Manipulating Interactions between T4 Phage Long Tail Fibers and Escherichia coli Receptors

ABSTRACT

Bacteriophages are the most abundant and diverse biological entities on Earth. Phages exhibit strict host specificity that is largely conferred by adsorption. However, the mechanism underlying this phage host specificity remains poorly understood. In this study, we examined the interaction between outer membrane protein C (OmpC), one of the Escherichia coli receptors, and the long tail fibers of bacteriophage T4. T4 phage uses OmpC of the K-12 strain, but not of the O157 strain, for adsorption, even though OmpCs from the two E. coli strains share 94% homology. We identified amino acids P177 and F182 in loop 4 of the K-12 OmpC as essential for T4 phage adsorption in the copresence of loops 1 and 5. Analyses of phage mutants capable of adsorbing to OmpC mutants demonstrated that amino acids at positions 937 and 942 of the gp37 protein, which is present in the distal tip (DT) region of the T4 long tail fibers, play an important role in adsorption. Furthermore, we created a T4 phage mutant library with artificial modifications in the DT region and isolated and characterized multiple phage mutants capable of adsorbing to OmpC of the O157 strain or lipopolysaccharide of the K-12 strain. These results shed light on the mechanism underlying the phage host specificity mediated by gp37 and OmpC and may be useful in the development of phage therapy via artificial modifications of the DT region of T4 phage.

IMPORTANCE Understanding the host specificity of phages will lead to the development of phage therapy. The interaction between outer membrane protein C (OmpC), one of the Escherichia coli receptors, and the gp37 protein present in the distal tip (DT) region of the long tail fibers of T4 bacteriophages largely determines their host specificity. Here, we elucidated the amino acid residues important for the interaction between gp37 and OmpC. This result suggests that the shapes of both proteins at the binding interface play important roles in their interactions, which are likely mediated by multiple residues of both binding partners. Additionally, we successfully isolated multiple phage mutants capable of adsorbing to a variety of E. coli receptors using a mutant T4 phage library with artificial modifications in the DT region, providing a foundation for the alteration of the host specificity.

INTRODUCTION

Bacteriophages, or phages, are the most abundant and diverse biological entities on Earth (1). Phage therapy has emerged as an effective treatment for antibiotic-resistant bacteria that threaten hundreds of thousands of lives each year. In contrast to antibiotics that kill a wide range of bacterial species, phages kill only their specific targets. For example, the T4 phage, one of the most well-characterized phages, infects only limited strains of Escherichia and Shigella (2, 3). This limited host range is primarily determined by adsorption, the first step in the phage life cycle (4).

Two different tail fibers, long and short, mediate T4 phage adsorption. First, at least three of the six long tail fibers reversibly bind to receptors on the surface of Escherichia coli cells. Next, the short tail fibers irreversibly bind to the receptors, which convert the baseplate from the hexagonal form into the star-shaped form necessary for injection of the phage DNA (5–8). The first step in this process plays an important role in determining the host range of T4 phages. The Krisch group identified several T4 phage mutants that change host specificity and determined that these mutations were located in gene 37, which encodes the distal tip (DT) region of the long tail fibers (9). Moreover, mutational analysis studies by our group and the Islam group demonstrated that the DT region is responsible for receptor binding (10, 11). The crystal structure of the C-terminal fragment of gp37 strongly supports this mechanism (12). Therefore, modification of the DT region can potentially alter the specificity of T4 phage adsorption, which may enable the development of phage therapy against pathogenic bacteria, including enterohemorrhagic E. coli O157.

Lipopolysaccharide (LPS) and outer membrane protein C (OmpC) are defined as receptors for T4 phage adsorption (13, 14). LPS is a component of the outer leaflet of the outer membrane of Gram-negative bacteria and comprises three regions: lipid A, inner core, and outer core. LPS structures vary depending on E. coli strains, and T4 phages utilize a specific LPS for entry into the hosts (10). The T4 phage recognizes LPS of E. coli strain B, which has two terminal glucose molecules in the outer core, but not the LPS of K-12 or O157 E. coli strains, which have additional sugar residues. The T4 phage is, however, able to adsorb to LPS mutants of the K-12 and O157 strains, which harbor a terminal glucose in the outer core (10), and it is possible that the DT region of the T4 phage interacts with the terminal glucose in the outer core.

The T4 phage also uses OmpC as another receptor because it cannot adsorb to a K-12 mutant lacking ompC (10). OmpC exists as a trimer in the outer membrane and forms a porin that mediates nonspecific diffusion of small molecules such as sugars, ions, and amino acids (15). Each OmpC monomer has a pore formed by 16-stranded antiparallel β-barrels, and seven of the β-strands are connected to the loops at the extracellular side (16). The T4 phage presumably interacts with the extracellular loops, but this has not been proven experimentally.

Herein, we identified the amino acids and loops of K-12 OmpC required for the adsorption of T4 phage using mutational analysis. We also isolated phage mutants capable of adsorbing to altered OmpC and identified amino acids in the DT region that are necessary for adsorption to OmpC. Furthermore, we introduced artificial modifications in the DT region and isolated and characterized T4 phage mutants that could adsorb to O157 OmpC or K-12 LPS. Our results lead to the understanding of the interaction between T4 phage tail fibers and E. coli receptors.

RESULTS

T4 phage can use OmpC_K12, but not OmpC_O157, for adsorption.

The amino acid sequences of OmpC partially differ between E. coli strains K-12 and O157. We examined whether OmpC from both strains could be used for T4 phage adsorption. OmpC from K-12 (OmpC_K12) or O157 (OmpC_O157) was expressed from a plasmid in an E. coli K-12 ompC deletion (ΔompC) mutant. The membrane fraction of each E. coli was extracted, and OmpC was detected using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Coomassie brilliant blue (CBB) staining showed that both OmpC_K12 and OmpC_O157 were present in the membrane fraction to the same extent as the endogenous one (see Fig. S1A in the supplemental material). We carried out spot test analyses to test the growth ability of T4 phage on ΔompC cells expressing OmpC_K12 or OmpC_O157 (Fig. 1A). Although T4 phage grew normally on ΔompC cells expressing OmpC_K12 at an efficiency of plating (EOP) similar to that on the wild-type K-12 strain, TY0807, no plaque was observed on ΔompC cells expressing OmpC_O157, even when 10⁵ PFU of phages was applied as spots. Next, the rate of adsorption was measured using the plaque-forming assay. As shown in Fig. 1B, T4 phage adsorbed efficiently to ΔompC cells expressing OmpC_K12, although its efficiency was slightly lower than that of TY0807. In contrast, T4 phage hardly adsorbed to ΔompC cells expressing OmpC_O157. These results indicate that the T4 phage cannot use OmpC_O157 for adsorption.

FIG 1.

Growth and adsorption of T4 phage on cells expressing OmpC_K12 or OmpC_O157. (A) The suspension containing the number of T4 phages indicated on the left was applied as spots onto an LB plate containing 0.2% l-Ara seeded with TY0807, O157:H7, TY0807 ΔompC, and TY0807 ΔompC harboring pBAD33-OmpC_K12 or pBAD33-OmpC_O157, and the plates were incubated overnight at 37°C. “–” indicates no plasmid. TY0807 was used as a wild-type Escherichia coli K-12 strain. (B) Adsorption analyses were performed as described in Materials and Methods. Symbols: ○, TY0807; □, O157:H7; ●, TY0807 ΔompC; ▴, TY0807 ΔompC harboring pBAD33-OmpC_K12; ◆, TY0807 ΔompC harboring pBAD33-OmpC_O157. The number of unadsorbed phages was determined by plating with TY0807 as an indicator.

Proline at position 177 and phenylalanine at position 182 in OmpC_K12 are required for the adsorption of T4 phage.

OmpC_K12 and OmpC_O157 share 94% amino acid sequence homology. There are a few nonconserved amino acids, including two amino acids in loop 1, four in loop 4, two in the transmembrane region between loops 4 and 5, four in loop 5, and one in loop 7. Additionally, there is an insertion of four amino acids (GFTS) specific to the K-12 strain in loop 4 as well as an insertion of four amino acids (GVIN) specific to the O157 strain in loop 7 (Fig. 2A). These differences in the amino acid sequence may determine whether the T4 phage can adsorb to OmpC. We replaced each OmpC_K12 amino acid with the corresponding OmpC_O157 amino acid, and each set of GFTS residues specifically inserted into OmpC_K12 was replaced with alanine. The growth of T4 phage on ΔompC cells expressing the OmpC mutants was examined (Fig. 2B). T4 phages formed plaques on cells expressing the OmpC mutants, except for OmpC_K12(P177V) and OmpC_K12(F182A), with similar EOP to the cells expressing the wild-type OmpC_K12. Furthermore, T4 phage could not adsorb to ΔompC cells expressing OmpC_K12(P177V) or OmpC_K12(F182A) (Fig. 2C). SDS-PAGE and CBB staining indicated that these OmpC mutants were abundantly present in the membrane fraction (Fig. S1B). These results demonstrate that P177 and F182 in OmpC_K12 are required for adsorption. Based on the crystal structure of OmpC, both amino acids are aligned vertically and are exposed toward the central part of the trimeric OmpC (Fig. 2D).

FIG 2.

Growth and adsorption of T4 phage on cells expressing OmpC mutants. (A) Alignment of amino acid sequence in OmpC from Escherichia coli strains K-12 and O157. The lines or the asterisks indicate amino acids forming extracellular loop structures in OmpC_K12 or different amino acids between OmpC_K12 and OmpC_O157. The black boxes represent a 4-amino-acid insertion specific to OmpC_K12 or OmpC_O157. (B) The suspension containing the number of T4 phages indicated on the left was applied as spots onto an LB plate containing 0.2% l-Ara seeded with TY0807 ΔompC harboring each plasmid expressing the OmpC mutants indicated on the top, and the plates were incubated overnight at 37°C. “–” indicates no plasmid. TY0807 was used as a wild-type E. coli K-12 strain. (C) Adsorption analyses were performed as described in Materials and Methods. Symbols: ○, TY0807 ΔompC harboring pBAD33; ●, TY0807 ΔompC harboring pBAD33-OmpC_K12; ◇, TY0807 ΔompC harboring pBAD33-OmpC_K12(P177V); ◆, TY0807 ΔompC harboring pBAD33-OmpC_K12(F182A). The number of unadsorbed phages was determined by plating with TY0807 as an indicator. (D) OmpC viewed from the top (left) and the side (right) is shown (PDB ID 2J1N). Proline at position 177 (red), phenylalanine at position 182 (blue), and loops 1 (green), 4 (orange), and 5 (purple) are highlighted. Gray bars indicate a predicted membrane boundary.

Loops 1 and 5 in OmpC_K12 are also required for the adsorption of T4 phage.

Amino acids P177 and F182 are both located in loop 4 of OmpC_K12. We examined whether T4 phage could adsorb to chimeric OmpC_O157, whose loop 4 is replaced with that of OmpC_K12. Before this experiment, we constructed a plasmid expressing OmpC_O157ΔGVIN, in which residues GVIN in loop 7 were deleted. Spot test and adsorption analyses reported that these amino acids were not the cause of T4 phage’s inability to grow on and adsorb to cells expressing OmpC_O157 (Fig. 3B and C). After chimeric OmpC_{O157ΔGVIN+K12Loop4} was observed in the membrane fraction (Fig. S1C), T4 phage growth was examined (Fig. 3B). Unexpectedly, no plaque was formed when 10⁴ PFU of phage was applied as spots on cells expressing OmpC_{O157ΔGVIN+K12Loop4}. Also, adsorption to these cells was not observed even 15 min after phage addition (Fig. 3C). These results suggest that not only loop 4 of OmpC_K12 but also other loops are necessary for T4 phage growth and adsorption. Next, we expressed OmpC_{O157ΔGVIN+K12Loop4,5}, which harbors an additional replacement of loop 5 with the OmpC_K12 sequence (Fig. S1C). With cells expressing OmpC_{O157ΔGVIN+K12Loop4,5}, T4 phage was as efficient in plaque formation as wild-type OmpC_K12 (Fig. 3B). Approximately 30, 50, and 70% of the T4 phage had adsorbed to these cells at 5, 10, and 15 min after phage addition, respectively (Fig. 3C). However, the rate of adsorption was slightly slower than on cells expressing wild-type OmpC_K12 (40, 60, and 80% at 5, 10, and 15 min after phage addition, respectively). Since this slight delay suggested the necessity of an additional loop, we also examined the effect of loop 1, which is adjacent to loop 4, on the rate of adsorption. When chimeric OmpC_{O157ΔGVIN+K12Loop1,4,5}, which harbors loops 1, 4, and 5 of K-12, was expressed (Fig. S1D), T4 phages formed plaques on this cell with the same EOP as that on OmpC_K12 (Fig. 3B), approximately 40, 70, and 80% of the phages adsorbed to the cells expressing chimeric OmpC_{O157ΔGVIN+K12Loop1,4,5} at 5, 10, and 15 min, respectively, and this rate was similar to that of cells expressing wild-type OmpC_K12 (Fig. 3C). These results indicate that loops 1 and 5, as well as loop 4 in OmpC_K12, are required for efficient adsorption. Collectively, they indicate that the correct placement of these three loops is important for T4 phage adsorption (Fig. 2D).

FIG 3.

Growth and adsorption of T4 phage on cells expressing chimeric OmpC_O157 with OmpC_K12 loops. (A) Schematic diagram of chimeric OmpC mutants. The lines above OmpC_K12 indicate regions of extracellular loop structures in OmpC. (B) The suspension containing the number of T4 phages indicated on the left was applied as spots onto an LB plate containing 0.2% l-Ara seeded with TY0807, TY0807 ΔompC, or TY0807 ΔompC harboring pBAD33-ompC_K12, pBAD33-OmpC_O157, pBAD33-OmpC_O157ΔGVIN, pBAD33-OmpC_{O157ΔGVIN+K12Loop4}, pBAD33-OmpC_{O157ΔGVIN+K12Loop4,5}, or pBAD33-OmpC_{O157ΔGVIN+K12Loop1,4,5}, and the plates were incubated overnight at 37°C. “–” indicates no plasmid. TY0807 was used as a wild-type Escherichia coli K-12 strain. (C) Adsorption analyses were performed as described in Materials and Methods. Symbols: ○, TY0807; ●, TY0807 ΔompC; △, TY0807 ΔompC harboring pBAD33-OmpC_K12; ▴, TY0807 ΔompC harboring pBAD33-OmpC_O157; □, TY0807 ΔompC harboring pBAD33-OmpC_O157ΔGVIN; ■, TY0807 ΔompC harboring pBAD33-OmpC_{O157ΔGVIN+K12Loop4}; ◇, TY0807 ΔompC harboring pBAD33-OmpC_{O157ΔGVIN+K12Loop4,5}; ◆, TY0807 ΔompC harboring pBAD33-OmpC_{O157ΔGVIN+K12Loop1,4,5}. The number of unadsorbed phages was determined by plating with TY0807 as an indicator.

Isolation and characterization of T4 phage mutants that adsorb to OmpC with P177V or F182A mutation.

To identify amino acids in the DT region involved in binding to OmpC, we attempted to isolate T4 phage mutants capable of growing on ΔompC cells expressing either OmpC_K12(P177V) or OmpC_K12(F182A). Clear and turbid plaques were observed when 2 × 10¹¹ PFU of T4 phages were plated with ΔompC cells expressing either OmpC_K12(P177V) or OmpC_K12(F182A) (see Fig. S2 in the supplemental material). These spontaneous mutants were amplified from five independent clearer plaques, and the sequences around the DT region (DNA positions 2580 to 3078; amino acid positions 860 to 1026) were analyzed. Each of the five mutants shared common mutations: OmpC_K12(P177V) mutants had a single base substitution, A2810G in gene 37, resulting in the N937S mutation, and OmpC_K12(F182A) mutants had a base substitution, G2824C, resulting in the G942R mutation. Mutants carrying the N937S or G942R mutation did not grow on ΔompC cells (Fig. 4A and B), indicating that these phage mutants can use the respective OmpC mutant for adsorption. Although the EOP of the phage mutant carrying the G942R mutation on cells expressing OmpC_K12(F182A) was similar to that on wild-type K-12 (Fig. 4B), EOP of the phage mutant carrying the N937S mutation on cells expressing OmpC_K12(P177V) was reduced to 0.1 (Fig. 4A). Based on adsorption analyses, the phage mutant with G942R mutation adsorbed to cells expressing wild-type OmpC_K12 or OmpC_K12(F182A) at a similar rate (Fig. 4D). In contrast, the phage mutant with N937S mutation adsorbed to cells expressing OmpC_K12(P177V) at a lower rate than in cells expressing OmpC_K12 (Fig. 4C), which may have resulted in the decreased EOP. These mutant phages also adsorbed to cells expressing wild-type OmpC_K12. Therefore, these phage mutants do not necessarily require the P177V or F182A mutation in OmpC and likely use other amino acids for binding. These results demonstrate that amino acids at positions 937 and 942 in gp37 play important roles in binding to OmpC. The crystal structure of the DT region showed that N937 and G942 are located on the lateral surface of the DT region and are adjacent to each other (Fig. 4E).

FIG 4.

Characterization of T4 phage mutants capable of growing on cells expressing OmpC_K12(P177V) or OmpC_K12(F182A). (A) The suspensions containing the number of wild-type (WT) or mutant (M1) T4 phages indicated on the left were applied as spots onto an LB plate containing 0.2% l-Ara seeded with TY0807, TY0807 ΔompC, and TY0807 ΔompC harboring pBAD33-OmpC_K12 or pBAD33-OmpC_K12(P177V), as indicated on the bottom, and the plates were incubated overnight at 37°C. (B) The suspensions containing the number of WT or mutant (M2) T4 phages indicated on the left were applied as spots onto an LB plate containing 0.2% l-Ara seeded with TY0807, TY0807 ΔompC, and TY0807 ΔompC harboring pBAD33-OmpC_K12 or pBAD33-OmpC_K12(F182A), as indicated on the bottom, and the plates were incubated overnight at 37°C. (C) Adsorption analyses of T4 phage mutant (N937S) were performed as described in Materials and Methods. Symbols: ●, TY0807 ΔompC harboring pBAD33; ■, TY0807 ΔompC harboring pBAD33-OmpC_K12; ▴, TY0807 ΔompC harboring pBAD33-OmpC_K12(P177V). The number of unadsorbed phages was determined by plating with TY0807 as an indicator. (D) Adsorption analyses of T4 phage mutant (G942R) were performed as described in Materials and Methods. Symbols: ●, TY0807 ΔompC harboring pBAD33; ■, TY0807 ΔompC harboring pBAD33-OmpC_K12; ▴, TY0807 ΔompC harboring pBAD33-OmpC_K12(F182A). The number of unadsorbed phages was determined by plating with TY0807 as an indicator. (E) The DT region of T4 phage long tail fiber (gp37) viewed from the side (left) and the top (right) is shown (PDB ID 2XGF). Note that the top view is enlarged. N937 (red) and G942 (blue) are highlighted.

Isolation and characterization of T4 phage mutants that adsorb to OmpC_O157.

T4 phages cannot use OmpC_O157 for adsorption (Fig. 1). We attempted to isolate T4 phage mutants capable of growing on ΔompC cells expressing OmpC_O157 using the same method as before. However, no plaques were observed. Therefore, we artificially introduced a mutation in the DT region using error-prone PCR and constructed a library of DT mutant phages, from which we searched for phages that could grow on cells expressing OmpC_O157. When 2 × 10¹⁰ PFU of phage from the library was plated with E. coli K-12 ΔompC cells expressing OmpC_O157, 25 clear plaques were observed (Fig. 5A, right). When the library was prepared using regular PCR instead of error-prone PCR, two plaques occurred with the same input of PFU (Fig. 5A, left). However, these two phages did not use OmpC_O157 for adsorption because both were able to grow on ΔompC cells (see Fig. S3 in the supplemental material). Out of 25 plaques obtained from the phage mutant library, those capable of growing on ΔompC cells were omitted, leaving two clones with different plaque sizes to be analyzed. One clone, forming a small plaque, had two base substitutions, A2836G and T2839C, in gene 37, resulting in M946V and S947P amino acid mutations. The other clone, forming a normal plaque, had three base substitutions, G2819T, A2836G, and T2839C, resulting in G940V, M946V, and S947P amino acid mutations, respectively. Both phage mutants could grow on ΔompC cells expressing OmpC_O157, but not on TY0807 or ΔompC cells (Fig. 5B). Also, the phage mutant forming a normal plaque efficiently adsorbed to ΔompC cells expressing OmpC_O157, but not to TY0807 or ΔompC cells (Fig. 5C). We could not examine the adsorption of the M946V S947P phage mutant on ΔompC cells expressing OmpC_O157, because of the inability to generate a sufficient quantity of phage stock. These results indicate that both phage mutants can adsorb to OmpC_O157 but not to OmpC_K12. Although M946 and S947 are located inside the DT region, G940, like N937 and G942, is located on the lateral surface (Fig. 5D).

FIG 5.

Isolation and characterization of T4 phage mutants capable of growing on cells expressing OmpC_O157. (A) The phage library (2 × 10¹⁰ PFU) prepared with regular PCR (left plate) or error-prone PCR (right plate) was plated onto an LB plate containing 0.2% l-Ara seeded with TY0807 ΔompC harboring pBAD33-OmpC_O157 and incubated overnight at 37°C. (B) The suspensions containing the number of wild-type (WT) or mutant (M1 and M2) T4 phages shown (10⁵, 10³, or 10¹ PFU for wild type and M1 and 10³, 10², or 10¹ PFU for M2) were applied as spots onto an LB plate containing 0.2% l-Ara seeded with TY0807, TY0807 ΔompC, TY0807 ΔompC harboring pBAD33-OmpC_O157 or pBAD33-OmpC_O157ΔGVIN, O157:H7, or BB, as indicated on the bottom, and the plates were incubated overnight at 37°C. (C) Adsorption analyses of T4 phage mutant (G940V, M946V, and S947P) were performed as described in Materials and Methods. Symbols: ●, TY0807; ○, TY0807 ΔompC; ■, TY0807 ΔompC harboring pBAD33-OmpC_O157; □, O157:H7. The number of unadsorbed phages was determined by plating with TY0807 ΔompC harboring pBAD33-OmpC_O157 as an indicator. (D) Top views of the DT region of gp37 showing the surface (left) and sliced interiors (right) (PDB ID 2XGF). G940 (blue), M946 (pink), and S947 (red) are highlighted.

The M946V S947P phage mutant could not grow on ΔompC cells expressing OmpC_O157ΔGVIN (Fig. 5B), indicating that this mutant adsorbs to the four amino acids (GVIN) in loop 7 specifically inserted into OmpC_O157. Interestingly, the third amino acid substitution of G940V caused the phage to grow on ΔompC cells expressing OmpC_O157ΔGVIN (Fig. 5B), indicating that G940V likely binds to amino acids other than GVIN. Finally, plaque formation and the adsorption of these mutants to E. coli O157 were examined, but neither was observed (Fig. 5B and C). As O157 has a repeating O-antigen chain on the outside of the outer core, the longer O157 LPS may interfere with the interaction between OmpC_O157 and T4 phage tail fibers.

Isolation and characterization of T4 phage mutants that bind to LPS.

T4 phages cannot grow on ΔompC cells because they cannot use K-12 LPS for adsorption. We searched for phage mutants that could grow on ΔompC cells using LPS for adsorption. Phages at 2 × 10¹⁰ PFU from the library were plated with ΔompC cells, resulting in 11 clear plaques (Fig. 6A, right). Using regular PCR instead of error-prone PCR, only one plaque was identified (Fig. 6A, left). Sequence analyses of 10 phages were performed after the growth of phages obtained from the library was checked again on ΔompC cells. One phage mutant had two base substitutions, T2857C and A2858G, resulting in the Y953R mutation. This mutant was identical to Arl, which we previously isolated and characterized as a mutant utilizing K-12 LPS for adsorption (10). Another phage mutant had two base substitutions, T2854A and C2855A, replacing serine at position 952 with lysine. The remaining eight phages had a single base substitution at T2857C, replacing tyrosine at position 953 with histidine. To verify that two phage mutants with S952K or Y953H used LPS for adsorption, we employed two sugar transferase-deficient mutants. The ΔwaaG, mutant, which lacks the Glc I transferase gene (waaG), lacks the outer core. The ΔwaaC, mutant, which lacks the Hep I transferase gene (waaC), lacks three heptoses in the inner core (10). Although both phage mutants could grow on ΔompC cells (Fig. 6B), the phage mutant with S952K could not grow on ΔompC ΔwaaC and ΔompC ΔwaaG cells. These results highlight the necessity of an intact outer core for adsorption. The EOP of the phage mutant with Y953H on ΔompC ΔwaaC cells was reduced to approximately 0.1 compared with that of ΔompC cells, suggesting that this mutant phage also uses LPS for adsorption.

FIG 6.

Isolation and characterization of the T4 phage mutants capable of growing on ΔompC cells. (A) The phage library (2 × 10¹⁰ PFU) prepared with regular PCR (left plate) or error-prone PCR (right plate) was plated onto an LB plate seeded with TY0807 ΔompC cells and incubated overnight at 37°C. (B) The suspensions containing the different numbers (10¹, 10², 10³, and 10⁴) of wild-type (WT) or mutant (M1 and M2) T4 phages indicated on the top were applied as spots onto an LB plate seeded with BB, TY0807, TY0807 ΔompC, TY0726, or TY0728, and the plates were incubated overnight at 37°C. TY0807 or BB was used as a wild-type Escherichia coli K-12 strain or B strain, respectively. (C) Top views of the DT region of gp37 showing the surface (PDB ID 2XGF). S952 (blue) and Y953 (red) are highlighted.

DISCUSSION

Here, we demonstrated that T4 phage requires proline at position 177 and phenylalanine at position 182 located in loop 4 of OmpC_K12 for adsorption to host cells. We also demonstrated that loops 1, 4, and 5 in OmpC_K12 are required for efficient adsorption. Loop 1 is shorter than loops 4 and 5 and is overshadowed by loop 4 in the adjacent barrel (Fig. 2D). Thus, loop 1 seems unlikely to directly interact with gp37. However, an additional experiment supported that loop 1 is involved in phage adsorption (see Fig. S4 in the supplemental material). While no plaque was formed when 10⁹ PFU of T4 phages was plated with cells expressing OmpC_O157ΔGVIN, turbid plaques were formed, although the EOP was extremely low, when 10⁵ to 10⁹ PFU of phages was applied as spots on cells expressing OmpC_{O157ΔGVIN+K12Loop1}. We anticipate that loops 1 and 5 of OmpC_K12 will aid in placing P177 and F182 in the correct positions so that the long tail fibers can dock, although we cannot rule out the possibility that these loops interact directly with the long tail fibers.

A previous study (12) and our docking model (Fig. 7) suggest that the DT region of gp37 docks into the central part of the trimeric OmpC. This is also supported by the structural observation where the size of the DT region (∼25 Å in diameter) is approximately the same as the size of the cavity formed by the three OmpC subunits. In this study, we demonstrated that amino acids N937 and G942 play important roles in gp37 binding to OmpC. As both N937 and G942 are located at the predicted interface between gp37 and OmpC, it is possible that these two residues directly mediate binding between the two proteins. Indeed, P177 and F182 in OmpC also face the internal cavity of the binding site. However, direct interactions between N937 and P177 or G942 and F182 are unlikely, considering that their mutant pairs (N937S P177V and G942R F182A) do not seem to promote favorable chemical interactions. Nevertheless, these mutations dramatically alter the side-chain properties, which likely evoke local conformational changes. Perhaps the shapes of both proteins at the binding interface play important roles in docking, which is likely mediated by multiple residues of both binding partners. Notably, previous studies have indicated that multiple amino acids in gp37 are involved in binding to OmpC, many of which are located at the interface (10, 11).

FIG 7.

DT-OmpC docking model. The DT region of gp37 (blue [modified from PDB ID 2XGF]) was docked with OmpC (white [PDB ID 2J1N]) using the ClusPro server. Gray bars indicate the predicted membrane boundary.

OmpC is a transmembrane protein embedded in the outer membrane and can be divided into transmembrane, extracellular, and periplasmic regions (16). Trojet et al. (17) demonstrated that T4-like phages use the extracellular loops of OmpC for adsorption. Alignment of OmpC sequences of six E. coli strains (K-12, O157:H7, UMNK88, MS60-1, avian pathogenic E. coli [APEC] O1, and B7A) indicated multiple amino acid differences in loops 1, 2, 4, 5, and 7 of OmpC_K12, but the residues in loops 3, 6, and 8 and the transmembrane and periplasmic regions are relatively conserved. Through the evolution of E. coli, amino acid sequences of E. coli receptor proteins could have repeatedly mutated to escape phage infections. Thus, the mutation-rich loops likely constitute sites that many phages use for adsorption. Our results showing that loops 1, 4, and 5 are required for adsorption support this prediction. Moreover, the K-12-specific 4-amino-acid insertion in loop 4 is necessary for adsorption. Similarly, an O157-specific 4-amino-acid insertion in loop 7 of OmpC_O157 is required for the adsorption of PP01 phage that specifically infects O157 (Y. Otsuka, unpublished data). Thus, each phage tends to use structural features specific to each E. coli strain for adsorption, which is responsible for host specificity.

We isolated two phage mutants that utilized OmpC_O157 for adsorption: one forming a smaller plaque carries two mutations (M946V and S947P), and another forming a normal plaque carries the same two mutations along with an extra mutation (G940V). As M946 and S947 are located inside the DT region, these residues will be structurally important, and the significant change of the structure by M946V and S947P mutations may cause the adsorption to OmpC_O157. In contrast, G940 is located on the surface and may play a role similar to that of N937 or G942. Interestingly, both phage mutants were unable to adsorb to OmpC_K12 as a trade-off. This result indicates that G940, M946, and/or S947 is required for T4 phage adsorption to OmpC_K12. These phages also failed to grow on the E. coli B strain (Fig. 5B). Because the B strain lacks OmpC, the wild-type T4 phage uses LPS as a receptor for adsorption to the B strain. Therefore, G940, M946, and/or S947 is involved in adsorption to both OmpC and LPS. In support of these results, Islam et al. reported that G940A or S947A mutation renders the phage unable to adsorb to strains K-12 and B (11).

Three phage mutants capable of growing on ΔompC cells were isolated, one of which was the same as Arl, previously isolated as a phage capable of growing on ΔwaaR ΔompC cells (10). Arl(Y953R) can adsorb to the wild-type K-12 LPS and the two LPSs generated by the ΔwaaR mutant lacking Glc III transferase, an LPS with Glc II at the outer terminus, and an LPS with GlcNAc-Hep IV branched from Glc II. We also isolated a mutant phage (Y953H). These results indicate that Y953 plays an important role in binding to LPS. The third phage mutant (S952K) failed to grow on ΔompC ΔwaaG cells lacking the entire outer core. All mutant phages capable of adsorbing to K-12 LPS have a positive charged amino acid at 952 or 953. Glycan residues in the inner core are phosphorylated or modified with phosphate-containing groups (10). Therefore, a positively charged amino acid at 952 or 953 may directly or indirectly interact with negatively charged glycan residues in the inner core. In support of this idea, shortening of glycan in the inner core resulted in a decrease in the EOP of the phage mutant with Y953H. Also, amino acids at 952 and 953 are located on the top surface of the DT region (Fig. 6C). This result supports our hypothesis that the top surface of the DT region is involved in the adsorption to LPS (10).

MATERIALS AND METHODS

E. coli strains and phages.

The wild-type phage is T4D (18). The E. coli strains used in this study are listed in Table 1. TY0807 ΔompC::kan was constructed using GT7-mediated transduction of a kanamycin resistance marker from the JW2203 strain, which was provided by the National BioResource Project (NIG, Japan).

TABLE 1.

Escherichia coli strains used in this study

Strain	Genotype	Source or reference
TY0807	sup0 araD139 hsdR ΔlacX74 rpsL araD+	Koga et al., 2011 (26)
TY0807 ΔompC	TY0807 ΔompC::Kan	This study
BW25113	rrnB3 ΔlacZ4787 hsdR514 Δ(araBAD)567 Δ(rhaBAD)568 rph-1	NBRP-E. coli at NIG
JW2203	BW25113 ΔompC::Kan	NBRP-E. coli at NIG
TY0726	BW25113 ΔwaaG ΔompC::Kan	Washizaki et al., 2016 (10)
TY0728	BW25113 ΔwaaC::Cm ΔompC::Kan	Washizaki et al., 2016 (10)
BB	B strain, wild-type, sup0	Kai et al., 1999 (27)
O157:H7	Wild type (ATCC 43888)	Morita et al., 2002 (28)

Construction of plasmids.

The primers used in this study are listed in Table 2. To construct pBAD33-OmpC_O157, a DNA fragment containing ompC was amplified via PCR using E. coli O157:H7 DNA as the template and the primers YO-608 and YO-609, digested with KpnI and HindIII, and ligated into the corresponding pBAD33 sites (19). Plasmids expressing either OmpC_K12 or OmpC_O157 mutants were constructed using the KOD-Plus− mutagenesis kit (Toyobo, Japan). Primers and template DNAs used for mutagenesis are listed in Table 3. Mutations were demonstrated through sequencing using primer YO-82 or YO-155.

TABLE 2.

Oligonucleotides used in this study

Primer name	Sequence (5′→3′)
For plasmid construction
YO-233	GCCGTTTTTACCCTGGTACTGAACAG
YO-234	AGCCCATCTGGTGAAGGCTTTACTAGTG
YO-235	AACGTATCTGGTGAAGGCTTTACTAGTG
YO-236	GCCACTAGTAAAGCCTTCACCAG
YO-237	ATGACTAACAACGGTCGTGACGCAC
YO-238	ACGACCGTTGTTAGTTACGCCAC
YO-239	GAAGCACTGCGTCAAAACGGCGACGGC
YO-240	ATCAGTACGTTTGGAGCTGGAGATC
YO-241	GATCAGAACACCGCTGCTTACATCGGTAAC
YO-242	GTTCTGAGCATCAGTACGTTTGGAGC
YO-243	AGCGCTGCTTACATCGGTAACGGCGACCG
YO-244	ACCCCGGCTTACATCGGTAACGGCGACCG
YO-245	ACCGCTCTTTACATCGGTAACGGCGACCG
YO-257	GATAAAGATGTAGATGGCGACCAGACC
YO-258	GTCAGAGAAATAGTGCAGGCCGTCTAC
YO-259	AACAAATCTGTAGATGGCGACCAGACC
YO-260	GCTGCGATCTCCAGCTCCAAACGTACTG
YO-261	ACCGATACCGAAACCTTCGTAATCATAAG
YO-262	GGTGCGGTCTCCAGCTCCAAACGTACTG
YO-263	AACTACGACGACGAAGATATCCTGAAATAT
YO-264	ACGACCCAGGTTTTTACCTTTAGACTGCAG
YO-403	GCGAATTCCCTCGCGATTTAATAGCAGGCAAAG
YO-404	GCGTCGACTTCGCCGGGCTCATCGAAAGTTATAC
YO-409	ATCCACTAGCACAAATGGTGAGCAC
YO-410	CCACTACCAGAGTGAGTGTGTCCAC
YO-419	TAGTGCTGGCGACCATTCCCACTCTG
YO-420	CTAGTCCCAAAAGAGAAAGTGTGAC
YO-425	GTGGATCCACTAGCACAAATGGTGAG
YO-426	CTACTAGTCCCAAAAGAGAAAGTGTG
YO-608	CCGGTACCTAAAAAAGCAAATAAAGGCA
YO-609	CCAAGCTTTGTACGCTGAAAACAATG
YO-610	ACTAGTAAAAGCTTCACCAGATGGG
YO-611	GGCGTAACTAACAACGGTCG
YO-612	ACTAGTAGCGCCTTCACCAGATGGG
YO-613	ACTAGCAAAGCCTTCACCAGATGGG
YO-614	AGCAGTAAAGCCTTCACCAGATGGG
YO-615	CAGGTTTTTACCTTTAGACT
YO-616	GGTCGTAACTACGACGACGA
YO-617	ACTAGTAAAGCCTTCACCAGATGGGTTACCGTTTTTGCCCTGGTACT
YO-618	GGCGTAACTAACAACGGTCGTGACGCACTGCGTCAGAACGGCGA
YO-619	GGTGTTCTGAGCATCAGTACGTTTGGAGCTGGA
YO-620	GCTGCTTACATCGGTAACGGCGACCGTGCT
YO-621	ATCTTTGTTGTCAGAGAAATAGTGCAGGCCGTC
YO-622	GTAGATGGCGACCAGACCTA
YO-623	ATCTTTGTTGTCAGAGAAATAGTGCAGGCCGTC
YO-624	GTAGATGGCGACCAGACCTA
For sequencing
37-2-up	AAGTCCGCATATCCAAAGTTAGCTGTTGC (for gene 37)
37-dw	TATATTTTCATATTTAGAAGGGCCGAAGC (for gene 37)
YO-82	AGATTAGCGGATCCTACCTG (for ompC)
YO-155	GATTTAATCTGTATCAGGC (for ompC)

TABLE 3.

Plasmids, primers, and template plasmid DNA used for mutagenesis

Plasmid	Primers	Template plasmid DNA
pBAD33-OmpCK12(N47D)	YO-257 and YO-258	pBAD33-OmpCK12a
pBAD33-OmpCK12(D49S)	YO-258 and YO-259	pBAD33-OmpCK12
pBAD33-OmpCK12(N176S)	YO-233 and YO-234	pBAD33-OmpCK12
pBAD33-OmpCK12(P177V)	YO-233 and YO-235	pBAD33-OmpCK12
pBAD33-OmpCK12(G181A)	YO-610 and YO-611	pBAD33-OmpCK12
pBAD33-OmpCK12(F182A)	YO-611 and YO-612	pBAD33-OmpCK12
pBAD33-OmpCK12(T183A)	YO-611 and YO-613	pBAD33-OmpCK12
pBAD33-OmpCK12(S184A)	YO-611 and YO-614	pBAD33-OmpCK12
pBAD33-OmpCK12(V186M)	YO-236 and YO-237	pBAD33-OmpCK12
pBAD33-OmpCK12(D192E)	YO-238 and YO-239	pBAD33-OmpCK12
pBAD33-OmpCK12(G216A)	YO-260 and YO-261	pBAD33-OmpCK12
pBAD33-OmpCK12(I218V)	YO-261 and YO-262	pBAD33-OmpCK12
pBAD33-OmpCK12(A226D)	YO-240 and YO-241	pBAD33-OmpCK12
pBAD33-OmpCK12(T229S)	YO-242 and YO-243	pBAD33-OmpCK12
pBAD33-OmpCK12(A230P)	YO-242 and YO-244	pBAD33-OmpCK12
pBAD33-OmpCK12(A231L)	YO-242 and YO-245	pBAD33-OmpCK12
pBAD33-OmpCK12(G309N)	YO-263 and YO-264	pBAD33-OmpCK12
pBAD33-OmpCO157ΔGVIN	YO-615 and YO-616	pBAD33-OmpCO157
pBAD33-OmpCO157ΔGVIN+K12(Loop4)	YO-617 and YO-618	pBAD33-OmpCO157ΔGVIN
pBAD33-OmpCO157ΔGVIN+K12(Loop4,5)	YO-619 and YO-620	pBAD33-OmpCO157ΔGVIN+K12(Loop4)
pBAD33-OmpCO157ΔGVIN+K12(Loop1,4,5)	YO-621 and YO-622	pBAD33-OmpCO157ΔGVIN+K12(Loop4+5)
pBAD33-OmpCO157ΔGVIN+K12(Loop1)	YO-623 and YO-624	pBAD33-OmpCO157ΔGVIN
pBR322-g37C-g38N-BamHI	YO-409 and YO-410	pBR322-g37C-g38N
pBR322-g37C-g38N-BamHI/SpeI	YO-419 and YO-420	pBR322-g37C-g38N-BamHI

^aWashizaki et al., 2016 (10).

Bacteriophage adsorption assay.

E. coli cells were grown in Luria-Bertani (LB) medium supplemented with or without 0.2% l-arabinose (l-Ara) at 37°C to a density of 3 × 10⁸ cells ml⁻¹, and T4 phage was added at a multiplicity of infection (MOI) of 0.01. At appropriate times, an aliquot was withdrawn and diluted with BS buffer (8% NaCl and 10 mM potassium phosphate buffer [pH 7.4]) saturated with chloroform. The number of unadsorbed phages was determined by plating with appropriate E. coli as an indicator cell. The fraction of unadsorbed phages was calculated using the number of input phages set to 100%. The adsorption assay was performed in triplicate, and data points represent the mean ± standard deviation (SD) from triplicate measurements.

Detection of OmpC in the membrane fraction.

E. coli cells were grown in 5 ml LB medium supplemented with 30 μg ml⁻¹ chloramphenicol and 0.2% l-Ara until the optical density at 600 nm (OD₆₀₀) reached approximately 0.5. Cells were harvested, washed once with phosphate-buffered saline, suspended in 1 ml of 10 mM Tris-HCl (pH 7.5), and lysed using sonication. The cell debris was removed by centrifugation at 2,300 × g for 5 min. The supernatant was again centrifuged at 20,000 × g for 1 h, and the pellet was suspended in 40 μl of 10 mM Tris-HCl (pH 7.5) and stored as a membrane fraction. Proteins in the membrane fraction were separated by SDS-PAGE using 12.5% gels and visualized using CBB staining.

Isolation of phage mutants that can grow on ΔompC cells expressing OmpC_P177V or OmpC_F182A.

The stock of T4 phages prepared with BB cells (2 × 10¹¹ PFU) was plated with TY0807 ΔompC cells harboring either pBAD33-OmpC_P177V or pBAD33-OmpC_F182A on LB agar plates containing 0.2% l-Ara and incubated overnight at 37°C. The five biggest and clearest plaques were selected, and these phages were again plated with TY0807 ΔompC cells harboring either pBAD33-OmpC_P177V or pBAD33-OmpC_F182A. Next, the single biggest and clearest plaque from each phage was used to prepare a high-titer stock after propagation on TY0807 cells in M9C medium (M9-glucose medium supplemented with 0.3% Casamino Acids, 1 μg ml⁻¹ thiamine, and 20 μg ml⁻¹ tryptophan). Sequence analysis for the DT regions was performed using two primers, 37-2-up and 37-dw.

Construction of T4 phage mutant library.

A 1,067-bp DNA fragment containing gene 37 and 38 segments (location on T4 phage genome at positions 158855 to 159921) was amplified using PCR. T4 genomic DNA was used as the template, and YO-403 and YO-404 as primers, digested with EcoRI and SalI, and ligated into the corresponding pBR322 sites to construct pBR322-g37C-g38N. pBR322-g37C-g38N-BamHI was generated using a KOD-Plus− mutagenesis kit with pBR322-g37C-g38N as the template and YO-409 and YO-410 as primers. In this plasmid, the GGTTCT sequence of gene 37 at positions 159302 to 159307 was changed (as underlined) to GGATCC, resulting in a BamHI site. Additionally, pBR322-g37C-g38N-BamHI/SpeI was generated using a KOD-Plus− mutagenesis kit with pBR322-g37C-g38N-BamHI as the template and YO-419 and YO-420 as primers. In this plasmid, the ACTAGC sequence of gene 37 at positions 159461 to 159466 was changed to ACTAGT, resulting in an SpeI site. These mutations are synonymous substitutions and therefore do not cause changes in the amino acids of gp37.

We first constructed a plasmid library containing one or more mutations in the DNA sequence corresponding to the DT region. Briefly, a DNA fragment was amplified by error-prone PCR using 10 ng of pBR322-g37C-g38N as the template and 2 μM each primer, YO-425 and YO-426, in 100 μl of PCR solution containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 8.9 mM MgCl₂, 0.5 mM MnCl₂, 0.2 mM dATP, 0.2 mM dGTP, 1 mM dCTP, 1 mM dTTP, and 5 U of Taq DNA polymerase (BioAcademia, Japan). A thermal cycle of 94°C for 1 min, 55°C for 1 min, and 72°C for 2 min was repeated 18 times. Amplified DNA was gel eluted, digested with BamHI and SpeI, and ligated into the corresponding pBR322-g37C-g38N-BamHI/SpeI sites. To establish if the mutation(s) occurred in the DT region, plasmids were purified from 11 independent colonies, and sequences were analyzed using two primers, 37-2-up and 37-dw. Of the 11 analyzed plasmids, three had a single mutation in the DT region, two had two different mutations, and the remaining plasmids had no mutation.

To generate a T4 phage mutant library containing the mutation(s) in the DT region of gene 37, the ligated DNAs were electroporated into 40 μl of TY0807 competent cells and incubated in 1 ml of LB medium for 1 h at 37°C. Then, 0.4 ml of cell culture was added to 40 ml of LB medium containing 50 μl ml⁻¹ of ampicillin, and transformants were grown until the OD₆₀₀ reached 0.3. T4 phage was added at an MOI of 0.01 and incubated overnight at 37°C. During phage amplification, homologous recombination occurred between the DT region of the phage genome and the DT region carrying the mutation(s) on the plasmid. Amplified phages were purified using polyethylene glycol (20) and stored in 1 ml of Dil2 buffer (0.01 M Tris-HCl [pH 7.5], 1 mM MgCl₂, 0.5% NaCl, and 0.001% gelatin). After titration with TY0807 cells as an indicator, approximately 1.0 × 10¹² PFU ml⁻¹ of the stock suspension was obtained and used as the phage mutant library.

Isolation of phage mutants that can grow on ΔompC cells expressing OmpC_O157.

The phage mutant library (2 × 10¹⁰ PFU) was plated with ΔompC cells harboring pBAD33-OmpC_O157 on an LB agar plate containing 0.2% l-Ara and incubated overnight at 37°C, and 25 clear plaques were observed. Phages recovered from plaques were transferred onto a plate seeded with ΔompC cells or ΔompC cells harboring pBAD33-OmpC_O157. After phages capable of growing on ΔompC cells were eliminated, two phages with different plaque sizes were isolated. Sequence analyses of the DT regions were performed using two primers, 37-2-up and 37-dw. A high-titer stock of the phage forming normal plaque was obtained, whereas that of another phage forming smaller plaque was not.

Isolation of phage mutants that can grow on ΔompC cells.

The phage mutant library (2 × 10¹⁰ PFU) was plated with ΔompC cells on an LB agar plate and incubated overnight at 37°C, and 11 clear plaques were observed. The phages were again plated with ΔompC cells to verify their growth, and sequence analyses of the DT regions were performed using two primers, 37-2-up and 37-dw. Three phage mutants were prepared as a high-titer stock after propagation in TY0807 cells in LB medium.

Protein-protein docking.

A Protein Data Bank (PDB) file of the DT region of gp37 was created by modifying the PDB file (PDB ID 2XGF) encoding a part of gp37 (amino acids 811 to 1026). This DT domain was docked with OmpC (PDB ID 2J1N) using the ClusPro server (21–25). To block interactions at the OmpC transmembrane region, 20 residues exposed to the membrane were used as repellants in the restriction mode. The model used was the top hit of 30 potential models.