Progress of Breast Cancer basic research in China

Breast cancer is the most commonly diagnosed and the most lethal cancer in females both in China and worldwide. Currently, the origin of cancer stem cells, the heterogeneity of cancer cells, the mechanism of cancer metastasis and drug resistance are the most important issues that need to be addressed. Chinese investigators have recently made new discoveries in basic breast cancer researches, especially regarding cancer stem cells, cancer metabolism, and microenvironments. These efforts have led to a deeper understanding of drug resistance and metastasis and have also indicated new biomarkers and therapeutic targets. These findings emphasized the importance of the cancer stem cells for targeted therapy. In this review, we summarized the latest important findings in this field in China.


Introduction
Breast cancer is the most common malignant disease in women worldwide. According to the data released in 2019, breast cancer accounted for 30% of newly diagnosed malignant tumors in females, and led to 15% of female deaths from cancer [1]. The incidence and mortality of breast cancer are still rising, both in developing and developed countries [2]. In China, breast cancer morbidity and mortality rank first among malignant diseases in females. Death rate reached about 6.9% among all female malignant tumors [3]. The lack of early-stage screening and detection methods and cost-effective therapies makes breast cancer as one of the most severe disease burden globally [4]. According to the international consensus guideline, the current treatment for breast cancer mainly included chemotherapy, radiotherapy, targeted therapy, immunity therapy, and endocrine therapy before and after surgery [5]. Nowadays, therapies targeted to cancer stem cells have been a hot spot in breast cancer treatment, which is a supplement to the traditional chemotherapeutic drugs that are unable to eradicate tumor dissemination and metastasis [6,7]. Great advances in cancer cell-oforigin, somatic mutation, epigenetic alteration, and tumor microenvironment were revealed in recent studies, both basic and clinical, leading to a better understanding of the mechanism, the pathogenesis, the diagnosis, and the treatment of breast cancer. Particularly, Chinese investigators have made great achievements in cancer stem cells, microenvironment, metastasis, drug resistance, tumor biomarkers, and new targets, as well as drugs for breast cancer therapy. In this review, we summarized the important findings on breast cancer basic research in China from 2019 to 2020.

Cancer stem cells (CSCs)
CSCs are defined as a group of undifferentiated cells that possess properties of self-renewal and pluripotent differentiation. Although breast cancer stem cells (BCSCs) make up only a small subpopulation within tumors, they are responsible for tumor initiation, recurrence, metastasis, and therapy resistance [8]. Previous studies revealed a panel of Ivyspring International Publisher biomarkers for BCSCs identification, among which CD44 high , CD24 low and aldehyde dehydrogenase 1 (ALDH1) + were the most commonly used biomarkers [9]. Triple-negative breast cancer (TNBC) contains more BCSCs than other subtypes, which is associated with worse outcomes [10,11]. In 2019, protein C receptor (PROCR) was identified as a novel marker for BCSCs, which could stratify TNBC into clinically relevant subgroups and be a therapeutic target (Fig. 1) [12]. NOTCH4 receptor was not only implicated in the regulation of BCSCs in TNBC, but also regarded as a better marker for BCSCs than CD24 -CD44 + in 2020 ( Fig.  1). Mechanistically, NOTCH4 drives mesenchymal-like BCSCs into a quiescent state and induces epithelial-mesenchymal transition (EMT) via upregulating GAS1 and SLUG, respectively. Zhou et al. also demonstrated the importance of NOTCH4-SLUG-GAS1 circuit in maintaining mesenchymal-like BCSCs [13].
Tumor microenvironment (TME) plays an essential role in regulating the activity of BCSCs [14]. Tumor-associated macrophages (TAMs) with elevated LSECtin expression interacts with its receptor BTN3A3, expressed in cancer cells, to promote the stemness of breast cancer cells in a cell-cell contact dependent manner ( Fig. 1) [15]. In addition, the Jag1 expression in endothelial cells activates Notch 1 to upregulate Zeb1 expression and to increase vascular endothelial growth factor A (VEGFA) production. In turn, VEGFA induces Jag1 expression in endothelial cells. Thus, the positive feedback loop in the tumor perivascular niche promotes the stemness of breast cancer cells [16].
transducer as well as activator of transcription 1 (STAT1) expression in basal-like breast cancer (BLBC) [20]. Additionally, loss of SIRT4 promoted self-renewal and expansion of BCSCs through suppressing glutamine metabolism in mitochondrial and SIRT1-mediated BRCA1 transcription in nucleus, which provided a novel cross-talk between mitochondrial and nuclear sirtuins [21].
BCSCs are divided into two categories according to the cell cycle rate: energetic BCSCs (e-BCSCs) [25] and quiescent BCSCs (q-BCSCs) [26]. Q-BCSCs play critical roles in resistance to chemoradiotherapy and disease relapse. SET domain-containing protein 4 (SETD4) is important in the maintenance of qBCSCs (Fig. 1). SETD4 facilitates heterochromatin formation via H4K20me3 (trimethylation of lysine 20 of histone 4) catalysis on certain promoter regions which leads to the silencing of qBCSCs-suppressing genes. Notably, SETD4-defined qBCSCs maintain quiescent state by asymmetric division, producing a small qBCSC and a big active daughter cell; the latter generates normal cancer cells [27]. These findings on stem cell status expanded our knowledge on the epigenetic determinants of q-BCSCs and provided new therapeutic targets for drug-resistant q-BCSCs.

Metabolism
Altered metabolism is an emerging hallmark of cancer. Unlike normal cells, cancer cells prefer aerobic glycolysis, which is accompanied by increased lactate production, also known as the Warburg effect [28]. Based on the law of conservation of electrons in chemical reactions, Li et al. built up an electron balance model to outline metabolic plasticity under hypoxia. According to the model, both proline biosynthesis and lipogenesis can act as alternative electron acceptor. Blocking them synergistically suppresses tumor growth [29]. Moreover, tumor associated macrophages (TAMs) enhance aerobic glycolysis and induce apoptosis resistance of breast cancer cells via extracellular vesicle (EV) transmission of HIF-1α-stabilizing long noncoding RNA (HISLA). Reciprocally, lactate released from glycolytic cancer cells upregulates HISLA in TAMs, constituting a feedforward loop between TAMs and cancer cells [30].
Psychological factors induced metabolic alterations was revealed to cancer progression. Lactate dehydrogenase A (LDHA) executes the final step of the Warburg effect by converting pyruvate to lactate. Breast cancer stem-like properties can be promoted by chronic stress-induced epinephrine via LDHA-dependent metabolic rewiring. Interestingly, vitamin C may reverse the chronic stress-induced cancer stem-like phenotype [31].

Tumor microenvironment (TME)
The TME of breast cancer is composed of multiple stromal cells, soluble factors, and physical properties, including intratumor and metastatic microenvironment representing local and distant lesions, respectively [32].
In malignant phyllodes tumors (PT), TAMs promote malignant progression by secreting large amount of CCL18, which then binds to its receptor PIPTNM3 on myofibroblasts [33]. Reciprocally, malignant PT recruits and induces macrophages to a TAM-like phenotype by secreting CCL5. Thus, a positive feedback loop between CCL5-CCR5 and CCL18-PIPTNM3 is constituted, which represents the communication of myofibroblasts with TAMs and plays a central role in tumorigenesis of PT ( Fig. 1) [34].
Tumor development is accompanied by the occurrence and persistence of immunosuppressive microenvironment. Myeloid-derived suppressor cells (MDSCs) and TAMs interact with CD4 + and CD8 + T lymphocytes and subsequently attenuate the anti-tumor immunity response, which is one of the factors that most responsible for immune evasion. Alisertib, a selective Aurora A kinase inhibitor, could reshape the immune microenvironment through selectively eliminating tumor-promoting myeloid cells, including MDSCs and TAMs, as well as restoring the anti-tumor immunity of T lymphocytes ( Fig. 1). Intriguingly, combining alisertib with antiprogrammed cell death-ligand 1 (PD-L1) therapy showed a synergistic efficacy in the treatment of advanced breast cancer [35].
According to the "seed and soil" theory, premetastatic niches in destination organs are key driving force for tumor dissemination and colonization [36]. Additionally, primary tumor is capable of inducing B cell accumulation in draining lymph nodes. These tumor-educated B cells produce pathogenic IgG that targets the tumor membrane antigen HSPA4 (heat shock protein family A member 4) to activate the HSPA4-binding protein ITGB5 (integrin β5) and the downstream Src/NF-κB pathway. These findings illustrated the role of B cells in promoting the establishment of premetastatic niches in draining lymph nodes and accelerating breast cancer lymph node metastasis [37].

Tumor metastasis
Distant metastasis is the main cause of breast cancer related death. A recent study deepened our understanding of EMT in metastasis by revealing that breast cancer cells were capable of recapitulating various epithelial and mesenchymal phenotypes. Epithelial-type circulating tumor cells (CTCs) and disseminated tumor cells (DTCs) with a restricted mesenchymal transition show the most metastatic traits, whereas mesenchymal-type CTCs and DTCs display limited metastatic ability [34]. Bone is the most common site of distant metastasis. To further explain the mechanisms underlying the predilection of bone metastasis has always been a focus of breast cancer research. It was proved that Forkhead box F2 (FOXF2) regulated the epithelium-to-osteomimicry transition (EOT) via pleiotropic transactivation of the BMP4/SMAD1 signaling pathway and bone-related genes (BRGs), thus giving the metastatic tendency to bone (Fig. 2) [38]. The same group also revealed that FOXF2 deficiency BLBC tended to transdifferentiate to a myofibroblast/cancer associated fibroblast (CAF)-like phenotype and metastasized to visceral organs. The underlying mechanism involves the reciprocal repression loop between FOXF2 and transforming growth factor-β (TGF-β) (Fig. 2) [39]. These groundbreaking studies provided theoretical basis for preventing breast cancer metastasis.
Metastasis-initiating cells (MIC) are a tiny population of cells, estimated less than 0.02% of disseminated tumor cells, and are responsible for forming secondary tumors. Yang et al found platelet-derived growth factor receptor (PDGFR) inhibition blocked AKT activation but not serum and glucocorticoid induced kinase 1 (SGK1) signaling in MIC, which resulted in suppressing lung metastasis of breast cancer; however, primary tumor burden was not affected. Co-targeting PDGFR and SGK1 showed synergistic anti-cancer effects and led to further inhibition of pulmonary metastases and primary breast tumors [40]. Reduction of cell-cell adhesion is crucial for cancer cells departing from the primary tumor. AMP-activated protein kinase (AMPK) is well known for maintaining energy homeostasis. A recent study revealed that PI3K and HER2 activation transcriptionally downregulated AMPKα1 expression via ΔNp63α and highlighted that transcriptional control was another layer of AMPK regulation, which suggested the pivotal role of AMPKα1 in cell-cell adhesion and cancer metastasis (Fig. 2) [41]. S100 calcium binding protein A14 (S100A14) promoted cancer metastasis by upregulating the expression and secretion of CCL2 and CXCL5 via RAGE/NF-κB pathway (Fig. 2) [42]. These findings provided new therapeutic strategies for inhibiting breast cancer metastasis.
In addition, Nie et al. raised new precautions in dexamethasone therapy for breast cancer patients. Glucocorticoids (GCs) could activate TEA domain transcription factor 4 (TEAD4) in the Hippo signaling pathway for cancer metastasis and chemo-resistance by enhancing the interaction of the glucocorticoid receptor (GR) with TEAD4 to form a transcriptional complex (Fig. 2) [46].

Drug resistance
Drug resistance, both intrinsic and acquired, remains the main obstacle for effective anticancer therapies.
In terms of chemotherapy, cancer cells can develop drug resistance by activating DNA damage repair signaling pathways [47]. Following DNA damage, poly ADP-ribose polymerase 1 (PARP1) recruits MORC2 (MORC family CW-type zinc finger 2) to DNA damage sites and catalyzes MORC2 PARylation, which enhances its chromatin remodeling activities, thereby facilitating efficient DNA repair. In turn, MORC2 stabilizes PARP1 through a crosstalk between N-acetyltransferase (NAT10)-mediated acetylation and CHFR-mediated ubiquitination [48]. DNA-damaging chemotherapeutic agents stimulate MORC2 acetylation in a NAT10 dependent manner, then acetylated MORC2 activates G2 DNA damage checkpoint, enhancing cell survival following DNA damage [49]. Therefore, depleting MORC2 or inhibiting NAT10 sensitizes cancer cells to chemotherapy by abrogating DNA damage repair. Moreover, Rac 1 enhances the activity of non-oxidative pentose phosphate pathway via activating aldolase A and ERK signaling pathway, thereby inhibiting the nucleotide metabolism and DNA damage caused by chemotherapy in breast cancer [50]. Additionally, Zhang et al revealed that a feed-forward circuit between serglycin (SRGN) and YES-associated protein (YAP) induced HDAC2 expression, promoting chemoresistance to 5-FU [51]. Dong et al. found glutathione S-transferases P1 (GSTP1) contributed to adriamycin resistance [52]. In TNBC, synaptotagmin-like 4 (SYTL4) contributes to taxane resistance via attenuating the stability of microtubule network and increasing the growth rate of microtubule [53] The antisense intronic lncRNA (ai-lncRNA) EGOT (eosinophil granule ontogeny transcript) sensitizes breast cancer cells to paclitaxel by enhancing autophagy [54]. MEF2-interacting transcriptional repressor (MITR), as a truncated isoform of HDAC9, induces resistance of paclitaxel via increasing interleukin-11 (IL11) expression and subsequent activating JAK/STAT3 signaling pathway [55].
In terms of endocrine therapy, antiestrogens stabilize MORC2 in a GPER1 (G protein coupled estrogen receptor 1) dependent manner and decrease MORC2 enhanced cellular sensitivity to tamoxifen and fulvestrant [56]. In estrogen receptor (ER) positive breast cancer, lncRNA BDNF-AS, driven by a MEF2A regulated enhancer, induces endocrine resistance of breast cancer by activating the RNH1 (ribonuclease inhibitor 1)/TRIM21/mTOR (mechanistic target of rapamycin) cascade [57]. This is a novel mechanism other than the canonical PI3K (phosphatidylinositol-3-kinase)/AKT/mTOR axis. LncRNA DILA1 contributes to tamoxifen resistance in breast cancer via interacting with Cyclin D1 and blocking its phosphorylation and degradation [58]. Ajuba recruits DBC1 and CBP/p300, forming a ternary complex, to increase transcriptional activity of ERα and to potentiate the ERα target gene expression, thereby contributing to tamoxifen resistance [59]. MiR-575 enhances ERα activity and tamoxifen resistance via targeting cyclin dependent kinase inhibitor 1B (CDKN1B) and BRCA1 [60]. Moreover, inactivation of neddylation with MLN4924 transcriptionally inhibits ERα and significantly improves the fulvestrant sensitivity [61]. However, in ER negative breast cancer, targeting ubiquitin carboxyl terminal hydrolase-L1 (UCH-L1) upregulates ERα expression via enhancing ubiquitination and degradation of EGFR, and subsequently increases the sensitivity of tamoxifen and fulvestrant [62].
In regard to immunotherapy, breast cancers have poor response to immunotherapy because of poor T-cell infiltration and heightened immunosuppression within TME [66]. A low-dose vascular endothelial growth factor receptor 2 (VEGFR2) blockade can increase immune cell infiltration and upregulate programmed cell death protein-1 (PD-1) expression on immune cells, thus sensitizing breast cancer to a PD-1 blockade [67]. PARP1 suppresses PD-L1 expression via interaction with nucleophosmin (NPM1), which abolishes the binding of NPM1 at the PD-L1 gene promoter in TNBC. Olaparib elevates PD-L1 expression and leads to better anti-cancer efficacy in combination with the anti-PD-L1 antibody [68]. In addition, as a strong candidate for autophagy deficiency mediated immunosuppression, Tenascin-C (TNC) is overexpressed and is inversely correlated with LC3B expression and CD8+ T cells in TNBC patients. Inhibition of TNC in autophagy impaired TNBC cells sensitizes T cell mediated tumor killing and boosts anti-tumor effects of anti-PD-1/PD-L1 therapy [69]. These findings provided new approaches to potentially combine immunotherapy with molecular-targeting agents. These drug resistance mechanisms aforementioned are summarized in Table 1 and Figure 3.

Tumor biomarkers
The prognosis of breast cancer mainly depends on early detection and intervention. Recently, accumulated evidence indicates that circulating tumor DNA (ctDNA) could be a sensitive and specific biomarker for monitoring breast cancer progression and predicting relapse in early-stage [70]. A study showed the positive detection rate of ctDNA in early-stage breast cancer could reach 74.2%. Moreover, positive ctDNA after surgery indicated potential recurrence and remote metastasis [71]. Serum LRP6 ectodomain (LRP6N) could be another promising metastatic diagnosis marker for its binding to CXCR4 and competitively preventing SDF-1/ CXCR4-induced lung metastasis [72].   Because of the highly heterogeneity of TNBC, appropriate classification will provide effective prediction for outcomes. TNBC were classified into three heterogeneous clusters based on the microenvironment phenotypes: cluster 1, the "immune-desert" cluster, with low microenvironment cells infiltration; cluster 2, the "innate immuneinactivated" cluster, with resting innate immune cells and nonimmune stromal cells infiltration; and cluster 3, the "immune-inflamed" cluster, with abundant adaptive and innate immune cells infiltration. Cluster 1 and 2 were both called "cold tumors" while cluster 3 was a "hot tumor". Immune checkpoint inhibitors might be effective for "immune-inflamed" clusters, indicating that these distinct phenotypes were potential biomarkers for predicting therapeutic efficacy. The transformation of "cold tumors" into "hot tumors" should also be considered when dealing with "immune-desert" and "innate immuneinactivated" clusters [73].
The molecular typing of breast cancer has been revealed deeply by novel detecting techniques. A study divided Chinese TNBC into four subtypes based on multi-omics data: luminal androgen receptor, immunomodulatory, basal-like immunesuppressed, and mesenchymal-like subtypes, which could serve as a reference in order to further advance the understanding and precision treatment of TNBC [74]. Further, the same group conducted a phase Ib/II trial, demonstrating the clinical benefit of molecular subtyping-based and targeted sequencing-based therapy for refractory metastatic TNBC. Specifically, nab-paclitaxel in combination with anti-PD-1 therapy or with anti-VEGFR therapy showed favorable outcomes on IM (immunomodulatory) subtype and BRCA1/2 wild type-BLIS (basal-like immunesuppressed) subtypes, respectively [75]. Moreover, based on distinct metabolic dysregulation, TNBC was classified into three subtypes via multi-omics database analysis. MPS1 (heterogeneous metabolicpathway-based subtype 1) is the lipogenic subtype with upregulated lipid metabolism. MPS2 is the glycolytic subtype with upregulated carbohydrate and nucleotide metabolism. MPS3 is the mixed subtype with partial pathway dysregulation. Different subtypes respond to metabolic inhibitors with distinct sensitivity which enables the development of personalized cancer therapy by targeting unique metabolic profiles [76].

Novel therapeutic targets and new drugs
With the increasing popularity of precision therapy for breast cancer, it is urgent to develop new therapeutic targets and drugs.
As we know, tumor metabolism is important in cancer progression. Nevertheless, the lack of effective and selective anti-metabolism drugs hinders clinical application. Chen et al reported a small molecule ZY-444 bound to pyruvate carboxylase, which was a key anaplerotic enzyme of the tricarboxylic acid cycle in promoting cancer growth and metastasis through Wnt/β-catenin/Snail signaling pathways. ZY-444 suppressed breast cancer cell proliferation with low cytotoxicity in normal cells by inactivating catalytic activity of pyruvate carboxylase [93]. Flubendazole, a broad-spectrum anthelmintic drug belonging to benzimidazole group, has been repurposed as a promising anti-cancer agent [94]. Zhen et al revealed flubendazole induced autophagic cell death by targeting Eva-1 homolog A (EVA1A) and suppressed TNBC proliferation and migration [95]. Moreover, resistomycin, QN-1 and emodin attenuates TNBC progression via binding to Pellino-1 to promote FBXO11-mediated SNAIL/SLUG degradation [96], down-regulating c-MYC transcription [97] and reducing VEGFA transcription [98], respectively. Accordingly, these studies provided alternative strategies for treating TNBC.

Conclusion
In summary, Chinese investigators have made significant progress in all areas of breast cancer research over the last two years, particularly in the field of BCSCs. A number of new BCSC membrane biomarkers, including PROCR, NOTCH4, SGCE, TSPAN8, and IL1R2, were identified. They may be used for diagnosis and potential therapeutic targets. In addition, stemness of BCSCs could be regulated by tumor microenvironment, such as TAM, CAF, and signaling molecules. Due to the increased patient numbers, limited treatment options, and unsatisfactory treatment efficacy in China, investigators focused on the exploration of drug resistance and metastasis mechanisms and made great efforts to identify new tumor biomarkers and therapeutic targets in order to improve the diagnosis and treatment of breast cancer. However, there is still a lack of milestone discovery in terms of breast cancer cell origin, pathological typing, and targeted metastatic therapy. Thus, multi-party cooperation is strongly recommended to make a breakthrough in basic research and clinical translation.