Advancements in Zebrafish Models for Breast Cancer Research: Unveiling Biomarkers, Targeted Therapies, and Personalized Medicine

Background

Breast cancer (BC) is a heterogeneous disease with different molecular subtypes [1]. Approximately 70% of breast tumors exhibit estrogen receptor (ER) expression; therefore, endocrine therapy is widely used for these patients. However, there is a high risk of recurrence in this therapy regimen [1]. Triple-negative breast cancer (TNBC) is the most aggressive subtype of BC, with associated chemoresistance. Due to the very heterogeneous character of TNBC, response to the existing therapeutic options is very limited. Therefore, effective personalized therapies with relatively low adverse effects are urgently required [2].

The use of animals in preclinical studies has been crucial in biomedical science development for decades [3]. In vivo models enable a better understanding of the pathological mechanisms driving disease and testing new therapy approaches (eg, drugs or their combinations, vaccination, irradiation) [4]. Zebrafish (Danio reiro) has become a useful tool in studies of human diseases like cancers [5], skeletal disorders [6], cardiovascular [7], neural diseases like Parkinson’s [8] or Alzheimer’s [9] or developmental disorders [10], and also in human angiogenesis analysis [11].

The functional and structural homology of around 76–82% of genes involved in human diseases is shared between humans and zebrafish (ZF) [12]. Compared to mice or rats, fish brings numerous advantages like fast development, a large number of offspring, low costs, and the possibility to carry out the experiment on a large scale (Figure 1) [5].

Numerous transgenic lines of ZF are commonly available, which significantly contributes to easier experimentation [4]. Among them, there are lines with labeled organism components (macrophages [13], neutrophils [14], or vascular system [11]), with immune system deficiency [15,16], or without pigmentation [17]. Especially the last species, called the Casper line, is a very powerful model in cancer research as its transparency combined with labeled cancer cells or proteins enables observation of tumor development, cell invasion, metastasis, or drug response [5] (Figure 2).

Research studies describe almost every type of cancer that has been already tested in the ZF model, including BC [18], gastric cancer [19], colorectal cancer [20], non-small cell lung adenocarcinoma [21], pancreatic ductal adenocarcinoma [22], melanoma [23], glioblastoma [24], and others [3,5].

This article aims to review the current status of the role of the ZF (Danio reiro) in preclinical and clinical models of biomarker identification and drug targeting, and developments in personalized medicine in BC.

Zebrafish Breast Cancer In Vivo Models

There are 2 directions to achieve a pathological condition in animal models: by introducing gene(s) to induce the disease (oncogenes) or to inhibit expression of targeted genes using genetic engineering methods (Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) or morpholino for knockdown), or by tumor transplantation (xenografts) [5]. In the case of BC, creating an environment similar to the patient using genetic engineering is difficult because of the lack of Breast Cancer Gene 1 (BRCA1) ortholog [25,26]. Some studies used the Breast cancer gene 2 (Brac2) fish gene as an ortholog for human Breast Cancer Gene 2 (BRAC2), but the homology is only 22% [27]. Based on that, a transgenic fish with brac2 knock down or knock out has been designed using morpholino. The same fish gene has been studied for analysis of deoxyribonucleic acid (DNA) damage repair (DDR) in BC [25].

Missing the mammary organs where BC develops is compensated for by xenotransplantation. Xenografts can be created from cell lines (CDX - cell lines-derived xenografts) or directly from surgery or biopsy of the patient (PDX - patient-derived xenografts). Numerous different BC cell lines have been successfully transplanted into ZF, eg, MCF-7 [28], MDA-MB-231 [16,29], MDA-MB-435 [30], MDA-MB-468 [31], and BT-474 [30]. Generally, 2 days after fertilization (dpf), fish are injected with ~800–1500 cells suspension. After 1 day, the tumoral mass is formed [32].

Some studies used three-dimensional (3D) spheroids to better mimic the initial stages of tumor angiogenesis and metastasis [33]. MCF7 and mammary adipose tissue-derived mesenchymal stromal/stem cells (MAT-MSCs) co-cultures created the spheroids that were transplanted into ZF. Injection of co-cultures showed higher tumor invasiveness compared to MCF7 monoculture. Another study presented a complex model based on the co-culture of MCF-7 or MDA-MB-231 human mammary tumor cells mixed with human preadipocytes (hMADS, human multipotent adipose-derived stem cells) to generate organoids [28].

For a better understanding of the molecular mechanisms in cancer development, mutated cell lines were transplanted. BC cells with silenced p53, Inhibitor of DNA Binding 4 (ID4), Serine/arginine-rich Splicing Factor 1 (SRSF1), MALAT1 genes were injected into ZF lead to attenuate the proangiogenic properties [34]. In turn, targeting High Mobility Group Box 3 (HMGB3) by microRNA-142-3p (miR-142-3p) inhibited migration, invasion, and metastatic potential in in vivo experiments performed on ZF [35]. Human TNBC tumor cells expressing Chemokine Receptor Type 4 (CXCR4) initiate early metastatic events by sensing ZF cognate ligands at the metastatic site. The CXCR4-CXC motif chemokine 12 (CXCL12) axis acts across ZF and humans and drives the formation of tumor micrometastases of human TNBC cells in ZF [36]. A BC oncogene – AIB1Δ4 (Amplified in Breast Cancer 1)-expressed in TNBC cells enabled the invasion of parental cells in ZF [37]. The expression of long non-coding RNA (lncRNA) H19 and microRNA-675 (miR-675) tested in ZF xenografts with human BC showed an increase in invasive capacities [38].

Breast Cancer Patient-Derived Xenografts

Cell lines are easily available and cheap. However, for medical reasons, using CDX is connected with some limitations. Multiple passaging leads to clonal selection and loss of tumor heterogeneity [39]. Differences in gene expression can be noticed even among the same cell line. An analysis of a commonly used BC cell line (MCF-7) indicated a parental cell line and its 3 subclones showed significant differences at the genomic and gene expression levels [40]. That might cause a different response for drug treatment of different strains among the same cell line. That effect was described in testes of 27 MCF7 strains when some of them indicated a drug response and some were non-responsive [41]. Moreover, some studies have found that ZF xenografts have poor predictive value in response to therapeutics. Tests of 39 compounds applied in both CDX models and phase II clinical trials at the National Cancer Institute’s Developmental Therapeutics Program did not indicate any close correlation [42]. A comprehensive comparison of molecular profiles in copy number variation (CNV), mutation, mRNA, and protein expression between 68 cell lines and 1375 primary breast tumors from The Cancer Genome Atlas (TCGA) showed that cell lines can mirror some molecular properties but not all of them. Among tested lines, T47D, BT483, and MDA-MB-453 showed the highest similarity with tumors [43].

Using cell lines in vitro or in vivo generally does not mimic the patient-specific environment and leads to differences in progression and treatment outcome. To get the most reliable results, using cells derived from patients is necessary. Patient-derived xenografts (PDXs), also called cancer “Avatars” [4], are generated by the implantation of human primary tumor cells or tissues, obtained from surgery or biopsy, into a host animal. For the first time, ZF was used as a PDX for leukemia [44]. Nowadays, PDXs are considered a very important tool in cancer therapy. Thanks to that, the most efficient treatment can be applied for an individual patient, which promotes the development of personalized medicine. For BC, mice have been described as the most effective PDX model [39,45–47]. Although much research has been done on BC CDX using ZF, the number of zPDX (zebrafish PDX) is still, surprisingly, limited [16,48,49].

Studies show that ZF is a good model to test patient breast tumors and is also a promising alternative to mice. The effective implantation success rate for zPDX can be very high (95%). Moreover, transplanted cancer cells were spread in ZF and correlated with clinical data on advancing tumor stage from stage 1 to stage 3 disease. Importantly, the response to drug treatment (docetaxel and trastuzumab) observed in fish was the same in 14 of 18 cases as in the mouse PDX models, clearly showing that the ZF model can be used successfully, like the mouse model [49]. Similar results were obtained by Yan et al in 2018 when they compared rates of proliferation and apoptosis of MDA-MB-231 TNBC cells between fish and mice [17]. Another study showed that ZF might be a promising model in preclinical BC therapy to identify prognostic markers and predict response to treatment. The behavior of zPDX with bone metastasis derived from BC patients has been compared to established BC cell lines. Primary cells from patients were extravasated from vessels and invaded the CHT (caudal hematopoietic tissues) of ZF. It was found that the patient’s primary cells can spread out, throwing blood vessels and invading the hematopoietic tissue of the ZF tail [50].

Troubleshooting In Vivo Models

One of the biggest problems, generally, with animal models is that their immune systems can cause rejection of transplanted cells; therefore, intensive studies have been conducted to develop a mutant with deficient immunity. A variety of immunocompromised ZF strains have been created using recombination activating gene 1 (rag1) [51], recombination activating gene 2 (rag2) [52], and protein kinase DNA-activated catalytic subunit (prkdc) [53] mutants. A major advance was made by the Langenau group, which created a prkdc−/−, ilrga−/− Casper strain [16]. The model lacks both adaptive and natural killers (NK) immune cells. Moreover, it enables the monitoring of engrafted human cancer cells at 37°C. This model has been successfully used for transplantation of therapy-resistant ER+/progesterone receptor (PR+)/human epidermal growth factor receptor 2 (Her2−) lobular BC (BRx-07) from patients. Screening drugs showed that cells were sensitive for olaparib and temozolomide and these results have been confirmed in the severe combined immunodeficiency (SCID) mouse model [16]. Intensive studies aimed at developing a perfect model is crucial for personalized medicine, which is so important in cancer therapy.

Epithelial-Mesenchymal Transition (EMT) in Tumor Metastasis

Metastasis is a complex and multi-stage process consisting of the spreading of cancer cells, infiltration, and extravasation, and finally colonization of metastatic cells in distant organs, which in turn leads to secondary lesions [54,55]. Epithelial-mesenchymal transition (EMT) is a major phenomenon that contributes to cancer metastasis and drug resistance [56]. During EMT, epithelial cells lose intercellular connections and gain the ability to migrate, becoming highly invasive mesenchymal cells. EMT is characterized by a loss of cellular cohesion mediated by the downregulation of adhesion molecules such as E-cadherin, β-catenin, occludin, and claudin-1. E-cadherin expression is regulated by numerous transcriptional factors, such as Zinc Finger Protein SNAI1 (SNAIL), Smad Interacting Protein 1 (SIP-1), δ-Crystallin Enhancer Factor 1 (δEF1), and Twist Family BHLH Transcription Factor 1 (TWIST 1). Additionally, during EMT, expression of mesenchymal markers such as N-cadherin or vimentin is increased [57].

Modeling Breast Cancer Metastasis with Zebrafish

Efforts to find therapeutic agents that target cancer metastases have often failed in clinical trials. Clinical failures are frequently due to the lack of modern, available, and easy-to-maintain preclinical models for metastasis research [55].

The transparent embryos make the ZF an ideal in vivo model to visualize migration of tumor cells [58]. Although ZF have an exceptionally strong innate immune system, they do not develop a mature immune system in the first 2 weeks. Therefore, the use of ZF in cancer research has recently expanded with the establishment of human cancer cell xenograft models [59]. Another advantage of using ZF xenografts in metastasis research is the fact that implanted cancer cells can be easily labeled with fluorescent dyes, making it possible to quickly identify neoplastic vs normal cells. As a result, studies on dynamic microtumor formation, cell invasion, and angiogenesis can be easily visualized in the body of a live fish [59].

A primary culture originating from a female patient with bone metastases from BC was cultured and then injected into ZF embryos to test its metastatic potential [50]. In the first step, the ZF model was validated by examining the colonization of 2 BC cell lines into 2-dpf ZF embryos. Experiments were performed on MDA-MB-231, a TNBC cell line, and MCF7, a hormone receptor-positive BC cell line. Interestingly, the MCF7 cells did not survive in the ZF embryo. In contrast, MDA-MB-231 cells spread and colonized various parts of the ZF, including tail hematopoietic tissue (CHT), indicating a strong migratory phenotype. Finally, the cells with the same hormonal receptor status as MCF7 cells were taken from the patient’s bone lesion and implanted in ZF [50]. The cells demonstrated an aggressive phenotype as they were extravasated from the vessels occupying various parts of the embryos [50].

In another study, ZF embryos were used to identify invasive subpopulations of MDA-MB-231 BC cells [55]. A subpopulation consistent with metastatic cells circulating in the human body was first created described by injecting MDA-MB-231 cells into the yolk sac 2 days after embryo fertilization and then collecting the cells that migrated to the tail [55]. Analysis of differential gene expression in an invasive subpopulation with the parental total population showed activation of EMT in circulating cells. It has been demonstrated that knockdown of the DNA Damage Inducible Transcript 4 (DDIT4), Cathepsin D (CTSD), Metallothionein 1X (MT1X), and Serpin Family E Member 1 (SERPINE1) genes was sufficient for minimizing MDA-MB-231 cellular invasion, thus creating the possibility of synthesizing drugs typically targeting these genes [55].

Importance of TGF-β in the Process of Metastasis

Transforming growth factor beta (TGF-β) plays an important role in controlling the development of BC [60]. It has been demonstrated that ZF xenografts are a good in vivo model for the pharmacological modulators and genetic perturbation of TGF-β signaling studies in human BC cells [60]. ZF embryo xenograft assays were described to investigate the mode of TGF-β signaling and its role in extravasation of BC cells, as well as regulation of angiogenesis. Tumor cells were fluorescently labeled with mCherry and then injected into a ZF embryo into Cuvier’s duct or perivitelline space. The invasive and migrating human cancer cells and the induction of new blood vessel formation were observed [60]. A ZF xenograft was generated by injecting MDA-MB-231 cells into the circulation of the embryo, and their migration to the avascular collagenous tail was observed. It was noted that BC cells dispersed very quickly and formed micrometastases in just 6 days. Blocking the TGF-β pathway through inhibition of TGF-β receptor kinase or knockdown of SMAD family member 4 (Smad4) leads to reduction of BC metastasis. Interestingly, similar results were obtained in a mouse metastasis model [61].

Role of PTEN in Invasion and Migration of Cancer Cells

Another analysis that was done in the BC metastasis context in the ZF model concerned the insufficiency of phosphatase and tensin homolog deleted on chromosome 10 (PTEN) protein, as a common marker in patients with malignant BC [62]. It was shown that PTEN deficiency induced an increase in MCF-7 cell invasion and migration by EMT induction. In addition, the knockdown of PTEN in the MCF-7 cell line induced cell migration in the ZF model. These findings indicate new therapeutic targets for PTEN-deficient BC patients [62].

Role of CXCR4-CXCL12 Chemokine Signaling Axis in Cancer Cell Migration

The CXCR4-CXCL12 chemokine signaling axis is a driver of cell migration in physiological and pathological conditions, including BC metastasis. Although CXCR4-CXCL12 axis-targeted therapies are in clinical trials, there still are no effective therapeutic approaches established for blocking CXCR4 in TNBC [36]. The CXCR4-CXCL12 axis acts across both ZF and human organisms, and drives the formation of tumor micrometastases of TNBC BC cells in zebrafish. IT1t is a small-molecule potent CXCR4 antagonist and orthosteric competitor of the CXCL12 N-terminal signaling peptide. IT1t impairs signaling activation by interfering with the docking of the ligand domain to the receptor. Treatment of MDA-MB-231 TNBC cells with IT1t as well as genetic silencing of CXCR4 significantly inhibited early metastasis xenotransplantation ZF model in vivo [36].

Metastatic Capacity of Circulating Tumor Cells

The metastatic competency of BC CTCs (circulating tumor cells) and CTC clusters in ZF embryos were analyzed [63]. CTCs are cancer cells that have sloughed off of the primary tumor and circulated in the blood [64]. Interestingly, CTC clusters derived from MDA-MB-231 and MCF7 cells exhibited more diverse behavior than single CTCs in in vitro metastasis analyses. In this study, CTC clusters were shown to have an advantage in terms of survival and proliferation over single CTCs, better adapting them to survival in circulation. However, the CTC clusters in the ZF experiments had less ability to invade and spread. Interestingly, the grouping of even a few cells increased their survival. Comparable results were obtained in a mouse model [63].

Study of Effect of the Microenvironment on Breast Cancer Cells

The newest research concerns environmental pollution, specifically persistent organic pollutants (POP), which, according to previous results, contributes to development of BC. Recent clinical studies have shown that increased POP concentration is associated with an increased risk of BC progression and metastases [28]. To study the effect of the microenvironment on BC cells, a co-culture model containing MCF-7 or MDA-MB-231 BC cells and hMADS preadipocytes as well as TCDD (2,3,7,8-tetrachlorodibenzo-p-dioxins) was used. Using the ZF larvae model, BC cells were shown to become metastatic when cultured with pre-adipocytes and exposed to a persistent organic pollutant, TCDD [28].

Zebrafish as a Preclinical Model for Drug Screening

Over the last decade, ZF have played a significant role in human diseases modeling and in finding novel therapeutic options for them [65,66]. For this purpose, ZF have been widely used in the creation of PDX cancer models to identify patient-specific responses, and in phenotypic drug screening of different antineoplastic agents [67,68]. Numerous therapeutic approaches that recently entered into clinical trials and clinics have their genesis in ZF [69]. Interestingly, ZF respond to different drugs and active agents at physiologically relevant dose ranges. In addition, drug activity can be studied at a single-cell resolution within the complexity of a whole animal after combination with cell- or tissue-specific reporters and gene editing technologies [69]. The unique opportunity to use ZF for drug screening is possible thanks to the many advantages of this model, like a fully sequenced genome, rapid embryonic development, short generation interval, transparent embryos (which make them amenable to microscopy-based screens usually restricted to cell cultures), low maintenance, and the ability to use large numbers due to cost-effectiveness (Figure 2) [68,70]. These advantages have recently helped to popularize ZF as a great model system for pharmaceutical and biomedical studies, including BC drug screening (Figure 3).

Monoclonal Antibodies Studies in Zebrafish Model

Olaparib is a poly ADP ribose polymerase (PARP) inhibitor used in the therapy of metastatic HER2-negative BC with an inherited BRCA1 or BRCA2 mutation [32]. To determine if ZF can discriminate sensitivity to olaparib, BC ZF xenografts with different BRCA statuses were established. To generate the xenografts, Hs578T and MDA-MB-468 BRCA wild-type (wt) and HCC1937 BRCA1 mutant BC cells were fluorescently labeled and injected into the perivitelline space of embryos 2 days after fertilization. Later, xenografts were treated with olaparib individually or in combination with ionizing radiation and characterized according to different hallmarks of cancer, such as cell death, cell proliferation, angiogenic potential, and metastasis. Interestingly, Hs578T BC cells revealed no decrease in tumor size, no alterations in proliferation rate, and no increase in active caspase-3 after olaparib treatment. The other MDA-MB-468 BRCA wt model demonstrated sensitivity to PARP inhibitor (PARPi) in a ZF xenograft model. Additionally, the combination of olaparib with ionizing radiation (IR) increased the cytotoxic effect of olaparib, enhanced the accumulation of DNA damage, and activated cell death signaling pathways. PARPi significantly increased aberrant mitosis in MDA-MB-468 xenografts; however, in the case of HCC1937 xenografts, only IR and the combined regimen led to an increase in mitotic aberrancies. The combination of olaparib with IR induced apoptosis in all xenograft models and demonstrated a synergistic effect on decreasing tumor size in BRCA wt Hs578T xenografts. Next, the influence of olaparib in combination with IR on the tumor-related vessel network was investigated. The impact on angiogenesis was only detected in the BRCAwt Hs578T tumors, which presented the strongest angiogenic potential in the panel. It has been demonstrated that tumor vessel density decreased after IR treatment in monotherapy and was further decreased in the combinatorial regimen. All these results suggest that the ZF xenograft model can be used as a sensitivity screening platform for checking the anti-cancer activity of olaparib used in monotherapy and combinatorial regimens (Table 1) [32].

Bevacizumab is a recombinant humanized monoclonal antibody directed against vascular endothelial growth factor (VEGF). The overall average response for bevacizumab as metastatic BC monotherapy is not satisfactory at only 9.3%. Additionally, bevacizumab treatment is linked with serious adverse effects such as renal toxicity, bleeding, and cardiovascular complications [71]. Depending on the individual tumor character, anti-VEGF therapy can either block or stimulate metastasis. Therefore, an assay for the prediction of individual response to therapy would have great value to guide treatment options. Zebrafish BC xenografts were used for determining the response to bevacizumab and its impact on angiogenesis and micrometastasis. Hs578T primary breast carcinoma and MDA-MB-468 pleural effusion metastasis-originated BC cells were injected into the perivitelline space of ZF larvae 2 days after fertilization and analyzed at 4 days after injection for testing the antiangiogenic and metastatic impact of bevacizumab in ZF xenografts. Cell death analysis demonstrated a significant induction of apoptosis in Hs578T TNBC xenografts accompanied by a 40% reduction in tumor size. Additionally, the Hs578T TNBC tumors that exhibited the highest angiogenic potential were also the most affected by bevacizumab treatment, with 20% reduction in vessel density and vessel infiltration (Table 1) [71].

Synthetic Drugs Studies in Zebrafish Model

The ZF model has also been used extensively in the study of synthetic drugs. Actein is a cycloartane triterpenoid that exhibits anti-cancer and anti-angiogenic activities [72]. A metastatic human BC cell xenograft model was established in transparent ZF embryos to determine the anti-migration effect of actein in vivo. Injection of MDA-MB-231 cells into the yolk sac was applied to test the activity of acetin. After quantification, obtained results demonstrated that the number of migratory cells in embryos treated with 60 μM of acetin significantly decreased (75%), confirming the anti-proliferation and anti-migration activity of actein in the ZF xenograft model. The survival rate of ZF embryos was not affected by actein, suggesting that actein at tested doses did not induce observable toxicity to the ZF embryos. The findings reveal that actein can be a potential anti-metastatic candidate drug for the treatment of advanced human BC (Table 1) [72].

1-O-octadecyl-2-O-methyl-sn-glycero-3-phosphocholine (edelfosine, ET) is an alkyl-lysophospholipid tested in phase II clinical trials in patients with inoperable brain tumors (or previously treated with other therapies with no positive results) [73]. Nanometric emulsion containing edelfosine (ET-NE) was prepared for the TNBC treatment to determine the biodistribution, toxicity, and anti-cancer efficacy in ZF embryos. In vivo results demonstrated that the ET-NE nanosystems can penetrate through the skin biological barrier into the yolk sac and significantly decrease highly aggressive and invasive tumoral cells’ proliferation in ZF embryos bearing MDA-MB-231 xenograft tumor, leading to its regression. Overall, the obtained results indicate that the developed ET-NEs formulation can be a promising therapy for treatment of TNBC (Table 1) [73].

Another drug, 5k, new N-2(5H)-furanonyl sulfonyl hydrazone derivative 2(5H)-furanone derivative significantly inhibited the proliferation of MCF7 BC cells and angiogenesis in the MCF7 ZF xenograft model as reflected in decreased tumor area and volume, and the number and length of ectopic vessels were reduced in a dose-dependent pattern. These results suggest that N-2(5H)-furanonyl sulfonyl hydrazine derivatives can be potent drugs for BC treatment (Table 1) [74].

The anti-cancer potential of [Ru(bipy)2(C12H8N6-N,N)][CF3SO3]2 Ru1 and [{Ru(bipy)2}2(μ-C12H8N6-N,N)][CF3SO3]4 Ru2 ruthenium complexes was studied in an MCF7 ZF xenograft model [75]. Both newly developed active compounds effectively prevented MCF7 proliferation and inhibited tumor growth compared to control in MCF7 BC cell ZF xenografts, with no significant signs of lethality observed. Therefore, both compounds show potential for further in vivo mouse studies, as well as for their use as scaffolds for further chemical modifications to improve their drug-like properties (Table 1) [75].

Natural Products Studies in Zebrafish Model

Natural products have been extensively studied in the last decade [76]. Fangjihuangqi decoction is a classic prescription in Chinese medicine that has been clinically used in therapy of fatigue, upper-extremity edema, and enhancement of immunity to prevent cancer recurrence in postoperative BC patients [76]. Fangjihuangqi decoction significantly inhibits tumor growth and reduces the number of cancer cells in the abdominal cavity and improves the abnormal morphology of tumor cells in the TNBC xenograft ZF model. To establish a human TNBC xenograft tumor in a ZF model, TGFβ1-treated MDA-MB-231 cells were microinjected into a 2dpf ZF yolk sac. TGFβ1-treated MDA-MB-231 cells presented a strong tumor formation ability. Fangjihuangqi decoction reduced the length of tumor neoplastic lymphatics and restrained the sprouts’ number of tumor neovascularization via the reverse EMT process. Western blotting analysis demonstrated that Fangjihuangqi decoction significantly increased E-cadherin and decreased Snail2, TWIST1, zinc finger E-box binding homeobox 2 (ZEB2), and TGFβ1 protein expression. All these results using the ZF model suggest that the Fangjihuangqi decoction could prevent TNBC metastasis through significant inhibition of tumor lymphangiogenesis and neovascularization (Table 1) [76].

Accumulating evidence suggests that breast cancer stem cells (CSCs) are related to tumor recurrence and drug resistance. Therefore, targeting CSCs can provide a useful approach to BC therapy [77]. It has been revealed that Chinese medicines with their natural active compounds can be involved in stem cell properties regulation. Gomisin M2 is a natural product isolated from the Schisandra viridis A. C. Smith herb that is used as an anti-cancer agent. Gomisin M2 reduced MDA-MB-231 and HCC1806 TNBC progenitor cell growth and proliferation in ZF embryos. Interestingly and unexpectedly, gomisin M2 blocked the CSCs growth in vivo in approximately 2-fold lower concentrations than these suppressing formation of tumor spheres, suggesting that gomisin M2 effectively targets cancer cells in the in vivo tumor microenvironment with lower doses than in vitro. Additionally, HCC1806 SCs underwent tail migration, although any fluorescent particles had migrated to the tail after gomisin M2 treatment. These findings suggest that gomisin M2 may be potentially used as a therapeutic active agent in targeting breast CSCs (Table 1) [77].

Bakuchiol is a meroterpene found in the Psoralea corylifolia L. dried ripe fruit and widely used in traditional Chinese herbal medicine [78]. The metastatic MCF7 breast CSCs were injected into 48-hpf (hours post-fertilization) ZF embryos and were exposed to 0.5 and 1 μg/ml of bakuchiol. Bakuchiol induced apoptosis and inhibited BCSCs metastasis in ZF embryos. The percentage of embryos without metastasis increased and the number of fluorescent particles decreased dose-dependently after bakuchiol treatment. Interestingly, the doses used for in vivo metastasis experiments were lower than for in vitro studies. These findings suggest that bakuchiol has great potential as anti-breast cancer therapy; however, further experiments are needed (Table 1) [78].

Furanodiene, a natural terpenoid isolated from Rhizoma Curcumae herb, demonstrates antineoplastic activity against different types of cancer cells [79]. ZF xenograft models with MCF7 and MDA-MB-231 human cancer cell lines were established and validated with anti-cancer agents used for treatment of cancer patients in clinics. It has been demonstrated that furanodiene is therapeutically effective for MCF7 BC cells xenotransplanted into ZF. Moreover, furanodiene showed a strong synergistic anti-cancer effect in combination with 5-fluorouracil (5-FU) in an MDA-MB-231 transplanted model. Additionally, ZF injected with drug-resistance cancer cells were used to determine the anti-multidrug resistance effects of furanodiene. Furanodiene reversed multiple drug resistance (MDR) in the ZF xenotransplanted with adriamycin-resistant MDA-MB-231 cells. The caspases 8, 9, and 3/7 activity were notably increased in the BC cell xenotransplanted ZF treated with furanodiene, suggesting that furanodiene-induced ZF apoptosis are caspase 8- and caspase 9- dependent. Additionally, furanodiene demonstrated its anti-neoplasm effects via inhibition of angiogenesis and induction of reactive oxygen species (ROS) production and DNA strand breaks. Furanodiene caused suppression of efflux transporter P-glycoprotein Pgb function and reduction of Pgp protein level, but not MDR1-Pgp-related gene expression [79]. All these results suggest that furanodiene is a potential sensitizing agent to chemotherapeutic drugs, probably mainly through blocking Pgp efflux transportation; however, further cellular, biochemical, and molecular mechanisms studies are needed to confirm furanodiene pharmacology in mammalian models (Table 1).

Jadomycin B is a secondary metabolite biosynthesized by the soil bacterium Streptomyces venezuelae ISP5230 under stress conditions [80]. It has been demonstrated that jadomycin B reduces MDA-MB-231 cell proliferation in a dose-dependent fashion 48 h after transplantation into the yolk sacks of ZF larvae; 60 μmol/l and 80 μmol/l of jadomycin B significantly decreased ZF larval viability after 120 h and 96 h, respectively. The MTD (the highest dose of a drug that does not cause unacceptable adverse effects) of jadomycin B was determined as 40 μmol/l in ZF larvae, which is approximately 10–20-fold higher than its IC50 in in vitro BC cell cytotoxicity experiments (Table 1) [80]

In summary, all these findings suggest that ZF has become an important vertebrate model for assessing BC drug effects.

Future Directions

The advantage of using ZF in BC therapy is its potential to provide a platform for drug discovery and testing. Researchers can use ZF to screen large numbers of compounds and identify those that are effective in treating BC. This can help to accelerate the drug discovery process and bring new therapies to patients. It is commonly known that the perfect therapy plan should be highly individualized, especially in the treatment of BC, where a huge heterogeneity of the tumor is observed. Using low-cost and ease-of-maintenance animal models has recently become a hot topic in medicine and genetics. Here, we summarized approaches like ZF genome editing, creating conditional immune compromised models, and selecting the best therapy that together have a significant impact on development of personalized medicine. There are still open questions, including the viability of immunodeficiency models or tumor rejection, but we believe that intensive research will solve these issues very soon. Overall, the use of ZF in BC research and therapy has tremendous potential to advance our understanding of this disease and to develop new, more effective treatments. Although ZF PDX still needs to be validated, we believe that this model gives information that is highly translatable to humans.

Conclusions

Cancer studies have rapidly moved away from autonomous tumor cells to a complex systemic approach. The systemic response plays an important role in tumor cell behavior, as well as affecting therapy and patient outcomes. Recently, ZF have been increasingly applied as a good in vivo model for human disease research with a particular focus on cancers. ZF BC models are currently widely used in carcinogenesis studies, monitoring of disease progression, and testing new standard or alternative therapies [5]. Here, we presented a rationale for using ZF as powerful in vivo models for BC mechanisms studies and testing new potential BC therapy agents. This review has presented the current status of the role of the ZF (Danio reiro) in preclinical and clinical models of biomarker identification and drug targeting, and developments in personalized medicine in BC.