Cancer Treatment Reviews 40 (2014) 739–749

Contents lists available at ScienceDirect

Cancer Treatment Reviews journal homepage: www.elsevierhealth.com/journals/ctrv

Laboratory-Clinic Interface

Retinoids and breast cancer: From basic studies to the clinic and back again Enrico Garattini a,⇑, Marco Bolis a, Silvio Ken Garattini a, Maddalena Fratelli a, Floriana Centritto a, Gabriela Paroni a, Maurizio Gianni’ a, Adriana Zanetti a, Anna Pagani b, James Neil Fisher a, Alberto Zambelli b, Mineko Terao a a b

Laboratory of Molecular Biology, IRCCS-Istituto di Ricerche Farmacologiche ‘‘Mario Negri’’, via La Masa 19, 20156 Milano, Italy Laboratory of Experimental Oncology and Pharmacogenomics, IRCCS-Fondazione S. Maugeri, Università degli Studi di Pavia, via Maugeri 10, 27100 Pavia, Italy

a r t i c l e

i n f o

Article history: Received 9 December 2013 Received in revised form 9 January 2014 Accepted 10 January 2014

Keywords: Retinoic acid Breast cancer RAR/RXR Gene pathways Chemo-prevention Treatment

a b s t r a c t All-trans retinoic acid (ATRA) is the most important active metabolite of vitamin A controlling segmentation in the developing organism and the homeostasis of various tissues in the adult. ATRA as well as natural and synthetic derivatives, collectively known as retinoids, are also promising agents in the treatment and chemoprevention of different types of neoplasia including breast cancer. The major aim of the present article is to review the basic knowledge acquired on the anti-tumor activity of classic retinoids, like ATRA, in mammary tumors, focusing on the underlying cellular and molecular mechanisms and the determinants of retinoid sensitivity/resistance. In the first part, an analysis of the large number of preclinical studies available is provided, stressing the point that this has resulted in a limited number of clinical trials. This is followed by an overview of the knowledge acquired on the role played by the retinoid nuclear receptors in the anti-tumor responses triggered by retinoids. The body of the article emphasizes the potential of ATRA and derivatives in modulating and in being influenced by some of the most relevant cellular pathways involved in the growth and progression of breast cancer. We review the studies centering on the cross-talk between retinoids and some of the growth-factor pathways which control the homeostasis of the mammary tumor cell. In addition, we consider the cross-talk with relevant intracellular second messenger pathways. The information provided lays the foundation for the development of rational and retinoid-based therapeutic strategies to be used for the management of breast cancer. Ó 2014 Elsevier Ltd. All rights reserved.

Introduction ATRA (All-trans-retinoic-acid) and 13-cisRA (13-cis-retinoic-acid) are the active metabolites of vitamin A, while the physiological significance of the other retinoic acid isomer, 9-cisRA (9-cisretinoic-acid), is debated [1]. ATRA is the first clinically useful cytodifferentiating agent, being employed in the treatment of acute promyelocytic leukemia (APL) [2]. There is interest in expanding the therapeutic use of ATRA and derivatives to breast cancer. The biological activity of classic retinoids (ATRA, 13-cisRA) is primarily mediated by nuclear retinoid-receptors, which are ligand-activated

⇑ Corresponding author. Tel.: +39 02 39014533. E-mail addresses: [email protected] (E. Garattini), [email protected] (M. Bolis), [email protected] (S.K. Garattini), [email protected] (M. Fratelli), fl[email protected] (F. Centritto), [email protected] (G. Paroni), [email protected] (M. Gianni’), [email protected] (A. Zanetti), [email protected] (A. Pagani), jamesneil.fi[email protected] (J.N. Fisher), [email protected] (A. Zambelli), [email protected] (M. Terao). 0305-7372/$ - see front matter Ó 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ctrv.2014.01.001

transcription factors. Retinoid receptors are divided into RARs (RARa, RARb and RARc) and RXRs (RXRa, RXRb and RXRc) [3], which are encoded by distinct loci producing alternative-splicing variants (Fig. 1). Active retinoid-receptors consist of RAR/RXR heterodimers, which bind to Retinoic-Acid-Responsive-Elements (RAREs) in retinoid-responsive genes [3]. ATRA and 13-cisRA are pan-RAR agonists activating all RAR-isoforms with similar efficiency. 9-cisRA binds both RARs and RXRs, although the mechanisms underlying the pharmacological/anti-tumor activity of the retinoid may stem from the activation of RAR/RXR, VDR/RXR and the elusive RXR/RXR homodimers [4–6]. RARs and RXRs are not the only nuclear receptors binding ATRA, as PPARb/d (Fig. 1) is also bound and activated by the retinoid [7,8]. Partitioning of ATRA between RARs and PPARb/d is controlled by the cytosolic retinoid-binding proteins, CRABP2 and FABP5, delivering ATRA to RARs and PPARb/d, respectively [8]. We intend to provide an overview of the anti-tumor activity of retinoids in breast cancer, concentrating on the cellular/molecular determinants and the underlying mechanisms of action. The main focus of our analysis is ATRA, the prototype of classic retinoids.

740

E. Garattini et al. / Cancer Treatment Reviews 40 (2014) 739–749

Fig. 1. Structure of the genes, mRNAs and proteins of the various retinoid-binding nuclear receptors. The figure illustrates the exonic structure of the human genes encoding RARa (RARA), RARb (RARB), RARc (RARG), RXRa (RXRA), RXRb (RXRB), RXRc (RXRG) and PPARb/d (PPPARD) along with the chromosomal location of each gene. Underneath each gene the structures of the known RAR, RXR and PPARb/d mRNAs resulting from differential splicing events are shown on the left side. The NCBI accession number of each splicing variant is indicated in parenthesis. It must be noticed that the transcript variant indicated as RARb1 is often referred to as RARb2 in the old literature. In addition, the accession No. of RARb5 is provisional, as the evidence for its existence is still incomplete. The vertical red lines indicate the position of the first ATG and the STOP codons. On the right, the structure of the encoded proteins is shown. The various domains constituting each protein are indicated with different colors. The NCBI accession number of each protein is indicated in parenthesis. A/B = N-terminal ligand-independent transactivating domains; C = DNA binding domain; D = hinge domain; E = ligand-binding domain; F = C-terminal domain of unknown function.

E. Garattini et al. / Cancer Treatment Reviews 40 (2014) 739–749

Retinoids and breast cancer: many pre-clinical studies and few clinical trials Interest in the use of retinoids in breast cancer is reflected by the vast scientific literature (>1300 articles) available (Fig. 2A). The large number of pre-clinical studies translated into few clinical trials (Fig. 2B and Table 1). Except for one report conducted with bexarotene (RXR agonist), all the known chemo-preventive trials involve fenretinide, which is not a bona fide retinoid. In invasive breast cancer, three therapeutic trials on retinoids, used as single agents, have been reported. The only ATRA-based trial is a phase-II study in pre-treated patients which failed to achieve the primary end-point [9]. Seven clinical trials using retinoid-containing combinations are available. ATRA+tamoxifen is the object of a dose-escalation phase I/II study conducted in patients with ER+ hormonerefractory tumors, which resulted in 9 objective responses or stable disease (SD) [10]. A pre-operative study in locally advanced breast cancer was conducted to ascertain the biologic effects and the minimal-effective-dose of ATRA with/without tamoxifen and IFNa2. Neither tamoxifen nor tamoxifen+IFNa2 potentiated ATRA [11]. A small pilot study was conducted to evaluate combinations of ATRA and paclitaxel in pre-treated metastatic breast cancer. In 17 evaluable patients, 3 showed partial remission (PR) and 10 presented with SD. The data suggest that this well-tolerated combination induces a modest frequency of PR but relatively high rates of SD [12]. Retinyl-palmitate, 9-cis-RA and 13-cis-RA [13] were also tested in combination with other agents. 13-cisRA+tamoxifen and

Fig. 2. Analysis of the available scientific literature. (A) The panel illustrates the temporal trend of the scientific articles published on retinoids in breast cancer from 1975 to 2013. The graph summarizes the results obtained after crossing the two keywords ‘‘Retinoic’’ and ‘‘Breast Cancer’’ in PUBMED ( Total). The figure indicates that the number of studies involving ATRA ( ATRA) exceeds by far those centering on 9-cisRA ( 9-cisRA) and 13-cisRA ( 13-cisRA). (B) The panel shows the number of clinical trials published with the indicated retinoids.

741

13-cisRA+IFNa2a were investigated in post-menopausal pretreated metastatic breast cancer. No difference in overall responses and overall survival was observed in the various treatment arms. These data stress the paucity and the disappointing nature of the clinical results. We surmise that the negativity of the results is predominantly due to the design of the clinical trials, which did not take into account heterogeneity and were conducted without selection for any particular sub-type of breast cancer. This highlights the importance of defining: (a) the molecular determinants of retinoid-sensitivity in distinct breast cancer sub-types; (b) the retinoid-receptor pathways involved in the anti-tumor activity of ATRA; (c) the intracellular pathways influenced by and influencing the activity of retinoids, as it is unlikely that retinoid monotherapy will ever represent a viable option [14]. In fact, scattered observations suggest that a number of factors contribute to natural or induced resistance to ATRA, however these factors have never been the object of systematic studies in breast cancer sub-types. For instance, high levels of CYP26 (cytochrome-P450isoform-26) causing metabolic inactivation of ATRA are proposed as determinants of ATRA resistance [15,16] in different neoplasias [17] and epigenetic silencing of RARb is also deemed to cause breast tumor resistance to the retinoid [18,19]. A pre-requisite to address points (a–c) is a thorough analysis of the available studies, which is the object of the next chapters. Retinoid receptors and retinoid sensitivity In many breast cancer cells, RARa, RARc, RXRa and RXRb [20] are expressed in basal conditions, while RARb is undetectable [21]. In ER+ cells, which are considered to be sensitive to retinoids, RARa may be the primary determinant of ATRA sensitivity. Indeed, RARa agonists inhibit the growth of ER+ cell lines, while silencing of RARa reduces the anti-proliferative effect of ATRA [22]. Although the retinoid receptor is slightly up-regulated in the neoplastic tissue relative to the normal counterpart, RARa levels are higher in ER+ than in ER- tumors (Fig. 3) consistent with the predicted retinoid sensitivity [22]. High expression of RARa in ER+ tumors [23] is likely due to ERa-dependent induction of RARA transcription via an Estrogen-Responsive-Element [24]. As for the mechanisms underlying the action of the retinoid-receptor in ER+ cells [25,26], unliganded RARa interacts with ERa and functions as a co-activator in the regulation of estrogen-dependent genes, paradoxically inducing proliferative effects [27]. In contrast, ATRA-bound RARa inhibits this interaction and blocks the estrogen-dependent activity of ERa. Although the results of these studies are somehow conflicting, ATRA is endowed with anti-estrogenic properties, which may explain why ER+ cells are generally sensitive to the retinoid. RARa expression may be relevant also in the context of ER-negativity. In fact as indicated by a subset of ER /HER2+ breast tumors characterized by RARA amplification is retinoid-sensitive [21]. Furthermore, RARa overexpression in ER- and ATRA-resistant MDA-MB231 cells restores ATRA sensitivity [4]. In summary, the available evidence indicates that RARa may represent a predictor of ATRA-sensitivity, despite the observed association with tamoxifen-resistance in ER+ breast tumors [28]. The relative importance of RARa1 and RARa2, the two major RARa isoforms (Fig. 1), in ATRA-sensitivity must be defined. Significantly, RARa1 knock-out animals show an increase in normal mammary gland stem-cells, suggesting that activation of the receptor has a negative impact on stem-cell renewal, which may contribute to the anti-tumor activity of ATRA [29]. Clarification of the RARa1 role in breast cancer inhibition by ATRA is a priority, as RARa2 is, instead, the crucial mediator of ATRA anti-proliferative activity in myeloma [30]. RARb is also suggested to be a determinant of ATRA-sensitivity [31], although its expression in tumor-associated stromal cells facilitates the progression of mouse breast tumors [32]. The

742

E. Garattini et al. / Cancer Treatment Reviews 40 (2014) 739–749

Table 1 Selected clinical trials of retinoids in breast cancer. Compound

End point

Trials (No.)

Clinical Phase

Reference

ATRA

BC prevention BC treatment

0 4

Ph I/II

[9–12]

9-cis-RA

BC prevention BC treatment

0 2

Ph I Ph I

[118] [119]

13-cis-RA

BC prevention BC treatment

0 1

Ph II

[13]

Bexarotene

BC prevention BC treatment

1 1

Ph I Ph II

[120] [121]

Retinyl palmitate

BC prevention BC treatment

0 1

Ph II

[122]

The table lists the published clinical trials on retinoids in breast cancer along with the corresponding references.

Fig. 3. Expression of RARs/RXRs and PPARb/d in ER+ and ER breast tumors. The figure illustrates the box plots of the expression levels of the mRNAs encoding the indicated RAR/RXR forms and PPARb/d in ER+ and ER breast cancers and normal mammary tissue. The results were obtained after analysis of the level 3 microarray expression data obtained with the Agilent platform derived from the TCGA portal (http://tcga-data.nci.nih.gov) (780 patients, 601 ER+ cases and 179 ER cases; 63 normal tissue samples, N; 531 primary tumors, T). However similar data were obtained from a number of other publicly available datasets. Only significant p values are indicated in red (p < 0.001). The results indicate that RARa shows significantly higher expression levels in the ER+ relative to the ER tumors, while RARb and PPARb/d show the opposite expression profile. In the comparison between normal and primary cancer samples, RARa, RARc and PPARb/d mRNA levels are significantly higher in the tumor tissue. An opposite effect is observed in the case of RARb and RXRc.

receptor is detectable in normal mammary tissue, while it is often absent in breast tumors (Fig. 3). ATRA-dependent induction of RARb correlates with growth-inhibition [33]. ER cells acquire ATRA sensitivity upon over-expression of RARb and the ATRAresponsiveness of ER+ cells is inhibited by RARb-selective antagonists/antisense-RNAs. However, the expression of RARb mRNA is significantly higher in ER relative to ER+ tumors. The presence of various RARb isoforms complicates the picture, as silencing of

RARb2/RARb5 in primary cell cultures causes sensitization to ATRA [34]. This suggests that activation of RARb1 and RARb2/RARb5 (Fig. 1) exert opposite effects and the relative expression of the receptors may contribute to retinoid sensitivity/resistance [35]. Being induced by ATRA, RARB is a direct retinoid-responsive gene. Thus, a signal-relay involving RARa and RARb may underly the anti-tumor action of ATRA, as suggested by ATRA-dependent induction of the growth inhibitory factor, IGFBP3, which requires

E. Garattini et al. / Cancer Treatment Reviews 40 (2014) 739–749

successive activation of the two receptors [36]. From a therapeutic perspective, confirmation of this signal-relay would support the use of ATRA, a pan-RAR agonist, instead of selective RARa/RARb agonists. RARc is part of a gene-signature associated with mammary tumors [37] and it provides a pro-oncogenic signal in a c-Myc-driven transgenic model of mammary cancer [38,39]. In addition, RARc favors the self-renewal and expansion of hematopoietic stem cells [40] and may exert a similar effect in the breast cancer counterparts [29,38]. These observations indicate that activation of RARc supports the growth/progression of mammary tumors with implications for the therapeutic use of ATRA. In fact, ATRA, being a panRAR agonist, is expected to cause simultaneous activation of RARc and RARa which should decrease the anti-tumor efficacy of the retinoid. Consistent with this, RARa agonists are often more effective than ATRA in retinoid-responsive cells [21,41]. PPARb/d is also a determinant of retinoid sensitivity/resistance in breast and other solid tumors [42]. Activation of the receptor by ATRA is associated with a carcinogenic effect that opposes the RAR-mediated anti-tumor activity of the retinoid [43]. Interestingly, PPARb/d mRNA levels are higher in ER than in ER+ tumors (Fig. 3). In addition the transcript is over-represented in cancerous relative to normal tissue (Fig. 3). Five potential PPARb/d forms arising from differential splicing of the same gene are known (Fig. 1), although none of them has been specifically involved in the tumorigenic action of ATRA. As binding/activation of PPARb/d and RARs by ATRA is controlled by FABP5 and CRABP2, respectively, disregulation of these cytosolic proteins may underly retinoid sensitivity/ resistance in different breast cancer subtypes [14]. Interestingly, FABP5 is upregulated in human breast cancers and ectopic expression of the protein augments the ability of PPARb/d to enhance cell proliferation, migration, and invasion [14]. Conversely, FABP5 deletion suppresses the development of HER2-dependent mouse mammary tumors probably via down-regulation of PPARb/d activity [14]. In the MMTV-neu and ATRA-resistant mouse model of breast cancer, the retinoid activates PPARb/d. This is due to an aberrantly high intratumor FABP5/CRABP-II ratio [43]. Decreasing this ratio diverts ATRA from PPARb/d to RAR suppressing tumor growth: This suggests that high FABP5/CRABP-II ratios and significant activation of PPARb/d by ATRA may underly retinoid-resistance in certain breast cancer subtypes.

Retinoids and the EGF/TGF/IGF pathways The most intensely studied growth factors in breast cancer are the epidermal-growth-factor (EGF) family members. EGF and EGF-like proteins bind to the ERBB/HER receptors (ERBB1-4 or HER1-4). The gene-amplification of ERBB2 defines HER2+ breast tumors. ATRA influences the EGF pathway and EGF modulates the RAR/RXR activity. In ER+ T47D cells, ATRA reduces the duration of EGF-induced ERBB1 phosphorylation via PKCa activation [44]. In ER-/HER2+ SKBR3 cells, showing co-amplification of the ERBB2/ RARA loci [21], ATRA decreases HER2 phosphorylation and prolongs its inhibition by the ERBB1/ERBB2 tyrosine kinase inhibitor, lapatinib, via RARRES3 (Retinoic-Acid-Receptor-ResponderProtein-3). These effects underly the anti-tumor synergism between ATRA and lapatinib [21]. In ER+ MCF-7 and SKBR3 cells, EGF induces the anti-apototic factor, Gene33, and the induction is blocked by ATRA [45]. ERBB1 inhibitors induce RARb in the triple-negative MDA-MB-468 cells [46] and this represents the only evidence for the influence of EGF on the RAR pathway. Interestingly, FABP5 is directly controlled and induced by ERBB1 activation with heregulin [47]. Thus, activation of ERBB1 shifts ATRA signaling towards PPARb/d and may induce ATRA-resistance. Classic retinoids control the morphogenesis of the mammary gland which is

743

also regulated by ERBB ligands. Thus, interference of ATRA with the EGF pathway may affect not only growth, but also other aspects of breast cancer homeostasis [48]. The ERBB ligand, NRG1 (Neuregulin-1) cooperates with retinoids in inducing differentiation. In malignant cells, NRG1 synergizes with RARa/RXR agonists and stimulates ductal branching, down-regulating ERBB2, ERBB3 and ERBB4 [48]. In conclusion, contemporaneous targeting of the ATRA and EGF is a rational therapeutic approach to breast cancer. Thus, the anti-tumor effect of retinoids in breast cancer may not simply involve antagonistic interactions with the ERBB pathway. TGFb1/TGFb2 bind to the tyrosine-kinase receptor, TGFBR1, while TGFa interacts with ERBB1. In the case of TGFbs, liganded TGFBR1 associates with TGFBR2 and activates SMAD (SMAD1-4) proteins, which enter the nucleus and act as transcription factors [49]. In breast cancer, TGFa is pro-oncogenic [50]. During the early phases of the disease, TGFb is tumor-suppressive, becoming prooncogenic at later stages [49]. Estrogens negatively regulate TGFb signaling and anti-estrogens exert an opposite effect in ER+ breast cancer [51]. In ER+ and retinoid-sensitive MCF-7 cells, TGFb1 potentiates the anti-tumor activity of ATRA [52], which inhibits estradiol-induced TGFa expression [53]. Some of the interactions observed between ATRA and TGFs converge on common mediators, like the IGF (Insulin-like-growth-factor)-binding-protein, IGFBP-3. Silencing of IGFBP3 suppresses the growth inhibitory action of ATRA and TGFb2 [54]. The link between retinoids and TGFs in ER+ cells is exemplified by a tamoxifen-resistant MCF-7 clone, which shows cross-resistance to ATRA and TGFb1 [55]. In ERbreast cancer cells, evidence of interactions between the retinoid and TGF pathways is limited [21]. In the retinoid-sensitive ER / HER2+ SKBR3 cells, ATRA up-regulates TGFb2, TGFBR2, and the downstream signal transducer SMAD3, while it down-regulates TGFBR1 [21]. Up-regulation of TGFb2 mediates ATRA-dependent phosphorylation/activation of SMAD3, contributing to the antiproliferative action of the retinoid [21]. Thus, possible strategies aimed at potentiating the retinoids’ anti-tumor activity may rely on the modulation of the TGF pathway. IGF1/IGF2 regulate the progression of breast tumors via membrane receptors (IGFR-1/IGFR-2) and their activity is stimulated by estrogens [56]. The action of IGF1/IGF2 is modulated by extracellular binding proteins (IGFBP1-7) in a positive or negative manner. In MCF-7 cells, ATRA inhibits the growth stimulatory activity of IGF-1/IGF-2 with [57] or without [58] suppression of synthesis/ secretion. This may be explained by the bimodal nature of the ATRA-dependent modulation of the IGF pathway. At low concentrations, ATRA stimulates IGF-induced proliferation of MCF-7 cells, while the opposite is true at high concentrations [56]. In ER+ cells [59], ATRA-dependent induction of IGFBP-3 may be relevant for the anti-estrogenic action of retinoids, as estradiol down-regulates IGFBP-3 [60]. However, IGFBP-3 can stimulate IGF-dependent proliferation and may also be a determinant of retinoid-resistance [61] by blocking RAR/RXR heterodimer formation. Finally, both IGF-1 and ATRA are endowed with anti-invasive properties mediated by distinct mechanisms [62].

Retinoids and WNT/NOTCH WNTs (wingless/integrated-growth-factors) control the homeostasis/development of mammary glands [63]. WNTs interact with the membrane receptors, FZD1-10 (frizzled1-10), activating the DSH (disheveled) kinase. DSH phosphorylates and inhibits GSKb (glycogen-synthase-kinase-beta), causing accumulation of b-catenin, which relocates to the nucleus and adherens junctions. In the nucleus, b-catenin acts as a transcription factor, causing a proliferative response, while migration to adherens-junctions favors cell–cell interactions, potentially inhibiting the metastatic process.

744

E. Garattini et al. / Cancer Treatment Reviews 40 (2014) 739–749

WNT ligands, b-catenin and GSK3b influence the activity of RARs/RXRs [63]. Conversely, ATRA modulates the function of WNTs, acting on b-catenin and corresponding target genes [63]. In MCF-7 cells, estradiol up-regulates WNT2, WNT5B, WNT10B [64,65] and FZD10 [66], while it down-regulates WNT2B2, WNT3 and WNT3A [67]. With the only exceptions of WNT5B and WNT10B, a very similar pattern of WNT and FZD up- and downregulation is afforded by ATRA in embryo-carcinoma cells [64,66,67]. Although confirmation in mammary tumors is necessary, it is interesting that an anti-estrogenic compound like ATRA alters the expression of specific WNTs in the same direction as estradiol. The effect of ATRA on b-catenin is complex. In the nucleus, ATRA-stimulated RARs and RXRs bind b-catenin and inhibit its transcriptional activity. Conversely, the transcription of retinoid-target genes by liganded RARs is stimulated by b-catenin [68]. Inhibition of nuclear b-catenin may contribute to the ATRAdependent anti-proliferative action. Despite functional inhibition in the nucleus, ATRA up-regulates b-catenin in a RARa-dependent manner [21] and re-localizes the protein to adherens-junctions stimulating the assembly of these structures [21]. Re-localization of the protein to adherens-junctions participates in the epithelial differentiation of breast cancer cells contributing to the anti-metastatic action of ATRA [21]. The last layer of interactions between the retinoid and the WNT pathways involves b-catenin targetgenes. Nuclear b-catenin induces cyclin D1 and c-MYC, two factors stimulating cell proliferation and down-regulated by ATRA [69–71] in ER+ cells. Another common target of WNTs and retinoids is PPARb/d [8,43]. Since ATRA-dependent activation of PPAR b/d stimulates the growth of breast cancer cells, increased expression of the receptor by WNT may prove detrimental for retinoids’ anti-tumor activity and may also underly natural or induced retinoid-resistance in certain types of breast cancer. NOTCH proteins (NOTCH 1–4) are cell-surface receptors interacting with specific ligands (Jagged1–2, Delta1–3) through cell– cell contacts. High expression of NOTCH ligands/receptors is linked to poor clinical outcome in breast cancer patients [72]. NOTCH-1 favors mammary cancer stem cell renewal [73] and influences invasiveness controlling epithelial-to-mesenchymal transition [74]. In MDA-MB231 breast cancer cells, high concentrations of ATRA down-regulate NOTCH3, up-regulate miR-200, and reduce invasiveness [74]. Interestingly, NOTCH and miR-200 control invasiveness in a positive and negative fashion, respectively. Retinoids and PI3K/AKT Class I PI3Ks (phosphoinositide-3-OH-kinase) are often deranged in cancer and are thoroughly studied enzymes, as they transform PIP2 into PIP3, the major phosphoinositide. Phosphorylation of PIP2 into PIP3 is reversed by PTEN (Phosphatase and Tensin Homolog). Many stimuli, including growth factors, activate PI3K (Fig. 4). As activation of the PI3K pathway contributes to the proliferation/survival/differentiation of neoplastic cells, ATRA is expected to inhibit PI3K directly or indirectly. In 3D cultures of breast cancer cells, the rapid, RAR-dependent and transcriptionindependent inhibition of PI3K by ATRA triggers differentiative responses [75]. Inhibition is ascribed to decreased binding of the p85a regulatory to the p110 catalytic subunit of PI3K. The effect of ATRA on PI3Ks is not simply inhibitory. ATRA-induced PI3K activation is observed in MCF-7 cells, where the phenomenon is also dependent on RARa binding to p85a [76]. Similar to RARa, RARb2 and RXRa also interact with p85a in breast cancer cells [77]. The downstream AKT kinases (AKT1–3) (Fig. 4) are invariably activated by PI3K [78] and play a fundamental role in the proliferation/survival of breast cancer cells. PIP3 anchors AKTs to the plasma membrane allowing phosphorylation/activation by PDK1 (Phosphoinositide-Dependent-Kinase-1). In breast cancer, there is

evidence for reciprocal influences between the AKT and the retinoid pathways. In HER2 /ER+ MCF-7 cells, ERBB2 over-expression causes retinoid resistance possibly via AKT-dependent down-regulation of RARa [79]. The AKT-dependent inhibitory action of ERBB2 involves also functional inhibition of RARa. In fact, the ERBB2 inhibitor, trastuzumab, increases RARE-binding of RARs/RXRs in MDA-MB453 cells via AKT stimulation [80]. Not only does AKT influence RAR/RXR activity, but there is also evidence for the reverse. In SKBR3 cells, ATRA and RARa agonists cause long-term inhibition of AKT phosphorylation/activation without affecting the steady-state levels of the kinase [21]. A possible mechanism underlying AKT inhibition by ATRA is IRS-1 (Insulin-Receptor-Substrate-1) down-regulation, which abrogates IGF1-induced stimulation of the kinase and concurs to growth inhibition [81]. A further mechanism involves the PI3K/AKT inhibitory phosphatase, PTEN. In MCF-7 cells, ATRA causes demethylation of the PTEN promoter and increased expression of the encoded protein. A similar effect is not observed in ATRA-resistant MDA-MB231 cells [82], suggesting the potential significance of PTEN suppression for ATRA-resistance. Further support to the potential involvement of the PI3K/ AKT pathway in the resistance of certain breast cancer types to ATRA is given by the observation that stimulation of the growth activator, PPARb/d, induces PDK1, which is a direct target of the nuclear receptor and contributes to mammary tumorigenesis [83,84]. An example of interaction between the retinoid and the AKT pathways at the level of the kinase target-genes is represented by the pro-apoptotic forkhead transcription factor, FKHRL1, which is induced by ATRA [85] and phosphorylated/inactivated by AKT. FKHRL1 represses ER-mediated and stimulates RAR-mediated transcription. Another example is the cell cycle inhibitor p27, which is phosphorylated/inactivated by AKT [86], whereas it is induced/ activated by ATRA [87]. The data available in breast cancer models indicate that retinoids generally antagonize AKT. As AKT inhibitors revert retinoid-resistance in AML models [88], combinations between AKT inhibitors and ATRA or derivatives should be tested in breast tumors refractory to retinoids. Retinoids and MAPKs The Mitogen-Activated Protein Kinases (MAPKs) are divided into 3 main groups: p38 MAPKs, Extracellular signal-Regulated Kinases (ERKs) and c-Jun NH2-terminal Kinases (JNKs). P38-MAPKs are known for their relevance in cytokine production and stress responses and they are activated by various stimuli, including growth factors. P38-MAPK downstream targets include RARa, RARc and other transcription factors. P38a physically interacts with unliganded RARa and inhibits ligand-dependent transactivation of the receptor [89]. Binding of p38a leads to a late and prolonged destabilization of RARa requiring proteasomedependent degradation. However, the action of the kinase on RARa is biphasic, as the inhibitory action is preceeded by an early and ATRA-dependent stimulatory effect, involving phosphorylation by Mitogen- and Stress-activated Protein Kinase 1 (MSK1) [90]. Phosphorylation favors the recruitment of RARa to RAREs. The rapid activation of p38-MAPK/MSK1 by ATRA is mediated by non-genomic effects requiring the fraction of RARa associated with membrane lipid rafts. RARc activity is also modulated by P38-MAPK, although the kinase exert opposite effects on this receptor and RARa. In fact, inhibition is RARa-specific, as p38-MAPK stimulates the transcriptional activity of RARc2 despite induction of the phosphorylation and turnover of the receptor [91]. Differential regulation of RARa and RARc by p38-MAPK may have important ramifications for the anti-tumor activity of ATRA in breast cancer, given the opposite action exerted by the two retinoid receptors on tumor cell growth (see Retinoids and breast cancer: many pre-clinical studies and few clinical trials). RARa and RARc are not the only components of the

E. Garattini et al. / Cancer Treatment Reviews 40 (2014) 739–749

745

Fig. 4. Cross-talk between the ATRA and the PI3K/AKT pathways. The figure illustrates the two-way cross-talk between the ATRA and the PI3K/AKT pathways. Green arrow = activation; red arrow = inhibition. The red and green lines ending with a semi-circle indicate direct physical interaction between the indicated proteins leading to inhibition and activation of PI3K, respectively. The dashed green lines indicate that the interactions between RARs and PI3K are linked to the activation of the downstream RhoA and Rac1 GTPases. The circular arrow associated with the ‘‘+’’ symbol shows that ATRA, PI3K or AKT promote the indicated interactions. Solid black arrows indicate intra-cellular relocation. The gray boxes indicate pathways rather than single elements of the pathway. TGASE = Transglutaminase gene. The binding of ATRA to the cytosolic CRABP2 and FABP5 proteins, leding is indicated. CRABP2 directs ATRA to the RAR/RXR nuclear complexes, while FABP5 directs the retinoid to PPARb/d/RXR heterodimers. P85a and TGASE are examples of genes whose expression is controlled by ATRA-dependent activation of the RAR/RXR complexes, while PDK1 exemplifies genes regulated by ATRA-dependent activation of the PPARb/d/RXR dimers. The symbols of RARs and RXRs in the solid square represent the pool of cytosolic retinoid receptors. The figure is an elaboration of the PI3K/AKT pathway available from Cell Signaling Technology, Inc.

746

E. Garattini et al. / Cancer Treatment Reviews 40 (2014) 739–749

RAR-containing transcription complexes which are phosphorylated by p38-MAPK, as the kinase also phosphorylates the SRC3 co-activator. SRC-3 phosphorylation causes an initial facilitation of RARatarget-genes activation, which is followed by an inhibition caused by SRC-3 degradation. Blockade of p38-MAPK-dependent SRC3 degradation enhances the transcription of ATRA-dependent genes. Interestingly, SRC-3 phosphorylation and degradation occur only within the context of RARa-containing complexes [92]. Besides controlling RARs, p38-MAPKs are also regulated by ATRA and this may have an impact on the anti-proliferative activity of the retinoid. Conflicting results on the effects exerted by ATRA on p38-MAPK in breast cancer cells are available. P38a and p38b have been shown to be activated by ATRA in MCF-7 cells [93], although an opposite effect associated with AP-1 inhibition and growth arrest has also been reported in the same cells [94]. In AML models, inhibition/silencing of p38-MAPK sensitizes the blast to the anti-leukemic activity of ATRA and RARa agonists. It would be important to investigate whether these data can be replicated in breast cancer cells to establish if combinations of retinoids and p38-MAPK inhibitors represent a viable therapeutic strategy. The point is of particular therapeutic significant, as activation rather than inhibition of p38-MAPK is associated with decreased breast cancer progression being a determinant of tumor cell dormancy [95]. ERKs (ERK1–2) contribute to the growth of breast cancer cells and are activated by many stimuli. Like p38-MAPKs, ERKs phosphorylate MSK, which controls the activity of RARs [90]. ERKs can also phosphorylate Signal Transducer and Activator of Transcription 1/3 (STAT1/3), ERa, c-JUN and c-FOS which are active in mammary tumor cells and are regulated by retinoids [26,96–98]. In general, the results obtained in breast cancer cells support the idea that ATRA inhibits the ERK pathway. In ER+ MCF-7 cells, ATRA reduces ERK phosphorylation, inhibiting AP-1 activity [94]. In the same model, ERKs control ATRA-dependent induction of the transcription factor, NRF2 (Nuclear-Factor-Erythroid-2-like-factor-2), which protects cells from oxidative damage. ERK inhibitors block the ATRA-dependent induction of NRF2, suggesting that combinations of these compounds and ATRA can be exploited to potentiate the cytotoxic activity of the retinoid [99]. These results suggest that combinations of ERK inhibitors and ATRA should be evaluated in ER+ tumors specifically, as ERa is activated by ERKs [100]. Another setting where combinations of ATRA and ERK inhibitors should be tested is represented by HER2+ tumors, since ATRA down-regulates the ERK protein and causes a late inhibitory action on ERK phosphorylation in ER /HER2+SKBR3 cells [21]. Contrasting results on the role played by JNKs in the process of breast cancer induction and progression are available [101]. Some studies emphasize the tumor-suppressor properties of JNK, while others support the idea that JNK is pro-oncogenic. Of further interest for the biology of breast cancer is the fact that JNK activation is associated with tamoxifen resistance in ER+ tumors. Only three studies focus on the action exerted by retinoids on the JNK pathways in breast cancer cells and two of them report no significant effects. The third one demonstrates that ATRA stimulates JNK, which contributes to the retinoid-dependent enhancement of the cytotoxic effect exerted by taxotere in the cells [102]. Consistent with the uncertainty as to the pro- or anti-oncogenic significance of JNK, the effects of ATRA on the activation state of the kinase in models other than breast cancer are contradictory. In non-small cell lung carcinoma cells, serum-induced JNK phosphorylation/ activation is inhibited by ATRA via both RAR/RXR-dependent and -independent effects [103]. In contrast, ATRA activates JNK in neuroblastoma cells [104]. JNK activation, which is sometimes observed in mammary cancer, may contribute to RAR dysfunction by phosphorylating RARa and inducing its degradation through the ubiquitin-proteasomal pathway, as observed in lung cancer [105].

Retinoids and PKA/PKC Intracellular cAMP accumulation stimulates cAMP-dependent Protein Kinases (PKAs). PKAs consist of two regulatory (R) and two catalytic (C) subunits and interact with many other signaling pathways. Increases in cAMP enhance the differentiative and anti-proliferative effects of ATRA in different cell types [106– 108]. Conversely, ATRA [109], 9-cisRA and 13-cisRA [110] activate PKAs in myeloid and lung cancer cells [109,110], which contributes tothe anti-proliferative effect of these retinoids. The cAMP/PKA pathway exerts a positive action on the retinoid system, as cAMP stimulates the transcriptional activity of the RAR/RXR complexes via PKA-dependent phosphorylation of RARs. In ER+ [60,111] breast cancer cells, activation of the cAMP pathway causes growth inhibition. In MCF-7 cells, this effect is accompanied by induction of IGFBP3. IGFBP3 induction and growth inhibition are triggered also by ATRA and enhanced by combinations of the retinoid and PKA activators. cAMP-dependent effects at the level of some estrogenresponsive genes contribute to the anti-estrogenic activity of retinoids. A relevant example is represented by the anti-apoptotic BCL2 gene, whose transcription is stimulated by estradiol via PKA. ATRA reverts this effect by inhibition of PKA translocation to the regulatory regions of the BCL-2 gene [112]. Another link between cAMP and retinoids in breast cancer cells is represented by the p85a subunit of PI3K (p85a-PI3K) [76]. In MCF-7 cells, unliganded RARa binds to p85a-PI3K [76,113]. Binding is activated by ATRA, PKA activators and more so by combinations of the two agents. The interaction between RARa and p85a-PI3K contributes to the anti-proliferative and anti-motility effects afforded by ATRA and/ or PKA activators. In conclusion, the results available in breast cancer indicate that combinations based on the use of retinoids and cAMP stimulating agents may represent a therapeutic opportunity particularly in the case of ER+ mammary tumor. PKCs phosphorylate serine/threonine residues in many target proteins and are divided into: conventional PKCs (c-PKCs), requiring calcium and diacylglycerol, novel PKCs (n-PKCs), requiring only diacylglycerol and atypical PKCs (a-PKC), requiring neither calcium nor diacylglycerol. PKCs phosphorylate various substrates, including RARa [114]. High levels and differential activation of specific PKCs are observed in breast cancer cells. ATRA induces c-PKCa and represses a-PKCi expression [115] in retinoid-sensitive ER+ T47D cells, while the retinoid has no effect in retinoid-resistant ER MDA-MB231 cells. The induction of c-PKCa, but not the reduction in a-PKCi, is RARa-dependent. In other retinoid-sensitive breast cancer cells, ATRA down-regulates c-PKCa, reducing ERK phosphorylation and G1 arrest [116]. Though conflicting, these results indicate that PKCs contribute to the the anti-tumor activity of ATRA. Combinations of ATRA and PKC inhibitors (GF109203X) trigger growth inhibition and apoptosis in the MDA-MB-231 cell line [117]. GF109203X down-regulates PKC activity and the effect is potentiated by ATRA. All this is accompanied by sustained activation of ERK1/2 by ATRA+GF109203X which is required to induce apoptosis. In MCF-7 cells, ATRA activates PKCd, via increased phosphorylation [114]. PKCd associates with RARa on RAREs and its activation contributes to ATRA-dependent transcription. Indeed, PKCd inhibitors diminish ATRA-induced transcription, while overexpression of the kinase exerts opposite effects. In the same cell line, these phenomena are accompanied by growth inhibitory effects, which involve PKCd-dependent inhibition of IGF-I synthesis [57].

Conclusion ATRA and derived retinoids are largely under-exploited therapeutic agents in breast cancer. Numerous pre-clinical studies

E. Garattini et al. / Cancer Treatment Reviews 40 (2014) 739–749

indicate two-way cross-talks between classic retinoids and various extra-cellular factors and intracellular pathways controlling the growth, survival and invasive/metastatic behavior of breast cancer cells. All this identifies these compounds as potential components of innovative and rational therapeutic combinations. In spite of the promising pre-clinical results, the published clinical studies on retinoids in mammary tumors are limited and generally disappointing. One possible explanation for this is the heterogeneous nature of breast cancer and the scarse knowledge on the disease subtypes which may really benefit from retinoid-based therapies. Thus, one of the future priorities for a rational use of retinoids in the clinics will be the identification of the sub-groups of tumors characterized by retinoid-sensitivity. At present, the only data available on the responsiveness of breast cancer subtypes to retinoids are based on the old classification and do not consider the recent segmentation of the disease into many more groups according to the profiles of global gene-expression. In addition, the scientific literature contains assumptions that are not necessarily correct. With respect to the last point, it is generally assumed that ER+ breast cancers are sensitive to ATRA, while the ER counterparts are resistant to the anti-tumor activity of these agents. However, the idea relies on scattered studies, which are dated and mostly based on results obtained in very few popular cell lines. Thus, confirming the association between ER-positivity and retinoid-sensitivity with systematic studies on large panels of cell lines characterized for their gene-expression profiles is a priority. However, it must be emphasized that the concept of heterogeneity unveiled by the whole-genome gene-expression studies applies equally well to the entire population and to the ER+ fraction of mammary tumors. Thus, it is important to evaluate whether ER+ tumors can be further subdivided into sub-groups showing preferential sensitivity to ATRA. In addition, it will be relevant to establish whether HER2 expression always plays a negative role in modulating retinoid sensitivity of HER2+/ER+mammary tumors, as suggested by some observations [79,80]. A step in this direction is the identification of a sub-group of ATRA-sensitive HER2+ breast cancers characterized by amplification of the RARA gene [21], where it is possible to envisage treatments based on combinations of ATRA and the HER2-targeting agent, lapatinib [21]. Finally, it will be crucial to confirm whether the refractoriness to retinoids is a general feature of triple-negative breast tumors, which currently lack effective therapeutic options. Conflict of interest All the authors declare that they have no financial or personal relationships with other people or organizations that could inappropriately influence (bias) the work. We confirm that the manuscript is not under consideration for publication elsewhere. No parts of the manuscript, including figures, have been sent by mail. All the authors have participated sufficiently to the redaction of different sections of the article and take public responsibility for its content. The authors can be contacted at the following e-mail addresses: Enrico [email protected]; Marco Bolis-marco.bolis@ marionegri.it; Silvio Ken [email protected]; Maddalena [email protected]; Floriana Centritto-fl[email protected]; Gabriela [email protected]; Maurizio Gianni’[email protected]; Adriana [email protected]; Anna Pagani-anna.pagani@ fsm.it; James Neil Fisher-jamesneil.fi[email protected]; Alberto [email protected]; Mineko Terao-mineko.terao@ marionegri.it. The publication of the present article is approved by all authors and tacitly by the responsible authorities of the IRCCS-Istituto di Ricerche Farmacologiche ‘‘Mario Negri’’ and the IRCCS-Fondazione S. Maugeri, where the work was carried out.

747

Acknowledgments The authors acknowledge the financial support of the Fondazione ‘‘Italo Monzino’’, the Associazione Italiana per la Ricerca contro il Cancro (AIRC) and the Weizmann-Negri Foundation that made this work possible. They also thank Silvio Garattini and Mario Salmona for critical reading of the manuscript as well as Felice Deceglie for the artwork. References [1] Wolf G. Is 9-cis-retinoic acid the endogenous ligand for the retinoic acid-X receptor? Nutr Rev 2006;64:532–8. [2] Gianni M, Kalac Y, Ponzanelli I, Rambaldi A, Terao M, Garattini E. Tyrosine kinase inhibitor STI571 potentiates the pharmacologic activity of retinoic acid in acute promyelocytic leukemia cells: effects on the degradation of RARalpha and PML-RARalpha. Blood 2001;97:3234–43. [3] Baumrucker CR, Schanbacher F, Shang Y, Green MH. Lactoferrin interaction with retinoid signaling: cell growth and apoptosis in mammary cells. Domest Anim Endocrinol 2006;30:289–303. [4] Sheikh MS, Shao ZM, Li XS, Dawson M, Jetten AM, Wu S, et al. Retinoidresistant estrogen receptor-negative human breast carcinoma cells transfected with retinoic acid receptor-alpha acquire sensitivity to growth inhibition by retinoids. J Biol Chem 1994;269:21440–7. [5] Anding AL, Nieves NJ, Abzianidze VV, Collins MD, Curley Jr RW, Clagett-Dame M. 4-Hydroxybenzyl modification of the highly teratogenic retinoid, 4-[(1E)2-(5,5,8,8-tetramethyl-5,6,7,8-tetrahydro-2-naphthalenyl)-1-propen-1yl]benzoic acid (TTNPB), yields a compound that induces apoptosis in breast cancer cells and shows reduced teratogenicity. Chem Res Toxicol 2011;24:1853–61. [6] Abrams JS, Moore TD, Friedman M. New chemotherapeutic agents for breast cancer. Br J Cancer 1994;74:1164–76. [7] Noy N. Ligand specificity of nuclear hormone receptors: sifting through promiscuity. Biochemistry 2007;46:13461–7. [8] Schug TT, Berry DC, Shaw NS, Travis SN, Noy N. Opposing effects of retinoic acid on cell growth result from alternate activation of two different nuclear receptors. Cell 2007;129:723–33. [9] Sutton LM, Warmuth MA, Petros WP, Winer EP. Pharmacokinetics and clinical impact of all-trans retinoic acid in metastatic breast cancer: a phase II trial. Cancer Chemother Pharmacol 1997;40:335–41. [10] Budd GT, Adamson PC, Gupta M, Homayoun P, Sandstrom SK, Murphy RF, et al. Phase I/II trial of all-trans retinoic acid and tamoxifen in patients with advanced breast cancer. Clin Cancer Res 1998;4:635–42. [11] Toma S, Raffo P, Nicolo G, Canavese G, Margallo E, Vecchio C, et al. Biological activity of all-trans-retinoic acid with and without tamoxifen and alphainterferon 2a in breast cancer patients. Int J Oncol 2000;17:991–1000. [12] Bryan M, Pulte ED, Toomey KC, Pliner L, Pavlick AC, Saunders T, et al. A pilot phase II trial of all-trans retinoic acid (Vesanoid) and paclitaxel (Taxol) in patients with recurrent or metastatic breast cancer. Invest New Drugs 2011;29:1482–7. [13] Chiesa MD, Passalacqua R, Michiara M, Franciosi V, Di Costanzo F, Bisagni G, et al. Tamoxifen vs Tamoxifen plus 13-cis-retinoic acid vs Tamoxifen plus Interferon alpha-2a as first-line endocrine treatments in advanced breast cancer: updated results of a phase II, prospective, randomised multicentre trial. Acta Biomed 2007;78:204–9. [14] Lo-Coco F, Avvisati G, Vignetti M, Thiede C, Orlando SM, Iacobelli S, et al. Retinoic acid and arsenic trioxide for acute promyelocytic leukemia. N Engl J Med 2013;369:111–21. [15] Nelson CH, Buttrick BR, Isoherranen N. Therapeutic potential of the inhibition of the retinoic acid hydroxylases CYP26A1 and CYP26B1 by xenobiotics. Curr Top Med Chem 2013;13:1402–28. [16] Ozpolat B, Mehta K, Tari AM, Lopez-Berestein G. All-trans-Retinoic acidinduced expression and regulation of retinoic acid 4-hydroxylase (CYP26) in human promyelocytic leukemia. Am J Hematol 2002;70:39–47. [17] Sonneveld E, van den Brink CE, van der Leede BM, Schulkes RK, Petkovich M, van der Burg B, et al. Human retinoic acid (RA) 4-hydroxylase (CYP26) is highly specific for all-trans-RA and can be induced through RA receptors in human breast and colon carcinoma cells. Cell Growth Differ 1998;9:629–37. [18] Sirchia SM, Ferguson AT, Sironi E, Subramanyan S, Orlandi R, Sukumar S, et al. Evidence of epigenetic changes affecting the chromatin state of the retinoic acid receptor beta2 promoter in breast cancer cells. Oncogene 2000;19:1556–63. [19] Sirchia SM, Ren M, Pili R, Sironi E, Somenzi G, Ghidoni R, et al. Endogenous reactivation of the RARbeta2 tumor suppressor gene epigenetically silenced in breast cancer. Breast Cancer Res 2002;62:2455–61. [20] Shao Z, Sheikh M, Chen J, Kute T, Aisner S, Schnaper L, et al. Expression of the retinoic Acid nuclear receptors (rars) and retinoid x-receptor (rxr) genes in estrogen-receptor positive and negative breast-cancer. Int J Oncol 1994;4:859–63. [21] Paroni G, Fratelli M, Gardini G, Bassano C, Flora M, Zanetti A, et al. Synergistic antitumor activity of lapatinib and retinoids on a novel subtype of breast cancer with coamplification of ERBB2 and RARA. Oncogene 2012;31:3431–43.

748

E. Garattini et al. / Cancer Treatment Reviews 40 (2014) 739–749

[22] Terao M, Fratelli M, Kurosaki M, Zanetti A, Guarnaccia V, Paroni G, et al. Induction of miR-21 by retinoic acid in estrogen receptor-positive breast carcinoma cells: biological correlates and molecular targets. J Biol Chem 2011;286:4027–42. [23] Lu M, Mira-y-Lopez R, Nakajo S, Nakaya K, Jing Y. Expression of estrogen receptor alpha, retinoic acid receptor alpha and cellular retinoic acid binding protein II genes is coordinately regulated in human breast cancer cells. Oncogene 2005;24:4362–9. [24] Elgort MG, Zou A, Marschke KB, Allegretto EA. Estrogen and estrogen receptor antagonists stimulate transcription from the human retinoic acid receptoralpha 1 promoter via a novel sequence. Mol Endocrinol 1996;10:477–87. [25] Hua S, Kittler R, White KP. Genomic antagonism between retinoic acid and estrogen signaling in breast cancer. Cell 2009;137:1259–71. [26] Ross-Innes CS, Stark R, Holmes KA, Schmidt D, Spyrou C, Russell R, et al. Cooperative interaction between retinoic acid receptor-alpha and estrogen receptor in breast cancer. Genes Dev 2010;24:171–82. [27] Salazar MD, Ratnam M, Patki M, Kisovic I, Trumbly R, Iman M. During hormone depletion or tamoxifen treatment of breast cancer cells the estrogen receptor apoprotein supports cell cycling through the retinoic acid receptor alpha1 apoprotein. Breast Cancer Res 2011;13:R18. [28] Johansson HJ, Sanchez BC, Mundt F, Forshed J, Kovacs A, Panizza E, et al. Retinoic acid receptor alpha is associated with tamoxifen resistance in breast cancer. Nat Commun 2013;4:2175. [29] Cohn E, Ossowski L, Bertran S, Marzan C, Farias EF. RARalpha1 control of mammary gland ductal morphogenesis and wnt1-tumorigenesis. Breast Cancer Res 2010;12:R79. [30] Wang S, Tricot G, Shi L, Xiong W, Zeng Z, Xu H, et al. RARalpha2 expression is associated with disease progression and plays a crucial role in efficacy of ATRA treatment in myeloma. Blood 2009;114:600–7. [31] Connolly RM, Nguyen NK, Sukumar S. Molecular pathways: current role and future directions of the retinoic acid pathway in cancer prevention and treatment. Clin Cancer Res 2013;19:1651–9. [32] Liu X, Nugoli M, Laferriere J, Saleh SM, Rodrigue-Gervais IG, Saleh M, et al. Stromal retinoic acid receptor beta promotes mammary gland tumorigenesis. Proc Natl Acad Sci USA 2011;108:774–9. [33] Liu Y, Lee MO, Wang HG, Li Y, Hashimoto Y, Klaus M, et al. Retinoic acid receptor beta mediates the growth-inhibitory effect of retinoic acid by promoting apoptosis in human breast cancer cells. Mol Cell Biol 1996;16:1138–49. [34] Peng X, Maruo T, Cao Y, Punj V, Mehta R, Das Gupta TK, et al. A novel RARbeta isoform directed by a distinct promoter P3 and mediated by retinoic acid in breast cancer cells. Cancer Res 2004;64:8911–8. [35] Hayashi K, Goodison S, Urquidi V, Tarin D, Lotan R, Tahara E. Differential effects of retinoic acid on the growth of isogenic metastatic and nonmetastatic breast cancer cell lines and their association with distinct expression of retinoic acid receptor beta isoforms 2 and 4. Int J Oncol 2003;22:623–9. [36] Shang Y, Baumrucker CR, Green MH. Signal relay by retinoic acid receptors alpha and beta in the retinoic acid-induced expression of insulin-like growth factor-binding protein-3 in breast cancer cells. J Biol Chem 1999;274:18005–10. [37] Muscat GE, Eriksson NA, Byth K, Loi S, Graham D, Jindal S, et al. Research resource: nuclear receptors as transcriptome: discriminant and prognostic value in breast cancer. Mol Endocrinol 2013;27:350–65. [38] Bosch A, Bertran SP, Lu Y, Garcia A, Jones AM, Dawson MI, et al. Reversal by RARalpha agonist Am 580 of c-Myc-induced imbalance in RARalpha/ RARgamma expression during MMTV-Myc tumorigenesis. Breast Cancer Res 2012;14:R121. [39] Lu Y, Bertran S, Samuels TA, Mira-y-Lopez R, Farias EF. Mechanism of inhibition of MMTV-neu and MMTV-wnt1 induced mammary oncogenesis by RARalpha agonist AM580. Oncogene 2010;29:3665–76. [40] Purton LE, Dworkin S, Olsen GH, Walkley CR, Fabb SA, Collins SJ, et al. RARgamma is critical for maintaining a balance between hematopoietic stem cell self-renewal and differentiation. J Exp Med 2006;203:1283–93. [41] Schneider SM, Offterdinger M, Huber H, Grunt TW. Activation of retinoic acid receptor alpha is sufficient for full induction of retinoid responses in SK-BR-3 and T47D human breast cancer cells. Breast Cancer Res 2000;60:5479–87. [42] Morgan E, KannanThulasiraman P, Noy N. Involvement of fatty acid binding protein 5 and PPARbeta/delta in prostate cancer cell growth. PPAR Res 2010;2010. [43] Schug TT, Berry DC, Toshkov IA, Cheng L, Nikitin AY, Noy N. Overcoming retinoic acid-resistance of mammary carcinomas by diverting retinoic acid from PPARbeta/delta to RAR. Proc Natl Acad Sci USA 2008;105:7546–51. [44] Tighe AP, Talmage DA. Retinoids arrest breast cancer cell proliferation: retinoic acid selectively reduces the duration of receptor tyrosine kinase signaling. Exp Cell Res 2004;301:147–57. [45] Xu J, Keeton AB, Wu L, Franklin JL, Cao X, Messina JL. Gene 33 inhibits apoptosis of breast cancer cells and increases poly(ADP-ribose) polymerase expression. Breast Cancer Res Treat 2005;91:207–15. [46] Grunt TW, Puckmair K, Tomek K, Kainz B, Gaiger A. An EGF receptor inhibitor induces RAR-beta expression in breast and ovarian cancer cells. Biochem Biophys Res Commun 2005;329:1253–9. [47] Kannan-Thulasiraman P, Seachrist DD, Mahabeleshwar GH, Jain MK, Noy N. Fatty acid-binding protein 5 and PPARbeta/delta are critical mediators of

[48]

[49] [50]

[51]

[52]

[53]

[54]

[55]

[56]

[57]

[58]

[59]

[60]

[61]

[62]

[63]

[64] [65]

[66]

[67] [68] [69]

[70]

[71]

[72] [73] [74]

epidermal growth factor receptor-induced carcinoma cell growth. J Biol Chem 2010;285:19106–15. Offterdinger M, Schneider SM, Grunt TW. Heregulin and retinoids synergistically induce branching morphogenesis of breast cancer cells cultivated in 3D collagen gels. J Cell Physiol 2003;195:260–75. Barcellos-Hoff MH, Akhurst RJ. Transforming growth factor-beta in breast cancer: too much, too late. Breast Cancer Res 2009;11:202. Ciardiello F, Kim N, McGeady ML, Liscia DS, Saeki T, Bianco C, et al. Expression of transforming growth factor alpha (TGF alpha) in breast cancer. Ann Oncol 1991;2:169–82. Kleuser B, Malek D, Gust R, Pertz HH, Potteck H. 17-Beta-estradiol inhibits transforming growth factor-beta signaling and function in breast cancer cells via activation of extracellular signal-regulated kinase through the G proteincoupled receptor 30. Mol Pharmacol 2008;74:1533–43. Valette A, Botanch C. Transforming growth factor beta (TGF-beta) potentiates the inhibitory effect of retinoic acid on human breast carcinoma (MCF-7) cell proliferation. Growth Factors 1990;2:283–7. Fontana JA, Nervi C, Shao ZM, Jetten AM. Retinoid antagonism of estrogenresponsive transforming growth factor alpha and pS2 gene expression in breast carcinoma cells. Breast Cancer Res 1992;52:3938–45. Gucev ZS, Oh Y, Kelley KM, Rosenfeld RG. Insulin-like growth factor binding protein 3 mediates retinoic acid- and transforming growth factor beta2induced growth inhibition in human breast cancer cells. Breast Cancer Res 1996;56:1545–50. Herman ME, Katzenellenbogen BS. Response-specific antiestrogen resistance in a newly characterized MCF-7 human breast cancer cell line resulting from long-term exposure to trans-hydroxytamoxifen. J Steroid Biochem Mol Biol 1996;59:121–34. Bentel JM, Lebwohl DE, Cullen KJ, Rubin MS, Rosen N, Mendelsohn J, et al. Insulin-like growth factors modulate the growth inhibitory effects of retinoic acid on MCF-7 breast cancer cells. J Cell Physiol 1995;165:212–21. Oh YI, Kim JH, Kang CW. Oxidative stress in MCF-7 cells is involved in the effects of retinoic acid-induced activation of protein kinase C-delta on insulin-like growth factor-I secretion and synthesis. Growth Horm IGF Res 2010;20:101–9. Fontana JA, Burrows-Mezu A, Clemmons DR, LeRoith D. Retinoid modulation of insulin-like growth factor-binding proteins and inhibition of breast carcinoma proliferation. Endocrinology 1991;128:1115–22. Sheikh MS, Shao ZM, Hussain A, Clemmons DR, Chen JC, Roberts Jr CT, et al. Regulation of insulin-like growth factor-binding-protein-1, 2, 3, 4, 5, and 6: synthesis, secretion, and gene expression in estrogen receptor-negative human breast carcinoma cells. J Cell Physiol 1993;155:556–67. Martin JL, Coverley JA, Pattison ST, Baxter RC. Insulin-like growth factorbinding protein-3 production by MCF-7 breast cancer cells: stimulation by retinoic acid and cyclic adenosine monophosphate and differential effects of estradiol. Endocrinology 1995;136:1219–26. Schedlich LJ, O’Han MK, Leong GM, Baxter RC. Insulin-like growth factor binding protein-3 prevents retinoid receptor heterodimerization: implications for retinoic acid-sensitivity in human breast cancer cells. Biochem Biophys Res Commun 2004;314:83–8. Vermeulen SJ, Bruyneel EA, van Roy FM, Mareel MM, Bracke ME. Activation of the E-cadherin/catenin complex in human MCF-7 breast cancer cells by alltrans-retinoic acid. Br J Cancer 1995;72:1447–53. Mulholland DJ, Dedhar S, Coetzee GA, Nelson CC. Interaction of nuclear receptors with the Wnt/beta-catenin/Tcf signaling axis: Wnt you like to know? Endocr Rev 2005;26:898–915. Katoh M. Differential regulation of WNT2 and WNT2B expression in human cancer. Int J Mol Med 2001;8:657–60. Kirikoshi H, Katoh M. Expression and regulation of WNT10B in human cancer: up-regulation of WNT10B in MCF-7 cells by beta-estradiol and downregulation of WNT10B in NT2 cells by retinoic acid. Int J Mol Med 2002;10:507–11. Saitoh T, Mine T, Katoh M. Up-regulation of Frizzled-10 (FZD10) by betaestradiol in MCF-7 cells and by retinoic acid in NT2 cells. Int J Oncol 2002;20:117–20. Katoh M. Regulation of WNT3 and WNT3A mRNAs in human cancer cell lines NT2, MCF-7, and MKN45. Int J Oncol 2002;20:373–7. Easwaran V, Pishvaian M, Salimuddin, Byers S. Cross-regulation of betacatenin-LEF/TCF and retinoid signaling pathways. Curr Biol 1999;9:1415–8. Pratt MA, Niu M, White D. Differential regulation of protein expression, growth and apoptosis by natural and synthetic retinoids. J Cell Biochem 2003;90:692–708. Wu K, DuPre E, Kim H, Tin UC, Bissonnette RP, Lamph WW, et al. Receptorselective retinoids inhibit the growth of normal and malignant breast cells by inducing G1 cell cycle blockade. Breast Cancer Res Treat 2006;96:147–57. Shang Y, Baumrucker CR, Green MH. C-Myc is a major mediator of the synergistic growth inhibitory effects of retinoic acid and interferon in breast cancer cells. J Biol Chem 1998;273:30608–13. Han J, Hendzel MJ, Allalunis-Turner J. Notch signaling as a therapeutic target for breast cancer treatment? Breast Cancer Res 2011;13:210. Gangopadhyay S, Nandy A, Hor P, Mukhopadhyay A. Breast cancer stem cells: a novel therapeutic target. Clin Breast Cancer 2013;13:7–15. Mezquita B, Mezquita J, Barrot C, Carvajal S, Pau M, Mezquita P, et al. A truncated Flt1 isoform that activates Src and promotes invasion in breast

E. Garattini et al. / Cancer Treatment Reviews 40 (2014) 739–749

[75]

[76]

[77]

[78]

[79]

[80]

[81]

[82]

[83]

[84]

[85]

[86]

[87]

[88]

[89]

[90]

[91]

[92]

[93]

[94] [95]

[96]

cancer cells is upregulated by Notch-1 and Notch-3 and downregulated by miR-200c and retinoic acid. J Cell Biochem 2014;115:52–61. Farias EF, Marzan C, Mira-y-Lopez R. Cellular retinol-binding protein-I inhibits PI3K/Akt signaling through a retinoic acid receptor-dependent mechanism that regulates p85–p110 heterodimerization. Oncogene 2005;24:1598–606. Donini CF, Di Zazzo E, Zuchegna C, Di Domenico M, D’Inzeo S, Nicolussi A, et al. The p85alpha regulatory subunit of PI3K mediates cAMP-PKA and retinoic acid biological effects on MCF7 cell growth and migration. Int J Oncol 2012;40:1627–35. Ohashi E, Kogai T, Kagechika H, Brent GA. Activation of the PI3 kinase pathway by retinoic acid mediates sodium/iodide symporter induction and iodide transport in MCF-7 breast cancer cells. Breast Cancer Res 2009;69:3443–50. Grunt TW, Mariani GL. Novel approaches for molecular targeted therapy of breast cancer: interfering with PI3K/AKT/mTOR signaling. Curr Cancer Drug Targets 2013;13:188–204. Tari AM, Lim SJ, Hung MC, Esteva FJ, Lopez-Berestein G. Her2/neu induces alltrans retinoic acid (ATRA) resistance in breast cancer cells. Oncogene 2002;21:5224–32. Siwak DR, Mendoza-Gamboa E, Tari AM. HER2/neu uses Akt to suppress retinoic acid response element binding activity in MDA-MB-453 breast cancer cells. Int J Oncol 2003;23:1739–45. del Rincon SV, Rousseau C, Samanta R, Miller Jr WH. Retinoic acid-induced growth arrest of MCF-7 cells involves the selective regulation of the IRS-1/PI 3-kinase/AKT pathway. Oncogene 2003;22:3353–60. Stefanska B, Salame P, Bednarek A, Fabianowska-Majewska K. Comparative effects of retinoic acid, vitamin D and resveratrol alone and in combination with adenosine analogues on methylation and expression of phosphatase and tensin homologue tumour suppressor gene in breast cancer cells. Br J Nutr 2012;107:781–90. Yu S, Levi L, Siegel R, Noy N. Retinoic acid induces neurogenesis by activating both retinoic acid receptors (RARs) and peroxisome proliferator-activated receptor beta/delta (PPARbeta/delta). J Biol Chem 2012;287:42195–205. Pollock CB, Yin Y, Yuan H, Zeng X, King S, Li X, et al. PPARdelta activation acts cooperatively with 3-phosphoinositide-dependent protein kinase-1 to enhance mammary tumorigenesis. PLoS One 2011;6:e16215. Zhao HH, Herrera RE, Coronado-Heinsohn E, Yang MC, Ludes-Meyers JH, Seybold-Tilson KJ, et al. Forkhead homologue in rhabdomyosarcoma functions as a bifunctional nuclear receptor-interacting protein with both coactivator and corepressor functions. J Biol Chem 2001;276:27907–12. Liang J, Zubovitz J, Petrocelli T, Kotchetkov R, Connor MK, Han K, et al. PKB/ Akt phosphorylates p27, impairs nuclear import of p27 and opposes p27mediated G1 arrest. Nat Med 2002;8:1153–60. Eto I. Upstream molecular signaling pathways of p27(Kip1) expression: effects of 4-hydroxytamoxifen, dexamethasone, and retinoic acids. Cancer Cell Int 2010;10:3. Martelli AM, Tazzari PL, Tabellini G, Bortul R, Billi AM, Manzoli L, et al. A new selective AKT pharmacological inhibitor reduces resistance to chemotherapeutic drugs, TRAIL, all-trans-retinoic acid, and ionizing radiation of human leukemia cells. Leukemia 2003;17:1794–805. Gianni M, Peviani M, Bruck N, Rambaldi A, Borleri G, Terao M, et al. P38alphaMAPK interacts with and inhibits RARalpha: suppression of the kinase enhances the therapeutic activity of retinoids in acute myeloid leukemia cells. Leukemia 2012;26:1850–61. Bruck N, Vitoux D, Ferry C, Duong V, Bauer A, de The H, et al. A coordinated phosphorylation cascade initiated by p38MAPK/MSK1 directs RARalpha to target promoters. EMBO J 2009;28:34–47. Gianni M, Kopf E, Bastien J, Oulad-Abdelghani M, Garattini E, Chambon P, et al. Down-regulation of the phosphatidylinositol 3-kinase/Akt pathway is involved in retinoic acid-induced phosphorylation, degradation, and transcriptional activity of retinoic acid receptor gamma 2. J Biol Chem 2002;277:24859–62. Gianni M, Parrella E, Raska Jr I, Gaillard E, Nigro EA, Gaudon C, et al. P38MAPK-dependent phosphorylation and degradation of SRC-3/AIB1 and RARalpha-mediated transcription. EMBO J 2006;25:739–51. Alsayed Y, Uddin S, Mahmud N, Lekmine F, Kalvakolanu DV, Minucci S, et al. Activation of Rac1 and the p38 mitogen-activated protein kinase pathway in response to all-trans-retinoic acid. J Biol Chem 2001;276:4012–9. Dedieu S, Lefebvre P. Retinoids interfere with the AP1 signalling pathway in human breast cancer cells. Cell Signal 2006;18:889–98. Sosa MS, Avivar-Valderas A, Bragado P, Wen HC, Aguirre-Ghiso JA. ERK1/2 and p38alpha/beta signaling in tumor cell quiescence: opportunities to control dormant residual disease. Clin Cancer Res 2011;17:5850–7. Gianni M, Terao M, Fortino I, LiCalzi M, Viggiano V, Barbui T, et al. Stat1 is induced and activated by all-trans retinoic acid in acute promyelocytic leukemia cells. Blood 1997;89:1001–12.

749

[97] Rousseau C, Pettersson F, Couture MC, Paquin A, Galipeau J, Mader S, et al. The N-terminal of the estrogen receptor (ERalpha) mediates transcriptional crosstalk with the retinoic acid receptor in human breast cancer cells. J Steroid Biochem Mol Biol 2003;86:1–14. [98] Schule R, Rangarajan P, Yang N, Kliewer S, Ransone LJ, Bolado J, et al. Retinoic acid is a negative regulator of AP-1-responsive genes. Proc Natl Acad Sci USA 1991;88:6092–6. [99] Tan KP, Kosuge K, Yang M, Ito S. NRF2 as a determinant of cellular resistance in retinoic acid cytotoxicity. Free Radic Biol Med 2008;45:1663–73. [100] Bratton MR, Antoon JW, Duong BN, Frigo DE, Tilghman S, Collins-Burow BM, et al. Galphao potentiates estrogen receptor alpha activity via the ERK signaling pathway. J Endocrinol 2012;214:45–54. [101] Whyte J, Bergin O, Bianchi A, McNally S, Martin F. Key signalling nodes in mammary gland development and cancer. Mitogen-activated protein kinase signalling in experimental models of breast cancer progression and in mammary gland development. Breast Cancer Res 2009;11:209. [102] Wang Q, Wieder R. All-trans retinoic acid potentiates Taxotere-induced cell death mediated by Jun N-terminal kinase in breast cancer cells. Oncogene 2004;23:426–33. [103] Lee HY, Sueoka N, Hong WK, Mangelsdorf DJ, Claret FX, Kurie JM. All-transretinoic acid inhibits Jun N-terminal kinase by increasing dual-specificity phosphatase activity. Mol Cell Biol 1999;19:1973–80. [104] Yu YM, Han PL, Lee JK. JNK pathway is required for retinoic acid-induced neurite outgrowth of human neuroblastoma, SH-SY5Y. NeuroReport 2003;14:941–5. [105] Srinivas H, Juroske DM, Kalyankrishna S, Cody DD, Price RE, Xu XC, et al. C-Jun N-terminal kinase contributes to aberrant retinoid signaling in lung cancer cells by phosphorylating and inducing proteasomal degradation of retinoic acid receptor alpha. Mol Cell Biol 2005;25:1054–69. [106] Gianni M, Terao M, Norio P, Barbui T, Rambaldi A, Garattini E. All-trans retinoic acid and cyclic adenosine monophosphate cooperate in the expression of leukocyte alkaline phosphatase in acute promyelocytic leukemia cells. Blood 1995;85:3619–35. [107] Gianni M, Zanotta S, Terao M, Garattini S, Garattini E. Effects of synthetic retinoids and retinoic acid isomers on the expression of alkaline phosphatase in F9 teratocarcinoma cells. Biochem Biophys Res Commun 1993;196: 252–9. [108] Parrella E, Gianni M, Cecconi V, Nigro E, Barzago MM, Rambaldi A, et al. Phosphodiesterase IV inhibition by piclamilast potentiates the cytodifferentiating action of retinoids in myeloid leukemia cells. Cross-talk between the cAMP and the retinoic acid signaling pathways. J Biol Chem 2004;279:42026–40. [109] Zhao Q, Tao J, Zhu Q, Jia PM, Dou AX, Li X, et al. Rapid induction of cAMP/PKA pathway during retinoic acid-induced acute promyelocytic leukemia cell differentiation. Leukemia 2004;18:285–92. [110] Al-Wadei HA, Schuller HM. Cyclic adenosine monophosphate-dependent cell type-specific modulation of mitogenic signaling by retinoids in normal and neoplastic lung cells. Cancer Detect Prev 2006;30:403–11. [111] Taverna D, Antoniotti S, Maggiora P, Dati C, De Bortoli M, Hynes NE. ErbB-2 expression in estrogen-receptor-positive breast-tumor cells is regulated by growth-modulatory reagents. Int J Cancer 1994;56:522–8. [112] Ombra MN, Di Santi A, Abbondanza C, Migliaccio A, Avvedimento EV, Perillo B. Retinoic acid impairs estrogen signaling in breast cancer cells by interfering with activation of LSD1 via PKA. Biochim Biophys Acta 2013;1829:480–6. [113] Masia S, Alvarez S, de Lera AR, Barettino D. Rapid, nongenomic actions of retinoic acid on phosphatidylinositol-3-kinase signaling pathway mediated by the retinoic acid receptor. Mol Endocrinol 2007;21:2391–402. [114] Cho Y, Tighe AP, Talmage DA. Retinoic acid induced growth arrest of human breast carcinoma cells requires protein kinase C alpha expression and activity. J Cell Physiol 1997;172:306–13. [115] Kambhampati S, Li Y, Verma A, Sassano A, Majchrzak B, Deb DK, et al. Activation of protein kinase C delta by all-trans-retinoic acid. J Biol Chem 2003;278:32544–51. [116] Nakagawa S, Fujii T, Yokoyama G, Kazanietz MG, Yamana H, Shirouzu K. Cell growth inhibition by all-trans retinoic acid in SKBR-3 breast cancer cells: involvement of protein kinase Calpha and extracellular signal-regulated kinase mitogen-activated protein kinase. Mol Carcinog 2003;38:106–16. [117] Pettersson F, Couture MC, Hanna N, Miller WH. Enhanced retinoid-induced apoptosis of MDA-MB-231 breast cancer cells by PKC inhibitors involves activation of ERK. Oncogene 2004;23:7053–66. [118] Kurie JM et al. Clin Cancer Res 1996;2:287–93. [119] Lawrence JA et al. J ClinOncol 2001;19:2754–63. [120] Brown P et al. Cancer Prev Res 2008;1:CN04–04. [121] Esteva FJ et al. J ClinOncol 2003;21:999–1006. [122] Recchia F et al. Oncol Rep 2009;21:1011–6.