Despite advances in targeted therapies, more than 70% of breast cancer patients cancer relapse within five years of treatment. Lapatinib, a type of targeted treatment, has been a pillar of therapy, often used after patients stop responding to another treatment that selectively targets HER2-positive breast cancer, called trastuzumab. Yet, resistance to lapatinib inevitably develops, and patients face limited options. Previous studies of resistant mechanisms focused on abnormal DNA changes or aberrant gene expression patterns, but cancers are not driven by changes to our genetic code alone.
The genome, epigenome, and proteome interact in dynamic and sometimes unpredictable ways, and we wanted to see what insights could be revealed by studying all three layers together. We chose a model system of HER2-positive breast cancer cells (SKBR3 cells) and their lapatinib-resistant counterparts (SKBR3-L). Our idea was simple but ambitious: use an integrative “multi-omics” strategy combining three state-of-the-art techniques that can interrogate cancer biology at multiple layers at high resolution. We used chromatin accessibility profiling (which investigates how DNA “opens” and “closes”), gene expression changes, and protein abundance (mass spectrometry–based proteomics) to capture the full picture of how resistance develops. This approach allowed us to go beyond the usual single-layer studies; by overlaying three molecular maps, we could identify the most robust and consistent changes, the ones most likely to be true drivers of resistance.
Conventional understanding suggests that aggressive cancer cells become more dangerous by opening their chromatin, making cancer causing genes accessible for transcription. To our surprise, resistant cells showed the opposite: a global reduction in chromatin (DNA) accessibility. This finding initially puzzled us – how could cells become more invasive and drug-resistant if they were closing off large parts of their genome? Nonetheless, the answer became clear when we delved deeper into each dataset and each gene. While overall chromatin accessibility decreased, specific key regions near transcription start sites remained highly open. These regions corresponded to genes that were consistently upregulated across all three datasets – a resistance “signature” that stood out from the background noise.
From the overlap of our three datasets, we identified nine markers consistently altered in resistant cells. EGFR and SCIN were already known players in breast cancer resistance, which reassured us that our approach was capturing meaningful results; but more exciting were the seven novel candidates, including HPGD, FASN, TPM1, CALD1, PCP4, AKR7A3, and KRT81. Many of these genes had not been previously linked to HER2-positive breast cancer or lapatinib resistance. Their functions highlighted processes such as actin remodelling, metabolic reprogramming, and stress response, all of which could help cancer cells adapt to treatment. When we validated two of these markers (FASN and HPGD) in an independent lung cancer cell model, we found that they were also upregulated during acquired lapatinib resistance. This suggests that the resistance signature we uncovered might be shared across different cancers, not only in breast cancer.
Looking at the behaviour of resistant cells reinforced what the multi-omics data suggested. Compared to their drug-sensitive counterparts, resistant cells were more irregular in shape, less spherical, and showed invasive protrusions when grown in 3D culture. In soft agar assays, they formed larger and more aggressive colonies, hallmarks of transformation as the resistant cells survived and adapted better to new and different microenvironment than their usual growth habits. These morphological changes aligned with the enrichment of pathways we identified in the omics data, particularly those related to epithelial–mesenchymal transition (EMT), KRAS/MAPK signalling, and DNA repair. In other words, resistant cells were rewiring themselves at multiple levels, trading off some gene expression flexibility for a specialised, invasive survival mode.
Our findings highlight the importance of looking beyond mutations when studying drug resistance. Epigenetic changes, such as chromatin alterations, can reprogram cancer cells in ways that make them harder to treat. The nine-marker resistance signature we identified could provide the scientific community further research incentives for developing new biomarkers to predict which patients are at risk of relapse, and if confirmed, designing targeted therapies that interrupt the specific pathways resistant cells depend on, applying integrative multi-omics approaches more broadly across different cancers and treatments. Ultimately, the hope is that resistance will no longer be an inevitable outcome, but rather a predictable process that can be prevented.