Resolution and Probabilistic Structural Models of Subcomponents Derived from CryoEM Maps of Mature P22 Bacteriophage

Abstract

CryoEM continues to produce density maps of larger and more complex assemblies with multiple protein components of mixed symmetries. Resolution is not always uniform throughout a cryoEM map, and thus it can be very useful to be able to estimate resolution in specific components. Also, measuring the uncertainty in structural models built based on density maps is becoming increasingly important especially at medium to low resolutions. Here, we present procedures to 1) estimate resolution in subcomponents by gold-standard FSC, 2) validate modeling procedures, particularly at medium resolutions, which can include loop modeling and flexible fitting, and 3) build probabilistic models that merge high-accuracy priors (such as crystallographic structures) with medium resolution information (the cryoEM densities they are fitted to). The main goal of such procedures is to better understand what information and insights can be obtained from cryoEM density maps, particularly at medium resolution. As an example, we apply these methods to new cryoEM maps of the mature bacteriophage P22 virion, reconstructed without imposing icosahedral symmetry. Resolution estimates based on gold-stadard FSC show the highest accuracies are in the coat region (7.6A), while portal, hub, adhesin, and needle components are seen at lower resolutions (9.2A, 8.5A, 10.9A, and 10.5A respectively). These differences are indicative of inherent structural heterogeneity and/or reconstruction accuracy in different subcomponents of the map. Using loop modeling and flexible fitting to augment and mold crystallographic models based on observed density, we then build probabilistic models of some of the subcomponents including the portal, hub, and adhesin proteins, leading to some new insights into the structure and composition of the P22 virion.