Prolactin and endocrine therapy resistance in breast cancer: The next potential hope for breast cancer treatment

Abstract Breast cancer, a hormone‐dependent tumour, generally includes four molecular subtypes (luminal A, luminal B, HER2 enriched and triple‐negative) based on oestrogen receptor, progesterone receptor and human epidermal growth factor receptor‐2. Multiple hormones in the body regulate the development of breast cancer. Endocrine therapy is one of the primary treatments for hormone‐receptor‐positive breast cancer, but endocrine resistance is the primary clinical cause of treatment failure. Prolactin (PRL) is a protein hormone secreted by the pituitary gland, mainly promoting mammary gland growth, stimulating and maintaining lactation. Previous studies suggest that high PRL levels can increase the risk of invasive breast cancer in women. The expression levels of PRL and PRLR in breast cancer cells and breast cancer tissues are elevated in most ER+ and ER− tumours. PRL activates downstream signalling pathways and affects endocrine therapy resistance by combining with prolactin receptor (PRLR). In this review, we illustrated and summarized the correlations between endocrine therapy resistance in breast cancer and PRL, as well as the pathophysiological mechanisms and clinical practices. The study on PRL and its receptor would help explore reversing endocrine therapy‐resistance for breast cancer.


| INTRODUC TI ON
The incidence of breast cancer ranks first among Chinese female malignant tumours. Even in women worldwide, breast cancer is also the malignant tumour with the highest incidence and mortality and accounts for about 11.6% of total cancer deaths. 1 In addition to operation and chemotherapy, endocrine therapy is also a common therapy for breast cancer. According to studies, endocrine therapy is effective in 50%-60% of patients with positive oestrogen receptor (ER); the response rate (RR) of endocrine therapy in patients with positive ER and progesterone receptor (PR) may be >75% and that in patients with negative ER and PR is about 10%. 2 Endocrine therapies for ER + breast cancer include selective ER modulators (SERMs) such as tamoxifen, aromatase inhibitors (AIs) such as anastrozole, ovarian function suspension (OFS) such as goserelin and selective ER downregulators (SERDs) such as fulvestrant. The main reason for endocrine therapy's failure is the primary or secondary resistance in patients during treatment. Primary resistance is common in breast cancer patients with negative ER and PR. Nevertheless, a considerable part of patients with positive ER also has primary resistance.
However, after a certain period of endocrine therapy, almost all patients may have secondary resistance. Therefore, endocrine resistance is a crucial problem to be solved in the treatment of breast cancer.
Prolactin's (PRL) role in breast cancer pathogenesis has been gaining increasing attention. PRL is a protein hormone primarily secreted by eosinophils in the anterior lobe of the pituitary gland.
PRL can stimulate DNA synthesis, epithelial cell proliferation and breast milk production by affecting the cell of origin or neighbouring cells in an autocrine/paracrine way and prolactin receptors. 3 PRLR, a member of the cytokine receptor family, mediates the growth regulation of PRL on the human breast. In breast cancer studies, these data demonstrated widespread expression of PRL and its receptor (>95%). 4 An epidemiological survey has proved that the elevated PRL level in the circulation before the confirmation of breast cancer is correlated to metastatic breast cancer (MBC). 5 Breast cells can promote cell proliferation and inhibit cell apoptosis through autocrine or paracrine PRL. PRLR is found in breast tissue. The expression levels of PRL and PRLR in breast cancer cells and breast cancer tissues are elevated in most ER + and ER − tumours. 6 The study of He Wei et al.
found that PRL can boost the proliferation and growth of human breast cancer T-47D cells and accelerate the transformation of the cell growth cycle and that the effect is dose-and time-dependent. 7 In the Nurse's Health Study and Nurse's Health Study II, Tvoroger et al. investigated the relationship between PRL and breast cancer risk. They measured PRL levels <10 and ≥10 years before the diagnosis of breast cancer. After 20 years of follow-up, the finding has demonstrated an association of prolactin levels <10 years before diagnosis and breast cancer risk of postmenopausal women, especially for ER(+) tumours and metastatic disease. 8 In another case control study, Tikk et al. analysed the association of PRL prediagnostic circulatory levels with the risk of breast cancer through menopause status, HRT therapy with the use of a blood donation system and the hormone receptor status of the breast tumours in 2250 breast cancer cases, the study found that the risk of breast cancer among postmenopausal women was higher because of prolactin overexpression. However, this increased risk was limited to women who used postmenopausal HRT when donating blood. 9 No evidence existed in other studies that combined hormone therapy increased serum PRL levels in hormone replacement therapy patients. 10,11 In addition, knocking down the long PRLR gene can inhibit breast cancer's metastasis in the lung and liver, indicating that the long PRLR gene plays a crucial role in breast cancer metastasis. 12 Researchers have demonstrated in Yonezawa's study that the long PRLR associate with breast cancer metastasis by knocking down the long PRLR in two breast cancer models. In both models, knockdown of long PRLR dramatically inhibited lungs and liver metastasis. 12 Sutherland et al. used quantitative immunohistochemistry to identify the associations between PRLR levels and time to bone metastasis, and they observed that PRL-PRLR can accelerate bone metastasis in breast cancer and the PRLR overexpression in the primary breast tumour results in a shorter time to metastasis. 13 Another study revealed no differences in serum PRL levels among visceral or bone metastases, so further studies are necessary to examine the association of PRL with breast cancer metastasis. 14 All these data indicated that PRL is but not in triple-negative patients. 15 They conducted a larger analysis in GOBO database (1881 patients) to assess PRLR gene expression level and proved that there was a significant association between PRLR expression and the luminal A subtype, but they also found no significant link between PRLR expression and hormone receptors. In the future, the relevance of PRL and molecular subtypes of breast cancer requires further study.
As is known, the independent activation of ER ligand is one of the mechanisms of endocrine therapy resistance. The evidence suggested that PRL may cause endocrine resistance through this mechanism in vitro. 16 The study of O'Leary et al. 17 showed that PRL may also activate ER in the absence of oestrogen ligand in vivo, causing endocrine resistance. In this study, the correlation between PRL and breast cancer endocrine resistance was reviewed.

| ENDOCRINE RE S IS TAN CE MECHANIS MS
Endocrine resistance is mainly divided into primary resistance and secondary resistance. The mechanisms of primary resistance mainly include: induction of apparent silence of ER by histone deacetylation modification, ER gene mutation (such as conversion of ERα351 site tyrosine to aspartic acid), promotion of tamoxifen (TAM) to act as an agonist and tumour growth, gene mutation of ER alpha (ESR) and p21-activated kinase 1 (Pak1) and aromatase gene polymorphism. 17 ESR1 and Pak1 gene mutation play a vital role in the endocrine therapy resistance of breast cancer with positive ER. 18,19 Secondary resistance is correlated to the expression and functional downregulation of ERα, the overexpression of ERβ of breast cancer with positive ERα, ligand-independent activation of growth factor receptor (including epidermal growth factor receptor [EGFR], human epidermal growth factor receptor-2 [HER-2] and insulin-like growth factor 1 receptor [IGF1-R]) or intracellular kinases (such as mitogenactivated protein kinase [MAPK]), and gene transcription caused by ER phosphorylation due to the activation of the downstream signal transduction pathway from protein kinase A (PKA) phosphorylation. Besides, the positive gene of fibroblast growth factor receptor 1 (FGFR1) is also correlated to breast cancer resistance. [20][21][22] In the mechanism of endocrine resistance, ER missing accounts for 15%-20% and ER mutation <1%. 23 Figure 1 briefly depicts the mechanism of endocrine resistance.

| Oestrogen receptor-mediated signalling
The interaction between ER signalling and other receptor-mediated signalling plays a crucial part in the therapeutic process for endocrine resistance. Blocking the oestrogen-mediated signalling and other over-activated signalling may reverse the resistance of endocrine therapy. The combination of Tamoxifen and growth factor receptor kinase inhibitor (RKI) is also one of the main therapeutic approaches for Tamoxifen-resistant breast cancer with overexpression of EGFR or HER-2. In addition, the abnormal expression of receptor tyrosine kinase (RTK), EGFR, HER-2, IGF-1R, fibroblast growth F I G U R E 1 Mechanisms of endocrine resistance. ESR1, oestrogen receptor 1; HER-2, human epidermal growth factor receptor 2; RET, rearranged during transfection factor receptors (FGF-R) and abnormal activation such as PI3K/ PTEN/AKT/mTOR signalling pathway, NF-kB signalling pathway may be involved in drug resistance ( Figure 2). 24

| HER-2
HER-2 is involved in resistance to endocrine therapy. Studies have found that tamoxifen-resistant cells overexpress HER-2. After tamoxifen treatment, the cells still proliferate malignantly, suggesting that HER-2 interacts with ERα. 25 Another study found that breast cancer amplified antigen 1 (AIB1) as an ERα co-regulator when HER-2 is expressed, and its increased expression is related to tamoxifen resistance. 26 YBX1 overexpressing breast cancer cells are resistant to tamoxifen and fulvestrant, which are related to decreased ER and elevated HER-2. Tamoxifen treatment can increase the ability of YBX1 to bind to the HER-2 promoter region, Induction of HER-2 transcriptional activation and increased expression ( Figure 3). 27

| RET activation
Protein arginine N-methyltransferase 2 (PRMT2) is an ERα coregulator, interacts with ERα66 and has the ability to inhibit breast cancer cell proliferation. Shen et al. found that after treating cells with tamoxifen, PRMT2 expression was reduced, ERα36 expression was increased, and tamoxifen resistance was mediated, while PRMT2 directly bound ERα36 and inhibited its activity, blocking PI3K/AKT and MAPK/ERK signalling pathway can reverse tamoxifen resistance. 28 Shimoda et al 29 found that aspartateβ-hydroxylase (ASPH) is related to the sensitivity of endocrine therapy. ASPH expression in tamoxifen-resistant breast cancer cells was up-regulated and MAPK and PI3K signalling pathways are involved in drug resistance regulation.
It has been well documented that ER-positive breast cancers present the functional receptor tyrosine kinase RET signalling activity. Breast cancers that are sensitive to endocrine therapy often lack RET ligands. In support, the RET ligand GDNF has been shown to induce endocrine therapy resistance ( Figure 4). 30 Another study found that oestrogen directly activates C-terminal Src kinase (CSK) expression in ER-positive breast cancer, activates p21-activated kinase2 (PAK2) and causes oestrogenindependent growth. ER-positive breast cancer with PAK2 overexpression is linked with resistance to endocrine therapy and poor prognosis. This study used PAK2 inhibitors and ER antagonists on drug-resistant cells to synergistically inhibit breast cancer growth. 31   involve the loss of oestrogen receptor (ER) alpha expression, which can be achieved by methylation of CpG islands or histone deacetylase activity in the ESR1 promoter. Tamoxifen-resistant growth is also stimulated by the upregulation of growth factor signalling pathways (HER2, IGF-IR and FGFR) and subsequent activation of the mitogen-activated protein kinase (MAPK) cascade or phosphoinositide 3-kinase (PI3K) pathway. Finally, tamoxifen has even been shown to stimulate the growth of breast cancer cells when bound to certain coactivators, such as AIB1, and this is especially true in HER2-expressing cells. (B) The mechanisms of aromatase inhibitor (AI) resistance share similarities with tamoxifen resistance, especially in terms of growth factor pathway upregulation. The enhanced activity of growth factors such as MAPK can result in oestrogen-independent phosphorylation and activation of ERα. In addition to growth factor signalling, interferon response genes and anti-apoptotic proteins have also been shown to have increased expression in AI-resistant cells. AIB1, amplified in breast cancer 1; FGFR1, fibroblast growth factor receptor 1; HER2, human epidermal growth factor receptor 2; IGFR1, insulin-like growth factor receptor 1. Reprinted from [24]. Copyright © 2015 Breast Cancer Research volume oestradiol (E2) hypersensitivity, the increased ability of DNA to bind to oestrogen response elements, and the activity of E2independent constitutive transactivation. Takeshita et al. 33  patients and found the frequency of these mutations was 12%. In preclinical models, these mutations were shown to bring about constitutive activity and relative resistance to endocrine therapy.
They believe that these mutations are responsible for driving endocrine resistance in ER + metastatic breast cancer. 35 Keren et al. Mutations in the CYP19A1 gene, which encodes a member of the cytochrome P450 enzyme superfamily, can result in increased or decreased aromatase activity. In a study by Magnani et al, they found that 21.5% of AI-treated relapsed patients acquired CYP19A1 gene amplification. CYP19A1 amplification not only increased aromatase activity, but also led to oestrogen-independent ERα binding to target genes, triggering CYP19A1 amp cells to show decreased sensitivity to AI treatment. 39 They also found that some patients whose disease progressed after reversible AI treatment occasionally responded to irreversible AI.

F I G U R E 3
Model depicting YBX1mediated resistance to anti-oestrogens of breast cancer cells. In oestrogendependent ER + breast cancer cells, YBX1induced ERBB2 expression is inhibited by YBX1 binding to active ER. Treatment with anti-oestrogens interferes with binding, and free, active YBX1 promotes ERBB2 expression. Reprinted from [27]. Copyright © 2017 Cancer Res F I G U R E 4 Schematic diagram of RET activation in endocrine sensitive and resistant tumours. Both endocrine sensitive and resistant breast cancer cells express all components of the RET signalling pathway, but endocrine sensitive breast cancer cells lack GDNF to initiate the resistance pathway. By contrast, endocrine resistant cells secret GNDF, which acts in an autocrine or paracrine fashion to promote endocrine resistance in nearby cells. Reprinted from [30]. and LY2) proliferation and inhibit LY2 cells' migration and tumour colony formation, but does not increase the sensitivity of LCC9 or LY2 cells to tamoxifen. 43 Studies have found that HOTAIR in lncRNA affects the regulation of HOXD by affecting polycomb repressor complex 2 (PRC2) binding to homeobox D (HOXD) cluster DNA. PRC2 can promote the histone H3K27 trimethylation (H3K27me3), thereby inhibiting transcription, leading to blocked differentiation and increase the metastasis and invasion of breast cancer cells. HOTAIR overexpression is thought to be related to oestrogen response elements in its promoter, and E2 induces HOTAIR expression in breast cancer. HOTAIR is upregulated in tamoxifen-resistant ER-positive breast cancer and further causes tamoxifen resistance. 44,45 Wu et al. 46 found that UCA1 in lncRNA enhanced breast cancer cells' resistance to tamoxifen by inhibiting the mTOR signalling pathway. UCA1 expression was significantly upregulated in tamoxifen-resistant cells, and LCC2 and LCC9 cells transfected with UCA1 siRNA had a higher apoptotic rate after using tamoxifen.
The study also found that UCA1 siRNA significantly reduced pAKT and p-mTOR protein levels in LCC2 and LCC9 cells, and MCF-7 cells over-expressed UCA1 could reduce tamoxifen-induced apoptosis.
The protective effect of UCA1 on tumour cells was weakened after

Hypoxia
Hypoxic microenvironments in solid tumours result from available oxygen being consumed by rapidly proliferating tumour cells, resulting in the oxygen levels around the tumour are significantly lower than in healthy tissues. Studies have also shown that hypoxia is related to endocrine resistance. 55 Hypoxia-inducible factor 1alpha (HIF-1α) can induce cancer cell resistance to tamoxifen and fulvestrant treatment, and its over-expression in ERα + patients is associated with poor survival to endocrine therapy. 56 Morotti et al. suggested that HIF-1αmakes up for the deficiency of ERα expression for keeping up the expression of SNAT2 under hypoxia or endocrine treatments. In vivo, SNAT2 overexpression produces complete resistance to anti-oestrogen therapy and induces tamoxifen resistance. Its expression is related to breast cancer patients' poor survival and resistance to endocrine therapy in ER-receptor + patients. 57 There are also other studies demonstrating that hypoxia and HIF-1α may play a significant role in endocrine treatment resistance. 55,58,59 Cancer-associated fibroblasts Cancer-associated fibroblast (CAF) is a cell type that, by initiating the extracellular matrix's remodelling or secreting cytokines, promotes tumourigenic features. CAFs provide pathways for aggressive cancers and promote invasion and metastasis by the biochemical alteration and cancer-related pathways' regulation. 60 CAFs are the largest stromal cell population in breast cancer and play a key role in breast cancer cells proliferation by producing cytokines and growth factors, remodelling of ECM and modulating immune cell function, further leading to endocrine resistance. 61,62 Brechbuhl et al. analysed the presence of CD146-positive and CD146-negative CAFs in ER + breast cancer patients' tissues and found that CD146-negative CAFs reduce ER expression and tumour cell sensitivity to oestrogen and tamoxifen sensitivity in ER breast cancer cells. They also demonstrated that CD146-negative CAFs are associated with poor drug response to tamoxifen and worse outcome in patients. 63

Extracellular matrix
The extracellular matrix (ECM) consists of extracellular macromolecules and minerals, and researchers have proved that it plays a crucial role in breast cancer progression and endocrine resistance. 64 Jansen et al. 65  demonstrated that CAFs can package intact mitochondrial genome into EVs, which is released and absorbed by dormant CSCs, subsequently transcribing donor mtDNA, contributing to the restoration of oxidative phosphorylation (OXPHOS) and CSCs' self-renovating and the endocrine therapy resistance. Therefore, we believe that EVs play a substantial role in endocrine therapy resistance.

| PRL IN ENDO CRINE THER APY RE S IS TAN CE
Prolactin is a protein hormone secreted by adenohypophysis gland eosinophil. Its major function is to stimulate the growth of breast and induce and maintain milk lactation. Its secretion is modulated by hormones such as hypothalamic prolactin-releasing inhibitory factor (PIF) and prolactin-releasing factor (PRF), and can be self-adjustment by short-loop feedback. Recent studies focusing on prolactin's roles and mechanisms and its receptor in HR-positive breast cancer were listed in Table 1. PRL participates in endocrine therapy resistance mainly by combining with PRLR and activating JAK2-STAT5, ras-raf-MEK-ERK1 /2, Ras-PI3K and other downstream signalling pathways. Figure 6 demonstrates the mechanism of prolactin in endocrine resistance.

| MicroRNA-339-5p-dependent pathway
Studies have found that micro RNA339-5p (miR-339-5p) target the gene B-cell lymphoma 6 (BCL6) and the expression of BCL6 protein was related to the breast cancer progression. 80 Hong Yan et al.

F I G U R E 7
Pathway circuitry dictates therapeutic response. (A) For tumours with defined genetic lesions, the ability to overcome a given targeted therapeutic lies in whether or not they need to acquire a secondary genetic mutation to overcome the effect of the drug on critical downstream biochemical effectors that are required for continued tumour cell growth, or whether they can simply upregulate existing alternative routes that lead to effectors already expressed in those cells. So, the drug places selection pressure to ramp up existing bypass routes. If there are no such routes to the critical downstream effectors, a specific mutation to upregulate those alternative routes or bypass the drug are required. In this example, a critical target for tumour cell growth and survival is the activation of eIF4E and HIF. Tumours with initiating mutations in RTKs, Ras or Raf have multiple routes to signal to eIF4E and HIF, so blocking mTOR with rapamycin does not inhibit these tumours. (B) In contrast, tumours with initiating lesions in PI(3)K or more direct regulators of mTOR (such as LKB1 and TSC) do not have alternative routes to activate eIF4E and HIF. Hence these tumours show greater response to rapamycin. (C) Similarly, the expression and use of specific adaptor proteins that enhance certain arms of pathway signalling will dictate the therapeutic response. In the example shown, human lung tumours expressing epidermal-growth-factor receptor (EGFR) are targeted with anti-EGFR drugs such as Iressa or Tarceva. In tumours expressing the ERBB3 heterodimerization partner, EGFR efficiently enhances PI (3) These findings indicate that CAML is thought to be a promising target for controlling breast cancer progression.

| ADVAN CE S IN ENDO CRINE THER APY RE S IS TAN CE
With the proposal of more and more endocrine therapy resistance mechanisms, more drugs can be used for the treatment of endocrine resistance. For instance, the increased activity of AKT/mTOR pathways is the main mechanism of the resistance to letrozole and flurvesant; and the mTOR inhibitor everolimus can block PI3K/AKT/ mTOR pathway. Preclinical studies showed that the median overall survival (OS) in tamoxifen + everolimus and exemestane + everolimus groups is longer than that in the control group and that the combination of mTOR inhibitor and endocrine therapy can be used as a therapy for patients with aromatase inhibitor resistance. 93 Cyclin-dependent kinase (CKD) is a new target for treating breast cancer. CDK4 /6 inhibitor arrests cell cycle. The FDA has approved Palbociclib and ribociclib as first-line therapy for advanced hormone receptor-positive breast cancer. TGFβ type 2 receptor impairs the oestrogen response and gives tamoxifen resistance. 94 The dual inhibitors for TGFβ and MAPK are being studied. Such inhibitors may be effective in patients with endocrine therapy resistance. 95  showed no antitumour activity as a monotherapy. 106 Studies showed that the new PRLR antibody drug conjugates (ADCs) rapidly release the cytotoxic drugs carried by ADCs into the intracellular lysosomes; and the active drugs released can also infiltrate from target cells into adjacent antigen-negative cells, which may cause the so-called bystander killing. The results showed that ADCs have antitumour activity in vivo and in vitro. 107 Some recently completed and ongoing clinical trials on breast cancer treatment targeting prolactin and its receptors were listed in Table 2.

| CON CLUS ION
In conclusion, PRL is closely correlated to the tumourigenesis, development and endocrine therapy resistance of breast cancer.
Therefore, inhibiting PRL can enhance the sensitivity of breast cancer cells to drugs. However, there are no clinical data on whether the reduction of PRL level affects the patients with endocrine resistance in breast cancer. Currently, PRL targeted therapies are still at the experimental stage and face many challenges before real clinical application. Therefore, keep exploring the role of PRL and its targeted therapies in the mechanism of endocrine therapy resistance in breast cancer plays a significant role in its becoming a method to overcome endocrine therapy resistance clinically. It is likely to become a very promising new therapeutic target for endocrine therapy for breast cancer.

ACK N OWLED G EM ENTS
Not applicable.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests. Supervision (equal). Yen-Hua Huang: Supervision (equal).

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing not applicable to this article as no datasets were generated or analysed during the current study.