Our lab focuses on big questions in cell biology, particularly: what are the proximate and ultimate causes of lipid movement within and between cells? What are the metabolic and neurodegenerative consequences of mistakes in lipid movement? How can biologists use microscopy to better understand how molecules and organelles traffic through cells? To answer these questions, the research goals of the Cohen Lab are focused on three interrelated projects:
Development of tools to study organelle dynamics and interactions. Dr. Sarah Cohen previously pioneered a live-cell multispectral imaging approach to study complex interactions between organelles (Valm & Cohen et al., Nature 2017; Cohen, Valm & Lippincott-Schwartz, Current Protocols in Cell Biology 2018). The Cohen Lab continues to develop computational tools for analyzing organelle morphology, dynamics, and interactions, using multispectral organelle imaging data. We created a Python package, infer-subc, to quantify these features in 3-D (available on GitHub). We collaborated with Dr. Serena Yeung’s computer vision lab (Stanford) to develop a machine learning approach to classify organelles according to shape; this allowed us to explore correlations between mitochondrial shape and contacts with other organelles (Burgess et al., Nature Comms. 2024). We also engineered a versatile suite of genetically encoded dimerization-dependent fluorescent probes (ddFPs) to specifically visualize membrane contact sites between any organelle pair in a reversible manner, allowing accurate and precise analysis of organelle contact dynamics (Miner et al., Contact 2024). We made these probes available to the cell biology community via Addgene, and they have already been requested by many other groups. We are continuing to develop new tools for visualizing organelle dynamics and interactions, as well as pushing the boundaries of “imaging ‘omics” methods to analyze and extract information from complex multidimensional microscopy data sets.
Composition and function of lipid droplet-organelle contact sites. Lipid droplets (LDs) are lipid storage organelles composed of a core of neutral lipids surrounded by a phospholipid monolayer. Because of their unique membrane structure, LDs are not connected to vesicular trafficking pathways. Thus, LDs communicate with other organelles primarily via membrane contact sites. We have focused on elucidating the composition and function of endoplasmic reticulum (ER)-LD and LD-mitochondria contact sites. During their biogenesis, LDs originate from the ER. Neutral lipids are synthesized in the ER and accumulate between the leaflets of the ER membrane; LDs then bud towards the cytoplasm and detach from the ER. After dissociating, they continue to interact with the ER via membrane contact sites to exchange lipids and proteins. We characterized the function of multiple C2 domain-containing transmembrane proteins (MCTPs) 1 and 2 at ER-LD contact sites. We found that the reticulon homology domain of MCTPs tubulates the ER membrane, promoting LD biogenesis, while the MCTP C2 domains tether LDs to the ER and increase LD size (Joshi et al., MBoC 2021). In a related study, we used multispectral imaging to determine the effect of mutant forms of the microtubule severing protein spastin on organelle contact sites, including ER-LD contacts (Arribat et al., PLOS Genetics 2020). In addition to the ER, LDs can exchange lipids with other organelles including mitochondria. We investigated the function of LD-mitochondria contact sites in myoblasts by focusing on the LD protein perilipin 5 (PLIN5). We discovered that both phosphorylation of PLIN5 and interaction of PLIN5 with the mitochondrial acyl-CoA synthase FATP4 at MCSs were necessary for efficient LD-to-mitochondria fatty acid flux during starvation (Miner et al., Dev. Cell 2023). This work provides insight into how cells tune their metabolism in response to changing environmental conditions. We are currently using spatial proteomics to identify novel proteins at LD-organelle contact sites. We will then follow up with functional assays to elucidate the diverse physiological roles of these proteins in regulating lipid and protein trafficking, metabolism, and cell signaling.
Organelle membrane contact sites in neurodevelopment and neurodegenerative disease. Organelle membrane contact sites are increasingly implicated in neurodegenerative diseases, including Alzheimer’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis. We helped benchmark the KOLF2.1J reference induced pluripotent stem cell (iPSC) line that serves as the parental line for the NIH-funded iPSC Neurodegenerative Disease Initiative (Pantazis et al., Cell Stem Cell 2022). We also discovered that apolipoprotein E (APOE), the protein encoded by the biggest genetic risk factor for late onset Alzheimer’s disease, can escape secretion and traffic from the ER to the cytoplasmic surface of LDs via ER-LD membrane contact sites in astrocytes (Windham et al., JCB 2024). Once on LDs, APOE modulates LD composition and morphology, with the E4 risk variant causing large unsaturated LDs that are sensitive to peroxidation. We are currently using multispectral imaging to elucidate how organelle morphodynamics and contacts rewire during neuronal differentiation and in response to neurodegenerative disease-associated stressors such as oxidative stress and ER stress (Rhoads et al., bioRxiv 2024).