The Neuropeptide Y Y1 Receptor: A Diagnostic Marker? Expression in MCF-7 Breast Cancer Cells Is Down-Regulated by Antiestrogens In Vitro and in Xenografts

The neuropeptide Y (NPY) Y1 receptor (Y1R) has been suggested as a tumor marker for in vivo imaging and as a therapeutic target. In view of the assumed link between estrogen receptor (ER) and Y1R in mammary carcinoma and with respect to the development of new diagnostic tools, we investigated the Y1R protein expression in human MCF-7 cell variants differing in ER content and sensitivity against antiestrogens. ER and Y1R expression were quantified by radioligand binding using [3H]-17β-estradiol and the Y1R selective antagonist [3H]-UR-MK114, respectively. The latter was used for cellular binding studies and for autoradiography of MCF-7 xenografts. The fluorescent ligands Cy5-pNPY (universal Y1R, Y2R and Y5R agonist) and UR-MK22 (selective Y1R antagonist), as well as the selective antagonists BIBP3226 (Y1R), BIIE0246 (Y2R) and CGP71683 (Y5R) were used to identify the NPY receptor subtype(s) by confocal microscopy. Y1R functionality was determined by mobilization of intracellular Ca2+. Sensitivity of MCF-7 cells against antiestrogen 4-hydroxytamoxifen correlated directly with the ER content. The exclusive expression of Y1Rs was confirmed by confocal microscopy. The Y1R protein was up-regulated (100%) by 17β-estradiol (EC50 20 pM) and the predominant role of ERα was demonstrated by using the ERα-selective agonist “propylpyrazole triol”. 17β-Estradiol-induced over-expression of functional Y1R protein was reverted by the antiestrogen fulvestrant (IC50 5 nM) in vitro. Furthermore, tamoxifen treatment of nude mice resulted in an almost total loss of Y1Rs in MCF-7 xenografts. In conclusion, the value of the Y1R as a target for therapy and imaging in breast cancer patients may be compromised due to Y1R down-regulation induced by hormonal (antiestrogen) treatment.


Introduction
Neuropeptide Y (NPY), a 36 amino acid peptide, is one of the most abundant peptides in the central and peripheral nervous system of mammals, involved in numerous (patho)physiological functions such as food intake, blood pressure, regulation of hormone secretion, anxiety and memory [1]. In humans NPY exerts its biological effects by interaction with at least four distinct G protein coupled receptors designated Y 1 (Y 1 R), Y 2 (Y 2 R), Y 4 (Y 4 R), and Y 5 (Y 5 R) [2]. The Y 1 R subtype was the first NPY binding receptor to be cloned [3]. Its constitutive expression and functionality in human erythroleukemia (HEL) cells [4] and in SK-N-MC neuroblastoma cells [5] is well established. Y 1 and Y 2 receptors were recently reported to be expressed in several human cancers and were therefore proposed as potential targets for diagnosis and treatment [6][7][8][9][10][11][12][13][14]. Mammary carcinomas revealed an 85% incidence of Y 1 R expression, whereas Y 2 R was shown to be the less expressed NPY receptor subtype [15]. An estrogen induced expression of Y 1 R mRNA in MCF-7 breast cancer cells was shown in a differential screening study [16]. Later, investigations confirmed the up-regulation of Y 1 R mRNA after estrogen treatment, and suggested a functional role of the Y 1 R in cell signaling and proliferation [17]. Very recently, a DOTA (1,4,7,10tetraazacyclododecane-1,4,7-10-tetraacetic acid) substituted Y 1 R selective peptide for radiolabeling with metallo positron emitters for PET imaging of breast cancer was described [18] and the use of a Y 1 R selective 99m Tc-labeled peptide in whole body scintimammography was reported [11].
In consideration of the assumed link between ER and Y 1 R in breast cancer and the potential value of new diagnostic tools we combined tumorpharmacological investigations with our work on receptor subtype-selective ligands for the detection of NPY receptors. Y 1 R selective fluorescence and radiolabeled compounds, recently developed in our laboratory, as well as a set of reference substances were used as pharmacological tools. To evaluate the working hypothesis that the Y 1 R is a potential diagnostic target in breast cancer, we performed preclinical investigations on ER and NPY receptor expression and function, taking into account the impact of standard therapies using antiestrogens or aromatase inhibitors.
The recently developed highly potent and selective tritiated Y 1 R antagonist [ 3 H]-UR-MK114 ( Fig. 1) [19], an (R)-argininamide derived from BIBP3226 [20], was applied to quantify Y 1 R protein expression in radioligand binding assays using adherent live cells. In the present study different subclones of MCF-7 breast cancer cells with different estrogen receptor (ER) content were analyzed with respect to a correlation between ER and Y 1 R expression. Furthermore, the influence of ER agonists and antagonists on the expression of the functional Y 1 R protein was determined in a fura-2 assay. In addition to in vitro studies, the Y 1 R expression was investigated by autoradiography of MCF-7 xenografts from nude mice supplemented with 17b-estradiol on one hand, and treated with tamoxifen on the other hand.

Ethics Statement
Animal studies. The use of animals in this study complies with the Guide for the Care and Use of Laboratory Animals (NIH publication no. 86-23, revised 1985) and the current German law on the protection of animals. The animal experiment was approved by the Regierung

Cell Culture
MCF-7 cells were grown in EMEM containing 5% FCS. HCC1806, HCC1937, and T-47-D cells were cultured in RPMI medium supplemented with 10% FCS. In the case of T-47-D cells, 10 mg/L of insulin (Sigma, Munich, Germany) were supplement- ed. To study (anti)estrogenic effects on Y 1 R expression, the medium was replaced with EMEM (or phenol red-free DMEM) supplemented with FCS twice treated with dextran-coated charcoal (ct-FCS) [24]. MDA-MB-231 cells were cultured in Mc Coy's 5A medium containing 5% FCS.

Proliferation Assay
The sensitivities of MCF-7 and MDA-MB-231 breast cancer cells against the antiestrogen 4-hydroxytamoxifen and the effect of pNPY on the growth of MCF-7 cells were determined in a previously described chemosensitivity assay [25]. Cells were grown in 96 well plates in the presence of increasing concentrations of 4hydroxytamoxifen or pNPY, respectively. Compounds were added as 1000-fold concentrates to the respective culture media. 16 wells were processed for each compound concentration and respective vehicle control. As readout of the cell mass per well the absorbance at 578 nm was determined at different time points after staining the cells with crystal violet.

Cytosol Preparation
Three different MCF-7 variants (H: high ER content (wild type); M: medium ER content; L: low ER content) and MDA-MB-231 cells (ER negative) were grown in 175-cm 2 culture flasks.
When cells were confluent, the medium was removed and the cells from 8-10 flasks were harvested after trypsination. The pooled cell suspensions were centrifuged at 1200 rpm for 7 min. The pellet was washed twice with PBS and suspended in 4-5 mL of TED-Mo-buffer (10 mM Tris-HCl, pH 7.4, 10 mM Na 2 MoO 4 (Sigma), 1 mM EDTA, 1 tablet of EDTA-free protease inhibitor cocktail (Roche, Basel, Switzerland) per 100 mL). Cells were lysed using an ultrasonic cell disrupter B15 (Branson, Danbury, CT, 3610 cycles, 10-20 s) under ice cooling. The suspension was centrifuged for 20 min at 5000 rpm. The supernatant cell extract was decanted carefully and stored at 270uC.
The protein content of the cytosols was determined after appropriate dilution by Bradford's protein assay [26] using Bradford dye reagent (BioRad Laboratories, Munich, Germany) following the manufacturer's protocol. Absorbance was measured in a Uvikon 930 spectrophotometer (Kontron, Neufahrn, Germany) at 595 nm. A calibration curve with human serum albumin (HSA, Behringwerke, Marburg, Germany) standards was recorded to assign absorbance values to protein concentrations.

[ 3 H]-17b-Estradiol Binding Assay
The [ 3 H]-17b-estradiol ([ 3 H]-E2) saturation binding assay was performed in 1.5 mL-reaction vessels (Eppendorf, Hamburg, Germany) under ice cooling. Mixtures of the radioligand (added as a 5-fold concentrated solution in Tris buffer (100 mL); final concentration range: 0.1-5.0 nM) and the respective cytosol (100 mL) were diluted to a final volume of 500 mL in buffer (10 mM Tris-HCl, pH 7.4). 17b-estradiol (final concentration: 1 mM) was added to determine nonspecific binding. Total and nonspecific binding were determined in triplicate. The samples were incubated for 16-20 h at 4uC under shaking. Non-bound radioactivity was removed by the dextran-coated charcoal (DCC) method. For this purpose 0.5 mL of a an ice-cold suspension containing 0.8% charcoal (Norit A; Serva, Heidelberg, Germany) and 0.008% dextran 60 (Serva) were added to each sample, followed by incubation at 4uC for 30 min under shaking. After centrifugation (10 min at 4000 rpm), 200 mL of the supernatant were transferred into minivials containing 3 mL of liquid scintillator (Rothiszint TM eco plus; Roth, Karlsruhe, Germany). The bound radioactivity was counted in a LS6500 liquid scintillation beta counter (Beckmann Instruments, Munich, Germany).

Whole Cell Y 1 R Radioligand Binding Assay
The maximum number of Y 1 Rs (B max ) was determined in saturation binding experiments using the radioligand [ 3 H]-UR-MK114 as previously described [19]. The average cell number per well was determined from identically processed control wells (n = 6) after counting the cells in a Neubauer improved hemocytometer.
For the determination of (anti)estrogenic effects on Y 1 R protein expression, MCF-7 cells were seeded in 48-well plates and grown in ct-FCS-containing medium until they had reached 70-80% confluence. 45-50 h prior to the Y 1 R binding assay, the medium was removed by suction and replaced with fresh medium (0.3 mL/well) containing the estrogens at the respective concentrations (by dilution of a 1000-fold concentrate in ethanol). For the analysis of the antagonistic effect of fulvestrant, the antiestrogen was added at multiple concentrations in the presence of 1 nM 17b-estradiol as stimulating agent. At least 6 wells per plate were processed for each (anti)estrogen concentration. All plates were prepared in duplicate as two identical sets. One set of 48 well plates was used for the Y 1 R radioligand binding assay to quantify Y 1 R expression: If not otherwise From each group of replicate wells (n = 6-8), one half was used for the determination of the total binding (radioligand alone) and the other half for the determination of unspecific binding (radioligand plus 300-fold excess of pNPY). In order to exclude dissociation of the radioligand [ 3 H]-UR-MK114 during the washing steps after incubation, additional experiments were performed with respect to the time period and the number of washing cycles (cf. Fig. S1).
The second set of plates was used as control to normalize the specifically bound radioactivity to the protein content. For this purpose, the cells of the control wells were lysed with a buffer (50-100 mL, volume dependent on the protein concentration), consisting of 25 mM Tricine (pH 7.8), 10% glycerol, 1% Triton TM X-100 (Serva) and 1 mM dithiothreitol (Sigma), for 30 min under shaking. 5 mL of each lysate were analyzed by the Bradford protein assay after appropriate dilution.

Confocal Microscopy
Images were acquired with a Zeiss Axiovert 200 M microscope equipped with the LSM 510 laser scanner. Two days before the experiment MCF-7 (L) cells were trypsinized and seeded in ibiTreat m-slide 8-well cover glasses (Ibidi, Planegg, Germany) in EMEM containing 1 nM 17-b-estradiol and 5% FCS. At a confluence of the cells of about 80% the culture medium was removed, the cells were washed with Leibowitz L15 culture medium (200 mL) and covered with L15 medium (100 mL) and Cy5-pNPY (100 mL of a two-fold concentrated solution in L15 medium) for total binding as well as with the competing agent (100 mL of a two-fold concentrated solution in L15 medium) and Cy5-pNPY (100 mL of a two-fold concentrated solution in L15 medium) for displacement. Images were acquired after an incubation period of 7-9 min (excitation at 633 nm (10% laser transmission), 650 nm long-pass filter).
Visualization of Y 1 Rs using the fluorescent Y 1 R-selective ligand UR-MK22 was performed as reported [27] with the following variations: on the day of the experiment confluence of the cells was about 70-80%. Images were acquired after an incubation period of 16 min (excitation at 488 nm (5.1% laser transmission), 560 nm long-pass filter).

Calcium Assay
The intracellular Ca 2+ concentration in MCF-7 (L) cells was measured by a spectrofluorimetric assay with the fluorescent Ca 2+ indicator fura-2. The assay was performed by analogy with a protocol established for HEL cells in our laboratory [28]. Prior to the assay, MCF-7 cells were incubated with 1 nM 17b-estradiol alone or in combination with 100 nM fulvestrant, or the respective vehicle, for 45 h. Calcium mobilization in MCF-7 cells was stimulated by 10 nM pNPY. To antagonize the Y 1 R mediated calcium mobilization, BIBP3226 (100 nM) was added 1 min prior to the addition of pNPY. The ratio R of fluorescence intensities at 510 nm after excitation at 340 and 380 nm was used for the calculation of the calcium concentration according to the Grynkiewicz equation [29]: ? SFB (K D : dissociation constant of the fura-2-Ca 2+ complex = 224 nM; R max : fluorescence ratio in presence of saturating Ca 2+ concentration (determined after the addition of 10 mL of digitonin solution (2% in water; Sigma), which caused lysis of the cells); R min : ratio in absence of free Ca 2+ , caused by addition of 50 mL of EGTA solution (600 mM in 1 M Tris buffer, pH 8.7) to lysed cells; SFB: correction factor; ratio of the fluorescence intensity (l ex = 380 nm, l em = 510 nm) of the Ca 2+ free and Ca 2+ saturated dye.

Autoradiography
About 4 million MCF-7 (L) cells (173 rd in vitro passage, suspended in 0.1 mL of PBS) were subcutaneously injected into 12 female NMRI (nu/nu) mice bearing subcutaneous 17ßestradiol depots [30] (implanted 14 days before). After 4 weeks of tumor growth, 6 animals, bearing tumors of comparable size (mean tumor area about 766 mm), were selected for control (3 mice) and tamoxifen treatment (3 mice). In case of the tamoxifen group, estrogen depots were explanted prior to tamoxifen administration. Tamoxifen citrate (12 mg/kg, dissolved in PEG400/1.8% NaCl 1:1 at a concentration of 2.4 mg/mL) was injected subcutaneously on day 2, 6 and 10. The control group was treated with the vehicle. 14 days after removal of the estrogen depots, tumors were excised, immediately frozen in Tissue-Tek and stored at 278uC. Cryosections (12 mm) were obtained at 216uC with a 2800 Frigocut E freezing microtome (Reichert-Jung/Leica, Germany). Adjacent sections were mounted on three microscopic slides (Superfrost Plus, 7562561 mm) and kept in a chamber of 100% humidity for 1-2 min. Two slides were used to determine total and non-specific binding, and the third slide immersed in an alcoholic formaldehyde fixative (37% (w/w) formaldehyde (40 mL), 95% (v/v) ethanol (360 mL) and calcium acetate (0.2 g)) for 20 s. For total binding the sections were covered with binding buffer (ca. 800 to 1000 mL) containing [ 3 H]-UR-MK114 (3 nM), and for unspecific binding with binding buffer, containing the radioligand (3 nM), pNPY (300 nM) and BIBP3226 (30 nM). The sections were incubated at room temperature (22-25uC) for a period of 8 min. After incubation, the binding buffer was removed, the slides immersed three times into ice-cold buffer split to 3 vessels (each 10 s) and finally immersed into ice-cold demineralised water (3 s). The slides were put uprightly on a paper towel for 1 min and then dried in horizontal position in a desiccator over P 4 O 10 . The slides were set in close contact with a tritium sensitive screen (PerkinElmer, 1926125 mm) using an X-ray film cassette and stored in a dark room for 15 d. The autoradiographic image was generated from the tritium screen using an imager (Cyclone Storage Phosphor System, Packard).
The fixed sections were stained according to Masson-Goldner    9.0 (Systat Software inc., Chicago, IL) using 4 parameter sigmoid and one site saturation binding fits, respectively. To calculate the number of receptors per cell, the B max value was divided by the mean cell number of six identically treated control wells. For the determination of (anti)estrogenic effects on Y 1 R expression, all mean values of specific binding (dpm/well) were normalized to the mean protein content (mg/well) and are given as percentage of the 17b-estradiol (1 nM) treated controls. Errors of calculated values determined by multiple parameters were estimated according to the Gaussian law of errors. Statistical significance was tested by Student's t-test. P,0.05 was accepted as statistically significant.  (Fig. 3B), HCC1806 and HCC1937 (data not shown) breast cancer cells. Fig. 3C shows the relative basal expression of Y 1 R and ER in the three investigated MCF-7 variants. Under identical culture conditions Y 1 R expression in MCF-7 (M) and MCF-7 (L) cells (91,00064,000 and 98,00069,000 sites/cell, respectively) was by more than a factor of two higher compared to the wild type (H) of the MCF-7 breast cancer cells (38,000610,000 sites/cell). From the phenotypical point of view, basal Y 1 R expression is inversely associated with basal ER expression. However, this does not reflect a functional correlation due to lacking agonist stimulation of both receptors.  The expression profile of NPY receptor subtypes in MCF-7 (L) cells was investigated by confocal laser scanning microscopy using fluorescent Cy5-pNPY [23], a universal ligand with comparable affinity (K i # 6 nM) at the Y 1 R, Y 2 R and Y 5 R (Fig. 4A-D). Cy5-pNPY (10 nM) was totally displaced by the Y 1 R selective antagonist BIBP3226 (1 mM), but neither by the Y 2 R selective antagonist BIIE0246 [21] (1 mM) nor by the Y 5 R selective antagonist CPG71683 [22] (1 mM). The displacement of Cy5-pNPY from Y 2 R and Y 5 R by BIBP3226 (1 mM) can be excluded due to high Y 1 R selectivity (K i values for Y 2 R and Y 5 R .40 mM [31][32][33]). Moreover, the sole expression of the Y 1 R was confirmed by the binding of the selective fluorescent Y 1 R antagonist UR-MK22 (Fig. 4E-F).

Effect of NPY on MCF-7 Cell Proliferation and ER Function
As the effect of NPY on tumor cell growth is controversially discussed in the literature [8], the influence of NPY on the growth of MCF-7 cells with particularly high Y 1 receptor status (tamoxifen low sensitive subclone (L)) was investigated in the kinetic chemosensitivity assay. As shown in Fig. 5, pNPY had no effect on the growth of this MCF-7 subclone when applied at concentrations up to 10 nM in the presence of 1 nM estradiol. A similar result was obtained in the absence of estradiol (data not shown). In a luciferase assay under the control of the ER responsive element [34] there was no unambiguous effect of NPY on the estrogenic activity of 17b-estradiol (cf. Fig. S3).

NPY Stimulated Mobilization of Intracellular Ca 2+ in MCF-7 Cells
To confirm the functionality of the Y 1 R expressed in MCF-7 (L) breast cancer cells in the absence and presence of ER stimulation, the coupling of the receptor to the calcium signaling cascade was investigated by a fura-2 assay. pNPY at a concentration of 10 nM induced an increase in the intracellular calcium level by a factor of four (Fig. 6). In the presence of the Y 1 R antagonist BIBP3226 (100 nM) the signal was depressed by < 80%, showing the Y 1 R specificity of the signaling. The calcium response was not affected, when cells were pretreated with 17b-estradiol (45 hours), but significantly decreased after pre-incubation of the cells with fulvestrant for 45 hours (Fig. 6).

Estrogen-dependent Expression of the Y 1 R Protein
To investigate, if estrogen receptor mediated up-regulation of Y 1 R mRNA in MCF-7 breast cancer cells reported by Amlal et al. [17] is paralleled at the protein level, the selective Y 1 R radioligand [ 3 H]-UR-MK114 was used for binding studies. Fig. 7A shows representative saturation binding curves for the specific binding of the radioligand to MCF-7 (L) cells pretreated with 17b-estradiol (1 nM) for 48 h or its vehicle. An increase in Y 1 R protein expression by approximately 250% was observed for the estrogen pre-incubated cells (B max = 1.8 and 0.51 fmol/ mg, resp.) The ratio of Y 1 Rs in estrogen treated vs. untreated cells was not significantly increased when the time of incubation was prolonged to 72 hours (data not shown). Consequently, 45 to 50 hours were considered as an appropriate incubation period for the treatment of MCF-7 cells with (anti)estrogens in all following experiments. In T-47-D breast cancer cells an upregulation of the Y 1 R after estrogen treatment occurred as well, but the expression was about 20-fold lower compared to MCF-7 (L) cells (data not shown).
To facilitate the analysis of Y 1 R regulation, the specifically bound radioactivity at a radioligand concentration of 12 nM was compared, whereupon the expression levels are presented as percentage of the control (cells treated with 1 nM 17b-estradiol). At this radioligand concentration, the saturation curves reveal an approximation of the specifically bound radioactivity to the B max value (cf. Fig. 3A).
The pH indicator phenol red was reported to bring along contaminants with weak estrogenic activity [35] and might therefore contribute to basal Y 1 R expression. However, the basal Y 1 R expression was not significantly different, when cells were maintained in phenol red-free DMEM and phenol red containing EMEM, respectively (Fig. S4). Fig. 8 shows concentration-response curves for the Y 1 R upregulation by a selection of ER agonists. 17b-estradiol was applied at picomolar to nanomolar concentrations, showing a sigmoidal concentration-response relationship with an EC 50 value of approximately 0.02 nM. Maximum Y 1 R up-regulation was achieved at a 17b-estradiol concentration of 0.5 nM (there was no further increase at concentrations of 10 and 50 nM; data not shown). PPT, an agonist with 400-fold selectivity for ERa over estrogen receptor b (ERb) [36], was applied to demonstrate the ERa subtype dependence of Y 1 R up-regulation. The compound showed an EC 50 value of 0.25 nM and 100% intrinsic activity compared to 17b-estradiol (Fig. 8). The non-selective, but ERb-preferring phytoestrogen genistein upregulated the Y 1 R protein to 70% compared to the maximum Figure 10. Y 1 R expression in MCF-7 xenografts is downregulated by antiestrogens in vivo. Effect of estradiol and tamoxifen on Y 1 R expression by MCF-7 (L) xenografts in vivo determined by autoradiography using the selective Y 1 R antagonist [ 3 H]-UR-MK114 (3 nM). Subcutaneously grown tumors from NMRI (nu/ nu) mice bearing subcutaneous 17b-estradiol depots. The control group (3 mice, C1-C3) was treated with the vehicle (PEG400/1.8% NaCl, 1:1). Tamoxifen group (3 mice, T1-T3): A cumulative dose of tamoxifen citrate (36 mg/kg, dissolved in PEG400/1.8% NaCl, 1:1, at a concentration of 2.4 mg/mL) was administered by injecting three times (on day 2, 6 and 10 after explantation of the estrogen depots) 12 mg/kg subcutaneously. doi:10.1371/journal.pone.0051032.g010 effect of 17b-estradiol. The EC 50 value was approximately 100 nM (Fig. 8).

Y 1 R up-and Down-regulation by ER Agonists and Antagonists
As depicted in Fig. 9A, the pure ER antagonist fulvestrant significantly down-regulated the Y 1 R expression below the basal expression level when co-incubated with 17b-estradiol. Fulvestrant inhibited the estradiol (1 nM) induced Y 1 R expression in a concentration-dependent manner with an IC 50 value of approximately 5 nM (Fig. 9B). To exclude adulterations of the determined Y 1 R expression due to anti-proliferative effects of antiestrogens or growth-stimulating effects of estrogenic agents, all specific binding values were normalized to the total protein content derived from an independently conducted protein assay (Bradford).
Complementary to these in vitro experiments the Y 1 R expression was studied by autoradiography in nude mice bearing MCF-7 (L) xenografts. As obvious from Fig. 10 the subcutaneously grown human breast cancer (control, C1-C3 in Fig. 10) demonstrated high specific binding of the Y 1 R selective antagonist [ 3 H]-UR-MK114. By contrast, the Y 1 R radioligand binding was extremely reduced in tumors (T1-T3) of tamoxifen treated mice. This is in agreement with Y 1 R down-regulation, because the histological grading corresponds to well differentiated adenocarcinomas of comparable size irrespective of tamoxifen treatment (histology cf. Fig. S5).

Discussion
NPY Y 1 and Y 2 receptors are reported to be expressed by various malignant tumors [8,15,[37][38][39]. The majority (85%) of human primary mammary carcinomas express the Y 1 R, whereas the Y 2 R is predominant in normal breast tissue [15]. More than 70% of breast cancers are classified as ER-positive [40] and estrogen-induced up-regulation of Y 1 R mRNA was reported previously [16,17]. Although the role of NPY receptors in tumor biology is a matter of debate [8], the Y 1 R has been considered as a diagnostic and therapeutic target. In view of the potential value of new diagnostic tools such as the recently reported Y 1 R selective 99m Tc-labeled peptide [11], we performed preclinical investigations on the expression of Y 1 Rs and ERs in breast cancer cells and tumors using well-established ER and NPY receptor agonists and antagonists. In particular, the influence of estrogens and antiestrogens on the expression and function of the Y 1 R protein was studied to explore the Y 1 R as a diagnostic target considering ER status and the impact of hormonal therapy with antiestrogens or aromatase inhibitors.
Among the investigated breast cancer cell types (ER-positive: three variants of MCF-7 cells, T-47-D cells; ER-negative: MDA-MB-231 cells and the triple-negative HCC1806 and HCC1937 cells), NPY receptors were only detected in ER-positive cells (Fig. 3  and 7) and identified as the Y 1 R subtype by confocal microscopy (Fig. 4) and radioligand binding ( Fig. 3 and 7). With approximately 40,000 receptors per cell, the basal Y 1 R protein density in wild type MCF-7 cells was found to be in the same range as in SK-N-MC neuroblastoma cells [19,41]. The Y 1 R protein expression was up-regulated by treatment with 1 nM 17b-estradiol in MCF-7 and -at a lower basal level -in T-47-D breast cancer cells. The estrogen induced Y 1 R protein expression reached its maximum after two days, which is indicative of a genomic process. The basal Y 1 R level in MCF-7 cells was 40-50% of that of the 17b-estradiol treated control when grown in medium containing hormonedepleted serum (ct-FCS) (Fig. 7B). Contrary to a previous finding [17], an effect of phenol red contaminants on Y 1 R expression was excluded by comparing the basal Y 1 R expression of MCF-7 cells grown in a phenol red containing and a phenol red-free medium, respectively (Fig. 7B). The Y 1 R expression was significantly downregulated by fulvestrant, a full ER antagonist described both, as an ER down-regulator [42] and an ER degrader [43], to approximately 25% of the basal level (Fig. 9A). As no estrogenic compounds were present in the medium supplement (ct-FCS), a ligand-independent ER activation mechanism may be involved to some extent in the basal Y 1 R expression. Ligand independent ER activation can be mediated by cross-talk activation pathways including protein kinase A and C or growth factor mediated signals [44]. In previous studies full ER antagonists such as fulvestrant were shown to be capable of blocking such signaling pathways [44].
The high expression and functionality of the Y 1 R supports speculations on a role of NPY in tumor growth, as suggested, for instance, for SK-N-MC [15,45] and MCF-7 cells [17]. Although the Y 1 R was demonstrated to be functionally active in MCF-7 cells (Fig. 6), NPY had no effect on cell proliferation (Fig. 5), which is in accordance with very recent results on human NCI-H295R adrenocortical carcinoma cells [46]. Y 1 R expression was stimulated by 17b-estradiol in a concentration-dependent manner (Fig. 8); the EC 50 value amounted to 20 pM. This is the first time that an up-regulation of the Y 1 R at physiologically relevant concentrations of 17bestradiol has been demonstrated at the protein level. These results are in accordance with the work of Amlal et al. [17], reporting an elevation of Y 1 R mRNA expression albeit at supra-physiological estradiol concentrations (10 and 100 nM). The EC 50 value of estradiol determined in the present study via Y 1 R up-regulation is in the same range as that reported for the up-regulation of the progesterone receptor mRNA in MCF-7 cells (44 pM; cf. [47]). As subtype selective ER antagonists are not available, appropriate agonists were used as pharmacological tools to identify the ER subtype involved. The high efficacy and potency of PPT suggests a predominant role of ERa in Y 1 R regulation, as PPT is devoid of any activity at ERb [36]. The EC 50 value is in good agreement with that reported for ERa from a co-transfection assay (< 0.1 nM, cf. [36]). Genistein, a phytoestrogen, was previously reported to be an ERb-preferring partial (50%) agonist and a weak full ERa agonist [48]. Genistein up-regulated the Y 1 R by 70% with an EC 50 value of 100 nM. This result matches with the reported data for ERa rather than for ERb, underlining that Y 1 R induction is ERa mediated.
The pure antiestrogen fulvestrant inhibited the ER-stimulated Y 1 R expression in a concentration-dependent manner (Fig. 9). The IC 50 value of 4.7 nM obtained for fulvestrant is in excellent accordance with data from a luciferase gene reporter assay [49]. Thus, the ERa-regulated expression of the Y 1 R is a suitable readout for the characterization of estrogens and antiestrogens.
The above-discussed results suggest a possible value of the Y 1 R as a surrogate marker of the ER status in breast cancer. Moreover, receptors of regulatory peptides such as NPY are in the focus of approaches to tumor targeting and molecular imaging of cancer [6][7][8][9][10][11]13,14,18]. Therefore, we investigated the effect of estradiol and tamoxifen treatment on the Y 1 R level in MCF-7 tumors growing subcutaneously in nude mice (Fig. 10). The regimen of antiestrogen treatment (cumulative dose of 36 mg/kg tamoxifen citrate) was adjusted over 14 days (three injections, 12 mg/kg), on one hand to stop tumor growth and on the other hand to prevent tumor regression and necrosis (cf. histology, Fig. S5). In accordance with the in vitro results autoradiography of the xenografts revealed strong expression of the Y 1 R in the presence of estradiol and an almost total down-regulation after antiestrogen treatment. Y 1 R protein expression in MCF-7 cells depends on the activation state of the ER. By analogy with these findings, very recently, fulvestrant treatment was reported to down-regulate the progesterone receptor levels, monitored by PET in STAT1deficient mammary tumors in mice [50], reflecting the response to endocrine therapy. In principle, a decrease in the expression of a membrane protein such as the Y 1 R might be exploited as a marker for the response to hormonal treatment as well. However, measuring a decrease, finally resulting in the lack of the signal is an unfavorable analytical parameter in view of high probability of false negative results. As Y 1 R down-regulation was a relatively fast process in vitro (,48 h) as well as in nude mice (,14 d) before tumor regression, negative PET results might be misinterpreted.
In conclusion, in view of the loss of the Y 1 R during tamoxifen treatment the suitability of this peptide receptor as a target for tumor therapy and imaging should be re-considered. In particular, in breast cancer patients the diagnostic value of the Y 1 R may be compromised due to Y 1 R down-regulation induced by therapeutically administered antiestrogens.

Dedication
Dedicated to Prof. Dr. Dr. Wolfgang Wiegrebe, Regensburg, on the occasion of his 80 th birthday.