Pheno-seq – linking visual features and gene expression in 3D cell culture systems

Stephan M. Tirier, Jeongbin Park, Friedrich Preußer, Lisa Amrhein, Zuguang Gu, Simon Steiger, Jan‐Philipp Mallm, Teresa G. Krieger, Marcel Waschow, Björn Eismann, Marta Gut, Karsten Rippe, Matthias Schlesner, Fabian J. Theis, Christiane Fuchs, Claudia R. Ball, Hanno Glimm, Roland Eils, Christian ConradScientific 2019 · DOI

Abstract

Patient-derived 3D cell culture systems are currently advancing cancer research since they potentiate the molecular analysis of tissue-like properties and drug response under well-defined conditions. However, our understanding of the relationship between the heterogeneity of morphological phenotypes and the underlying transcriptome is still limited. To address this issue, we here introduce “pheno-seq” to directly link visual features of 3D cell culture systems with profiling their transcriptome. As prototypic applications breast and colorectal cancer (CRC) spheroids were analyzed by pheno-seq. We identified characteristic gene expression signatures of epithelial-to-mesenchymal transition that are associated with invasive growth behavior of clonal breast cancer spheroids. Furthermore, we linked long-term proliferative capacity in a patient-derived model of CRC to a lowly abundant PROX1-positive cancer stem cell subtype. We anticipate that the ability to integrate transcriptome analysis and morphological patho-phenotypes of cancer cells will provide novel insight on the molecular origins of intratumor heterogeneity.

Introduction

Cancer is notoriously heterogeneous; even cells from the same tumor can exhibit vastly different behaviors and treatment responses. To study this complexity, researchers increasingly rely on 3D culture systems, such as spheroids and organoids, which better mimic human tissue than traditional 2D cultures. However, existing analytical methods often force a trade-off: single-cell RNA-seq (scRNA-seq) provides high-resolution transcriptomics but requires dissociating the 3D structure, losing all morphological context. Conversely, spatial transcriptomics can preserve structure but is difficult to scale.

To bridge this gap, researchers developed Pheno-seq, a high-throughput workflow that directly couples automated imaging of intact 3D cell cultures with transcriptomic profiling. This method allows for the simultaneous analysis of a spheroid’s physical "face" (morphology) and its molecular "instruction manual" (transcriptome).

Methodology

The Pheno-seq pipeline integrates automated handling, high-resolution imaging, and advanced transcriptomics:

  • Automated Handling: The team adapted the Wafergen iCELL8 system to dispense intact 3D spheroids (up to 150 µm) into barcoded nanowells.
  • High-Resolution Imaging: Before lysis, spheroids are imaged via automated confocal microscopy to quantify features like circularity, volume, and cell count.
  • Transcriptomics: The system performs 3’-end RNA-seq on the imaged samples.
  • Computational Deconvolution: To address the multi-cellular nature of spheroids, the authors used a maximum likelihood inference algorithm (stochprofML) to mathematically "unmix" the data and infer single-cell regulatory states from the multi-cell profiles.

Key Insights: Linking Morphology to Molecular Drivers

A primary finding of the study was that visual morphology serves as a reliable proxy for a cell's molecular state. In breast cancer models (MCF10CA), "aberrant" or non-round shapes were directly linked to specific gene programs.

Gene set enrichment analysis (GSEA) identifying EMT and signaling pathways associated with aberrant, invasive spheroid phenotypes.

As shown in the enrichment analysis, these aberrant phenotypes are driven by Epithelial-to-Mesenchymal Transition (EMT) signatures, as well as TNFA and KRAS signaling. This confirms that visual "invasiveness" in 3D culture is the physical manifestation of a specific mesenchymal gene program.

In colorectal cancer (CRC) models, the team discovered that spheroid size was a key indicator of stemness. Larger spheroids possessed a distinct "stem/progenitor" signature, while smaller ones were enriched for differentiation markers like TFF3, suggesting they represent "transit-amplifying" cells with limited growth potential. Through deconvolution, they identified a rare PROX1-positive cancer stem cell subtype responsible for long-term growth.

Key Results and Validation

The Pheno-seq workflow demonstrated high throughput and sensitivity, successfully profiling 210 breast cancer and 95 CRC spheroids. A critical validation of the method was the alignment between transcriptomic clusters and physical features.

t-SNE visualization of spheroid transcriptomes colored by circularity, demonstrating the link between morphology and gene expression.

As illustrated in the t-SNE plot, spheroids naturally clustered based on their visual circularity. This statistical separation proves that the transcriptome is intrinsically tied to the morphological phenotype. Furthermore, Pheno-seq achieved higher gene detection rates than standard scRNA-seq because it utilizes the higher RNA input of a whole spheroid, avoiding the "dissociation bias" often caused by breaking down 3D structures.

Limitations and Future Directions

While powerful, the current Pheno-seq hardware is limited to spheroids smaller than 150 µm. Future iterations may utilize different microfluidic systems to accommodate larger organoids. Additionally, the authors suggest expanding the workflow to include time-lapse microscopy to track spheroid evolution over time or integrating other "omics" like ATAC-seq to study chromatin accessibility. This evolution would move cancer research from static snapshots to a dynamic understanding of tumor progression.

Summary updated

Abstract

Abstract Patient-derived 3D cell culture systems are currently advancing cancer research since they potentiate the molecular analysis of tissue-like properties and drug response under well-defined conditions. However, our understanding of the relationship between the heterogeneity of morphological phenotypes and the underlying transcriptome is still limited. To address this issue, we here introduce “pheno-seq” to directly link visual features of 3D cell culture systems with profiling their transcriptome. As prototypic applications breast and colorectal cancer (CRC) spheroids were analyzed by pheno-seq. We identified characteristic gene expression signatures of epithelial-to-mesenchymal transition that are associated with invasive growth behavior of clonal breast cancer spheroids. Furthermore, we linked long-term proliferative capacity in a patient-derived model of CRC to a lowly abundant PROX1-positive cancer stem cell subtype. We anticipate that the ability to integrate transcriptome analysis and morphological patho-phenotypes of cancer cells will provide novel insight on the molecular origins of intratumor heterogeneity.
BibTeX
@inproceedings{tirier2019pheno,
  author = {Stephan M. Tirier and Jeongbin Park and Friedrich Preußer and Lisa Amrhein and Zuguang Gu and Simon Steiger and Jan‐Philipp Mallm and Teresa G. Krieger and Marcel Waschow and Björn Eismann and Marta Gut and Marta Gut and Karsten Rippe and Matthias Schlesner and Fabian J. Theis and Christiane Fuchs and Claudia R. Ball and Hanno Glimm and Roland Eils and Christian Conrad},
  title = {{Pheno-seq – linking visual features and gene expression in 3D cell culture systems}},
  booktitle = {Scientific},
  year = {2019},
  doi = {10.1038/s41598-019-48771-4},
}
Figure