Expression of a Novel Factor, com1, Is Regulated by 1,25-Dihydroxyvitamin D₃ in Breast Cancer Cells¹

After malignant transformation, tumor cells and their surrounding stroma produce a variety of factors within the tumor cell environment to promote tumor growth and metastasis (1). Moreover,current evidence suggests that metastasis formation is primarily the result of the ability of disseminated tumor cells to initiate and continue growth in the target organ (2, 3). We recently reported a novel approach, comparing the phenotypes of human breast carcinoma cells acquired from an early step in clinical tumor progression and the cell population isolated from experimental metastases formed by these tumor cells, to identify properties that might prevail in the metastatic cells. A factor termed com1 was identified as up-regulated in the metastatic cell population. com1 presumably represents a helix-turn-helix-type DNA-binding protein and may participate in the response of breast carcinoma cells to growth conditions offered by the target organ on formation of distant metastases (4).

In addition to its role in the calcium metabolism, vitamin D also promotes tissue differentiation and inhibits cellular proliferation. Several reports have described antiproliferative effects of vitamin D in breast cancer cells in culture (5, 6, 7), and it has been suggested that noncalcemic vitamin D analogues may have a clinical potential in the treatment of breast cancer (8, 9). Furthermore, epidemiological studies have shown an inverse correlation between average annual sunlight exposure and the incidence of breast cancer, suggesting an association between endogenous vitamin D production and breast carcinogenesis (10, 11).

The aim of the present study was to examine whether cellular com1 expression may be associated with vitamin d-dependent growth regulation of breast cancer cells. We used the human cell line MCF-7 and its derivative subline MCF7/LCC2, which reflect phenotypes of the carcinoma cells observed during clinical progression of breast cancer. The com1-negative MCF-7 cells (4) are highly responsive to estrogens but poorly tumorigenic in animal models(12). In contrast, the estrogen-independent MCF7/LCC2 cells, which are also tamoxifen resistant and possess a tumorigenic phenotype (12, 13), show constitutive com1 expression(4). The active metabolite of vitamin D,1,25(OH)₂D₃,³promoted growth inhibition of the MCF7/LCC2 cells mediated by induction of the cdk inhibitor p21 and concomitant hypophosphorylation of pRB. These events, ultimately leading to cell cycle arrest in the G₁ phase, were preceded by a transient induction of cellular com1 expression. Because expression of com1 mRNA was also inversely correlated to clonal growth of MCF-7 cell lines, we postulate that com1 may participate in the regulatory pathway involved in cellular growth control when recruited by inhibitory signals.

The MCF-7 human breast cancer cells were routinely grown in MEM containing phenol red (Life Technologies, Inc., Rockville, MD)supplemented with 5% FCS (Life Technologies, Inc.), insulin (5.0μ m; Sigma, St. Louis, MO), and glutamine (2.0 mm; Life Technologies, Inc.). The subline MCF7/LCC2 was routinely grown in MEM without phenol red (Life Technologies, Inc.)supplemented with insulin and glutamine as described above, in addition to 5% steroid-depleted FCS. The serum was stripped of endogenous steroids by treatment with charcoal as described previously(14). The MCF7/com1 and MCF7/pMAM cells were grown in the same medium as MCF-7 supplemented with Geneticin (G418; Sigma). Cell cultures were kept at 37°C in a humidified 5%CO₂ atmosphere and refed every 3–4 days. Twenty-four h before the start of all experiments, media were exchanged, and cells were further incubated in the absence of insulin(defined as experimental media). All experiments were conducted on cells in exponential growth phase.

The MCF-7 and MCF7/LCC2 cells were seeded in experimental media on culture dishes (6-cm in diameter; 2.5 × 10⁵ cells/dish). After 24 h (at the start of the experimental period), medium was changed, and the cells were grown in the absence or presence of 1,25(OH)₂D₃(10⁻⁷ m; a generous gift from Dr. L. Binderup; Leo Pharmaceutical Products,Ballerup, Denmark) for up to an additional 96 h. For both cell lines, the seeding of this particular cell number resulted in near confluent growth at the end of the incubation period. At time points 0,24, 48, 72, and 96 h, respectively, control and treated cells were harvested by trypsination, and trypan blue-excluding cells were counted using a Bürker’s counting chamber. All cell number determinations were performed in triplicate, and the internal variation in counts was <5%. Three separate sets of experiments with three parallel samples for each condition were performed.

Cells were harvested, fixed in 1% paraformaldehyde, and subsequently resuspended in 100% methanol for storage at −20°C. The staining procedure was performed for each sample in a 50-μl solution consisting of 5 units of biotinylated terminal transferase (Boehringer Mannheim, Mannheim, Germany), 0.5 nmol of biotin-16-dUTP (Boehringer Mannheim), 1.5 mm CoCl₂, and 0.1 mm DTT. After incubation for 30 min at 37°C, cells were washed with PBS and incubated for another 30 min in 50 μl of streptavidin-conjugated FITC (Amersham Pharmacia Biotech, Uppsala,Sweden) diluted 1:50 in PBS with 0.1% Triton X-100. Cell pellets were finally resuspended in 500 μl of PBS containing 0.1% Triton X-100,100 μg/ml RNase A, and 5 μg/ml propidium iodide. Stained cells were analyzed in a FACStar+ laser flow cytometer with excitation at 488 nm,and DNA content (integrated propidium iodide fluorescence, collected using linear amplification) versus replicative cell fractions (FITC fluorescence intensity, collected using logarithmic amplification) was measured.

Cells were harvested and homogenized in ice-cold lysis buffer [250 mm NaCl, 2 mm EDTA, 0.1% NP40, 1 mm DTT, 1 mm NaF, 1 mmorthovanadate, 60 mm β-glycerophosphate, 50 μg/ml phenylmethylsulfonyl fluoride, and 2 μg/ml each of aprotinin and leupeptin (pH 7.2); Ref. 15)]. Aliquots of 30 μg of total protein were separated on 7.5% (for detection of pRB) or 12%(for detection of the other proteins) SDS-PAGE and transferred to nitrocellulose membranes (Amersham Pharmacia Biotech) by standard methods. The membranes were blocked in TBS-T containing 5% nonfat dry milk for 1 h at room temperature and subsequently incubated with 1μg/ml dilutions of mouse antihuman antibodies in TBS-T for 2 h at room temperature. The antibodies were anti-pRB [G3–254; purchased from PharMingen (San Diego, CA)] and anti-p21 (C-19), anti-p27 (C-19),anti-cdk4 (C-22), and anti-cyclin D1 C-20 [all obtained from Santa Cruz Biotechnology (Santa Cruz, CA)]. After four washes with TBS-T,the membranes were incubated with a 1:6000 dilution of horseradish peroxidase-linked secondary antibody (Bio-Rad, Hercules, CA), and immunoreactive proteins were visualized with the enhanced chemiluminescence detection system (Amersham Pharmacia Biotech).

A PCR-generated DNA fragment containing the complete coding region of com1 (4) was ligated in-frame into the EcoRI site behind the Dex-inducible promoter of the mammalian expression vector pMAMneo (Clontech, Hampshire, United Kingdom). This construct was transfected into MCF-7 cells by electroporation. Individual stable transfectants were selected in media containing 2 mg/ml G418 and maintained in media containing 500 μg/ml G418. Control MCF7/pMAM transfectants were generated correspondingly,but with the use of an empty pMAMneo vector instead. The com1 transcript was found to be inducible by Dex in the MCF7/com1 cells.

Total RNA was extracted and analyzed by standard Northern blotting techniques. Samples of 10 μg of RNA were resolved by gel electrophoresis before transfer onto Hybond-N+ membranes (Amersham Pharmacia Biotech). The cDNA probes were labeled with[α-³²P]dCTP (Amersham Pharmacia Biotech) by using the random priming technique, and standard Church hybridization conditions were used. To evaluate the amounts of RNA loaded, the filters were rehybridized to a kinase-labeled oligonucleotide probe complementary to nucleotides 287–305 of human 18S rRNA. Finally, the autoradiographs were subjected to densitometric measurements in a Molecular Dynamics 300A laser densitometer, and the mRNA expression levels relative to 18S rRNA were calculated.

Soft agar cultures were performed in tubes by adding 0.2 ml of nude rat blood diluted 1:8 in experimental media, 0.6 ml of 0.5% agar (Difco Laboratories, Ltd., Surrey, United Kingdom) in experimental media, and a 0.2-ml suspension of cells grown for 24 h in experimental media(500 cells/culture sample). The cultures contained a final concentration of 5% steroid-depleted FCS for the MCF7/LCC2 cells or 15% FCS for the MCF-7, MCF7/pMAM, and MCF7/com1 cells. The tubes were incubated at 37°C in 5% CO₂, 5%O₂, and 90% N₂. Experimental media (1 ml) with the proper concentrations of 1,25(OH)₂D₃, Dex, or ethanol vehicle were added on days 7 and 14. Colonies of >2 mm were scored after 21 days of incubation using a Nikon stereomicroscope. Experiments with six to nine parallel samples for each condition were performed.

The statistical analyses were performed using the SigmaStat software program (Jandel, Erkrath, Germany) with a significance level of P < 0.05.

The MCF-7 and MCF7/LCC2 cells were cultured in the absence or presence of 1,25(OH)₂D₃(10^-7 m), and growth was measured after 0–96 h of incubation (Fig. 1). Both cell lines revealed an exponential growth pattern of untreated cells (a 6–8-fold increase in cell number after 96 h), whereas a striking difference was observed between the cell lines in the presence of 1,25(OH)₂D₃. The MCF-7 cells still showed exponential growth, although the growth rate was significantly reduced after 72–96 h of treatment. In the MCF7/LCC2 cells, the exponential growth pattern was abolished in the presence of 1,25(OH)₂D₃. However, a significant inhibitory effect compared with the control situation was not observed before 72 h.

To characterize the mechanism of growth inhibition by 1,25(OH)₂D₃, the cells were cultured in the absence or presence of 1,25(OH)₂D₃(10⁻⁷ m) for 48 h. First, cell cycle profiles were analyzed by flow cytometry(Fig. 2). The cell cycle distribution showed 1,25(OH)₂D₃-dependent accumulation of MCF7/LCC2 cells in the G₁ phase. Signals corresponding to apoptotic cells were almost absent, as shown by gates R2 (displaying apoptotic cell fractions) of the histograms. Subsequently, the phosphorylation status of pRB was examined (Fig. 3). Consistent with the flow cytometry data, an accumulation of the hypophosphorylated form of pRB was observed in the 1,25(OH)₂D₃-treated MCF7/LCC2 cells.

Among the factors involved in regulation of the cell cycle, p21 has been shown to be a primary target for transcriptional regulation by vitamin D (9, 16). Hence,mRNA expression of p21 and other factors involved in regulation of the G₁ phase was analyzed (Fig. 4,a). Both cell lines showed constitutive mRNA expression of the cdk inhibitors p21 and p27, as well as of cdk4, cyclin D1, and p53, all of which represent factors involved in G₁ cell cycle control(17). Of these, only p21 was significantly up-regulated (>2-fold; p21 mRNA: 18S rRNA ratio) by 1,25(OH)₂D₃, but again,this was observed solely in the MCF7/LCC2 cells. However, it has also been reported that vitamin d-dependent control of the involved cell cycle factors is generally exerted at the level of the proteins (9), particularly regulated degradation of cyclin D1 (18). Hence, protein expression of the key factors was determined (Fig. 4 b). Yet again, both cell lines showed constitutive expression of p21, p27, cdk4, and cyclin D1, but only p21 seemed to be a target for 1,25(OH)₂D₃-dependent regulation. The induction of p21 at the protein level in the MCF7/LCC2 cells was closely identical with the corresponding regulation of its mRNA.

Finally, neither baseline nor 1,25(OH)₂D₃-dependent expression of mRNAs for cdk2, the cdk4 inhibitors p15 and p16 (17), or the tumor suppressor BRCA1, which may trans-activate the 5′-flanking region of the p21 gene (19), was observed in any cell line (data not shown).

To examine whether cellular com1 expression may be associated with the observed growth inhibition, com1 mRNA was measured in the MCF-7 and MCF7/LCC2 cells after treatment with 1,25(OH)₂D₃(10⁻⁷m) for 24 h. For comparison, the cells were also incubated in the presence of 10⁻⁷m each of E₂, Dex, or RA because these steroids are also known to regulate breast cancer growth(Fig. 5). In the com1-negative MCF-7 cells, expression of com1 mRNA was below the level of detection, irrespective of the mode of treatment. Whereas E₂, Dex, or RA did not significantly alter the baseline expression of com1 mRNA in the MCF7/LCC2 cells, a surprising up-regulation of com1 mRNA was observed in the presence of 1,25(OH)₂D₃.

Based on the previous experiments, the MCF7/LCC2 cells were incubated in the presence of 1,25(OH)₂D₃(10⁻⁷ m) for 0–48 h, and com1 mRNA was analyzed (Fig. 6). The expression of com1 mRNA showed a biphasic induction(5–8-fold) with a peak level at 24 h followed by a gradual decline after 36–48 h of 1,25(OH)₂D₃ treatment. Hence, the kinetics of this induction was much faster than the cell cycle events leading to the growth arrest of the MCF7/LCC2 cells. After 36–48 h of incubation, an up-regulation of com1 mRNA was also observed in the control cells.

The 1,25(OH)₂D₃-dependent regulation of com1 seemed to be an early regulatory event in growth inhibition of the MCF7/LCC2 cells. The question of whether com1 might also be involved in long-term growth control was assessed by a colony formation assay. First, MCF7/LCC2 cells dispersed in soft agar were incubated with increasing concentrations of 1,25(OH)₂D₃(10⁻¹¹ to 10⁻⁷ m; Fig. 7). Whereas the lower concentrations(10⁻¹¹ to 10⁻⁹ m) were associated with significant stimulatory effects, half-maximal inhibition of clonal growth was observed at a concentration of∼10⁻⁸ m, and complete suppression of colony formation was obtained at 10⁻⁷ m.

To strengthen the association between com1 expression and clonal growth of the MCF-7 cell lines, the com1-negative MCF-7 cells were stably transfected with a Dex-inducible com1 construct, and the resulting MCF7/com1 cells were compared with the wild-type parental cells and MCF7/pMAM control transfectants with regard to colony formation in soft agar (Fig. 8). Dex at concentrations of 10⁻⁷ to 10⁻⁵m caused inhibition (30–35%) of clonal growth of both the MCF-7 and MCF7/pMAM cells (Fig. 8,b). In contrast, clonal growth of the MCF-7/com1 cells was inhibited by 65–75% in the presence of Dex (Fig. 8,b). Importantly,the transfected cells expressing Dex-inducible com1 mRNA(Fig. 8,a) also revealed significant down-regulation(50–60%) of colony formation when compared with the MCF-7 and MCF7/pMAM cells treated with Dex (Fig. 8 b).

In the present study we demonstrated that 1,25(OH)₂D₃ inhibited in vitro growth of the com1-positive MCF7/LCC2 cells. This was associated with up-regulation of p21 and the subsequent accumulation of the hypophosphorylated form of pRB, ultimately leading to cell cycle arrest in G₁ phase. Moreover, a 1,25(OH)₂D₃-dependent induction in com1 mRNA expression was closely associated with the growth-inhibitory effect, suggesting that com1 may participate in the regulatory pathway involved in cellular growth inhibition.

Vitamin D is a recognized inhibitor of cell proliferation in model systems of several human malignancies (5, 6, 7, 9, 20, 21, 22, 23, 24). The cell cycle events observed in the 1,25(OH)₂D₃-treated MCF7/LCC2 cells are consistent with the identification of a vitamin D response element in the 5′-flanking region of the p21 gene and with the notion that the VDR trans-activates p21 independently of the tumor suppressor p53(16).

We recently identified com1 as a nuclear factor that may mediate the proliferative response of breast carcinoma cells on establishment in secondary organs during experimental metastasis formation(4). Based on these results, it is rather intriguing that treatment of the MCF7/LCC2 cells with 1,25(OH)₂D₃ caused an up-regulation of the level of com1 mRNA. This vitamin d-dependent effect seemed to be rather specific because the other steroid hormones and steroid-like factors tested(E₂, Dex, and RA), all of which are known to modulate experimental or clinical progression of breast cancer, did not alter the expression of com1 mRNA.

The VDR is a ligand-operated transcription factor, which principally acts in a heterodimer complex that stimulates target gene transcription via vitamin D response elements (25). Thus far, only a few primary vitamin D-responding genes have been identified(9), and p21 is one of these(16). The possibility that the com1 gene is transcriptionally activated by the VDR is particularly appealing but still unproven and demands comprehensive experiments on the still uncharacterized 5′-flanking region of the human com1 gene.

The facts that the kinetics of the transient 1,25(OH)₂D₃-dependent induction of com1 mRNA was much faster than the cell cycle events leading to growth arrest of the MCF7/LCC2 cells and that the growth rate of the com1-negative MCF-7 cells was also inhibited by 1,25(OH)₂D₃ argue against a direct involvement of com1 in cell cycle regulation. Our preliminary observation that the level of p21 mRNA in the MCF7/com1 cell line does not apparently differ from that in the MCF-7 wild-type or MCF7/pMAM control cells (data not shown) is another argument. The increase in the level of com1 mRNA in control MCF7/LCC2 cells after 36–48 h of incubation, which is followed by a shift toward hypophosphorylated pRB and G₁ arrest after 96 h (data not shown), illustrates the same concept. This growth retardation of the control cells is presumably due to consumption and a subsequent lack of nutrients in the medium exchanged no later than at time 0. Nevertheless, the striking inverse correlation between com1 mRNA expression and colony formation in the MCF7/com1 cells supports the notion of some growth-regulatory effect of com1,although the suppression of clonal growth of the com1-transfected cell line was incomplete.

The gene sequence for com1 is earlier described as the rat cDNA analogue p8 by Mallo et al.(26). The authors found that pancreatic expression of p8 was strongly enhanced after stimuli that induce apoptosis(26). The kinetics of these p8 mRNA responses seems closely comparable with our observations on 1,25(OH)₂D₃-dependent regulation of com1 mRNA. Several reports have shown that 1,25(OH)₂D₃ and vitamin D analogues activate apoptotic death pathways in breast cancer cells(27, 28, 29, 30). In contrast to this, our flow cytometry analysis revealed hardly any signals corresponding to apoptotic cells. However,we did not specifically examine other apoptotic features of the 1,25(OH)₂D₃-treated MCF7/LCC2 cells.

Mallo et al. (26) also demonstrated high levels of p8 mRNA expression in developing and regenerating rat pancreas and liver. The authors also showed increased activity in cellular growth assays, but no alterations were seen in response to apoptotic stimuli on cellular overexpression of p8 by transfection (26, 31). Our own initial identification of human com1 was from actively proliferating metastatic tumors (4), representing biological entities in which the relative fraction of apoptotic cells is low (32). We have also found that the expression levels of com1 mRNA in human breast tumors are consistently and significantly higher than those in the adjacent normal breast tissue (33). Notably, the com1 gene has a localization within chromosome 16 at position p11.2 (31), a chromosomal region occasionally amplified in primary breast tumors (34, 35, 36), which supports the assumption that com1 may function as a growth-promoting factor in the development of malignant breast tumors.

It is conceivable to assume that com1/p8 plays some regulatory role in cellular growth, although the previously published (4, 26, 31, 33) and present data are apparently conflicting. We speculate that com1 may mediate growth stimulation as well as inhibition by being recruited by preferentially stimulatory or inhibitory signals and perhaps by acting with different nuclear partners in a cell- or tissue-specific manner. To analyze the effect of com1 alone, we established the MCF7/com1 transfectants from the com1-negative MCF-7 cell line. Although we used a construct demanding the use of a potent corticosteroid analogue with an inhibitory effect on cellular proliferation, the induced MCF7/com1 cells revealed significant inhibition of colony formation in the soft agar assay compared with both the wild-type MCF-7 cells and MCF7/pMAM control transfectants treated with Dex.

In conclusion, the present study demonstrated that 1,25(OH)₂D₃ exerts highly regulated antiproliferative control of the com1-positive MCF7/LCC2 cell line. The regulatory mechanism was confirmed to involve the cell cycle inhibitor p21, and a striking association with a preceding com1 mRNA induction was observed. The MCF7/com1 cell line established from the com1-negative MCF-7 cells seemed to mimic 1,25(OH)₂D₃-treated MCF7/LCC2 cells in a clonal growth assay. These results apparently indicate that rather than promoting growth, com1 may mediate growth inhibition of MCF-7 cell lines.

We thank Dr. N. Brünner (Finsen Laboratory,Rigshospitalet, Copenhagen, Denmark) for the gift of the MCF7/LCC2 cells; Dr. L. Binderup (Leo Pharmaceutical Products, Ballerup, Denmark)for the gift of the vitamin D metabolite; Dr. B. Vogelstein (The John Hopkins University School of Medicine, Baltimore, MD) for providing p21 cDNA; Dr. O. Myklebost for providing the cDNAs for p27, p15, and p16 and Dr. B. Smith-Sørensen for providing p53 cDNA (both at the Norwegian Radium Hospital, Oslo, Norway); Dr. P. Meltzer (NIH,Bethesda, MD) for providing cdk4 cDNA; Dr. D. Beach (Howard Hughes Medical Institute, Cold Spring Harbor, NY) for providing cyclin D1 and cdk2 cDNAs; and Dr. P. Devilee(Medical Genetics Center South-West Netherlands, Leiden, the Netherlands) for providing BRCA-1 cDNA.