MCF-7 Breast Cancer Cells Transfected with Protein Kinase C-a Exhibit
 Altered Expression of Other Protein Kinase C Isoforms and Display a More
 Aggressive Neoplastic Phenotype
 D. Kirk Ways,* Cynthia A. Kukoly,* James deVente,* Jerry L. Hooker,* Winifred 0. Bryant,* Karla J. Posekany,t
 Donald J. Fletcher,* Paul P. Cook,* and Peter J. Parker1l
 Departments of *Medicine, WMicrobiology/Immunology, and §Anatomy/Cell Biology, East Carolina University School of Medicine,
 Greenville, North Carolina 27858; and 11 Imperial Cancer Research Fund, Protein Phosphorylation Laboratory, London WC2A 3PX,
 United Kingdom
 Abstract
 Increased protein kinase C (PKC) activity in malignant
 breast tissue and positive correlations between PKC activity
 and expression of a more aggressive phenotype in breast
 cancer cell lines suggest a role for this signal transduction
 pathway in the pathogenesis and/or progression of breast
 cancer. To examine the role of PKC in the progression of
 breast cancer, human MCF-7 breast cancer cells were
 transfected with PKC-a, and a group of heterogenous cells
 stably overexpressing PKC-a were isolated (MCF-7-PKC-
 a). MCF-7-PKC-a cells expressed fivefold higher levels of
 PKC-a as compared to parental or vector-transfected MCF-
 7 cells. MCF-7-PKC-a cells also displayed a substantial
 increase in endogenous expression of PKC-8 and decreases
 in expression of the novel 6- and q-PKC isoforms.
 MCF-7-PKC-a cells displayed an enhanced prolifera-
 tive rate, anchorage-independent growth, dramatic morpho-
 logic alterations including loss of an epithelioid appearance,
 and increased tumorigenicity in nude mice. MCF-7-PKC-
 a cells exhibited a significant reduction in estrogen receptor
 expression and decreases in estrogen-dependent gene ex-
 pression. These findings suggest that the PKC pathway may
 modulate progression of breast cancer to a more aggressive
 neoplastic process. (J. Clin. Invest. 1995. 95:1906-1915.)
 Key words: PKC isoforms * vimentin * estrogen receptor.
 anchorage independent * tumorigenicity
 Introduction
 Protein kinase C (PKC)' is a gene family consisting of several
 subfamilies (conventional: a, /3k, 31, y, novel: 6, E, 9, 77; and
 This work was presented in part at the American Federation for Clinical
 Research in Baltimore, MD, 29 April-2 May 1994 and published in
 abstract form (1994. Clin. Res. 42:206a).
 Address correspondence to Dr. Kirk Ways, Department of Medicine,
 Section of Endocrinology & Metabolism, East Carolina University
 School of Medicine, Greenville, NC 27858. Phone: 919-816-2567; FAX:
 919-816-3096.
 Received for publication 26 September 1994 and in revised form
 29 November 1994.
 1. Abbreviations used in this paper: CAT, chloramphenicol acetyltrans-
 ferase; MDR, multidrug resistant; PKC, protein kinase C; TPA, tetrade-
 canoyl-13-phorbol acetate; vit-tk-CAT, estrogen-responsive vitellogenin
 promoter-CAT reporter construct.
 atypical: X, 4, t) that lie on signal transduction pathways regu-
 lating growth and differentiation in a large and diverse group
 of cell types ( 1, 2).
 A role for members of the PKC gene family in regulating
 growth and differentiation of breast cancer has been suggested
 by several separate observations. In extracts prepared from sam-
 ples derived from patients undergoing breast biopsy, PKC activ-
 ity was elevated in malignant breast tumors relative to normal
 breast tissue (3). In cultured breast cancer cell lines, a positive
 correlation also existed between PKC activity and the more
 biologically aggressive estrogen receptor negative phenotype
 (4, 5). In the MCF-7 and other breast cancer cell lines, phorbol
 ester treatment that induced downregulation of PKC activity
 inhibited growth and enhanced expression of parameters associ-
 ated with a more differentiated phenotype (6-9). Studies exam-
 ining the effects of bryostatin-1 and tetradecanoyl-13-phorbol
 acetate (TPA) on growth of breast cancer cell lines have impli-
 cated a role for the conventional PKC-a isoform in mediating
 TPA-induced growth inhibition (8). Desensitization to phorbol
 ester-induced growth inhibition has been associated with in-
 creases in PKC content and activity (6). Additionally, inhibitors
 of PKC-associated kinase activity, such as the staurosporine
 analogue UCN-01, inhibit growth of the MCF-7 cell (10).
 These findings suggest that aberrations in the PKC signal trans-
 duction pathway may be involved in the pathogenesis and/or
 progression of breast cancer, and that PKC may be involved in
 the regulation of estrogen receptor expression, a parameter of
 prognostic and therapeutic importance.
 Given the possible role for the PKC signal transduction
 pathway, specifically PKC-a, in the pathogenesis of breast can-
 cer, we directly examined the effects of stably overexpressing
 PKC-a on the MCF-7 human breast cancer cell line. Stable
 overexpression of PKC-a led to an enhanced rate of prolifera-
 tion, more efficient anchorage-independent growth, dramatic
 alterations in cellular morphology manifested by loss of an
 epithelioid appearance with a marked increase in vimentin ex-
 pression, a marked decrease in estrogen receptor mRNA tran-
 scripts and estrogen-dependent gene expression, enhanced tu-
 morigenicity with metastatic capacity when injected into nude
 mice, and significant increases in the endogenous expression
 of PKC-/3 accompanied by decreases in the content of PKC-77
 and -6.
 These findings demonstrate that overexpression of PKC-
 a, either directly or indirectly, possibly via increases in the
 endogenous expression of PKC-,3 or decreases in the 77 or 6
 isoforms, leads to the acquisition of a more fully transformed
 phenotype in the MCF-7 breast cancer cell, and provide further
 evidence for a possible role of the PKC signaling pathway
 in the pathogenesis and/or progression of breast cancer and
 regulation of estrogen receptor status.
 1906 Ways et al.
 J. Clin. Invest.
 © The American Society for Clinical Investigation, Inc.
 0021-9738195104/1906/10 $2.00
 Volume 95, April 1995, 1906-1915
Methods
 Cell culture. MCF-7 cells were purchased from the American Type
 Culture Collection (Rockville, MD) and grown in DME supplemented
 with 2 mM glutamine, 10 mM Hepes, 10% FCS, 50 U/ml penicillin,
 and 50 jig/ml streptomycin. For cell growth experiments, 7.5 X 103
 cells/ml were incubated in 24-well plates. Cell viability was assessed
 using trypan blue exclusion. Counting of cells was performed using a
 Coulter counter (Coulter Corp., Hialeah, FL).
 Generation of MCF-7 cells overexpressing PKC-a. Bovine PKC-a
 ( 1 ) was subcloned into the pSV2M(2)6 plasmid (20 pig) and cotrans-
 fected with 2 jtg pMAM-neo plasmid (Clonetech, Palo Alto, CA) by
 calcium phosphate precipitation. The pSV2M(2)6 vector was gener-
 ously provided by Dr. Marilyn Sleigh (Commonwealth Scientific and
 Industrial Research Organization, Sydney, Australia [12]). 72 h after
 transfection the cells were incubated in 0.75 mg/ml G418. After a 6-
 wk selection period, multiple clones of cells overexpressing PKC-a
 (MCF-7-PKC-a) and vector without the PKC insert were detected.
 The results presented herein are derived from pooling these separate
 clones into heterogeneous groups of vector- or PKC-a-transfected cells.
 Although transfected PKC-a is under control of a metallothionein pro-
 moter, cells stably transfected with this construct displayed marked
 phenotypic changes and overexpressed the a isoform approximately
 fivefold in the absence of supplemental heavy metals beyond the quanti-
 ties contained in FCS (see Results). Because of the basal induction of
 PKC-a, experiments were done in the absence of further heavy metal
 supplementation. The MCF-7-PKC-a and vector-transfected cells have
 maintained stable phenotypic characteristics over a 9-mo period.
 PKC activity and Western blot analysis. Cytosolic, solubilized par-
 ticulate, or whole-cell solubilized extracts, were prepared as previously
 described (13) and used for quantitation of PKC activity and Western
 blot analysis. Protein content was determined by the Bradford technique
 (14). Using equal protein concentrations (10 jtg), PKC activity was
 assessed as previously described ( 15 ) with histone III-s, a peptide corre-
 sponding to residues 4-14 of myelin basic protein (16), and a serine-
 substituted PKC-6 pseudosubstrate site peptide ( 15) as substrates. PKC
 activity was defined as the difference in cpm incorporated into the
 substrates in the absence and presence of TPA and phosphatidylserine.
 Western blot analysis was done as previously described using anti-
 sera specific for the a, /3, c, and 4 isoforms (13, 15). Antipeptide
 antiserum against the carboxy terminus of PKC-77 (INQDEFRNFSYV-
 SPELQL) was prepared and used in this study at a dilution of 1:1,000.
 This antiserum recognized a protein doublet at approximately the pre-
 dicted molecular weight for PKC-77 and a slightly lower molecular
 weight species. The specificity of these proteins was confirmed by the
 ability of the immunizing peptide to abolish detection of these bands.
 A separate PKC-77 antiserum, purchased from Santa Cruz Biotechnol-
 ogy, Inc., (Santa Cruz, CA) gave identical results to those obtained
 using our 7 antipeptide antibody. Antiserum to PKC-6 was purchased
 from Transduction Labs (Lexington, KY) and was used as per their
 recommendations. Detection was performed using goat anti-mouse an-
 tiserum and enhanced chemiluminescence (15). Densitometric analysis
 of the autoradiograms was performed using a laser densitometer (Ul-
 trascan XL; Pharmacia LKB Biotechnology, Inc., Piscataway, NJ).
 Northern blot analysis. RNA was extracted using a single-step acid
 guanidinium thiocyanate-phenol extraction technique (17). RNA (20
 jig) was electrophoresed, transferred to nylon filters, and hybridized to
 random primer labeled cDNA probes (18). Ethidium bromide staining
 was used to confirm the integrity of the RNA and assess the equivalency
 of loading. The probes used for PKC-fl, -77, and -6 were isolated by P.
 Parker and represent the full length cDNAs. The vimentin cDNA was
 generously provided by Dr. Susan Rittling (Rutgers University). Probes
 for the estrogen receptor, cathepsin D, and pS2 genes were purchased
 from the American Type Culture Collection.
 Microscopy. Cells maintained in culture were photographed at a
 magnification of 400 with an inverted microscope (Diaphot; Nikon,
 Inc., Melville, NY). For structural analysis, cells were fixed for 30 min
 in 2.5% (vol/vol) glutaraldehyde and 0.2 M phosphate buffer (pH
 7.4). After postfixation in 1% (wt/vol) osmium tetroxide, cells were
 dehydrated through a graded series of alcohol and propylene oxide and
 embedded in Epon (Electron Microscopy Sciences, Fort Washington,
 PA). Thin sections (< 90 nm) were stained with uranyl acetate and
 lead citrate and examined with an electron microscope ( 1200 EX; JEOL
 USA Inc., Peabody, MA).
 Cell cycle analysis. 30 min before harvesting, cells were incubated
 with 10 ,M bromodeoxyuridine. Cells were then trypsinized, washed,
 resuspended in PBS, and fixed in 70% ethanol at 40C for 30 min. The
 fixed cells were treated at room temperature for 6 min with 6 N HCL
 containing 0.5% Triton X-100. After washing with PBS, the cells were
 resuspended in serum-free RPMI 1640 media and incubated for 15
 min at room temperature with FITC-conjugated antibromodeoxyuridine
 antiserum (Becton Dickinson Immunocytometry Systems, Mountain
 View, CA). The samples were washed, resuspended in 50 jug/ml propid-
 ium iodide contained in a citrate buffer, and immediately analyzed on
 a flow cytometer (Facscan®; Becton Dickinson Immunocytometry Sys-
 tems). Linear red and green fluorescence and logarithmic forward and
 side scatter signals were obtained on 104 cells. For statistical analysis,
 polygonal regions were drawn on a propidium iodide versus bromodeox-
 yuridine 2-parameter histogram display defining cells in GO/GI, S, and
 G2M phases. The percentage of cells falling within these regions was
 calculated using the Facscan® statistical analysis package (Becton Dick-
 inson Immunocytometry Systems) and normalized to 100%.
 Transfection and reporter construct analysis. Cells were transiently
 transfected using calcium phosphate precipitation with 3 Isg p-cytomeg-
 alovirus /3-galactosidase plasmid (Clonetech Laboratories, Palo Alto,
 CA), 4 ,g estrogen-responsive vitellogenin promoter-chloramphenicol
 acetyltransferase reporter construct (vit-tk-CAT), and 13 ig of carrier
 DNA. After an overnight incubation, the cells were washed and grown
 in media with 10% FCS. 40 h later, fi-galactosidase and CAT activity
 were analyzed by standard methodologies and CAT measurements were
 normalized to ,6-galactosidase activity. The estrogen-responsive, vit-tk-
 CAT construct was kindly provided by Dr. Gerhardt Ryffel (Kern-
 forschungszentrum, Karlsruhe Institut fur Genetik und Toxikolo-
 gie [19]).
 Growth in soft agar. In a 35-mm dish, 1 ml of 0.6% agar in DME
 with 10% FCS was plated and allowed to harden. Top agar, 0.36%,
 containing 103 cells/ml was layered over the hardened agar. Cultures
 were incubated at 37° C in 5% CO2 for 10-12 d. Colonies > 50 cells
 were visually enumerated.
 Tumor and metastatic evaluation in nude mice. BALB/c nu/nu
 female mice, 6-8 wk old, were inoculated with 7 x 106 cells in 0.1
 ml of PBS into the mammary fat pad. Because of the rapid primary
 tumor growth in mice inoculated with the MCF-7-PKC-a cells, animals
 were killed only 4 wk after inoculation. Tumor volume was assessed
 by measuring tumor size in three dimensions. Histologic examination
 of the primary tumor and other organs was done using hematoxylin and
 eosin staining.
 Results
 Overexpressing of PKC-a increased PKC activity and altered
 expression of endogenous PKC isofonns. MCF-7 cells were
 cotransfected with the neomycin resistance plasmid and either
 the pSV2M(2)6 vector without the insert or containing a full-
 length cDNA encoding PKC-a. Cells harboring the neomycin
 resistance gene were identified by selection in G418. Rather
 than single cell cloning, a mixed group of neomycin-resistant
 cells was isolated. A heterogenous population of transfected
 cells was used to minimize the possible artifacts that can be
 associated with single cell cloning. Analysis of PKC-a expres-
 sion is shown in Fig. 1. As compared to either parental MCF-
 7 cells or cells stably transfected with the neomycin gene and
 vector without the PKC insert, PKC-a was overexpressed ap-
 proximately three- to fivefold, as assessed by densitometric
 MCF-7 Cells Transfected with Protein Kinase C- a 1907
120 r-
 100 1-
 A
 Cell: Con PKCox
 Cell Fraction: C P C P
 a
 -
 80 k
 60k
 ma
 CP CP CP CP CP CP CP
 Con ax Con a Con a
 MBP 4-14 Histone 111- s Protamine
 Figure 1. PKC-a content, and PKC activity in MCF-7-PKC-a. Equal
 amounts of protein (40 [tg) derived from cytosolic (C) and solubilized
 particulate (P) fractions from parental (con) and MCF-7-PKC-a
 (PKC-a) cells were subjected to Western blot analysis using PKC-a
 antiserum (inset). PKC activity of the same cells is shown below. The
 substrates used are indicated beneath the figure. PKC activity is defined
 as the difference in cpm incorporated into substrate in the absence and
 presence of phosphatidylserine and TPA in parental (con) and MCF-
 7-PKC-a (a) cells. These experiments were repeated on three separate
 occasions with similar results.
 analysis, in the cells stably transfected with this isoform (MCF-
 7-PKC-a cells). Corroborating the increased content of this
 isoform in PKC-a cells was a 5-16-fold increase, depending
 upon the substrate used, in PKC activity in the MCF-7-PKC-
 a cells as compared to the parental MCF-7 cell (Fig. 1).
 Endogenous expression of other PKC isoforms in MCF-7-
 PKC-a cells was examined by Western blot analysis (Fig. 2
 A). Parental MCF-7 cells contained readily apparent amounts
 of a, E, 6, and 4 isoforms (Figs. 1 and 2). Using a antiserum
 directed against the V3 region of the molecule, we were unable
 to detect significant PKC-,/ expression in the parental MCF-7
 cell. The PKC-r7 antiserum specifically detected three bands
 with molecular weights of 86, 82, and 60 kD. Relative to the
 parental MCF-7 cells, MCF-7-PKC-a cells displayed a large
 increase in protein content of 6 and similar quantities of c and
 4. Levels of 6 were reduced modestly in MCF-7-PKC-a cells.
 Levels of the 86- and 82-kD proteins detected by the PKC-i7
 antiserum were negligible in MCF-7-PKC-a cells. Content of
 the lower 60-kD protein detected by the 7 antiserum was also
 substantially reduced in MCF-7-PKC-a cells. The identity of
 the 60-kD protein and its relation to the predicted molecular
 weight form of remains unresolved. Densitometric scanning
 of cytosolic and solubilized particulate fractions subjected to
 Western blot analysis indicated the following subcellular distri-
 bution of PKC isoforms in MCF-7 cells: a, 65/35 (percentage
 of total isoform content in cytosolic/solubilized particulate frac-
 tions); c, 60/40; 4, 52/48; r 86-kD species, 39/61; i1 82-kD
 species, 53/47; and rq 60-kD species, 80/20. In MCF-7-PKC-a
 cells, the cytosolic/solubilized particulate fraction distributions
 .:,\ CG
 .4 A wr4r.Ut
 a mu
 r_
 84 m
 58 -
 B
 8.7 ?a;
 3.4
 2.6
 f 3.3
 3.1
 EtBr
 Figure 2. Altered endog-
 enous PKC isoform con-
 tent in MCF-7-PKC-a
 cells. (A) Western blot
 analysis was performed
 on equal protein concen-
 trations of whole-cell
 solubilized fractions de-
 rived from parental and
 UC V C MCF-7-PKC-a cells.The antisera used are in-
 dicated (left). Molecular
 weight standards are
 shown to the left of the
 ;... PKC-r1 autoradiogram.
 Vector-transfected cells
 V;k'M displayed identical PKC
 isoform expression as did
 parental MCF-7 cells.
 (B) Northern blot analy-
 : * ,s using total RNA (20
 MIF-g)prepared from pa-
 rental (C), vector (V),
 and MCF-7-PKC-a (a)
 cells was performed with
 PKC-,8, -r7, and -6 DNA
 probes. The approximate
 molecular weight of the
 detected mRNA tran-
 scripts is indicated to the
 side of the autoradio-
 gram. Ethidium bromide
 (EtBr) staining is shown
 below the autoradio-
 grams. These experi-
 ments were performed on
 a separate occasions with
 similar results.
 were as follows: a, 46/54; /3, 81/19; E, 56/44; ;, 58/42; and
 r1 60-kD species, 60/40. Because of their negligible amounts,
 PKC-,/ and the 86/82-kD species of PKC-r1 could not be quanti-
 tated in MCF-7 and MCF-7-PKC-a cells, respectively. Be-
 cause of these changes in the endogenous expression of the /,
 6, and isoforms in MCF-7-PKC-a cells, Northern blot analy-
 sis was performed to examine the content of their mRNA tran-
 scripts. A concomitant increase in mRNA transcripts for the
 isoform and decreases in the 6 and rq isoform mRNA transcripts
 were noted in the MCF-7-PKC-a cells (Fig. 2 B). Thus, either
 in a direct fashion or secondary to the phenotypic changes dem-
 onstrated in MCF-7-PKC-a cells (see below), PKC-a overex-
 1908 Ways et al.
 D
 to
 .0
 0
 r_
 'O
 0
 'a
 CI
 0
 0
 5:r
 U
 0
 09
 40 -
 20 k
 Fraction: C P
 Cell: Con a
 Substrate: 8 Peptide
1~ ~ ~ ~~i1
 A
 Figure 3. Dissimilar morphology of MCF-7-PKC-a and parental MCF-7 cells. Exponentially growing parental (A), vector-transfected (B), and
 MCF-7-PKC-a (C) cells were photographed at a magnification of 400 using a Diaphot inverted microscope.
 pression modulated expression of other endogenous PKC family
 members.
 Overexpression of PKC-a drastically altered morphology of
 the MCF-7 cell. MCF-7-PKC-a cells demonstrated dramatic
 morphologic alterations as compared to parental MCF-7 cells
 (Fig. 3). Parental and vector-transfected MCF-7 cells displayed
 an epithelioid appearance growing adherent to the plastic surface
 (Fig. 3, A and B, respectively). In marked contrast, MCF-7-
 PKC-a cells assumed a spherical shape, could proliferate in sus-
 pension, especially as cell density increased, and occasionally
 formed giant cells (Fig. 3 C). Electron microscopy indicated that
 parental MCF-7 cells contained primary lysosomal granules (Fig.
 4 A, black triangular indicators), while MCF-7-PKC-a cells
 contained an abundance of secondary lysosomal granules (Fig.
 4 B, thin arrows). The nuclear to cytoplasmic ratio was signifi-
 cantly greater in MCF-7-PKC-a cells. While not observed in
 MCF-7 cells, areas enriched in intracellular fibrils were noted in
 MCF-7-PKC-a cells (Fig. 4 X, thick arrows). At a magnifica-
 tion of 50,000 these fibrils had a diameter consistent with interme-
 diate filaments (data not shown). As shown in Fig. 4 C, Northern
 blot analysis revealed a dramatic increase in mRNA transcripts
 for vimentin, an intermediate filament that is not normally ex-
 pressed in MCF-7 cells (20) and whose expression is a poor
 prognostic indicator in breast cancer (21, 22).
 MCF-7-PKC-a cells are estrogen receptor negative and
 display a loss ofestrogen-dependent gene expression. In previous
 studies an association between increases in PKC activity and
 decreases in estrogen receptor expression have been noted (4, 5).
 Given this association and the more morphologically aggressive
 appearance of the MCF-7-PKC-a cell, estrogen receptor status
 was analyzed (Fig. 5 A). By Northern blot analysis, parental
 MCF-7 cells were shown to contain transcripts for estrogen recep-
 tor mRNA. In contrast, mRNA transcripts for the estrogen recep-
 tor were not detected in MCF-7-PKC-a cells. To assess the
 functional importance of this finding, expression of estrogen-
 responsive genes and the activity of an estrogen-responsive genes
 and the activity of an estrogen-responsive reporter construct were
 analyzed. Expression of estrogen-responsive genes, pS2, and ca-
 thepsin D (23, 24), were markedly decreased in MCF-7-PKC-
 a cells grown in the presence of 10% FCS (Fig. 5 A). Activity
 of the estrogen-responsive vit-tk-CAT reporter construct in-
 duced by estrogens contained in 10% FCS was apparent in paren-
 tal MCF-7 cells (Fig. 5 B). However, this estrogen-dependent
 activity was minimal in MCF-7-PKC-a cells. Thus, the MCF-
 7-PKC-a cells share a similar relationship between decreased
 estrogen receptor expression and increased PKC activity, as has
 previously been described in clinical samples from patients with
 breast cancer (4).
 MCF-7-PKC-a cells demonstrate a higher proliferative
 rate and enhanced anchorage-independent growth. MCF-7-
 PKC-a cells proliferated at a higher rate than did parental or
 vector-transfected cells (Fig. 6 A). As MCF-7-PKC-a in-
 creased in density, they detached from the plastic surface and
 proliferated in suspension (Fig. 6 B). The nonadherent popula-
 tion of the MCF-7-PKC-a were viable as assessed by trypan
 blue exclusion (data not shown). In contrast, parental MCF-7
 cells proliferated in an adherent fashion (Fig. 6 B). MCF-7-
 PKC-a and parental or vector-transfected cells reached satura-
 tion density at 1.2 x 106/ml and 0.5 X 106/ml, respectively
 (data not shown). The importance of the higher cell density at
 confluence in MCF-7-PKC-a cells is somewhat uncertain
 given the ability of these cells, but not the parental MCF-7
 cells, to grow in suspension.
 Given the rapid proliferative rate of MCF-7-PKC-a cells
 and the poor prognostic value in clinical specimens of a high
 S phase value (25), the proportion of exponentially growing
 cells in S phase was evaluated. In parental MCF-7 cells, the
 cells were proportioned as follows: Go/GI: 61± 1.6; S: 23±1.3;
 and G2/M: 16±1.5 (X±SEM, n = 10). In MCF-7-PKC-a
 cells, the cell cycle distribution was as follows: GO/GI: 44±1.2;
 S: 39+2.4; and G2/M: 17±1.5 (X±SEM, n = 8). Thus, MCF-
 7-PKC-a cells displayed an increase in S phase cells with a
 MCF-7 Cells Transfected with Protein Kinase C- a 1909
 ''"
 1.
 At
A B
 I
 ,VP
 OF 4. i .
 .f, i>t
 C;
 C,
 -p
 concomitant decrease of cells in Go/G1. Using a two-tailed
 Student's t test, these differences between the parental MCF-7
 and MCF-7-PKC-a cells were statistically significant at a level
 of <0.05.
 MCF-7-PKC-a cells readily grew in an anchorage-indepen-
 dent manner in soft agar (Fig. 7). However, under these experi-
 mental conditions, parental MCF-7 cells displayed negligible
 anchorage-independent growth with only a few cell clusters
 observed and no colonies of > 50 cells detected (Fig. 7).
 MCF-7-PKC-a cells were tumorigenic in nude mice with
 a propensity to metastasize. When injected into the mammary
 fat pad, MCF-7-PKC-a cells rapidly formed large primary tu-
 mors in all mice by 4 wk postinoculation (Table I). The single
 mouse that did not form a tumor when inoculated with MCF-
 7-PKC-a cells was inadvertently injected outside the mammary
 fat pad and into the subcutaneous space. The tumors in mice
 inoculated with MCF-7-PKC-a cells were well circumscribed
 and bordered by a thin fibrous tissue network. Histologically,
 the MCF-7-PKC-a primary tumors were composed of solid
 sheets of tightly packed, anaplastic, round to polygonal cells
 with scant basophilic to pale eosinophilic cytoplasm and mild
 cellular pleomorphism (Fig. 8, A and B). Marked nuclear pleo-
 morphism, increased content of nucleoli, and numerous mitotic
 figures were also noted in the primary MCF-7-PKC-a tumors.
 In contrast, parental MCF-7 cells formed only microscopic,
 Figure 4. Electron microscopy of MCF-7 and
 MCF-7-PKC-a cells. Electron microscopy
 was performed on parental MCF-7 (A) and
 MCF-7-PKC-a (B) cells at x5,000. In A,
 the black triangular indicators mark primary
 lysosomal granules. In B, the thin arrows
 point to secondary lysosomal granules while
 the thicker arrows are directed at a bundles
 of intracellular fibrils. Northern blot analysis
 of vimentin expression in parental MCF-7
 and MCF-7-PKC-a cells is shown in C.
 Ethidium bromide staining demonstrated that
 equivalent amounts of RNA were loaded per
 lane (data not shown).
 nonpalpable primary tumors in 2/10 animals with no detectable
 metastases. Western blot analysis of the primary MCF-7-PKC-
 a tumors demonstrated the continued expression of a, A, 6, E,
 77, and 4 isoforms (data not shown). Due to the microscopic
 size of the tumors in mice injected with parental MCF-7 cells,
 a direct comparison between the relative in vivo expression of
 isoforms in parental MCF-7 and MCF-7-PKC-a cells could
 not be made. In mice inoculated with MCF-7-PKC-a cells,
 metastases were observed in lymph nodes, lung, and perinephric
 fat (Table I). Metastases in mice inoculated with MCF-7-PKC-
 a cells were microscopic in size. The full extent of the meta-
 static potential of the MCF-7-PKC-a cell may have been un-
 derestimated because of the relatively short period from inocula-
 tion to necropsy (4 wk) that was necessitated by the rapid
 growth of the primary MCF-7-PKC-a tumor. Given the posi-
 tive correlation between metastatic capacity and the size of the
 primary breast tumor (26), it is uncertain whether MCF-7-
 PKC-a cells had a higher metastatic potential simply because
 of their high growth rate or whether these cells acquired separate
 biologic programs, rendering them with an intrinsically in-
 creased metastatic capacity.
 Discussion
 This study demonstrates that human MCF-7 breast cancer cells
 transfected with and overexpressing PKC-a display enhanced
 1910 Ways et al.
 I ..'
 ,-
 -,
 "-Al. I t" 1.V A6 `.-nice 711%` -It-
8
 A
 MCF-7 MCF-7-PKCa 7
 ER
 6
 5
 pS2
 cathepsin D
 2
 B is
 11
 00
 o 9
 a
 D
 01
 4
 3
 2
 I
 MUCF-7 MCF-7-P~
 Figure 5. Diminished estrogen receptor content and function in MCF-
 7-PKC-a cells. (A) Total RNA, 20 .g, prepared from parental and
 MCF-7-PKC-a cells, was subjected to Northern blot analysis using
 cDNA probes corresponding to the estrogen receptor (ER), pS2, and
 cathepsin D as shown to the left of the autoradiograms. Equivalent
 amounts of RNA were loaded per lane as assessed by ethidium bromide
 staining (data not shown). (B) vit-tk-CAT activity normalized to f3-
 galactosidase measurements (ordinate) was analyzed in parental and
 MCF-7-PKC-a cells transiently transfected with this reporter construct.
 The bars represent SEM. Each of these experiments was repeated with
 similar results.
 growth rate with a higher percentage of cells in S phase, anchor-
 age-independent proliferation, a transition from an epithelial
 to a mesenchymal-like morphology, loss of estrogen receptor
 content, and increased tumor formation with metastatic capacity
 in nude mice. The transition to a more mesenchymal appear-
 ance, the corresponding increase in vimentin expression, a high
 S phase fraction, and the loss of estrogen receptor expression
 with the attendant decrease in levels of the pS2 mRNA tran-
 scripts occurring in the MCF-7-PKC-a cell are observed in
 breast cancer patients with a more malignant course and less
 favorable prognosis (21-23, 27, 28). Interestingly, mRNA
 transcripts for cathepsin D were markedly reduced in the MCF-
 7-PKC-a cell. While cathepsin D is an estrogen-responsive
 gene, clinical studies have demonstrated that cathepsin D ex-
 pression is a poor prognostic indicator, with high levels pre-
 dicting a poor clinical outcome in breast cancer patients (29).
 Thus, the reduced cathepsin D content in MCF-7-PKC-a cells
 seems at odds with the obvious progression of this cell to a
 more aggressive phenotype. Perhaps explaining this discrepancy
 is the finding that in clinical specimens the cathepsin D content
 may be derived from cells other than mammary epithelium cells
 (30). Thus, the MCF-7-PKC-a cell seems to recapitulate a
 very similar phenotype to that found in more aggressive lesions
 derived from patients with breast cancer. This close recapitula-
 tion of phenotypic characteristics substantiates a possible role
 for alterations in the PKC pathway in modulating the in vivo
 progression of human breast cancer and suggests the potential
 usefulness of the MCF-7-PKC-a cell in screening antineoplas-
 tic agents used for treating advanced forms of breast cancer.
 A
 a
 Is
 i
 10
 I1
 z
 7
 The (days) Tkme (days)
 Figure 6. MCF-7-PKC-a cells exhibit a higher proliferative rate than
 MCF-7 cells. (A) MCF-7 (o0o), vector-transfected (.-*), and
 MCF-7-PKC-a (A-A) cells were plated in 24-well plates at 7.5
 x 103/ml, and total cell number was assessed at various times (abscissa)
 after plating. (B) In a separate experiment, the growth of nonadherent
 MCF-7 (o0o) and MCF-7-PKC-a (A - - A) cells is shown. The num-
 ber of nonadherent cells is expressed as a percentage of total cells. Each
 point is derived from six separate samples. These experiments were
 repeated on several occasions with similar results.
 A negative correlation between PKC activity and estrogen
 receptor status in breast cancer has been previously noted (4,
 5). The mechanism responsible for this association has yet to
 be fully elucidated. Exposure of breast cancer cells to phorbol
 esters that both activate and downregulate PKC activity results
 in a significant reduction in mRNA transcripts of the estrogen
 receptor (30-32). Given that the status of alterations within
 the PKC pathway was not evaluated in these investigations (31-
 33), the specific mechanism by which phorbol esters interacted
 with the PKC pathway to decrease the level of mRNA tran-
 scripts for the estrogen receptor gene is unclear. Thus, estrogen
 receptor expression could be negatively modulated by phorbol
 ester-induced activation of a specific PKC isoform(s) or be-
 cause of downregulation of an isoform( s) necessary for mainte-
 nance of estrogen receptor mRNA levels. In the MCF-7-PKC-
 a cell, the predominant alterations in the PKC pathway were
 an increase in the conventional a and isoforms. It is tempting
 to speculate that the increased content of these isoforms led
 directly to an enhanced basal kinase activity, which in turn
 activated a cascade that decreased content of estrogen receptor
 mRNA transcripts. Given the increased level of PKC-a in the
 estrogen receptor-negative MCF-7-PKC-a cell and the in-
 creased PKC-a levels with a reduction in PKC-,B content in the
 estrogen receptor-negative, multidrug-resistant (MDR) MCF-
 7-MDR cells (33), a role for PKC-a in negatively modulating
 estrogen receptor expression seems possible. Obviously, further
 studies will need to be undertaken to prove such a hypothesis.
 The significant morphologic alterations and higher degree
 of phenotypic transformation observed in the MCF-7-PKC-a
 cell contrast with the MCF-7-MDR clones that also overex-
 press PKC-a. In the MCF-7-MDR subclones, PKC-a overex-
 pression, relative to parental MCF-7 cells, is at least as much
 as or greater than that seen in MCF-7-PKC-a cells (34). The
 lack of similar morphologic and proliferative alterations in
 MCF-7-MDR cells suggests that PKC-a may not be directly
 responsible for the entirety of the changes observed in the MCF-
 MCF-7 Cells Transfected with Protein Kinase C-a 1911
 B
Cells: Vector MCF-7-PKCca
 Colonies: 0 227 ± 8.9
 ( +±SEM)
 Figure 7. Anchorage-independent growth of MCF-7-PKC-a cells. Growth in soft agar of vector-transfected and MCF-7-PKC-a cells was performed
 as described in Methods. A colony was defined as a group of > 50 cells. Colony growth is expressed as X +SEM, n = 4. This experiment was
 repeated with similar results.
 7-PKC-a cells. Possibly explaining this discrepancy are differ-
 ences in expression of other PKC isoforms in the MCF-7-PKC-
 a and MCF-7-MDR cells. In MCF-7-MDR cells, PKC-e is
 undetectable and significant decreases in the /3 and 6 isoforms
 are detected (34). In contrast, PKC-/3 expression is enhanced,
 levels of PKC-6 are only modestly decreased, and PKC-e con-
 tent is unchanged in MCF-7-PKC-a cells. In addition, content
 of PKC-iq is markedly decreased in MCF-7-PKC-a cells, and
 to our knowledge the status of this isoform in MCF-7-MDR
 has not yet been reported. The decreased PKC-27 content in
 MCF-7-PKC-a cells. may be related to the less differentiated
 status of this breast epithelial cell. In other epithelial cell types,
 a positive correlation exists between the level of PKC-77 expres-
 sion and a more differentiated epithelial phenotype (35). Thus,
 the dissimilar phenotypes displayed by MCF-7-PKC-a and
 MCF-7-MDR cells could result from the differences in the
 array of expression of other PKC isoforms or an interplay be-
 tween the content of an individual PKC isoform or subfamily
 relative to that of another isoform or subfamily. Alternatively,
 other phenotypic alterations occurring in the MCF-7-MDR cell
 resulting from selection in doxorubicin could modify the re-
 sponse of this cell to the elevated levels of PKC-a. This latter
 possibility is suggested by the finding that overexpression of
 PKC-a in an MCF-7 cell line that was clonally isolated on the
 basis of its resistance to doxorubicin was reported to proliferate
 at approximately the same rate as the parental MCF-7 cell (36).
 An alternative explanation for the different phenotype ex-
 pressed by the MCF-7-PKC-a cell is that exposure to G418
 Table L Tumorigenicity of MCF- 7 and MCF- 7-PKC-a Cells in Nude Mice
 Site of metastasis
 Animals with Volume of primary Number of animals Lymph Perinephric
 Inoculated cell type primary tumor tumor with metastasis Lung nodes fat
 MCF-7 2/10* < 1 m3 0/10 0/10 0/10 0/10
 MCF-7-PKC-a 9/10t 3,383 mm3±937' 6/1011 4/101 2/7 3/10
 * Animals having either micro- or macroscopic primary tumors at 4 wk after inoculation with 107 cells. * The animal not exhibiting a primary
 tumor had cells inadvertently injected into the subcutaneous space. I X±SEM; n = 9; range: 630-7,344 mm3; vol was defined as the product of
 length x width x diameter in mm. 11 In one mouse, the contiguous muscle was extensively invaded by the primary tumor. 1 Metastases at all sites
 were microscopic in size. A denominator less than the 10 animals in each group indicated that the specific anatomic site was not examined in all
 animals.
 1912 Ways et al.
Figure 8. Histologic appearance of primary and metastatic lesions in nude mice. The primary tumors derived from mice injected with 7 x 106
 parental MCF-7 and MCF-7-PKC-a cells for 4 wk are shown in A and B, respectively.
 resulted in selection of a neomycin-resistant cell with dramati-
 cally altered phenotypic characteristics that fortuitously overex-
 pressed PKC-a. While it is impossible to totally exclude this
 possibility, we believe that this scenario is unlikely. In mock
 or vector-transfected MCF-7 cells, we have not observed the
 selection of a cell with a MCF-7-PKC-a-like phenotype. All
 G418-resistant cells from MCF-7 cells transfected with the neo-
 mycin resistance gene display an epithelioid appearance and
 have proliferative characteristics similar to parental MCF-7
 cells. To our knowledge, examples of epithelial cells selected
 in G418 exhibiting MCF-7-PKC-a-like phenotypic character-
 istics and concomitantly overexpressing PKC-a have not been
 reported. Thus, we believe that the phenotypic alterations oc-
 curring in the MCF-7-PKC-a cell were initiated by and are
 related to overexpression of PKC-a.
 While enhancing tumorigenicity in breast cancer cells, PKC-
 a overexpression in murine fibroblasts and R6 cells inhibits
 proliferation and does not lead to transformation (37, 38). In
 hematopoietic cells, increases in the endogenous expression of
 PKC-a are associated with differentiation and cessation of
 growth (13). These divergent effects indicate that alterations
 in proliferation or differentiation induced by PKC-a overex-
 pression are not an intrinsic property of this isoform, but are
 modulated by a dynamic interaction between cell-specific and
 maturationally related factors.
 The mechanism by which overexpression of PKC-a alters
 morphology, enhances proliferation, increases tumorigenicity,
 and enhances metastatic potential remains to be determined.
 The increased mass of PKC-a would be predicted to increase
 the basal level of kinase activation. An enhanced basal level of
 kinase activation leading to increased phosphorylation of cellu-
 lar substrates could induce the phenotype observed in the MCF-
 7-PKC-a cells. However, activation of PKC-a and other en-
 dogenous MCF-7 isoforms in the parental MCF-7 cell by either
 TPA or cell permeant diacylglycerol derivatives causes growth
 inhibition and differentiation (6-9). Thus, if an increase in the
 basal kinase activity was responsible for the observed changes
 in the MCF-7-PKC-a cell, then qualitatively different cellular
 responses are elicited by PKC-associated kinase activity stimu-
 lated by autocrine mechanisms or factors contained in serum
 (e.g., enhanced proliferation in MCF-7-PKC-a cells). Alterna-
 tively, PKC-a overexpression may use a kinase-independent
 mechanism to elicit the phenotypic alterations observed in the
 MCF-7-PKC-a cell. A kinase-independent mechanism is sup-
 ported by the temporal association of an increase in an under-
 phosphorylated, kinase-inactive species of PKC-a in TPA-
 treated MCF-7 cells that have become desensitized to TPA-
 induced growth inhibition and have resumed proliferation (6).
 Therefore, PKC-a overexpression, whether via direct protein-
 protein interactions or through a potential association with DNA
 by the cysteine-rich zinc fingers contained in the C1 domain,
 has the theoretical potential to elicit signal transduction by a
 kinase-independent mechanism. Further studies with kinase-de-
 fective PKC-a mutants will be required to directly assess this
 hypothesis. Lastly, the phenotypic alterations in MCF-7-PKC-
 a cells could be due in part to the high level of PKC-#B expres-
 sion or reduction in PKC-77 and -6 content. Given the low level
 of #3 expression in the parental cell, the signal transmitted by
 phorbol ester-induced activation of this isoform would be lim-
 ited relative to that elicited by stimulation of the other TPA-
 responsive isoforms (a, 6, 71, and e). Thus, the biologic re-
 sponses in TPA-treated parental MCF-7 cells would largely
 reflect activation of the a, 6, rq, and e isoforms. In MCF-7-
 PKC-a cells, PKC-,/ is highly overexpressed and would be
 predicted to contribute substantially to the basal level of kinase
 activity generated by PKC family members, while the contribu-
 tion from PKC-r1 and -6 would be less. As has been postulated
 by others ( 1, 2), differences in substrate specificity or subcellu-
 lar distribution of PKC-,/ could elicit qualitatively dissimilar
 cellular responses from those induced by the a, 6, i7, and E
 isoforms. Such differences between isoforms and the high de-
 gree of, overexpression with the concomitant reduction in r,
 and 6 expression in MCF-7-PKC-a cells could provide a possi-
 ble mechanism explaining the divergent growth and phenotypic
 characteristics of parental MCF-7 and MCF-7-PKC-a cells.
 The anchorage-independent growth of MCF-7-PKC-a cells
 demonstrates that anoikis that normally occurs upon detachment
 of epithelial cells from their matrix has been abrogated (39).
 A role for the PKC pathway in negatively modulating initiation
 of anoikis has been previously demonstrated in phorbol ester-
 treated Madin-Darby canine kidney epithelial cells (39). Our
 results extend this observation regarding the inhibitory effects
 of the PKC pathway on induction of anoikis to mammary epithe-
 lial cells and suggest a potential role for the conventional PKC
 MCF-7 Cells Transfected with Protein Kinase C-a 1913
isoforms in regulating this process and its clinical counterpart,
 metastasis.
 Expression of PKC-/3 was enhanced in the MCF-7-PKC-
 a cell. The human PKC-f promoter contains multiple enhancer
 elements, including AP-1, AP-2, SpI, E box, and octamer mo-
 tifs, and also contains a sequence involved in transcriptional
 silencing (40). Interestingly, in K562 erythroleukemia cells,
 both basal and TPA-inducible expression of PKC-,6 only require
 a sequence containing two SpI sites, an E box, and an octamer
 motif (40). Given that the AP-1 and AP-2 sites are not neces-
 sary for induction of this gene, it has been postulated that tran-
 scriptional activation occurs via a non-AP-1-dependent mecha-
 nism (40). Consistent with these results, our findings show that
 TPA treatment for a 24-h period does not increase expression
 of PKC-/3 mRNA transcripts in parental MCF-7 cells (data not
 shown). These findings suggest that, in MCF-7-PKC-a cells,
 induction of PKC-,6 expression occurs either as a direct effect
 of PKC-a overexpression by a kinase-independent mechanism
 or secondary to phenotypic changes associated with the MCF-
 7-PKC-a cell. The factors, and the level at which they act to
 modulate the endogenous expression of the ,f isoform, are under
 investigation in our laboratories.
 The clinical significance of these findings in the prognostic
 evaluation of malignant breast lesions and in developing novel
 therapeutic modalities seems promising. Assessing the level and
 array of specific PKC isoform expression in primary breast
 masses could be useful in gauging the metastatic potential and/
 or biological aggressiveness of the lesion. Such information
 could be of use in determining the appropriate degree of radia-
 tion or chemotherapeutic intervention, especially in circum-
 stances such as axillary node-negative disease. The ability of
 PKC-a overexpression to enhance proliferation and to induce
 a more biologically aggressive phenotype, either directly or
 indirectly via alterations in expression of other endogenous PKC
 isoforms, also indicates the potential promise of therapeutic
 modalities directed at inhibiting isoform(s) expression or func-
 tion. Currently, such methodology is available using antisense
 or ribozyme therapy directed against individual PKC isoforms
 or by downregulating PKCs with bryostatin-1 treatment (41,
 42). Thus, these findings open exciting new possibilities that
 could improve the evaluation and treatment of patients with
 certain forms of breast cancer.
 Acknowledgments
 We wish to thank Nancy Hamm and June Long for expert assistance
 in preparation of this manuscript, Dr. Jack Brinn and Dr. Alvin Volkman
 (both from East Carolina University School of Medicine) for their
 advice on morphologic interpretations, Dr. John Bradfield and the De-
 partment of Comparative Medicine (East Carolina University School of
 Medicine) for assistance in performing the nude mice experiments, and
 Dr. Susan Rittling and Dr. Gerhardt Ryffel for providing the plasmids
 used in this study.
 This work was partially supported by a National Institutes of Health
 grant (CA43023).
 References
 1. Nishizuka, Y. 1988. The molecular heterogeneity of protein kinase C and
 its implications for cellular regulation. Nature (Lond.). 334:661-665.
 2. Parker, P., G. Kour, R. Marais, F. Mitchell, C. Pears, S. Stabel, and C.
 Webster. 1989. Protein kinase C: a family affair. Mol. Cell. Endocrinol. 65:1-
 11.
 3. O'Brian, C., V. Vogel, S. Singletary, and E. Ward. 1989. Elevated protein
 kinase C expression in human breast tumor biopsies relative to normal breast
 tissue. Cancer Res. 49:3215-3217.
 4. Borner, C., R. Wyss, R. Regazzi, U. Eppenberger, and D. Fabbro. 1987.
 Immunological quantitation of phospholipid/Ca2+-dependent protein kinase of
 human mammary carcinoma cells: inverse relationships to estrogen receptors. Int.
 J. Cancer. 40:344-348.
 5. Lee, S., J. Karaszkiewicz, and W. Anderson. 1992. Elevated level of nuclear
 protein kinase C in multidrug-resistant MCF-7 human breast carcinoma cells.
 Cancer Res. 52:3750-3759.
 6. Fabbro, D., R. Regazzi, S. Costa, C. Borner, and U. Eppenberger. 1986.
 Protein kinase C desensitization by phorbol esters and its impact on growth of
 human breast cancer cells. Biochem. Biophys. Res. Commun. 135:65-73.
 7. Issandou, M., F. Bayard, and J. Darbon. 1988. Inhibition of MCF-7 cell
 growth by 12-0-tetradecanoylphorbol-13-acetate and 1,2 dioctanoyl-SN-glycerol:
 distinct effects on protein kinase C activity. Cancer Res. 48:6943-6950.
 8. Kennedy, M., L. Presligiacoma, G. Tyler, W. May, and N. Davidson. 1992.
 Differentiation effects of bryostatin 1 and phorbol ester on human breast cancer
 cell lines. Cancer Res. 52:1278-1283.
 9. Valette, A., N. Gas, F. Roubinet, M. Dupont, and F. Bayard. 1987. Influence
 of 12-0-tetradecanoylphorbol- 13-acetate on proliferation and maturation of human
 breast carcinoma cells (MCF-7): relationship to cell cycle events. Cancer Res.
 47:1615-1620.
 10. Seynaeve, C., M. Stetler-Stevenson, S. Sebers, G. Kaur, E. Sausville, and
 P. Worland. 1993. Cell cycle arrest and growth inhibition by the protein kinase
 antagonist UNC-01 in human breast carcinoma cells. Cancer Res. 53:2081-2086.
 11. Pears, C., G. Kour, C. House, B. Kemp, and P. Parker. 1990. Mutagenesis
 of the pseudosubstrate site of protein kinase C leads to activation. Eur. J. Biochem.
 194:89-94.
 12. McNeall, J., A. Sanchez, P. Gray, C. Chesterman, and M. Sleigh. 1989.
 Hyperinducible gene expression from a metallothionein promoter containing addi-
 tional metal-responsive elements. Gene (Amst.). 78:81-88.
 13. Ways, D., B. Messer, T. Garris, W. Qin, P. Cook, and P. Parker. 1992.
 Modulation of protein kinase C e by phorbol esters in the monoblastoid U937
 cell. Cancer Res. 52:5604-5609.
 14. Bradford, M. 1976. A rapid and sensitive method for the quantitation of
 microgram quantities of protein utilizing the principles of protein-dye binding.
 Anal. Biochem. 72:248-254.
 15. Ways, D., P. Cook, C. Webster, and P. Parker. 1992. Effect of phorbol
 esters on PKC-4. J. Biol. Chem. 267:4799-4805.
 16. Yasuda, I., A. Kishimoto, S. Tanaka, M. Tominaga, A. Sakurai, and Y.
 Nishizuka. 1990. A synthetic peptide substrate for selective assay of protein kinase
 C. Biochem. Biophys. Res. Commun. 166:1220-1227.
 17. Chomczynski, P., and N. Sacchi. 1987. Single-step method of RNA isola-
 tion by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Bio-
 chem. 162:156-159.
 18. McCubrey, J., L. Steelman, G. Sandlin, R. Riddle, and D. Ways. 1990.
 Effects of phorbol esters on an interleukin-3-dependent cell line. Blood. 76:63-
 72.
 19. Klein-Hitpass, L., M. Schorpp, V. Wagner, and G. Ryffel. 1986. An
 estrogen-responsive element derived from the 5' flanking region of Xenopus
 vitellogenin A2 gene functions in transfected human cells. Cell. 46:1053-1061.
 20. Heuijerjans, J., F. Pieper, F. Ramaekers, L. Timmermans, H. Kuijpers, H.
 Bloemendal, and W. Venrooij. 1989. Association of mRNA and elF-2a with the
 cytoskeleton in cells lacking vimentin. Exp. Cell Res. 181:317-330.
 21. Raymond, W., and A. Leong. 1989. Co-expression of cytokeratin and
 vimentin intermediate filament proteins in benign and neoplastic breast epithelium.
 J. Pathol. 157:299-306.
 22. Raymond, W., and A. Leong. 1989. Vimentin: a new prognostic parameter
 in breast carcinoma? J. Pathol. 158:107-114.
 23. Rio, M., J. Bellocq, B. Gairard, U. Rasmussen, A. Krust, C. Keohl, H.
 Calderoli, V. Schiff, R. Renaud, and P. Chambon. 1987. Specific expression of
 the pS2 gene in subclasses of breast cancers in comparison with expression of
 the estrogen and progesterone receptors and the oncogene ERB Z. Proc. Natl.
 Acad. Sci. USA 84:9243-9247.
 24. Westley, B., and H. Rochefort. 1980. A secreted glycoprotein induced by
 estrogen in human breast cancer cell lines. Cell. 20:353-362.
 25. Muss, H., and T. Kute. 1982. Flow cytometry in the management of breast
 cancer. In High-Risk Breast Cancer: Diagnosis. J. Ragaz and I. Ariel, editors.
 Springer-Verlag, Berlin. 103-119.
 26. Kurebayashi, J., S. McLeskey, M. Johnson, M. Lippman, R. Dickson, and
 F. Kern. 1993. Quantitative demonstration of spontaneous metastasis by MCF-7
 human breast cancer cells cotransfected with fibroblast growth factor-4 and LacZ
 Cancer Res. 53:2178-2187.
 27. Bae, S., G. Arand, H. Azzam, P. Pavasant, J. Torri, T. Frandsen, and E.
 Thompson. 1993. Molecular and cellular analysis of basement membrane invasion
 by human breast cancer cells in Matrigel-based in vitro assays. Breast Cancer
 Res. Treat. 24:241-255.
 28. Knight, W., R. Livingston, E. Gregory, and W. McGuire. 1977. Estrogen
 1914 Ways et al.
receptor as an independent prognostic factor for early recurrence in breast cancer.
 Cancer Res. 37:4669-4671.
 29. Rochefort, H. 1990. Cathepsin D in breast cancer. Breast Cancer Res.
 Treat. 16:3-13.
 30. Johnson, M., J. Torri, M. Lippman, and R. Dickson. 1993. The role of
 cathepsin D in the invasiveness of human breast cancer cells. Cancer Res. 53:873 -
 877.
 31. Ree, A., B. Landmark, S. Walaas, H. Lahooti, L. Eikvar, W. Eskild, and
 V. Hansson. 1991. Down regulation of messenger ribonucleic acid and protein
 levels for estrogen receptors by phorbol ester and calcium in MCF-7 cells. Endo-
 crinology. 129:339-344.
 32. Saceda, M., C. Knabbe, R. Dickson, M. Lippman, D. Bronzert, R. Lindsey,
 M. Gottardis, and M. Martin. 1991. Post-transcriptional destabilization of estrogen
 receptor mRNA in MCF-7 cells by 12-0-tetradecanoylphorbol-13-acetate. J. Biol.
 Chem 266:17809-17814.
 33. Tzukerman, M., X. Zhang, and M. Pfahl. 1991. Inhibition of estrogen
 receptor activity by the tumor promoter 12-0-tetradecanoylphorbol-13-acetate: a
 molecular analysis. Mol. Endocrinol. 5:1983-1992.
 34. Blobe, G., C. Sachs, W. Khan, D. Fabbro, S. Stabel, W. Wetsel, L. Obeid,
 R. Fine, and Y. Hannun. 1993. Selective regulation of expression of protein kinase
 C (PKC) isoenzymes in multidrug-resistant MCF-7 cells. J. Biol. Chem 268:658-
 664.
 35. Osada, S., Y. Hashimoto, S. Nomura, Y. Kohno, K. Chida, 0. Tajima, K.
 Kubo, K. Askimoto, H. Koizumi, Y. Kitamura, K. Suzuki, S. Ohno, and T. Kuroki.
 1993. Predominant expression of nPKC-77, a Ca2" independent isoform of protein
 kinase C in epithelial tissues, in association with epithelial differentiation. Cell
 Growth & Differ. 4:167-175.
 36. Yu, G., S. Ahmad, A. Aquino, C. Fairchild, J. Trepel, S. Ohno, K. Suzuki,
 T. Tsuruo, K. Cowan, and R. Glazer. 1991. Transection with protein kinase C-a
 confers increased multidrug resistance to MCF-7 cells expressing P-glycoprotein.
 Cancer Commun. 3:181-189.
 37. Borner, C., I. Fillipuzzii, I. Weinstein, and R. Imber. 1991. Failure of
 wild-type or a mutant form of protein kinase C-a to transform fibroblasts. Nature
 (Lond.). 353:78-80.
 38. Cacace, A., S. Guadango, R. Krauss, D. Fabbro, and I. Weinstein. 1993.
 The epsilon isoform of protein kinase C is an oncogene when overexpressed in
 rat fibroblasts. Oncogene. 8:2095-2014.
 39. Frisch, S., and H. Francis. 1994. Disruption of epithelial cell-matrix inter-
 actions induces apoptosis. J. Cell Biol. 124:619-625.
 40. Obeid, L., G. Blobe, L. Karolak, and Y. Hannun. 1992. Cloning and
 characterization of the major promoter of the human protein kinase C P gene. J.
 BioL Chem. 267:20804-20810.
 41. Prendiville, J., D. Crowther, N. Thatcher, P. Woll, B. Fox, A. McGowan,
 N. Testa, P. Stem, R. McDermott, M. Potter, and G. Pettit. 1993. A phase one
 study of intravenous bryostatin 1 in patients with advanced cancer. Br. J. Cancer.
 68:418-425.
 42. Philip, P., D. Rea, P. Thavasu, J. Carmichael, N. Stuart, H. Rockett, D.
 Talbot, T. Ganesan, G. Pettit, F. Balkwill, and A. Harris. 1993. Phase I study of
 bryostatin 1: assessment of interleukin 6 and tumor necrosis factor a induction
 in vitro. J. Nati. Cancer Inst. 85:1812-1818.
 MCF-7 Cells Transfected with Protein Kinase C-a 1915