Stabilization of prolactin receptor in breast cancer cells

Article metrics

Abstract

The role of the hormone prolactin (PRL) in the pathogenesis of breast cancer is mediated by its cognate receptor (PRLr). Ubiquitin-dependent degradation of the PRLr that negatively regulates PRL signaling is triggered by PRL-mediated phosphorylation of PRLr on Ser349 followed by the recruitment of the beta-transducin repeats-containing protein (β-TrCP) ubiquitin-protein isopeptide ligase. We report here for the first time that interaction between PRLr and β-TrCP is less efficient in human breast cancer cells than in non-tumorigenic human mammary epithelial cells. Furthermore, we demonstrate that both PRLr degradation and PRLr phosphorylation on Ser349 are impaired in breast tumor cells and tissues, an observation that directly correlates with enhanced expression of the PRLr in malignant breast epithelium. These findings represent a novel mechanism through which altered PRLr stability may directly influence the pathogenesis of breast cancer.

Introduction

The polypeptide hormone prolactin (PRL) is essential for mammary gland development and acts as a potent prosurvival factor and motility factor for mammary tissues (Clevenger et al., 1995, 2003; Das and Vonderhaar, 1997; Perks et al., 2004). Epidemiologic studies in humans (Tworoger et al., 2004; Harvey, 2005) and results obtained in transgenic mice (Wennbo et al., 1997; Wennbo and Tornell, 2000; Rose-Hellekant et al., 2003) point to a significant role for PRL during the pathogenesis of breast cancer. PRL mediates its activities by engaging the PRL receptor. Unlike the short and intermediate isoforms of this receptor, which either mediate a partial response to PRL or inhibit PRL-dependent signaling, the long and the ΔS1 isoforms confer the entire repertoire of PRL-induced signaling pathways (Clevenger et al., 2003). Given that the ΔS1 isoform responds only to increased concentrations of PRL (Kline et al., 2002), the regulation of abundance and activities of the long form prolactin receptor (PRLrL) is probably most important for PRL signaling outcomes, especially in human breast cancers where this form is predominantly expressed at the mRNA levels (Meng et al., 2004).

As a critical counter-regulatory mechanism, the ligand-induced downregulation of PRLrL limits the extent of PRL signaling. PRL triggers the interaction of PRLrL with the beta-transducin repeats-containing protein (β-TrCP) ubiquitin-protein isopeptide ligase (E3). This interaction depends on phosphorylation of Ser349 within a phospho-degron, which is highly conserved between the intracellular domains (ICDs) of PRLrL from different species. Such interaction results in ubiquitination and lysosomal degradation of the PRLrL and leading to the restriction of the magnitude and duration of PRL-induced signaling events (Li et al., 2004).

Whereas PRL is elaborated by both the pituitary and mammary epithelium, the importance of paracrine/autocrine effects of locally secreted PRL are emphasized by the inefficiency of inhibitors of pituitary PRL release in breast cancer therapy (Clevenger et al., 2003). Antagonists of PRL/PRLr kill human breast cancer cells in vitro and abrogate the tumorigenesis in the xenograft models demonstrating that persistent signaling induced by locally secreted PRL is essential for growth and survival of these cells (Goffin et al., 2005). However, it is not clear how these effects of PRL may persist given that, in normal tissues, PRL triggers rapid degradation and downregulation of PRLr and limits the extent of PRL signaling.

Our previous studies using an antibody directed against all forms of PRLr showed that the levels of PRLr are decreased in the breast cancer intratumoral stromal compartment compared to the stroma from benign tissue. This decrease is attributed to the downregulation of PRL receptor in response to its ligand secreted by tumor cells (Reynolds et al., 1997). However, the levels of PRLr in tumor cells are not decreased in comparison with benign mammary cells (Clevenger et al., 1995; Reynolds et al., 1997) suggesting a possibility that downregulation and degradation of PRLr in tumor cells might be impaired. Here, we show that phosphorylation of PRLrL on Ser349 is decreased in breast cancer cell lines and primary cancer tissues that exhibit stabilization and accumulation of PRLrL.

Results

Previous studies have shown that, in comparison to surrounding normal stroma, malignant breast epithelium demonstrates increased levels of PRLr expression, suggesting altered regulation and degradation of the PRLr in tumor cells. To test this hypothesis, we measured the rate of PRLrL turnover in human mammary cell lines. A greater stability of PRLrL in T47D breast cancer cells as compared to non-tumorigenic human mammary epithelial cells (HMEC) was observed in a pulse chase analysis (Figure 1a). Further analysis of PRLrL degradation demonstrated that two other human breast cancer cell lines (MCF7 and MDA-MB-468) also exhibited inefficient proteolysis of PRLrL (Figure 1b). These results suggest that downregulation and degradation of PRLrL is indeed impaired in human breast cancer cells.

Figure 1
figure1

Impaired degradation of PRLrL in breast cancer cells. (a) Pulse chase analysis of endogenous PRLrL degradation in human mammary cells metabolically labeled with 35S-methionine/35S-cysteine, and chased by incubation in the medium containing 2 mM of unlabeled methionine and cysteine in the presence of 0.5% FBS and PRL (20 ng/ml). Cells were harvested at different time points of chase and the lysates were immunoprecipitated with U5 antibody against PRLr and analysed by autoradiography. (b) The graph depicts % of remaining PRLrL at each time point of the chase quantified from three independent experiments. Cell lines used: HMEC, closed squares; T47D, open squares; MDA-MB468, closed circles; MCF7, open circles. An average of three experiments is shown.

We previously found that proteolytic turnover of PRLrL depends on the interaction between PLRrL and β-TrCP E3 ubiquitin ligase in 293T human embryo kidney cells (Li et al., 2004). Thus, to investigate the mechanisms underlying the delayed kinetics of PRLrL degradation in breast cancer cells, we assessed binding of PRLrL to β-Trcp by coimmunoprecipitation in human mammary cells. We found that the interaction between these proteins was much more efficient in HMEC than in three human cancer cell lines (Figure 2a, upper panel). It is important to note that decreased association of PRLrL with β-TrCP was detected despite high levels of PRLrL in tumor cells, which is consistent with stabilization of PRLrL in these cells seen in Figure 1. Furthermore, inefficient coimmunoprecipitation of β-TrCP with PRLrL cannot be attributed to low levels of β-TrCP in the whole-cell extracts from breast cancer cells. On contrary, consistent with the earlier observations that β-TrCP levels are induced in human breast cancers (Spiegelman et al., 2002), these levels were increased in tumor cells compared to HMEC (Figure 2a). These data suggest that increased stability of PRLrL may result from the impaired recruitment of E3 ubiquitin ligase that is expected to result in inefficient ubiquitination and degradation of PRLrL.

Figure 2
figure2

Impaired interaction between PRLrL and β-TrCP in breast cancer cells. (a) Lysates from serum-starved HMEC and breast cancer cells (3 mg each) were analysed by immunoprecipitation with U5 monoclonal anti-PRLr antibody or irrelevant monoclonal antibody (M2, ‘Fl’) and immunoblotting with sc-20992 PRLr antibody to detect PRLrL (M.w. 90 kD) or with β-TrCP antibody to detect different β-TrCP species (including β-TrCP1 and β-TrCP2, M.w. 58–68 kD). Lower panels depict immunoblotting analysis of β-TrCP and α-tubulin (Tub) levels in the whole-cell lysates (WCE, 50 μg) from HMEC and cancer cells. A representative of four experiments is shown. (b) Phosphorylation-binding assay. GST-PRLrL proteins were incubated with extracts from indicated cells (grown in the presence of 0.5% serum and pretreated with PRL (20 ng/ml) for 30 min) in the presence ATP, except in the first lane where 293T extract was not supplemented with ATP. Binding of in vitro-translated and 35S-labeled β-Trcp2 to immobilized GST-PRLrL proteins was analysed by means of SDS–PAGE and autoradiography (upper panel). Input of radioactive β-Trcp2 is also shown (‘Input’). Aliquots of these reactions were analysed by immunoblotting with GST antibody to control the loading of GST-PRLrL proteins (lower panel). A representative of three experiments is shown.

Stabilization and impaired interaction with β-TrCP in cancer cells due to a decreased phosphorylation within a specific destruction motif is characteristic of oncogenic activation of another β-TrCP substrate, β-catenin (Polakis, 1999). In some cases, this impaired phosphorylation results from mutations in Ser phospho-acceptor site residues in β-catenin that abrogate its binding to β-TrCP (Polakis, 2000). We have sequenced the PRLr DNA corresponding to the sequence that encodes portion of PRLr cytoplasmic tail encompassing Ser349 but found no mutations in either HMEC or all three cancer cell lines (MCF7, MDA-MB468, and T47D). These results are in line with reported lack of mutations within the ICD of PRLrL in primary breast cancers (Glasow et al., 2001; Canbay et al., 2004).

We previously found that Ser349Ala substitution in PRLrL results in impaired recruitment of β-TrCP. This result indicated that interaction between β-TrCP and PRLrL depends on phosphorylation of the latter protein on Ser349 within its conserved phospho-degron. Such phosphorylation can be indirectly measured in vitro by phosphorylation-binding assays using cell lysates (as a source of kinase activity), bacterially produced GST-PRLrL protein (as a substrate) and binding of an in vitro translated β-TrCP (as a detection mode) (Li et al., 2004). Using this assay to compare the ability of lysates from different cell lines to phosphorylate GST-PRLrL and enable its interaction with β-TrCP in vitro, we found that the lysates from human breast cancer cells exhibited a noticeably lesser ability to mediate such binding as compared either to the lysates from HMEC or to 293T cell lines (Figure 2b), both of which exhibit efficient interaction between PRLrL and β-TrCP in vivo (Li et al. (2004) and Figure 1a). These data suggest that, in breast cancer cells, inefficient binding of β-TrCP to PRLrL and concurrent stabilization of PRLrL may result from impaired activity of a yet to be identified PRLrL kinase.

In order to corroborate these data and directly assess the extent of PRLrL phosphorylation on Ser349 we developed a phospho-specific antibody, which efficiently recognized recombinant human PRLrL but not its S349A mutant (Figure 3a). Consistent with our previously published results of phosphorylation-binding assay in vitro (Li et al., 2004), analysis of endogenous PRLrL in non-tumorigenic mammary cells showed that phosphorylation of PLRr on Ser349 is stimulated by PRL treatment (Figure 3b). These data together with our earlier results suggest a mechanism for the ligand-dependent downregulation of PRLrL that includes PRL-dependent activation of a yet to be identified PRLrL kinase, phosphorylation of PRLrL on Ser349, recruitment of β-TrCP, ubiquitination and degradation of PRLrL.

Figure 3
figure3

Characterization of pS349 antibody. (a) 293T cells were transfected with either empty vector (pCDNA3, lane 1) or with Flag-tagged PRLrLwt (lane 2), or with Flag-PRLrLS349A mutant (lane 3) as indicated. At 48 h after transfection, cells were treated with PRL (200 ng/ml for 30 min) and harvested. PRLr proteins were purified by immunoprecipitation with Flag antibody and analysed by immunoblotting with Flag antibody (lower panel) or phospho-PRLrL antibody (pS349, upper panel). (b) Phosphorylation of PRLrL within the β-TrCP recognition motif is ligand-dependent. HMEC cells were serum starved and treated with PRL (200 ng/ml) for indicated times. PRLrL was immunoprecipitated from cell lysates and analysed by immunoblotting with antibodies against either pS349 (upper panel). The blots were reprobed using anti-ICD PRLr antibody (lower panel). A representative of four experiments is shown.

We then compared phosphorylation of endogenous PRLrL in tumorigenic and non-tumorigenic mammary cells by immunoprecipitation-immunoblotting assay. Given that tumor cells express much more PRLrL (Figure 2a), the initial amounts of lysates taken into the immunoprecipitation reactions were adjusted to produce comparable total levels of PRLrL. This analysis showed that phosphorylation of PRLrL on Ser349 is much less efficient in human breast cancer cells compared to the non-tumorigenic HMEC (Figure 4a). Given that our previous studies in vitro and in nonmammary cells have already established that phosphorylation of the PRLrL on Ser349 is required for PRLrL interaction with β-TrCP, as well as for ubiquitination and degradation of PRLrL (Li et al., 2004), these data suggest that impaired phosphorylation of PRLrL in breast cancer cells may result in inefficient β-TrCP recruitment (as seen in Figure 2) and PRLrL stabilization (as seen in Figure 1). Furthermore, in vitro kinase assay using the lysates from mammary cells as a source of kinase activity, GST-PRLrL as a substrate and immunoblotting with phospho-Ser349-specific antibody as a detection mode showed that breast cancer cells exhibited lesser kinase activity compared to HMEC (Figure 4b). This data together with our sequencing results showing the lack of mutations in PLRrL suggest that inefficient phosphorylation of PRLrL Ser349 (Figure 4a) along with decreased interaction with β-TrCP in cells and in vitro (Figure 2) might result from inhibited activity of yet undetermined PRLrL kinase.

Figure 4
figure4

Phosphorylation of PRLrL in human mammary cell lines. (a) Phosphorylation of PRLrL within the β-TrCP recognition motif is impaired in breast cancer cells. Lysates from HMEC (6 mg) or from breast cancer cells (2 mg for each cell line) pretreated with PRL (200 ng/ml for 30 min before harvesting) were analysed by immunoprecipitation of endogenous PRLrL using U5 antibody followed by immunoblotting analysis with pS349 antibody (upper panel). The blots were re-probed using anti-ICD PRLr antibody (lower panel). A representative of four experiments is shown. (b) Phosphorylation of GST-PRLr in vitro by the lysates from indicated cell was analysed by immunoblotting with phospho-specific pS349 antibody. The experiment was carried out as indicated in the legend to Figure 2b. ATP was omitted from the reaction shown in the left lane. The blots were re-probed with GST antibody to control the loading of GST-PRLrL proteins (lower panel). A representative of four experiments is shown.

We further assessed the extent of PRLrL phosphorylation on Ser349 (determined as a ratio between S349-phospho-specific signal and total PRLrL levels analysed by immunoprecipitation followed by immunoblotting with respective antibody) as an indirect inverse indicator of PRLr stability in human primary breast cancer tissues. Analysis of unmatched samples from tumor or apparently normal tissues from patients with mammary malignancies showed that PRLrL was hypo-phosphorylated on Ser349 in breast cancer tissues compared to the benign samples (Figure 5a–b). The relative ratio of phosphorylated PRLrL to the total PRLrL was 1.47±0.48 in normal tissues and 0.75±0.38 in cancer samples (P<0.05).

Figure 5
figure5

Phosphorylation of PRLrL on Ser349 and accumulation of PRLrL in human breast tissues. (a) Lysates from samples of unmatched primary apparently normal or malignant mammary tissues (2 mg each) were analysed by immunoprecipitation of endogenous PRLrL using U5 antibody followed by immunoblotting analysis with pS349 antibody (upper panel). The blots were reprobed using anti-ICD PRLr antibody (lower panel). (b) Scatter plot graph of ratios between signal from immunoblot with pS349 antibody and signal from anti-ICD PRLr antibody calculated from (a). (c) Lysates from samples of matched primary apparently normal (N) or malignant (T) mammary tissues (2 mg each) were analysed by immunoprecipitation of endogenous PRLr using U5 antibody followed by immunoblotting analysis with pS349 antibody (upper panel). The blots were re-probed using anti-ICD PRLr antibody (lower panel). (d) Scatter plot graph of ratios between signal from immunoblot with pS349 antibody and signal from anti-ICD PRLr antibody calculated from (c).

Given that levels of phosphorylation and expression of PRLrL may substantially vary between the individuals, a comparison between malignant and benign tissue within the same patient was carried out via the analysis of matched tumor and benign tissues from a separate patient cohort. As seen from Figures 5c–d, the PRLrL is hypophosphorylated and demonstrates increased levels in cancer tissues compared to the adjacent normal tissues in six out of eight primary breast cancer patients. These results are in line with the data obtained from the unmatched samples and indicate that phosphorylation and degradation of PRLrL might be impaired in human breast cancers.

To corroborate the obtained biochemical data demonstrating a reduction in phosphorylation at S349 in the PRLr, immunohistochemistry on primary malignant and adjacent normal human breast tissues was performed with the affinity purified anti-pS349 antisera. Initial studies were performed on infiltrating ductal carcinomas of the breast that included normal adjacent surrounds, that were subsequently augmented through the use of a breast tissue progression microarray (courtesy of the CHTN, NIH) that contained seven normal breast tissues and seven primary infiltrating ductal carcinomas. For these studies, antigen–antibody complexes on the tissue surface were detected through the use of a secondary biotinylated antibody, followed by streptavidin-horseradish peroxidase and diaminobenzidine. Control experiments (with or without antigenic phospho-peptide as a blocking agent) demonstrate that this antibody stained the epithelial compartment of mammary tissue in a specific manner (Figure 6a). Anti-pS349 label on normal tissues was noted to be intense and was largely localized to perinuclear vacuoles, suggesting of localization of the label to either lysosomes or the Golgi apparatus (all seven normal specimens demonstrating 100% labeling of the epithelial compartment with 2–3+ intensity). Confirming our biochemical analysis, the anti-pS349 label on malignant breast tissues was significantly reduced in both intensity and extent (three specimens demonstrating 0% labeling of the epithelial compartment; four specimens demonstrating 80–90% labeling with only 1+ intensity). A representative image is shown in Figure 6b.

Figure 6
figure6

Immunohistochemical analysis of phosphorylation of PRLrL on Ser349 and PRLrL levels in human breast tissues. (a) Immunohistochemical labeling of normal breast tissue with anti-pS349 PRLr (1:25 dilution) incubated in the absence or presence of PRLrL-derived phospho-Ser349-specific peptide (5 μg/ml). (b) Immunohistochemical labeling of normal and malignant breast tissues with anti-pS349 PRLr (1:25 dilution). N=normal adjacent breast tissue, T=infiltrating ductal carcinoma. Intense labeling of perinuclear vacuoles is noted in the normal ductal epithelium; both the intensity and extent of this label are reduced in the malignant tissues. Magnification of both photomicrographs is × 400. (c) Immunohistochemical labeling of normal and malignant breast tissues with anti-ICD PRLr (sc-20992, H-300). N=normal adjacent breast tissue, T=infiltrating ductal carcinoma. Both nuclear and cytoplasmic labeling of normal and malignant tissues is noted. Magnification of both photomicrographs is × 400.

Additional immunohistochemical analysis of breast tissues using an anti-PRLr antibody directed against the ICD of the long and deltaS1 PRLr (anti-ICD PRLr) revealed that the majority of the PRLr was intracellular, consistent with prior studies (Reynolds et al., 1997). While both 100% of the epithelial compartment in normal and malignant breast tissues demonstrated label, significant differences in the labeling intensity were noted; all seven normal specimens demonstrated only 1+ intensity; whereas all seven malignant breast specimens demonstrated 3+ intensity. A representative image is shown in Figure 6c. Taken together, the immunohistochemistry with the anti-pS349 and –ICD PRLr antibodies in primary breast tissues supports the initial biochemical analysis and the hypothesis that resulted from these data; namely, that phosphorylation of the PRLr is decreased at position S349 in malignant tissues, resulting in enhanced levels of long and delta S1 PRLr expression.

Discussion

Our data presented here indicate that, in human breast cancer cells, the activity of a yet to be identified kinase that phosphorylates PRLrL on Ser349 is impaired (Figure 2b). As a result, PRLrL phosphorylation within its phospho-degron is decreased (Figure 4) leading to inefficient recruitment of the β-TrCP E3 ubiquitin ligase (Figure 2a), stabilization (Figure 1) and accumulation (Figure 2a) of PRLrL. Hypo-phosphorylation of PRLrL and high levels of its expression is also detected in primary human breast cancer tissues (Figures 5 and 6). Taken together, these data support the hypothesis that phosphorylation of the PRLr on Ser349 is decreased in malignant breast cells resulting in stabilization and accumulation of PRLr. Such stabilized receptor should be resistant to downregulation by locally secreted ligand and capable of persistent augmentation of PRL-induced effects (including accelerated growth, increased survival and motility) in breast cancer cells. Whereas future prospective studies should determine whether impaired phosphorylation of PRLrL on Ser349 may serve as prognostic biomarker, our results presented here provide the first in vivo demonstration of altered PRLrL expression and phosphorylation in human breast cancer and suggest that the widespread upregulation of this isoform in malignant breast tissues is of biologic significance.

Abrogation of phosphorylation of PRLrL within its phospho-degron is expected to result in a receptor that is resistant to downregulation upon interacting with its ligand. We previously observed that CHO hamster cells harboring human PRLrLS349A mutant exhibit an elevated level of signaling via signal transducer and activator of transcription (Stat)5 and of human PRL-dependent cell proliferation as compared to the cells expressing the wild-type receptor (Li et al., 2004). It is plausible that impaired phosphorylation of PRLrL on Ser349 in breast cancer cells should also result in a stabilized receptor that is likely to augment the effects of PRL (including accelerated growth, increased survival and motility) in these cells. Our findings that PRLr degradation is impaired in cancer cells provide a mechanistic insight as per how breast cancer cells may withstand the phosphorylation- and β-TrCP-dependent negative regulation of PRL signaling and maintain the constitutive activation of PRL-induced pathways. Indeed, a constitutive activation of a number of pathways that are known to be downstream of PRLr (including Stat5 (Clevenger, 2004)) has been found in human breast cancers. Future studies will determine the role of PRLrL phosphorylation and stability in mammary tumorigenesis.

Impaired phosphorylation and degradation of PRLrL in cancer cells is reminiscent of alterations that are known to occur with some proto-oncogenic proteins, whose stability is regulated by phosphorylation including cyclin D1 (Diehl, 2002) and β-catenin (Polakis, 1999). Interestingly, whereas deletions of or somatic mutations within the phospho-degrons of these proteins have been reported, there is no current evidence that PRLrL is mutated in either primary human breast cancers (Glasow et al., 2001; Canbay et al., 2004) or breast cancer cell lines (this study). Instead, our current data indicate that a PRLrL kinase activity is inhibited in human breast cancer cells (Figures 2b, 4b). Future studies will be aimed at the identification of the PRL-induced kinase that phosphorylates the PRLrL on Ser349 within the PRLrL phospho-degron as well as at the activation pathways for such kinase. These studies should provide a mechanistic understanding of the altered regulation of the PRLrL in human breast cancer cells and may potentially identify novel targets for breast cancer therapy.

Materials and methods

Cell lines and primary tissue samples

293T human embryo kidney cells and human breast cancer T47D, MDA-MB468 and MCF7 cells were kindly provided by Dr Z Ronai (Burnham Institute, San Diego, CA, USA). Immortalized non-tumorigenic human mammary epithelial cells (Kim et al., 2002) were a generous gift of Dr DN Foster (Department of Animal Science, University of Minnesota, St Paul, MN, USA). Anonymized matched and nonmatched normal and malignant breast tissues and tissue microarray were obtained from the Cooperative Human Tissue Network and from the Tissue Resource Core of the Breast SPORE of the Robert H Lurie Comprehensive Cancer Center of Northwestern University (Chicago, IL, USA) under appropriate IRB protocols. Cells were cultured as previously described (Kim et al., 2002; Tang et al., 2005) and treated with human recombinant PRL (kindly provided by Dr AF Parlow, National Hormone and Peptide Program, Torrance, CA, USA) as indicated. Genomic DNA from cells was extracted using the phenol-chloroform procedure, PCR amplified using 5′-IndexTermGCCTTGGGATGCCAAGACTTTCCTCCC-3′ and 5′-IndexTermCAACAGAGTGGCCGGTGCACCTGC-3′ primers and subjected to dideoxy sequencing using a nested 5′-IndexTermTTGTGGTAAGAGGATCTGG-3′ primer.

Antibodies and immunotechniques

Cells were harvested and the lysates were prepared using Tris-HCl buffer (50 mM, pH 7.4). containing 150 mM NaCl, 50 mM NaF, 1% Nonidet P-40 and a protease inhibitors cocktail (all reagents from Sigma, St Louis, MO, USA). Primary tissues were pulverized in liquid N2 and extracted with the same buffer. The antibodies against Flag (M2, Sigma), PRLr (U5 from Affinity BioReagents, Golden, CO, or sc-20992 from Santa Cruz Biotechnology, Santa Cruz, CA, USA), glutathione S-transferase (GST, Santa Cruz, CA, USA) were purchased. Rabbit polyclonal antibody against β-TrCP were previously described (Spiegelman et al., 2002). Ser349-phospho-specific antibodies (pS349) were raised in rabbits in collaboration with PhosphoSolutions, Inc (Aurora, CO, USA) against the phospho-peptide YLDPDTDpSGRGSCD. Reactive sera were passed over a column bearing a nonphosphorylated version of this peptide and finally purified using a phospho-peptide (containing the sequence shown above) affinity column. Immunoprecipitation and immunoblotting procedures were described elsewhere (Li et al., 2004). For the pulse chase analysis, cells were grown in 100 mm dishes and metabolically labeled with a 35S-methionine/35S-cysteine mixture (Perkin Elmer, Boston, MA, USA). Cells were harvested at each time point of the chase with complete DMEM supplemented with FBS (0.5%), PRL (20 ng/ml) and unlabeled methionine and cysteine (2 mM). Harvested cells were lysed and PRLr proteins were immunoprecipitated with PRLr antibody, separated on SDS–PAGE and analysed by autoradiography. Immunohistochemistry was carried out using the pS349 PRLr antibody described above, or with an anti-PRLr antibody directed against the ICD of the long and deltaS1 PRLr (1:200 dilution; anti-ICD PRLr antibody from Santa Cruz H-300, sc-20992). A streptavidin horseradish immunoperoxidase method reported previously was used with some modifications (Cartun and Pedersen, 1989). Sections of 5 μm from formalin-fixed paraffin-embedded tissue microarrays were deparaffinized in xylene and rehydrated in graded alcohol. After deparaffinization, heat induced antigen retrieval by boiling slides in 10 mM citrate buffer pH 6 for 20 min was carried out. The sections were blocked with a peroxidase blocking system (Dako, Carpinteria CA, USA) for 10 min. Sequential incubations included the following: 2% normal goat serum for 20 min; affinity purified S349 rabbit polyclonal primary antibody at 1:25 dilution for 60 min at room temperature; secondary biotinylated anti-rabbit IgG for 30 min; and, finally, the streptavidin-biotinylated horseradish peroxidase complex reagent (Dako) for 30 min at RT. After each incubation, the slides were washed three times in buffer of 3 min each. Sections were then exposed to the chromagen DABplus (Dako) for 5 min and were counterstained in hematoxylin, dehydrated, cleared and mounted.

In vitro phosphorylation-binding assay

In vitro phosphorylation-binding assay was carried out as previously described (Li et al., 2004). Briefly, recombinant GST-PRLr (cytoplasmic tail) proteins expressed in bacteria and immobilized on glutathione beads were phosphorylated with lysates (25 μg) from the indicated cells in the presence or absence of ATP for 30 min at 30°C followed by stringent washing with stripping buffer and reequilibration with binding buffer. Immobilized GST-PRLr proteins were then incubated with in vitro translated and 35S-labeled β-TrCP for 60 min at 4°C. The beads were then extensively washed with binding buffer and associated proteins were analysed using SDS–PAGE and autoradiography and immunoblotting with GST antibody.

Abbreviations

PRL:

prolactin

PRLr:

prolactin receptor

β-TrCP:

beta-transducin repeats-containing protein

E3:

ubiquitin-protein isopeptide ligase

HMEC:

human mammary epithelial cells

Stat:

signal transducer and activator of transcription

ICD:

intracellular domain

References

  1. Canbay E, Degerli N, Gulluoglu BM, Kaya H, Sen M, Bardakci F . (2004). Curr Med Res Opin 20: 533–540.

  2. Cartun RW, Pedersen CA . (1989). J Histotechnol 12: 273–280.

  3. Clevenger CV . (2004). Am J Pathol 165: 1449–1460.

  4. Clevenger CV, Chang WP, Ngo W, Pasha TL, Montone KT, Tomaszewski JE . (1995). Am J Pathol 146: 695–705.

  5. Clevenger CV, Furth PA, Hankinson SE, Schuler LA . (2003). Endocr Rev 24: 1–27.

  6. Das R, Vonderhaar BK . (1997). J Mammary Gland Biol Neoplasia 2: 29–39.

  7. Diehl JA . (2002). Cancer Biol Ther 1: 226–231.

  8. Glasow A, Horn LC, Taymans SE, Stratakis CA, Kelly PA, Kohler U et al. (2001). J Clin Endocrinol Metab 86: 3826–3832.

  9. Goffin V, Bernichtein S, Touraine P, Kelly PA . (2005). Endocr Rev 26: 400–422.

  10. Harvey PW . (2005). J Appl Toxicol 25: 179–183.

  11. Kim H, Farris J, Christman SA, Kong BW, Foster LK, O’Grady SM et al. (2002). Biochem J 365: 765–772.

  12. Kline JB, Rycyzyn MA, Clevenger CV . (2002). Mol Endocrinol 16: 2310–2322.

  13. Li Y, Kumar KG, Tang W, Spiegelman VS, Fuchs SY . (2004). Mol Cell Biol 24: 4038–4048.

  14. Meng J, Tsai-Morris CH, Dufau ML . (2004). Cancer Res 64: 5677–5682.

  15. Perks CM, Keith AJ, Goodhew KL, Savage PB, Winters ZE, Holly JM . (2004). Br J Cancer 91: 305–311.

  16. Polakis P . (1999). Curr Opin Genet Dev 9: 15–21.

  17. Polakis P . (2000). Genes Dev 14: 1837–1851.

  18. Reynolds C, Montone KT, Powell CM, Tomaszewski JE, Clevenger CV . (1997). Endocrinology 138: 5555–5560.

  19. Rose-Hellekant TA, Arendt LM, Schroeder MD, Gilchrist K, Sandgren EP, Schuler LA . (2003). Oncogene 22: 4664–4674.

  20. Spiegelman VS, Tang W, Chan AM, Igarashi M, Aaronson SA, Sassoon DA et al. (2002). J Biol Chem 277: 36624–36630.

  21. Tang W, Li Y, Yu D, Thomas-Tikhonenko A, Spiegelman VS, Fuchs SY . (2005). Cancer Res 65: 1904–1908.

  22. Tworoger SS, Eliassen AH, Rosner B, Sluss P, Hankinson SE . (2004). Cancer Res 64: 6814–6819.

  23. Wennbo H, Gebre-Medhin M, Gritli-Linde A, Ohlsson C, Isaksson OG, Tornell J . (1997). J Clin Invest 100: 2744–2751.

  24. Wennbo H, Tornell J . (2000). Oncogene 19: 1072–1076.

Download references

Acknowledgements

We thank Drs Ronai, Foster, and Parlow for providing reagents and CHTN and the Tissue Resource Core of the Breast SPORE of the Robert H Lurie Comprehensive Cancer Center of Northwestern University (Chicago, IL, USA) for primary tissues and the array. This work was supported in part by The University of Pennsylvania Cancer Center Pilot Grant and The Susan G Komen Breast Cancer Foundation grant BCTR0504447 (to SYF) and NCI Grant CA069294 (to CVC).

Author information

Correspondence to S Y Fuchs.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Li, Y., Clevenger, C., Minkovsky, N. et al. Stabilization of prolactin receptor in breast cancer cells. Oncogene 25, 1896–1902 (2006) doi:10.1038/sj.onc.1209214

Download citation

Keywords

  • prolactin
  • receptor
  • β-TrCP
  • ubiquitin
  • E3 ligase
  • breast cancer

Further reading