediumdeprived of estrogen, treatment with estradici increased thesteady state levels of IGF-II mRNA in T47D. Maximal increases in the level of the mRNA was seen at 8 h postexposureto 1 nM 17/3-estradiol (Fig. 5A, Lane 8) and fell at 24 h. Therise and fall in IGF-II message preceded the mitogenic effect of

estrogen as assayed by monolayer growth. The weak estrogen,phenol red (31 ), also could increase levels of IGF-II message(Fig. 5B, Lane 6). This increase was partially blocked by treatment with 0.1 MMconcentrations of the antiestrogen 4-OHtamoxifen (Fig. 5Ä,Lane 7). Treatment of MCF-7 with 170-estradiol did not induce IGF-II message (data not shown). SinceIGF-II message is induced by estrogen stimulation, IGF-II maybe a mediator of the mitogenic effects of estrogen in T47D.

Tissue Expression of IGF-II mRNA. We next determined theexpression of IGF-II mRNA in adult normal and canceroustissues. Breast cancer tissue RNA (24/26), normal breast tissuedissected away from adjacent cancer (7/7), and benign fibroad-enomas (5/5) all expressed IGF-II mRNA (Fig. 6, A and A, anddata not shown). The level of expression varied approximately20-fold between different cancer tissues (Fig. 6Ä).IGF-II message was also prominently expressed in normal liver, kidney,

muscle, and adrenal gland (Fig. 6A and data not shown). Exceptfor adrenal tissue, RNA derived from breast tissue had thehighest level of IGF-II mRNA expression seen in the adult.

Because IGF-II was expressed in both normal breast andcancer tissue RNA and in so few transformed breast epitheliallines, it seemed likely that IGF-II message originated fromstromal elements. We studied normal fibroblasts and pairedcancer and uninvolved breast samples taken from the samepatient. Fig. 7A shows that five out of six fibroblast cell linesexamined expressed IGF-II mRNA. The human foreskin fibroblast, AG1523 (Fig. 4, Fig. 7A, Lane 3) expressed the highestlevel, however fibroblasts obtained from adults and childrenalso expressed IGF-II message.

When paired samples from the same patient were studied,IGF-II expression was seen in the primary breast tumor, insurrounding normal breast tissue, and in lymph node métastases(Fig. IB). In these samples, IGF-II mRNA in normal tissuewas equal to or greater than the message found in the cancer.Therefore, the presence of RNA derived from malignant epithelial cells did not increase detection of the IGF-II message.Since the IGF-II message can also be found in normal fibroblasts, these data suggest that IGF-II expression in breast cancermay more commonly originate in the stroma.

Effect of IGF-II on Breast Cancer Cell Lines. When theestrogen receptor positive breast cancer cell lines T47D andMCF-7 were plated in phenol red free, serum free medium at adensity of 50 x IO3cells/well, 10 nM IGF-II caused a 2.5- to 4-

fold increase in monolayer growth (Fig. 8). The stimulation ofgrowth seen at this concentration equaled that caused by 1 nM17/3-estradiol (Fig. 8). The mitogenic effect was dose dependentas 5 HM IGF-II caused an intermediate mitogenic response,while 1 nM did not increase cell growth over control levels (Fig.8 and data not shown). The IGF-II concentrations needed formitogenesis are twofold higher than those needed with IGF-I.The mitogenic effects of the IGF-II preparation are not likelydue to contaminating IGF-I since we could detect virtually noIGF-I W-terminal amino acid sequences in our preparation, yetonly twice the concentration of IGF-II was needed to producean equivalent effect. Recombinant human IGF-II has also been

Fig. 6. Tissue expression of IGF-II message. Samples were hybridized with the 292basepair Acc\-Psti probe. A, 30 ng of totalRNA derived from various tissues. MJ-CA,breast cancer sample from patient M. J. Radiograph exposure was for 16 h; B, 30 up oftotal RNA derived from breast cancer tissue.Each set of initials represents an individualpatient sample. Radiograph exposure was for16 h.

QJ

aCOem

co

" 2i Ico 5

ooO

a

oCCD

soÌ5

03

OCd>13o

rao o

jQo'•••ft

•¿� nLUco O co o I

tru9ap

I

6694

Research. on December 9, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

IGF-II AND BREAST CANCER

A.

ÃŽTB.

Fig. 7. Expression of IGF-II message in normal fibroblasts and paired cancer/normal specimens. Samples were hybridized with the Rsal-Pstl probe. A, fibro-blast expression of IGF-II. /, probe alone; 2, 30 /¿gHepG2; 3, 20 »igAG 1523; 4,30 ng GM3440; 5, 60 MgGM3440; 6, 30 ng GM0498; 7, 60 ng GM0498; 8, 30ng GM3118; 9, 30 ¿igGM408; 10, 30 ng GM3524; //, 30 »igtRNA. Radiographexposure was for 48 h. It. 20 *tg of total RNA derived from paired tumor andnormal tissue samples. /. Patient A-cancer; 2, Patient A-adjacent normal breast;-i. Patient A-lymph node metastasis; 4, Patient B-cancer; 5, Patient B-adjacentnormal breast; 6, Patient B-lymph node metastasis; 7, Patient C-cancer; A,PatientC-adjacent normal breast; 9, HepG2; 10, tRNA; //, probe. Radiograph exposurewas 16 h.

300000

.300000

100000

200000

2DAY

2

DAY

Fig. 8. Effects of IGF-II on monolayer growth of MCF-7 and T47D. Cellcounts were done at Day 2 and 4. A, T47D grown in serum free media (O), l nM17/3-estradiol (•),5 nin IGF-I (A), 1 nM IGF-II (A), 10 nM IGF-II P); B, MCF-7 grown in serum free media (O), 1 nM 17/3-estradiol (•),5 nM IGF-I (A), 1 nMIGF-II (A), 10 nM IGF-II (D). Bars represent standard error of the mean.

reported to stimulate DNA synthesis in breast cancer cell lines(32). Therefore, breast cancer cell lines respond to IGF-II withincreased growth.

DISCUSSION

We have used RNase protection assays and Northern analysisto study expression of IGF-II message. The RNase protectionassay is a sensitive method of detecting mRNA species exactlycomplementary to the transcribed probe. Since some of thebands seen in Northern analysis of IGF-II message may represent cross-hybridizing species, Northern analysis or dot blotanalysis alone may be inadequate to evaluate IGF-II message.*n the Wilms' tumor cell line, G401, a prominent 4.8-kilobase

band is easily detected on Northern analysis of total RNA, yetno authentic IGF-II mRNA can be found in the RNase protection assay (data not shown). Previous work demonstrating

enhanced levels of IGF-II transcript in certain carcinomas using10 n dot blot analysis alone may overestimate the level of IGF-II

transcript (13, 14).Using this technique we have found expression of IGF-II

message in two transformed breast epithelial cell lines, normalfibroblasts, and nearly all tissues derived from the breast. Additionally we have found that IGF-II is stimulatory for thebreast cancer cell lines MCF-7 and T47D.

T47D has been shown to be responsive to pituitary extractand prolactin. Some evidence exists to suggest that the activepeptide in pituitary extract is IGF-II (33). We have shown that10 nM human IGF-II was clearly mitogenic for T47D and

10 11 another estrogen receptor positive cell line MCF-7. The IGF-II preparation we used was highly purified and it was extremelyunlikely that the mitogenic response we saw was due to contaminating IGF-I. We have observed that IGF-II was half as potentas recombinant IGF-I in stimulating mitogenesis. IGF-II canbind to both the type I and II IGF receptors. The type IIreceptor, which has a higher affinity for IGF-II than IGF-I,does not appear to mediate a mitogenic response (24). It is feltthat the mitogenic responses caused by both IGF-I and IGF-IIare mediated through the type I receptor and IGF-II has a loweraffinity for this receptor than IGF-I (6, 7). Since breast cancercells express both type I and type II receptors (35),3 IGF-IIcould be a relatively less potent mitogen than IGF-I for thesecells. However, both IGF-I and IGF-II are effective at nano-molar concentrations and local in vivo concentrations of IGF-II may be sufficient to cause growth of malignant epithelialcells. Therefore, both IGF-II and IGF-I could be involved in

the growth of mammary cancer.The breast cancer cell line T47D expressed IGF-II mRNA,

and the message was regulated by estradiol. Therefore, in thiscell line, IGF-II may be an autocrine growth factor. Moreover,since IGF-II message was increased by exposure to estradiol,IGF-II may be part of the mitogenic response caused by estrogen in this cell. Although this is not true for other estrogenreceptor containing cell lines, development of IGF-II expressionin T47D may have contributed to its malignant progression.

When RNA derived from normal breast tissue, benign breastdisease, or breast cancer was examined we found that virtuallyall samples contained IGF-II mRNA. These data agree withour in vitro findings that both normal fibroblasts and breastcancer cells can express IGF-II message. When paired samplesfrom normal and cancerous areas were assayed, expression wasnot increased in the malignant tissue. Preliminary in situ hybridization studies demonstrate that in normal breast tissueand fibroadenoma, IGF-II mRNA is contained entirely in cellsof the stroma.4 These findings are consistent with those of Hanet al. who used in situ hybridizations to show that IGF-IImessage is expressed only in fetal mesenchymal cells (10).Expression of IGF-II by stromal cells adjacent to epithelialtumor cells that can respond to this factor indicates that IGF-II may function as a paracrine growth factor in some breasttumors.

Some samples of breast cancer tissue RNA contained highlevels of IGF-II mRNA (Fig. 6Ä,Patient J. P.). It is not clearif IGF-II mRNA was elevated in these samples because theycontain significant amounts of stroma, or because the malignantepithelial cells expressed message. When the stromal/epithelialratios of these samples on histológica! sections were comparedto the level of IGF-II message, the highest producer of IGF-IImessage (J. P.) had virtually no stroma in the tumor (data not

3K. J. Cullen, unpublished data.4 S. Paik, unpublished data.

6695

Research. on December 9, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

IGF-II AND BREAST CANCER

shown). Since we have shown that some transformed breastepithelial cells can produce IGF-II message in culture, theseobservations are presumptive evidence that IGF-II can be produced by malignant epithelial cells in the tumor. Preliminaryin situ hybridizations show that breast cancer tissues showsignificant stromal hybridization, but there is the suggestionthat some malignant epithelial cells also express IGF-II mRNA.We are currently analyzing more tissues and correlating thelevel of IGF-II mRNA expression as determined by RNaseprotection with the cell of origin by in situ hybridization.

IGF-II mRNA has been found in Wilms' tumor cell lines and

tissue. This disease is associated with a homozygous deletionin chromosome Ilpl3 implying the existence of a recessivegene important in the etiology of this tumor (36, 37). A deletionmapping to this area on chromosome 11 has also been describedin human breast cancer (28), and it is possible that mechanismof IGF-II message expression in Wilms' tumor and breast

cancer are similar. Investigation into the mechanisms of breastcancer IGF-II mRNA expression, characterization of the IGF-II protein produced by breast cancer cells and examination ofthe biological effects of blockade of the type I and type IIreceptor may yield new insights into the growth regulation ofbreast cancer.

ACKNOWLEDGMENTS

We thank AndréVeillette for technical advice and for the gift ofseveral RNAs, Susan Bates for cancer RNAs, and Graeme Bell forkindly supplying the plasmid phigf2.

REFERENCES

1. Sporn, M. B., and Roberts, A. B. Autocrine growth factors and cancer.Nature (Lond.), 313: 745-747, 1985.

2. Salomon. D. S., Zweibel, J. A., Bano. M., Losonczy, I., Felnel, P., andKidwell, W. R. Presence of transforming growth factors in human breastcancer cells. Cancer Res., 44: 4069-4077, 1984.

3. Bronzen, D. A., Pantazis, P.. Antoniades, H. N., Kasid, A., Davidson. N..Dickson. R. B., and Lippman, M. E. Synthesis and secretion of platelet-derived growth factor by human breast cancer cell lines. Proc. Nati. Acad.Sci. USA, 84: 5763-5767, 1987.

4. Vignon. F., Capony, F., Chambón. M., Freiss, G., Garcia, M., and Rochefort,H. Autocrine growth stimulation of the MCF-7 breast cancer cells by theestrogen regulated 52Kd protein. Endocrinology, 118: 1537-1545, 1986.

5. Huff, K. K., Kaufman. D., Gabbay. K. H.. Spencer, E. M., Lippman, M. E.,and Dickson, R. B. Human breast cancer cells secrete an insulin-like growthfactor-1 related polypeptide. Cancer Res., 46: 4613-4619, 1986.

6. Rechler, M. M., and Nissley. S. P. Insulin-like growth factor (IGF)/soma-tomedin receptor subtypes: structure, function, and relationships to insulinreceptors and IGF carrier proteins. Hormone Res.. 24: 152-159, 1986.

7. Baxter. R. C. The somatomedins: insulin-like growth factors. Adv. Clin.Chem., 25:49-115. 1986.

8. Rotwein, P., Pollock, K. M., Watson, M., and Milbrandt. J. D. Insulin-likegrowth factor gene expression during rat embryonic development. Endocrinology, 121: 2141-2144, 1987.

9. Adams, S. O., Nissley, S. P., Greenstein. L. A., Yang, Y. W-H., and Rechter.M. M. Synthesis of multiplication-stimulating activity (rat insulin-likegrowth factor II) by rat embryo fibroblasts. Endocrinology. 112: 979-987,1983.

10. Han, V. K. M., D'Ercole, A. J., and Lund, P. K. Cellular localization of

somatomedin messenger RNA in the human fetus. Science (Wash. DC), 236:193-197. 1987.

11. Scott, J., Cowell. J., Robertson, M. E., Priestley, L. M., Wadey, R.. Hopkins,B., Pritchard. J.. Bell, G. !.. Rail, L. B., Graham. D. F., and Knotl, T. J.Insulin-like growth factor-II gene expression in Wilms' tumour and embryonic tissues. Nature (Lond.), 317: 260-263, 1985.

12. Peres, R., Betsholtz. C.. Westermark, B., and Heldin, C-H. Frequent expression of growth factors for mesenchymal cells in human mammary carcinomacell lines. Cancer Res., 47: 3425-3429, 1987.

13. Tricoli, J. V., Rail, L. B., Karakousis, C. P., Herrera, L., Petrelli, N. J., Bell,G. L, and Shows, T. B. Enhanced levels of insulin-like growth factor messenger RNA in human colon carcinomas and liposarcomas. Cancer Res., 46:6169-6173, 1986.

14. Haselbacher, G. K., Irminger, J-C, Zapf, J., Ziegler, W. H., and Humbel, R.E. Insulin-like growth factor II in human adrenal pheochromocytomas andWilms tumors: expression at the mRNA and protein level. Proc. Nati. Acad.Sci. USA, 84:1104-1106, 1987.

15. Myal, Y., Shiu, R. P. C., Bhaumick, B., and Baia, M. Receptor binding andgrowth-promoting activity of insulin-like growth factors in human breastcancer cells (T-47D) in culture. Cancer Res., 44: 5486-5490, 1984.

16. Soule, H. D., Vazquez, J., Long, A., Albert, S., and Brennan, M. A humancell line from a pleural effusion derived from a breast carcinoma. J. Nati.Cancer Inst., 51: 1409-1416, 1973.

17. Engel, L. W., Young, N. A., Tralka, T. S., Lippman, M. E., O'Brien, S. J.,

and Joyce, M. J. Establishment and characterization of three new continuouscell lines derived from human breast cancinoma. Cancer Res., 38: 3352-3364, 1978.

18. Hackett, A. J., Smith, H. S., Springer. E. L., Owens, R. B., Nelson-Rees, W.A., Riggs, J. L., and Gradner, M. B. Two syngeneic cell lines from humanbreast tissue: the aneuploid mammary epithelial (Hs578T) and the diploidmyoepithelial (Hs578T) cell lines. J. Nati. Cancer Inst., 58: 1795-1806,1977.

19. Darbre, P., Yates, J., Curtis. S., and King, R. J. B. Effects of estradici andtamoxifen on human breast cancer cells in serum-free culture. Cancer Res.,«.-349-354, 1983.

20. Gowan, L. K., Hampton, B., Hill, D. J., Schlueter, R. J., and Perdue, J. F.Purification and characterization of a unique high molecular weight form ofinsulin-like growth factor II. Endocrinology, 121:449-458. 1987.

21. Chirgwin, J. M., Przybla, A. E., MacDonald, R. J., and Rutter, W. J. Isolationof biologically active ribonucleic acid from sources enriched in ribonucleases.Biochemistry, 18: 5294-5299, 1979.

22. Davis, L. G., Dibner, M. D., and Battey, J. F. Basic Methods in MolecularBiology, pp. 139-142. New York: Elsevier, 1986.

23. Davis, L. G., Dibner, M. D., and Battey, J. F. Basic Methods in MolecularBiology, pp. 44-50. New York: Elsevier, 1986.

24. Bell, G. I., Merryweather, J. P., Sánchez-Pescador, R., Stempien, M. M.,Priestley. L., Scott, J., and Rail, L. B. Sequence of a cDNA clone encodinghuman preproinsulin-like growth factor II. Nature (Lond.), 310: 775-777,1984.

25. Melton. D. A., Krieg. P. A., Rebagliati, M. R., and Maniatis, T. Efficient invitro synthesis of biologically active RNA and RNA hybridization probesfrom plasmids containing a bacteriophage SP6 promoter. Nucleic Acid Res.,25: 7035-7056, 1984.

26. dePagter-Holthuizen. P., Jansen, M., van Schaik, F. M. A., van der Kämmen,R., Oosterwijk. C., Van den Brande. J. L., and Sussenbach, J. S. The humaninsulin-like growth factor 11gene contains two development-specific promo-tors. FEBS Lett., 214: 259-264, 1987.

27. Irminger, J-C, Rosen, K. M., Humbei, R. E., Villa-Komaroff, L. Tissue-specific expression of insulin-like growth factor II mRNAs with distinct 5'untranslated regions. Proc. Nati. Acad. Sci. USA, 84: 6330-6334, 1987.

28. Forrester, K., Almoguera, C., Han, K.. Grizzle, W. E., and Perucho, M.Detection of high incidence of K-ros oncogenes during human colon minor¡genesis.Nature (Lond.), 327: 298-303, 1987.

29. Carónde Fromental, D., Nardeux, P. C., Soussi, T., Lavialle, C., Estrade,E., Cartoni, G., Chandrasekaran, K., and Cassingena, R. Epithelial Ulto100 cell line derived from milk of an apparently healthy woman harboursSV40 genetic information. Exp. Cell Res.. 160: 83-94, 1985.

30. Tricoli, J. V., Rail, L. B., Scott, J.. Bell, G. I., and Shows, T. B. Localizationof insulin-like growth factor genes to human chromosomes 11 and 12. Nature(Lond.). 310: 784-786, 1984.

31. IU-iili'us. Y., Katzenellenbogen, J. A., and Katzenellenbogen, B. S. Phenolred in tissue culture media is a weak estrogen: implications concerning thestudy of estrogen-responsive cells in culture. Proc. Nati. Acad. Sci. USA, 83:2496-2500,1986.

32. Hirsch, K. S.. Sledge, Jr., G. W., Swift, R. A., and Falcone, J. F. Binding,processing, and effects of IGF-II on normal and malignant mammary epithelial cells. J. Cell Biochem. S12A Abs »C6211988.

33. Shiu, R. P. C., Murphy, L. C, Myal, Y., Dembinski. T. C, Tsuyuki, D., andIwasiow, B. M. Actions of pituitary prolactin and insulin-like growth factorII in human breast cancer In: M. E. Lippman and R. B. Dickson (eds.).Breast Cancer: Cellular and Molecular Biology. Boston: Kluwer AcademicPublishers, 1988.

34. Furlanetto, R. W., DiCarlo, J. N.. and Wisehart, C. The type II insulin-likegrowth factor receptor does not mediate deoxyribonucleic acid synthesis inhuman fibroblasts. J. Clin. Endocrinol. Metab., 64: 1142-1149. 1987.

35. DeLeon, D. D.. Bakker, B., Wilson, D. M., Hintz, R. L., and Rosenfeld, R.G. Demonstration of insulin-like growth factor (IGF-I and -II) receptors andbinding protein in human breast cancer cell lines. Biochem. Biophys. Res.Commun., 752: 398-405, 1988.

36. Reeve, A. E., Eccles, M. R., Wilkins. R. J., Bell, G. I., and Millow, L. J.Expression of insulin-like growth faclor-II transcripts in Wilms tumour.Nature (Lond.). 317: 258-260, 1985.

37. Scott, J., Cowell, J., Robertson, M. E.. Priestley, L. M., Wadey, R., Hopkins,B., Pritchard, J., Bell, G. L, Rail, L. B., Graham, C. F., and Knott, T. J.Insulin-like growth factor-II gene expression in Wilms' tumour and embryonic tissues. Nature (Lond.), 317: 260-262, 1985.

38. Ali, I. U., Lidereau. R., Theillet. C., and Callahan. R. Reduction to homo-zygosity of genes on chromosome 11 in human breast neoplasia. Science(Wash. DC), 238: 185-188, 1987.

6696

Research. on December 9, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

1988;48:6691-6696. Cancer Res   Douglas Yee, Kevin J. Cullen, Soonmyoung Paik, et al.   CancerInsulin-like Growth Factor II mRNA Expression in Human Breast

  Updated version

  http://cancerres.aacrjournals.org/content/48/23/6691

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/48/23/6691To request permission to re-use all or part of this article, use this link

Research. on December 9, 2020. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from