Cancer Susceptibility to Stapled Oncolytic Peptides is Dictated by Membrane Cholesterol and Inflammatory Signaling

Relapsed and refractory cancers effectively overcome diverse modalities of cancer treatment, whose principal targets are nucleic acids and proteins. The plasma membranes of cancer cells represent an alternative and underutilized target, with the potential for membrane lysis to induce a rapid, pro-inflammatory cell death that circumvents the challenges of intratumor heterogeneity and immune evasion. Here, we repurposed StAMP51.2, a stapled Magainin II peptide previously optimized for selective membrane lysis of gram-negative bacteria, to target cancer cell membranes. Using PRISM, a high throughput cancer cytotoxicity screen, we identified cancer cells most vulnerable to StAMP51.2 and biomarkers of susceptibility, specifically reduced cholesteryl esters and elevated triacylglycerols. We validated this signature in a pair of sensitive (OCI-AML3) and resistant (K562) leukemia cell lines, correlating their differential responses to distinct lipidomic profiles. Susceptibility of OCI-AML3 cells in culture extended to the in vivo context, where StAMP51.2 suppressed leukemic growth in orthotopic and intraperitoneal models. To further characterize the mechanism of action, we undertook the challenge of generating StAMP51.2-resistant OCI-AML3 cells, which required four months of low-level exposure. Strikingly, drug-resistant OCI-AML3 cells recapitulated the lipidomic phenotype of naturally-resistant K562 cells. Transcriptomic analyses further revealed that lipid reprogramming was accompanied by pervasive downregulation of inflammatory signaling. Thus, in advancing StAMP51.2 as an oncolytic prototype, we uncovered an immune regulatory axis that links membrane integrity to inflammatory signaling. Overall design: 3x10^6 OCI-AML3 sensitive and resistant cells, as well as K562 cells, were harvested by centrifugation and washed once with 1x PBS. Supernatants were discarded, and total RNA was isolated using the Qiagen RNeasy Mini Kit, following the manufacturer's protocol. Purified RNA samples were sent to the Molecular Biology Core Facility (MBCF) at Dana-Farber Cancer Institute for library preparation and sequencing. Libraries were generated from 200 ng of total RNA using the Roche Kapa mRNA HyperPrep strand-specific kit according to the manufacturer's guidelines, automated on a Beckman Coulter Biomek i7. Double-stranded DNA libraries were quantified using a Qubit fluorometer and Agilent TapeStation 4200. Dual-indexed libraries were pooled in equimolar amounts and initially shallow-sequenced on an Illumina MiSeq to assess library concentration and pooling accuracy. The final library pool was sequenced on an Illumina NovaSeq6000 with a target of 40 million 100 bp paired-end reads per sample. Sequencing reads were aligned to the hg38 human genome, and gene expression counts were obtained. Differential expression analysis was conducted using DESeq2 v1.22.1 (Love et al., 2014). The entire RNA-seq data analysis workflow was executed using the VIPER Snakemake pipeline (Cornwell et al., 2018).