Fig. 2: Structural diversity of DNA-shell-coated vesicles.

a Relative abundance of particle subspecies as a function of cholesterol (chol) count. Besides octahedral and icosahedral DNA-shell-coated vesicles (DCVs), pentagonal DCVs—likely representing incomplete (‘scarred’) icosahedra—were frequently observed across all samples. Triangles with 3 chol/triangle yielded more empty shells and DCVs whose vesicles do not fill up the shell cavity (‘partial DCVs’), suggesting weaker membrane binding. While some may be staining artefacts, their enrichment at low cholesterol supports a genuine trend. Particles with unclear morphology were excluded. Numbers in brackets indicate absolute particle counts (n). b Identification of shell subspecies by gel extraction and transmission electron microscopy (TEM). Electrophoretic mobility reflects particle size: icosahedral shells migrate slowest (top band), octahedral fastest (bottom), with intermediates (e.g. pentamers, half-shells) in between. Monomeric triangles run below shell assemblies but are often faint. DCV formation is assessed by agarose gel electrophoresis using fluorescently labelled DNA or ethidium bromide (white) and a small fraction of labelled DOPE lipids in the vesicle mixture (magenta). Lipid signal colocalises only with assembled shells, indicating membrane material migrates primarily as part of DCVs. DNA retained in the gel pocket reflects vesicle-bound triangles; the presence of a monomer band indicates saturated membranes. Pockets were oversaturated to visualise faint bands. c Summed z-slices depicting the top, middle and bottom segments of DCVs and an empty shell obtained by electron tomography (ET) of negatively-stained (ns) or vitrified (cryo) samples. The lipid vesicle is visible in the middle segments of DCVs, confirming engulfment. In the vitrified example, a gap in the DNA shell (arrows) also exposes the vesicle in the bottom segment. Cryo-images were Gaussian-blurred to enhance contrast. d DCVs occasionally appear incomplete with ‘scars’ of missing triangles (arrows). Steric hindrance at the bud neck can prevent closure, leaving gaps in the shell that remain visible after neck scission. Top: ns-TEM; Bottom: cryoEM. Scale bars: 100 nm. Source data are provided as a Source Data file.