P22 Bacteriophage (Molecular Biology)

The Salmonella typhimurium phage P22 is a double-stranded DNA phage of the Podoviridae family that has played a key role in the development of molecular biology. Studies with P22 have contributed to our understanding of genetic exchange, control of gene expression, and biological self-assembly. It has proven itself to be a valuable laboratory tool for genetic manipulation of its host, Salmonella. P22 is classified as a lambdoid phage because it shares life cycle similarities with the Escherichia coli lambda phage. Comparison of the similarities and differences between P22 and the other lambdoid phage has served to illuminate both the common themes and the range of diversity that nature employs in problem solving (1).

P22 was originally isolated as a lysogen of Salmonella, and it has played a key role in the development of bacterial genetics. In 1952, Zinder and Lederberg discovered the phenomenon of generalized transduction of Salmonella genes by P22 (2). This observation led to studies on the mechanism of the establishment of lysogeny, which contributed to our current understanding of repressors and anti-repressors. Studies of the assembly of P22 have been instrumental in framing the concepts that guide our thinking about control mechanisms in biological assembly (3). In addition, the well-defined and readily manipulable genetics of P22 have allowed it to lead the way in the development of genetic analysis of protein folding and assembly (3-6).


Bacteriophage P22 is a temperate virus, meaning that following infection two distinct life cycles are available to it: the lytic or the lysogenic life cycles. In the lytic life cycle, a series of viral genes are expressed that ultimately result in the production of several hundred progeny virions per cell. The progeny virions are released approximately one hour after infection, through lysis of the host cell (Fig. 1). In the alternative pathway, termed the lysogenic pathway, the phage chromosome integrates into the host chromosome, and expression of the phage structural genes is suppressed (see Lysogeny). The integrated phage is called a prophage, which is replicated and passed to the daughter cells during cell division. In response to an appropriate environmental signal, such as irradiation with ultraviolet light, the prophage can excise itself, enter the lytic cycle, and generate progeny phage. The decision between whether an infecting phage enters the lytic or lysogenic pathway depends on a number of factors, the most important of which seems to be the number of infecting virions, or multiplicity of infection.

Figure 1. Electron micrograph of a section through a bacteriophage P22-infected Salmonella typhimurium cell approximately 30 min after infection. The electron-dense progeny phage within the cell contains double-stranded DNA, whereas the infecting phage that remain attached to the cell at 1 and 8 o’clock do not.

Electron micrograph of a section through a bacteriophage P22-infected Salmonella typhimurium cell approximately 30 min after infection. The electron-dense progeny phage within the cell contains double-stranded DNA, whereas the infecting phage that remain attached to the cell at 1 and 8 o'clock do not.

1. Structure

The P22 virion consists of two distinct structural elements: the capsid and the tail. The dsDNA phage genome is contained within the capsid and is thereby protected from environmental insult. The capsid of the P22 virion is a T = 7 icosahedral symmetry lattice approximately 600 A in diameter, composed of ~420 identical copies of the 47-kDa gene product 5 (gp5) coat protein (7). The capsid itself is a very stable structure, resistant to elevated temperature and drying, as well as to nucleases and proteinases. Interestingly, this stability does not arise directly from the coat protein subunits, but rather from their intersubunit contacts (8). The coat protein subunits themselves, prior to polymerization into the capsid, are only modestly stable. The tail is composed of up to six trimers of the 72-kDa gp9 tailspike protein. The tailspike protein serves as the organelle for attachment of the virus to the host cell during infection and is itself also a very stable structure (9).

The 27*106-kDa linear dsDNA genome is contained within the capsid of the phage in a highly condensed liquid-crystalline state. The close packing of the phosphate groups in the condensed form requires charge neutralization, which is accomplished through the binding of Mg ions. Although the precise packing of the DNA within the capsid is not known, the B form of DNA structure appears to be retained (10).

The structure of the tailed vertex of the phage is surprisingly complex. The tailspike trimers are attached to the capsid through a connector region. The connector is composed of a macromolecular complex composed of 12 subunits of the gp1 portal protein, as well as the connector proteins, gp4, gp10, and gp26. The 83-kDa gp1 portal protein is arranged as a dodecameric complex, approximately 180 A in diameter with a central channel approximately 30 A in diameter (11). It is thought that the DNA enters and exits the head through the central channel of the portal complex. Portal complexes have been identified in nearly all the dsDNA-containing bacteriophage. Although there has been debate as to whether portal complexes are composed of 12 or 13 subunits, the existence of the portal complex at an icosahedral vertex poses an interesting structural dilemma: The portal complex exhibits 12- or 13-fold symmetry, but is located at a fivefold symmetrical vertex, resulting in a symmetry mismatch. This means that there cannot be a simple 1:1 interaction between portal and capsid subunits during assembly and function. Although the exact manner in which the portal protein interacts with the capsid is unknown, it has been suggested that such an arrangement would make it possible for the portal complex to rotate within the capsid during DNA translocation

The tailspike protein folds and assembles independently of the capsid, and then adds subsequently to the capsid to render the phage infectious. The unique biochemical properties of the tailspike protein, coupled with the well-defined genetics of the bacteriophage system, have made it an ideal subject for in vivo studies of protein folding. The tailspike folds in a multistep pathway, in which partially folded intermediates associate to form an SDS-sensitive protrimer (13). The protrimer then matures into a very stable SDS-insensitive trimer. The reason for this stability is that the central region of each subunit of the trimer is folded into a right-handed b-helix structure, and the C-terminal regions interdigitate to form b-sheet structures (14,15). The ability to identify readily and quantify the ratio of partially folded and fully folded proteins by the simple technique of SDS-PAGE analysis has allowed the isolation and characterization of a series of mutations that destabilize the partially folded intermediates. Because these mutations have no effect on the stability of the final fully folded trimer, these mutants have been termed temperature sensitive for folding (tsf) mutants. Analysis of these mutants has lent support to the idea that protein folding and assembly takes advantage of transient interactions between amino acid residues that do not play a critical role in the stability of the final structure. At restrictive temperatures, the tsf mutant proteins form aggregates, indicating that aggregation occurs through the interaction of partially folded protein molecules and that alterations in the lifetime of the protein folding intermediates can result in aggregation (6, 16-18).

2. The Infection Process

P22 initially infects the host Salmonella cell through the binding of the tailspike to the O-antigenic repeating units of the bacterial lipopolysaccharide. The interaction of the tailspike protein with this receptor defines the host range. Once bound, the tailspike slowly destroys the receptor by cleaving the au (1, 3)-0-glycosidic bond between rhamnose and galactose units. The virion carries up to six tailspikes, but only three are used for binding during infection (9, 19, 20). Once bound, the phage moves laterally across the cell surface through repeated cycles of binding and release, presumably until it locates a second receptor, at which point binding becomes irreversible (21).

The dsDNA is then injected into the host cell in a linear fashion (22). The phage-encoded pilot proteins gp7, gp16, and gp20 are required to ensure that the DNA crosses the plasma membrane in an infectious form. Although the mechanism of their action remains unknown, they translocate from the phage into the host cell during infection (23). Following infection, the dsDNA circularizes by homologous recombination between the terminal redundancy regions, yielding a circular dsDNA that represents a single copy of the phage genome (24).

3. The Lysogenic Pathway

In the lysogenic state, the bacterial chromosome carries an integrated copy of the bacteriophage chromosome, termed a prophage. In the prophage, the genes responsible for growth of the phage and subsequent death of the host are repressed through the action of a specific repressor system. This repression results in a phenomenon known as immunity. If a host cell carrying a lysogen is subsequently superinfected by a phage carrying the same repressor system as the prophage, it too will be unable to grow. P22 maintains the prophage in a quiescent state through the action of two repressor proteins (Fig. 2). In this regard, P22 differs from l phage, which has only a single repressor, cI. The two P22 repressor proteins are called c2 and mnt, and they give rise to two immunity systems. The immC region is similar to the immunity region of l. The P22 c2 repressor acts at Ol and Or to prevent transcription of the early genes. The unique immI region includes an antirepressor, ant, and its regulators mnt and arc. Mnt acts to repress the transcription of the ant antirepressor. If mnt is turned off, ant is synthesized. Ant binds to and inactivates the c2 repressor protein, with the result that the prophage enters the lytic pathway. The production of ant itself is regulated through the action of the arc gene. In arc mutants, high levels of ant are produced, and these levels block phage development. The arc and mnt proteins are two members of a small family of proteins that use anti-parallel b-sheet motifs rather than helix-turn-helix motifs to bind their operator DNA (25, 26) (see DNA-Binding Proteins).

Figure 2. Genetic map of bacteriophage P22. The genes are positioned above the line (not to scale), with their function indicated above. Promoter regions and the loci of repressor action are indicated below the line. The c2 repressor protein inhibits synthesis of the early genes, P1 and Pr. The anti-terminators 24 and 23 result in expression of the early and late genes, respectively. Notice that genes with related functions are clustered.

Genetic map of bacteriophage P22. The genes are positioned above the line (not to scale), with their function indicated above. Promoter regions and the loci of repressor action are indicated below the line. The c2 repressor protein inhibits synthesis of the early genes, P1 and Pr. The anti-terminators 24 and 23 result in expression of the early and late genes, respectively. Notice that genes with related functions are clustered.

The ant gene lies downstream of Plate and is transcribed during the lytic cycle as part of the late operon, but Ant protein is not synthesized. An additional promoter Psar lies within the ant gene and directs the synthesis of a 69-nucleotide long antisense RNA that binds to the ant messenger RNA and inhibits its translation (27, 28).

In order to form a stable lysogen, the phage genome has to integrate into the host; otherwise it would be diluted out during cell division. Integration occurs preferentially at a particular site, attB, in the host chromosome and requires the action of the phage-encoded integrase (Int) protein and the host-encoded integration host factor (IHF). The circular phage chromosome carries a similar site, termed attP (29). A single crossover event catalyzed by the Int protein results in integration. For entry into the lytic cycle, excision proceeds by the reverse mechanism but, to ensure that excision does not take place prematurely, requires the additional function of the xis gene product, which overlaps int in the genetic map (30).

Entry into the lysogenic pathway is controlled by the genes c1 and c3. These gene products stimulate the production of the c2 repressor protein immediately after infection. C1 binds to the Pre (promoter repressor establishment) promoter as a tetramer and stimulates c2 synthesis (31). High-level expression of c2 appears to be required for entry into the lysogenic pathway, and this is the probable reason that a high multiplicity of infection favors lysogeny.

In addition to the immunity conferred by the immI and immC repressor systems, the phage employs three other mechanisms to prevent superinfection. The first is a modification of the host O antigen, which inhibits the adsorption of P22. This function is encoded by the a1 gene (32). The genes sieA and sieB (superinfection exclusion) block superinfection. Expression of SieB appears to lead to abortive infection; while the superinfecting phage enters the cell, and early functions are normally expressed, late function expression is inhibited. The sieA protein is an 18-kDa protein that partitions into the membrane and acts to block DNA from crossing the periplasmic space and entering the host cell (33). It therefore excludes virion DNA from entering the cells, regardless of the identity of the DNA.

4. The Lytic Pathway

In the lytic pathway, P22 enters the Salmonella host cell and begins the process of reproduction. In P22, as in l, the proteins involved in the lytic cycle are expressed sequentially. Control of the timing of the production of mRNA transcripts is achieved through the use of anti-termination proteins that allow the RNA polymerase to ignore encoded termination signals. To take advantage of this, the P22 genome is organized so that these regulatory proteins can exert their influence on a large number of genes (34). The genes themselves are clustered into functional units; for example, all the genes involved in assembly are clustered into the late operon and are under the control of a single regulatory protein. Like lambda, the early genes of P22 are arranged into two operons, Pl and Pr, which flank the P22 c2 repressor gene. The genes in these operons encode proteins involved in DNA replication, recombination, integration, and the regulation of gene expression. Gene 24, which is transcribed off Pl, is similar in function to the lambda N gene. It functions to prevent the termination of transcription at genetically defined sites in the chromosome and allows for complete transcription of the early genes (35, 36). Gene 23, which is transcribed off Pr, is analogous to gene Q in lambda and allows transcription of the late genes (37). The late genes that are driven off the Plate promoter code for the proteins necessary for head and tail assembly.

5. Generalized Transduction

Transduction is a process by which bacterial DNA is carried from one cell to another by a phage particle. Generalized transducing particles contain DNA derived entirely from the host-cell chromosome. Bacteriophage P22 transducing particles originate when the host chromosome, rather than the phage chromosome, serves as the substrate for packaging. During a lytic infection, the phage chromosome is replicated as a concatamer, and the concatamers are resolved as the double-stranded DNA is packaged by a headful mechanism. The products of phage genes 2 and 3 are required for packaging and act as a complex. Genetic evidence suggests that gp3 is responsible for recognizing a unique site or region termed a pac site. Packaging proceeds in an ATP-dependent reaction until the head is full, at which point the gp2/gp3 complex cuts the DNA; packaging then continues into another empty prohead. This model is the "sequential packaging model." The P22 head can package 105% of a phage genome, accounting for the observed terminal redundancy and circular permutation of the genetic map (38-41).

Within the Salmonella chromosome, there are a minimum of five to six sites, called pseudo-pac sites, which are recognized by the gp2/gp3 complex as pac sites. The nonrandom distribution of pseudo-pac sites therefore leads to a nonrandom distribution of transduced markers (42). Mutants that map in gp3, termed HT mutants, have been isolated in which up to 50% of the phage heads carry host DNA. These mutants have proven extremely valuable for studying the genetics of Salmonella. The 44-kbp length of the packaged DNA corresponds to about 50 genes and has made possible linkage analysis and genetic mapping of the Salmonella chromosome.

6. Specialized Transduction

Specialized transducing particles carry both host and phage DNA linked in a continuous stretch in a single particle. There are two mechanisms by which specialized transducing particles can arise: from aberrant prophage excision or by insertion into the phage genome of host translocatable elements. P22 has more latitude in its ability to mediate specialized transduction than does bacteriophage l. Because P22 packaging does not use unique end sites, it can package extremely long stretches of DNA by segmenting them into multiple heads. If a host cell is multiply infected, an intact oversized genome can be generated by recombination.

7. Capsid Morphogenesis

During P22 morphogenesis, properly folded and associated tailspike trimers add to the dsDNA-containing head, to generate an infectious virion. However, the filled capsid is not the first structure that is assembled (Fig. 3). The gp5 coat protein subunits are first assembled into a structure termed a procapsid (37). Like the mature virion, the procapsid has T = 7 icosahedral symmetry, and the portal protein is present at a single vertex in the form of a dodecameric complex (43, 44). However, the procapsid does not contain DNA. The interior volume is filled with approximately 300 copies of the gp8 scaffolding protein. Studies with conditional lethal mutants have demonstrated that the presence of the scaffolding protein facilitates coat protein polymerization and ensures a proper structure. In the absence of scaffolding protein, a variety of morphologically aberrant polymers are formed, while assembly proceeds with high fidelity (45). The procapsid is approximately 10% smaller in diameter and is less stable than the mature head. During DNA packaging, the scaffolding protein exits, and the lattice expands. Three-dimensional single-particle reconstructions from cryoelectron microscopy of the procapsid and mature form revealed that there are ~25 A diameter holes situated at the center of the hexavalent capsomeres in the procapsid and that these holes are closed during expansion. No proteins are added to the procapsid to close the holes; instead, they are closed by domain movements within the coat protein subunits (46). These holes are likely candidates for the exit ports for the scaffolding protein during DNA packaging (43). The freed scaffolding protein is recycled to participate in further rounds of assembly (47). Thus, the scaffolding protein functions as an "assembly molecular chaperone" whose presence is transiently required to provide positional information. The use of scaffolding proteins is a common theme in the assembly of the lambdoid phages and is also required for the assembly of the herpesviruses.

Figure 3. The morphogenetic pathway of the bacteriophage P22 capsid. The first structure formed is a procapsid, which coat, scaffolding, and portal proteins. The "pilot" proteins are also incorporated at this stage. The concatameric DNA is \ with the exit of the scaffolding protein.

The morphogenetic pathway of the bacteriophage P22 capsid. The first structure formed is a procapsid, which coat, scaffolding, and portal proteins. The "pilot" proteins are also incorporated at this stage. The concatameric DNA is \ with the exit of the scaffolding protein.

The scaffolding protein of P22, as well as all well-characterized scaffolding proteins, is a highly a-helical molecule that forms oligomers in solution (48). In the case of P22, dimerization of the scaffolding protein plays a key role in assembly. In both P22 and the herpes virus, the region of interaction between the coat and scaffolding protein has been mapped to the C-terminal end of the molecule. In addition to its role in assembly, it has been suggested that the scaffolding protein may function to exclude the chance incorporation of cellular proteins during head assembly, because the presence of cellular proteins within the capsid would result in a decreased internal volume and preclude the encapsidation of a complete P22 genome. P22 scaffolding protein regulates its own levels of biosynthesis, which is decreased if there is a large free pool, but up-regulated if all the scaffolding protein is incorporated into procapsid (49-51). This post-transcriptional control might be a mechanism to insure that all procapsids contain a full complement of scaffolding protein, despite the fact that a full complement is not strictly required for assembly (52).

During morphogenesis, the dsDNA is packed into the procapsid, the scaffolding protein exits, and the capsid lattice expands. The exact sequence of these events is not known, but the expansion is an exothermic process, suggesting that the capsid is "spring loaded" (53). Interestingly, the portal protein is part of the gauge that determines when the head is full; mutants in the gp1 portal protein have been identified that result in packaging a piece of P22 DNA up to 5% larger than normal (54). Following DNA packaging, the portal vertex is closed by the addition of the protein products of genes 4, 10, and 26. These proteins, which add sequentially, serve to stabilize the DNA within the capsid and also to provide the site for tail attachment (55).

There are two genes that are essential for lysis, genes 13 and 19. Gene 13 encodes an 11-kDa protein whose function is to disrupt the cell membrane by forming pores and to allow the hydrolytic enzymes access to the cell wall. Mutations in gene 13 have been isolated that delay lysis for up to several hours, despite the fact that the phage-encoded lysozyme is being produced. Gene 19 encodes the P22 lysozyme, a 16-kDa monomeric protein that attacks and degrades the peptidoglycan layer of the cell wall, leading to cell lysis. The bacteriophage P22 lysozyme is very similar to the well-studied T4 lysozyme with respect to structure and function.

Procapsid-like particles of P22 can be assembled in vitro from purified proteins (52). This has made it possible to perform sophisticated physical chemical studies of virus assembly. These studies have revealed that capsid assembly is a nucleation-limited reaction (56), proceeds along a well-directed pathway (56), and can be inhibited by the binding of small molecules to the coat protein subunits (57).

Over the past 30 years, the bacteriophage P22 system has contributed to our understanding of gene expression, protein folding and assembly, and basic virology. It appears that the lessons that we can learn from these systems is limited only by the creativity of the investigator.

Next post:

Previous post: