HNH proteins are a widespread component of phage DNA packaging machines

HK97 gp74 Plays an Essential Role in Phage Head Morphogenesis.

To determine whether gp74 plays a role in phage assembly, we created an HK97 phage mutant bearing an amber nonsense mutation in gene 74 (HK97 74am). Although this mutant produced a titer of ∼10⁹ plaque-forming units (pfu) per milliliter when grown in a permissive E. coli strain, <100 pfu/mL were produced when it was grown in a nonpermissive strain. This demonstrates that gp74 is essential for phage replication. To determine whether any phage assembly intermediates were produced in the absence of gp74, an HK97 74am lysate was mixed with an HK97 13am lysate, which contains normal levels of DNA-filled heads, but no active tails (gene 13 encodes the essential tail assembly chaperone) (21). This complementation reaction produced functional phage particles at a concentration of 10⁹ pfu/mL. By contrast, a complementation reaction containing an HK97 74am lysate combined with an HK97 6am lysate, which contains normal tails with no functional DNA-filled heads (22), produced fewer than 100 pfu/mL. These complementation data show that HK97 74am phage produce normal active tails and no active heads, as might be expected if gp74 is required for terminase activity.

Electron micrographs of negatively stained HK97 74am lysates revealed an abundance of free tails and head-like structures, but no fully assembled phage particles. Judging by their small size, spherical shape, and thick walls, the head-like structures likely correspond to empty proheads (Fig. 1A) (23). For comparison, we engineered an HK97 mutant prophage where gene 2, which encodes the large terminase subunit, was replaced by a kanamycin resistance cassette (HK97 2^kan). As can be seen in Fig. 1B, the head-like structures present in lysates produced by induction of the HK97 2^kan lysogen, which are empty proheads resulting from the absence of DNA packaging, are identical to those seen in the HK97 74am lysate. The proheads present in the 74am and 2^kan lysates are distinctly different in appearance from the icosahedral DNA-filled heads seen in a 13am lysate (Fig. 1C) and the empty heads seen in a 6am lysate (Fig. 1D). gp13, the tail assembly chaperone of phage HK97, produces normal heads but no tails (24), and the 6am lysate contains empty heads that result from genome packaging and subsequent loss due to lack of the gp6 head–tail connector protein (22). They display dark internal staining and uniformly thin walls. Thus, the electron microscopy data support the conclusion that the absence of gp74 disrupts DNA packaging and results in the accumulation of empty proheads.

To ensure that the effects on phage assembly resulting from the 74am mutation were a result of the lack of gp74 and not another phenomenon, such as a polar effect on a downstream gene, we assessed the ability of gp74 expressed from a plasmid to rescue the HK97 74am phenotype using an in vivo complementation assay. Serial dilutions of HK97 74am phage particles produced in an amber suppressor E. coli strain were spotted onto a lawn of nonsuppressor cells expressing wild-type (WT) gp74 from a plasmid. This resulted in a phage titer of 5 × 10⁹ pfu/mL, similar to plating of the same lysate on an amber suppressor strain. These data confirm that the effects on phage viability of the 74am mutation were due solely to the absence of gp74.

Gp74 Stimulates in Vitro cos Cleavage and Interacts Directly with TerL.

To directly assess the role of gp74 on terminase function, we performed in vitro cos cleavage assays using purified gp74 and HK97 terminase (TerS + TerL). Cleavage of a linearized plasmid containing a cos site (Fig. 2A) was monitored by agarose gel electophoresis (Fig. 2B). When the reaction contained terminase alone at a concentration of 5 μM, there was a very small amount of cos cleavage activity observed. No DNA cleavage was noted when gp74 alone was incubated with the cos site substrate. When 0.2 μM gp74 was added to a reaction containing 5 μM terminase, there was a stimulation in cutting of the plasmid at the cos site, resulting in the production of two bands of the expected size. When the concentration of gp74 in the reaction was increased to 5 μM, the cos cleavage increased, illustrating a concentration dependence of the reaction.

As efficient cos cleavage activity was dependent on the presence of both terminase and gp74 in the reaction, we examined their ability to form a complex. A lysate containing His-tagged gp74 was mixed with lysates containing untagged TerL and TerS and was applied to a Ni- nitrilotriacetic acid (NTA) column. Following extensive washing, the products were eluted from the column and analyzed by SDS/PAGE. As can be seen in Fig. 2C, a band corresponding in size to TerL coeluted with the His-tagged gp74. This band was excised from the gel and we used mass spectrometry to positively identify it as TerL, with 62 unique peptides covering 56% of the TerL protein sequence identified. No other phage proteins were identified by more than three peptides and no band at this size was seen in a control experiment performed under the identical conditions in which a His-tagged version of the phage λ tail tube protein, gpV (25), was applied to the Ni-NTA column in place of gp74. Most of the other prominent background bands were observed in the eluted fractions of both gp74 and gpV. These experiments demonstrate that TerL and gp74 interact.

A Putative HNH Active-Site Histidine Is Required for gp74 Activity in Vitro and in Vivo.

As proteins containing an HNH motif have been found to bind to nucleic acids and possess endonuclease activity, we investigated the importance of the putative HNH motif found in gp74 for cos site cleavage. The HNH motif is ∼35 aa long and is characterized by the presence of two highly conserved histidines and one asparagine. Previously determined structures of HNH containing proteins revealed that a divalent metal ion is bound to the center of the HNH structure by interactions with the side chain of the second conserved His residue, which corresponds to His82 in gp74. Using site-directed mutagenesis we created a variant gp74 with an alanine substitution at position gp74^H82A and assessed its ability to cleave the cos site in vitro. The presence of the H82A mutation abrogated the stimulation of terminase activity of gp74 at both 0.2 and 5 μM concentrations (Fig. 2).

To assess the effects of the H82A mutation on phage propagation, we used an in vivo complementation assay. Serial dilutions of HK97 74am phage were top plated on nonsuppressor cells expressing gp74^H82A from a plasmid, and revealed a titer of 4 × 10⁴ pfu/mL, which is ∼10⁵-fold lower than WT. To ensure that the loss of activity of gp74^H82A was not a result of a protein-folding defect, we purified the protein via Ni-affinity chromatography and used circular dichroism (CD) spectroscopy to assess its structure and stability (Fig. 3A). The CD spectrum of WT gp74 displays minima at 222 and 208 nm, which is characteristic of a predominantly α-helical protein. The spectrum of gp74^H82A is identical to that of gp74^WT, illustrating that gp74^H82A is folded at 25 °C. To ensure that the stability of gp74^H82A is not affected under the conditions of the in vivo assay (37 °C), we carried out temperature denaturation experiments with WT gp74 and gp74^H82A. The thermal denaturation of gp74 is not reversible as the protein aggregates at high temperature, so the curves cannot be fitted to reveal thermodynamic parameters. However, the relative stability of the mutant with respect to WT protein can be assessed, and as can be seen in Fig. 3B, gp74^WT and gp74^H82A both possess a transition temperature of thermal denaturation of ∼53 °C.

Conservation of Terminase-Associated HNH Proteins in Long-Tailed Phages.

To determine whether the reliance of terminases on HNH proteins is a common phenomenon, we analyzed the complete genomes of 604 long-tailed phages (Siphoviridae and Myoviridae) present in the National Center for Biotechnology Information (NCBI) Reference Sequence (RefSeq) Database. Using HMMER (26) to search with the hidden Markov model (HMM) corresponding to the Pfam HNH family (PF01844), we detected 319 HNH proteins. We also identified the genes encoding the TerL subunits of 546 (90%) of these phages using six different HMMs from the Pfam and TIGRFAMs (27) databases that each correspond to distinct families of TerL proteins (Table S1). We were then able to determine that 135 genes encoding HNH proteins lay within 5 ORFs of genes encoding TerL subunits (Table S2), suggesting that HNH proteins are involved in the function of a considerable number of terminases. These terminase-associated HNH proteins were encoded in the genomes of a wide variety of both siphophages and myophages infecting Gram-negative and -positive bacteria. Importantly, we found that 90% of the HNH-associated TerL subunits were members of the Terminase_1 (PF03354) family even though only 28% of the total number of identified TerL subunits were members of this family. Furthermore, of the 154 genomes encoding Terminase_1 TerL subunits, 80% possessed an HNH-encoding gene within 5 ORFs of the TerL gene. An additional 15% of genomes encoding Terminase_1 TerL subunits encoded an HNH protein somewhere within their genomes. These results imply that the function of the Terminase_1 family of TerL subunits uniquely requires an HNH protein.

Because the terminase-associated HNH proteins appear to possess a unique function compared with other HNH proteins, we expected that they may conserve sequence features among them that are not seen in the greater family of HNH proteins. Accordingly, we aligned a diverse selection of terminase-associated HNH proteins (Fig. 4) and created an HMM from this alignment. Using this HMM to scan the RefSeq Database, we found that 95% of the 200 most significant hits were encoded by genes located in phages or prophages within 10 ORFs of a gene encoding a TerL subunit. By contrast, of the top 200 hits obtained searching with the Pfam HNH HMM, only 16% were encoded close to TerL genes. These results suggest that the terminase-associated HNH proteins constitute a distinct functional subfamily, which is consistent with their performing a unique biological role.

Plasmid Construction, Protein Purification, and Biophysical Characterization.

Genes 1, 2, and 74 from phage HK97 were PCR amplified and cloned into a modified pET expression plasmid with a Tobacco etch virus-cleavable N-terminal 6-His tag. The H82A mutation was introduced into the plasmid containing full-length gene 74 by the QuikChange (Stratagen) approach. The proteins were overexpressed in E. coli ER2566 and purified from crude extracts by using Ni-NTA affinity chromatography following the standard native purification protocol (Qiagen). As gp1 and gp2 had a tendency to precipitate, crude lysates of the proteins were mixed and copurified. They were eluted from the column and maintained in a high salt buffer [50 mM Tris (pH 7), 250 mM imidazole, 250 mM NaCl, 5 mM β-mercaptoethanol]. gp74 was dialyzed into the same buffer with no imidazole. For the CD studies gp74^WT and gp74^H82A were dialyzed into 20 mM sodium phosphate (pH 7.0), 150 mM NaCl, and 5 mM β-mercaptoethanol. The coelution experiments were performed by mixing crude lysates of His-tagged gp74 (or λ gpV as a negative control) with untagged TerL and TerS subunits in binding buffer [10 mM Tris (pH 7), 500 mM NaCl, 5 mM imidazole]. The cleared lysates were applied to a Ni-NTA column and washed 10 times with 10 column volumes of 10 mM Tris (pH 7), 500 mM NaCl, and 30 mM imidazole. The bound proteins were eluted with 10 mM Tris (pH 7), 250 mM NaCl, and 250 mM imidazole; and the samples were analyzed by SDS/PAGE.

Construction of the HK97 74am and 2^kan Phages.

HK97 lysogens, in which genes 74 and 2 were replaced by a kanamycin resistance cassette (HK97 74^kan or 2^kan), were generated as previously described (33). The HK97 74am phage was generated by induction of the 74^kan lysogen by mitomycin C in the presence of a plasmid carrying an amber nonsense mutation in gene 74 (pCDF-74am). The resulting phage lysate was plated on a lawn of E. coli suppressor cells (QD5003). Plaques resulted only from phages that recombined with the plasmid as gene 74 is essential. Single lysogens were streaked out from the center of a plaque, and the 74am substitution was confirmed by DNA sequencing.

Phage Complementation Assays.

Phage lysogens bearing amber mutations in genes 74, 13, and 6 were induced by mitomycin C from nonsuppressor E. coli 594 cells. The induction was allowed to proceed for 3 h, at which point the cells were lysed by the addition of chloroform and the lysates were cleared by centrifugation. The resulting lysates were mixed in equal proportion and incubated at 37 °C for 15 min to allow the reaction to proceed. Serial dilutions of these reactions were mixed with cells in top agar, pour-plated, incubated overnight at 37 °C, and the number of plaques produced were quantified. For the plasmid complementation assay, E. coli ER2566 cells transformed with the plasmid carrying either the WT or mutant 74 gene were suspended in molten 0.7% agar and top plated on LB/agar. Serial dilutions of a 74am lysate (from a suppressor strain) were spotted in 3-μL aliquots on top of the cells and incubated at 37 °C overnight. Complementation ability was quantified by observation of plaques at various phage dilutions.

Electron Microscopy.

Samples were prepared by the induction of lysogens of HK97 74am, 6am, 13am, or 2^kan from E. coli nonsuppressor strain 594. Subcultures of a saturated overnight culture were grown to an OD₆₀₀ of 0.6, induced with mitomycin C (1 μg/mL) at 37 °C for 3 h, and lysed with chloroform. The resulting phage lysate was concentrated approximately threefold using a centrifugal filter device, filtered through a 0.2-μm filter, and applied to the surface of a carbon film coated electron microscopy grid and stained with 2% (wt/vol) uranyl acetate. Grids were examined with a Hitachi H-7000 microscope.

In Vitro cos Cleavage Assay.

A pCDF-based plasmid (EMD Millipore) containing a region of the HK97 genome encompassing the cos site (pHK97cos) was linearized by NheI and used as the substrate. A 30-μL reaction containing 2 nM DNA substrate was incubated with the proteins in a buffer containing 20 mM Tris (pH 7.5), 5 mM MgCl₂, 2 mM spermidine, 4 mM putrescine, 5 mM β-mercaptoethanol, 0.5 mM EDTA, and 1.5 mM ATP. gp1 and gp2 were used at a concentration of 5 μM and gp74 (WT or H82A mutant) was used at 0.2 or 5 μM as indicated. The reaction was incubated at 37 °C for 1 h before inactivation by the addition of 0.3 mg/mL proteinase K and 0.6% SDS. The resulting cleavage products were visualized by agarose gel electrophoresis followed by staining with ethidium bromide.

Bioinformatic Studies.

To identify HNH protein encoding genes in close proximity to terminase genes we performed HMM-based searches against the NCBI RefSeq Database of complete phage genomes with HMMER3 (26). HNH proteins were detected using the HMM corresponding to the Pfam HNH family (PF01844). Genes encoding TerL were identified using the HMMs listed in Table S1. To generate an HMM corresponding only to terminase-associated HNH proteins, a diverse selection of these proteins was aligned using MAFFT (34) and HMMER3 was used to generate the HMM. The NCBI RefSeq Database of complete eubacterial genomes was queried with this HMM. To determine whether these identified genes are contained within a prophage, a set of HMMs from the Pfam (11) database that are specific for morphogenetic functions found in all long-tailed phages (TerL, portal, major capsid, tail tube, head–tail connector, tail terminator, tail assembly chaperone, and tape measure proteins) were used to query the same database. If the HNH hits were close to the TerL and multiple hits to the other morphogenetic protein HMMs then this was assumed to be a prophage.