Identification of PSB15, a phage-encoded protein factor essential for assembly. (A) ϕLf-UK PS T340A/C341G mutant was used for suppressor screening, whose procedures are described in detail in Supplemental Figure S1. (B) Summary of suppressor mutations. Numbers of each isolated suppressor are shown in parentheses (total 100 sequences). Suppressor T5291C (Thr7Ala) and T5246C (Thr22Ala) mutations in PSB15 are labeled in red. (C) (Blue) Phage yields of ϕLf-UK infection in Xcc-TcR and Xcc-TcR-TrxA G93D bacteria. (Red) ϕLf-UK-T340A/C341G and its suppressor T5291C and T5246C mutants released in the medium were measured using qPCR. (Green and organe) In vivo complement assay was carried out by pPSB15 plasmid, which encodes PSB15 protein. pPSB15 was co-transformed with ϕLf-UK-τιPSB15 (green) and T340A/C341G-τιPSB15 RF (orange), followed by qPCR to measure phage production. (D) AlphaFold prediction of PSB15 structure (UniProt: Q8P905). Model confidence based on a per-residue confidence score (pLDDT) is shown by color. (E) RF replication (attP) and site-specific integration of ϕLf-UK-τιPSB15 phage DNA (attL and attR) into the host chromosome was analyzed by PCR. The PCR primers, detail methods, and ϕLf-UK integration (upper) were described previously (24).

DNA binding properties of PSB15. (A) Gel filtration chromatography of recombinant PSB15. The peak fractions were collected and analyzed by 13.5% SDS-PAGE and Coomassie blue staining. (B and C) Oligonucleotides of ϕLf-UK PS and its mutant (B), and Xanthomonas filamentous phages PS (24) were used for EMSA (C). PS DNA-PSB15 complex was separated in 1.5% agarose gel and stained for SYBR Gold. The free DNA probe is at the bottom. (D) Sequence alignment of PSB15 N-terminus of Xanthomonas and Stenotrophomonas filamentous phages. Mutations are listed in the box. (E) EMSA of PSB15 mutants with <Lf-UK PS and PS (-). PSB15 tryptophan 35 and 46 [purple asterisks in (D)] were replaced with alanine (2WA). The free DNA probe is at the bottom. (F) PS-PSB15 complex (arrowhead, F) and a supershift by PSB15 antibody (asterisk, F) were shown. (G and H) PSB15-PS DNA binding kinetics in (C) and (E) from 3 independent experiments.

Identification and conformation changes of PSB15 DNA binding motif. (A) AlphaFold prediction of PSB15 N-terminus LTG (blue) and WAGF (green) motif. F15 is labeled in red. Model confidence is shown at left. (B) UV cross-linking of recombinant PSB15 and 32P-labeled PS DNA. Samples were analyzed in 15% SDS-PAGE (lower panel) and autoradiography (upper panel). (C) UV-crosslinked PSB15-PS complex was trypsinized, and 32P-labeled peptides were purified by HPLC then subjected to Edman sequencing. The recovery yield of phenylthiohydantoin (PTH)-amino acids is shown. (D-G) Intrinsic tryptophan fluorescence (W10) of PSB15 WT(2WA) (“WT”, D and E) and T7A(2WA) (“T7A”, F and G) upon PS and PS(-) DNA binding. Samples were excited at 295 nm, and fluorescence emission spectra was recorded between 310 to 400 nm.

Membrane and phospholipid binding of PSB15. (A) Xcc-TcR bacteria was transformed with pPSB15 plasmid and fractionated as periplasmic (P), cytoplasmic (C), IM, and OM. Samples were analyzed in 15% SDS-PAGE and immunoblotting. MsbA and OprF are the IM and OM marker, respectively. (B) The IM and OM of <Lf-UK-infected bacteria was fractionated on a step sucrose gradient (25 to 61% [wt/wt]) centrifugation, and the fractions were analyzed by immunoblotting. Sucrose density was determined from the refractive index. (C and D) Alignment of PBS15 hydrophobic linker, basic helix 2 (C) and C-terminal tail (D) of Xanthomonas filamentous phages. Mutations are listed in the box. (E) AlphaFold prediction of PSB15 basic helix 2 structure. (F) Putative translocation motif (TrM) for Xanthomonas type III effector proteins (33). (G) pPSB15 mutant plasmids were transformed into Xcc-TcR bacteria. The IM was fractionated as (A) and immunoblotted. Total lysate (total) was probed with PSB15 as the control. (H) Liposome flotation assay. 5 μM PSB15 proteins were incubated with 1mM liposome (PE+PG+CL+PC=52%: 13%: 30%: 5%; PE+CL: 50%: 50%; PE+PG+PC: 40%: 40%: 20%). The mixtures were centrifuged in a step sucrose gradient (2M, 0.75 M, 0 M). The floated liposome-bound (SUV) fractions were collected and analyzed by SDS-PAGE and immunoblotting. Inputs were loaded as the control. (I and J) Binding kinetics of GFP-PSB15 and FM4-64-labeled liposome (PE: PG: CL: PC=52%: 13%: 30%: 5%). Flotation assays were carried out as (H). Fluorescence intensity of GFP-PSB15 was measured and normalized with FM4-64 intensity. (K) Xcc-TcR was transformed with pPSB15 mutant plasmids as indicated, and bacteria were infected with ϕLf-UK-ϕPSB15 phage. The IM was fractionated and cross-linked by formaldehyde, and phage DNA was immunoprecipitated using anti-PSB15 antibody. Precipitated phage DNA was recovered and amplified by PCR (nt 190 to 476 of ϕLf-UK) and analyzed in 1% agarose gel. DNA from cytoplasmic fraction was PCR amplified by the same primers and served as the control (total).

Trx binding and its effects on PSB15. (A) AlphaFold prediction of PSB15 hydrophobic linker and helix 2. (B) Pull down assay of GST-TrxA with PSB15. TrxA mutant G93D was used as the negative control (white arrowhead). (C) TrxA-PSB15-PS complex (black arrowhead) was analyzed in EMSA. The supershift by anti-GST antibody (asterisk) was shown. Lanes with TrxA-G93D were labeled with white arrowhead. (D) Thioredoxin was added with PSB15 and liposome (PE:PG:CL:PC= 52: 13: 30: 5) for 30 min. Mixtures were subjected to a floatation assay. Liposome-bound PSB15 and input of TrxA (and G93D mutant) were analyzed by immunoblotting. (E and F) Kinetics of liposome-bound GFP-PSB15 (E) and FITC-labeled PS DNA (F) in the flotation assay in response to TrxA and TrxA G93D concentration. F15A acts as a negative control, which confirmed that PS DNA is bound to liposomes via PSB15.

PSB15 dynamics in vivo. (A) Live cell imaging of GFP-PSB15 under TIRF microscopy. pGII-transformed Xcc-TcR bacteria were infected with <Lf-UK GFP-PSB15 or PS(-)-GFP-PSB15 phages at a multiplicity of infections of 1 for 30 min. Snapshots of time-lapse movies were shown. The timestamps are labeled on the upper right corner (sec) in the corresponding movies (Movies S1-S7) and the IM was labeled by FM4-64 dye and margined. Bar = 1 μm. (B-E) Example of brightness of GFP-PSB15 spots at cell membrane was recorded in (A). Dwelling time is defined as the duration of GFP signal more than 10 arbitrary unit (A.U.; pink line)

Working model of filamentous phage assembly and its checkpoint control. PSB15 efficiently targets phage DNA to the IM through PS binding at the priming site in the cell pole. Other essential factors (chaperons, scaffolding factors etc.) may participate in the prerequisite reactions and/or the formation of the “initiation complex” to satisfy the assembly checkpoint. Once the checkpoint is satisfied, Trx releases PSB15 and phage DNA from the priming site, then phage DNA is loaded to pI export complex. pV is replaced by major coat protein pVIII, driven by ATP hydrolysis of pI. Virus particles assemble and pass through the OM (likely via the host secretin complex, such as type II secretion system T2SS). pVI and pIII are added to terminate the assembly process, and virions are released out of the host. The structures were illustrated based on AlphaFold or PDB as the following: PSB15 (Uniport: Q8P905), pIII (Uniport: Q3BSR6), pVI (Uniport: Q3BSR8), pVII (Uniport: Q3BSR4), pVIII (PDB:2IFO), thioredoxin (PDB: 2TRX), phage (PDB: 2IFO, 8B3P, 8B3O, 8IXK, 8IXL). Both ends of phage are adopted from f1 and M13 phage structures (Conners et al., 2023; Jia and Xiang, 2023).