The Role of the Tumor Microenvironment in CSC Enrichment and Chemoresistance: 3D Co-culture Methods

Abstract

Cancer stem-like cells (CSC) are responsible for tumor progression, chemoresistance, recurrence, and poor outcomes in many cancers, making them critical research and therapeutic targets. One of the critical components potentiating CSC chemoresistance is the interactions between CSC and the surrounding cells in the tumor microenvironment. Our lab has developed several 3D co-culture models to study ovarian CSC interactions with stromal or immune cells found in ovarian tumor microenvironments. In this chapter, we use ovarian cancer as a model to describe the methodologies developed in our lab; however, these techniques are applicable to a wide range of cancers. First, we discuss our method for isolating CSC from heterogeneous tumors and for creating 3D self-assembled tumoroids in hanging drop plates, in either monoculture or co-culture with mesenchymal stem cells or monocytes/macrophages. We then discuss methods for analyzing these models with a focus on isolating cell-type-specific changes and mechanism investigation. Specifically, we describe lentiviral transduction and flow cytometry as established and robust methods to identify and separate each cell type for downstream analysis. We then describe methods to examine CSC functionality with transwell migration assays and colorimetric MTS-based proliferation assays. Finally, we demonstrate enzyme-linked immunosorbent assays (ELISA ) and quantitative polymerase chain reaction (qPCR) methods as mechanistic investigation tools to decouple paracrine and juxtacrine interactions. These methods have wide-reaching applications in cancer research from basic biological investigations, to drug discovery, and personalized drug screening for precision medicine.

Keywords: Cancer stem-like cells; Chemoresistance; Co-cultures; Juxtacrine; Ovarian cancers; Paracrine; Three-dimensional; Tumor microenvironment.

Figure 1.. Flow cytometry gating scheme to sort for ALDH⁺CD133⁺ ovarian cancer stem cells (CSC).

Ovarian CSC can be identified via expression of various markers including CD44, CD117, CD24, Nestin, Nanog, and Oct3/4 as well as CD133 and high ALDH expression. Cells that express both CD133 and high ALDH are commonly used to isolate ovarian CSC as expression of both CD133 and high ALDH represents less differentiated versions of CSC expressing only CD133 or high ALDH. Isolation of CSC based on these markers using flow cytometry allows for downstream experiments that can isolate the interactions between CSC and other non-tumor cells in the tumor microenvironment, such as MSC and macrophages. A) The unstained control tube can be used to select for the cell population as well as the proportion of that population that is made up of single cells (unusually large cells, indicating the presence of two cells, will appear towards the right of the FSC1-H / FSC1-W plot). B). The DAPI control tube can be used to select for the single cells that are alive (live cells exclude DAPI from entering the cell). C) Non-specific staining of the APC-isotype on live cells can be set in the APC-isotype control tube. The APC signal is plotted on the y-axis against the ALDH signal on the x-axis to allow for evaluation of APC and ALDH signal simultaneously. A quadrant gate is used to plot so that cells expressing CD133 and high levels of ALDH can be identified. The non-specific staining for the APC-isotype control is set to ~0.5% by adjusting the position of the horizontal line along the y-axis. D) The same gate from the APC-isotype control is applied to the live single cells in the DEAB control tube and the vertical line is adjusted to allow 0.15% non-specific ALDH signal. E) The same gate is then applied to the live single cells in the ALDH⁺CD133⁺ sorting tube. Events in the top right quadrant are ALDH⁺CD133⁺ ovarian CSC. Events in the top left quadrant and the bottom right quadrant are ALDH⁻CD133⁺ and ALDH⁺CD133⁻ ovarian CSC respectively.

Figure 2.. Interactions between ovarian cancer stem cells (CSC) and stromal and immune cells can be investigated in 3D suspension hanging drop arrays.

Pro-tumoral stromal and immune cells that support CSC in the tumor microenvironments, such as mesenchymal stem cells (MSC) and monocytes/macrophages, can be easily incorporated on this 3D platform. Each cell type can be stably transduced with lentivirus to express distinct fluorescent proteins, such as green fluorescent protein (GFP) or dtTomato/mCherry, in order to track organization of each cell type in a co-culture and facilitate downstream analysis. A) Phase contrast images of 3D self-assembled tumoroids generated with CSC alone, CSC with human adipose-derived MSC, or CSC with monocytes. B) Fluorescent images of co-culture cellular interactions on the 3D hanging drop platform including a cancer cell only tumoroid, a CSC-MSC co-culture tumoroid, and a CSC-monocyte co-culture tumoroid. GFP labeled high grade serous ovarian cancer cell line OVCAR3, mTomato-expressing CSC, GFP expressing MSC, and mCherry expressing monocytes (U937), can be seen in the images. Scale bar = 200 μm.

Figure 3.. Fluorescence activated cell sorting (FACS) coupled with lentiviral transduction allows for analysis of cell type specific viability, and separation of multiple cell types to investigate enrichment of stemness in stromal and immune co-cultures.

GFP tagged CSC were co-cultured with MSC, and evaluated for elevated ALDH activity (a CSC marker) measured with AldeRed (Millipore Sigma) after 7 days in a hanging drop suspension. A) To analyze viability of the CSC population within the co-culture, the single cells are isolated based on gates made in the unstained control, and then GFP+ CSC are identified from the single cell population for evaluation of viability through DAPI exclusion in the DAPI control. B) To evaluate ALDH activity (AldeRed) in the CSC population from a CSC-MSC co-culture, the unstained control is first used to isolate single cells. Then the DAPI control is used to select the single cells that are alive (DAPI-), which are then used to select the GFP+ CSC population. Finally, the DEAB control is used to set the ALDH+ gate to allow 0.15 % non-specific signal, and the viable GFP+ CSC are analyzed for high ALDH activity. This scheme demonstrates how FACS and lentiviral transduction can be utilized to parse apart phenotypes of multiple cell types in the analysis of viability and stemness in co-cultures.

Figure 4.. 3D suspension hanging drop arrays are compatible with on-plate viability quantification methods, such as MTS assays, making them highly suitable for high throughput drug screening.

A) Phase contrast images of untreated CSC tumoroids, CSC-MSC co-culture tumoroids, and CSC-U937 monocyte co-culture tumoroids on day 7. B) Phase contrast images of the same experimental groups as above, that were treated with 200 μM carboplatin for 48 hours. C) MTS viability of mono-culture tumoroids and co-culture tumoroids treated with either 200 μM or 300 μM carboplatin, represented as percent viability normalized against untreated controls. Dark grey bars represent a CSC-MSC co-culture experiment, and light grey bars represents a CSC co-culture with M∅-like macrophages or M2-like alternately activated macrophages (M2-like AAM). Scale bar = 200 μM. ** p<0.01 – One-way ANOVA with Tukey’s multiple comparisons test. MTS data adapted from Raghavan 2019 (light grey) and Raghavan 2020 (dark grey).

Figure 5.. Functional changes, such as the increased migration capacity associated with epithelial-to-mesenchymal transition (EMT) can be investigated using FACS-sorted cells following suspension co-culture.

A) Representative FACS plots showing selection of GFP tagged CSC from CSC-MSC co-culture tumoroids for FACS collection. Sorted GFP+ CSC can then be plated on top of 8 μm porous transwell membranes and allowed to migrate along a chemotactic gradient. GFP+ cells that successfully migrate through the membrane can be imaged for quantification. B) Quantification of CSC migration following co-culture with MSC, M∅-like macrophages from PBMC or U937 monocytes, or with M2-like alternately activated macrophages from PBMC or U937 monocytes. Migration is shown as a percentage of the number of GFP+ CSC that migrated through the membrane following mono-culture in hanging drop suspension. ** p<0.01, *** p<0.001, **** p<0.0001 – One-way ANOVA with Tukey’s multiple comparisons test. Transwell data adapted from Raghavan 2019 (B: grey bars) and Raghavan 2020 (A: images; B: black bar).

Figure 6.. 3D hanging drop suspension co-cultures are highly amenable to mechanistic investigations using techniques such as ELISA to evaluate paracrine signaling and qPCR to quantify changes in transcriptome.

Enzyme linked immunosorbent assays (ELISA) can be performed to further probe which soluble factors may be causing functional changes in CSC from co-culture tumoroids. A) Quantification of IL-6 in the collected conditioned medium of monoculture tumoroids and CSC co-cultures with M∅-like macrophages or M2-like AAM showing increased IL-6 paracrine signaling in co-cultures. B) Since CSC-MSC co-culture produces more migratory CSC, quantitative polymerase chain reaction (qPCR) was performed to examine the fold change in EMT associated transcripts (TWIST, SNAIL, SLUG, ZEB1, and ZEB2) in CSC from CSC-MSC co-cultures compared to CSC mono-cultures. CSC-MSC co-cultures produced CSCs with 20–30X greater expression of EMT transcripts. Data adapted from Raghavan 2019 (A) and Raghavan 2020 (B).