| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Molecular Pharmacology, Albert Einstein College of Medicine, Bronx, New York 10461
Address all correspondence and requests for reprints to: Nancy Carrasco, Department of Molecular Pharmacology, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461. E-mail: carrasco{at}aecom.yu.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The prime example of targeted radiation therapy via a selectively expressed plasma membrane transporter is radioiodide therapy. This is the most effective anticancer targeted radiotherapy available today (2) and has been employed for more than 60 yr to destroy thyroid cancer remnants and/or metastases after thyroidectomy (3). The presence of the sodium/iodide symporter (NIS) in thyroid cancer cells ensures that administered radioiodide is selectively accumulated in these cells, causing little damage to other cells and only minimal side effects (2). Thus far, radioiodide therapy has been viewed as applicable only to thyroid cancer. Nevertheless, recent observations have raised the possibility of applying radioiodide therapy to breast cancer (4, 5, 6), and potentially to other cancers as well, by introducing NIS into the tumor via viral vectors (7, 8, 9) or up-regulating the tumors’ endogenous NIS expression, if present (2, 4, 10, 11).
NIS was cloned in 1996 in our laboratory from the highly functional rat thyroid-derived FRTL-5 cells (12). It was subsequently demonstrated that the same NIS protein mediates I– transport in all tissues that concentrate I–, including salivary glands, gastric mucosa, and lactating mammary gland (4, 13, 14). Interestingly, NIS is regulated differently in each of these tissues (4, 14, 15). Our secondary structure model for NIS predicts 13 transmembrane segments (16); we have experimentally demonstrated that the amino terminus faces extracellularly and the carboxy terminus faces intracellularly (17).
Driven by the Na+ gradient generated by the Na+/K+ ATPase, NIS couples the inward downhill transport of Na+ down its concentration gradient to the inward uphill translocation of I– against its electrochemical gradient. NIS transports 2 Na+ per each I– (18). Importantly, in addition to radioiodide, NIS also transports other radionuclides with high therapeutic potential, such as 188ReO4 (19) and 211At (20).
We have demonstrated in healthy mammary tissue that NIS is expressed in lactating but not in nonlactating breast and have shown that NIS expression is regulated by lactogenic hormones (4). By comparison, in the thyroid NIS is expressed continuously and is primarily regulated by TSH (14, 15, 17). We have observed NIS-mediated I– uptake in mammary adenocarcinomas in transgenic mice bearing ras or neu oncogenes and, even more significantly, we found NIS expression in more than 80% of human breast cancers by immunohistochemistry (4, 6, 21), whereas there was no NIS expression in the surrounding noncancerous breast tissue. In short, some regulatory mechanisms cause NIS to be expressed only in lactating breast and breast cancer but not in noncancerous nonlactating breast. Furthermore, we have recently demonstrated in vivo that NIS-mediated I– uptake occurs in human breast cancer metastases, and that thyroidal I– uptake can be selectively down-regulated by administering thyroid hormones, which suppress TSH production in the anterior hypophysis (6). This would protect the thyroid from radioiodide administered to destroy breast cancer cells. To reach maximal therapeutic efficiency, patients with metastatic thyroid disease are given a low iodide diet and recombinant TSH (22). This greatly increases functional NIS expression and radioiodide accumulation in malignant thyroid tissue. Similarly, it is imperative to find comparable ways to selectively induce or increase existing endogenous functional NIS expression in breast cancer. Kogai et al. (11) have reported the induction of functional NIS expression by all-trans-retinoic acid (tRa) in the human adenocarcinoma-derived MCF-7 cell line. MCF-7 cells are an excellent and widely used in vitro system for studying NIS regulation in human breast cancer cells. Here we analyzed the effects of hormones and other factors on NIS expression in MCF-7 cells. We found that glucocorticoids and purinergic regulatory mechanisms greatly enhanced tRa-induced functional NIS expression in MCF-7 cells. Interestingly, factors that increased NIS expression in MCF-7 cells have the opposite effect in FRTL-5 cells, i.e. they down-regulate NIS expression. Our findings open new strategies for carrying out selective pharmacological modulation of radioiodide uptake in breast cancer.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
, horizontally striped bar) when compared with tRa treatment alone (Fig. 1A, diagonally striped bar), leading to I– accumulation levels comparable to those in thyroid cells (Fig. 1A, gray bar). In contrast, prolactin, even at doses as high as 100 ng/ml to 1µg/ml, had no effect on steady-state I– uptake in cells simultaneously treated with either tRa alone or tRaH (data not shown).
|
). These findings indicate that hydrocortisone increases tRa-induced I– accumulation in MCF-7 cells solely by increasing NIS-mediated I– transport, without affecting I– efflux.
To further explore the effect of different hormonal treatments on NIS function in MCF-7 cells, we performed kinetic analyses of NIS-mediated I– uptake (Fig. 1, C and D). We measured the effect of varying concentrations of I– (ranging from 0.625 to 80 µM) on the initial rates (2 min) of I– transport (Fig. 1C). In all cases saturation was reached at an I– concentration of approximately 40 µM. The apparent Vmax value of NIS was significantly higher (2-fold) when the combined tRaH treatment was used, as compared with tRa alone [Vmax (I–tRa) = 4.5 ± 0.1; Vmax (I–tRaH) = 9.2 ± 0.4 pmol I–/µg DNA/2 min]. In contrast, the calculated Michaelis-Menten constant (Km) values for I– were similar in both conditions [Km (I–tRa) = 7.5± 1.2; Km (I–tRaH) = 10.6 ± 1.0 µM]. We also investigated the effect of varying concentrations of Na+ (ranging from 0–140 mM) on initial I– uptake rates in MCF-7 cells treated with either tRa or tRaH (Fig. 1D). Osmolarity was kept constant with choline chloride. We found that, similar to I–, the Km of NIS for Na+ was the same in cells subjected to either treatment [Km (Na+-tRa) = 48.3 ± 3.2, Km (Na+-tRaH) = 42.4 ± 1.6 mM], whereas the Vmax of Na+ transport was also increased 2-fold [Vmax (Na+-tRa) = 3.0 ± 0.1, Vmax (Na+-tRaH) = 6.1 ± 0.1 pmol I–/µg DNA/2 min] when the combined treatment was employed. These data suggest that the addition of hydrocortisone to tRa results in the presence of more functional NIS molecules in the plasma membrane of MCF-7 cells.
The Combination of tRa and Hydrocortisone Stimulates NIS Protein Expression and Targeting to the Plasma Membrane
Having established that the combination of tRa and hydrocortisone increases I– uptake, we sought to discern the underlying mechanism involved. We assessed NIS expression and its subcellular localization in MCF-7 cells treated with either tRa alone or tRaH (Fig. 1, E–I). When present, the NIS protein migrated as an approximately 100-kDa mature and an approximately 60-kDa partially glycosylated polypeptide (17), as determined by immunoblot analysis (Fig. 1E). We found that both NIS protein expression and targeting to the plasma membrane increased significantly in cells treated with the tRaH combination compared with those treated with tRa alone (Fig. 1, E–F). NIS targeting to the plasma membrane was assessed by surface biotinylation experiments carried out with the amino-specific membrane-impermeable reagent Sulfo-NHS-SS-biotin. The entire biotinylated fraction was isolated with streptavidin-coated beads and immunoblotted with anti-hNIS Ab. Whereas an approximately 100-kDa immunoreactive band corresponding to mature NIS was evident in the immunoblot, the approximately 60-kDa partially glycosylated band was not observed because it does not reach the plasma membrane (Fig. 1F). Prolactin had no effect on NIS protein expression, independently of whether it was added alone or together with the tRaH combination (data not shown).
The MCF-7 cell line is heterogeneous (26). Hence, whereas NIS was clearly revealed by immunofluorescence to be localized in the plasma membrane of tRaH-treated MCF-7 cells (Fig. 1I), not all cells expressed NIS. By flow cytometry [fluorescence-activated cell sorting (FACS)], we quantitated the percentage of cells expressing NIS after hormonal treatment (Fig. 1, J–L), and observed that tRa induced NIS expression in only 10% of the cells (Fig. 1K). By comparison, the tRaH combination induced NIS expression in as many as 30% of the cells (Fig. 1L), indicating that the addition of hydrocortisone led not only to higher protein levels in tRa-responsive MCF-7 cells, but also enhanced the stimulatory effect of tRa. Virtually identical results were obtained with two other Abs directed against different NIS epitopes.
We determined the time course and dose dependence of the hydrocortisone effect on tRa-induced I– uptake and NIS protein expression in MCF-7 cells (Fig. 2). The presence of 10 µM hydrocortisone for more than 12 h significantly increased I– uptake (Fig. 2, A and B), NIS protein expression (Fig. 2, C and D), and plasma membrane targeting (Fig. 2, E and F). Incubation times up to 48 h resulted in no further increase in NIS protein expression and I– uptake (Fig. 2). In the presence of tRaH, I– uptake and NIS protein expression were maintained at a constant level, even after 6 d of treatment (data not shown). In contrast to hydrocortisone, none of the other lactogenic hormones with known roles in NIS regulation in mammary glands in mice, namely oxytocin, prolactin, and estrogen, alone or in combination, induced NIS expression or had any effect on tRa- or tRaH-induced NIS expression or NIS trafficking, or on I– uptake in MCF-7 cells (data not shown). Insulin or IGF-I treatment resulted in a slight decrease of tRa- or tRaH-induced I– uptake (data not shown).
|
). Surprisingly, IBMX by itself caused a significant increase (
30%) of I– uptake in both tRa- and tRaH-treated MCF-7 cells (Fig. 3A). Immunoblot and surface biotinylation experiments also showed that the tRaH-IBMX treatment resulted in both higher overall NIS protein expression and a higher number of NIS molecules in the plasma membrane, as compared with tRaH-treated cells (Fig. 3, B and C).
|
, G and H). The Km values of NIS for both I– and Na+ were very similar [Km (I–tRaH) = 2.89 ± 0.76 µM; Km (I–tRaH-IBMX) = 5.1 ± 1.16 µM Km (Na+tRaH) = 41.7 ± 7.5 mM; Km (Na+tRaH-IBMX) = 41.28 ± 4.9 mM], whereas the Vmax values for both I– and Na+ increased significantly when IBMX was added along with tRaH [Vmax (I–tRaH) = 2.67 ± 0.14 pmol; Vmax (I–tRaH-IBMX) = 7.82 ± 0.46 pmol; Vmax (Na+tRaH) = 3.44 ± 0.28 pmol; Vmax (Na+tHRa-IBMX) = 5.98 ± 0.34 pmol].
Although forskolin (20 µM) (an adenyl cyclase agonist) induces functional NIS expression in TSH-deprived FRTL-5 cells (27, 30), it significantly decreased tRaH-induced NIS activity in MCF-7 cells, even after a short administration (30 min) (Fig. 3D). These results are different from the lack of effect of forskolin in MCF-7 cells reported by Kogai et al. (10).
ATP Increases I– Uptake in tRaH-Treated MCF-7 Cells
That cAMP slightly decreased IBMX stimulation of tRaH-induced I– uptake and forskolin decreased tRaH-induced I– uptake suggested that the IBMX effect was independent of the intracellular levels of cAMP. 8-Br-cGMP (up to 500 µM) had no effect on tRaH-induced I– uptake in MCF-7 cells (data not shown). Hence, the stimulatory effect of IBMX on I– transport may be related to the interaction of IBMX with xanthine-sensitive purinergic P2 receptors (31). When we administered ATP along with tRaH, we observed a moderate but statistically significant increase in NIS protein expression (Fig. 4, C and D) and plasma membrane targeting (Fig. 4, E and F), consistent with the increase in I– uptake (Fig. 4, A and B). Maximal effect was achieved by 500 µM ATP administered for more than 18 h (Fig. 4, A–F). ATP (500 µM) administered alone, without tRa or hydrocortisone, did not induce functional NIS expression in MCF-7 cells (data not shown). We measured initial rates of I– uptake in MCF-7 cells treated with tRaH and ATP and found a similar increase in Vmax to the one observed in cells treated with tRaH and IBMX (Fig. 4G).
|
). All experiments were carried out in tRaH-treated MFC-7 cells. The effect of administering IBMX and ATP together was indistinguishable from the effects of administering IBMX or ATP alone (Fig. 5, A–C). Two ATP analogs were tested, namely ATP-
-S, which interacts with the P2Y1, P2Y2, P2Y11, and P2Y12 receptors, and ADP-ß-S, which interacts with the P2Y1, P2Y11, and P2Y12 receptors. Whereas treatment with 100 µM ATP--S had a stimulatory effect similar to that of 500 µM ATP, 100 µM ADP-ß-S had no effect (Fig. 5, A–C). Because ATP--S binds with equally high affinity with both P2Y1 and P2Y2, to determine which of these two receptors was involved, we tested UTP, which interacts with P2Y2 but not with P2Y1. We found that UTP (100 or 500 µM) modestly increased I– uptake and NIS protein expression (Fig. 5, D–F). This suggests that the receptor involved is P2Y2.
|
NIS Regulation Differs Markedly in Thyroid vs. Breast Cancer Cells
Functional NIS expression in FRTL-5 cells requires TSH (10, 27), whereas in MCF-7 cells it requires tRa (10). Interestingly, Schmutzler et al. (33) have reported that 1 µM tRa actually down-regulated NIS mRNA levels and decreased I– uptake in FRTL-5 cells, i.e. tRa had the opposite effect on NIS in thyroid (FRTL-5) cells compared with its effect in breast (MCF-7) cells. Here we have shown, for the first time, that beyond its effect on NIS mRNA and I– uptake, tRa also decreased TSH-induced NIS protein expression and plasma membrane targeting in FRTL-5 cells (Fig. 6, A–C). Hydrocortisone, for its part, has been reported to lower TSH-induced functional NIS expression in thyroid cells (34), an observation we have confirmed by measuring decreased TSH-induced steady-state I– uptake in FRTL-5 cells in response to hydrocortisone treatment (data not shown). In contrast, we showed that hydrocortisone markedly increased tRa-induced I– uptake in MCF-7 cells (Fig. 1). NIS expression in thyroid and breast cancer cells is regulated by different purinergic signals. The presence of multiple P2Y receptor subtypes has been proven in FRTL-5 cells, and these receptors have been suggested to participate in regulation of DNA synthesis (35, 36).
|
and 5) probably via P2Y2 receptors. Surprisingly, when we tested the effect of ATP on NIS expression in FRTL-5 cells, we found that, in contrast to MCF-7 cells, ATP significantly reduced TSH-induced NIS protein expression, plasma membrane targeting, and steady-state I– uptake (Fig. 6, A–C). IBMX increased I– uptake both in FRTL-5 (30) and MCF-7 cells (Fig. 4), probably by different mechanisms.
Other than TSH, I– is the main regulator of NIS expression and thus of its own transport in the thyroid. The Wolff-Chaikoff effect, first described in 1948, is the blocking of thyroid I– organification (covalent incorporation of I– into thyroglobulin) by high concentrations of I– itself (37). Shortly after this condition is established, the thyroid gland protects itself against I– overload by down-regulating the transport of the anion, leading to restoration of I– organification. This latter physiological phenomenon is called the "escape" from the acute Wolff-Chaikoff effect (38, 39). With the availability of the NIS cDNA and NIS Abs, it was shown that high I– doses down-regulate NIS protein expression, and consequently I– transport, mostly by a posttranslational mechanism that remains to be fully characterized (39, 40, 41, 42). To examine whether I– modulates NIS activity in breast epithelial cells, FRTL-5 and MCF-7 cells were incubated with 100 µM Na125I overnight, and then assayed for I– transport and NIS protein expression. Surprisingly, I– preincubation had no effect either on I– transport or NIS protein expression in MCF-7 cells (Fig. 6, D–F, right panels). This contrasted both with the expected and previously reported (39, 40, 41, 42) inhibition of I– uptake by I– preincubation in thyroid cells (Fig. 6D, left panel), and with the presently observed decrease in expression at the plasma membrane by I– preincubation (Fig. 6F, left panel), an effect that had not been shown before.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
200,000 newly diagnosed cases per year in the United States), endogenous NIS expression is a distinct advantage of this disease over other malignancies for the purpose of treating it with radioiodide therapy, because it is a cancer that only requires optimization of its existing intrinsic NIS activity without the need to introduce NIS into tumor cells by means of gene transfer techniques (4, 5, 6). An understanding of NIS regulation is a key requirement for optimizing radioiodide therapy in breast cancer, as has been done in thyroid cancer. Therefore, we have investigated the regulation of NIS activity by a variety of factors in MCF-7 cells, a frequently used in vitro model system to study regulatory effects in human breast adenocarcinoma, and observed that NIS regulation in these cells is markedly different from that in thyroid FRTL-5 cells. Kogai et al. (10, 43) have demonstrated that tRa (or some tRA analogs, like 9-cis-retinoic acid and AGN190168) is an absolute requirement to induce NIS expression in MCF-7 cells, an observation that we have confirmed in this study. In contrast, Arturi et al. (44) reported that iodide uptake in MCF-7 cells could be induced by insulin, IGF-I, IGF-II, and prolactin in the absence of tRa, a finding that we could not reproduce either under serum-starved or nonstarved conditions. Moreover, when we added these growth factors simultaneously with tRa or tRaH, insulin or IGF-I actually had a slight inhibitory effect (data not shown).
Being convinced that tRa treatment is an absolute requirement to elicit NIS expression in MCF-7 cells, we sought to test modulators that might increase the tRa effect, such as hydrocortisone, a hormone that induces milk protein synthesis in cell and organ cultures when given in combination with insulin and prolactin (23, 24, 25). Indeed, we found that hydrocortisone significantly enhanced tRa-induced I– transport in MCF-7 cells (Fig. 1A). In a time- and dose-dependent fashion, hydrocortisone increased the Vmax (Fig. 1, C and D) of tRa-induced I– transport in MCF-7cells by increasing total (Fig. 1E) and cell surface NIS protein expression (Fig. 1F), without affecting I– efflux (Fig. 1B). Hydrocortisone alone (i.e. without tRa) did not have a significant effect (Fig. 1A). Our observations confirm and extend those from a recent report by Kogai et al. (43) showing that dexamethasone increased tRa-induced I– uptake in MCF-7 cells. These authors suggested that hydrocortisone stabilizes tRa-induced NIS mRNA (43). Future studies are necessary to elucidate the mechanism of the tRa-hydrocortisone effect.
MCF-7 cells are known to be heterogenous (26). When treated with tRa, only 10% of cells exhibited NIS expression (Fig. 1K). In contrast, when cells were treated with the tRaH combination, as many as 30% of the cells expressed NIS protein (Fig. 1L). This clearly indicates that hydrocortisone significantly enhances the turn-on effect of tRa on NIS expression. Not much is known about the mechanism by which tRa up-regulates NIS expression in MCF-7 cells. Recently, Dentice et al. (45) observed that, in MCF-7 cells, tRa treatment induces the expression of Nkx-2.5, a cardiac homeobox transcription factor that, in turn, is a potent inducer of the NIS promoter.
In the thyroid, a major modulator that markedly up-regulates NIS expression is cAMP (27, 28, 29). Hence, we tested 8-Br cAMP, a membrane-permeable analog of cAMP; forskolin, an adenyl cyclase activator; and IBMX, a phosphodiesterase inhibitor in MCF-7 cells. All three lead to increased intracellular cAMP levels. Interestingly, far from stimulating it, 8-Br-cAMP caused a very slight decrease in both tRa- and tRaH-induced I– uptake. This finding alone indicates that cAMP does not play the same up-regulating role in NIS expression in breast cancer cells as it does in thyroid cells (15, 27). Kogai et al. (10) reported no effect of forskolin alone (i.e. without tRa) on NIS expression in MCF-7 cells, an observation we have confirmed (data not shown), demonstrating that, in the absence of tRa, cAMP alone does not up-regulate NIS in these cells. Like 8-Br-cAMP, forskolin also caused a decrease in both tRa- and tRaH-induced I– uptake in MCF-7 cells (Fig. 3D), although the decrease was much more pronounced than that with 8-Br-cAMP, and most likely was not due to higher intracellular cAMP levels. In stark contrast, Arturi et al. (44) reported that forskolin and dibutyryl-cAMP treatment produced a minimal increase of I– uptake in MCF-7 cells.
Given that both 8-Br-cAMP and forskolin caused a decrease in both tRa- and tRaH-induced NIS activity, it was most surprising to find that IBMX significantly increased, rather than decreased, I– uptake (both initial rates and steady state) and NIS protein expression in these cells (Fig. 3), an effect that was clearly independent of phosphodiesterase inhibition and of cAMP levels.
Although the above observations appear to be in conflict with a report by Knostman et al. (46) stating that cAMP induces NIS expression in breast cancer, there is no real discrepancy with our findings: these authors observed that 8-Br-cAMP treatment stimulated NIS mRNA expression in MCF-7 cells, but there was no concomitant increase in I– uptake. Like Kogai et al. (10) and our group, Knostman et al. (46) also found that forskolin alone did not increase I– uptake in MCF-7 cells. Thus, in light of our current findings, we conclude that cAMP does not induce NIS protein expression and function in MCF-7 cells. In contrast, IBMX up-regulates both I– uptake and NIS protein expression in these cells, albeit by a mechanism that does not involve either phosphodiesterase inhibition or an increase in intracellular cAMP levels. Therefore, we suggest that the up-regulating effect of IBMX on NIS expression in MCF-7 cells, also reported by Knostman et al., is unrelated to cAMP.
Having ruled out the possibility that the surprising non-cAMP-dependent stimulatory effect of IBMX on NIS expression was mediated by an increase in cGMP (data not shown), and considering that the effect was clearly not mediated by phosphodiesterase inhibition, we analyzed the potential involvement of purinergic signaling. Methylxanthines are known antagonists of adenosine receptors, and the possible existence of methylxanthine-sensitive purinergic receptors has also been described (31). We found that ATP, like IBMX, significantly increased tRaH-induced I– uptake in MCF-7 cells. A given epithelial cell type commonly expresses a mixture of various purinergic receptors, and the combination of the different receptors that are present determines how the signal is transduced from the extracellularly released ATP and/or its metabolites. Purinergic receptors are divided into two classes: G protein-coupled P1 adenosine receptors and nucleotide P2 receptors, which bind ATP, ADP, UTP, and UDP. Yoshioka et al. (47) have recently reported that the A1 and P2Y1 receptors can oligomerize, thus resulting in A1 receptor ligand-induced P2Y1 receptor signaling. In addition to the versatility of receptors capable of transmitting purinergic signals, other factors influencing their signaling are the metastable nature of ATP and the role of its metabolites. Extracellularly, ATP is degraded rapidly by different ecto-enzymes, yielding ADP, 5'-AMP, and adenosine. We observed no effect of stable adenosine analogs on tRaH-induced I– uptake (data not shown), ruling out the possibility that IBMX acts through adenosine receptors or that the effect of ATP is mediated by its metabolite adenosine.
The complex interactions described above make it difficult to identify with certainty the purinergic receptor that mediates the stimulation of NIS expression by ATP. Still, we addressed this point by administering various ATP analogs and other nucleotides with different receptor specificities and determining their effectiveness in stimulating I– transport in MCF-7 cells (Fig. 5). ATP and UTP are known to elevate intracellular calcium and stimulate proliferation via activation of the P2Y2 receptor in MCF-7 cells. These cells express P2Y2 receptors and proliferate in response to both ATP and UTP (48). We found that the nonhydrolyzable ATP analog ATPS was as effective as ATP itself in stimulating I– transport, whereas UTP was noticeably less so and ADPßS had no effect (Fig. 5). These findings suggest that the receptor involved is probably P2Y2 (32).
We have previously reported that functional NIS expression in the mammary gland appears during the second half of gestation and is maintained thereafter by suckling during lactation (4). We have also shown that, whereas oxytocin alone up-regulates NIS expression in the mammary glands of nubile mice, the combination of estrogen, oxytocin, and prolactin induces the highest NIS protein expression (4). The lack of any effect of prolactin and oxytocin on NIS expression in MCF-7 cells (data not shown) could be explained either by the malignant status of these cells or by the absence of the required microenvironment, particularly myoepithelial cells. Furthermore, it has been observed that tRa down-regulates prolactin receptor expression in MCF-7 cells (49). In these cells, ATP and UTP promote cell proliferation via stimulation of the P2Y2 receptors (48, 50). In the lactating mammary gland, P2Y2-expressing secretory epithelial and P2Y1-expressing myoepithelial cells are believed to interact via released ATP to enhance milk ejection (51, 52, 53). Mechanical stretch (e.g. alveoli filling with milk) induces ATP release from epithelial cells (51, 52, 53). Oxytocin, which by itself is able to up-regulate NIS expression in nubile mice, most likely exerts its effect through the myoepithelial cells. Myoepithelial cell contraction in response to oxytocin leads to ATP release, which probably increases NIS expression and I– transport into the epithelial cells. Our study of the purinergic regulation of NIS expression, reported here for the first time, shows that ATP up-regulates NIS expression in MCF-7 cells (Fig. 4) but decreases it in FRTL-5 cells (Fig. 6, A–C). The main requirement for NIS expression in FRTL-5 cells is TSH, whereas in MCF-7 cells it is tRa. TSH has no effect in MCF-7 cells. In FRTL-5 cells, cAMP is the main positive regulator of NIS expression, but in MCF-7 cells, cAMP had no effect on NIS expression (Fig. 3, A–C). Clearly, the regulation of NIS expression is tissue specific, and some regulatory factors have opposite effects on NIS expression in different tissues. tRa significantly decreased the amount of total and plasma membrane-localized NIS protein in FRTL-5 cells (Fig. 6, A–C); in contrast, tRa had the opposite effect in MCF-7 cells (Fig. 1). Hydrocortisone, which is added to the FRTL-5 media to improve surface attachment (34), decreases TSH-induced I– uptake in FRTL-5 cells, whereas it significantly enhanced tRa-induced NIS expression in MCF-7 cells (Fig. 1). It has long been known that I– at high concentrations down-regulates its own transport in thyroid cells. Here we have shown, for the first time, that I–, even at high concentrations, had no effect on its own transport in MCF-7 cells (Fig. 6).
In conclusion, having previously reported NIS protein expression in more than 80% of human breast cancers (4) and functional NIS expression in human breast cancer metastases in vivo (6) without expression in healthy nonlactating mammary tissue (4), we have now demonstrated that: 1) hydrocortisone and ATP stimulate tRa-induced functional NIS expression in human breast carcinoma MCF-7 cells; 2) IBMX increases tRa-induced NIS expression in these cells by a purinergic signaling pathway; and 3) regulatory factors of NIS expression exert different and sometimes even opposite effects on breast cancer as compared with thyroid cells. The differences in regulatory mechanisms of NIS expression between these tissues provide a key advantage for the development of selective treatment strategies that would destroy breast cancer cells without harming the thyroid. As MCF-7 and FRTL-5 cells have been extensively used for in vitro investigations of in vivo regulatory processes of breast cancer and healthy thyroid, respectively, data obtained in these cell lines are widely regarded as being potentially physiologically relevant. Thus, our results may be a significant step toward the goal of effectively applying radioiodide in the diagnosis and therapy of breast cancer. Nevertheless, these findings on selective up-regulation of NIS expression in breast cancer cells by tRa, hydrocortisone, and IBMX will require subsequent in vivo validation in future animal and, eventually, human studies.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-S, ADP-ß-S, UTP, adenosine, N6-cyclopentyladenosine, N6-(3-iodobenzyl)-9[5-(methylcarbamoyl-ß-D-ribofuranosyl]adenine, N-ethylcarboxamidoadenosine, estradiol, human recombinant prolactin, and oxytocin were from Sigma-Aldrich (St. Louis, MO). Forskolin was from Calbiochem (San Diego, CA).
Cell Culture
MCF-7 cells (lot 2188915) were purchased from American Type Culture Collection (Manassas, VA), and they were grown in Eagle’s MEM with Earle’s balanced salt solution and 2 mM L-glutamine, containing 1 mM sodium pyruvate, 0.1 mM nonessential amino acids, 1.5 g/liter sodium bicarbonate, and supplemented with 0.01 mg/ml bovine insulin and 10% fetal bovine serum. Before (48 h) and during the hormonal treatments, the medium was switched to medium containing only 2.5% charcoal dextran-stripped serum (Gemini Bio-Products, Woodland, CA). FRTL-5 cells were maintained as described previously (15).
Iodide Uptake
After the culture medium was aspirated, cells were washed twice with 1 ml of a modified Hanks’ balanced salt solution (buffered HBSS) with the following composition: 137 mM NaCl, 5.4 mM KCl, 1.3 mM CaCl2, 0.4 mM MgSO4 · 7 H2O, 0.5 mM MgCl2, 0.4 mM Na2HPO4 · 7 H2O, 0.44 mM KH2PO4, and 5.55 mM glucose with 10 mM HEPES buffer (pH 7.3). Cells grown in 24-well plates were incubated with buffered HBSS containing 20 µM NaI supplemented with 10 µCi/µl carrier-free Na125I to give a specific activity of 100 mCi/mmol. For steady-state experiments, incubations proceeded at 37 C for 45 min in a humidified atmosphere and were terminated by aspirating the radioactive medium and washing twice with 1 ml ice-cold HBSS. For kinetic analysis, cells were incubated for 2 min with 0.625, 1.25, 2.5, 5, 10, 20, 40, or 80 µM NaI, and uptake reactions were terminated as indicated above. Data were processed using the equation: v = Vmax · [I]/(Km+[I]). To assess Na+ dependence of I– uptake, cells were incubated for 2 min with 0, 5, 10, 20, 30, 40, 80, or 140 mM NaCl; isotonicity was maintained constant with choline Cl. Na+ dependence data were analyzed using the equation: v = Vmax · [Na+] 2/Km+[Na+]2. Data were fitted by nonlinear least squares using Gnuplot software (www.gnuplot.info). Nonspecific background, as measured in nontreated cells, was subtracted.
To determine the amount of 125I accumulated in the cells, 500 µl 95% ethanol was added to each well for 20 min at 4 C and then quantitated in a -counter. The DNA content of each well was determined on the material not extracted by ethanol after trichloroacetic acid precipitation, by the diphenylamine method, as previously described (54). I– uptake was expressed as picomoles per µg DNA. All parameters were determined at least in triplicate.
Immunoblot Analysis of Total and Biotinylated Plasma Membrane Proteins
Cells were grown in six-well plates to 80% confluence. Cells were rinsed at 4 C twice with PBS/Ca2+/Mg2+ (138 mM NaCl; 2.7 mM KCl; 1.5 mM KH2PO4; 9.6 mM Na2HPO4; 1 mM MgCl2; 0.1 mM CaCl2, pH 7.4). Cells were next incubated with 500 µl/well of a solution containing 1.5 mg/ml sulfo-NH-SS-biotin (Pierce Chemical Co., Rockford, IL) in biotinylation buffer [HEPES 20 mM (pH 8.5), CaCl2 2 mM, NaCl 150 mM]; for 20 min at 4 C with gentle shaking. The biotinylation solution was removed by two washes in PBS Ca2+/Mg2+ containing 100 mM glycine. Cells were lysed with 1 ml of lysis buffer (50 mM Tris-HCl, pH 7.5; 150 mM NaCl; 5 mM EDTA; 1% Triton X-100; 1% sodium dodecyl sulfate; protease inhibitors) at 4 C for 15 min. A 200-µl aliquot was taken for assessing total NIS expression by immunoblot. The remaining cell lysate (800 µl) was incubated overnight at 4 C with 50 µl of streptoavidin agarose beads (Pierce). Beads were centrifuged at 5000 x g for 2 min and rinsed three times with lysis buffer, twice with high-salt, and once with low-salt solution. Adsorbed proteins were eluted from the beads with sample buffer containing 10 mM dithiothreitol, at 75 C for 5 min, and analyzed by immunoblot. Ab purification was performed as described previously (17).
PAGE and electroblotting to nitrocellulose were performed as previously described (17). Immunoblot analysis was also carried out as described (17), with affinity-purified 4 nM antihuman NIS or 4 nM antirat NIS Ab, and a 1:2000 dilution of a horseradish peroxidase-linked donkey antirabbit IgG (Chemicon, Temecula, CA). Both incubations were performed for 1 h. Proteins were visualized by the enhanced chemiluminescence Western blot detection system (Amersham Pharmacia Biotech, Arlington Heights, IL). Equal loading was assessed by reprobing the same blot with a mAb against the 
-subunit of the Na+/K+ ATPase (Affinity BioReagents, Inc., Golden, CO). Protein expression was quantitated using ImageQuant software (Molecular Dynamics, Inc., Sunnyvale, CA) and standardized for the loading control.
Immunofluorescence Analysis
MCF-7 cells were seeded onto polylysine-coated cover slips. They were fixed with 2% paraformaldehyde for 30 min at room temperature and then rinsed with PBS containing 0.1 mM CaCl2 and 1 mM MgCl2, subsequently referred to as PBS-C-M. Cells were permeabilized with PBS-C-M-0.2% BSA-0.1% Triton for 10 min and then quenched with 50 mM NH4Cl in PBS-C-M for 10 min. Cells were rinsed with PBS-C-M-BSA-Triton and incubated for 1 h at room temperature with 4 nM affinity-purified anti-hNIS Ab in PBS-C-M-BSA-Triton. Subsequently, cells were washed three times for 10 min with PBS-C-M-BSA-Triton, incubated for 1 h in the dark with antirabbit fluorescein-conjugated secondary Ab (Vector Laboratories, Inc., Burlingame, CA) (1:1000 dilution), and washed as above. Cover slips were mounted onto slides with Slow Fade antifade reagent (Molecular Probes, Inc., Eugene, OR), sealed with quick-dry nail polish, and allowed to dry in the dark for 2 h at room temperature and then stored at 4 C in the dark. Cells were visualized in a Bio-Rad Radiance 2000 Laser Scanning Confocal Microscope (Bio-Rad Laboratories, Inc., Hercules, CA).
Flow Cytometry
MCF-7 cells were detached with trypsin, counted, fixed in PBS containing 2% paraformaldehyde, and then permeabilized with 0.2% saponin in PBS supplemented with 0.1% BSA and transferred into 5-ml polystyrene tubes (200,000 cells per tube). Cells were incubated for 1 h at room temperature with 100 µl PBS-0.1% BSA-0.2% Sap containing 4 nM anti-hNIS Ab. Cells were washed with 1 ml PBS-BSA-saponin and centrifuged as above. Next, they were incubated for 30 min at room temperature in the dark with fluorescein-conjugated antirabbit IgG (Vector Laboratories) at a 1:1000 dilution in the same buffer. Cells were washed once again, centrifuged, and resuspended in 300 µl PBS. The fluorescence of 10,000 cells per tube was assayed by a FACScan flow cytometer (Becton Dickinson and Co., Franklin Lakes, NJ).
Statistical Analysis
Statistical analyses were performed using Student’s t test. Each experimental condition was compared with its respective control group; P values below 0.05, below 0.01, and below 0.001 are represented as *, **, and ***, respectively, in all figures.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Disclosure summary: all authors have nothing to declare.
First Published Online January 26, 2006
Abbreviations: Abs, Antibodies; 8-Br-cAMP, 8-bromo-cAMP; FACS, fluorescence-activated cell sorting; HBSS, Hank’s balanced salt solution; IBMX, isobuthylmethylxantine; mAbs, monoclonal antibodies; NIS, sodium/iodide symporter; NUE, NIS upstream enhancer; PBS-C-M, PBS containing 0.1 mM CaCl2 and 1 mM MgCl2; tRA, all-trans-retinoic acid; tRAH, hydrocortisone combined with tRA.
Received for publication September 13, 2005. Accepted for publication January 17, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
, tRa-treated MCF-7 cells; 
, tRaH-treated MCF-7 cells; 
, FRTL-5 cells. C and D, Hydrocortisone increases the Vmax of NIS-mediated I– transport in tRa-treated MCF-7 cells without changing the Km of the transporter for either I– (panel C) or Na+ (panel D). Initial rates (2-min time points) of I– uptake were determined at the indicated concentrations of I– (as described in Materials and Methods). Calculated curves were generated with the equation v = Vmax·[I]/(Km+[I]), using Gnuplot software. Cells were incubated for 2 min with the indicated concentrations of Na+; isotonicity was maintained constant with choline chloride. Na+ dependence data were analyzed with the equation v = Vmax· [Na+]2/Km+[Na+]2). Data were fitted by nonlinear least squares using Gnuplot software. E, Total NIS protein expression in nontreated, tRa-, or tRaH-treated (48 h) MCF-7 cells. Experiments were performed at least in triplicate; representative data are shown. Upper panel, MCF-7 cells were treated for 48 h with 1 µM tRa or with the combination of 1 µM tRa and 10 µM hydrocortisone (panel H). Cells were lysed and subjected to immunoblot analysis with 4 nM anti-hNIS Ab. Total protein (20 µg) was loaded onto each lane. Middle panel, Equal loading was assessed by reprobing the same blot with an Ab against the 
