Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Integrated and Functional Genomics Analysis Validates the Relevance of the Nuclear Variant ErbB380kDa in Prostate Cancer Progression

  • Mahmoud El Maassarani,

    Affiliation Equipe 2RCT, Université de Poitiers, Faculté des Sciences Fondamentales, Pôle Biologie- Santé, 1 rue G. Bonnet, 86073, Poitiers cedex 9, France

    ⨯
  • Alice Barbarin,

    Affiliation Equipe 2RCT, Université de Poitiers, Faculté des Sciences Fondamentales, Pôle Biologie- Santé, 1 rue G. Bonnet, 86073, Poitiers cedex 9, France

    ⨯
  • Gaëlle Fromont,

    Affiliation Centre Hospitalier Universitaire Bretonneau, Laboratoire d’Anatomopathologie, INSERM U1069, 37000 Tours, France

    ⨯
  • Ouafae Kaissi,

    Affiliation LTI Laboratory, Abdelmalek Essaadi University, ENSAT, BP 1818, 90 000 Tangier, Morocco

    ⨯
  • Margot Lebbe,

    Affiliation Equipe 2RCT, Université de Poitiers, Faculté des Sciences Fondamentales, Pôle Biologie- Santé, 1 rue G. Bonnet, 86073, Poitiers cedex 9, France

    ⨯
  • Brigitte Vannier,

    Affiliation Equipe 2RCT, Université de Poitiers, Faculté des Sciences Fondamentales, Pôle Biologie- Santé, 1 rue G. Bonnet, 86073, Poitiers cedex 9, France

    ⨯
  • Ahmed Moussa,

    Affiliation LTI Laboratory, Abdelmalek Essaadi University, ENSAT, BP 1818, 90 000 Tangier, Morocco

    ⨯
  • Paule Séité

    paule.seite@univ-poitiers.fr

    Affiliation Equipe 2RCT, Université de Poitiers, Faculté des Sciences Fondamentales, Pôle Biologie- Santé, 1 rue G. Bonnet, 86073, Poitiers cedex 9, France

    ⨯

Abstract

The EGF-family of tyrosine-kinase receptors activates cytoplasmic pathways involved in cell proliferation, migration and differentiation in response to specific extracellular ligands. Beside these canonical pathways, the nuclear localization of the ErbB receptors in primary tumours and cancer cell lines led to investigate their role as transcriptional regulators of cancer genes. The nuclear localization of ErbB3 has been reported in various cancer tissues and cell lines but the nuclear functions and the putative correlation with tumour progression and resistance to therapy remain unclear. We first assessed ErbB3 expression in normal and tumour prostate tissues. The nuclear staining was mainly due to an isoform matching the C-terminus domain of the full length ErbB3185kDa receptor. Nuclear staining was also restricted to cancer cells and was increased in advanced castration-resistant prostate cancer when compared to localized tumours, suggesting it could be involved in the progression of prostate cancer up to the terminal castration-resistant stage. ChIP-on-chip experiments were performed on immortalized and tumour cell lines selected upon characterization of endogenous nuclear expression of an ErbB380kDa isoform. Among the 1840 target promoters identified, 26 were selected before ErbB380kDa-dependent gene expression was evaluated by real-time quantitative RT-PCR, providing evidence that ErbB380kDa exerted transcriptional control on those genes. Some targets are already known to be involved in prostate cancer progression even though no link was previously established with ErbB3 membrane and/or nuclear signalling. Many others, not yet associated with prostate cancer, could provide new therapeutic possibilities for patients expressing ErbB380kDa. Detecting ErbB380kDa could thus constitute a useful marker of prognosis and response to therapy.

Citation: El Maassarani M, Barbarin A, Fromont G, Kaissi O, Lebbe M, Vannier B, et al. (2016) Integrated and Functional Genomics Analysis Validates the Relevance of the Nuclear Variant ErbB380kDa in Prostate Cancer Progression. PLoS ONE 11(5): e0155950. https://doi.org/10.1371/journal.pone.0155950

Editor: Zoran Culig, Innsbruck Medical University, AUSTRIA

Received: July 20, 2015; Accepted: May 7, 2016; Published: May 18, 2016

Copyright: © 2016 El Maassarani et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: Microarray data are available at ArrayExpress: E-MTAB-3499, E-MTAB-3504.

Funding: This work was supported by Ligue contre le cancer, LCC PS/2012-13-14 to PS; Groupe Pasteur Mutuality to AB; and Ministère de l'éducation nationale de la recherche et des technologies 2010-2014 to MEM. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

The EGF-receptors family consists of four tyrosine kinase (TK) receptors (EGFR/ErbB1, HER2/ErbB2, HER3/ErbB3 and HER4/ErbB4) with an extracellular ligand-binding domain, a transmembrane domain and an intracytoplasmic domain. Upon specific ligand binding at cell surface, ErbB receptors dimerize and trigger intracellular pathways involved in cell proliferation, migration and differentiation [1]. The deregulation of ErbB downstream pathways, through three mains mechanisms -increased receptor expression, increased ligand expression or activating mutation of the TK domain- is frequently involved in tumorigenesis [2]. In addition to their localization at the plasma membrane, ErbB receptors have also been described in the nucleus and various functions have been ascribed to the translocated ErbB receptors among which are cell proliferation, DNA repair and transcription control [3–8]. In this regard, the full-length ErbB1 receptor in the nucleus is associated with increased expression of proliferative and metastatic genes [9–11]. The expression of the pro-apoptotic and pro-angiogenic COX-2 protein is modulated upon ErbB2-binding on the COX-2 promoter in mammary tumour cells [12]. ErbB2 also functions as a transcriptional co-activator of the Stat3 protein on the CCND1 (Cyclin D1) promoter [13] and collaborates with ErbB1 to up-regulate STAT1 gene expression [14]. The full length ErbB3185kDa protein or short isoforms have been described in the nucleus of normal mammary epithelial cells [15] and Schwann cells [16] as well as in various human cancer cell lines: breast (SKBr3, MCF-7 et BT474), lung (H226 et SCC6) head and neck (SCC6 et SCC1) and colon (LoVo et CaCO2) [15, 17–19]. In H358 non-small cell lung cancer (NSCLC) and SKBR3 breast cell lines, ErbB3 has been shown to exert function as a co-transcriptional activator for CCND1 gene expression [17,19]. The strong transactivation potential of ErbB3 relies almost exclusively on the C terminal domain of the receptor [19].

Prostate cancer (PCa) is the most common urologic malignancy and the second leading cause of cancer death in the industrialized countries. Tumour recurrence after surgical ablation or radiation therapy frequently occurs in a significant fraction of patients that develop disseminated aggressive disease. As PCa cells rely on androgen receptor (AR) activity for growth and proliferation, androgen withdrawal therapy has been endorsed in those advanced cases [20]. Beneficial effects are observed for about 24 months before the patient develops fatal castration-resistant-prostate-cancer (CRPC). Molecular mechanisms involved in therapy resistance still remain unclear. One explanation for a patient’s relapse after treatment could come from cross-talk between AR signalling and the epithelial growth factor family of receptors (EGFR) pathways [21]. The ErbB receptors and their ligands are expressed in the normal adult prostate to maintain the homeostasis of the organ and ErbB1, ErbB2 and ErbB3 are frequently increased during PCa progression [22]. In CRPC, increased ErbB2 receptor and/or ErbB1, ErbB2 or ErbB3-ligands expression is associated with poor prognosis and metastasis [23–25].

To by-pass the mechanisms of resistance to androgen-therapy, ErbB1 and ErbB2 TK-inhibitors have been tested alongside hormone-based therapy in patients with high grade, CRPC tumours. However, trials were not successful as tumour cells developed compensatory pathways by regulating ErbB3 expression and localization [18,26]. Indeed, several studies have reported the nuclear localization of ErbB3 in prostate tissues and PCa cell lines potentially correlated with disease progression [18,27].The nuclear ErbB3 protein could thus play a pivotal role in PCa progression and therapy resistance, but the molecular mechanisms involved still remain unexplored.

The current study demonstrates the functional relevance of ErbB3 nuclear localization in PCa and provides new clues for the understanding of pathways associated with cancer progression and therapy resistance.

Materials and Methods

Patients Tissue microarrays

Written informed consents were obtained from patients in accordance with the requirements of the institutional ethic committee of Centre Hospitalo-Universitaire de Poitiers that approved the study. 169 formalin-fixed, paraffin-embedded, archival tissues from patients treated at the “Centre Hospitalier Universitaire de Poitiers” for PCa, were included. 137 tissue samples were obtained from hormone-sensitive (HS) patients that underwent radical prostatectomy for clinically localised cancer, none of these patients received hormone deprivation therapy before surgery. Of these, 97 were stage pT2 and 40 pT3. The Gleason scoring was respectively 6 in 38 patients, 7 in 83 patients, and 8 or more in 16 patients. 32 tissue samples were obtained by trans-uretral resection from CRPC patients. Normal appearing prostate tissue from 50 patients treated by radical prostatectomy were also included. Immunostaining was performed using the following antibodies: ErbB3 C-ter (rabbit polyclonal sc-285, Santa Cruz Biotechnology 1/200), ErbB3 N-ter (Mouse monoclonal Ab-5 Clone H3.105.5, Neomarkers, 1/50), CCND1 (Rabbit monoclonal, clone EP12, DAKO, 1/50) and Ki67 (Mouse monoclonal, clone MIB-1, DAKO, 1/100). Negative controls were obtained by the use of an irrelevant antibody. Tissue from mantle zone lymphoma was used as positive control for Ki67 and CCND1. Slides were analysed by 2 observers in a blind fashion. For ErbB3C-ter, CCND1 and Ki67 staining, each tissue core was given a continuous value representing the percentage of tumour cells exhibiting nuclear staining. For each patient, the core with the highest percentage was selected.

Cell lines

Human prostatic RWPE cells (ATCC® CRL11609™), normal epithelial cells immortalized by HPV-18, were maintained in Keratinocyte Serum Free Medium supplemented with bovine pituitary extract (BPE, 0,05mg/ml) and human recombinant epidermal growth factor (EGF, 5ng/ml) (Invitrogen). Human prostatic LNCaP cell line (ATCC® CRL1740™) derived from lymph node and PC3 (ATCC® CRL1435™) cell line derived from bone metastasis, were grown in DMEM supplemented with 10% foetal calf serum and 1% penicillin/streptomycin.

Western blotting

Protein extracts and western blot analysis were performed as described in [17]. Nuclear and non-nuclear extracts were prepared as follows: cells were resuspended in Tris–HCl 10 mM pH7.5, NaCl 120 mM and lysed with NP40 to a final concentration of 0.3%. After centrifugation for 5 min at 2500 g, supernatants were supplemented with EDTA 1 mM, DTT 1 mM, SDS 0.1%, PIC 1/50 and PMSF 1/100 to constitute the non-nuclear fraction. The pellet corresponding to nuclei was washed three times in Tris–HCl 10 mM pH 7.5, NaCl 120 mM, NP40 0.3% to eliminate cytoplasmic residues, then lysed as above.

ChIP-on-chip

Cells were treated with 1% formaldehyde and chromatin extraction was performed with the Simple ChIP Enzymatic Chromatin IP kit (Cell Signaling Technology, CST) according to the manufacturer’s recommendation. The recovered chromatin was digested by Micrococcal nuclease to obtain chromatin fragments around 500pb size. At this step, a total DNA sample (Input) was collected for subsequent use. Immunoprecipitation was performed using anti ErbB3 C-ter antibodies (sc-285 or sc-7390, SCBT) or anti-H3 (positive control, Histone H3 antibody ChIP formulated #2650, CST) and for incubation with the non-relevant normal rabbit IgG (ChIP negative control). Total DNA and ChIP-DNA were amplified using the Whole Genome Amplification Kit (WGA-2, Sigma-Aldrich) and respectively labelled with Cy-3 or Cy-5 fluorochromes using BioPrime Array Genomic Labelling kit (Invitrogen). Labelled samples were hybridized on Human Promoters Microarrays (Agilent's Human Promoter ChIP-on-Chip Microarray Set 244K, P/N G4489A) according to the manufacturer’s recommendation.

Normalization and analysis of microarray data

Data provided by Agilent high resolution DNA Microarray scanner were analysed using Feature Extraction 10.1 Software (Agilent Technologies). The fluorescence spots that failed to pass quality control were excluded. Signals of enrichment ratios (ChIP/Input) related to merged replicates were calculated using CoCAS software [28]. ErbB3-enriched promoters from three independent experiments were isolated using the peak detection algorithm in CoCAS followed by the enrichment scores calculated by measuring the effective peak area.

Real-time quantitative PCR (qPCR) analysis

qPCR experiments were performed on ErbB3-immunoprecipitated or IgG-control DNA, using the GoTaq qPCR Master Mix (Promega) according to the manufacturer’s instructions on a 7500 Fast Real Time PCR system (Applied Biosystems). Relative enrichment was estimated after quantification of PCR signal by measuring the ratio IP ErbB3/IgG. List of ChIP-qPCR primers is given in Table 1.

thumbnail
Download:
Table 1. Primers sequences for ChIP-qPCR analysis.

https://doi.org/10.1371/journal.pone.0155950.t001

RNA isolation and RT-qPCR

LNCaP and PC3 cells were transfected with 20pmol of control or ErbB3-specific siRNAs (Eurogentec) before total RNA was extracted using RNAble (Eurobio). cDNA were synthetized from 2–6μg of total RNA and amplified by qPCR as described above. GAPDH was used as internal control to normalize gene expression. List of cDNA RT-qPCR primers is given in Table 2.

thumbnail
Download:
Table 2. Primers for RT-qPCR analysis.

https://doi.org/10.1371/journal.pone.0155950.t002

Results

ErbB3 expression and localization in normal and tumour prostate tissues

137 HS, 32 CRPC and 50 normally appearing tumour-adjacent tissues were analysed by immunohistochemistry (Fig 1). In all samples, staining for ErbB3 N-ter, when present, was located mainly at the membrane, and no nuclear staining was observed (Fig 1A). All the 169 samples displayed cytoplasmic staining with ErbB3 C-ter. (Fig 1B). In contrast, nuclear staining was observed with ErbB3 C-ter in 60 of the 169 (35%) PCa samples, but in none of the non-tumour samples. Nuclear ErbB3 expression was most often observed in CRPC (17/ 32, 53%) than in HS tumours (43/ 137, 31%) (p = 0.03, Chi 2 test). In addition, when present, the percentage of PCa expressing ErbB3 in the nucleus was higher in CRPC (median 10%, range 2–95%) than in HS cancers (median 2%, range 1–20%) (p = 0.007, Mann Whitney test). The median expression of CCND1 was 15% of HS prostate cancer cells (range 0–70%, not shown) and 20% of cancer cells in CRPC tissues (range 0–80%) (Fig 1D and 1G). The median proliferation rate assessed by Ki67 staining was 1% of PCa cells in HS tissues (range 0–10%, not shown) and 17.5% in CRPC samples (Fig 1E and 1H). In HS, no significant association was observed between nuclear ErbB3, CCND1 (p = 0.09) and proliferation rate (p = 0.1, Spearman test). In CRPC, the nuclear expression of ErbB3 was significantly associated with CCND1 expression, with a positive correlation (p = 0.01, Spearman test) (Fig 1C and 1D vs 1F and 1G). To resume, nuclear ErbB3 is highly and frequently expressed in CRPC and correlates with CCND1 expression, whereas no significant correlation was observed in HS tumours.

thumbnail
Download:
Fig 1. ErbB3 expression and localization in normal and tumour prostate tissues.

Staining with ErbB3 N-ter antibody in PCa tissues was located at the membrane, without nuclear expression (a). No ErbB3 C-ter nuclear staining was detected in “normal” prostate tissues (b). Nuclear ErbB3 expression was observed in PCa cells, and more often in CRPC advanced tumours when compared to clinically localized cancers. In CRPC, high nuclear staining was associated with high CCND1 expression and tended to be associated with higher proliferation (Ki67) (CRPC 1, c-e), whereas most tumours without nuclear staining displayed low CCND1 expression and tended to show lower proliferation (CRPC 2, f-h).

https://doi.org/10.1371/journal.pone.0155950.g001

ErbB3 expression and localization in prostate cell lines

We next assessed endogenous ErbB3 expression and localization in prostate cell lines using ErbB3 C-ter antibodies. By indirect immunofluorescence, ErbB3 staining was mainly present at the plasma membrane and in the cytoplasm of RWPE cells whereas a high nuclear staining was detected in tumour cell lines (Fig 2A). Western blot analysis of total protein extracts showed that the three cell lines expressed similar levels of the full-length ErbB3185kDa protein. An additional 80kDa protein was specifically detected in the hormone-sensitive (LNCaP) and hormone-resistant (PC3) tumour cell lines (Fig 2B). Next, western blotting was performed on non-nuclear vs nuclear protein fractions. The quality of the extracts was validated by the specific detection of the cytosolic PAK1 protein in the non-nuclear extracts and the C23 (nucleolin) protein in the nuclear extracts. As shown, ErbB3185kDa was mainly expressed in the non-nuclear fractions and was absent or slightly detectable in the nuclear fractions of all cell types. Instead, the 80kDa protein was almost exclusively detected in the nuclear fractions of LNCaP and PC3 cells (Fig 2C). These data suggest that the nuclear protein detected in PCa cells could mainly correspond to the ErbB380kDa nuclear isoform previously described in lung cancer cell lines [17].

thumbnail
Download:
Fig 2. Characterization of an endogenous nuclear ErbB3 variant in prostate cell lines.

(A) Both RWPE and PC3 cells display cytoplasmic and membrane ErbB3 localization. In addition, native PC3 cells exhibit a strong nuclear localization detected by the ErbB3 C-ter antibody (sc-285, Santa-cruz Biotechnology). (B) In total protein extracts, the ErbB3 C-ter antibody detects the full length ErbB3185kDa protein in all cell lines and the 80kDa protein in the LNCaP and PC3 tumour cell lines. (C) The full length ErbB3185kDa protein is almost only detected in the non-nuclear fractions that include plasma membrane and cytoplasmic proteins whereas a strong 80kDa signal is observed in the nuclear fractions of the tumour lines. (D) A 719 bp amplification product is detected in the three cell lines with the primers designed to amplify both the full length mRNA (NM_001982.3) and the AK300909.1 variant (PCR1), whereas PCR2 and PCR3 primers specific to AK124710.1 and AK1250281 respectively, failed to amplify any sequence. (E) Quantitative analysis with specific primers for ErbB3 transcripts indicate that NM_001982.3 and AK300909.1 are expressed at comparable levels in LNCaP and PC3 cell lines whereas AK300909.1 is slightly detectable in RWPE cells.

https://doi.org/10.1371/journal.pone.0155950.g002

Among the putative variants reported in public databases that could be transcribed from the ERBB3 gene, only three could correspond to the nuclear ErbB380kDa protein. They lack the 5’ sequence encoding the ligand-binding and transmembrane domains of the receptor and possess predicted nuclear localisation signals (NLS) [29]. To discriminate between those variants, we conducted RT-PCR with primers around the ATG starting codon of the mRNAs. In the three cell lines, a 719 bp PCR product was detected with the primers pair (PCR1) enabling the amplification of both the full length NM_001982.3 and the variant AK300909.1 transcripts that share the same nucleotide sequence. This experiment led us to eliminate AK124710.1 and AK1250281 and to focus on the AK300909.1 transcript (Fig 2D). This variant is predicted to encode a 699aa protein, whose sequence is strictly identical to the intracellular domain of ErbB3185kDa. To discriminate between NM_001982.3 and AK300909.1 expression in the cell lines, RT-qPCR was conducted using primers that amplify both transcripts or primers specific to the 5’ coding sequence of NM_001982.3 (Fig 2E). As shown, both transcripts were expressed at the same level in LNCaP and PC3 cells, whereas AK300909.1 was hardly detectable in RWPE cells (Fig 2E).

Thus, LNCaP and PC3 prostate tumour cell lines encode two different protein isoforms ErbB3185kDa and ErbB380kDa, corresponding to the NM_001982.3 and AK300909.1 mRNA transcripts respectively.

Genome-wide analysis of ErbB3-target promoters in prostate cancer cells

To identify ErbB3-target genes, we next performed ChIP-on-chip assays using ChIP-validated ErbB3 C-ter antibodies [17] according to the workflow described (Fig 3A). ErbB3-bound genomic regions were first investigated using the standalone CoCAS V2.4 software [28]. After pre-processing step, peak detection algorithm detected high probe signals corresponding to significant changes in Cy3/Cy5 ratio of enriched genomic regions. A total of 68% of ErbB3-target probes were located into the core promoters, while 30% were inside the genes and 2% downstream of the transcription start site (TSS). Significantly enriched probes were identified above a threshold adjusted to 0.998, leading to finally identify 1840 ErbB3-target promoters distributed as follows: 317 promoters in RWPE, 606 in LNCaP and 1297 in PC3 cells, among which 8 were found in the 3 cell lines and 361 were shared by the LNCaP and PC3 cell lines (Fig 3B). The data were further confirmed with the Agilent Workbench 7.0 software. The distribution of the signal along the promoters was used to design ChIP-qPCR primers in the regions that displayed the highest concentration of ErbB3-enriched probes with the highest log-ratio (Fig 3C). The number of target promoters was significantly higher in hormone-resistant PC3 cells than in hormone-sensitive LNCaP cells, suggesting the involvement of ErbB380kDa in PCa progression and aggressiveness, in accordance with the data obtained in patients (Fig 1). Gene lists derived from ChIP-on-chip experiments were further analysed with DAVID (Database for Annotation, Visualization and Integrated Discovery) [30]. Functional annotation confirmed the involvement of ErbB380kDa targets in cellular processes highly deregulated in tumour mechanisms (Fig 3D). A significant number of ErbB380kDa targets in PC3 or PC3 and LNCaP tumour cell lines were already known to be associated with PCa, or other solid tumours, suggesting that ErbB380kDa could regulate in a more general way the mechanisms of tumour progression (Fig 3D).

thumbnail
Download:
Fig 3. ChIP-on-chip workflow and Venn diagram.

(A) Chromatin immunoprecipitation was performed upon endogenous conditions. Nuclear lysates were prepared from native cells grown in complete medium and without any androgen. ChIP-DNA and total DNA (Input) labelled with Cy5 and Cy3 fluorophores respectively, were co-hybridized on Agilent promoters tilling arrays. (B) Data analysis with CoCAS software led to select 1840 ErbB3-target promoters in the three cell lines. (C) Agilent Worbench 7.0 software analysis illustrating the distribution of positive probes (ErbB3-bound DNA/red dots) and negative probes (Input/ green dots) throughout the genes sequence (X-axis). Signal intensity is shown on the Y-axis. CCND1 and MMP7 correspond to positive and negative controls, respectively. (D) Gene ontology analysis of ErbB3-targets with DAVID.

https://doi.org/10.1371/journal.pone.0155950.g003

ChIP-qPCR validation of nuclear ErbB3 target promoters in prostate cancer cells

Among the 1840 targets, 26 genes were selected for ChIP-qPCR validation (Fig 4). Four non-enriched promoters (ERBB3, EBP1, PSA and MMP7) were used as negative controls and displayed no significant amplification, whereas the CCND1 promoter displayed 7 to 12 enrichment fold in LNCaP and PC3 cells, respectively (Fig 4). The 3 LNCaP-specific targets exhibited significant enrichment fold in LNCaP cells and not in PC3 cells, whereas a strong and specific enrichment was observed for the 6 PC3-specific targets in the PC3 cell line (Fig 4A vs 4B). Enrichment fold observed for the shared promoters varied depending on the target and the cell line but was always higher in hormone-resistant PC3 cells than in LNCaP cells.

thumbnail
Download:
Fig 4. ChIP-qPCR validation.

On the basis of biocomputing data, 16 randomly selected ErbB3-target promoters from LNCaP cells (A) and 19 randomly selected ErbB3-target promoters from PC3 cells (B) were tested. All showed a significant ErbB3 enrichment compared to the negative control ChIP-IgG (value normalized to 1). The CCND1 promoter was used as a positive control whereas non-enriched promoters ERBB3, EBP1, PSA, and MMP7 were used as negative controls. The data are representative of 3 independent experiments and are the means of triplicate samples. *P<0.05, **P<0.01, ns = non-significant, Student’s t test.

https://doi.org/10.1371/journal.pone.0155950.g004

Transcriptional regulation of ErbB3-dependent target genes

Given the structure of the AK300909.1 transcript, no siRNAs specific to this variant can be designed. So, we addressed the transcriptional effects of ErbB380kDa by subtractive analysis: two types of siRNAs were used, targeting either the 5’ coding sequence of the NM_001982.3 ERBB3 transcript (siErbB3A) or the 3’ coding sequence of both NM_001982.3 and AK300909.1 ERBB3 transcripts (siErbB3B) as indicated on Fig 5A. The siRNAs were validated by RT-qPCR (not shown) and western blotting analysis: ErbB3185kDa expression was inhibited by both siRNAs, whereas ErbB380kDa expression was only decreased by siErbB3B in LNCaP and PC3 cell lines (Fig 5A). Under these conditions, we next analysed the expression of 15 ErbB380kDa-targets genes (Fig 5B). As expected, the expression of CCND1 was significantly reduced in LNCaP and PC3 cells transfected with siErbB3B vs siErbB3A, the latter displayed no difference with the control siRNA. The same applied for GATA2, RUNX2, MAP3K14 and ErbB2 in the two cell lines, and for ARID1A in PC3 cells. Conversely, siErbB3B induced higher expression for PPIG and MXI1 in the two cell lines and for ID3 and PC3-only targets in PC3 cells (Fig 5B). To simplify data analysis, the percentage of expression fold change that can be specifically attributed to the nuclear isoform was calculated with the formula (relative gene expression upon siErbB3B treatment)/(relative gene expression upon siErbB3A treatment) for each target gene (Fig 5C). As shown, ErbB380kDa transcriptionally activates CCND1, GATA2, RUNX2, MAP3K14 and inhibits PPIG, MXI1 in both LNCaP and PC3 cells, whereas it inhibits TBCC, SIN3A, ACE, ID3, RAD52, CXCR3 in PC3 cells only. Western blot analysis performed on ErbB380kDa-targets confirmed the transcriptional effects of the nuclear ErbB3 isoform at the protein level (S1 Fig).

thumbnail
Download:
Fig 5. Gene expression profiling.

(A) Two siRNAs were used to specifically inhibit the NM_001982.3 transcript (siErbB3A) or both the NM_001982.3 and AK300909.1 transcripts (siErbB3B) and validated by western blotting using the ErbB3 C-ter antibody. (B) LNCaP and PC3 cells were transiently transfected with siErbB3A or siErbB3B before mRNAs were reverse transcribed and amplified. RT-qPCR was performed on a selection of ErbB3-target genes. Expression fold change was normalised to control siRNA transfected samples. Transfections were done in triplicate, and the RT-qPCR results reported are from at least three independent experiments. (C) Expression fold change consecutive to the inhibition of the ErbB380kDa isoform in each cell line and for each target gene is calculated by the ratio (relative gene expression upon siErbB3B treatment)/(relative gene expression upon siErbB3A treatment). ErbB380kDa-dependent transcriptional activation of GATA2, RUNX2, MAP3K14, ErbB2 and CCND1 and inactivation of PPIG, MXI1 were similar in LNCaP and PC3 whereas ARID1A, TBCC, FYN, SIN3A, TOR3A, ACE, RAD52, CXCR3A appeared to be differentially regulated in the two cell lines. *P<0.05, **P<0.01, ns = non-significant, Student’s t test.

https://doi.org/10.1371/journal.pone.0155950.g005

The data presented here can be integrated to propose a model for nuclear ErbB380kDa functions in prostate cancer progression (Fig 6).

thumbnail
Download:
Fig 6. Proposed model for ErbB380kDa functions in prostate cancer progression.

Blocking RA-dependent signalling by androgen withdrawal therapy (AWT) first inhibits tumour growth, but soon alternative proliferative pathways are reactivated inducing tumour progression. Increased ErbB3 expression and subsequent reactivation of the PI3K pathway in response to AWT is a major mechanism of tumour resistance. Alternatively, we report here the nuclear accumulation of ErbB380kDa, a variant lacking the ligand-binding and transmembrane domain of ErbB3180kDa, in advanced PCa resistant to therapy (CRPC). As a co-regulator for the transcription of genes associated with cell proliferation, differentiation, apoptosis, oncogenesis, ErbB380kDa is likely to be strongly involved in therapy resistance and progression to fatal PCa. DHT: dihydrotestosterone; AR: androgen receptor; ARE: androgen responsive elements.

https://doi.org/10.1371/journal.pone.0155950.g006

Discussion

The goal of this study was to bring out molecular pathways involving ErbB3 in tumour progression and metastatic spreading, in order to characterize new potential therapeutic targets for prostate cancer. Recent therapies approved for recurrent prostate cancer delay metastatic progression and extend overall survival, but still fail to inhibit tumour progression to fatal CRPC. Inappropriate expression and/or activation of the ErbB family of tyrosine kinase receptors has been reported in advanced PCa, leading to consider ErbB1, ErbB2 and ErbB3 as potential diagnostic and prognostic markers [22,31]. Accordingly, increased ErbB3 expression has been reported in androgen-dependent PCa treated by androgen withdrawal and was associated with poor prognosis. This overexpression correlates with cell cycle progression and reactivation of the transcriptional activity of the AR in absence of androgens [26]. In addition, the nuclear compartmentalization of ErbB3 has been suggested to play a role in PCa progression but the mechanisms involved still remain unknown [18,27,32].

We first investigated the expression of ErbB3 on normal or tumour prostate tissues. Staining for ErbB3 N-ter was directed at the membrane, without nuclear expression. Using the ErbB3 C-ter antibody, the nuclear staining we observed was restricted to PCa cells and was increased in advanced castration-resistant prostate cancer when compared to localized tumours. These data strengthen the involvement the nuclear ErbB3 protein in PCa progression up to terminal castration-resistant stage [18]. As we failed to detect nuclear staining in CRPC with anti ErbB3 N-ter antibodies, we assumed that the nuclear ErbB3 protein could be a variant, on the model of the previously described ErBB380kDa or ErbB350kDa isoforms [16, 17]. ErbB3 protein expression assessed on normal and tumour prostate cell lines confirmed our hypothesis, with the full length receptor ErbB3185kDa mainly expressed in the non-nuclear protein fraction of all cell lines and an ErbB380kDa isoform mainly detected in the nuclear protein fraction of the tumour cell lines. To date, only the full length ErbB3185kDa protein was reported in the nucleus of prostate tumours and cell lines [15,19]. The mechanisms of nuclear localization analysed so far involved routing from endosomes and/or membrane cleavage following receptor activation [5–7, 32–36]. Since ErbB3185kDa is tyrosine kinase deficient, its phosphorylation occurs upon dimerization with other members of the ErbB family [37]. However, heregulin1ß treatment of PCa cell lines led to phosphorylate cytoplasmic but not nuclear ErbB3 protein [18]. Furthermore, nuclear translocation of ErbB1 or ErbB2 together with ErbB3, has never been reported in the nucleus of PCa cells, suggesting that the mechanisms underlying the nuclear localization of the full length ErbB3185kDa receptor may differ from those already described [5,6]. Alternatively, the nuclear ErbB3 protein detected in PCa cells would rather correspond to ErbB380kDa, on the same model as ErbB355kDa was specifically detected in the nucleus of normal Schwann cells [16].

ChIP-on-chip experiments performed to address the function of ErbB380kDa in PCa progression revealed that the number of ErbB380kDa target promoters was significantly higher in aggressive PC3 cells than in LNCaP hormone-sensitive cells, matching with the increased ErbB3 nuclear staining observed in castration resistant prostate tumours (Fig 1 and [18]). Genes set enrichment analysis also indicated that ErbB380kDa target genes were involved in a broad range of cellular processes such as transcription, proliferation, differentiation, apoptosis, and most of these genes are deregulated in neoplastic tissues. Interestingly, PC3-specific target genes were directly related to oncogenic processes in functional annotate on analysis, suggesting that these targets could be leading players in progression up to fatal PCa. Bioinformatics data were strengthened by the validation of 26 selected genes as effective ErbB380kDa-bound targets. To do so, we had to deal with the fact that CCND1 in lung and breast tumour cell lines, EZR (Ezrin) and HMGB1 in normal Schwann cells, were the only published nuclear targets of ErbB3 [16,17,19]. The CCND1 promoter we had previously identified as a transcriptional target of ErbB380kDa in the H358 lung carcinoma cell line, was found significantly enriched in LNCaP and PC3 cell lines and was considered further as positive control for ErbB380kDa binding [17]. On this basis, ChIP-qPCR analysis first indicated that the enrichment fold was promoter- and cell line- dependent. Indeed, no enrichment was found for the PC3-specific targets TBCC, FYN or ACE in LNCaP cells (Fig 4A), therefore no variation was observed at the expression level (Fig 5B). More, the amount of ErbB380kDa linked to the promoters was significantly higher in PC3 cells than in LNCaP. The latter point might be explained by the fact that ErbB380kDa doesn’t present any feature of a transcription factor. Thus, ErbB380kDa is more likely to bind the target promoters in an indirect manner. The recruiting factors may be different or differentially expressed between the LNCaP and PC3 cells. This could explain the discrepancies observed in the amount of ErbB380kDa on target promoters and also in the number of targets. Characterization of ErbB380kDa partners involved in the complex could offer interesting insights for the development of new therapies.

Targets characterized in this study can be classified into two categories whether they are already known to be involved in PCa or not. In this way, RUNX2, CXCR3 or GATA2 seem to be major players of progression and metastatic dissemination in PCa [38–43] whereas MXI1 has been recently identified as a new biomarker of aggressive disease [44]. In contrast, MAP3K14, also known as NF-ĸB-inducing kinase (or NIK) is a major player of the non-canonical NF-κB pathway mainly associated with lymphoid malignancies [45,46]. Interestingly, the non-canonical NF-κB involves specific membrane receptors among which the receptor activator for nuclear factor kB (RANK) [46].

Bone metastasis is the primary cause of mortality in patients with prostate cancer. Normal bone remodelling relies on a balance between bone resorption by osteoclasts and bone formation by osteoblasts. The RANK/RANKL pathway controls precursor’s differentiation into mature osteoclasts that promote bone resorption. Metastatic prostate tumour cells alter this balance by producing paracrine factors that stimulate neighbouring osteoclasts to resorb bone and in turn, bone resorption stimulates tumour cells growth [47]. Thus, increased MAP3K14 transcription by ErbB380kDa is strongly to be involved in bone resorption during prostate cancer progression, through constitutive activation of the non-canonical NF-ĸB signalling pathway.

These data strengthen the validity of our study and raise the question of the relevance of therapies in advanced PCa. At the time of personalized therapy, targeting ErbB3 could be attractive. Neutralizing antibodies specific for the extracellular domain of ErbB3 receptors are currently under evaluation in phase I or II clinical trials in various tumours except prostate. As prostate tumours develop resistance therapy through compensatory pathways regulating ErbB3 expression and localization, anti-ErbB3 therapeutic antibodies could be used for tumours displaying ErbB3185kDa overexpression and/or nuclear localization but not for tumour expressing ErbB380kDa [17, 25]. Indeed, as devoid of extracellular ligand-binding domain, ErbB380kDa is untargeted by ErbB3 therapeutic antibodies currently under Phase I or II clinical assays in various tumours, we have tested so far (unpublished data). The potential use of siRNA targeting specifically the ErbB380kDa isoform is not possible here as the full-length RNA would also be downregulated. Therefore, patients with ErbB380kDa positive tumours should be proposed more appropriate treatments aiming at inactivate the downstream targets rather than the nuclear ErbB380kDa protein. In this context, recently described markers such as GATA2, a poor prognosis factor associated with early relapse, could be of interest [42]. Recently, nanoparticles carrying siRNA targeting GATA2 significantly reduced tumour growth in a xenograft murine model of NSCLC harbouring oncogenic KRAS mutations [48].These promising results require further exploration. Targeting the non-canonical NF-kB could also be explored. Inhibition of the RANK-RANKL axis in association with the bone morphogenetic protein (BMP) activity has yet been reported in a murine model of lytic prostate cancer lesion in bone. The combine treatment led to inhibit osteoclast-dependent bone resorption and to reduce bone lesion formation [49].

Therefore, multitargeted approaches aiming at inhibit several ErbB380kDa-regulated factors according to the pattern of the tumour, could offer interesting individual perspectives for patients.

Supporting Information

S1 Fig. Transcriptional effects of the nuclear ErbB380kDa isoform at the protein level.

PC3 cells were transiently transfected with si control, siErbB3A or siErbB3B 48h before cytosolic (C) and nuclear (N) extraction. Western blotting was performed using antibodies to PPIG (sc-100699, SCBT), GATA2 (sc-9008, SCBT), RUNX2 (sc-10758, SCBT), CCND1 (#2922, CST) or histone H3 (#2650, CST) as a control for nuclear extraction.

https://doi.org/10.1371/journal.pone.0155950.s001

(PPTX)

Author Contributions

Conceived and designed the experiments: PS MEM GF. Performed the experiments: MEM ML AB PS GF OK. Analyzed the data: MEM ML AB PS GF BV AM OK. Contributed reagents/materials/analysis tools: PS GF. Wrote the paper: PS GF AM.

References

  1. 1. Yarden Y, Pines G. The ERBB network: at last, cancer therapy meets systems biology. Nat Rev Cancer. 2012 Aug;12(8):553–63. pmid:22785351
  2. 2. Roskoski R. The ErbB/HER family of protein-tyrosine kinases and cancer. Pharmacol Res Off J Ital Pharmacol Soc. 2014 Jan;79:34–74.
  3. 3. Carpenter G. Nuclear localization and possible functions of receptor tyrosine kinases. Curr Opin Cell Biol. 2003 Apr;15(2):143–8. pmid:12648669
  4. 4. Linggi B, Carpenter G. ErbB receptors: new insights on mechanisms and biology. Trends Cell Biol. 2006 Dec;16(12):649–56. pmid:17085050
  5. 5. Carpenter G, Liao H-J. Trafficking of receptor tyrosine kinases to the nucleus. Exp Cell Res. 2009 May 15;315(9):1556–66. pmid:18951890
  6. 6. Wang S-C, Hung M-C. Nuclear translocation of the epidermal growth factor receptor family membrane tyrosine kinase receptors. Clin Cancer Res Off J Am Assoc Cancer Res. 2009 Nov 1;15(21):6484–9.
  7. 7. Wang Y-N, Hung M-C. Nuclear functions and subcellular trafficking mechanisms of the epidermal growth factor receptor family. Cell Biosci. 2012;2(1):13. pmid:22520625
  8. 8. Mills IG. Nuclear translocation and functions of growth factor receptors. Semin Cell Dev Biol. 2012 Apr;23(2):165–71. pmid:21930229
  9. 9. Lin SY, Makino K, Xia W, Matin A, Wen Y, Kwong KY, et al. Nuclear localization of EGF receptor and its potential new role as a transcription factor. Nat Cell Biol. 2001 Sep;3(9):802–8. pmid:11533659
  10. 10. Lo H- W, Hsu S- C, Ali-Seyed M, Gunduz M, Xia W, Wei Y, et al. Nuclear interaction of EGFR and STAT3 in the activation of the iNOS/NO pathway. Cancer Cell. 2005 Jun;7(6):575–89. pmid:15950906
  11. 11. Hanada N, Lo H-W, Day C-P, Pan Y, Nakajima Y, Hung M- C. Co-regulation of B-Myb expression by E2F1 and EGF receptor. Mol Carcinog. 2006 Jan;45(1):10–7. pmid:16299810
  12. 12. Wang S-C, Lien H-C, Xia W, Chen I-F, Lo H-W, Wang Z, et al. Binding at and transactivation of the COX-2 promoter by nuclear tyrosine kinase receptor ErbB-2. Cancer Cell. 2004 Sep;6(3):251–61. pmid:15380516
  13. 13. Béguelin W, Díaz Flaqué MC, Proietti CJ, Cayrol F, Rivas MA, Tkach M, et al. Progesterone receptor induces ErbB-2 nuclear translocation to promote breast cancer growth via a novel transcriptional effect: ErbB-2 function as a coactivator of Stat3. Mol Cell Biol. 2010 Dec;30(23):5456–72. pmid:20876300
  14. 14. Han W, Carpenter RL, Cao X, Lo H-W. STAT1 gene expression is enhanced by nuclear EGFR and HER2 via cooperation with STAT3. Mol Carcinog. 2013 Dec;52(12):959–69. pmid:22693070
  15. 15. Offterdinger M, Schöfer C, Weipoltshammer K, Grunt TW. c-erbB-3: a nuclear protein in mammary epithelial cells. J Cell Biol. 2002 Jun 10;157(6):929–39. pmid:12045181
  16. 16. Adilakshmi T, Ness-Myers J, Madrid-Aliste C, Fiser A, Tapinos N. A nuclear variant of ErbB3 receptor tyrosine kinase regulates ezrin distribution and Schwann cell myelination. J Neurosci Off J Soc Neurosci. 2011 Mar 30;31(13):5106–19.
  17. 17. Andrique L, Fauvin D, El Maassarani M, Colasson H, Vannier B, Séité P. ErbB3(80 kDa), a nuclear variant of the ErbB3 receptor, binds to the Cyclin D1 promoter to activate cell proliferation but is negatively controlled by p14ARF. Cell Signal. 2012 May;24(5):1074–85. pmid:22261253
  18. 18. Koumakpayi IH, Diallo J-S, Le Page C, Lessard L, Gleave M, Bégin LR, et al. Expression and nuclear localization of ErbB3 in prostate cancer. Clin Cancer Res Off J Am Assoc Cancer Res. 2006 May 1;12(9):2730–7.
  19. 19. Brand TM, Iida M, Luthar N, Wleklinski MJ, Starr MM, Wheeler DL. Mapping C-terminal transactivation domains of the nuclear HER family receptor tyrosine kinase HER3. PloS One. 2013 Jan;8(8):e71518. pmid:23951180
  20. 20. Scher HI, Sawyers CL. Biology of progressive, castration-resistant prostate cancer: directed therapies targeting the androgen-receptor signaling axis. J Clin Oncol Off J Am Soc Clin Oncol. 2005 Nov 10;23(32):8253–61.
  21. 21. Mellinghoff IK, Vivanco I, Kwon A, Tran C, Wongvipat J, Sawyers CL. HER2/neu kinase-dependent modulation of androgen receptor function through effects on DNA binding and stability. Cancer Cell. 2004 Nov;6(5):517–27. pmid:15542435
  22. 22. Grasso AW, Wen D, Miller CM, Rhim JS, Pretlow TG, Kung HJ. ErbB kinases and NDF signaling in human prostate cancer cells. Oncogene. 1997 Nov 27;15(22):2705–16. pmid:9400997
  23. 23. Signoretti S, Montironi R, Manola J, Altimari A, Tam C, Bubley G, et al. Her-2-neu expression and progression toward androgen independence in human prostate cancer. J Natl Cancer Inst. 2000 Dec 6;92(23):1918–25. pmid:11106683
  24. 24. Osman I, Scher HI, Drobnjak M, Verbel D, Morris M, Agus D, et al. HER-2/neu (p185neu) protein expression in the natural or treated history of prostate cancer. Clin Cancer Res Off J Am Assoc Cancer Res. 2001 Sep;7(9):2643–7.
  25. 25. Shah RB, Ghosh D, Elder JT. Epidermal growth factor receptor (ErbB1) expression in prostate cancer progression: correlation with androgen independence. The Prostate. 2006 Sep 15;66(13):1437–44. pmid:16741920
  26. 26. Chen L, Siddiqui S, Bose S, Mooso B, Asuncion A, Bedolla RG, et al. Nrdp1-mediated regulation of ErbB3 expression by the androgen receptor in androgen-dependent but not castrate-resistant prostate cancer cells. Cancer Res. 2010 Jul 15;70(14):5994–6003. pmid:20587519
  27. 27. Koumakpayi IH, Diallo J-S, Le Page C, Lessard L, Filali-Mouhim A, Bégin LR, et al. Low nuclear ErbB3 predicts biochemical recurrence in patients with prostate cancer. BJU Int. 2007 Aug;100(2):303–9. pmid:17532856
  28. 28. Benoukraf T, Cauchy P, Fenouil R, Jeanniard A, Koch F, Jaeger S, et al. CoCAS: a ChIP-on-chip analysis suite. Bioinforma Oxf Engl. 2009 Apr 1;25(7):954–5.
  29. 29. Hsu S- C, Hung M- C. Characterization of a Novel Tripartite Nuclear Localization Sequence in the EGFR Family. J Biol Chem. 2007 Apr 6;282(14):10432–40. pmid:17283074
  30. 30. Dennis G, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, et al. DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol. 2003;4(5):P3. pmid:12734009
  31. 31. Gross ME, Jo S, Agus DB. Update on HER-kinase-directed therapy in prostate cancer. Clin Adv Hematol Oncol HO. 2004 Jan;2(1):53–6, 64.
  32. 32. Cheng C-J, Ye X, Vakar-Lopez F, Kim J, Tu S-M, Chen D-T, et al. Bone microenvironment and androgen status modulate subcellular localization of ErbB3 in prostate cancer cells. Mol Cancer Res MCR. 2007 Jul;5(7):675–84. pmid:17634423
  33. 33. Scaltriti M, Rojo F, Ocaña A, Anido J, Guzman M, Cortes J, et al. Expression of p95HER2, a truncated form of the HER2 receptor, and response to anti-HER2 therapies in breast cancer. J Natl Cancer Inst. 2007 Apr 18;99(8):628–38. pmid:17440164
  34. 34. Xia W, Liu Z, Zong R, Liu L, Zhao S, Bacus SS, et al. Truncated ErbB2 expressed in tumor cell nuclei contributes to acquired therapeutic resistance to ErbB2 kinase inhibitors. Mol Cancer Ther. 2011 Aug;10(8):1367–74. pmid:21673090
  35. 35. Zhou W, Carpenter G. Heregulin-dependent trafficking and cleavage of ErbB-4. J Biol Chem. 2000 Nov 3;275(44):34737–43. pmid:10944525
  36. 36. Ni CY, Murphy MP, Golde TE, Carpenter G. gamma -Secretase cleavage and nuclear localization of ErbB-4 receptor tyrosine kinase. Science. 2001 Dec 7;294(5549):2179–81. pmid:11679632
  37. 37. Pinkas-Kramarski R, Soussan L, Waterman H, Levkowitz G, Alroy I, Klapper L, et al. Diversification of Neu differentiation factor and epidermal growth factor signaling by combinatorial receptor interactions. EMBO J. 1996 May 15;15(10):2452–67. pmid:8665853
  38. 38. Akech J, Wixted JJ, Bedard K, van der Deen M, Hussain S, Guise TA, et al. Runx2 association with progression of prostate cancer in patients: mechanisms mediating bone osteolysis and osteoblastic metastatic lesions. Oncogene. 2010 Feb 11;29(6):811–21. pmid:19915614
  39. 39. Wu Q, Dhir R, Wells A. Altered CXCR3 isoform expression regulates prostate cancer cell migration and invasion. Mol Cancer. 2012;11:3. pmid:22236567
  40. 40. Shah S, Prasad S, Knudsen KE. Targeting pioneering factor and hormone receptor cooperative pathways to suppress tumor progression. Cancer Res. 2012 Mar 1;72(5):1248–59. pmid:22258452
  41. 41. Wu D, Sunkel B, Chen Z, Liu X, Ye Z, Li Q, et al. Three-tiered role of the pioneer factor GATA2 in promoting androgen-dependent gene expression in prostate cancer. Nucleic Acids Res. 2014 Apr;42(6):3607–22. pmid:24423874
  42. 42. Chiang YT, Wang K, Fazli L, Qi RZ, Gleave ME, Collins CC, et al. GATA2 as a potential metastasis-driving gene in prostate cancer. Oncotarget. 2014 Jan 30;5(2):451–61. pmid:24448395
  43. 43. Plymate SR, Bhatt RS, Balk SP. Taxane resistance in prostate cancer mediated by AR-independent GATA2 regulation of IGF2. Cancer Cell. 2015 Feb 9;27(2):158–9. pmid:25670076
  44. 44. Long Q, Xu J, Osunkoya AO, Sannigrahi S, Johnson BA, Zhou W, et al. Global transcriptome analysis of formalin-fixed prostate cancer specimens identifies biomarkers of disease recurrence. Cancer Res. 2014 Jun 15;74(12):3228–37. pmid:24713434
  45. 45. Staudt LM. Oncogenic activation of NF-kappaB. Cold Spring Harb Perspect Biol. 2010 Jun;2(6):a000109. pmid:20516126
  46. 46. Sun S- C. Non-canonical NF-κB signaling pathway. Cell Res. 2011 Jan;21(1):71–85. pmid:21173796
  47. 47. Rucci N, Angelucci A. Prostate cancer and bone: the elective affinities. BioMed Res Int. 2014;2014:167035. pmid:24971315
  48. 48. Shen S, Mao C-Q, Yang X-Z, Du X-J, Liu Y, Zhu Y-H, et al. Cationic lipid-assisted polymeric nanoparticle mediated GATA2 siRNA delivery for synthetic lethal therapy of KRAS mutant non-small-cell lung carcinoma. Mol Pharm. 2014 Aug 4;11(8):2612–22. pmid:24521262
  49. 49. Virk MS, Alaee F, Petrigliano FA, Sugiyama O, Chatziioannou AF, Stout D, et al. Combined inhibition of the BMP pathway and the RANK-RANKL axis in a mixed lytic/blastic prostate cancer lesion. Bone. 2011 Mar 1;48(3):578–87. pmid:21073986