Agenda

The RAS family of GTPases (HRAS, KRAS4A, KRAS4B, NRAS) are the most frequently mutated proteins in human cancer. Their low endogenous abundance, labile post-translational modifications (PTMs), and high sequence identity all provide formidable challenges to standard proteomic workflows. The NCI RAS Initiative has thus employed top-down mass spectrometry (TDMS), which measures intact and modified protein forms (proteoforms), to precisely identify and confidently localize mutations, PTMs, and compounds within each RAS isoform. We have optimized novel denaturing top-down proteomic assays to evaluate RAS proteoforms on Orbitrap Fusion Lumos, Orbitrap Exploris 480, and Exactive Plus EMR platforms (Thermo), along with automated data acquisition and analysis methods including BioPharma Finder and OptiMSe software (Thermo). We have also performed extensive manual data analysis and RAS proteoform validation for all assays by employing the Xtract algorithm (Thermo) and ProSight Lite (NRTDP) to ensure the accuracy of our findings. In doing so, we have evaluated the specificity and selectivity of KRAS-targeting compounds by targeted TDMS, performed high-throughput targeted TDMS analysis of recombinant RAS proteoforms, and identified a wealth of novel RAS proteoforms within a panel of malignant cell lines. Our results demonstrate the ability of top-down proteomics to improve our understanding of RAS proteoforms while laying the groundwork for development of future RAS-targeting compounds.

Radical Protein Footprinting (RPF) is a powerful tool to probe protein higher order structure. It is mostly performed in vitro, but recent advances have enabled its use in live cells, nematodes, and 3D cultures. However, application in living mammalian tissues has not been accomplished. Here, we present the first successful use of radical protein footprinting in mammalian whole blood from wildtype (WT) and a Cg Dock7m +/+ Leprdb/J age-matched type 2 diabetes mouse model (n=4 for each group). The diabetic mouse model showed significant weight gain and glucose elevation over WT controls.

Using 200 mM persulfate photoactivated with the FOX Photolysis System immediately after mixing with EDTA-stabilized whole blood, we achieved effective protein labeling without significant disruption to blood cell morphology as observed by both bright field microscopy and hemolysis assay. An optimized quenching protocol eliminated background labeling. A significant increase in labeling of extracellular proteins was achieved at room temperature compared to 0V controls, with intracellular labeling being much lower. We observed improved labeling of intracellular proteins at higher temperatures, indicating that labeling specificity can be somewhat controlled by labeling temperature. Targeted analysis of proteins was used to allow for more detailed HRPF information of the top 11 proteins detected. Multiple proteins in the top 11 were found to have disease-associated conformational changes. Among them, complement c3 showed conformational changes consistent with complement activation while serotransferrin showed conformational changes consistent with a higher amount of iron binding. These conformational changes were confirmed by orthogonal assays, and are consistent with previously published reports from type 2 diabetes models and patients, while giving additional information regarding specific changes in protein topography that occurs in these disease states. Work is ongoing to further automate this method for application to clinical samples, as well as adaptation to CSF and other clinical biofluids.

Pulmonary fibrosis encompasses a heterogeneous group of interstitial lung diseases, of which idiopathic pulmonary fibrosis (IPF) is the most common and deadly. Despite advances in antifibrotic therapy, outcomes remain poor and treatment selection remains empiric. Traditional radiographic and physiologic measures incompletely capture the biologic diversity of these patients, limiting our ability to tailor therapy.

Proteomics provides an opportunity to move beyond descriptive disease labels by identifying molecular endophenotypes—biologically distinct subgroups that transcend conventional clinical classifications. By profiling the bronchoalveolar lavage fluid and lung tissue proteome across IPF and autoimmune-associated ILD, we aim to define shared and divergent molecular programs driving fibrogenesis, inflammation, and immune dysregulation.

In this presentation, I will highlight recent work applying unbiased, high-resolution mass spectrometry to large, well-characterized ILD cohorts. Using clustering and systems biology approaches, we demonstrate reproducible molecular signatures associated with disease progression, treatment response, and survival. Specific proteomic endophenotypes suggest the presence of “hypoinflammatory” versus “immune-activated” fibrotic states, offering a framework for precision medicine approaches.

Ultimately, proteomics-derived molecular endophenotypes can refine prognosis, guide rational treatment selection, and identify novel therapeutic targets. By integrating these signatures with quantitative imaging, genomics, and clinical data, we move toward a biologically anchored taxonomy of pulmonary fibrosis that has direct translational relevance.

The intestinal epithelium is a critical barrier that confines the gut microbiota to the lumen, preventing these generally beneficial microbes from invading host tissues and triggering overt inflammation. While gut homeostasis requires this physical separation of the microbiota and host, it also depends on symbiotic interactions between them, which often occur through small molecule metabolites. For instance, by converting indigestible fiber into short-chain fatty acids (SCFA), the microbiota provides colonic epithelial cells with a fuel source to power fatty acid oxidation (FAO) and energy production. Accordingly, intestinal tissues from germ-free mice show signs of energy deprivation, and the clinical significance of this metabolic integration is evident from studies correlating inflammatory bowel disease (IBD) with impaired FAO and changes in microbiota composition. However, the mechanisms by which these alterations could contribute to IBD remain unclear, highlighting the need to understand the pathways linking epithelial cell metabolism with microbial metabolism and whether disruptions in this symbiosis could promote inflammation. In a recent study to investigate metabolic contributions from the microbiota to the host, we performed stable isotope tracing in a mouse model and found that acetylated histones and numerous metabolites in epithelial cells contain microbiota-derived carbon. Treatment with DSS, a chemical irritant that induces inflammation, suppressed the isotopic labeling of molecules bearing fatty acid chains, coinciding with shifts in microbiota composition and decreased expression of genes related to FAO. In ongoing work, we are employing cell and organoid models to trace the anabolic and catabolic pathways by which epithelial cells metabolize microbiota-derived carbon. Overall, we aim to generate mechanistic insight into how metabolic interactions between the gut microbiota and host promote tissue homeostasis and how a disruption in this symbiosis may contribute to the development of IBD.

Accurate cellular differentiation relies on precise regulation of RNA processing and decay that coordinates transcript maturation and turnover across developmental transitions. RNA surveillance machinery safeguards the transcriptome to ensure these transitions proceed correctly. Although many RNA surveillance systems are ubiquitously expressed, their dysregulation often causes tissue-specific neuropathology. At the center of this control is the RNA exosome, an essential 3’-5’ ribonuclease complex implicated in stem cell differentiation. Recessive mutations in EXOSC3, a structural subunit gene of the RNA exosome complex, cause Pontocerebellar Hypoplasia Type 1b (PCH1b), a severe neurodevelopmental disorder characterized by degeneration of the pons and cerebellum. The role of RNA surveillance in human cerebellum development, and the specific contribution of EXOSC3, has remained unclear. Here, we establish a human in vitro model of PCH1b by CRISPR-editing human induced pluripotent stem cells (hiPSCs) to introduce patient-derived EXOSC3 alleles spanning mild and severe disease, and differentiated into cerebellar organoids in 3D culture. Immunofluorescence and single-cell transcriptomics benchmarked to the human fetal cerebellar atlas show that our organoids recapitulate the expected cell-state landscape and developmental gradients, thus providing a platform to study RNA exosome-linked cerebellar phenotypes. Our findings reveal that pathogenic EXOSC3 variants do not broadly collapse lineages; instead, the principal effect is selective: the Purkinje trajectory is perturbed. Cells initiate the Purkinje trajectory but fail to consolidate late maturation features, accumulating at intermediate states. These results are consistent with a maturation checkpoint on the Purkinje branch, while other lineages progress comparatively well. The magnitude and direction of these alterations track with allelic severity, consistent with a graded requirement for RNA exosome function during high-flux developmental transitions. Together, these findings define the RNA  exosome as a gatekeeper of Purkinje maturation, clarify the cellular basis of PCH1b, and establish a platform to study how RNA quality control shapes brain development and disease.

Proteoforms—the diverse protein products arising from genetic variation, alternative splicing, and post-translational modifications—are the functional units that drive biological processes and disease. Top-down proteomics (TDP) enables direct analysis of intact proteoforms, providing molecular resolution essential for mechanistic insight and precision medicine. However, widespread adoption of TDP has been limited by challenges including protein solubility, proteome complexity, and analytical sensitivity. In this presentation, I will describe a multifaceted strategy developed in our laboratory to overcome these barriers. We have engineered new MS-compatible cleavable surfactants to improve protein solubilization, introduced novel materials and methodologies for enhanced multidimensional chromatographic separations, and developed nanomaterials for enrichment of low-abundance proteins. I will also highlight recent advances in high-sensitivity LC–MS workflows, including a functionally integrated platform for single-cell proteoform analysis that reveals cellular heterogeneity. In addition, we are developing multidimensional LC strategies incorporating orthogonal separation modes to expand proteome coverage and enable detection of proteins and protein complexes directly from human heart tissues for native TDP. I will further describe a newly developed native nanoproteomics approach for enriching and characterizing endogenous cardiac troponin complexes from human myocardium, along with a comprehensive, user-friendly software platform for both denatured and native TDP data analysis. Collectively, these advances address key analytical challenges in TDP and enable broader adoption of proteoform-resolved proteomics for elucidating complex biological systems and disease mechanisms.

Across a series of proteomics-driven studies, the Le Roch Lab, in collaboration with the Florens Lab, has employed an integrated suite of mass spectrometry and complementary omics technologies to uncover how the human malaria parasite Plasmodium falciparum regulates its genes through chromatin, RNA, and protein-complex interactions.

Early work applied quantitative proteomics to chromatin to map the parasite’s histone modifications during the red blood cell cycle. Using multidimensional protein identification technology (MudPIT), the team discovered more than 200 distinct post-translational histone marks, many unique to Plasmodium that change dynamically with developmental stage. These findings revealed a distinctive and highly combinatorial “histone code” linking epigenetic state to transcription, DNA replication, and virulence regulation. Building on this work, the labs used chromatin enrichment for proteomics (ChEP) to identify proteins controlling epigenetic features during parasite development and to define the chromatin-bound proteome, revealing both conserved and parasite-specific chromatin regulators. Together, these results highlighted that chromatin organization and associated protein complexes are central to transcriptional control in P. falciparum. Subsequent proteome-wide studies expanded the focus from DNA- to RNA-level regulation, including the use of the R-DeeP sucrose-gradient mass spectrometry method to catalog nearly 900 RNA-dependent proteins. This analysis revealed that non-coding RNAs scaffold many of the parasite’s regulatory complexes, influencing not only chromatin structure and epigenetic regulation but also RNA processing and translation. Targeted functional proteomics of DNA-binding and RNA-binding protein families, notably the MORC and RAP proteins, connected specific factors to chromatin structure and organellar RNA processing in the mitochondrion and apicoplast, demonstrating how parasite-specific mechanisms of gene regulation contribute to survival.

Together, these proteomic frameworks illustrate how the Le Roch and Florens Labs have, across multiple studies, defined the complex interface between chromatin and RNA–protein interactions, uncovering the networks that govern gene expression in the malaria parasite, and providing new mechanistic insights and potential therapeutic targets.

Liver diagnostic assays in the clinic have changed little since the deployment of the liver enzyme panel in the 1960s. This is striking due to the large proteomic interplay between blood and the liver, with as much as 70% of blood plasma proteins originating, in some way, from the liver. Single cell proteomics (SCP) is a flashy new tool that can be used to obtain unprecedented depth and granularity in the phenotypes of biological systems. In this talk I will detail our lab's ongoing work to characterize wasted material from both liver donors and transplant recipients following surgeries at the Pittsburgh Liver Research Center. Human hepatocytes have proven remarkably compatible with SCP methods due to their relatively large size and suspiciously limited proteomic complexity on a cell-by-cell basis. At the point of writing this abstract we are routinely quantifying more than 3,000 protein groups from single human hepatocytes derived from patient samples at a throughput approaching 100 cells per instrument per day with label free methods.