The Tail Sheath of Bacteriophage N4 Interacts with the Escherichia coli Receptor

Abstract

Unlike other characterized phages, the lytic coliphage N4 must inject the 360-kDa virion RNA polymerase (vRNAP), in addition to its 72-kbp genome, into the host for successful infection. The process of adsorption to the host sets up and elicits the necessary conformational changes in the virion to allow genome and vRNAP injection. Infection of suppressor and nonsuppressor strains, Escherichia coli W3350 supF and E. coli W3350, with a mutant N4 isolate (N4am229) harboring an amber mutation in Orf65 yielded virions containing (N4gp65⁺) and lacking (N4gp65⁻) gp65, respectively. N4gp65⁺ but not N4gp65⁻ phage was able to adsorb to the host. Recombinant gp65 with a hexahistidine tag at the N terminus or hexahistidine and c-myc tags at the C terminus was able to complement N4gp65⁻ virions in vivo and in vitro. Immunogold detection of gp65 in vivo complemented virions revealed its localization at the N4 tail. Finally, we show both in vitro and in vivo that gp65 interacts with the previously determined N4 outer membrane receptor, NfrA.

Adsorption, the recognition of and docking to a host cell exterior, constitutes the first critical and essential step of all viral infections. Only upon successful adsorption is a virus properly posed to deliver its genetic material into the host cell cytoplasm, where the infection cycle can continue. Bacteriophages rely on adsorption not only for stable docking to the host but also as a signaling event for DNA injection. Bacteriophages that infect gram-positive bacteria must inject their DNA through a cell envelope composed of a thick peptidoglycan meshwork and cytoplasmic membrane, whereas bacteriophage infection of gram-negative bacteria requires DNA injection through the host outer membrane, periplasm, and inner membrane. The mechanism of genome injection, beginning with adsorption to the host and ending with complete delivery of genomic material, remains largely uncharacterized for many bacteriophages.

N4, a bacteriophage that infects the gram-negative bacterium Escherichia coli K-12, presents a notable challenge early in the infection process. Specifically, N4 encodes and encapsidates a DNA-dependent RNA polymerase (RNAP), a 3,500-amino-acid (3,500-aa) nonprocessed polypeptide present at 4 ± 1 copies per virion (5). Virion RNAP (vRNAP) is required for the injection and transcription of the early region of the genome (9; A. Demidenko and L. B. Rothman-Denes, unpublished data). Previous investigations of the initial steps of N4 infection focused on the E. coli host requirements. Mapping of spontaneous E. coli K-12 mutants resistant to N4 infection led to the identification and characterization of an outer membrane protein, NfrA (96 kDa), and an inner membrane protein, NfrB (69.5 kDa), as necessary for N4 adsorption (15, 16); mutations in NfrB do not affect the synthesis or localization of NfrA (15).

N4 virions are characterized by an icosahedral head with T=9 quasisymmetry, a short tail, and 12 appendages projecting from a neck connecting the head and tail (5). The N4 virion is comprised of 10 proteins and a concentrically arranged 72-kbp genome encoding 3 tRNAs and 72 open reading frames (ORFs). Here we show that the second largest N4 virion protein, gp65, which constitutes a sheath surrounding the tail tube (5), is required for adsorption to the host. Moreover, we show in vivo and in vitro that gp65 interacts with the outer membrane receptor NfrA.

MATERIALS AND METHODS

Bacterial strains and media.

E. coli W3350 and E. coli W3350 supF were the nonpermissive and permissive strains used, respectively. In some experiments, E. coli W3350(pNfrA/B), overexpressing the E. coli NfrA and NfrB proteins, was used. E. coli BL21, resistant to phage N4, and E. coli BL21(pNfrA/B) were used to characterize the interaction of gp65 with NfrA. Cells were grown at 37°C in Luria-Bertani (LB) broth, unless otherwise stated, supplemented with 20 μg/ml chloramphenicol for retention of pNfrA/B or with 100 μg/ml ampicillin for retention of pBAD/His Borf65, pBAD/Myc-His Borf65, or pMAL-c4xnfrA Δ1-27.

Characterization of an amber mutation in Orf65.

Candidate phage from an N4 amber phage library were suspended in TM buffer (15 mM Tris-HCl, pH 8.0, 10 mM MgCl₂) at a concentration of 2 × 10⁸ phage/μl. An equal volume of phenol was added, and the mixture was vortexed and centrifuged at 14,000 × g. The top, aqueous layer was phenol extracted and chloroform treated. DNA was ethanol precipitated, washed twice with 70% ethanol, allowed to dry, and resuspended in water. Orf65 was PCR amplified with Pfu DNA polymerase (Stratagene, La Jolla, CA), using the following primers: F, 5′-CGTGTTCAGGTTAAGTTCAG-3′; and R, 5′-GAATCTCCCTAATCTGTTCCC-3′. The Orf65 amplicon was then sequenced with the following primers, beginning from the 5′ end of Orf65: (i) 5′-CGTGTTCAGGTTAAGTTCAG-3′, (ii) 5′-CGTCATAATCCTGATGAACC-3′, (iii) 5′-GTAATGCTCAGGCAGCAGAG-3′, (iv) 5′-GTGCATACCCTGACCGTGGC-3′, (v) 5′-CCTATTCGTACAGGATTACC-3′, (vi) 5′-GCCTGTTAATGTAGCTGCTG-3′, (vii) 5′-GCCATTGAACTAGGTGAAGC-3′, (viii) 5′-CTCTAACATGGACTGTTGCAG-3′, and (ix) 5′-GTTGGACAGGGCTTTGCTAAG-3′.

Isolation of N4 virions containing (N4gp65⁺) or lacking (N4gp65⁻) gp65.

E. coli W3350 supF or E. coli W3350 cells, grown to an optical density at 600 nm (OD₆₀₀) of 0.2, were infected with N4am229 at a multiplicity of infection (MOI) of 10. After incubation for 3 h, cells were lysed by the addition of chloroform. Virions were purified and concentrated by glycerol gradient centrifugation, cesium chloride buoyant density centrifugation, and a final glycerol gradient centrifugation step. Virion proteins were analyzed by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and visualized by silver staining.

Isolation of ³H-labeled N4gp65⁺ and N4gp65⁻ virions.

E. coli W3350 or E. coli W3350 supF was grown in LB to an OD₆₀₀ of 0.2 and infected with N4am229 at an MOI of 10 for 10 min. Cells were then pelleted and resuspended in M9-Casamino Acids medium containing 40 μCi/ml [methyl-³H]thymidine (Amersham, United Kingdom) (2). Infection continued for 3 h before lysis with chloroform. The resultant labeled virions were purified and analyzed as described above.

Adsorption assays.

E. coli W3350(pNfrA/B) cells grown to an OD₆₀₀ of 0.2 were concentrated 10-fold, and 250-μl aliquots were infected with ³H-labeled gp65⁺ or gp65⁻ virions at an MOI of 0.5 for the specified amounts of time to allow adsorption. Cells were then pelleted at 14,000 × g for 1 min. Radioactivities measured in the supernatant and pellet corresponded to unadsorbed and adsorbed phage, respectively. For adsorption competition experiments, cells were concentrated 10-fold, and 250-μl aliquots were infected with a mixture of ³H-labeled gp65⁺ virions and unlabeled gp65⁺ or gp65⁻ virions at the indicated ratios at an MOI of 0.5. Adsorption was allowed for 7 min, at which time the levels of unadsorbed and adsorbed phage were determined as described above.

Cloning of N4 Orf65 and purification of gp65.

N4 Orf65 was PCR amplified using Pfu DNA polymerase and cloned into the arabinose-inducible pBAD/His B or pBAD/Myc-His B vector (Invitrogen, Carlsbad, CA) such that it was hexahistidine tagged at the N terminus (N-tagged gp65) or C terminus (C-tagged gp65), respectively. The following primers were used to clone Orf65 into pBAD/His B, yielding pBAD/His Borf65: Xho F, 5′-CGACTCGAGATCCATTGAAGATTACCTCAAAGGC-3′ (underlined XhoI site); and KpnI R, 5′-GGACTGGGTACCTTAAAGAGTGGCTGTCATATCAAATTG-3′ (underlined KpnI site). The recombinant protein has the N-terminal vector-encoded 36-residue sequence MGGSHHHHHHGMASMTGGQQMGRDLYDDDDKDPSSR. To clone Orf65 into the pBAD/Myc-His B vector, yielding pBAD/Myc-His Borf65, the following primers were used: NcoI F, 5′-GGACTGCCATGGCATCCATTGAAGATTACCTCAAAGGC-3′ (underlined NcoI site); and HindIII R, 5′-GGACTGAAGCTTGCAAGAGTGGCTGTCATATC-3′ (underlined HindIII site). The recombinant protein has the C-terminal vector-encoded 25-residue sequence ASFLEQKLISEEDLNSAVDHHHHHH. The plasmids were transformed into E. coli BL21. E. coli BL21(pBAD/His Borf65) or E. coli BL21(pBAD/Myc-His Borf65) cells were grown to an OD₆₀₀ of 0.2, and the synthesis of recombinant gp65 was induced for 45 min with 0.2% arabinose. Cells were pelleted at 4°C for 10 min at 4,200 × g and then resuspended in sonication buffer (20 mM Tris-HCl, pH 8.0, 20 mM NaCl, EDTA-free protease inhibitor [Roche, Mannheim, Germany]). Cells were sonicated in 15-s pulses followed by 15-s ice incubations until lysis was observed. Cellular debris was pelleted by centrifugation at 4°C for 30 min at 31,600 × g. The tagged recombinant proteins were then purified from the supernatant by immobilized metal ion affinity chromatography (IMAC). Specifically, the supernatants were loaded onto Talon Co²⁺-IMAC resin (Clontech, Mountain View, CA) and permitted to batch bind with rotation at 4°C for 1 h. After 3 bed volume washes with low salt (20 mM Tris-HCl, pH 8.0, 20 mM NaCl), 3 bed volume washes with high salt (20 mM Tris-HCl, pH 8.0, 1 M NaCl), and 3 bed volume washes with low-salt buffer, the recombinant protein was eluted with 100 mM imidazole in low-salt buffer. IMAC-purified gp65 was further purified by binding to Mono Q Sepharose ion-exchange resin (GE Healthcare, Piscataway, NJ) at 4°C with rotation for 20 min. N-terminally tagged or C-terminally tagged gp65 was eluted from the Mono Q ion-exchange resin with 3 bed volumes of high-salt buffer.

Incorporation of recombinant gp65 into virions in vivo and in vitro.

Cultures of E. coli W3350(pBAD/His Borf65) or E. coli W3350(pBAD/Myc-His Borf65) were grown to an OD₆₀₀ of 0.2. Cells were infected with N4am229 at an MOI of 10 for 10 min, after which the infected cells were spun down and resuspended in LB supplemented with 100 μg/ml ampicillin and 0.2% arabinose to induce the synthesis of either N-terminally tagged or C-terminally tagged gp65. After 3 h of incubation, the cells were lysed by the addition of chloroform and the resultant phage progeny was purified as described above. The virion proteins were subjected to 10% SDS-PAGE and silver staining or immunoblot detection using anti-histidine-glycine (anti-His-G) and anti-myc primary antibodies (Invitrogen, Carlsbad, CA) after transfer to Hybond ECL nitrocellulose membranes (GE Healthcare, Piscataway, NJ). For in vitro incorporation, recombinant purified N-terminally tagged or C-terminally tagged gp65 was incubated at a final concentration of 0.67 μg/μl with approximately 4.5 × 10¹⁰ N4gp65⁻ virions for 30 min at 37°C. The resultant virions were purified through a CsCl gradient and analyzed as described above.

Immunogold detection of gp65 in virions.

N4 virions containing recombinant C-terminally c-myc-epitope-tagged gp65 (gp65C*) and N4 virions containing recombinant N-terminally His-G-tagged gp65 (gp65N*) were incubated with primary antibodies against the c-myc epitope and His-G epitope, respectively. Incubation with primary antibody was performed overnight at room temperature. Next, 5-nm-gold-conjugated goat anti-mouse immunoglobulin G secondary antibody (Ted Pella, Inc., Redding, CA) was added for 2 h at room temperature, followed by the addition of 0.1% glutaraldehyde. After a 30-min incubation, virions were negatively stained with 2% ammonium molybdenate and visualized by electron microscopy using an FEI Technai 30 electron microscope. N4gp65⁻ virions, treated as described above, and N4 virions containing gp65N* or gp65C*, treated as described above but excluding incubation with primary antibody, were used as controls.

Pull-down of recombinant N-terminally tagged gp65 by cells overexpressing NfrA and NfrB.

E. coli BL21, E. coli BL21(pNfrA/B), and E. coli W3350(pNfrA/B) cultures were grown to an OD₆₀₀ of 0.2 in LB or LB supplemented with 20 μg/ml chloramphenicol. The cultures were then concentrated 10-fold. Purified N-terminally tagged gp65 in low-salt buffer (20 mM Tris-HCl, pH 8.0, 20 mM NaCl) was added at 40.5 μg/ml to one-half of the cultures, while the other half received the same volume of low-salt buffer. Cells were incubated with the purified protein or low-salt buffer at 37°C for 20 min, after which the cells were spun down and washed five times with LB. The cells were pelleted at 14,000 × g for 10 min at 4°C, resuspended in 6× protein loading buffer (2), and subjected to immunoblot detection. Expression of NfrA was detected using polyclonal rabbit anti-NfrA primary antibodies, and pull-down of N-terminally tagged recombinant gp65 was detected using anti-His-G primary antibody.

Blockage of phage adsorption using purified gp65 protein.

E. coli W3350(pNfrA/B) cells were grown to an OD₆₀₀ of 0.2 and concentrated 10-fold, and purified recombinant N-terminally tagged or C-terminally tagged gp65 at 29 μg/ml, 58 μg/ml, and 115 μg/ml or bovine serum albumin at a final concentration of 385 μg/ml was added to 250-μl aliquots of concentrated cells. The mixtures were incubated for 5 min at 37°C, followed by infection with ³H-labeled gp65⁺ phage at an MOI of 0.5 for 7 min. Cells were then pelleted at 14,000 × g for 1 min. Radioactivities measured in the supernatant and pellet corresponded to unadsorbed and adsorbed phage, respectively.

Cloning and purification of MBP-NfrA.

E. coli K-12 W3350 nfrA was PCR amplified using Pfu DNA polymerase, omitting the Sec pathway secretion signal sequence (aa 1 to 27), and inserted into the isopropyl-β-d-thiogalactopyranoside (IPTG)-inducible plasmid pMAL-c4x (New England Biolabs, Ipswich, MA) such that it was fused at the N terminus to maltose binding protein (MBP). The following primers were used: EcoRI F, 5′-GGACTGGAATTCGACAATATCGGCACCAGCGC-3′ (EcoRI site is underlined); and SalI R, 5′-GGACTGGTCGACTTACCAGTGCACTCCAATGGTGAG-3′ (SalI site is underlined). The resulting plasmid (pMAL-c4xnfrAΔ1-27) was transformed into E. coli BL21 for expression [E. coli BL21(pMAL-c4xnfrAΔ1-27)]. Two liters of cells was grown to an OD₆₀₀ of 0.3, and the synthesis of recombinant MBP-NfrA was induced for 3 h with 0.3 mM IPTG at 4°C. Cells were pelleted at 4°C for 10 min at 4,200 × g and then resuspended in 16 ml sonication buffer (20 mM Tris-HCl, pH 7.4, 20 mM NaCl, EDTA-free protease inhibitor). Cells were sonicated in 15-s pulses followed by 15-s ice incubations until lysis was observed. Cellular debris was pelleted by centrifugation at 4°C for 30 min at 31,600 × g. The supernatant was loaded onto an amylose resin column (New England Biolabs, Ipswich, MA) and permitted to batch bind with rotation at 4°C for 4 h. After being washed with 6 bed volumes of low-salt buffer (20 mM Tris-HCl, pH 7.4, 20 mM NaCl), the recombinant protein was eluted from the resin with 10 mM maltose in low-salt buffer. Eluted protein was concentrated by centrifugation at 4°C, using Amicon Ultra 100 K centrifugal filtration columns (Millipore, Ireland).

In vitro interaction of purified recombinant gp65 and MBP-NfrA.

Purified N-terminally tagged gp65 and MBP-NfrA were preincubated for 30 min at room temperature with rotation prior to being loaded onto Talon resin. Resin binding occurred for 30 min at room temperature, with rotation, and was followed by 9 bed volume washes with low-salt buffer (20 mM Tris-HCl, pH 8.0, 20 mM NaCl). Complexes were eluted with 100 mM imidazole in low-salt buffer and analyzed by immunoblotting using anti-His-G and anti-MBP antibodies (New England Biolabs, Ipswich, MA) after transfer to a Hybond ECL nitrocellulose membrane. Binding of MBP-paramyosin (New England Biolabs, Ipswich, MA) to N-terminally tagged gp65 was used as a negative control.

RESULTS

Identification of N4 phage carrying an amber mutation in Orf65.

Insight into virion components involved in N4 adsorption was provided by a comparison of the sequences of the bacteriophage N4 and NC10 genomes. Bacteriophage NC10 infects the bacterial pathogen Yersinia ruckeri, the causative agent of enteric red mouth disease in fish (26). Comparisons of the 72 N4 ORFs and the 77 NC10 ORFs indicated that 44 ORFs share 60 to 90% homology and are present in their respective genomes in similar arrangements (T. Welch, personal communication). NC10 orf69, which encodes an endo-N-acetylneuraminidase required for adsorption to Y. ruckeri (Welch, personal communication), shares no homology to ORFs in the corresponding region of the N4 genome, which includes ORFs 64, 65, and 66. ORFs 65 and 66 encode virion proteins gp65 and gp66 (5). Given the extensive homology shared between the N4 and NC10 genomes yet their different host ranges, we hypothesized that virion proteins gp65 and/or gp66 might be involved in N4 adsorption to E. coli.

To investigate the possible roles of N4 gp65 and gp66 in adsorption, an N4 amber mutant phage library was screened for phage containing a mutation in either Orf65 or Orf66. An HpaI restriction digest library of the N4 genome was used to define complementation groups. N4am229 was rescued by the HpaI E fragment, which encodes the N-terminal 136 residues of gp64, the entirety of gp65, and the C-terminal 136 residues of gp66 (20). PCR amplification of this region of the N4am229 genome, followed by sequence analysis, revealed that the amber mutation was located in Orf65, resulting from a G-to-T transversion at nucleotide 1297 and a consequent mutation of a glutamine codon (aa 433) to a stop codon (GAG → TAG).

Role of virion protein gp65 in N4 infection.

To our surprise, N4am229 infection of E. coli W3350 yielded progeny (N4gp65⁻ phage) that lacked only the gp65 virion protein and was not infectious. As expected, N4am229 infection of E. coli W3350 supF yielded infectious progeny (N4gp65⁺ phage) containing all N4 virion proteins (Fig. 1). N4gp65⁻ and N4gp65⁺ phage yields were similar, as estimated by OD₂₈₀/OD₂₆₀ measurements.

FIG. 1.

N4am229 infection of E. coli W3350 yields noninfectious progeny lacking only gp65. N4 phage harboring an amber mutation in Orf65 were identified from a collection of phage with mutations mapping to the N4 HpaI E fragment. Mutant phage were used to infect E. coli W3350 supF and E. coli W3350. Noninfectious virions were recovered from E. coli W3350 infection and then purified. Virion proteins were analyzed by 10% SDS-PAGE and silver staining.

To determine whether gp65 plays a role in adsorption to E. coli, DNA-radiolabeled gp65⁻ and gp65⁺ phage were isolated from standard phage infections of E. coli W3350 and E. coli W3350 supF, respectively, in the presence of [³H]thymidine. Adsorption assays were performed using E. coli W3350 cells overexpressing the NfrA and NfrB proteins [E. coli W3350(pNfrA/B)]. Adsorption of N4 phage to E. coli W3350(pNfrA/B) for 5, 7, or 10 min indicated that N4gp65⁺ phage, but not N4gp65⁻ phage, were able to adsorb to the host (Fig. 2A). Therefore, gp65 is essential for N4 adsorption to the host.

FIG. 2.

Characterization of virions lacking gp65. (A) N4 virions lacking gp65 cannot adsorb to the host. ³H-labeled N4gp65⁺ or N4gp65⁻ virions were incubated at 37°C for the indicated times with cells overexpressing NfrA and NfrB, and the amounts of radioactivity in the supernatant (open bars) and in the pellet (gray bars) were determined after centrifugation. (B) N4gp65⁻ phage does not compete for host receptors. Mixtures of N4gp65⁻ or N4gp65⁺ phage with N4 ³H-labeled N4gp65⁺ phage were prepared at increasing ratios of unlabeled to labeled phage. The phage mixtures were then added to E. coli W3350(pNfrA/B) and incubated for 7 min at 37°C. The amounts of radioactivity in the supernatant (open bars) and in the pellet (gray bars) were measured after centrifugation.

A competition assay between N4gp65⁺ and N4gp65⁻ phage was performed to corroborate the adsorption phenotype observed. Increasing concentrations of unlabeled N4gp65⁺ or N4gp65⁻ phage were mixed with ³H-labeled N4gp65⁺ phage prior to infection of E. coli W3350(pNfrA/B) at saturating levels of phage to host receptors (Fig. 2B). As the ratio of unlabeled N4gp65⁺ phage to labeled N4gp65⁺ phage increased, the adsorption of labeled N4gp65⁺ phage decreased. In contrast, the adsorption of ³H-labeled N4gp65⁺ phage was not significantly affected when phage were mixed with increasing amounts of unlabeled N4gp65⁻ phage, as evidenced by similar distributions of radioactivity in the supernatant and pellet. These results indicate that N4gp65⁻ phage do not adsorb efficiently to E. coli and therefore do not interfere with N4gp65⁺ phage adsorption.

Recombinant gp65 is incorporated into virions in vitro and in vivo.

In order to determine the localization of gp65 in virions, N4 Orf65 was cloned into the arabinose-inducible pBAD vector such that gp65 was hexahistidine tagged at its N terminus or hexahistidine and c-myc tagged at its C terminus. Both recombinant constructs were overexpressed in E. coli BL21 and subsequently purified using IMAC (Fig. 3A). N4gp65⁻ virions were then incubated at 37°C for 30 min with purified N-terminally tagged or C-terminally tagged gp65. The resultant virions were subjected to CsCl buoyant density gradient centrifugation. Progeny virion proteins were separated by SDS-PAGE and visualized by silver staining as well as immunoblotting, using antisera to the His-G or c-myc epitope (Fig. 3B). Both N-terminally tagged and C-terminally tagged gp65 proteins were able to complement N4gp65⁻ virions in vitro, indicating that gp65 localizes to the surface of the N4 virion and suggesting that it is added during the final step of N4 virion assembly.

FIG. 3.

Purified recombinant gp65 complements N4gp65⁻ virions in vitro. (A) Purification of gp65 protein. N4 Orf65 was cloned into the arabinose-inducible pBAD vector such that it was hexahistidine and c-myc tagged at the C terminus or hexahistidine tagged at the N terminus. gp65 was overexpressed in E. coli BL21 and purified using IMAC and ion-exchange chromatography. Recombinant N-terminally tagged and C-terminally tagged gp65 proteins were visualized after 10% SDS-PAGE by silver staining and immunodetection. (B) Purified recombinant gp65 is incorporated into N4gp65⁻ virions. Purified N-terminally tagged or C-terminally tagged gp65 was incubated with N4gp65⁻ virions for 30 min at 37°C. The resultant virions were purified through CsCl gradient centrifugation, followed by 10% SDS-PAGE, silver staining, and immunodetection.

Attempts to incorporate recombinant N-terminally tagged and C-terminally tagged gp65 in vivo provided insights into the gp65 requirements for incorporation into virions during assembly. Expression of either recombinant construct in the nonsuppressor strain, E. coli W3350, followed by infection with N4am229 yielded N4 progeny as expected. The resultant progeny was purified by CsCl centrifugation, subjected to SDS-PAGE, and analyzed by silver staining and immunoblotting (Fig. 4A). Both recombinant proteins were able to complement N4gp65⁻ virions in vivo. Two gp65 antibody-reacting polypeptides were detected in cells expressing C-terminally tagged gp65 and infected with N4am229: the full-length polypeptide and an N-terminal degradation product (Fig. 4A, C-tagged gp65, α-myc panel). However, the N4 progeny from this infection contained only the full-length C-terminally tagged gp65, suggesting that N-terminal gp65 sequences are required for its incorporation or for proper folding of an incorporation determinant. Several faint gp65 C-terminal degradation products were detected in N4am229-infected cells expressing the N-terminally tagged gp65 construct (Fig. 4A, N-tagged gp65, α-his-G panel). In contrast to the C-terminally tagged gp65 in vivo complemented virions, these virions contained full-length N-terminally tagged gp65 as well as C-terminal degradation products. This finding suggests that sequences at the C terminus are not necessary for incorporation of the protein into N4 virions and again supports the localization of an incorporation determinant at the N terminus.

FIG. 4.

Virion localization of gp65. (A) Recombinant gp65 complements N4gp65⁻ virions in vivo. E. coli W3350 expressing recombinant N-terminally tagged or C-terminally tagged gp65 was infected with N4am229. The resultant virions were purified and analyzed by 10% SDS-PAGE followed by silver staining and immunodetection. (B) Localization of gp65 in N4 virions. In vivo complemented virions containing C-terminally tagged gp65 were incubated overnight with primary antibody against the c-myc epitope. Secondary 5-nm-gold-conjugated antibody was added and incubated with the virions for 2 h, followed by fixation with 0.1% glutaraldehyde. Labeled phage were stained with 2% ammonium molybdenate and visualized by electron microscopy. Gold particles are visible at the distal part of the tails.

Localization of gp65 in virions.

To determine the location of gp65 in the N4 virion, we used in vivo complemented virions containing recombinant N-terminally tagged or C-terminally tagged gp65. To that end, primary antibodies were used to detect either the His-G or c-myc epitope at the N or C terminus of the recombinant gp65 proteins, respectively. Secondary gold-conjugated antibodies were then added to the virions incubated with antiserum to the His-G or c-myc epitope. Finally, treated virions were negatively stained with ammonium molybdenate and visualized by electron microscopy.

The resulting images of C-terminally tagged gp65-complemented N4 virions revealed that gp65 is located at the N4 tail (Fig. 4B). Notably, virions containing recombinant N-terminally tagged gp65 showed no labeling, suggesting that the N terminus of gp65 either is not exposed or interacts with another N4 virion protein, whereas the C terminus of gp65 is exposed and available to react with the antisera. No immunogold labeling was detected in N4gp65⁻ virions treated with the same antibody regimen or in N4 virions containing N-terminally tagged or C-terminally tagged gp65 when the primary antibody treatment was omitted.

The detection of gp65 at the N4 tail is in agreement with the in vitro complementation results indicating that gp65 is located externally on the N4 virion and added last during virion assembly. Unequivocal localization of gp65 in the N4 virion was established by comparing the three-dimensional structures of N4gp65⁺ and N4gp65⁻ virions, as determined by cryo-electron microscopy (5). Protein density differences between the two reconstructions revealed that gp65 is present in six copies, forming a tail sheath surrounding the N4 tail tube.

Purified gp65 interacts with E. coli NfrA.

NfrA is a large E. coli outer membrane protein of unknown function in E. coli. A predicted transmembrane helix is present at the NfrA N terminus, which is part of a signal peptide that directs NfrA to the inner membrane. The signal peptide (V9 to A27) is cleaved by the signal peptidase, which releases proteins sorted by the Sec machinery into the periplasm; the precursor and processed forms of NfrA have been detected in maxicells (15). The processed form of NfrA is then translocated to the outer membrane (15). The requirement of the E. coli outer membrane protein NfrA for N4 adsorption suggests that NfrA is the N4 receptor (15). Therefore, we surmised that gp65 must interact with NfrA. To investigate the interaction of gp65 with NfrA, we used three approaches. First, we tested the ability of E. coli strains—E. coli BL21, to which N4 does not adsorb, and E. coli BL21(pNfrA/B), expressing the E. coli W3350 outer membrane NfrA and inner membrane NfrB proteins—to pull down purified gp65. Both strains were grown to exponential growth phase and concentrated before the addition of purified N-terminally tagged gp65. Following incubation at 37°C for 20 min, cells were washed extensively with LB, pelleted, and analyzed by SDS-PAGE and immunoblotting (Fig. 5A). The reactivity to polyclonal antibodies against NfrA indicated that NfrA is indeed expressed in E. coli BL21(pNfrA/B) but not in E. coli BL21. Antibodies against the His-G epitope in N-terminally tagged gp65 revealed that E. coli BL21(pNfrA/B), but not E. coli BL21, was able to pull down gp65. These results suggest that purified gp65 interacts with the E. coli W3350 outer membrane receptor NfrA.

FIG. 5.

Purified recombinant gp65 interacts with NfrA in vivo and in vitro. (A) E. coli BL21 expressing the N4 outer membrane receptor NfrA is able to pull down purified gp65. E. coli BL21 and E. coli BL21(pNfrA/B), expressing the outer membrane receptor NfrA and the inner membrane protein NfrB, were incubated with purified gp65. Cells were then washed several times and analyzed by 10% SDS-PAGE and immunoblot detection. (B) Purified N-terminally tagged and C-terminally tagged gp65 proteins inhibit the ability of N4gp65⁺ phage to adsorb to the host. Purified N-terminally tagged or C-terminally tagged gp65 was preincubated with E. coli W3350 cells expressing the N4 outer membrane receptor NfrA and the inner membrane protein NfrB for 5 min at 37°C. The cells were then infected by ³H-labeled N4gp65⁺ phage for 7 min at 37°C. After centrifugation, the amounts of radioactivity in the supernatant (open bars) and in the pellet (gray bars) were measured. (C) In vitro interaction between MBP-NfrA and N-terminally tagged gp65. MBP-NfrA was mixed at a molar ratio of 2:1 or 10:1 with N-terminally tagged gp65 (L) at room temperature for 30 min. The mixture was then loaded onto a Talon column and incubated for 30 min with rotation. The flowthrough (FT), wash (W), and eluate (E) were collected and analyzed by 10% SDS-PAGE and immunoblotting, using primary antibodies against MBP (to detect MBP-NfrA or MBP-paramyosin) and His-G (to detect N-terminally His-G-tagged gp65). MBP-paramyosin was used as a negative control.

Second, we tested the ability of purified gp65 to block adsorption of ³H-labeled N4gp65⁺ phage to the host. Purified N-terminally tagged gp65 or C-terminally tagged gp65 was preincubated for 5 min at 37°C with E. coli W3350(pNfrA/B). ³H-labeled N4gp65⁺ virions were then added to the cells for 7 min, and their ability to adsorb to the host was determined. As the amount of purified N-terminally tagged or C-terminally tagged gp65 preincubated with the host increased, the ability of N4gp65⁺ phage to adsorb to the host decreased, as evidenced by increasing amounts of radioactivity remaining in the supernatant (Fig. 5B). These results indicate that gp65 effectively blocks the adsorption of N4 phage to the host, presumably by interacting with the N4 outer membrane receptor NfrA and occluding it from infecting phage.

In vitro pull-down assays of recombinant gp65 and recombinant NfrA provided unequivocal evidence of their interaction. NfrA lacking its N-terminal Sec pathway secretion signal was cloned into the IPTG-inducible pMAL-c4x vector, such that MBP was fused at its N terminus (MBP-NfrA). Purified MBP-NfrA and N-terminally tagged gp65, at 2:1 and 10:1 molar ratios, were incubated for 30 min at room temperature prior to binding to Talon resin. Resin binding occurred for 30 min at room temperature with rotation, followed by several low-salt washes; all input gp65 was retained in the column, and no proteins were detected in the low-salt buffer washes. Bound proteins were eluted with 100 mM imidazole. At a 2:1 ratio of MBP-NfrA to N-terminally tagged gp65, detection of MBP-NfrA was faint but evident, while at a 10:1 ratio, the presence of MBP-NfrA in the elution fraction was obvious (Fig. 5C). MBP-paramyosin, used as a control, did not coelute with N-terminally tagged gp65. The excess of MBP-NfrA relative to N-terminally tagged gp65 required to detect their interaction was not unexpected since the amount of active NfrA in the purified preparation was unknown. In addition, gp65 forms a hexamer surrounding the tail tube in the N4 virion, as visualized by cryo-electron microscopy reconstruction of wild-type N4 (5). It may be that an oligomeric form of gp65 is required for interaction with NfrA and that the Talon-bound recombinant gp65 protein is not in the optimal orientation and/or oligomeric form required for binding to NfrA.

DISCUSSION

Most often, the process of bacteriophage adsorption to the host involves the recognition of specific components on the bacterial outer surface by a protein(s) located on the distal end of the phage tail and/or a protein(s) comprising phage tail fibers or spikes. Bacteriophages are divided into the following three major families, based on virion tail morphology: Siphoviridae (long, noncontractile tails), Myoviridae (long, contractile tails), and Podoviridae (short, noncontractile tails), to which bacteriophage N4 belongs (1).

Phages lambda and T5, both belonging to the Siphoviridae family, recognize the E. coli host through distally located tail proteins. Lambda phage adsorption involves initial interactions between side tail fibers, located at the distal end of the long, noncontractile tail, and a prevalent surface component, most likely the OmpC porin protein (12). Irreversible interaction follows via the C-terminal portion of the central tail fiber protein, gpJ, and the E. coli outer membrane trimeric protein LamB maltoporin (3, 12, 23, 24). Similarly, T5 adsorption commences with reversible interactions between three L-shaped tail fibers and an abundant cell surface component, the E. coli lipopolysaccharide (LPS) (10). Irreversible adsorption involves the distally located tail protein pb5 and the E. coli iron-ferrichrome transporter FhuA (11, 17, 23).

Phages that use tail fiber proteins exclusively to recognize the E. coli surface include T4 and T7, both of which interact with the E. coli LPS (18, 22). In the case of the Myoviridae family phage T4, reversible adsorption to the host is characterized by the binding of at least three of the long tail fibers to the core region of the LPS or to the outer membrane protein OmpC (14). Irreversible adsorption then requires binding of the remaining long tail fibers and unfolding of the short tail fibers from underneath the T4 baseplate, located at the distal end of the tail tube. The extended short tail fibers bind to the LPS, bringing the baseplate closer to the cell surface. Concurrently, the baseplate changes from a hexagonal to an “open star” conformation, driving tail sheath contraction and subsequent protrusion and penetration of the tail tube through the cell envelope for genome delivery (13, 19, 25).

T7, a member of the Podoviridae family, is not equipped with a long tail able to traverse the entire cell envelope. It has been proposed that after successful irreversible adsorption to the LPS, proteins within the T7 phage capsid, including a peptidoglycan hydrolase, are extruded through the head-tail connector and tail into the cell envelope, where they form a channel. Presumably, the genome would pass through the channel protected from DNA-hydrolyzing proteins in the periplasm (22).

In the present study, we elucidated the mechanism of adsorption for the Podoviridae family coliphage N4. In contrast to the case for other phages, early steps of N4 infection must compensate not only for injection obstacles imparted by having a short tail but also for the requirement of vRNAP injection. The results presented indicate that (i) the N4 virion protein gp65 is essential for phage adsorption to the host and (ii) N4 gp65 interacts directly with NfrA, an E. coli outer membrane protein required for N4 adsorption. Cryo-electron microscopy and three-dimensional reconstruction revealed that the virion protein gp65 constitutes a tail sheath surrounding the N4 tail tube (5). Specifically, gp65 forms elongated ribbons which are 265 Å in length, 230 Å wide at the neck, and 75 Å wide at the distal end of the tail, is present in six copies, and makes two contacts with the tail tube (5). Moreover, gp65 is the first tail sheath protein in the Podoviridae family of viruses found to be involved in adsorption to the host (5). gp65 shares no extensive homology to any proteins with defined function but does have an N-terminal region with partial homology over ∼250 residues to hypothetical bacterial proteins (∼700 aa in length) found in the genomes of several strains, including Erythrobacter (Erythrobacter litoralis HTCC2594 [GenBank accession no. YP_459684], 6e−09), Acinetobacter (Acinetobacter baumannii [YP_001713743], 3e−04), Novosphingobium aromaticivorans DMS 12444 (YP_498398, 2e−06), Sphingopyxis alaskensis RB2256 (YP_617036, 1e−04), and Sphingomonas sp. strain SKA58 (ZP_01303909, 7e−04). It is worth noting that N4 Orf64 (417 aa), present downstream of Orf65 and absent from the NC10 genome, shares homology with the same hypothetical proteins (∼300 residues; e−40 to e−30) at their C-terminal halves.

Interaction between phage tail components and the bacterial surface during adsorption to the host sets up and immediately precedes genome injection. Adsorption is the critical signal transduction event informing an infecting phage that it has recognized and adsorbed to the proper host and can proceed to DNA injection. We speculate that direct interaction of NfrA with gp65 is responsible for the initial signaling event leading to N4 DNA and vRNAP injection. Our current model of N4 genome and vRNAP injection includes the following steps. First, in a manner similar to that of phages T4 and T7, nonspecific reversible adsorption to the host involves the interaction of some number of the 12 N4 appendages with the E. coli LPS or enterobacterial common antigen (15). Based on primary and secondary structures as well as stoichiometry, we have proposed that each appendage is composed of a trimer of gp66 (59 kDa) (5). Furthermore, Orf66 is present in the N4 genome in a region that shares no homology with the corresponding region in the NC10 genome, which contains the NC10 host specificity determinant, gp69. Recognition of the host through these noncovalent interactions would allow N4 to “tickle” around the outer membrane until gp65 encounters and interacts with NfrA irreversibly, possibly via a tetratricopeptide (TPR) domain present in NfrA (determined by BLAST search). TPR motifs are involved in many protein-protein interactions, including chaperone, protein oligomerization, cell cycle, transcription, and protein transport complexes (4, 6). This strong, irreversible interaction would transmit a signal eliciting conformational changes in the tail, including release of the tail plug at the distal end of the tail as well as a peptidoglycan-hydrolyzing protein, which may be the plug itself (5). Moak and Molineux, using zymogram analysis, have shown that many bacteriophage virions contain peptidoglycan-hydrolyzing proteins (21). A protein with peptidoglycan-hydrolyzing activity was detected in N4 virions; however, its apparent size did not correlate with the known size of any of the N4 virion proteins. This may indicate that a larger N4 virion protein may undergo processing in order to activate a dormant peptidoglycan-hydrolyzing activity (21). NfrA may then interact with additional phage or host proteins, including the inner membrane protein NfrB, to create a channel traversing the periplasm, through which vRNAP and the N4 genome are injected. The N-terminal domain of vRNAP is required for the injection of the first 500 bp of the genome (Demidenko and Rothman-Denes, unpublished data); moreover, vRNAP remains associated with the bacterial inner membrane upon injection (7, 8). Whether this vRNAP domain is involved in channel formation remains to be determined. Finally, our ability to purify active NfrA provides a tool for investigating conformational changes in virion structure upon N4-NfrA interaction.