Human breast cancer epithelial cell lines, MDA MB-231 (HTB26), and MCF-7 (HTB-22), and human non-tumorigenic breast epithelial cell line, MCF-10 A (RL-10317), were purchased from ATCC, United States. All the cell lines, including KAIMRC1, were maintained in advanced Dulbecco’s modified Eagle medium (DMEM) (Invitrogen) supplemented with 10% fetal bovine serum (FBS) (FBS-DMEM), 50 U/ml penicillin (Gibco) and 50 μg/ml streptomycin (Gibco), and 2 mM L-glutamine (Gibco) at 37°C in a humidified 5% CO₂ atmosphere. MCF-10A was maintained in 1:1 DMEM/F12 media (Invitrogen) supplemented with 5% equine serum, 0.5 μg/ml hydrocortisone, 100 ng/ml cholera toxin, 10 μg/ml insulin, 20 ng/ml epidermal growth factor (EGF), 50 U/ml penicillin and 50 μg/ml streptomycin, and 2 mM L-glutamine. Cells were cultured at 37°C in a humidified 5% CO₂ atmosphere.

KAIMRC1 cells were cultured in a 96-well cell culture plate (Nunc, Thermo Fisher Scientific) at a density of 10,000 cells/well. One half of the plate was incubated with advanced DMEM supplemented with FBS to generate KAIMRC1 monolayer, and the other half was incubated with advanced DMEM supplemented with 10% new born calf serum (NBCS), (NBCS-DMEM), to generate KAIMRC1 spheroids, at 37°C in a humidified 5% CO₂ atmosphere. The following day, cell culture plate was monitored under a microscope to look for spheroids formation before any downstream experiments.

Human breast cancer tumor samples were collected after obtaining the informed written patients’ consent following the guidelines set by the KAIMRC Ethics Committee. The sample was washed several times with phosphate-buffered saline (PBS), minced and cut into smaller pieces, and incubated normal cell culture medium in 24-well cell culture plates at 37°C in a humidified environment. The tissue culture was monitored microscopically on a daily basis to look for adherent fibroblast-like cells protruding out of the tissue pieces. After several days of culture, tissue pieces were removed, and the adherent cells were trypsinized and subcultured. These cells were incubated with fibroblast monoclonal antibody (D7-FIB, MA5-16642, Invitrogen) and α-smooth muscle actin (1 μg/ml, Abcam, UK), overnight at 4°C. It was followed by Alexa Fluor 488-conjugated secondary antibody at room temperature for 1 h. The cells were analyzed using flow cytometry. These cells were further passaged and used in the co-culture studies.

Primary breast CAFs were isolated as described above and cultured in a 96-well plate at 5,000 cells/well. At 90% confluency, cells were stained with Live CellTracker™ Orange CMTMR dye (10 µM, Cat. No. C2927, Invitrogen) for 1.5 h in a humidified incubator. At the same time, KAIMRC1 cells were subcultured and stained with Live CellTracker™ Green CMFDA dye (2.5 µM, Cat. No. C7025, Invitrogen) for 45 min in a humidified incubator. Both CAFs and KAIMRC1 were washed 2x with PBS and stained with HOECHST 33342, trihydrocholoride, trihydrate (1 μg/ml, Cat. No. H3570, Invitrogen) for 15 and 5 min, respectively. KAIMRC1 cells were then seeded on the CAFs monolayer at a density of 2,500 cells/well. This co-culture was then monitored every day for 6 days, and images were acquired using the Zeiss LSM780 microscope system.

Single KAIMRC1 cell spheroids were cultured as described above in a six-well plate, and images were acquired every 24 h for 14 days to measure the size distribution of the spheroids under an EVOS FL auto system equipped with live-cell imaging capabilities. We counted at least 40 spheroids per well to demonstrate the size distribution of KAIMRC1 spheroids.

To study the effect of compound treatment, KAIMRC1 monolayer and spheroids were treated with a large panel of commercially available compounds. Briefly, KAIMRC1 cell monolayer and spheroids were seeded at 5,000 cells/well into 96-well plates and incubated at a 5% CO₂ incubator overnight. Subsequently, cells were treated with compounds in serial dilution followed by incubation for 48 h. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was performed after 48 h. Five microliters of MTT (5 mg/ml) was added to the cell culture wells and incubated for 4 h. The plate was centrifuged at 500xg for 5 min to pull down all the floating spheroids. Later, the media was aspirated carefully avoiding the disruption of spheroids. The plates were then dried on paper towels followed by the addition of dimethyl sulfoxide (DMSO). The plates were incubated with DMSO on a shaker for 1.5 h; the spheroid wells were observed under a microscope to look for any unsolubilized spheroids. The absorbance was read at 560 nm using a Molecular Devices Spectra Max Plus 384 spectrophotometer. Background readings were collected at 620 nm. Absorbance readings were normalized and expressed as a relative percentage. The data were analyzed in GraphPad Prism 8 software, and the half-maximal inhibitory concentration (IC₅₀) was determined. Error bars denote standard deviation (SD). The statistical significance of differences was determined using the Students’ t-test. p < 0.05 was considered statistically significant.

KAIMRC1 spheroids and monolayers were plated in 96-well plates at a density of 5,000 cells per well. Cells were treated with graded concentrations of Mitoxantrone ranging from 50 to 0 µM for 48 h. Post-treatment, cells were stained with calcein AM (2 μg/ml), HOECHST33342 (2.5 μg/ml), and propidium iodide (2.5 μg/ml) for 10 min at 37°C and 5% CO₂. Plates were imaged using a Molecular Devices ImageXpress^® Micro and analyzed using MetaXpress^® software, Molecular Devices, Downingtown, PA, United States. Live/Dead module available in the MetaXpress software was used to count live and dead cells. The acquired values were fed in the GraphPad Prism 8 software to calculate IC₅₀. All experiments were performed in triplicates.

Total RNA isolation from KAIMRC1 cells and spheroids was performed using an RNA Isolation Kit (Ambion) following manufacturer instructions, and RNA was quantified with a Nanodrop 8000 spectrophotometer (Thermo Scientific). One microgram of total RNA was reverse transcribed into cDNA using High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) and was diluted 1:2 in ddH₂O.

The manufacturer’s protocol was followed for both the panels. The PCR amplification was performed on a 7900HT Fast Real-Time PCR System (Applied Biosystems). The data was analyzed in RQ Manager 1.2.1 software (Applied Biosystems), and relative changes in mRNA expression levels were calculated by the ΔΔ_Ct method in the RT2 Profiler PCR Array Data Analysis Spreadsheet Version 5.1, 4/2019. It is available on Qiagen web portal.

Cells were seeded 24 h before the experiment in µ-slide eight-well Ibitreat chamber slides (Cat. No. 80,826, Ibidi, Germany). Cells were fixed, permeabilized, and incubated overnight at 4°C with the diluted primary mouse monoclonal antibody to β-3-tubulin (TU-20; Cat. No. 4466s), mouse monoclonal antibody to a-tubulin (DM1A; Cat # 3873s), rabbit monoclonal antibody to caveolin-1 (D46G3; Cat. No. 3267s), rabbit monoclonal antibody to E-cadherin (24E10; Cat. No. 3195s), rabbit monoclonal antibody to Phopho-Akt (Ser473; D9E XP^®; Cat. No. 4060s), and β-actin (BA3R; Cat. No. MA5-15739, Thermo Fisher Scientific; 1:500). All antibodies were purchased from Cell Signaling. Secondary antirabbit Alexa Fluor^® 488 (H + L; Ref. No. A11034) and antimouse Alexa Fluor™ 647 (H + L; Ref. No. A21236) from Life Technologies were used to detect antibody fluorescence. Slides were imaged using cLSM 780 (Carl Zeiss, Germany). Cell Tracker™ Green (Life Technologies) and Alexa Fluor 488 were detected using argon laser at 488 nm/520–530 (ex/em), whereas Alexa Fluor 647 was detected using HeNe laser at 633 nm/650 (ex/em). The counterstained nucleus was stained with HOECHST 33342 and detected using a UV laser at 350 nm/460 nm (ex/em).

Flow cytometry was performed using the FACS Canto II cytometer (BD), and subsequent analysis was performed using the FACS Diva software (BD). Briefly, cells and spheroids resuspended in PBS were pelleted. Fluorescently conjugated primary antibodies α-smooth muscle 1A4 (Abcam), fibroblast monoclonal antibody (D7-FIB, MA5-16642, Invitrogen), CD24, CD234 (E-Cadherin), CD44, CD227, EpCAM (EBA-1), CD49c, CD116, SSEA-1, CD45, CD47, ALDH1A1 (LSBio), cytokeratin 18 (CK18), and SSEA-1 (CD15) were purchased from BD Pharmingen unless otherwise specified. Cells were incubated with antibodies for an hour and then analyzed on the flow cytometer. For the flow cytometry analysis, cells were gated based on FSC and SSC properties, and interval gates on histogram plots were set up for positive cell populations in FITC, PE, and APC-Cy7 channels.

Spheroids were grown and fixed as described above. Cells were then dehydrated with graded concentrations of ethanol (Sigma) and transferred to appropriate carbon taped stubs (Ted Pella, USA) for SEM. To enhance the electron conductivity, samples were coated with gold/palladium (Au/Pd) by sputter coating and examined on an FEI NanoSEM 450 scanning electron microscope at 10 kV.

Monolayer and spheroids cell culture was performed as described above. The specimen was fixed in 4% glutaraldehyde in 0.1 M PBS (pH 7.4) for 1 h. After washing in the same buffer, the sample was post-fixed in 1% OsO₄ in 0.1 M PBS (pH 7.4) for 1 h, followed by washing with PBS (3x). The specimen was then dehydrated in a series of acetone solutions and infiltrated in acetone/resin (1:1) for 1 h and then with acetone/resin (1:2) for 3 h. The specimen was then embedded in epoxy resin (Araldite) and placed in an oven at 80°C overnight to polymerize. Ultrathin sections were obtained with ultramicrotome (RMS), mounted on copper grids, and stained for contrast with heavy metal stains (uranyl acetate and lead citrate). TEM images were then collected on a JEOL-JEM 1400 operating at 120 kV using a Gatan camera with Digital Micrograph Imaging software.

KAIMRC1 monolayer and spheroids samples were fixed in 4% formaldehyde, placed in cassettes, and mounted into an automated vacuum tissue processor (Leica Biosystems, Germany). The samples were embedded and blocked in paraffin wax, and thin sections were made using a microtome (Accu-Cut^® SRM™, Sakura, Netherlands). Sections were stained with Mayer’s hematoxylin solution and counterstained in eosin–phloxine solution. Slides were examined and imaged at 10x and 40x objectives under an upright microscope (Eclipse TS100, Nikon, Japan).

KAIMRC1 monolayer and spheroids cultures were lysed, and total protein was extracted. The Western blotting analysis was performed with primary antibodies against cell–cell and cell–matrix contact proteins, namely, rabbit monoclonal antibody to integrin β4 (Cat. No. 4707s), E-cadherin (24E10; Cat. No. 3195s), FAK (D2R2E; Cat. No. 13009s), caveolin-1 (D46G3; Cat. No. 3267s), and survivin (71G4B7; Cat. No. 2808). All the antibodies were purchased from Cell Signaling and used at 1:1,000 dilution. Signals were detected using a ChemiDoc MP System (Bio-Rad) and analyzed on ImageLab software. Sample loading was examined by probing the same membrane with an anti-GAPDH antibody.

To further profile the proteins and their phosphorylation status involved in spheroids formation, several Human Proteome Profilers were utilized. Human Phospho-Kinase Array (Catalog No. ARY003B), Human Pluripotent Stem Cell Array (Catalog No. ARY010), Human Phospho-RTK Array (Catalog No. ARY001B), and Human Phospho-Immunoreceptor Array (Catalog No. ARY004B). All were purchased from R&D Systems, Minneapolis, MN, USA, and used according to the manufacturer’s protocol.

The growth kinetics of KAIMRC1 spheroids at passages 30–60 was studied. KAIMRC1 cells were seeded in six-well plates in FBS-DMEM until confluence was achieved. Then, the media was changed to NBCS-DMEM. Initially, within 4 h of incubation, the cell monolayer started to retract and formed tube-like structures, followed by the formation of big organized spheroids in the next 24 h. To confirm that the spheroid formation is a reversible process, we again switched to the FBS-DMEM growth media. Interestingly, spheroids started to adhere and spread once again on the plate surface within a few hours. Finally, spheroids were completely dispersed back as a monolayer in 48 h of culture (Figure 1B).

To identify the optimal percentage of NBCS required to initiate spheroid formation, we incubated KAIMRC1 cells with a graded concentration of NBCS-DMEM, i.e., 1, 2.5, 5, 10, and 20% NBCS. The first two concentrations, 1% and 2.5%, showed monolayer and aggregation of KAIMRC1 cells, but no detachment from the plate surface; 5% showed many detached spheroids along with attached aggregates; whereas 10% and 20% concentrations showed complete detachment and floating spheroids of KAIMRC1 cells (data not shown).

When cultured in the NBCS-DMEM condition, KAIMRC1 cells initially aggregated together. During the aggregation process, cells fuse, forming strings, and later these strings round up to form spheroids. These spheroids then increased in size for several days before reaching a critical point, i.e., around the seventh day when necrosis is visible at the center of the spheroids. At this point, spheroids are no longer organized and start disintegrating by shedding cells (Figure 2A). Most of these excluded cells are attributed to having lost essential cell–cell adhesion molecules (Stadler et al., 2018). However, few among these single cells were able to divide and make smaller spheroids (Figure 2B). Single cells are pointed out with a white arrow in Figure 2B showing spheroid formation over time.

Spheroids size distribution assay exhibited that the size of KAIMRC1 spheroids ranges from 50 to 700 µm with a mean size of 55.8 µm (SD ± 3.5), and the maximum size can be reached within 14 days of culture (Figure 2C). Notably, a necrotic center can be seen after 6 days of culture. This result suggests that an approximate size of 300 µm that can be reached in 3–4 days might be optimal to be used in drug screening applications.

To study the effect of NBCS-DMEM on other well-established breast cancer cell lines, MDA MB-231, MCF-7, and MCF-10A, we incubated these cell lines both with FBS-DMEM and NBCS-DMEM for 48 h. Only KAIMRC1 cells showed spheroid formation in NBCS-DMEM, whereas other cell lines did not show any signs of spheroid formation (Figure 3A). Therefore, it was confirmed that NBCS is triggering spheroid formation in KAIMRC1 cells only.

Western blot analysis of monolayer and spheroids was performed on several key cell–cell interactions and adhesion protein. Results demonstrated an increase in the expression of E-cadherin, a member of adherens junction transmembrane protein; focal adhesion kinases (FAK), a cytoplasmic tyrosine kinase protein involved in cell survival and migration; Caveolin-1, a plasma membrane protein involved in cell adhesion; Survivin, an antiapoptotic protein highly expressed in malignant cells; and integrin β4, a heterodimeric cell surface receptor involved in cell adhesion, growth, and survival. (Figure 3B). Protein profiling and immunocytochemistry also confirmed the overexpression of E-cadherin, caveolin-1, and FAK in KAIMRC1 spheroids (Figure 4 and Figure 5A).

We also performed time-dependent flow cytometric analysis of breast cancer and stem-cell-specific markers. Cells were collected at three time points 0 h (monolayer) and 3 and 48 h. Interestingly, there was an increase in the CD44⁺ cells at 48 h timepoint and significant decrease in CD49c at 3 and 48 h. CD324 and cytokeratin 18 expression also increase at 48 h.

\In comparison to the monolayer, protein profiling of spheroids also showed notable decrease in the expression of ERK1/2, GSK3 α/β, CREB, insulin receptor, and IGF-I receptor, whereas an increase in the expression of p53, c-Jun, ChK-2 (Figure 4).

Gene expression profiling of spheroids revealed that most of the key cell adhesion proteins were downregulated during the initial phase of cell aggregation, i.e., 3 h except GJC2. However, high upregulation of key cell junction genes was noticed when spheroids were 48 h old. For instance, CDH1, a gene that makes E-cadherin protein, was significantly downregulated at 3 h, whereas it was highly upregulated at 48 h (Figure 6C). Our immunophenotyping results have also confirmed the upregulation of E-cadherin on protein level.

Interestingly, most of the genes were downregulated during the initial phase of cell aggregation (3 h) except GJC2. High upregulation of key cell junction genes was noticed when spheroids were 48 h old. For instance, CDH1 was significantly downregulated at 3 h, whereas it was highly upregulated at 48 h.

3D cell culture systems with more than one type of cells mimic more closely the real in vivo environment. Here, we have demonstrated the ability of KAIMRC1 spheroids to make 3D co-cultures with other cell types, especially with primary breast cancer-associated fibroblasts (CAFs). We co-cultured KAIMRC1 cells with primary CAFs exhibiting the potential of KAIMRC1 spheroids to grow in co-culture with other cell types (Figure 5B and Supplementary Figure S1). The ultimate goal was to use these co-cultured spheroids in drug discovery applications.

Next, we performed drug treatment and high content imaging of KAIRMC1 monolayer and spheroids. Commercially available toolbox compounds categorized into epigenetics, stem cells, and kinase inhibitors were used to treat KAIMRC1 cells in monolayer and spheroids at 10 µM concentration. The bar graph shows percent survival of cells after treatment for 24 h. It is evident from the results that KAIMRC1 spheroids showed better survival against most of the compounds. Interestingly, several compounds showed more specificity towards KAIMRC1 spheroids than monolayer (Figure 7A). Figure 7B presents high content imaging of KAIRMC1 monolayer and spheroids after Mitoxantrone treatment. KAIMRC1 cells in monolayer and spheroids were treated with a graded concentration of Mitoxantrone for 24 h. High-content image acquisition and analysis software, MetaXpress, was used to analyze monolayer and spheroid images using standard application modules to measure live and dead cells. KIAMRC1 spheroids can be used under environmental control for days with live-cell trackers to monitor cell health kinetics. For instance, we stained KAIMRC1 cells with Live Cell Tracker Green and HOECHST 33342 in monolayer and later changed the serum condition to NBCS to yield spheroids. Post 24 h, cells were treated with drug candidates. The cells in both monolayer and spheroids conformation were imaged for 4 days to monitor the health kinetics.