The past decade has brought significant progress in cancer research, from a detailed understanding of the myriad genomic changes within an individual tumor to the development of novel immunotherapies.
These advances have contributed to the promise of precision medicine, with therapies tailored to match the molecular makeup of each patient’s disease. However, there are still significant challenges across all areas of the oncology pipeline, from drug discovery and target validation through to the identification of clinically relevant biomarkers for predicting and monitoring therapeutic response.
Genomics can provide deep insights into the genetic drivers of cancer, while transcriptomics sheds light on dysregulated gene regulation. But only the proteome — the ultimate output of the genome — provides deeper insight into what is going on at the molecular, tissue, and whole-body level.
In recent years, the field of proteomics has undergone a transformation, enabling robust and reproducible quantitative analysis of more than 10,000 proteins across thousands of samples. At the same time, advances in structural chemoproteomics techniques allow researchers to identify and map drug binding sites with unprecedented resolution.
Co-founded in 2008 by ETH Zurich proteomics pioneer Ruedi Aebersold, Biognosys is an innovator in the field of next-generation proteomics, developing a suite of technologies, products, and services based on unbiased quantitative mass spectrometry methods.
Building on this expertise, a rapidly growing body of evidence published in leading journals and presented at major international meetings has demonstrated that proteomics has the potential to transform all aspects of oncology — from drug development to trials and clinical decision-making.
Applying proteomics in oncology drug discovery
The majority of novel cancer therapeutics still fail at some point along the journey from bench to bedside, often due to lack of efficacy or unacceptable side effects. Target identification and validation, along with the identification of likely off-target effects, are therefore key challenges in oncology drug discovery.
Conventional phenotypic drug discovery approaches, such as large-scale chemical screens, can identify promising molecules but reveal relatively little about the underlying target or mechanism of action.
Coming from the other direction, therapies identified through rational target-driven drug discovery or functional genetic screens must be validated to ensure they work in the anticipated manner and to identify and minimize undesirable off-target effects.
A promising approach to solving these challenges is probing the interactions between compounds and proteins within the context of the whole proteome in an unbiased way.
Limited proteolysis combined with mass spectrometry (LiP-MS) is a novel protein structural analysis technique that differentiates drug-bound proteins from unbound proteins using quantitative mass spectrometry, revealing all the proteins bound by a compound within a complex cell or tissue lysate.
Importantly, it is one of the very few chemoproteomics techniques that can identify protein targets and ligand interactions without chemical modification or labelling of drugs, and can detect low-abundance targets in complex proteomes from mammalian cells.
Accurately identifying all the targets of a novel compound has significant utility in the drug discovery process, accelerating hit-to-lead time and flagging of potential harmful off-target effects early in the journey.
Furthermore, recently developed novel high-resolution methods (LiP-HR), presented at the 2020 ENA Symposium, enable accurate characterization and prediction of on- and off-target binding sites and drug-induced conformational changes with peptide-level resolution. This offers a powerful approach for revealing mechanism of action and target validation in the oncology drug discovery toolbox.
Additional power comes from combining LiP-MS with automated machine learning to identify compound-binding sites in complex proteomes, allowing unbiased peptide-level target deconvolution and ranking of potential drug candidates.
This approach has been validated by Biognosys and collaborators, who showed in a study published in Nature that they were able to identify and quantify novel drug targets with a broad range of affinities in complex cell lysates, including inhibitors of relevant oncological targets such as kinases, phosphatases, regulatory proteins, and membrane proteins. Additionally, they were able to identify novel off-target binding sites for the supposedly highly specific MEK1 inhibitor selumetinib.
Identifying proteomic signatures for clinical biomarker discovery
It has become abundantly clear in recent years that a “one-size-fits-all” approach for cancer therapy does not work, with individual patients responding differently to therapy depending on the underlying molecular makeup of their disease.
Clinically relevant biomarkers that can predict which patients are likely to respond to specific treatments increase the chances of clinical trial success. This is particularly true for highly promising immune checkpoint inhibitors, where there are still relatively few reliable biomarkers for predicting response.
However, biomarker discovery approaches based on genomics or transcriptomics alone fail to capture the full extent of what is happening within the body at the phenotypic level.
Large-scale, high-throughput discovery proteomics offers deep and unbiased quantification of tumor tissue or plasma proteomes, identifying signatures of response and novel biomarkers in an unbiased manner from a wide range of complex clinical samples including tissue, tumor, and biofluids such as urine and blood plasma.
This provides a better understanding of the tumor microenvironment and host immune response, helping to predict and optimize therapeutic response and progression-free survival.
In collaboration with researchers at Institut Curie, Biognosys used Hyper Reaction Monitoring mass spectrometry (HRM-MS) to analyze 125 samples of blood plasma from two cohorts of patients with non-small cell lung cancer (NSCLC) treated with anti-PD-1 therapy.
By identifying and quantifying hundreds of proteins across the samples, the team identified a robust proteomic signature associated with response to therapy and good progression-free survival outcomes. This includes known markers and therapeutic targets and molecules involved in cell adhesion, cell cycle, inflammation, immune response, and metabolic processes.
In another example, presented at the 2020 SITC congress, Biognosys and collaborators at INT-Pascale used unbiased discovery proteomics based on label-free data acquisition mass-spectrometry to analyze formalin fixed paraffin-embedded (FFPE) melanoma samples from treatment-naive patients treated with first-line anti-PD-1 immunotherapy.
The team identified a robust proteomic signature associated with response to anti-PD-1 immunotherapy, including a novel biomarker, ganglioside GM2 activator (GM2A), associated with poor response to the treatment. GM2A has previously been shown to perturb T-cell function, pointing to a potential target for drugs that enhance the efficacy of PD-1 blockade.
Embedding proteomics in oncology at scale
Today’s proteomics technologies offer deep and unbiased quantitative analysis of thousands of proteins across multiple samples, leveraging parallelization, automation, and data science to enable discovery at scale.
By harnessing insights from the proteome, cancer researchers can go beyond conventional genetic and biochemical techniques, accelerating drug discovery and development and providing deeper understanding of the molecular processes underpinning disease and therapeutic response.
In turn, this will translate into greater chances of navigating the journey from lab to clinic successfully, supporting the development and application of precision cancer therapy and ultimately leading to improved patient outcomes.