In situ cryo-EM of protein condensates in the early secretory pathway of neurons

Neurons are highly polarized cells that exhibit distinct moieties that serve to receive (dendrites) and transmit (axons) electrochemical signals. In a variety of neurodegenerative diseases, loss of neuronal polarity represents a hallmark of pathogenesis, and coincides with loss of the endomembrane organization in neurons. Despite a growing inventory of compartment-specific factors and machinery, we lack a fundamental understanding how neuronal polarity is established and maintained. Specifically, a vast knowledge gap exists with regard to how bulk routing of cargo is achieved in neurons. Biosynthetic cargo, such as plasma membrane receptors and channels involved in relaying the electrochemical signals between neurons, must be selectively delivered to either dendritic or axonal surfaces to support neuron function. Biosynthesis of secretory cargo in neurons occurs in the endoplasmic reticulum (ER), a compartment that extends throughout dendritic and axonal parts of the neuron. At nanoscale domains of the ER, the ER exit sites (ERES), cargo is routed via membranous carriers toward the Golgi apparatus. Golgi acceptor membranes remain near ERES, forming a 300-nm, nearly spherical interface that exhibits a high extent of long-range order. In addition to a ribbon-like Golgi in the perinuclear region of the cell soma, recently, additional Golgi elements termed Golgi outposts, or Golgi satellites, have been detected at the base of dendritic arbors – prompting the view that Golgi outposts could route cargo into the dendrite, while putatively, somatic Golgi might route cargo into the axon. In the past years, the powerful framework of biomolecular condensation has gained attention as a mechanism putatively capable of compartmentalizing cytosol and structuring membrane compartments, with proteins forming dynamic and flexible networks by the means of a liquid-liquid phase separation (LLPS). Recent evidence points to the ability of multiple proteins in the early secretory pathway to undergo LLPS – supporting the view that the early secretory pathway may be structured by self-organizing protein collectives. We will differentiate human induced pluripotent stem cells (hiPSCs) into mature cortical neurons that are stably transfected with fluorescently tagged ER-Golgi interface components as molecular ‘beacons’ to enable cryo-correlative light and elecron microscopy (cryo- CLEM) to guide cryo-focused ion beam (cryo-FIB) milling in conjunction with cryo-electron tomography (cryo- ET). Our goal is to dissect the molecular architecture of the dendritic and somatic early secretory pathway at molecular resolution in situ. This will yield an unprecedented opportunity to systematically compare the molecular composition of individual secretory hubs, and to identify the role of condensates in the organization of these critical cellular structures while revealing putative adaptations that might enable selective routing of cargo. Our approach will fill critical knowledge gaps in the organization of secretory compartments in neurons, and lead to the innovation of experimental and computational tools for the community. It will further provide a reference point for the analysis of neuron dysfunction under various neurodegenerative conditions.