Identification of the bacteriophage nucleus protein interaction network

Abstract

In the arms race between bacteria and bacteriophages (phages), some large-genome jumbo phages have evolved a protein shell that encloses their replicating genome to protect it against DNA-targeting immune factors. By segregating the genome from the host cytoplasm, however, the "phage nucleus" introduces the need to specifically transport mRNA and proteins through the nuclear shell, and to dock capsids on the shell for genome packaging. Here, we use proximity labeling and localization mapping to systematically identify proteins associated with the major nuclear shell protein chimallin (ChmA) and other distinctive structures assembled by these phages. We identify six uncharacterized nuclear shell-associated proteins, one of which directly interacts with self-assembled ChmA. The structure and protein-protein interaction network of this protein, which we term ChmB, suggests that it forms pores in the ChmA lattice that serve as docking sites for capsid genome packaging, and may also participate in mRNA and/or protein transport.

Figure 1.. Identification of jumbo phage nuclear shell-associated proteins

(a) Subcellular localization of GFP-tagged ΦPA3 RecA (gp175) and ChmA (gp53) in uninfected (left) and ΦPA3-infected (right) P. aeruginosa cells. GFP is shown in green, FM4–64 (to visualize membranes) in red, and DAPI (to visualize nucleic acids) in blue. Scale bar = 2 μm. (b) Experimental schematic for identification of jumbo phage nuclear or nuclear-shell-associated genes, by proximity labeling with miniTurboID-fused RecA or ChmA in ΦPA3-infected P. aeruginosa cells. See Figure S1a for fusion construct design, and Figure S1b for localization of miniTurboID-fused proteins. See Tables 1 and 2 for top 25 identified proteins, Tables S1 and S2 for full protein lists, and Figure S1c–d for diagrams showing overlap between independent mass spectrometry datasets. (c-e) Subcellular localization of selected proteins identified by proximity labeling, with panel (c) showing nuclear-localized proteins, (d) showing nuclear shell-associated proteins, and (e) showing phage bouquet-associated proteins. See Figure S2 for further data, and Table 3 for a collated list of localizations. Scale bar = 2 μm.

Figure 2.. gp2 is an interaction hub in the jumbo phage nuclear shell

(a) Colocalization of mCherry-fused ΦPA3 ChmA (red) and GFP-fused ΦPA3 gp2 (green) in P. aeruginosa cells. Scale bar = 2 μm. (b) Colocalization of mCherry-fused 201Φ2–1 ChmA (gp105; red) and GFP-fused 201Φ2–1 gp2 (green) in P. chlororaphis cells. Scale bar = 2 μm. (c) Silver stain SDS-PAGE analysis of GFP pulldown experiments. EV: empty vector. Dotted box indicated the gel slice that was cut out (of the same bands in a Coomassie blue-stained gel) for tryptic mass spectrometry protein identification (see Table S3). (d) Interaction network of the jumbo phage nuclear shell, with blue arrows indicating interactions identified by ChmA miniTurboID and green arrows indicating interactions identified in GFP pulldowns (see Figure S4a for SDS-PAGE gels of all analyzed GFP pulldown samples, and Table S4 for full data). nvRNAP: non-virion RNA polymerase. (e) Ni²⁺ pulldown analysis of E. coli-coexpressed 201Φ2–1 gp2 (His₆-tagged) and ChmA (full-length or truncated: ΔN missing residues 1–63, ΔC missing residues 583–631, and ΔN+C missing residues 1–63 and 583–631). Doublet bands for ChmA arise from a methionine codon at position 33 of the annotated gene. Orange marks show the presence of ChmA in the lysates. See Figure S4b for control pulldown. (f) Ni²⁺ pulldown analysis of E. coli-coexpressed ΦPA3 gp2 (His₆-tagged) and portal (gp148).

Figure 3.. Structure of gp2

(a) Size exclusion chromatography coupled to multi-angle light scattering (SEC-MALS) of ΦPA3 gp2, showing that it is homodimeric in solution (monomer molecular weight = 22.5 kDa). See Figure S5a–c for SEC-MALS analysis of other jumbo phage gp2 proteins. (b) Structure of the PA1C gp2 homodimer, with one protomer colored gray and the other colored as a rainbow from N-terminus (blue) to C-terminus (red).

Figure 4.. gp2 mutations cause defects in phage nucleus formation and morphology

(a) Fluorescence imaging of ΦPA3-infected P. aeruginosa cells expressing no additional proteins (EV: empty vector), GFP-tagged wild-type gp2, or GFP-tagged gp2 point mutants. Un-deconvolved images that were used for DAPI quantitation (panels b and c) are shown. GFP is shown in green, FM4–64 (to visualize membranes) in red, and DAPI (to visualize nucleic acids) in blue. Scale bar = 2 μm. (b) Phage nuclear area of ΦPA3-infected P. aeruginosa cells expressing no additional proteins (EV: empty vector), GFP-tagged wild-type gp2, or GFP-tagged gp2 point mutants. n=100 for all samples; error bars represent mean +/− standard deviation. P-values: ns: p>0.05 (not significant); **:p<0.01; ****:p<0.0001. (c) Total nuclear DNA in ΦPA3-infected P. aeruginosa cells expressing no additional proteins (EV: empty vector), GFP-tagged wild-type gp2, or GFP-tagged gp2 point mutants, calculated by multiplying each cell’s average DAPI signal within the nucleus by that cell’s nuclear area (panel (b)). (d) Visual phenotypes observed in ΦPA3-infected P. aeruginosa cells expressing GFP-tagged wild-type gp2 (n=100 cells). See Figure S6c for additional examples. (e) Visual phenotypes observed in ΦPA3-infected P. aeruginosa cells expressing GFP-tagged gp2 A159D (n=100 cells). See Figure S6D for additional examples, and Figure S5e for examples of similar phenotypes from gp2 Q53A.

Figure 5.. Model for jumbo phage protein localization and nuclear shell architecture and function.

(a) Schematic of a ΦPA3-infected P. aeruginosa cell with assembled phage nucleus (blue) bounded by a ChmA lattice (orange). Proteins that we find localize to the nucleus, nuclear shell, phage bouquet, or cytoplasm are listed. ChmB (pink) is shown integrated into the ChmA lattice, where it may mediate the docking of phage capsids by binding the portal protein, for genomic packaging. Further interactions with gp61/gp63/gp64 (light yellow) or other shell-associated proteins could accommodate mRNA export or specific protein import. (b) Schematic of the ChmA lattice derived from cryoET analysis of intact 201Φ2–1 and Goslar nuclear shells. ChmA is shown in orange with N-terminal and C-terminal tails shown blue and red, respectively. Removal of four contiguous ChmA protomers from the lattice would leave a cavity ~11.5 × 11.5 nm (two possibilities shown), which could be filled by an assembly of ChmB to generate a pore.