Introduction

ErbB-2, a member of the ErbB family of receptor tyrosine kinases (EGF-R/ErbB-1, ErbB-2, ErbB-3, and ErbB-4), is a major player in the breast cancer (BC) scenario. Approximately 15–20% of diagnosed BC show either membrane overexpression of ErbB-2 (MErbB-2) or ERBB2 gene amplification [1]. In the molecular classification of BC, this subtype is called ErbB-2-positive [1]. Until the development of anti-ErbB-2 agents, this BC subtype was associated with higher metastatic potential and poor prognosis [1]. Notably, the overall survival and cure rates were improved by said agents [2, 3]. Current therapeutic options for patients with ErbB-2-positive BC include monoclonal antibodies (trastuzumab and pertuzumab), tyrosine kinase inhibitors (TKI, lapatinib and neratinib), and trastuzumab-DM1, an antibody-drug conjugate. Despite a significant clinical response to trastuzumab (TZ), the first approved anti-ErbB-2 therapy, nearly 40–60% of patients with ErbB-2-positive metastatic BC do not respond to TZ, showing either intrinsic or acquired resistance [2, 4]. Although lapatinib provides clinical benefit to patients progressing on TZ, less than 25% achieve an objective response, and the majority develop lapatinib resistance [5]. Phase 3 TH3RESA trial, the study performed in metastatic BC patients who had previously received regimens with both trastuzumab and lapatinib, demonstrated that although trastuzumab-DM1 improved the progression-free survival in this cohort, only 31% of the patients achieved an objective response [6]. These data reveal that resistance to ErbB-2-targeted therapy remains a major clinical issue. Multiple mechanisms were found to be involved in TZ resistance, such as hyperactivation of the PI3K/AKT pathway, presence of ligand-induced ErbB-2/ErbB-3 heterodimers or of ErbB-2/insulin-like growth factor receptor I (IGF1-R) dimerization, and expression of constitutively active truncated ErbB-2 forms (reviewed in [7]). Many clinical trials are assessing the efficacy of new treatments with PI3K/AKT inhibitors, anti-IGF-1R antibodies, and TKI afatinib [8]. A better understanding of the ErbB-2 pathway constitutes a critical task in the battle against BC. A major contribution to the understanding of ErbB-2 biology was the demonstration that MErbB-2 migrates to the nucleus (nuclear ErbB-2 (NErbB-2)), where it modulates gene expression and biological responses in BC [9]. Here, we will review NErbB-2 function in BC and will particularly discuss the role of NErbB-2 as a novel target for therapy in ErbB-2-positive BC.

ErbB-2 canonical action

ErbBs ligands include all isoforms of heregulins (HRG), which bind to ErbB-3 and ErbB-4 and recognize EGF-R and ErbB-2 as co-receptors [10]. Upon HRG binding, ErbBs form homo- and heterodimers. This stimulates their tyrosine kinase activity and results in the activation of signaling pathways which transduce ErbB’s biological effects. Although ErbB-2 is an orphan receptor, it dimerizes with other ErbBs [11]. In MErbB-2-overexpressing BC cells, ErbB-2 also forms homo- and heterodimers in the absence of ligand [12]. ErbB-3 critical role as ErbB-2 coreceptor mediating ligand-independent and -dependent ErbB-2 signaling has been reported [13,14,15,16]. Among ErbB-2-activated cascades, the best characterized are the p42/p44 mitogen-activated protein kinases (MAPKs) and the phosphatidylinositol-3-kinase (PI3K)/AKT pathways [17]. It has also been reported that ErbB-2 activates the signal transducer and activator of transcription 3 (Stat3) pathway. Indeed, Stat3 activation by ErbB-2 signaling drives BC growth and metastasis [16, 18]. Figure 1a illustrates the signaling pathways activated by ErbB-2 canonical action.

Fig. 1
figure 1

Canonical and non-canonical actions of ErbB-2 in breast cancer (BC). a ErbB-2 canonical signaling pathway. Upon ligand binding, ErbBs form homodimers or heterodimers, which activate downstream signaling cascades and transduce ErbBs effects. Heregulin-induced ErbB-2/ErbB-3 heterodimers are illustrated. The two key signaling pathways activated are p42/p44 mitogen-activated protein kinases (MAPKs) and phosphatidylinositol 3-kinase (PI3K)/AKT. Activation of the MAPK pathway leads to the transcription of genes that drive cellular proliferation, migration, and angiogenesis, while activation of the PI3K/AKT pathway leads to downstream cellular endpoints including cell survival and apoptosis suppression (reviewed in [17]). Altogether, activation of these signaling cascades promotes BC growth and metastasis. b ErbB-2 non-canonical action. ErbB-2 non-canonical action involves nuclear translocation of membrane ErbB-2. Heregulin-induced ErbB-2/ErbB-3 heterodimerization and nuclear translocation is depicted [19]. Once in the nucleus, ErbB-2 may directly bind to its response elements and function as a transcription factor (TF), or may tether to other TFs and act as a transcriptional coactivator. An example of ErbB-2 acting as a transcriptional coactivator is shown. NErbB-2 actions also drive BC growth and metastasis

Full size image

ErbB-2 non-canonical mechanism of action

Multiple pieces of evidence indicate that ErbB-2 is present in the nucleus, where it acts as a transcriptional regulator to modulate biological responses in BC [9, 19,20,21,22,23,24,25,26,27] (Fig. 1b). Moreover, clinical studies identified NErbB-2 positivity as an independent prognostic factor of poor clinical outcome in patients bearing MErbB-2-positive tumors [28]. The molecular mechanisms underlying ErbB-2 nuclear trafficking have been extensively investigated and reviewed elsewhere [29,30,31]. ErbB-2 contains a nuclear localization sequence (NLS, 676-KRRQQKIRKYTMRR-689) located adjacent to the transmembrane domain, which is crucial for NErbB-2 trafficking [32]. Indeed, during ErbB-2 nuclear transport, it associates with importin β1 (also called karyopherin β1: Kpnβ1) through the ErbB-2-NLS. Thereafter, the interaction of this ErbB-2/importin β1 complex with the nuclear pore protein Nup358 leads to ErbB-2 nuclear translocation [32]. In spite of multiple findings highlighting NErbB-2’s role in BC and providing insights into the mechanisms of ErbB-2 nuclear trafficking, there is no drug designed to specifically block nuclear ErbB-2 expression and/or function. Interestingly, a recent study showed that Kpnβ1 knockdown inhibited NErbB-2 presence and in vitro proliferation of MErbB-2-overexpressing BC cells [33]. Moreover, the authors demonstrated that Kpnβ1 expression was associated with poor prognosis and identified Kpnβ1 as a possible novel therapeutic target [33]. Further investigations are required to test the effects of Kpnβ1 knockdown in BC growth. Since Kpnβ1 is involved in the nuclear transport of other proteins such as the ribosomal proteins [34], it also remains to be explored if Kpnβ1 knockdown could affect nuclear transportation of other proteins in non-malignant cells.

Molecular and biological functions of NErbB-2 in breast cancer

The dogma of ErbB-2 mechanism of action has been challenged by the demonstration that MErbB-2 migrates to the nucleus where it binds DNA at specific sequences, called HER-2-associated sequences (HASs) [9]. In this pioneering work, ErbB-2 tyrosine phosphorylation was required for the nuclear localization of the full-length ErbB-2 protein. Cyclooxygenase-2 (COX-2) expression was modulated via NErbB-2’s role as a transcription factor (TF) [9]. Inhibition of COX-2 activity suppressed the invasive activity of ErbB-2-overexpressing BC cells [9], suggesting that the role of NErbB-2 as a TF may induce BC metastasis. Since ErbB-2 lacks a putative DNA-binding domain, it was proposed that NErbB-2 acts as a coactivator of other TFs. NErbB-2’s function as a transcriptional coactivator was identified in BC models that express estrogen (ER) and progesterone receptors (PR) and overexpress MErbB-2 [20]. Bidirectional interactions between PR and MErbB-2 classical signaling pathways had long been investigated in those BC models [15, 35]. Stat3 was revealed as the convergence point between said pathways [16, 36]. A novel class of bidirectional interaction between PR and ErbB-2 was described at the nucleus. Indeed, PR induced the nuclear translocation of full-length ErbB-2 which acts as a coactivator of the TF Stat3 bound to its response elements (GAS sites) at the cyclin D1 (CCND1) promoter [20]. Interestingly, the PR was also assembled along with Stat3 and ErbB-2 in this transcriptional complex. The biological action of NErbB-2 in ER-, PR-, and ErbB-2-positive BC models was investigated by transfection of cells with hErbB-2ΔNLS, a human ErbB-2 nuclear localization domain mutant engineered by Hung and co-workers, which is unable to translocate to the nucleus and functions as a dominant negative inhibitor of endogenous ErbB-2 nuclear migration [20, 32]. The hErbB-2ΔNLS mutant retains ErbB-2 intrinsic tyrosine kinase activity [9, 20, 32]. Notably, blockade of NErbB-2 action by transfection with hErbB-2ΔNLS disrupted the assembly of nuclear Stat3/ErbB-2/PR complexes at the CCND1 promoter and inhibited progestin-induced BC growth [20]. Progestins also induced the assembly of nuclear Stat3/ErbB-2/PR complexes to the GAS sites at the p21CIP1 promoter, thus inducing p21CIP1 expression and growth of PR-positive, MErbB-2-overexpressing BC cells [27]. Another study provided further evidence of the role of NErbB-2 as coactivator of transcription factors in BC growth. Indeed, it was demonstrated that progestins induce the assembly of a transcriptional complex among activator protein 1 (AP-1), Stat3, PR, and ErbB-2 at a region of the CCND1 promoter containing AP-1 response elements (TRE) and GAS sites [26]. It was also reported that NErbB-2 activates transcription of ribosomal RNA genes through its association with RNA polymerase-I and β-actin at the ribosomal DNA (rDNA), leading to ribosome biosynthesis [21]. In this article, the authors demonstrated that NErbB-2 function increases total protein synthesis and cell size, indicating that it may play a role in BC growth [21]. A recent study revealed that NErbB-2, in its role as a Stat3 coactivator and also in its direct role as TF, upregulates microRNA-21, a microRNA regulated by membrane ErbB-2 action [37], and promotes BC metastasis [23]. This reveals a novel NErbB-2 function as a regulator of microRNAs expression. Enhanced levels of microRNA-21 inhibited the expression of programmed cell death 4 (PDCD4) [23], which is a metastasis-suppressor protein and also a well acknowledged microRNA-21 target gene [37]. Consistent with a startling study [9], these findings support NErbB-2’s role in BC metastasis. Table 1 summarizes NErbB-2’s molecular and biological functions described. Although most of the reports indicate that NErbB-2 acts as a transcriptional regulator, ErbB-2 may also exert its tyrosine kinase activity in the nucleus. Indeed, ErbB-2 phosphorylates the cyclin-dependent kinase Cdc2, leading to inhibition of Cdc2 activation and resistance to taxol-induced apoptosis in MErbB-2-overexpressing BC cells [24]. Notably, NErbB-2 colocalized with the Cdc2/cyclin B complexes, raising the possibility that NErbB-2 function as a protein kinase may contribute to Cdc2 inhibition and to taxol resistance [24, 38].

Table 1 Molecular and biological functions of Nuclear ErbB-2 in breast cancer
Full size table

NErbB-2 role in response to anti-MErbB-2 therapies

Full-length nuclear ErbB-2 was found to be involved in the mechanisms of BC resistance to anti-MErbB-2 agents. Basal levels of nuclear ErbB-2, ErbB-3, and Stat3, as well as of nuclear ErbB-2/ErbB-3 and nuclear ErbB-2/Stat3 complexes were detected in trastuzumab (TZ)-sensitive and -resistant ErbB-2-positive BC lines [19]. Heregulin treatment further enhanced nuclear migration and colocalization of all three proteins. Interestingly, basal MErbB-2/MErbB-3 dimers were more numerous in TZ-responsive than in -resistant cells, whereas nuclear dimers were more abundant in resistant cells, highlighting a role of nuclear ErbB dimers in the response to anti-ErbB-2 agent TZ [19]. Moreover, heregulin induced the assembly of a Stat3/ErbB-2/ErbB-3 transcriptional complex at the GAS sites of the CCND1 promoter, where both ErbB-2 and ErbB-3 act as Stat3 coactivators [19] (Table 1). This study identifies the first nuclear function of ErbB-2/ErbB-3 dimers. Furthermore, blockade of NErbB-2 action by transfection with the hErbB-2ΔNLS mutant abrogated growth of BC cells, sensitive and resistant to TZ, in a scenario in which ErbB-2/ErbB-3 dimers are formed and the PI3K/AKT pathway is activated [19], conditions where TZ is inefficient [39,40,41,42]. Disruption of the Stat3/ErbB-2/ErbB-3 nuclear complex driving CCND1 expression was revealed as the differential molecular signature underlying hErbB-2ΔNLS growth inhibitory effects in TZ-resistant cells [19]. All these findings identified NErbB-2 as a major proliferation driver in TZ-resistant BC and highlighted NErbB-2 as a novel target to overcome TZ resistance. A series of truncated ErbB-2 variants known as p95ErbB-2 were both found at the cytoplasm and the nucleus of BC cells [43, 44]. Although the exact molecular mechanism by which these NErbB-2 fragments exert their action remains to be established, it has been shown that nuclear p95ErbB-2 variants induce BC growth [43, 44] and are also involved in the response to anti-MErbB-2 therapies. Indeed, treatment of BC cells with lapatinib increased the expression of a tyrosine phosphorylated p95ErbB-2 form (p95L) located in the nucleus, which rendered these cells resistant to lapatinib [45]. Furthermore, enhanced expression of nuclear p95L was detected in biopsies from metastatic BC sites that had developed while patients were on lapatinib therapy [45], indicating a role of NErbB-2 fragments in acquired resistance to lapatinib. However, these findings need to be reconciled with other reports which found that the antiproliferative effects of lapatinib are mediated by its ability to inhibit phosphorylation of membrane ErbB-2 and its nuclear migration [25, 46]. ErbB-2 activating mutation at leucine 755 (L755), which is located at the ErbB-2 kinase domain and is near a putative nuclear export signal, has been associated with lapatinib resistance [47]. Interestingly, ectopic expression of ErbB-2 L755 mutants (L755P and L755S) enhanced NErbB-2 localization in ErbB-2-positive BC cells and increased their ability to form mammospheres in vitro, a key feature of cancer stem cells, highlighting NErbB-2 function in cancer stemness [48]. Although the mechanistic relationship among NErbB-2, lapatinib resistance, and cancer stemness needs to be elucidated, this work further supports the importance of targeting NErbB-2 presence or function to overcome resistance to anti-ErbB-2 therapies.

Conclusions

Accumulating findings have proven ErbB-2 localization and function at the nucleus of ErbB-2-positive BC cells. NErbB-2 is involved in BC growth and metastasis, both key features of this disease. NErbB-2 is also involved in BC de novo and acquired resistance to ErbB-2-targeted therapies. Most importantly, blockade of NErbB-2 presence abrogates BC growth in ErbB-2-positive BC, revealing NErbB-2 as a novel therapeutic target. Although both full-length ErbB-2 molecules and fragments were identified in the nucleus of BC cells, the exact molecular mechanisms by which NErbB-2 fragments exert their role remains to be established. An open question that also needs to be confirmed is whether ErbB-2 exerts its tyrosine kinase activity in the nucleus. In light of the key role of NErbB-2 in the resistance to anti-MErbB-2 agents, the clinical significance of NErbB-2 in response to these therapies should be explored. The identification of NErbB-2 as a novel biomarker for predicting responses to anti-ErbB-2 therapies at diagnosis, allowing the identification of patients who will fail to respond to these therapies, will be extremely important to modify the current therapeutic protocols and include NErbB-2-targeted therapy. Finally, research efforts should be focused on developing drugs designed to target NErbB-2 function.