Background
Breast cancer is the most common cancer of woman in many countries [
1]. Human breast carcinomas represent a collection of diverse tumors that vary in their natural history and responsiveness to therapy. Because of its heterogeneity, many tumor tissue biomarkers have been identified and used in breast cancer diagnosis to classify subsets, indicate specific therapies and predict tumor behavior. Today, biomarkers such as estrogen receptor (ER), progesterone receptor (PR), p53, Ki-67 and human epidermal growth factor receptor type 2 (HER2) guide treatment decisions and prognosis.
HER2 is a proto-oncogene and is a member of the
HER gene family, which includes HER1 (epidermal growth factor receptor, EGFR/erbB1), HER2, HER3 (erbB3) and HER4 (erbB4). The receptor is amplified or overexpressed (or both) in approximately 18–20% of breast cancers [
2]. HER2 promotes cell proliferation and angiogenesis and inhibits apoptosis via several pathways [
2], and HER2-positive status is a negative prognostic factor [
3,
4]. Treating HER2-positive breast cancer with anti-HER2 monoclonal antibodies, such as trastuzumab, has markedly improved the outcome of this disease [
5]. One major challenge to targeted therapy, however, is acquired and primary resistance. Acquired resistance eventually develops in most patients in the advanced disease setting [
6]. Advances in molecular biology have led to the identification of potential markers of prognostic and therapeutic importance in breast cancers.
Methylation of histones by protein arginine methyltransferases (PRMTs) is increasingly being acknowledged as an important aspect for the dynamic regulation of gene expression. CARM1 (coactivator-associated arginine methyltransferase 1) is a kind of type I protein arginine methyltransferase that catalyzes the formation of asymmetric dimethylarginine [
7]. It initially was described as a transcriptional activator of the p160 family of nuclear receptor-associated proteins [
8]. The p160 family includes steroid receptor coactivators-1(SRC-1), SRC-2 and SRC-3/AIB1 (amplified in breast cancer 1) [
8]. CARM1 functions as a coactivator for many nuclear receptors (NRs) [
9,
10], including ERα [
11]. ERα plays a pivotal role in promoting the proliferation of several types of estrogen-stimulated breast cancer [
12]. CARM1 has also been shown to be a molecular switch that controls multiple classes of gene-specific transcription factors, including p53, NF-κB, LEF1/TCF4, E2Fs, and cyclin E1 [
9,
13‐
16]. These suggest this enzyme plays pleiotropic roles in cell proliferation and survival. Some researchers had investigated the expression of CARM1 in many kinds of malignant tumors [
17‐
19]. Aberrant expression of CARM1 has been linked to human breast cancer tissue in a few reports [
13,
16,
17]; however, current studies are contradictory and incomplete. The mRNA level of CARM1 was found to be elevated in grade 3 breast tumors in a cohort of 81 human breast carcinomas of various types [
16]. While another study demonstrated there was inverse correlation between CARM1 expression and tumor grade in ER + and LN-breast cancer cases [
13]. Kim YR et al. reported CARM1 overexpression was noted only in small number of breast cancer patients (27%) [
17]. All these reports suggest CARM1 is an important factor involved in progression and may affect prognostication of breast cancer. However, many of these studies were limited either by low n values of breast cancer patients or by a special tumor type. It still remains unclear whether CARM1 expression is correlated with clinicopathological features, molecular subtype and prognosis.
The aim of this study was to characterize the CARM1 expression pattern in invasive breast carcinoma and to analyze its relationship with clinicopathologic characteristics, including the expression of ER, PR, HER2, p53 and Ki-67 index. Additionally, we compared the expression of CARM1 in different molecular subtypes to assess its potential value in improving patient stratification and guiding personal patient management.
Discussion
Breast cancer is a heterogeneous disease encompassing multiple subgroups with differing molecular signatures, prognoses, and responses to therapies [
27]. Other reasons for heterogeneity may include differences in the studied population (e.g., ethnicity, menopausal status), or it might be due to interaction with other risk factors (e.g., BRCA variants) [
28]. El FH found that molecular classification and biological profile may be different according to geographical distribution [
29]. Zhang Q’s study showed the ectopic expression of BRCA1 was associated with the genesis, progression, and prognosis in young breast cancer patients [
30]. Finding of the sources of heterogeneity would contribute to patient’s stratification and personalized treatment. Our study demonstrated that nuclear CARM1 expression was associated with a younger age at diagnosis; a higher tumor grade; a higher rate of HER2, p53, and ki-67 expression; and a lower rate of ER and PR expression in breast cancer patients of Chinese women. We also found that nuclear CARM1 expression was significant different among four molecular subtypes.
In this study, we demonstrated that CARM1 expression, both in the cytoplasm and the nucleus, were more remarkable in younger patients than in patients older than 40 years of age. Young age at the time of diagnosis of breast cancer is an independent factor of poor prognosis for reasons that are not fully understood [
31,
32]. Some studies have demonstrated breast cancer at a younger age might be associated with higher grade, ER-negativity, a more advanced stage of the disease, ectopic expression of BRCA1 [
30], and higher levels of HER2. Previously, it has been shown that embryonic stem cells overexpressing CARM1 were more resistant to differentiation [
33]. In our study, the CARM1 overexpression in younger patients was more common compared with older patients, and might contribute to the clinical characteristics of younger patients, such as lower differentiation. However, we believe it may not be the only factor.
Regarding the subcellular distribution of CARM1, it is predominantly localized in the nucleus as a transcriptional coactivator [
34]. That is to say, most of the known functions of CARM1 are related to its nuclear localization. A few studies also revealed that CARM1 accumulates in the cytoplasm during mitosis [
14,
34]. CARM1 S217E mutant protein and a small percentage of wild-type CARM1 are also localized in the cytosol [
34]. This suggested CARM1 may play an unknown function in the cytoplasm [
34]. Because nucleus is the principal subcellular localization of CARM1 activation, we will focus on nuclear CARM1 expression hereinafter.
Our results indicated that CARM1 expression positively correlated with HER2 expression and grade, and negatively correlated with hormone receptors in separate analyses with universal molecular makers. The molecular subtypes were also classified according to a panel of ER, PR, and HER2 biomarkers combined with grade in our study [
31]. We will discuss the correlation between these factors and CARM1 expression in different molecular subtypes.
The luminal A subtype, known as the hormone subtype, showed the lowest rate of CARM1 expression compared with that in the other subtypes. Previous reports have shown that CARM1 plays an essential role in estrogen-mediated transcriptional activation [
35,
36], and is necessary for the estradiol (E2)-induced proliferation of breast cancer cells [
15]. The co-regulator requirement of CARM1 can be highly tissue- and context-dependent [
8,
13,
37]. Furthermore, CARM1 transcriptional coactivating functions are not restricted to nuclear receptors [
16,
33]. Consistent with this concept, our result also shows only a small part of luminal A subtype cancer tissue overexpress nuclear CARM1.
Our study demonstrated that CARM1 overexpression in breast cancer was associated with the overexpression of HER2. Both HER-2 subtype (69.6%) and luminal B subtype (59.6%) showed higher rate of CARM1 expression compared with that in luminal A tumors. Within luminal B subtype, HER2 signaling is dominant, as demonstrated by the poor response of such tumors to endocrine therapy alone. HER2 overexpression confers intrinsic or primary resistance to hormone-based therapy despite the presence of hormone receptors [
38,
39]. In this study, we showed that there was a strong positive correlation between CARM1 and HER2 expression. This suggests that CARM1 may be useful as a predictor of clinical outcomes in patients with HER2-positive tumors.
The mechanism of CARM1 and HER2 interaction, or through which pathway they crosstalk, has not yet been elucidated. One potential mechanism may be the transcriptional coactivation mediated by CARM1 and p160. CARM1 was initially described as a transcriptional coactivator of the p160 nuclear receptor family [
8]. All members of the p160 family are natural substrates of CARM1, which can bind and recruit CARM1 to synergistically exert transcriptional co-activating functions of target genes [
40]. Multiple studies have demonstrated that AIB1 (a member of the p160 family) mRNA and protein expression in breast cancer is associated with the expression of HER2. AIB1 was shown to play a role in the regulation of the HER2 pathway [
41]. Although AIB1 expression was not examined in our study, data from El Messaoudi S et al. showed that AIB1 and CARM1 mRNA levels were both elevated in breast cancer, notably in grade 3 [
16]. Our finding that CARM1 overexpression in breast cancer correlated with high HER2 expression supports the hypothesis that CARM1 might play an important role in the regulation of the HER2 pathway, probably through transcriptional coactivation with the p160 family.
TN breast cancer is associated with poor prognosis because it lacks the benefit of specific therapy. The TN subtype group encompasses a number of distinct entities with defined gene expression profiles and outcomes [
39]. Our results revealed that over half of the TN tumors overexpressed nuclear CARM1. These results suggest that monitoring the level of CARM1 expression might be valuable to distinguish different entities of TN tumors. Future research should explore this hypothesis as well as its clinical applicability
In summary, based on previous work and the results presented here, CARM1 may promote tumor cell growth by activating nuclear receptors and multiple growth factor signaling cascades in breast cancer. However, the predominance of which pathway is regulated by CARM1 depends on the tumor phenotype. CARM1 requires its enzymatic activity for all of its known nuclear functions. Thus, specific and potent small molecule inhibitors of CARM1 will incapacitate all of its nuclear functions [
42]. Additionally, chemotherapeutic drugs targeted at CARM1 will likely interfere with several pathways concomitantly. Therefore, targeted therapy of CARM1 is promising and warrants further exploration.
One limitation of the study is that we didn’t know the relationship of CARM1 overexpression and the prognosis of breast cancer. It still remains unclear about the accurate regulative pathways mentioned above in cancer cells. All these need to be explored by further study.
Competing interests
The authors declare no conflict of interest.
Authors’ contributions
HC collected clinical data, evaluated the immunohistochemical stainings, performed the statistical analyses and drafted the manuscript. YQ assisted with the design of the study, evaluation of the immunohistochemical stainings and pathological diagnosis. HF carried out the immunoassays and fluorescent in situ hybridization analysis. PS assisted with the immunoassays and the collection of clinical data. XZ and HZ were involved in pathological diagnosis and evaluated the immunohistochemical stainings. GZ conceived the study, was involved in the design, and edited the manuscript for intellectual content. All authors read and approved the final manuscript.