Treacherous Apoptosis- Cancer Cells Sacrifice Themselves at the Altar of Heterogeneity

Hepatocellular carcinoma (HCC) displays striking levels of heterogeneity at inter-patient, inter-tumor and intratumoral levels (1). This high level of heterogeneity impedes biomarker development for patient stratification, and also limits therapeutic efficacy of targeted therapies or immune checkpoint inhibitors (2). Thus, it remains critical to unravel the mechanisms promoting molecular heterogeneity in order to develop more effective biomarkers and therapeutics for HCC. Our understanding of tumor heterogeneity is relatively limited partly due to the low resolution of bulk tumor sequencing studies where the molecular signals from the various cancer cell subclones, immune cells and stromal cells are averaged. However, the evolution of technologies which enable the study of cancer at a single cell resolution is revolutionizing our understanding of cancer heterogeneity (3). Exciting new spatial biology technologies using multiplexed fluorescence labeling to study DNA, RNA or protein expression of single cells in their native spatial context are facilitating a more comprehensive understanding of intratumoral heterogeneity. The optimization of some of these spatial biology technologies, like in situ hybridization (ISH), for formalin fixed paraffin embedded (FFPE) tissues has the potential to enable their rapid adoption to existing clinical workflows.

In this current issue, Khatib et al use multiplexed single-cell RNA in situ hybridization (RNAScope) to identify drivers of HCC heterogeneity (4). Using a computational approach to filter through previously published single-cell RNAseq (scRNAseq) data of liver cancer samples, authors identify two genes, cyclin dependent kinase inhibitor 3 (CDKN3) and and protein regulator of cytokinesis 1 (PRC1), to be surrogate markers to assess intratumor heterogeneity. Caspase 3 (CASP3) expression was used as a marker of apoptotic cell death. Multiplex RNA-ISH was used to measure the cellular expression of these three genes in 254 HCC samples present on a tissue microarray. This enabled the authors to demonstrate that increased apoptotic cell death was associated with increased tumor heterogeneity. This finding is in line with previous work from the same group which showed that increased cellular apoptosis correlated with increased tumor diversity in scRNAseq data (5). The strength of the current study is the power of spatial context. Since RNA-ISH technologies preserve spatial architecture at a single-cell level, authors were able to demonstrate that tumor diversity was highest in the close proximity of apoptotic cell islands than further away. Lastly, authors generated an inducible CRISPR/Cas9-based system to conditionally knockout anti-apoptosis gene MCL1 and induce varying levels of cell death. Using this controlled system authors validated findings from RNA-ISH and demonstrated that higher levels of cell death led to increased therapy resistance, reflecting the therapeutic implications of cellular heterogeneity.

Programmed cell death or apoptosis has generally been viewed as a protective mechanism which suppresses tumor growth and prevents tumor progression (6). However, recent studies have shown that apoptosis can also have a paradoxical pro-tumoral role by increasing genomic instability, creating an immunosuppressive microenvironment, or driving therapy resistance (7). These oncogenic functions are mediated by dysfunctional mitochondria and caspases and provide an explanation for the observation that tumors with high apoptotic index tend to have poor prognosis (8,9). The current study identifies a new putative pro-tumorigenic function of apoptosis. The study demonstrates that islands of apoptotic cell death are strongly associated with tumor heterogeneity in their spatial proximity in vivo and induce therapy resistance in vitro. These findings raise the possibility that apoptotic pathways may be hijacked by cancer cells in a coordinated fashion to render a survival advantage. More importantly, the study identifies apoptosis as a novel factor driving tumor heterogeneity, apart from known factors like tumor mutational burden, genomic instability or defective DNA repair. These findings have clear therapeutic implications and suggest that inducing apoptotic cell death as a therapeutic strategy could have a conflicting deleterious effect in certain contexts. The findings of this study may also provide an explanation for the observed limited efficacy of apoptosis inducers in clinical trials for cancer (10).

Lastly, the study highlights how knowledge of spatial cellular context at a single-cell level can empower us to gain innovative molecular insights with clear therapeutic and prognostic implications. Eliciting transcriptomic data at a single-cell level while preserving spatial architectural elements will be key to deciphering the complex molecular landscape of therapy-resistance in HCC. Thus, further studies harnessing the power of spatially resolved transcriptomic and proteomic profiling carry the potential to finally enable us to comprehensively decipher molecular heterogeneity in HCC.

Figure 1: Apoptotic Cell Death as a Driver of Tumor Heterogeneity.

A. The panel on the left demonstrates the increased diversity of cancer cell subclones in the close proximity of apoptotic cell death islands than farther away. B. The panel on the right demonstrates how this heterogeneity associated with apoptosis can potentially lead to emergence of resistant cancer cell subclones under therapeutic selective pressure.