Patient-Derived Breast Cancer Bone Metastasis In Vitro Model Using Bone-Mimetic Nanoclay Scaffolds

. Te unavailability of reliable models for studying breast cancer bone metastasis is the major challenge associated with poor prognosis in advanced-stage breast cancer patients. Breast cancer cells tend to preferentially disseminate to bone and colonize within the remodeling bone to cause bone metastasis. To improve the outcome of patients with breast cancer bone metastasis, we have previously developed a 3D in vitro breast cancer bone metastasis model using human mesenchymal stem cells (hMSCs) and primary breast cancer cell lines (MCF-7 and MDAMB231), recapitulating late-stage of breast cancer metastasis to bone. In the present study, we have tested our model using hMSCs and patient-derived breast cancer cell lines (NT013 and NT023) exhibiting diferent characteristics. We investigated the efect of breast cancer metastasis on bone growth using this 3D in vitro model and compared our results with previous studies. Te results showed that NT013 and NT023cells exhibiting hormone-positive and triple-negative characteristics underwent mesenchymal to epithelial transition (MET) and formed tumors in the presence of bone microenvironment, in line with our previous results with MCF-7 and MDAMB231cell lines. In addition, the results showed upregulation of Wnt-related genes in hMSCs, cultured in the presence of excessive ET-1 cytokine released by NT013cells, while downregulation of Wnt-related genes in the presence of excessive DKK-1, released by NT023 cells, leading to stimulation and abrogation of the osteogenic pathway, respectively, ultimately mimicking diferent types of bone lesions in breast cancer patients.


Introduction
Breast cancer is a leading cause of cancer-related deaths in women worldwide [1], causing fatal skeletal failure at their advanced stage [2].Invasive breast cancer is the most common cancer afecting women in the United States, and 287,850 individuals are expected to be diagnosed in 2022.Of these, 43,250 are estimated to die from it in 2022 [3].Breast cancer metastasis accounts for the majority of deaths from breast cancer.Bone is a common site of metastases for breast cancer [4] and can cause signifcant complications such as pain, pathologic fracture, hypercalcemia, and spinal cord compression [5,6].Tese complications can also lead to death from breast cancer.Detection of breast cancer and treatment of metastasis at the earliest stage is important to decrease mortality [7].Metastatic breast cancer is still an incurable disease.Median survival is about 3-5 years for hormonal receptor-positive metastatic breast cancer [8,9] and 1-2 years for triple-negative metastatic breast cancer [10,11].Due to complex cellular heterogeneity within cancer cells [12] and the low success rate of novel drugs for metastasized breast cancer in clinical trials [13], efective treatment for advanced-stage breast cancer remains a challenge for researchers.New models are required wherein a personalized approach to selecting the best treatment for a patient can be determined in a timely manner.Using patient-derived breast cancer cell lines to create 3D in vitro models for personalized drug selection in the treatment of metastatic breast cancer is a major step forward in this direction.
While some preclinical models, such as two-dimensional (2D) monolayer cell culture models and in vivo mice models, have been utilized by researchers for preclinical cancer research, these have several limitations.2D models are known to poorly recapitulate the in vivo complexity due to a lack of cell-microenvironment interactions, while in vivo models often fail to develop into metastatic disease [14].Tus, there is a need to create new robust preclinical models that better recapitulate human tumor biology at their advanced stage and can be used for high-throughput drug screening.Increasing evidence has shown that three-dimensional (3D) disease models derived from patients' own healthy and tumor tissue could better predict the pathogenesis of cancer cells and provide a more accurate measurement of potential drugs than existing models because these models retain characteristic features of cancer cells derived from individual patient's cells [15].Despite classifying breast cancer cells into three categories based on their cell surface receptors and growth behavior, breast cancer patients within each category can have markedly diferent disease outcomes and therapeutic responses [16].Tus, models derived from patients' cancer cells could help researchers better predict therapeutic responses.
Several attempts at developing 3D in vitro models of metastasized breast cancer have been made [17][18][19][20][21].However, such models have utilized only cancer cells to resemble the metastatic stage.In addition, eforts have been made to recapitulate bone metastasis of breast cancer by coculturing breast cancer cells with osteoblasts [22][23][24].However, such models failed to mimic the ideal in vivo conditions of breast cancer metastasis to bone due to inaccurate representation of the bone microenvironment where cancer cells interact with remodeled bone.Tus, to address this issue, we developed a novel 3D in vitro bone metastatic scafold model using a tissue-engineered approach, where bone marrowderived hMSCs diferentiate into bone cells and generate extracellular matrix (ECM) for breast cancer dissemination to better recapitulate breast cancer bone metastasis [25,26].Tese bone metastatic scafolds possess high porosity (86.1%) with a pore size range between 100 and 300 μm and exhibit a high compressive modulus of 2.495 MPa, essential for hard tissue growth [27].Previously, we utilized primary human breast cancer cells-MCF-7 and MDAMB231 to develop this 3D in vitro breast cancer bone metastasis model [26] and investigated the role of the Wnt/β-catenin pathway in osteogenic diferentiation of hMSCs on the scafold surface during breast cancer bone metastasis [28].
Te present study aims to understand the metastases of patient-derived breast cancer cells to bone and their role in hMSCs osteogenic diferentiation.We evaluated the efect of cancer on bone growth via the Wnt/β-catenin pathway and compared our results with previous studies.

Cell Lines and Cell
Culture.Human mesenchymal stem cells (hMSCs) were purchased from Lonza (PT-2501) and cultured in MSCGM BulletKit medium (Lonza, PT3001).Human breast cancer cell lines NT013 and NT023 were derived from the patient tissue samples obtained from Sanford Roger Maris Cancer Center, Fargo.Te ethical committee approved the study, and before surgery, all patients provided written informed consent to allow any excess tissue to be used for research.If there was extra breast cancer tissue that would not be needed for clinical diagnosis and management, it was submitted for the study to the NDSU team for further study.Te tissue samples were excised from the primary site (breasts) of patients.After surgery, the pathologist reviewed the excised breast tissue to confrm the presence of cancer cells.Te cancer cells present in NT013 and NT023 breast tissue specimens were characterized as hormone-positive (ER/PR positive) and triple-negative (ER/PR/HER2 negative), respectively, by the clinical pathologist.Patient tissue samples were transported to the research lab using a transportation medium containing DMEM, 1% of pen/strep mix (100x), gentamicin (10 mg/ml), and amphotericin B (250 μg/ml).Breast cancer cells were isolated using a cell isolation kit (Miltenyi Biotec-130-095-929) following the manufacturer's protocol and cocultured with irradiated 3T3-J2 feeder cells (Kerafast) (Figure 1(a)).Finally, cells were maintained at 37 °C and 5% CO 2 in high glucose DMEM containing 5 μg/ml insulin, 250 ng/ml amphotericin B, 10 μg/ml gentamicin, 0.1 nM cholera toxin, 0.125 ng/ml epidermal growth factor (EGF), 25 ng/ml hydrocortisone, ROCK inhibitor Y-27632 10 μM, 10% (v/v) FBS, and 100 U/mL penicillin and 100 μg/mL streptomycin.

Cell
Seeding.Scafolds were sterilized in 70% ethanol for 24 hours, further sterilized under UV light for 45 min, washed twice in phosphate-bufered saline (PBS), and fnally immersed in the culture medium and incubated for 24 hours in a humidifed 5% CO 2 incubator at 37 °C.hMSCs were seeded at a density of 1 × 10 5 cells per scafold and cultured for 23 days to obtain tissue-engineered bone on the scafold surface.Next, patient-derived breast cancer cells were seeded at a density of 1 × 10 5 cells per scafold on the tissue-engineered bone and maintained in the breast cancer cells medium (Figure 2(a)).Te media was changed every two days during both hMSCs and sequential culture of breast cancer cells on the scafold surface.

2
Journal of Tissue Engineering and Regenerative Medicine 2.4.Immunofuorescence Staining.Both 8-well chambers (Termo scientifc) seeded and scafold seeded cells were washed twice in PBS and fxed in 4% paraformaldehyde (PFA) for 30 min.Next, cells were permeabilized with 0.2% TritonX-100 in PBS for 5 min, followed by blocking with blocking bufer (0.2% fsh skin gelatin (FSG) with 0.02% Tween20) for 1 hour.Furthermore, the cells were incubated with the primary antibody overnight at 4 °C.Te primary antibodies were diluted in the blocking bufer using dilutions given in Table S1.Finally, cells were incubated with conjugated secondary antibodies corresponding to the species of used primary antibodies at 1 : 200 dilutions and incubated for 45 minutes at room temperature (RT).Te nuclei were counterstained with 4,6-diamidino-2-phenylindole (DAPI), and immunofuorescence images were taken under a confocal microscope (Zeiss Axio Observer Z1 LSM 700).

Gene Expression by RT-qPCR.
RNA was isolated from cells grown on TCPS (2D) and scafolds using a Direct-zol RNA MiniPrep kit (Zymo Research) following the protocol described elsewhere [33].Briefy, 1000 ng of RNA was reversed transcribed to cDNA using random primers and M-MLV reverse transcriptase (Promega) in a thermal cycler using a thermal profle-70 °C for 5 minutes (Applied Biosystems).Next, the qPCR experiment was performed using a 7500 Fast Real-Time PCR instrument (Applied Biosystems) using a thermal profle with a holding stage (5 min at 95 °C) and a cycling stage (40 cycles of 30 s at 95 °C and 1 min at 55 °C).Te mRNA expressions of genes (listed in Table S2) were quantifed using their respective primers and normalized to the housekeeping gene glyceraldehyde-3phosphate dehydrogenase (GAPDH).Finally, fold change in target gene expressions was calculated using the comparative Ct method (2 −ΔΔCt ).
2.6.ELISA Assays.Te released DKK-1 and ET-1 cytokines concentration was measured in serum-free cell culture media using high-sensitivity ELISA kits of DKK-1 (Ray-Biotech) and ET-1 (RayBiotech) as per the manufacturer's protocol.Te cell-seeded scafolds were kept in a serum-free medium for 48 hours before collecting the medium for sample preparation (Figure 3(a)).Next, the medium was centrifuged at 350 × g for 10 minutes at 4 °C to remove cell debris, and supernatants were stored at −20 °C until analysis.

Statistical Analysis.
Data were presented as the mean value ± standard deviation.Statistical signifcance between the two groups was determined by an unpaired Student's ttest or one-way or two-way ANOVA followed by Tukey's post-test using GraphPad Prism v7.04 software.Te signifcance level was set at p ≤ 0.05."n" represents the technical replicates of each experiment.that cancer cells isolated from patient tissue samples retain their tumor-associated features.Breast cancer cells are broadly classifed into three categories; hormone-positive, triple-negative, and HER2-positive, based on the expression of hormone receptors-estrogen (ER) and progesterone (PR) and human epidermal growth factor receptor 2 (HER2) [16].

Isolated Patient-Derived Cancer Cells
After characterizing NT013 patient-derived cells, we observed that NT013 retained their hormone-positive characteristics by expressing ER.Interestingly, we also observed positive HER2 expression in NT013 cells.However, other studies on NT013 patient tissue by pathologists confrmed reduced HER2 levels (data not shown), suggesting that NT013 cells can be categorized into hormone-positive breast cancer cells.Next, we evaluated similar protein expression in NT023 cells and observed that NT023 cells do not express ER, PR, and HER2 receptors, retaining their triple-negative characteristics.
Cytokeratin-19 (CK19) is also a suitable marker for identifying breast cancer cells [34,35].CK19 is an epithelial cell marker, and its expression is seen in more than 90% of breast cancer cases.It is also reported that luminal-type hormonal positive cells exhibit higher positive rates of CK19 than triple-negative cells [34,36].We observed protein expression of CK19 in both cell lines; however, the expression was more intense in NT013 cells compared to NT023 cells.Next, we analyzed Epithelial to Mesenchymal transition (EMT) markers for NT013 and NT023 cells to identify their invasive nature.Epithelial cells can be  identifed by expressing the epithelial protein E-cadherin on their cell surface.In contrast, mesenchymal cells can be identifed by various protein expressions such as Ncadherin, vimentin, and twist [37].We observed that NT013 cells expressed both E-cadherin and vimentin protein expressions, while NT023 cells mostly expressed high vimentin levels, indicating that NT023 cells are more mesenchymal in nature compared to NT013 cells (Figure 1).

Bone Microenvironment-Induced MET in Breast Cancer
Cells.EMT/MET processes represent the invasiveness of cancer cells where cancer cells leave their primary site to acquire migratory phenotype during EMT.At the same time, MET potentiates cancer cells to regain their epithelial characteristics and adapt to the new environment at their secondary site [38][39][40].To investigate the efect of the bone microenvironment on patient-derived breast cancer cells' invasiveness, we analyzed their mRNA levels related to EMT/MET biomarkers such as E-cadherin and N-cadherin and compared our results with cells grown on a 2D surface.E-cadherin is a cell surface protein that participates in forming homotypic junctions across epithelial cells [41].Te loss of E-cadherin, while the gain in N-cadherin levels is associated with the EMT process of cancer cells [37], and vice-versa is valid for the MET process [39].Previously, we observed that the primary breast cells-MCF-7 and MDAMB231 underwent mesenchymal to epithelial transition due to upregulation of E-cadherin and downregulation of vimentin and twist levels in the presence of bone [26].Likewise, we observed a signifcant increase in E-cadherin levels ( * * * p < 0.001) in NT013 (∼2-fold) and NT023 (∼4fold) breast cancer cells grown in the bone microenvironment compared to their respective 2D cell cultures.We also observed signifcant downregulation ( * p < 0.05) in Ncadherin levels in NT013 cells and insignifcant change in N-cadherin levels in NT023 cells, indicating that both NT013 and NT023 cells acquired more epithelial characteristics in the bone microenvironment (Figure 2).

Te Bone Microenvironment Induces Aggressiveness and
Angiogenesis in Patient-Derived Cells.Wnt5a is an important member of the Wnt pathway and acts as either tumorsuppressive or tumor-promoting in diferent cancer types [42].Lower levels of Wnt-5A expression are signifcantly associated with poor prognosis and more aggressive behavior in triple-negative breast cancer [43,44].Similarly, β-catenin is highly expressed in breast cancer patients [45] and is signifcantly associated with poor clinical outcomes in invasive breast cancer [46].To evaluate the aggressive behavior of breast cancer cells in the presence of bone, we quantifed their Wnt-5A and β-catenin levels.Our results showed signifcant downregulation ( * * * p < 0.001) in Wnt-5A levels in NT023 growing on the bone compared to 2D culture, while we didn't observe any signifcant change in Wnt-5A in NT013 cells.However, we observed signifcant upregulation ( * * * p < 0.001) in β-catenin levels in both NT023 and NT013 cells, indicating increased aggressiveness in the presence of bone.
VEGF is a well-known marker of angiogenesis, highly expressed in many solid tumors resulting in a poor prognosis of the disease [47].We observed upregulation in the VEGF mRNA levels in both NT013 ( * p < 0.05) and NT023 ( * * * p < 0.001) cells grown on bone compared to their 2D cultures, respectively, indicating increased angiogenesis in both cell types in the presence of bone (Figure 2).

Tumor Formation by Patient-Derived Cell Lines on Bone
Niche.To investigate the morphology of cancer cells on tissue-engineered bone after MET, we stained the cells with cancer-specifc protein, EpCAM, and compared our results with cells grown on a 2D surface.EpCAM is a transmembrane protein signifcantly overexpressed in breast cancer tissues [48].We observed that both NT013 and NT023 cell lines formed tumors on tissue-engineered bone.In contrast, cancer cells in their monoculture did not form any tumors.We also noticed that NT013 cells formed compact tumors on the bone microenvironment, exhibiting distinguishable cellular boundaries with strong cell-cell interactions, while NT023 cancer cells grouped into clusters (with moderate cell-cell interactions) instead of forming compact tumors.Previously, we observed that hormone-positive MCF-7 cells formed dense tumors on bone scafolds [26], similar to NT013 breast cancer cells.In contrast, triple-negative MDAMB 231 cells formed loose aggregates (with poor cell-cell interaction) [26], diferent from NT023 cells (Figure 2(b)), indicating that inherent characteristic diferences among two diferent triplenegative cells could alter their tumor-forming ability after interacting with the bone microenvironment.

DKK-1 and ET-1 Factors Released by Breast Cancer Cells
Regulate the Osteogenic Wnt/β-Catenin Pathway.ET-1 and DKK-1 are well-known markers of bone metastases that induce osteoblastic or osteolytic lesions, respectively, in breast cancer patients resulting in the poor mechanical stability of the bone [49,50].It is reported that serum DKK-1 levels are higher in patients with bone metastasized breast cancer than at other metastatic sites [51].Previously, we observed that MCF-7 cells grown on bone released high levels of ET-1 in serum-free media, whereas MDAMB231 cells released high levels of DKK-1, leading to stimulation and inhibition of osteogenesis via the Wnt/ β-catenin pathway [25].To investigate the efect of NT013 and NT023 released cytokines on bone health via Wnt signaling, we frst quantifed ET-1 and DKK-1 levels in the sequential culture of NT013 and NT023 and observed that NT013 cells released high levels of ET-1 ( * * * p < 0.001) while NT023 cells released high levels of DKK-1 ( * * * p < 0.001) in line with our results with MCF-7 and MDAMB231 cell lines, respectively.Next, to determine that ET-1 and DKK-1 released by patient-derived cells are involved in the regulation of the Wnt/β-catenin signaling mediated bone osteogenesis, we analyzed the expressions of Wnt-related genes of MSCs on Day (23 + 10) maintained under diferent conditioned mediums.We observed upregulation in both Wnt 5a ( * * * p < 0.001) and β-catenin ( * * p < 0.01) expressions in 6 Journal of Tissue Engineering and Regenerative Medicine hMSCs cultured with conditioned media of sequentially cultured NT013 cells containing high ET-1 levels.In contrast, hMSCs cultured with conditioned media of sequentially cultured NT023 cells containing high DKK-1 levels showed downregulation of Wnt 5a ( * * * p < 0.001) and β-catenin ( * * p < 0.01) levels compared to control MSCs samples (Day 33).We also assessed the expression of a latestage osteogenic marker, osteocalcin (OCN) in hMSCs cultured under diferent conditioned media w.r.t to control samples.We noticed a signifcant increase ( * * * p < 0.001) in mRNA levels of OCN in hMSCs cultured with conditioned media of sequentially cultured NT013 cells while downregulation in OCN levels ( * * * p < 0.001) in hMSCs cultured with conditioned media of sequentially cultured NT023 cells.Overall, the results suggested that cytokines released by NT013 cells stimulate Wnt/β-catenin signaling in hMSCs while cytokines released by NT023 cells abrogate the Wnt/β-catenin pathway, thus promoting and inhibiting osteogenesis, respectively (Figure 3).

Discussion
Many 3D in vitro models have been developed to recapitulate breast cancer bone metastasis disease conditions [52].However, existing breast cancer bone metastatic models attempted to mimic late-stage breast cancer by coculturing breast cancer cells with osteoblasts that do not resemble the ideal conditions of breast cancer metastasis to bone in vivo [22][23][24].In these coculture systems, diferent cell types were seeded together on the scafold surface; however, in ideal conditions, breast cancer cells migrate to the remodeling bone, where cancer cells interact with diferentiated bone cells and bone microenvironment to disseminate.Our 3D in vitro breast cancer bone metastatic model accurately represents the late stage of breast cancer metastasis to the bone, where we implemented a sequential culture system (Figure 2(a)).In the sequential culture, hMSCs diferentiated into osteoblastic lineage on a nano clay-based scafold along with calcium deposition [53] and collagen formation [27], thus generating a remodeling bone microenvironment for breast cancer metastasis.Previously, we have successfully developed a 3D in vitro bone metastatic model using primary breast cancer cell lines-MCF-7 and MDAMB231 [26].Our results showed that MCF-7 and MDAMB231 breast cancer cells underwent MET and formed tumors in the bone microenvironment.Moreover, their interaction with bone cells induces the release of cytokines that further infuence bone growth via the Wnt/ β-catenin pathway [25].In the present study, we have developed a 3D in vitro breast cancer bone metastatic model using patient-derived breast cancer cells replacing primary cell lines to predict the metastasis of a patient's breast cancer cells to bone originating from the primary site of the breast.Te isolated NT013 and NT023 breast cancer cells from the patient specimens were characterized as hormone-positive and triple-negative, respectively.In line with our previous results [26], we observed the occurrence of MET in both patient-derived breast cell lines in the presence of bone due to the upregulation of E-cadherin and the downregulation of Ncadherin levels.However, we also noticed a dissimilarity in fold change of E-cadherin levels in NT013 and NT023 breast cancer cells grown in the bone microenvironment.Te possible reason for a variation in fold change of E-cadherin levels can be attributed to inherent low levels of E-cadherin expression in NT023 that upregulated substantially in the presence of bone microenvironment.In contrast, NT013 cells inherently exhibit a high E-cadherin expression; thus, fold change was not so high.MDAMB231 cells also express low Ecadherin levels inherently [54].Previously we observed that MDAMB231 cells formed loose aggregates (with poor cell-cell interactions) in the presence of bone because fold change in upregulated E-cadherin levels was not so high [25].However, in the present study, we observed that NT023 cells formed clustered tumors (with moderate cell-cell interactions) in the presence of bone, suggesting that relatively high E-cadherin levels of NT023 cells stimulate them to form tumors. Te Wnt/β-catenin pathway has been well-known for regulating bone formation in vivo and osteoblast diferentiation in vitro [55,56].Our results showed that excessive release of ET-1 and DKK-1 by hormone-positive NT013 cells and triplenegative NT023 cells stimulated and abrogated the Wnt/ β-catenin pathway, respectively.Our previous results also revealed a similar trend by hormone-positive MCF-7 and triplenegative MDAMB231 cells where stimulation and abrogation of the Wnt/β-catenin pathway occurred by ET-1 (released by MCF-7) and DKK-1(released by MDAMB231), respectively [25].Tus, our current results are in good agreement with our previous studies [25] and the reported literature [49,50].We have also demonstrated upregulation in OCN levels in the presence of ET-1, indicating excessive bone formation due to increased hMSCs osteogenesis, while downregulation in OCN levels due to the inhibitory efect of DKK-1 leading to inhibited bone formation, in line with our results of primary cell lines [25] and reported studies on bone formation in vivo [49,[57][58][59].
In summary, our 3D in vitro model showed both excessive and inhibitory bone growth by the cytokines released from patient-derived cell lines NT013 and NT023 exhibiting diferent cell characteristics by altering the physiological Wnt/β-catenin signaling pathway in a healthy bone.Terefore, this model is suitable for investigating the metastases of cancer to bone and underlying signaling mechanisms during bone metastasis.For more advancement or to better recapitulate bone metastases, we have planned to utilize the patient's hMSCs for model development.In addition, future studies are designed to screen potential drugs to target bone metastasized breast cancer.

Conclusion
A better understanding of the complex interactions between breast cancer cells and the bone microenvironment is of paramount importance for improving the outcome for latestage breast cancer patients.One of the critical challenges associated with poor prognosis is the lack of reliable models for studying breast cancer at its advanced stage.In the present study, we utilized a 3D in vitro nanoclay-based breast cancer bone metastatic model, previously developed using primary breast cancer cell lines, to investigate the efect of patient-derived breast cancer cells on bone Journal of Tissue Engineering and Regenerative Medicine growth.We demonstrate that patient-derived breast cancer cells retain their idiosyncratic characteristics after isolating, using the most efcient method for cancer cell isolation from solid tumors.Te model can mimic the MET process of breast cancer metastasis and reveal excessive and inhibitory bone growth by breast cancer cell lines of diferent characteristics via Wnt/β-catenin signaling, mimicking bone lesions observed in breast cancer patients in their late stages.Te 3D in vitro breast cancer models using patient-derived cells recapitulate the metastatic ability of breast cancer cells to bone.However, future studies are planned to utilize the patient-derived MSCs to develop bone on these scafolds for more advancement.Tese models could be a viable tool for future breast cancer studies, including investigating metastatic molecular mechanisms and screening novel drugs.

Figure 1 :
Figure 1: (a) Schematic showing the steps in isolation of breast cancer cells from the patient tissue samples (b), (c) representative immunofuorescence microscope images of NT013 and NT023 cells cultured in 2D culture.Scale bar: 20 μm.n � 3 (b) and (c).