Eukaryotic cells utilize the secretory pathway to transport proteins and lipids. Anterograde transport from the endoplasmic reticulum (ER) to the Golgi apparatus is mediated by COPII-coated vesicles. COPII, comprised of Sar1, Sec23, Sec24, Sec13, and Sec31, assembles on the ER membrane, generating curvature and recruiting cargo. Sar1 initiates assembly by inserting into the ER membrane, recruiting Sec23-Sec24 (inner coat), followed by Sec13-Sec31 (outer coat) assembly. GTP hydrolysis by Sar1, stimulated by Sec23 and Sec31, is thought to destabilize the coat, though the dynamics are unclear. Previous studies using giant unilamellar vesicles (GUVs) showed COPII forms tubules; however, in vivo observations suggest spherical vesicles. This study aims to understand how native membrane composition affects COPII assembly and budding morphology, and how cargo binding is compatible with the tightly packed inner coat.

Prior research has elucidated the structural components and assembly mechanism of the COPII coat. The roles of Sar1, Sec23/Sec24, and Sec13/Sec31 in initiating and building the coat have been well-established, including the cargo-binding function of Sec24. Previous studies utilizing cryo-electron tomography (cryo-ET) and subtomogram averaging (STA) on reconstituted COPII coats on GUVs revealed a tubular morphology and provided insights into coat interactions. These studies allowed the design of coat mutants and showed that membrane curvature is generated by interactions across both coat layers. However, the in vivo morphology of COPII carriers and the influence of native membrane components and cargo remain unclear, creating a knowledge gap addressed in this study.

The researchers reconstituted COPII budding in vitro using purified *S. cerevisiae* COPII proteins and native ER-enriched microsomes (as a source of native membranes). They used a non-hydrolyzable GTP analog (GMP-PNP) to prevent coat disassembly. Cryo-ET and STA were employed to visualize and analyze the resulting coated carriers. For comparison, they also reconstituted budding using GUVs enriched with the cytosolic domain of a small cargo protein, Sed5. Mass spectrometry (MS) was used to analyze the composition of the microsomal membranes. High-resolution STA was performed on both microsome-derived vesicles and Sed5-enriched GUV-derived tubes to determine the structure of the inner and outer coats at high resolution.

In contrast to the tubular structures observed on GUVs, COPII primarily formed pseudospherical vesicles (96.3%) on microsomes. The inner coat assembled in small, randomly oriented patches on the vesicle surface, with cargo density detected between inner coat subunits, suggesting that small, flexible cargo can be accommodated within the lattice. STA revealed that the inner coat arrangement within these patches resembles the pseudohelical lattice seen on tubules. The outer coat showed significant heterogeneity, with diverse cage-like structures comprising Sec13-Sec31 rods forming various vertices (four-way, five-way, T-junctions) and possessing considerable flexibility with hinge regions identified. The inner and outer coat layers showed no fixed alignment, suggesting the Sec31 linker allows for translational and rotational freedom between the two layers. Experiments using Sed5-enriched GUVs confirmed that small, flexible cargo can bind to the inner coat without disrupting lattice formation; however, it was hypothesized that the presence of larger cargoes in microsomes disrupts the lattice formation, leading to vesicle formation.

The results challenge the prevailing notion that the outer coat cage assembly is the primary driver of membrane curvature. Instead, the study proposes a model where inner coat assembly dictates membrane shape. Extensive inner coat polymerization leads to tubules, while the presence of bulky cargo proteins on native membranes prevents extensive polymerization, resulting in small, randomly oriented patches and pseudospherical vesicle formation. The outer coat's adaptability to varying curvatures likely stabilizes the vesicles. This model is supported by previous studies using COPII mutants with weakened inner coat interfaces. The findings highlight the importance of native membrane components and cargo in shaping COPII-mediated vesicle formation.

This study provides a detailed structural analysis of COPII assembly on cargo-containing membranes, revealing a more complex and dynamic process than previously appreciated. The finding that inner coat assembly is the primary determinant of membrane curvature offers a new perspective on COPII-mediated vesicle formation. Future research could explore the mechanisms regulating inner coat polymerization on native membranes and the roles of specific cargo proteins in determining vesicle shape and scission.

The study primarily utilized in vitro reconstitution. While microsomes provide a more native-like environment than GUVs, they may not perfectly recapitulate all aspects of the in vivo process. The diversity of cargo proteins in the microsomes made it challenging to resolve the structure of individual cargo molecules. Furthermore, the use of a non-hydrolyzable GTP analogue may have affected the scission process.