Mapping the molecular architecture of biomolecular condensates with crosslinking mass spectrometry

The National Institute of General Medical Sciences awarded The Johns Hopkins University $414,709 in project grant funding on May 1, 2026, under Biomedical Research and Research Training (CFDA 93.859) to develop and apply crosslinking mass spectrometry methods for mapping the molecular architecture of biomolecular condensates. The work spans three broad thrusts: advancing XL-MS tools and methodology to capture protein-RNA interactions through photo-crosslinking and developing computational workflows combining artificial intelligence, crosslinking data, and molecular dynamics for integrative modeling of large assemblies; and applying these methods to map the structure of two specific condensate systems—the Caulobacter crescentus PopZ granule and the mammalian nucleolus. Performance takes place in Baltimore, Maryland, with a completion date of February 28, 2031. Biomolecular condensates are dynamic cellular compartments implicated in aging and neurodegeneration, but lack detailed structural models; this award supports in vivo XL-MS applied to map native condensates in situ, leveraging the investigator's combined expertise in XL-MS methodology, protein folding, and biophysics.

Abstract

Biomolecular condensates have become the subject of intense interest across the biological, biophysical, and biomedical communities because they represent an elegant form of higher-order organization that cells can use to create dynamic compartments; they also are connected to a wide range of diseases, particularly those related to ageing and neurodegeneration. However, we lack a detailed structural model for any natural biomolecular condensate. Mapping the architecture of these assemblies is challenging because they are heterogeneous, liquid-like, prominently feature intrinsically disordered regions (idrs), and are exquisitely sensitive to their environment. Crosslinking mass spectrometry (XL-MS) is uniquely suited to map the architecture of biomolecular condensates because IT can freeze structural information about proteins when they are in their native cellular context and is equally adept at probing folded and disordered regions. In vivo XL-MS has not yet been applied to map native condensates in situ. The investigator has a unique combination of technical expertise in XL-MS and background in protein folding and biophysics; hence, is particularly well suited to this task. This proposal consists of three broad thrusts. Firstly, we seek to advance tools and methodology to map condensate structure, by (i) expanding the capacity of existing methods to capture protein-RNA interactions through photo-crosslinking and (II) developing computational workflows that combine artificial intelligence, crosslinking data (as experimental restraints), and molecular dynamics for integrative/hybrid modeling of large assemblies. The other two thrusts focus on two specific condensate systems, the caulobacter crescentus popz granule and the mammalian nucleolus. Popz forms a polar 150-NM microdomain that regulates cell division and physiological changes during caulobacter's lifecycle. We aim to understand how this condensate's structural properties evolve over the course of cellular development, and hypothesize that structural models of this assembly will provide a molecular basis to the processes responsible for differentiating the cell's two poles. The nucleolus is a larger, more complex condensate responsible for ribosome biogenesis in eukaryotes. Whilst some proteins within IT are well-understood, including the structured ribosome assembly factors and a few high-abundance scaffolding proteins, many of its proteins are quite disordered and poorly characterized. We will obtain high- coverage crosslinking maps of the nucleolus with a range of techniques that will additionally focus on rrna- protein crosslinking. These experiments will also characterize interactions among ribosome assembly intermediates and the whole host of nucleolar proteins, thereby uncovering potentially new ribosome biogenesis factors and functional roles for idrs. A further goal is to map contacts between scaffold proteins (fibrillarin and NPM1) to other nucleolar proteins to interrogate the role of heterotypic interactions in supporting the organelle's unique tripartite substructure. To summarize, the PI seeks to lead efforts to apply structural proteomics to build high-resolution models of native biomolecular condensates. The PI anticipates this will represent a third of his research program, with other complementary efforts on protein folding biophysics and aging/neurodegeneration.