NMR assignments for the telokin-like domain of bacteriophage P22 coat protein

Abstract

The bacteriophage P22 virion is assembled from identical coat protein monomers in a complex reaction that is generally conserved among tailed, double-stranded DNA bacteriophages and viruses. Many coat proteins of dsDNA viruses have structures based on the HK97 fold, but in some viruses and phages there are additional domains. In the P22 coat protein, a "telokin-like" domain was recently identified, whose structure has not yet been characterized at high-resolution. Two recently published low-resolution cryo-EM reconstructions suggest markedly different folds for the telokin-like domain that lead to alternative conclusions about its function in capsid assembly and stability. Here we report (1)H, (15)N, and (13)C NMR resonance assignments for the telokin-like domain. The secondary structure predicted from the chemical shift values obtained in this work shows significant discrepancies from both cryo-EM models but agrees better with one of the models. In particular, the functionally important "D-loop" in one model shows chemical shifts and solvent exchange protection more consistent with β-sheet structure. Our work will set the basis for a high-resolution NMR structure determination of the telokin-like domain that will help improve the cryo-EM models, and in turn lead to a better understanding of how coat protein monomers assemble into the icosahedral capsids required for virulence.

Figure 1

Assigned ¹H-¹⁵N HSQC spectrum of the TL-domain. The spectrum was collected on a 1.5 mM sample of the TL-domain in 20 mM phosphate buffer, pH 6.0, at a temperature of 37 °C. The amino acid sequence numbering scheme is that of the full-length P22 coat protein. Backbone resonance assignments are indicated in blue and side chain amide resonances with red labels and the superscript SC. Peaks that are visible only at lower contour levels than shown are indicated by squares.

Figure 2

Secondary structure of the TL-domain (A) The amino acid sequence of the TL-domain. (B) Amide protons protected from exchange after 2 h in D₂O (dots). (C) Secondary structure prediction using the Jpred-PSSM program (Cole et al 2008). (D) Secondary structure calculated with the Talos+ program (Cornilescu et al 1999) based on the backbone chemical shift assignments reported in this work. (E) The cryo-EM model of Parent et al. (2010) for amino acids 229–315 only, colored according to the Talos+ results. Arrows are drawn for the regions assigned to b-strands in the cryo-EM model. (F) The cryo-EM model of Chen et al. (2011), colored according to the Talos+ results. TheChen et al. (2011) structure is a Ca-only model, and no information on secondary structure is available. For both models regions predicted by the Talos+ analysis to be in b-strands are colored in cyan, a-helix in red, and segments for which no prediction was made in grey.