Multiplexed Immunofluorescence Imaging for Characterizing Compositional Heterogeneity in Tissue-Engineered 3D in vitro Breast Cancer Tumor Aggregates

Introduction:: Intra-tumoral heterogeneity (ITH) in breast cancer, or heterogeneity in both biomarker composition and spatial distribution of cancer cells, has been shown to correlate with poor prognosis and reduced therapy response.¹ This drives a demand for next generation in vitro tumor models able to incorporate such compositional and spatial heterogeneities into their design and fabrication. Our lab strives to manufacture such “histology-grade” 3D tumor systems with requisite cellular and spatial resolution, utilizing a laser direct-write (LDW) bioprinting approach.² To validate this workflow, we seek to establish an imaging approach to identify the different cell types in heterogeneous tissue-engineered dense 3D tumor aggregates, via their unique biomarker expressions. Herein, we present a novel method to robustly characterize the cellular composition and spatial distribution of tissue-engineered 3D in vitro tumor models using Cell DIVE multiplexed immunofluorescence (MxIF) imaging.³


Materials and Methods:: Homogeneous and heterogeneous 3D tumor aggregates, comprised of GFP-labeled MDA-MB-231 (triple negative) and ZsGreen-labeled AU565 (HER2+) breast cancer cells, were fabricated in core-shelled microcapsules created via electrostatic bead generation,² utilizing a 2% alginate bioink (Sigma, #9005-38-3) suspending either one of the breast cancer cell types, or a co-culture of both. To create microcapsules, a controlled electric field was applied to the alginate bioink as it was extruded through a needle via syringe pump into a 2% CaCl₂ bath, resulting in a charged stream of small droplets that rapidly crosslink in the oppositely-charged solution (CaCl₂) to form cell-loaded calcium alginate hydrogel microbeads. The printed microbeads were then coated with chitosan (Spectrum MFG, C1569) which complexes to the alginate to form a shell, then the alginate core was chelated with sodium citrate to produce core-shelled microcapsules within which the cells self-assemble to form dense 3D tumor aggregates. Upon maturation, the tumor aggregates underwent MxIF imaging to generate single-cell intensity data for their tagged fluorophores and several biomarkers related to the HER2 pathway.


Results, Conclusions, and Discussions:: Results

Single-cell analysis of the MxIF data of biofabricated 3D tumor aggregates highlights the potential of this approach by displaying correlations between different biomarkers (Fig. 1). For example, ZsGreen correlates with HER2, but not with vimentin, illustrating that vimentin is expressed primarily in MDA-MB-231 cells, and not in AU565 HER2+ cells. Use of spatial single-cell analysis algorithms provides quantitative measurements of multiplexed biomarkers and spatial location of individual cells in biofabricated tumor constructs. Hence, rather than a single biomarker, cell types can be distinguished by their unique expression of multiple biomarkers, such as the nine biomarkers explored herein (Fig. 1).



Discussion

Unlike standard immunohistochemistry methods, which are limited to 1-5 markers in a single section, Cell DIVE MxIF is a cyclical staining, imaging, and signal removal workflow that can provide single-cell data on more than 60 proteins in a sample, including cell spatial coordinates, thereby allowing analysis of co-expressed biomarkers and relationships between cell types. Our future work in fabricating “histology-grade” living 3D in vitro tumor constructs will leverage this technique to first characterize patient tumor sections, thresholding and discretizing biomarker expression to inform a CAD-design reflecting composition and spatial distribution. This design would be reproduced in a living construct via LDW bioprinting, using cell lines matching the unique biomarker expression combinations. The resulting construct would then be validated for print and biologic fidelity using the workflow for which we lay the foundation herein. Fabrication of reproducible in vitro models able to incorporate compositional and spatial heterogeneities is paramount for exploring the role of ITH on early drug response in breast cancer.



Conclusions

Utilizing MxIF for characterization, we were able to successfully demonstrate fabrication of a dense 3D tumor aggregate comprised of two distinct breast cancer cell lines, as well as the ability to discern the two cell types via their biomarker expression signature. Our future studies will utilize LDW bioprinting to create heterogeneous tumor constructs, and leverage Cell DIVE MxIF imaging to assess agreement with CAD input design (“print fidelity”), and biomarker spatial distribution (“biologic fidelity”).


Acknowledgements (Optional): : Supported by NIH R01 BRG CA207725 (DTC/MB) and NIH R01 CA233188 (MB).

References (Optional): : 1. Caswell-Jin JL et al. (2020) Annu Rev Cancer Biol. 2020;5:79–94.
2. Kingsley DM et al. (2013) Biofab. 5(4): 045006.
3. Gerdes MJ et al. (2013) Proc Natl Acad Sci 110:11982–7.