Heteromulticellular Stromal Cells in Scaffold-free 3D Cultures of Epithelial Cancer Cells to Drive Invasion

Summary

There is a critical need for 3D cancer models that capture heterocellular crosstalk to study cancer metastasis. Our study presents the generation of heteromulticellular stromal-epithelial in a scaffold and scaffold-free environment that can be used to study invasion and cellular spatial distributions.

Abstract

Breast cancer is the second leading cause of cancer-related death among women in the U.S. Organoid models of solid tumors have been shown to faithfully recapitulate aspects of cancer progression such as proliferation and invasion. Although patient-derived organoids and patient-derived xenograft organoids are pathophysiologically relevant, they are costly to propagate, difficult to manipulate, and comprised primarily of the most proliferative cell types within the tumor microenvironment (TME). These limitations prevent their use for elucidating cellular mechanisms of disease progression that depend upon tumor-associated stromal cells which are found within the TME and known to contribute to metastasis and therapy resistance.

Here, we report on methods for cultivating epithelial-stromal multicellular 3D cultures. The advantages of these methods include a cost-effective system for rapidly generating organoid-like 3D cultures within scaffold-free environments that can be used to track invasion at single-cell resolution within hydrogel scaffolds. Specifically, we demonstrate how to generate these heteromulticellular 3D cultures using BT-474 breast cancer cells in combination with fibroblasts (BJ-5ta), monocyte-like cells (THP-1), and/or endothelial cells (EA.hy926). Additionally, differential fluorescent labeling of cell populations enables time-lapse microscopy to define 3D culture assembly and invasion dynamics.

Notably, the addition of any two stromal cell combinations to 3D cultures of BT-474 cells significantly reduces circularity of the 3D cultures, consistent with the presence of organoid-like or secondary spheroid structures. In tracker dye experiments, fibroblasts and endothelial cells co-localize in the peripheral organoid-like protrusions and are spatially segregated from the primary BT-474 spheroid. Finally, heteromulticellular 3D cultures of BT-474 cells have increased hydrogel invasion capacity. Since we observed these protrusive structures in heteromulticellular 3D cultures of both non-tumorigenic and tumorigenic breast epithelial cells, this work provides an efficient and reproducible method for generating organoid-like 3D cultures in a scaffold-free environment for subsequent analyses of phenotypes associated with solid tumor progression.

Introduction

Cancer progression is now recognized to be dependent on two major factors: the genetic/epigenetic changes in the tumor cells and a myriad of interaction with non-tumor cells in the tumor microenvironment (TME)¹. While genetic changes in cells are acknowledged to be necessary for tumor initiation, such alterations alone are not sufficient for tumor progression and metastasis². Components of the TME, originally thought to be silent bystanders, are now known to actively promote cancer progression via mutual and dynamic crosstalk with tumor cells³. The TME composition differs depending on the tissue the tumor originates from, the tumor stage, and patient characteristics, but hallmark features include stromal cancer-associated fibroblasts (CAFs), extracellular matrix (ECM), vascular endothelial cells, as well as adaptive and myeloid immune cells^1,4.

Stromal CAFs in the TME are composed of fibroblast subtypes of diverse origins and functions⁵. Such CAFs are key components of the TME as they interact with tumor cells at several interfaces. CAFs secrete ECM proteins that alter the stiffness of the matrix, which may either limit drug delivery via excessive deposition of collagen, proteoglycans, and fibronectin or allow tumor cells to invade from the primary tumor site via secretion of ECM-degrading matrix metalloproteinases (MMPs)^6,7. In addition, CAFs promote tumor growth, migration, and vascularization via the secretion of a variety of growth factors, cytokines, and angiogenic factors such as epidermal growth factor (EGF), transforming growth factor β (TGF-β), and vascular endothelial growth factor (VEGF), respectively^1,6. In parallel, endothelial cells, prompted by the hypoxic TME, also promote tumor vascularization and suppress immune cell functions via the increased secretion of angiogenic factors and lowered secretion of leukocyte adhesion molecules^1,8.

With the apparent intricate complexity of cancer progression, it has become essential to incorporate TME stromal components in basic cancer research. However, the establishment of models that faithfully recapitulate known tumor pathophysiology is still a significant unmet need^9,10. While traditional two-dimensional (2D) cell culture models are easy to handle, rapidly cultured, and are highly reproducible, it is comprised of only rapidly proliferating cancer cell clones and do not reflect the cellular heterogeneity found in tumors^10,11,12. In a similar manner, transgenic mouse models also do not capture human tumor biology due to low genetic heterogeneity from inbreeding, significant differences in the immune system, and histological complexity^13,14. Due to such limitations, therapeutics developed from classical cancer models often fail to translate to clinical settings.

Patient-derived cancer models such as patient-derived xenografts and patient-derived organoids can address drawbacks of conventional cancer models by capturing in-situ tumor molecular features, genetic background, and cellular organization^10,11,15. However, such patient-derived xenografts and organoids require complicated engraftment procedures and long culturing time^16,17. Combined with variation in tumor acquisition and sampling sites and poor efficiency in cryopreservation, there is a need to develop models that act as a bridge between classical 2D cell cultures and patient-derived cancer models^11,18. In this regard, 3D models of cell culture can serve as models that can be cultured rapidly and capture important in vivo tumor features such as cell-cell interaction, cell-ECM interaction, hypoxia, angiogenesis, and the production of ECM^19,20.

3D cell culture models are categorized into scaffold-free and scaffold-based model systems. In scaffold-free systems, cells are induced to self-aggregate into a spherical shape by using specific low-attachment cell culture plates or by manipulating the physical parameters of culturing methods. Established methods to obtain scaffold-free 3D spheroids range from simple cell pellet cultures by centrifugation to hanging microplate drops, magnetic levitation, and dynamic bioreactor and microfluidic systems^20,21. Scaffold-based 3D cell cultures are established by the addition of polymer- or hydrogel-based scaffolds to imitate the physiological extracellular matrix^19,22. Such models hold immense potential to model in vivo cellular organization, topology, matrix attachment, migration, and drug response.

In addition to scaffold manipulation for ECM composition models in disease states, 3D cell cultures can also be used to model heterogeneous cellular populations in a TME. 3D cell cultures composed of cancer cells, and stromal fibroblasts or endothelial cells, have been used to study the interaction of cancer and individual non-tumor cell lines^23,24,25. Reproducible and cost-effective methods for expanding such 3D cell cultures composed of multiple heterogeneous cell lines would help researchers elucidate tumor progression. Here, we report on methods for cultivating epithelial-stromal multicellular 3D cultures to study proliferation, invasion, and cell state plasticity. The protocol describes scaffold-free and basement membrane extract scaffold-based 3D cultures of breast cancer cells co-cultured with a combination of stromal cells ranging from fibroblasts (BJ-5ta), endothelial (Ea.hy926) cells, and monocytes-like cells (THP-1). Breast cancer is currently the second most common cancer worldwide and the most diagnosed cancer in women in the USA²⁶. Death from breast cancer is largely due to the metastatic and therapy-resistant nature of the disease as overall and metastasis-free survivorship is significantly reduced in patients diagnosed with aggressive HER2-enriched and basal-like breast cancer subtypes²⁷. Our described 3D cell culture protocols may assist in developing cost-effective, rapid, and reproducible culturing methods that may be paired with formalin-fixed paraffin-embedded tissue preservation methods and subsequent spatial biology applications.

Protocol

1. Cell culture medium

NOTE: Prepare all media inside a biosafety cabinet.

1. To prepare cell culture medium for BJ-5ta, BT474, EA.hy926, and MDA-MB-468, supplement 500 mL of Dulbecco's Modified Eagle Medium (DMEM) high-glucose with 10% heat-inactivated fetal bovine serum (FBS) and 1% penicillin-streptomycin with a pipette. Incorporate 0.1% gentamicin with a micropipette.
2. To prepare cell culture medium for MCF10A, supplement 500 mL of DMEM/F12 with 5% horse serum and 1% penicillin-streptomycin with a pipette. Incorporate 0.1% gentamicin, 1 mL of 1 µg/mL hydrocortisone, 500 µL of 10 µg/mL insulin, 50 µL of 100 ng/mL cholera toxin, and 10 µL of 20 ng/mL epidermal growth factor with a micropipette.
3. To prepare cell culture medium for MCF10Ca1h, supplement 500 mL of DMEM/F12 with 5% horse serum and 1% penicillin-streptomycin with a pipette. Incorporate 0.1% gentamicin with a micropipette.
4. To prepare cell culture medium for THP-1, supplement 500 mL of RPMI 1640 with 10% FBS and 1% penicillin-streptomycin with a pipette. Incorporate 0.1% gentamicin with a micropipette.
5. Change the cell culture medium every 2-3 days for BJ-5ta, BT474, EA.hy926, MDA-MB-468, MCF10A, and MCF10Ca1h until the cells reach 70-80% confluency by assessing their growth daily with a brightfield microscope. Change the cell culture medium once a week for THP-1 cells.
6. Grow BJ-5ta, BT474, EA.hy926, MDA-MB-468, MCF10A, and MCF10Ca1h cells in 100 mm dishes with cell culture-treated surfaces in standard cell culture incubators with 5% CO₂.
7. Grow THP-1 cells in T75 Cell Culture Treated Flasks in standard cell culture incubators with 5% CO₂.

2. Cell collection

1. Turn on the UV light to sanitize the interior of the biosafety cabinet for 15 min.
2. Open the biosafety cabinet window sash to stabilize the airflow and turn on the vacuum aspiration system.
3. Clean the interior hood surface and the tubing of the vacuum aspiration system with 70% ethanol.
4. Prepare fresh serum-free cell culture medium inside the biosafety cabinet: supplement 500 mL of Dulbecco's modified Eagle's medium (DMEM) high-glucose with 1% penicillin/streptomycin using a pipette. Incorporate 0.1% Gentamicin with a micropipette.
5. Warm the cell culture medium, phosphate-buffered saline (PBS), and trypsin-EDTA (0.25%) to 37 °C by placing the items in a bead bath before starting the experiment.
6. Ensure that the cells are 70-80% confluent through visual inspection through a microscope.
7. Aspirate and discard the culture medium from the plated cells with the vacuum aspirator. Wash the remaining medium once with 2 mL of PBS with a pipette, then aspirate and discard the PBS with the vacuum aspirator.
   NOTE: THP-1 cells are grown in suspension; steps 2.7-2.9 can be omitted.
8. Add 1 mL of trypsin to the cell culture dish using a micropipette, and place the plate inside a 5% CO₂ incubator at 37 °C for 5 min.
9. Inactivate the trypsin by adding 1 mL of the soybean trypsin inhibitor (in 1x PBS) to the plate with a micropipette. Disperse cell clusters by pipetting the liquid mixture with a P1000 micropipette and collect the cell suspension from the bottom of the plate. Transfer the cell suspension with a micropipette to a 15 mL conical tube and centrifuge at 100 × g for 5 min at room temperature. Discard the supernatant with the vacuum aspirator.
10. Continue to cell counting.

3. Preparation of working dye solution and staining cells in suspension

1. Before opening the vial of cell tracker molecular fluorescent probes (see Table of Materials), allow the product to warm to room temperature for 15 min in a bead bath set at 37 °C.
2. Dissolve the lyophilized cell tracker blue dye (mass = 5 mg, molecular weight = 209.6 g/mol) to a final concentration of 10 mM with 2.385 mL of DMSO using a micropipette.
3. Dissolve the lyophilized cell tracker orange dye (mass = 50 µg, molecular weight = 550.4 g/mol) to a final concentration of 10 mM with 9.084 µL of DMSO using a micropipette.
4. Dissolve the lyophilized cell tracker deep red dye (mass = 15 µg, molecular weight = 698.3 g/mol) to a final concentration of 1 mM with 20 µL of DMSO using a micropipette.
5. Prepare the working cell tracker blue dye media solution (5 µM) by diluting 1 µL of the dye in 2 mL of serum-free DMEM medium with a micropipette.
6. Prepare the working cell tracker orange dye media solution (5 µM) by diluting 1 µL of the dye in 2 mL of serum-free DMEM medium with a micropipette.
7. Prepare the working cell tracker deep red dye media solution (1 µM) by diluting 2 µL of the dye in 2 mL of serum-free DMEM medium with a micropipette.
8. Resuspend the epithelial cells, BJ-5ta fibroblasts, and Ea.hy926 endothelial cells in the prepared working cell tracker blue, orange, and deep red dye media solutions (2 mL), respectively with a micropipette.
9. Incubate the tubes at 37 °C in the 5% CO₂ incubator for 30 min.
10. After 30 min of incubation, centrifuge the tubes at 100 x g for 5 min at room temperature.
11. Aspirate and discard the supernatant with the vacuum aspirator and resuspend the pellet thoroughly in 1 mL of 10% FBS serum-containing DMEM medium with a micropipette.
12. Continue to cell counting.

4. Cell counting

1. Collect 10 µL of the cell suspension and transfer it to a microtube with a micropipette.
2. Mix with 10 µL of trypan blue and pipette thoroughly.
3. Transfer 20 µL of the cell-trypan solution with a micropipette to a cell counting chamber slide. Insert and count the cells using an automated cell counter.
4. Calculate the average total live cell number from two readings.

5. Calculations

1. Prepare working cell stocks for each cell type at a concentration of 6.67 × 10³ cells/mL equivalent to 2,000 cells/300 µL.
   NOTE: The total stock volume will depend on the experimental sample size.
2. For mono-culture spheroids, make sure each independent sample consists of 300 µL of the epithelial cell working stock equivalent to 2,000 epithelial cells.
3. For co-culture spheroids (two cell types), ensure that each independent sample consists of 150 µL of epithelial cell working stock and 150 µL of stromal cell working stock. The sample will contain 1,000 epithelial cells and 1,000 stromal cells.
4. For co-culture spheroids (three cell types), make sure each independent sample consists of 150 µL of epithelial cell working stock, 75 µL of stromal cell #1 working stock, and 75 µL of stromal cell #2 working stock. The sample will contain 1,000 epithelial cells, 500 stromal cells of working stock #1, and 500 stromal cells of working stock #2.

6. Plating

1. Transfer the required volume with a micropipette for three technical replicates, plus one extra, into a microtube. Mix thoroughly with a pipette, then transfer 300 µL of the sample to a well of a U-shaped-bottom, ultralow attachment 96-well microplate.
2. Repeat the process for each additional technical replicate.
3. Place the 96-well plate in an incubator at 37 °C.

7. Brightfield imaging

1. Observe spheroid growth and morphology every 24 h up to 96 h using a microscope.
2. Image spheroids using a phase-contrast microscope.

8. Widefield imaging protocol setup for spheroids stained with tracker dyes and spheroids overlayed with basement membrane extract

1. Turn on the imaging and automated incubator devices listed in the Table of Materials, and create a new imaging protocol in the task manager of the imaging software. Under the Procedure tab, set the automated incubator temperature set point to 37 °C, and allow the incubator to equilibrate CO₂% and temperature before continuing with the next step.
   NOTE: An alternative fluorescent imager can be used if it has the appropriate filters for each of the dyes (blue, deep red, and orange) used.
2. Set the image settings to the following specifications: Magnification: 4X PL FL Phase, Field of View: 3185 X 3185 µm, Full WFOV.
3. Use the following specifications for the channels: DAPI: 377/447 nm, illumination = 10, integration time = 107 ms, gain = 10, RFP: 531/593 nm, illumination = 10, integration time = 137 ms, gain = 10, CY5: 628/685 nm, illumination = 10, integration time = 137 ms, gain = 10.
4. For imaging spheroids in a basement membrane extract solution overlay, use the following brightfield specifications: illumination = 10, integration time = 5 ms, and gain = 17.1.
5. Select the desired wells for imaging and approve the specification changes by clicking the select wells icon.
6. Navigate to the Data Reduction tab to adjust the cellular analysis settings.
7. Set the threshold value to 19,500 with a light background, and select Fill holes in masks.
8. For object selection, set the minimum object size to 100 µm and the maximum object size to 1,000 µm, and select Analyze entire image.
   NOTE: Only a primary mask and object count are needed for this analysis.
9. Save all changes and open the imager application.
10. Place the experimental U-shaped-bottom microplate in the incubator by opening the drawer, then close it using the imager software.
11. To run the protocol, click on the Procedure Info tab, add a user, and choose the protocol.
12. Ensure the correct plate type is selected and set the imaging time to 30 min per plate.
13. Select the desired imaging interval, indicate whether the plate has a lid, and adjust the imaging start time and duration.
14. Click Schedule Plate/Vessel to start the imaging process.

9. Basement membrane extract solution overlay (optional)

NOTE: Basement membrane extract solution can be applied to spheroids 24 h post plating.

1. Fill an ice bucket with ice to keep the basement membrane extract solution cold and store it at 4 °C when not in use.
2. Aspirate approximately 170 µL of medium with a multichannel pipette.
3. Use a magnifying glass and a mini lightbox to closely observe the small spheroids. Place the 96-well spheroid plate over the lightbox and position the magnifying glass overhead.
4. Set the P200 pipette to 30 µL and collect the basement membrane extract to create three microdroplets.
5. Add one microdroplet over each spheroid of the three technical replicates with a micropipette.
6. Ensure the 96-well plate is flat and position the pipette vertically above the spheroid. Release the droplet without touching the bottom of the well.
7. Place the plate in an incubator at 37 °C for 20 min.
8. Overlay the spheroids with an additional 50 µL of the basement membrane extract solution per well with a micropipette.
9. Reposition spheroids to the center of the well using a pipette tip.
10. Incubate at 37 °C for 30 min.
11. Add 100 µL of cell culture medium to each well.

10. Quantification

1. Download and open the ImageJ software. Upload the spheroid images.
2. Set measurements by clicking on Analyze | Set Measurements and select Area | Centroid. Measure the spheroids by using the lasso tool to trace the spheroid and measure the object.
3. Record the values for three biological replicates and calculate the average area and circularity.
4. Convert the area from pixels to square meters using the following formula:
   ​Area (m²) = ((550/504) × √ (area in pixels)) ²
5. Analyze data.
   1. Select Enter replicate values, stacked into columns.
   2. Input the data into the columns and select the monoculture spheroid control and the co-culture treatment spheroid samples.
   3. Perform a one-way ANOVA column analysis.
6. Use the following ANOVA parameters:
   1. Avoid matching or pairing.
   2. Assume a Gaussian distribution of residuals.
   3. Assume equal standard deviations.
   4. Use Tukey's multiple comparison test to compare the mean of each column with the mean of every other column.

11. Widefield immunofluorescent image processing

1. Save the images directly from the imager software as PNG or Tiff file format.
2. Select each individual channel or overlapping channels and save as PNG file.
3. Adjust brightness and contrast in the imager software or in ImageJ if images are exported as Tiff files.

Results

In this study, we developed a cell culture system to generate heteromulticellular 3D spheroids consisting of epithelial and stromal cells with organoid-like morphology. Spheroids were established by plating 2,000 epithelial cells in monoculture conditions. In co-culture conditions of two cell types, spheroids were established by plating 1,000 epithelial cells and 1,000 stromal cells. In co-culture conditions of three cell types, spheroids were established by plating 1,000 epithelial cells and two different stromal cell types of 500 cells of each cell type. Upstream of spheroid establishment, cells can be stained with fluorescent cell tracker dyes that allow for monitoring cellular spatial organization. After 24 h of initial spheroid formation, downstream applications include pharmacological perturbations, imaging, and sample collection. Time-lapse imaging is useful for assessing changes in spheroid behavior and morphology, including area and circularity (Figure 1A). At 24 h post plating, spheroids can be embedded in a scaffold environment, and time-lapse imaging can be used to assess the onset of invasive structures from the spheroid (Figure 1B). The collection of heteromulticellular spheroid samples has many applications, including genomic and proteomic profiling at the global and single-cell levels through experimental techniques such as RNA sequencing, single-cell RNA sequencing, proteomic sequencing, and cyclic immunofluorescence.

Figure 1: Schematic representation of the 3D cell culture process and potential applications. (A) Cell suspension of epithelial cells with non-epithelial stromal cells is pipetted into 3D ultralow attachment plates to form spheroids. Spheroids are imaged by brightfield microscopy every 24 h over 96 h. Cell lines can be stained with fluorescent cell tracker dyes for downstream widefield microscopy before inducing spheroid formation or can be perturbed pharmacologically after spheroid formation. Spheroid parameters such as morphology, area, circularity, and organization can be analyzed. (B) Spheroids of epithelial and epithelial cells with stromal cells are formed using the protocol from A. Spheroids were established by plating 2,000 epithelial cells in monoculture conditions. In co-culture conditions of two cell types, spheroids were established by plating 1,000 epithelial cells and 1,000 stromal cells. In co-culture conditions of three cell types, spheroids were established by plating 1,000 epithelial cells and two different stromal cell types of 500 cells of each cell type. After 24 h, a scaffold-like basement membrane extract solution is overlaid, and images are captured in brightfield microscopy every 24 h for 120 h. (C) Established scaffold-free and scaffold-based hetero-multicellular 3D cultures can be used for a variety of downstream applications such as cyclic immunofluorescence, single-cell RNA sequencing, and single-cell proteomics. Please click here to view a larger version of this figure.

MCF10A, MCF10Ca1h, and BT-474 monoculture spheroids maintain a compact spherical phenotype for up to 96 h post plating. When co-cultured with EA.hy926 microvascular endothelial cells, BJ-5ta fibroblasts, and/or THP-1 monocyte-like cells, the spheroids developed cellular protrusions at the periphery, which became more pronounced at 96 h (Figure 2A-C). Importantly, these protrusions, budding and compaction phenomena of spheroids represent variations of cellular organization of the cancer spheroid and the co-cultured stromal cells that correlated with 3D invasiveness following addition of a basement membrane-based hydrogel like Matrigel. The budding morphology ranged from solid to loose cell aggregates, resembling organoid morphology. In contrast, MDA-MB-468 monoculture spheroids appeared as large, loose cell aggregates. However, when MDA-MB-468 cells were co-cultured with EA.hy926, BJ-5ta, and/or THP-1, they formed compact spheroids (Figure 2D).

Figure 2: The range of architectures and morphologies adopted by combinations of non-epithelial stromal cells in combination when combined with epithelial tumor/non-tumor cells in scaffold-free 3D cultures. Representative brightfield images of (A) MCF10A, (B) MCF10Ca1h, (C) BT-474, (D) MDA-MB-468 spheroids in monoculture or in co-culture conditions with stromal BJ-5ta fibroblasts/Ea.hy926 microvascular endothelial cells or Ea.hy926/THP-1 monocyte-like cella or BJ-5ta/THP-1 cells across 96 h. Each spheroid was formed by plating 2,000 cells. Spheroids in monoculture conditions were formed using 2,000 epithelial cells. Spheroids in co-culture conditions were formed using 1,000 epithelial cells and two different stromal cell types of 500 cells. Formation of budding or aggregated organoid-like structures was initiated in spheroid co-culture conditions 24 h post plating in MCF10A, MCF10Ca1h, and BT-474 epithelial cells. Compaction of MDA-MB-468 cells was observed in spheroid co-culture conditions 25 h post plating. Scale bars = 100 µm. Abbreviation: hpp = hours post plating. Please click here to view a larger version of this figure.

At 72 h post plating, MCF10A, MCF10Ca1h, and BT-474 cells co-cultured with EA.hy926 and THP-1, or with BJ-5ta and THP-1, exhibited a significant increase in spheroid area compared to monoculture epithelial spheroids. MCF10Ca1h also showed a significant increase in the spheroid area when co-cultured with EA.hy926 and BJ-5ta. The onset of budding structures in co-cultured spheroids led to a significant decrease in spheroid circularity for MCF10A, MCF10Ca1h, and BT-474 co-cultured with EA.hy926 and THP-1, or with BJ-5ta and THP-1. Similar effects were observed for BT-474 co-cultured with both EA.hy926 and BJ-5ta (Figure 3A,B). In contrast, MDA-MB-468 cells co-cultured with EA.hy926 and BJ-5ta, EA.hy926 and THP-1, or BJ-5ta and THP-1, showed a significant decrease in spheroid area compared to monoculture MDA-MB-468 spheroids; yet, there was no effect on circularity (Figure 3A,B).

Figure 3: Area and circularity analysis graphs of mono-cultured and heteromulticellular stromal 3D scaffold-free cultures at 72 h post plating. (A) Average area (cm²) and (B) average circularity of MCF10A, MCF10Ca1h, BT-474, and MDA-MB-468 in mono-cultured and heteromulticellular stromal spheroid cultures 72 h post plating. Data reported are representative of at least three independent biological replicates and are reported as technical replicate averages ± SEM, unless otherwise indicated. *, **, *** or **** represent p values < 0.05, 0.01, 0.001, or 0.0001, respectively, unless otherwise noted. Please click here to view a larger version of this figure.

The application of cell tracker dye to BT-474 tumorigenic epithelial cells and stromal cells prior to spheroid establishment demonstrated that stromal cells, including EA.hy926 and BJ-5ta, formed the budding structures at the perimeter of the central BT474 spheroids (Figure 4, Supplemental Video S1, Supplemental Video S2, Supplemental Video S3, and Supplemental Video S4). At 48 h post plating, individual widefield fluorescence images of spheroids co-cultured with fibroblasts reveal that fibroblast spheroids co-localized together with endothelial cells but did not co-localize with BT-474 spheroids. A minority of endothelial cells were also found to co-localize with BT-474 spheroids in co-culture conditions. This suggests that the arrangement of stromal cells within the spheroid is correlated with an organoid-like morphology.

Figure 4: Wide-field fluorescence still images of differentially dyed stromal and BT-474 cells in heteromulticellular 3D cultures at 48 h post plating. BT-474 spheroids are stained with blue cell tracker fluorescent dye. BJ-5ta fibroblasts are stained with orange cell tracker dye, represented in red color. Ea.hy926 endothelial cells are stainedwith deep red cell tracker dye, represented in green color. Each spheroid was formed by plating 2,000 cells. Spheroids in monoculture conditions were formed using 2,000 epithelial cells. Spheroids in co-culture conditions (BT-474/BJ-5ta, BT-474/Ea.hy926) were formed using 1,000 epithelial cells and 1,000 stromal cells. Spheroids in double co-culture conditions (BT-474/BJ-5ta/Ea.hy926) were formed using 1,000 epithelial cells and 500 cells of each stromal cell type. Figures are representative of at least three biological replicates. Scale bar = 100 µm. Please click here to view a larger version of this figure.

To assess the biological relevance of our organoid spheroid model, spheroids were overlaid with basement membrane extract solution 24 h after plating. Monoculture BT-474 spheroids displayed no invasive properties 120 h post plating. However, BT-474 spheroids co-cultured with BJ-5ta or EA.hy926 developed structures at the periphery of the spheroid, which invaded the scaffold basement membrane extract solution environment. The number and length of these protrusions were significantly enhanced in BT-474 spheroids co-cultured with both BJ-5ta and EA.hy926 by 48 h through 120 h post plating (Figure 5, Supplemental Video S5, Supplemental Video S6, Supplemental Video S7, and Supplemental Video S8).

Figure 5: Brightfield still images of BT-474 cells in heteromulticellular 3D cultures with basement membrane extract solution overlay at 5 days post overlay. Each spheroid was formed by plating 2,000 cells. Basement membrane extract solution was overlaid over spheroids 24 h post plating. Spheroids in monoculture conditions were formed using 2,000 epithelial cells. Spheroids in co-culture conditions (BT-474/BJ-5ta and BT-474/Ea.hy926) were formed using 1,000 epithelial cells and 1,000 stromal cells. Spheroids in double co-culture conditions (BT-474/BJ-5ta/Ea.hy926) were formed using 1,000 epithelial cells and 500 cells of each stromal cell type. (A) Invasive structures protruding from the cancer spheroid embedded in basement membrane extract solution can be observed in co-culture conditions. (B) Qualification of invasive protrusion count at time 48 hpp. (C) Qualification of invasive protrusion length at time 48 hpp. Figures are representative of at least three biological replicates. Scale bar = 200 µm. Abbreviation: hpp = hours post plating. Please click here to view a larger version of this figure.

Supplemental Video S1: BT-474 monoculture spheroid with cell tracker blue dye. BT-474 (2,000) cells were resuspended in a U-shaped-bottom, ultralow attachment 96-well microplate to form spheroids. An overlap of brightfield and widefield fluorescence images was captured over 48 h. Abbreviation: ULA = ultralow attachment. Please click here to download this Video.

Supplemental Video S2: BT-474 spheroids co-cultured with BJ-5ta fibroblasts. BT-474 (1,000) and 1,000 BJ-5ta fibroblasts were resuspended in ULA plates to form spheroids. BT-474 cells were incubated with cell tracker blue dye and BJ-5ta fibroblasts with cell tracker orange dye. An overlap of brightfield and widefield fluorescence (blue = BT-474, red = BJ-5ta) images was captured over 48 h. Abbreviation: ULA = ultralow attachment. Please click here to download this Video.

Supplemental Video S3: BT-474 spheroids co-cultured with Ea.hy926 endothelial cells. BT-474 (1,000) and 1,000 Ea.hy926 endothelial cells were resuspended in ULA plates to form spheroids. BT-474 cells were incubated with cell tracker blue dye and Ea.hy926 endothelial cells with cell tracker deep red dye. An overlap of brightfield and widefield fluorescent (blue = BT-474, green = Ea.hy926) images was captured over 48 h. Abbreviation: ULA = ultralow attachment. Please click here to download this Video.

Supplemental Video S4: BT-474 spheroids co-cultured with BJ-5ta fibroblasts and Ea.hy926 endothelial cells. BT-474 (1,000), 500 BJ-5ta fibroblasts, and 500 Ea.hy926 endothelial cells were resuspended in ULA plates to form spheroids. BT-474 cells were incubated with cell tracker blue dye. BJ-5ta fibroblasts and Ea.hy926 endothelial cells were incubated with cell tracker orange dye and deep red dye, respectively. An overlap of brightfield and widefield fluorescent (blue = BT-474, red = BJ-5ta, green = Ea.hy926) images was captured over 48 h. Abbreviation: ULA = ultralow attachment. Please click here to download this Video.

Supplemental Video S5: BT-474 monoculture spheroid in basement membrane extract solution. BT-474 cells (2,000) were resuspended in ULA plates to form spheroids, and the spheroids were embedded in a basement membrane solution 24 h post plating. Brightfield images were taken over 60 h. Abbreviation: ULA = ultralow attachment. Please click here to download this Video.

Supplemental Video S6: BT-474 spheroids co-cultured with BJ-5ta fibroblasts in basement membrane extract solution. BT-474 (1,000) and 1,000 BJ-5ta fibroblasts were resuspended in ULA plates to form spheroids, and the spheroids were embedded in a basement membrane extract solution 24 h post plating. Brightfield images were taken over 60 h. Abbreviation: ULA = ultralow attachment. Please click here to download this Video.

Supplemental Video S7: BT-474 spheroids co-cultured with Ea.hy926 endothelial cells in basement membrane extract solution. BT-474 (1,000) and 1,000 Ea.hy926 endothelial cells were resuspended in ULA plates to form spheroids, and the spheroids were embedded in a basement membrane extract solution 24 h post plating. Brightfield images were taken over 60 h. Abbreviation: ULA = ultralow attachment. Please click here to download this Video.

Supplemental Video S8: BT-474 spheroids co-cultured with BJ-5ta fibroblasts and Ea.hy926 endothelial cells in basement membrane extract solution. BT-474 (1,000), 500 BJ-5ta fibroblasts, and 500 Ea.hy926 endothelial cells were resuspended in ULA plates to form spheroids, and the spheroids were embedded in a basement membrane extract solution 24 h post plating. Brightfield images were taken over 60 h. Please click here to download this Video.

Discussion

Our hetero-multicellular spheroid model demonstrates that epithelial-stromal cell interactions drive stromal cell budding in scaffold-free 3D culture conditions and the formation of invasive structures in scaffold-based 3D culture conditions. We observed consistent budding structure formation in both tumorigenic cell lines (MCF10Ca1h and BT-474) and the non-tumorigenic epithelial cell line MCF10A (Figure 2A-C). Interestingly, while spheroids from cell lines MCF10A, MCF10Ca1h, and BT-474, whether in monoculture or co-culture with fibroblasts and endothelial cells, formed circular and compact budding structures in scaffold-free environments, spheroids from cell line MDA-MB-468 showed less compaction (Figure 2D). This was unexpected, as MDA-MB-468 cells were anticipated to form circular spheroids, like the other tested epithelial cell lines. Cell line SUM-149, a hybrid epithelial-mesenchymal primary breast cancer cell line, was found to form compact spheroids like MCF10A, MCF10Ca1h, and BT-474 (data not shown)²⁸. Future research should investigate the signaling pathways mediating the formation of less compact spheroids in the MDA-MB-468 cell line since such anomalies may indicate cellular vulnerabilities that may be leveraged for the development of new therapeutic strategies.

Regarding the combination of different cell lines, THP-1 monocyte-like cells induce the formation of loose aggregate structures in all co-culture conditions of spheroids. Less compaction is observed in budding structures in co-culture conditions with THP-1/Ea.hy926 cells compared to those in co-culture conditions with THP-1/BJ-5ta cells (Figure 2A-D). Therefore, the compaction of spheroids is influenced by the stromal cell composition in the model^29,30.

Inconsistent plating of the hetero-cell combinations can result in incomplete spheroid development or variability in spheroid size, as compared to their technical replicates. Ensuring consistency in cell plating is crucial, as the spatial arrangement of cells that determines spheroid phenotype and development occurs primarily within the first 48 h³¹. During this critical period, cell ratios play a crucial role, with the proportion of stromal cells affecting both spheroid area and circularity.

Loose spheroid aggregates, which are dependent on cell type, can easily disassemble during manual handling. However, manual manipulation is necessary when overlaying spheroids with a scaffold-like basement membrane extract solution or when collecting whole spheroid samples. To mitigate this issue, it is advisable to work at a slower pace when collecting spheroids via pipette aspiration and to minimize manual handling whenever possible. During the basement membrane extract solution overlay, spheroids are prone to shifting within the well, often settling at the well's edge. A basement membrane extract solution-to-media ratio is important for maintaining the matrix properties of the 3D system; however, the spheroids cannot be fully enveloped due to the viscosity of the basement membrane extract solution, so manual intervention is required³². Non-centered spheroid positioning poses a challenge for imaging, as light refraction near the plastic edge of the well can interfere with image clarity.

The spheroids used in this study were composed of human cancer cell lines; the methods described here have not yet been applied to primary cell lines. Our results suggest that specific cell-cell interactions can drive the formation of organoid-like structures, indicating the potential applicability of this method to a wide range of tumorigenic cell lines. However, while our model attempts to mimic the complexity of the tumor microenvironment, the differing rates of cellular proliferation present a challenge for long-term studies³³. Rapidly proliferating cells tend to outgrow slower proliferating ones, making this model less suitable for extended experiments. In addition, different growth media type requirements for cell lines can introduce additional factors that may lead to spheroid phenotypes. Most cell lines in this study were grown in DMEM media. Studies with cell lines grown in different media may require different culture conditions testing a variety of media ratios; yet, this effect is observed in long-term experimentation³⁰.

Our model offers several advantages over other 3D cancer models. The spheroids we generate have a uniform and reproducible initial morphology and size, making it easier to compare treatment outcomes with controls. In contrast, some models using single cells suspended in media or scaffolds can result in variable cellular spatial distributions, leading to inconsistency. This model demonstrates that interactions between distinct cell populations can be effectively studied.

Traditionally, organoid and spheroid models have been composed of monoculture cancer cells or cancer cell co-cultures to investigate gradients of nutrients, hypoxia, or cellular arrangement³⁴. Additionally, this model can be studied in a scaffolded environment within 24 h post plating, providing an opportunity to explore early invasion events driven by diverse cell-cell interactions. While alternative 3D cancer approaches employ different scaffold matrix components to study changes in spheroid morphology and invasiveness, this model complements these by allowing the investigation of these changes in the context of both cell population dynamics and scaffold composition³⁵.

Most spheroids produced using our methods retain their structural integrity after manual handling, allowing for sample collection in downstream applications. Conventional genetic modifications such as CRISPR-Cas9, lentiviral shRNA transduction, and siRNA interference can be introduced upstream of spheroid formation. The complex behaviors observed in these spheroids suggest dynamic alterations in gene expression, which can be further investigated using RNA-seq. Advanced techniques such as CycIF, scRNA-seq, CosMx, and Visium can also be employed to study genomics and proteomics at spatial or single-cell levels. This spheroid model has the potential to mimic in vivo studies of targeted tumor therapeutics by capturing interactions between stromal cell buds and therapeutic agents before it reaches the cancer cells. This is significant because it can help determine whether stromal cells influence the drug's potency, potentially protecting the cancer cells from the therapeutic effects.

Materials

Name	Company	Catalog Number	Comments
15 mL Polypropylene Centrifuge Tubes	Falcon	14-959-53A
96-Well, Treated, U-Shaped-Bottom Microplate	Thermo Scientific	12566432
100 mm Dish Cell Culture Treated Surface	Thermo Scientific	130182
170 L CO2 Incubators	VWR	VWR51014991(10810902)
200 Self-Sealing Barrier Pipet Tips	Fisher Scientific	02-682-255
500 mL DMEM/Hams F-12, [+] L-glutamine	Corning	10-090-CV
1000 Self-Sealing Barrier Pipet Tips	Molecular Biology Products Inc.	2179HR
Animal-Free Recombinant Human EGF	PeproTech	AF-100-15
Automated Pipette Repeater E3x	Eppendorf	4987000410
Basement membrane solution	Huntsman Cancer Institute	https://healthcare.utah.edu/huntsmancancerinstitute/
Bead Bath	Lab Armor	74300706
Biorender	Biorender	Biorender.com
Biosafety Laminar Flow Cabinets	Thermo Scientific	300590389
BioSpa 8 Automated Incubator	Agilent	23082120
Cell Counter	Invitrogen	AMQAF2000
Cell Counting Chamber Slides	Thermo Scientific	C10283
CellTracker Deep Red	Thermo Fisher Scientific	C34565
Centrifuge	Thermo Scientific	13100675
Cholera Toxin B subunit	Sigma-Aldrich	C9903-1MG
Combitips advanced, Sterile, 5.0 mL	Fisher Scientific	13-683-714
Cytation 10 Reader Imager	Agilent	2105245
Defined Trypsin Inhibitor	Gibco	R007100	soybean trypsin inhibitor in 1x PBS
Donor Horse Serum, 500 mL, United States Origin	Corning	35-030-CV
Dulbecco's Modified Eagle's Medium	Corning	10-013-CV
Excel Sofware	Microsoft office 365	https://www.microsoft.com/en-us/microsoft-365/excel
Fetal Bovine Serum	Thermo Scientific	26140079
Fetal Bovine Serum, qualified, US origin	Thermo Scientific	26140079
Gentamicin Sulfate, Liquid	Corning	30-005-CR
GraphPad Prism
Hydrocortisone solution,50 µM, sterile-filtered, BioXtra	Sigma-Aldrich	H6909-10ML
Image J Software	Fiji	https://imagej.net/ij/
Insulin solution human	Sigma-Aldrich	I9278-5ML
Magnifying Glass	Fisher Scientific	01-182-392
Microcentrifuge Tube	Costar	3621
Microscope	Olympus	CKX53SF
Mini Lightbox	Fisher Scientific	361044708
Molecular Probes CellTracker Blue CMAC Dye	Fisher Scientific	C2110
Molecular Probes CellTracker Orange CMRA Dye	Fisher Scientific	C34551
Penicillin-Streptomycin Solution	Cytiva	SV30010
Penicillin-Streptomycin Solution	Cytiva	SV30010
Phase-Contrast Microscope	Echo	RVSF1000
Phosphate Buffered Saline	Fisher Scientific	SH3025601
Pipet Controller	Corning	4099
Pipette 100 – 1,000 µL	Eppendorf	3123000063
Pipette 20 – 200 µL	Eppendorf	3123000055
Prism	GraphPad Software	https://www.graphpad.com/features
RPMI 1640 with L-glutamine and 25 mM HEPES	Corning	10-041-CV
Serological Pipette, 10 mL	Thermo Scientific	170356N
Serological Pipette, 25 mL	Thermo Scientific	170357N
Serological Pipette, 5 mL	Thermo Scientific	170355N
Trypan Blue Solution, 0.4%	Corning	25900CI
Trypsin-EDTA, 0.25%	Corning	25053CI