23rd Human Proteome Organization World Congress

Biophysical proteomics

Keynote speaker

Keynote lecture

Chemical proteomic platforms for diversifying drug mechanism of action

Assistant Professor

Characterizing the target space of molecular glues and bifunctional degraders in cells and animal models by bottom-up proteomics

Chemical Biology Insights

Targeted protein degradation facilitates the development of selective therapies against protein targets that were previously deemed &quot;undruggable&quot; by conventional inhibitors, for instance because they lack a distinctive active site. Characterizing the biological activity of degrader compounds, including proteolysis-targeting chimeras (PROTACs®) and molecular glues, is both easy and hard: While degrader targets can be readily identified by global quantitative proteomics, the effects can be highly tissue-specific, as they depend not only on the main target but also on the simultaneous expression, activity, and localization of a ubiquitin ligase complex. Therefore, it is desirable to validate target degradation and search for potential off-targets across many different biological models ranging from cell culture and primary cells to tissues from animal models. Such experiments with large cohorts and often limited sample amounts require high throughput analytics with good sensitivity.Here, we present a streamlined and largely automated bottom-up proteomics workflow quantifying up to 9,000 proteins at a throughput of 30 samples per day. As a proof-of-concept, we confirmed the (off-) targets of the clinical compounds Indisulam, Lenalidomide, and Pomalidomide in a human T-cell model. To narrow down the main targets and rule out secondary effects, we performed further experiments in dose-response, with a time course, and with pharmacological co-treatments to selectively inhibit proteasomal degradation or Cullin-RING ligases.To demonstrate applicability for more complex biological models, we studied a bifunctional degrader in murine liver, spleen, and adipose tissue and in human peripheral blood mononuclear cells (PBMCs). Our data successfully confirmed target engagement in all four matrices including the sample-limited PBMCs and was key in the discovery of a previously unknown off-target. Furthermore, we characterized a PROTAC® in a patient-derived xenograft (PDX) model at four time points, each at four different doses. After 48 h of MS analysis time, we obtained global proteome data with time- and dose-resolution that helped us to validate target degradation and identify potential off-targets.Taken together, we established a global degrader target identification method that can be universally applied to a range of relevant disease models and offers relatively high throughput und sensitivity. We anticipate that our method will accelerate drug discovery by providing a detailed picture of the target landscapes of degrader modalities, regardless whether they come from phenotypic screens or from rational drug design.

Oral presentation

Professor

Combining chemoproteomics and phenotypic screening for identifying STING modulators

The cGAS–STING pathway is a crucial component of the innate immune system that functions to detect the presence of cytosolic DNA and, in response, trigger the expression of inflammatory genes, therefore playing a critical role in autoimmune diseases, inflammation, and cancer. Yet, there are no approved STING inhibitors in the clinic. By high-throughput phenotypic screening of thousands of small covalent molecule fragments, we were able to identify &gt;30 compounds that inhibit STING pathway activation. To understand the compound&quot;s mechanism we investigated the cysteines to which the compounds bind using activity-based protein profiling (ABPP), followed by an in-depth quantitative whole proteome and phospho-proteome analyses to identify the proteome-wide changes by the compounds. Although different compounds demonstrated varying levels of STING modulation, by the profiling of the cysteines to which the compounds bind, we identified that many of them target STING cysteine 91 (C91), resulting in inhibition of STING palmitoylation. The phospho-proteome and whole proteome analyses complemented this observation and validated the changes resulting from STING inhibition.In addition to identifying the binding site of the compounds, the cysteine profiling approach revealed changes in the reactivity of other cysteines. We observed changes in the reactivity of STING C148 and C206, suggesting higher availability of these positions to the desthiobiotin iodoacetamide probe due to STING conformational changes. As well as changes in the availability of cysteines in STING interactors, for example, IRF3 C371 which resides in the proximity of STING-IRF3 interaction. These examples demonstrate the utility of this approach for the understanding of the global changes in protein interactions and conformation by the compounds.Identifying structurally variable compounds in the screen directs us to improve the specificity of the compounds to STING. In addition, ABPP results enable us to identify off target proteins to which the compounds bind. Together, this information led us to design a library of elaborated compounds and tested their binding specificity by ABPP and their activity using the functional assay, and identifying compounds with improved specificity and binding to STING.The combination of high-throughput phenotypic screening and proteomic approaches facilitates the identification of functionally relevant inhibitors and their mechanisms.

Prof

Rationale prediction of drug combinations based on large-scale perturbation proteomics

The pharmaceutical industry is hindered by high failure rates, steep expenses, and sluggish drug development. Drug repositioning can alleviate these challenges by slashing costs, expediting development, and unveiling new targets for existing drugs, but standalone efficacy is often limited by low plasma concentrations. Drug combinations may enhance repositioning success, but the vast number of possible pairings is a significant obstacle. Current computational methods, which rely heavily on genomic and transcriptomic data, fall short in systematically identifying successful drug repositioning combinations. We propose integrating a drug repurposing library and post-perturbation proteomic data to enhance the success rate of drug combination repositioning. In our study, we focused on three cell lines, namely NCI-H1299, HCT116, and 786-O. To identify effective drug combinations, we implemented a 2x7 concentration matrix approach, using 25 cancer-specific anticancer drugs as 'anchor' compounds, each at two optimized concentrations. Leveraging the molecular representation learning model, Uni-Mol, we carried out unsupervised clustering on 3000 repurposed drugs, and picked 192 library drugs, each set at seven concentration points across a 1000-fold range. We conducted perturbed proteomic analysis of 3000 drugs, and 300 each of effective and ineffective drug combinations on these cell lines at two time points (6 and 24 hrs), using data-independent acquisition on Orbitrap Exploris™ 480 Mass Spectrometers. We generated over 40,000 viability measurements for over 4000 pairwise drug combinations across three cell lines—the largest dataset of its kind focusing on repurposed drugs for anticancer regimens. Combinations demonstrating shifts beyond Bliss in potency (ΔIC50 &gt; 3) or efficacy (ΔEmax &gt; 0.2) were classified as synergistic. Notably, less than 10% of the combination-cell line tuples exhibited synergy, underlining the rarity of effective combinations. Furthermore, synergistic drug combinations showed a strong correlation with tissue origin. Next, we produced over 20,000 perturbed proteomic data, quantifying over 10,000 proteins with high confidence and technical reproducibility. Unsupervised clustering showed cell line-specific proteome responses to perturbations, highlighting the cell context-dependence of the perturbation proteome. We compared perturbed proteomic data between synergistic and ineffective drug combinations to uncover potential molecular mechanisms of synergy effects. Utilizing baseline genomic data, perturbed proteomic data, and SMILE as training inputs, we developed a neural Ordinary Differential Equations model to predict the efficacy of independent drug combinations from a pool of repurposing drugs and anticancer drugs. The model achieved an AUC above 0.80. We further validated a few representative drug synergies in cell line models and explored the targets of the selected drug combinations by thermal proteome profiling experiments.

Conference room 5-6

Session

Advancing drug discovery

Session
Advancing drug discovery

Appointment

Chairs

Program

SessionAdvancing drug discovery

Appointment

Chairs

Program

Session
Advancing drug discovery