Cloning of the bacteriophage T4 uvsX gene and purification and characterization of the T4 uvsX recombination protein

Bacteriophage T4 Escapes CRISPR Attack by Minihomology Recombination and Repair

Bacteria and bacteriophages (phages) have evolved potent defense and counterdefense mechanisms that allowed their survival and greatest abundance on Earth. CRISPR (clustered regularly interspaced short palindromic repeat)-Cas (CRISPR-associated) is a bacterial defense system that inactivates the invading phage genome by introducing double-strand breaks at targeted sequences. While the mechanisms of CRISPR defense have been extensively investigated, the counterdefense mechanisms employed by phages are poorly understood. Here, we report a novel counterdefense mechanism by which phage T4 restores the genomes broken by CRISPR cleavages. Catalyzed by the phage-encoded recombinase UvsX, this mechanism pairs very short stretches of sequence identity (minihomology sites), as few as 3 or 4 nucleotides in the flanking regions of the cleaved site, allowing replication, repair, and stitching of genomic fragments. Consequently, a series of deletions are created at the targeted site, making the progeny genomes completely resistant to CRISPR attack. Our results demonstrate that this is a general mechanism operating against both type II (Cas9) and type V (Cas12a) CRISPR-Cas systems. These studies uncovered a new type of counterdefense mechanism evolved by T4 phage where subtle functional tuning of preexisting DNA metabolism leads to profound impact on phage survival. IMPORTANCE Bacteriophages (phages) are viruses that infect bacteria and use them as replication factories to assemble progeny phages. Bacteria have evolved powerful defense mechanisms to destroy the invading phages by severing their genomes soon after entry into cells. We discovered a counterdefense mechanism evolved by phage T4 to stitch back the broken genomes and restore viral infection. In this process, a small amount of genetic material is deleted or another mutation is introduced, making the phage resistant to future bacterial attack. The mutant virus might also gain survival advantages against other restriction conditions or DNA damaging events. Thus, bacterial attack not only triggers counterdefenses but also provides opportunities to generate more fit phages. Such defense and counterdefense mechanisms over the millennia led to the extraordinary diversity and the greatest abundance of bacteriophages on Earth. Understanding these mechanisms will open new avenues for engineering recombinant phages for biomedical applications.

Herpes simplex virus 1 ICP8 mutant lacking annealing activity is deficient for viral DNA replication

Most DNA viruses that use recombination-dependent mechanisms to replicate their DNA encode a single-strand annealing protein (SSAP). The herpes simplex virus (HSV) single-strand DNA binding protein (SSB), ICP8, is the central player in all stages of DNA replication. ICP8 is a classical replicative SSB and interacts physically and/or functionally with the other viral replication proteins. Additionally, ICP8 can promote efficient annealing of complementary ssDNA and is thus considered to be a member of the SSAP family. The role of annealing during HSV infection has been difficult to assess in part, because it has not been possible to distinguish between the role of ICP8 as an SSAP from its role as a replicative SSB during viral replication. In this paper, we have characterized an ICP8 mutant, Q706A/F707A (QF), that lacks annealing activity but retains many other functions characteristic of replicative SSBs. Like WT ICP8, the QF mutant protein forms filaments in vitro, binds ssDNA cooperatively, and stimulates the activities of other replication proteins including the viral polymerase, helicase-primase complex, and the origin binding protein. Interestingly, the QF mutant does not complement an ICP8-null virus for viral growth, replication compartment formation, or DNA replication. Thus, we have been able to separate the activities of ICP8 as a replicative SSB from its annealing activity. Taken together, our data indicate that the annealing activity of ICP8 is essential for viral DNA replication in the context of infection and support the notion that HSV-1 uses recombination-dependent mechanisms during DNA replication.

Coordination and processing of DNA ends during double-strand break repair: the role of the bacteriophage T4 Mre11/Rad50 (MR) complex

The in vivo functions of the bacteriophage T4 Mre11/Rad50 (MR) complex (gp46/47) in double-strand-end processing, double-strand break repair, and recombination-dependent replication were investigated. The complex is essential for T4 growth, but we wanted to investigate the in vivo function during productive infections. We therefore generated a suppressed triple amber mutant in the Rad50 subunit to substantially reduce the level of complex and thereby reduce phage growth. Growth-limiting amounts of the complex caused a concordant decrease in phage genomic recombination-dependent replication. However, the efficiencies of double-strand break repair and of plasmid-based recombination-dependent replication remained relatively normal. Genetic analyses of linked markers indicated that double-strand ends were less protected from nuclease erosion in the depleted infection and also that end coordination during repair was compromised. We discuss models for why phage genomic recombination-dependent replication is more dependent on Mre11/Rad50 levels when compared to plasmid recombination-dependent replication. We also tested the importance of the conserved histidine residue in nuclease motif I of the T4 Mre11 protein. Substitution with multiple different amino acids (including serine) failed to support phage growth, completely blocked plasmid recombination-dependent replication, and led to the stabilization of double-strand ends. We also constructed and expressed an Mre11 mutant protein with the conserved histidine changed to serine. The mutant protein was found to be completely defective for nuclease activities, but retained the ability to bind the Rad50 subunit and double-stranded DNA. These results indicate that the nuclease activity of Mre11 is critical for phage growth and recombination-dependent replication during T4 infections.

Double-strand break repair and recombination-dependent replication of DNA in bacteriophage T4 in the absence of UvsX recombinase: replicative resolution pathway

The effects of mutations in bacteriophage T4 genes uvsX and 49 on the double-strand break (DSB)-promoted recombination were studied in crosses, in which DSBs were induced site-specifically within the rIIB gene by SegC endonuclease in the DNA of only one of the parents. Frequency of rII+ recombinants was measured in two-factor crosses of the type i×ets1 and in three-factor crosses of the type i×ets1 a6, where ets1 is an insertion in the rIIB gene carrying the cleavage site for SegC; i's are rIIB or rIIA point mutations located at various distances (12-2040 bp) from the ets1 site, and a6 is rIIA point mutation located at 2040 bp from ets1. The frequency/distance relationships were obtained in crosses of the wild-type phage and of the amber mutant S17 (gene uvsX) and the double mutant S17 E727 (genes uvsX and 49). These data provide information about the frequency and distance distribution of the single-exchange (splices) and double-exchange (patches) events. The extended variant of the splice/patch coupling (SPC) model of recombination, which includes transition to the replication resolution (RR) alternative is substantiated and used for interpretation of the frequency/distance relationships. We conclude that the uvsX mutant executes recombination-dependent replication but does it by a qualitatively different way. In the absence of UvsX function, the DSB repair runs largely through the RR subpathway because of inability of the mutant to form a Holliday junction. In the two-factor crosses, the double uvsX 49- is recombinationally more proficient than the single uvsX mutant (partial suppression of the uvsX deficiency), while the patch-related double exchanges are virtually eliminated in this background.

Biochemical characterization of bacteriophage T4 Mre11-Rad50 complex

The Mre11-Rad50 complex (MR) from bacteriophage T4 (gp46/47) is involved in the processing of DNA double-strand breaks. Here, we describe the activities of the T4 MR complex and its modulation by proteins involved in homologous recombination. T4 Mre11 is a Rad50- and Mn(2+)-dependent dsDNA exonuclease and ssDNA endonuclease. ATP hydrolysis is required for the removal of multiple nucleotides via dsDNA exonuclease activity but not for the removal of the first nucleotide or for ssDNA endonuclease activity, indicating ATP hydrolysis is only required for repetitive nucleotide removal. By itself, Rad50 is a relatively inefficient ATPase, but the presence of Mre11 and dsDNA increases ATP hydrolysis by 20-fold. The ATP hydrolysis reaction exhibits positive cooperativity with Hill coefficients ranging from 1.4 for Rad50 alone to 2.4 for the Rad50-Mre11-DNA complex. Kinetic assays suggest that approximately four nucleotides are removed per ATP hydrolyzed. Directionality assays indicate that the prevailing activity is a 3' to 5' dsDNA exonuclease, which is incompatible with the proposed role of MR in the production of 3' ssDNA ends. Interestingly, we found that in the presence of a recombination mediator protein (UvsY) and ssDNA-binding protein (gp32), Mre11 is capable of using Mg(2+) as a cofactor for its nuclease activity. Additionally, the Mg(2+)-dependent nuclease activity, activated by UvsY and gp32, results in the formation of endonuclease reaction products. These results suggest that gp32 and UvsY may alter divalent cation preference and facilitate the formation of a 3' ssDNA overhang, which is a necessary intermediate for recombination-mediated double-strand break repair.

Crystal structure of the phage T4 recombinase UvsX and its functional interaction with the T4 SF2 helicase UvsW

Bacteriophage T4 provides an important model system for studying the mechanism of homologous recombination. We have determined the crystal structure of the T4 UvsX recombinase, and the overall architecture and fold closely resemble those of RecA, including a highly conserved ATP binding site. Based on this new structure, we reanalyzed electron microscopy reconstructions of UvsX-DNA filaments and docked the UvsX crystal structure into two different filament forms: a compressed filament generated in the presence of ADP and an elongated filament generated in the presence of ATP and aluminum fluoride. In these reconstructions, the ATP binding site sits at the protomer interface, as in the RecA filament crystal structure. However, the environment of the ATP binding site is altered in the two filament reconstructions, suggesting that nucleotide cannot be as easily accommodated at the protomer interface of the compressed filament. Finally, we show that the phage helicase UvsW completes the UvsX-promoted strand-exchange reaction, allowing the generation of a simple nicked circular product rather than complex networks of partially exchanged substrates.

Role of allosteric switch residue histidine 195 in maintaining active-site asymmetry in presynaptic filaments of bacteriophage T4 UvsX recombinase

Recombinases of the highly conserved RecA/Rad51 family play central roles in homologous recombination and DNA double-stranded break repair. RecA/Rad51 enzymes form presynaptic filaments on single-stranded DNA (ssDNA) that are allosterically activated to catalyze ATPase and DNA strand-exchange reactions. Information is conveyed between DNA- and ATP-binding sites, in part, by a highly conserved glutamine residue (Gln194 in Escherichia coli RecA) that acts as an allosteric switch. The T4 UvsX protein is a divergent RecA ortholog and contains histidine (His195) in place of glutamine at the allosteric switch position. UvsX and RecA catalyze similar strand-exchange reactions, but differ in other properties. UvsX produces both ADP and AMP as products of its ssDNA-dependent ATPase activity--a property that is unique among characterized recombinases. Details of the kinetics of ssDNA-dependent ATP hydrolysis reactions indicate that UvsX-ssDNA presynaptic filaments are asymmetric and contain two classes of ATPase active sites: one that generates ADP, and another that generates AMP. Active-site asymmetry is reduced by mutations at the His195 position, since UvsX-H195Q and UvsX-H195A mutants both exhibit stronger ssDNA-dependent ATPase activity, with lower cooperativity and markedly higher ADP/AMP product ratios, than wild-type UvsX. Reduced active-site asymmetry correlates strongly with reduced ssDNA-binding affinity and DNA strand-exchange activity in both H195Q and H195A mutants. These and other results support a model in which allosteric switch residue His195 controls the formation of an asymmetric conformation of UvsX-ssDNA filaments that is active in DNA strand exchange. The implications of our findings for UvsX recombination functions, and for RecA functions in general, are discussed.

Mutations in a conserved motif inhibit single-stranded DNA binding and recombination mediator activities of bacteriophage T4 UvsY protein

The UvsY recombination mediator protein is critical for homologous recombination in bacteriophage T4. UvsY uses both protein-protein and protein-DNA interactions to mediate the assembly of the T4 UvsX recombinase onto single-stranded (ss) DNA, forming presynaptic filaments that initiate DNA strand exchange. UvsY helps UvsX compete with Gp32, the T4 ssDNA-binding protein, for binding sites on ssDNA, in part by destabilizing Gp32-ssDNA interactions, and in part by stabilizing UvsX-ssDNA interactions. The relative contributions of UvsY-ssDNA, UvsY-Gp32, UvsY-UvsX, and UvsY-UvsY interactions to these processes are only partially understood. The goal of this study was to isolate mutant forms of UvsY protein that are specifically defective in UvsY-ssDNA interactions, so that the contribution of this activity to recombination processes could be assessed independent of other factors. A conserved motif of UvsY found in other DNA-binding proteins was targeted for mutagenesis. Two missense mutants of UvsY were isolated in which ssDNA binding activity is compromised. These mutants retain self-association activity, and form stable associations with UvsX and Gp32 proteins in patterns similar to wild-type UvsY. Both mutants are partially, but not totally, defective in stimulating UvsX-catalyzed recombination functions including ssDNA-dependent ATP hydrolysis and DNA strand exchange. The data are consistent with a model in which UvsY plays bipartite roles in presynaptic filament assembly. Its protein-ssDNA interactions are suggested to moderate the destabilization of Gp32-ssDNA, whereas its protein-protein contacts induce a conformational change of the UvsX protein, giving UvsX a higher affinity for the ssDNA and allowing it to compete more effectively with Gp32 for binding sites.

Properties of bacteriophage T4 proteins deficient in replication repair

An epistasis group of mutations engendering increased sensitivity to diverse DNA-damaging agents was described previously in bacteriophage T4. These mutations are alleles of genes 32 and 41, which, respectively, encode a single-stranded DNA-binding protein (gp32) and the replicative DNA helicase (gp41). The mechanism by which the lethality of DNA damage is mitigated is unknown but seems not to involve the direct reversal of damage, excision repair, conventional recombination repair, or translesion synthesis. Here we explore the hypothesis that the mechanism involves a switch in DNA primer extension from the cognate template to an alternative template, the just-synthesized daughter strand of the other parental strand. The activities of the mutant proteins are reduced about 2-fold (for gp32) or 4-fold (for gp41) in replication complexes catalyzing coordinated synthesis of leading and lagging strands, in binding single-stranded DNA, promoting DNA annealing, and promoting branch migration. In striking contrast, the mutant proteins are strongly impaired in promoting template switching, thus supporting the hypothesis of survival by template switching.

Bacteriophage T4 genome

Phage T4 has provided countless contributions to the paradigms of genetics and biochemistry. Its complete genome sequence of 168,903 bp encodes about 300 gene products. T4 biology and its genomic sequence provide the best-understood model for modern functional genomics and proteomics. Variations on gene expression, including overlapping genes, internal translation initiation, spliced genes, translational bypassing, and RNA processing, alert us to the caveats of purely computational methods. The T4 transcriptional pattern reflects its dependence on the host RNA polymerase and the use of phage-encoded proteins that sequentially modify RNA polymerase; transcriptional activator proteins, a phage sigma factor, anti-sigma, and sigma decoy proteins also act to specify early, middle, and late promoter recognition. Posttranscriptional controls by T4 provide excellent systems for the study of RNA-dependent processes, particularly at the structural level. The redundancy of DNA replication and recombination systems of T4 reveals how phage and other genomes are stably replicated and repaired in different environments, providing insight into genome evolution and adaptations to new hosts and growth environments. Moreover, genomic sequence analysis has provided new insights into tail fiber variation, lysis, gene duplications, and membrane localization of proteins, while high-resolution structural determination of the "cell-puncturing device," combined with the three-dimensional image reconstruction of the baseplate, has revealed the mechanism of penetration during infection. Despite these advances, nearly 130 potential T4 genes remain uncharacterized. Current phage-sequencing initiatives are now revealing the similarities and differences among members of the T4 family, including those that infect bacteria other than Escherichia coli. T4 functional genomics will aid in the interpretation of these newly sequenced T4-related genomes and in broadening our understanding of the complex evolution and ecology of phages-the most abundant and among the most ancient biological entities on Earth.

Analysis of regions within the bacteriophage T4 AsiA protein involved in its binding to the sigma70 subunit of E. coli RNA polymerase and its role as a transcriptional inhibitor and co-activator

Bacteriophage T4 AsiA, a protein of 90 amino acid residues, binds to the sigma(70) subunit of Escherichia coli RNA polymerase and inhibits host or T4 early transcription or, together with the T4 MotA protein, activates T4 middle transcription. To investigate which regions within AsiA are involved in forming a complex with sigma(70) and in providing transcriptional functions we generated random mutations throughout AsiA and targeted mutations within the C-terminal region. We tested mutant proteins for their ability to complement the growth of T4 asiA am phage under non-suppressing conditions, to inhibit E. coli growth, to interact with sigma(70) region 4 in a two-hybrid assay, to bind to sigma(70) in a native protein gel, and to inhibit or activate transcription in vitro using a T4 middle promoter that is active with RNA polymerase alone, is inhibited by AsiA, and is activated by MotA/AsiA. We find that substitutions within the N-terminal half of AsiA, at amino acid residues V14, L18, and I40, rendered the protein defective for binding to sigma(70). These residues reside at the monomer-monomer interface in recent NMR structures of the AsiA dimer. In contrast, AsiA missing the C-terminal 44 amino acid residues interacted well with sigma(70) region 4 in the two-hybrid assay, and AsiA missing the C-terminal 17 amino acid residues (Delta74-90) bound to sigma(70) and was fully competent in standard in vitro transcription assays. However, the presence of the C-terminal region delayed formation of transcriptionally competent species when the AsiA/polymerase complex was pre-incubated with the promoter in the absence of MotA. Our results suggest that amino acid residues within the N-terminal half of AsiA are involved in forming or maintaining the AsiA/sigma(70) complex. The C-terminal region of AsiA, while not absolutely required for inhibition or co-activation, aids inhibition by slowing the formation of transcription complexes between a promoter and the AsiA/polymerase complex.

Focused genetic recombination of bacteriophage t4 initiated by double-strand breaks

A model system for studying double-strand-break (DSB)-induced genetic recombination in vivo based on the ets1 segCDelta strain of bacteriophage T4 was developed. The ets1, a 66-bp DNA fragment of phage T2L containing the cleavage site for the T4 SegC site-specific endonuclease, was inserted into the proximal part of the T4 rIIB gene. Under segC(+) conditions, the ets1 behaves as a recombination hotspot. Crosses of the ets1 against rII markers located to the left and to the right of ets1 gave similar results, thus demonstrating the equal and symmetrical initiation of recombination by either part of the broken chromosome. Frequency/distance relationships were studied in a series of two- and three-factor crosses with other rIIB and rIIA mutants (all segC(+)) separated from ets1 by 12-2100 bp. The observed relationships were readily interpretable in terms of the modified splice/patch coupling model. The advantages of this localized or focused recombination over that distributed along the chromosome, as a model for studying the recombination-replication pathway in T4 in vivo, are discussed.

Domain 1.1 of the sigma(70) subunit of Escherichia coli RNA polymerase modulates the formation of stable polymerase/promoter complexes

The sigma 70 (sigma(70)) subunit of Escherichia coli RNA polymerase specifies transcription from promoters that are responsible for basal gene expression during vegetative growth. When sigma(70) is present within polymerase holoenzyme, two of its domains, 2.4 and 4.2, interact with sequences within the -10 and -35 regions, respectively, of promoter DNA. However, in free sigma(70), DNA binding is prevented by domain 1.1, the N-terminal domain of the protein. Previous work has demonstrated that the presence of domain 1.1 is required for efficient transcription initiation at the lambda promoter P(R). To investigate whether this is a general property of domain 1.1, we have used five promoters to compare polymerases with and without domain 1.1 in in vitro transcription assays, and in assays assessing the formation and decay of stable, pretranscription complexes. We find that the absence of domain 1.1 does not render the polymerase defective at all of these promoters. Depending on the promoter, the absence of domain 1.1 can promote or inhibit transcription initiation by affecting the formation of stable pretranscription complexes. However, domain 1.1 does not affect the stability of these complexes once they are formed. For polymerases containing domain 1.1, the efficiency of stable complex formation correlates with how well the -10 and -35 regions of a promoter match the ideal sigma(70) recognition sequences. However, when domain 1.1 is absent, having this match becomes less important in determining how efficiently stable complexes are made. We suggest that domain 1.1 influences initiation by constraining polymerase to assess a promoter primarily by the fitness of its -10 and -35 regions to the canonical sequences.

Repair of topoisomerase-mediated DNA damage in bacteriophage T4

Type II topoisomerase inhibitors are used to treat both tumors and bacterial infections. These inhibitors stabilize covalent DNA-topoisomerase cleavage complexes that ultimately cause lethal DNA damage. A functional recombinational repair apparatus decreases sensitivity to these drugs, suggesting that topoisomerase-mediated DNA damage is amenable to such repair. Using a bacteriophage T4 model system, we have developed a novel in vivo plasmid-based assay that allows physical analysis of the repair products from one particular topoisomerase cleavage site. We show that the antitumor agent 4'-(9-acridinylamino)methanesulphon-m-anisidide (m-AMSA) stabilizes the T4 type II topoisomerase at the strong topoisomerase cleavage site on the plasmid, thereby stimulating recombinational repair. The resulting m-AMSA-dependent repair products do not form in the absence of functional topoisomerase and appear at lower drug concentrations with a drug-hypersensitive topoisomerase mutant. The appearance of repair products requires that the plasmid contain a T4 origin of replication. Finally, genetic analyses demonstrate that repair product formation is absolutely dependent on genes 32 and 46, largely dependent on genes uvsX and uvsY, and only partly dependent on gene 49. Very similar genetic requirements are observed for repair of endonuclease-generated double-strand breaks, suggesting mechanistic similarity between the two repair pathways.

Recombination-dependent DNA replication in phage T4

Studies in the 1960s implied that bacteriophage T4 tightly couples DNA replication to genetic recombination. This contradicted the prevailing wisdom of the time, which staunchly supported recombination as a simple cut-and-paste process. More-recent investigations have shown how recombination triggers DNA synthesis and why the coupling of these two processes is important. Results from T4 were instrumental in our understanding of many important replication and recombination proteins, including the newly recognized replication/recombination mediator proteins. Recombination-dependent DNA replication is crucial to the T4 life cycle as it is the major mode of DNA replication and is also central to the repair of DNA breaks and other damage.

Relationship between hexamerization and ssDNA binding affinity in the uvsY recombination protein of bacteriophage T4

In bacteriophage T4, homologous genetic recombination events are catalyzed by a presynaptic filament containing stoichiometric quantities of the T4 uvsX recombinase bound cooperatively to single-stranded DNA (ssDNA). The formation of this filament requires the displacement of cooperatively bound gp32 (the T4 ssDNA-binding protein) from the ssDNA, a thermodynamically unfavorable reaction. This displacement is mediated by the T4 uvsY protein (15.8 kDa, 137 amino acids), which interacts with both uvsX- and gp32-ssDNA complexes and modulates their properties. Previously, we showed that uvsY exists as a hexamer under physiological conditions and that uvsY hexamers bind noncooperatively but with high affinity to ssDNA. We also showed that a fusion protein containing the N-terminal 101 amino acid residues of uvsY lacks interactions with uvsX and gp32 but retains both weak ssDNA-binding activity and a residual ability to stimulate uvsX-catalyzed recombination functions. Here, we present quantitative data on the oligomeric structure and ssDNA-binding properties of a closely related fusion protein designated uvsY. Sedimentation velocity and equilibrium results establish that uvsY, unlike native uvsY, behaves as a monomer in solution (M(app) = 14.2 kDa, = 2.1). Like native uvsY, uvsY binds noncooperatively to an etheno-DNA (epsilonDNA) lattice with a binding site size of 4 nucleotides/monomer; however at physiological ionic strength, the association constant for uvsY-epsilonDNA is decreased 10(4)-fold relative to native uvsY. Nevertheless, the magnitude of the salt effect on the association constant (K) is essentially unchanged between uvsY and uvsY, indicating that disruption of the C-terminus does not disrupt the electrostatic ssDNA-binding determinants found within each protomer of uvsY. Instead, the large difference in ssDNA-binding affinities reflects the loss of hexamerization ability by uvsY, suggesting that a form of intrahexamer synergism or cooperativity between binding sites within the uvsY hexamer leads to its high observed affinity for ssDNA.

Biochemical interactions within a ternary complex of the bacteriophage T4 recombination proteins uvsY and gp32 bound to single-stranded DNA

The presynaptic phase of homologous recombination requires the formation of a filament of single-stranded DNA (ssDNA) coated with a recombinase enzyme. In bacteriophage T4, at least three proteins are required for the assembly of this presynaptic filament. In addition to the T4 recombinase, uvsX protein, the T4 ssDNA binding protein (gp32), and the uvsY recombination accessory protein are also required. Here we report on a detailed analysis of a tripartite filament containing ssDNA bound by stoichiometric quantities of both uvsY and gp32, which appears to be an important intermediate in the assembly of the T4 presynaptic filament. We demonstrate that uvsY and gp32 simultaneously co-occupy the ssDNA in a noncompetitive fashion. In addition, we show that protein-protein interactions between uvsY and gp32 are not required for the assembly of this ternary complex and do not affect the affinity of uvsY for the ssDNA lattice. Finally, we demonstrate that the interaction of gp32 with the ssDNA is destabilized within this complex, in a manner which is independent of gp32-uvsY interactions. The data suggest that the uvsY protein acts to remodel the gp32-ssDNA complex via uvsY-ssDNA interactions. The implications of these findings for the mechanism of presynapsis in the T4 recombination system are discussed.

Single-stranded DNA binding properties of the UvsX recombinase of bacteriophage T4: binding parameters and effects of nucleotides

Bacteriophage T4 provides an important model for the biochemistry and genetics of DNA metabolism. Phage-encoded proteins conduct all essential steps of T4 DNA replication, repair, and recombination. Central to these three processes is the T4 UvsX protein, a member of the filamentous, ATP-dependent class of general recombination enzymes typified by the Escherichia coli RecA protein. Like RecA, UvsX forms presynaptic filaments on single-stranded (ss) DNA, which are the obligatory nucleoprotein intermediates in recombination. Aspects of the T4 presynaptic filament are explored by quantitative characterization of the UvsX-ssDNA interaction using an etheno-derivitized single-stranded DNA molecule, epsilonDNA, whose fluorescence is enhanced by UvsX binding. Studies with this model lattice show that UvsX exhibits a moderate level of cooperativity (omega=100) when binding to epsilonDNA with a binding-site size (n) equal to four nucleotide residues. Salt-stability studies of this complex reveal that the non-hydrolyzable ATP analog, ATPgammaS, induces a high-affinity binding mode that is distinguishable from complexes formed with ADP or in the absence of a nucleotide cofactor. With this new information, both functional relationships between the UvsX and RecA recombinases, and implications for UvsX interactions with the other proteins of the T4 presynaptic filament (UvsY and gp32) may be further explored.

Characterization of an amino-terminal fragment of the bacteriophage T4 uvsY recombination protein

The uvsY protein plays essential roles in homologous genetic recombination processes in the bacteriophage T4. In vitro, uvsY promotes the formation of presynaptic filaments containing stoichiometric amounts of the T4 uvsX recombinase bound to single-stranded DNA. uvsY protein has intrinsic binding activities towards ssDNA, uvsX, and gp32, the T4-encoded SSB, however, it has not been directly determined which of these activities are essential for uvsY's role in presynapsis. We have therefore sought to generate altered forms of uvsY deficient in uvsX- and/or gp32-binding, in order to assess whether these specific protein-protein interactions are essential for uvsY recombination functions. Limited chymotrypsinolysis of the 16 kDa uvsY protein generates two major fragments: an 11.5 kDa fragment containing the N-terminus of uvsY, and a 4.5 kDa C-terminal fragment. We have expressed and purified the large fragment as a fusion protein containing the N-terminal 101 amino acids of uvsY. We show that this truncated uvsY species, which we call uvsYNT, retains ssDNA-binding activity, but is devoid of both uvsX- and gp32-binding activities. Like native uvsY, uvsYNT stimulates the ssDNA-dependent ATPase activity of the uvsX protein, however, the synergistic effects observed between uvsY, uvsX, and gp32 are not observed with uvsYNT. In addition, uvsYNT weakly stimulates uvsX-catalyzed DNA strand exchange reactions. The latter result is surprising since it suggests that specific interactions with uvsX and/or gp32 are not absolutely essential for uvsY recombination functions. Taken together, the data are consistent with a model in which uvsY-ssDNA interactions alone are capable of promoting the assembly of functional uvsX-ssDNA complexes, while uvsY-protein interactions stabilize uvsX-ssDNA complexes.

Role of the bacteriophage T7 and T4 single-stranded DNA-binding proteins in the formation of joint molecules and DNA helicase-catalyzed polar branch migration

Bacteriophage T7 gene 2.5 single-stranded DNA-binding protein and gene 4 DNA helicase together promote pairing of two homologous DNA molecules and subsequent polar branch migration (Kong, D., and Richardson, C. C. (1996) EMBO J. 15, 2010-2019). In this report, we show that gene 2.5 protein is not required for the initiation or propagation of strand transfer once a joint molecule has been formed between the two DNA partners, a reaction that is mediated by the gene 2.5 protein alone. A mutant gene 2.5 protein, gene 2.5-Delta21C protein, lacking 21 amino acid residues at its C terminus, cannot physically interact with gene 4 protein. Although it does bind to single-stranded DNA and promote the formation of joint molecule via homologous base pairing, subsequent strand transfer by gene 4 helicase is inhibited by the presence of the gene 2.5-Delta21C protein. Bacteriophage T4 gene 32 protein likewise inhibits T7 gene 4 protein-mediated strand transfer, whereas Escherichia coli single-stranded DNA-binding protein does not. The 63-kDa gene 4 protein of phage T7 is also a DNA primase in that it catalyzes the synthesis of oligonucleotides at specific sequences during translocation on single-stranded DNA. We find that neither the rate nor extent of strand transfer is significantly affected by concurrent primer synthesis. The bacteriophage T4 gene 41 helicase has been shown to catalyze polar branch migration after the T4 gene 59 helicase assembly protein loads the helicase onto joint molecules formed by the T4 UvsX and gene 32 proteins (Salinas, F., and Kodadek, T. (1995) Cell 82, 111-119). We find that gene 32 protein alone forms joint molecules between partially single-stranded homologous DNA partners and that subsequent branch migration requires this single-stranded DNA-binding protein in addition to the gene 41 helicase and the gene 59 helicase assembly protein. Similar to the strand transfer reaction, strand displacement DNA synthesis catalyzed by T4 DNA polymerase also requires the presence of gene 32 protein in addition to the gene 41 and 59 proteins.

Inhibition of Holliday structure resolving endonuclease VII of bacteriophage T4 by recombination enzymes UvsX and UvsY

Proteins UvsX, UvsY and Endonuclease VII (Endo VII) of bacteriophage T4 are required for DNA recombination, replication and repair. Endo VII is the product of gene 49 (gp49) and essential for resolution of branches from newly made DNA, prior to packaging into preformed heads. The ability of Endo VII to resolve Holliday structures in vitro suggested an in vivo function for the resolution of recombination intermediates, generated by UvsX and UvsY during the early infection cycle. Here we report results which contrast with this hypothesis. It is shown that the potent endonucleolytic activity of Endo VII with branched DNAs is inhibited in strand transfer reactions by the strand transferase UvsX, and more strongly by the accessory protein UvsY in vitro. The inhibitory effect of UvsX or UvsY is also seen in reactions with Endo VII using two synthetic cruciform DNAs and a C/C-mismatch containing substrate. Low concentrations of UvsY protein (12 ng or 0,76 pmol) were sufficient to reduce the cleavage efficiency of 30 units of Endo VII (about 16 fmol) to 50%. The inhibition is due to a direct protein-protein interaction between Endo VII, UvsX and UvsY as suggested by electrophoretic mobility shift assays (EMSAs). These results were confirmed through affinity chromatography, where UvsX and UvsY bound to Endo VII, immobilized on a NHS-activated Sepharose matrix. This is the first identification of phage-encoded proteins which modulate the potent endonucleolytic activity of gp49 in vitro.

The gene 32 single-stranded DNA-binding protein is not bound stably to the phage T4 presynaptic filament

A central reaction in homologous recombination is synapsis, which involves invasion of duplex DNA by a homologous single strand. A key intermediate in this process is the presynaptic filament, a protein-DNA complex composed of a "strand transferase" polymerized along the invading single strand. In this report, the organization and mechanism of assembly of the bacteriophage T4 presynaptic filament are explored. Three T4 proteins, encoded by the uvsX, uvsY and 32 genes, are involved in this process. It is demonstrated that a well-defined series of events involving multiple protein-DNA and protein-protein interactions is required to mediate a transition from an initial gene 32-DNA complex to a mature presynaptic filament in which the UvsX and UvsY proteins are in contact with the DNA and each other, while most or all of the gene 32 protein is removed from the complex.

Phage T4 homologous strand exchange: a DNA helicase, not the strand transferase, drives polar branch migration

Homologous strand exchange is a central step in general genetic recombination. A multiprotein complex composed of five purified bacteriophage T4 proteins (the products of the uvsX, uvsY, 32, 41, and 59 genes) that mediates strand exchange under physiologically relevant conditions has been reconstituted. One of these proteins, the product of the uvsY gene, is required for homologous pairing but strongly inhibits branch migration catalyzed by UvsX protein, the phage RecA analog. Branch migration is completely dependent on the gene 41 protein, a DNA helicase that also functions in phage replication. The helicase is delivered to the strand exchange complex by the gene 59 accessory protein in a strand-specific fashion through direct interactions between the gene 59 and gene 32 proteins. These data suggest that strand transferases such as UvsX protein are essential for homologous pairing in vivo, but that a DNA helicase drives polar branch migration.

Recombination apparatus of T4 phage

The substantial process of general DNA recombination consists of production of ssDNA, exchange of the ssDNA and its homologous strand in a duplex, and cleavage of branched DNA to maturate recombination intermediates. Ten genes of T4 phage are involved in general recombination and apparently encode all of the proteins required for its own recombination. Several proteins among them interact with each other in a highly specific manner based on a protein-protein affinity and constitute a multicomponent protein machine to create an ssDNA gap essential for production of recombinogenic ssDNA, a machine to supply recombinogenic ssDNA which has a free end, or a machine to transfer the recombinogenic single strand into a homologous duplex.

A bacteriophage model system for studying topoisomerase inhibitors

The bacteriophage T4 provides a unique and informative system in which to study the mechanism of action of antitumor agents that inhibit type II DNA topoisomerases. The evolutionary conservation of inhibitor sensitivity provides a strong argument for a conserved inhibitor binding site at or very near the active site of the enzyme-DNA complex. Studies of the wild-type and drug-resistant T4 topoisomerases have provided several important arguments that the drug binding site is located very near the phosphodiester bonds that are cleaved by the topoisomerase. One reasonable model is that the inhibitors intercalate between the two bases on each side of the cleaved phosphodiester bond and physically block the resealing reaction. Finally, genetic analyses using the T4 system have provided some of the most detailed information concerning the role of type II topoisomerase in various aspects of DNA metabolism. The topoisomerase is involved in two distinct pathways of mutagenesis, one that generates frameshift mutations and the other involving gross DNA rearrangements. Both pathways operate precisely at the DNA sites that are cleaved by the enzyme in the presence of inhibitors. Furthermore, recombinational repair can apparently correct lesions that are generated upon inhibition of the T4 topoisomerase, and these inhibitors correspondingly stimulate homologous recombination in phage-infected cells. A complete description of the action of antitumor agents that inhibit type II topoisomerases clearly involves many diverse aspects of nucleic acid metabolism.

Renaturation of DNA catalysed by yeast DNA repair and recombination protein RAD10

The RAD10 gene of Saccharomyces cerevisiae is required for the incision step of excision repair of ultraviolet-damaged DNA, and it functions in mitotic recombination. RAD10 has homology to the human excision repair gene ERCC-1. Here we describe the purification of the protein encoded by RAD10 and show that it is a DNA-binding protein with a strong preference for single-stranded DNA. We also show that RAD10 promotes the renaturation of complementary DNA strands.

An overview of homologous pairing and DNA strand exchange proteins

Processes fundamental to all models of genetic recombination include the homologous pairing and subsequent exchange of DNA strands. Biochemical analysis of these events has been conducted primarily on the recA protein of Escherichia coli, although proteins which can promote such reactions have been purified from many sources, both prokaryotic and eukaryotic. The activities of these homologous pairing and DNA strand exchange proteins are either ATP-dependent, as predicted based on the recA protein paradigm, or, more unexpectedly, ATP-independent. This review examines the reactions promoted by both classes of proteins and highlights their similarities and differences. The mechanistic implications of the apparent existence of 2 classes of strand exchange protein are discussed.

Functional interactions between phage T4 and E. coli DNA-binding proteins during the presynapsis phase of homologous recombination

The "protein machine" for phage T4 homologous recombination has begun to be assembled in vitro. A particularly heavily studied reaction has been the uvsX protein (a RecA-like strand transferase)-mediated homologous pairing reaction between single and double-stranded DNAs, a key step in the recombination cycle in vivo. A necessary prerequisite for uvsX protein-mediated pairing is the polymerization of this factor along the invading single strand, a process known as presynapsis. Recent work has indicated that at least two other T4 recombination factors are involved in this process as well, the uvsY and gene 32 products. These proteins are also ssDNA-binding factors and exhibit an affinity for UvsX and each other. In order to begin to sort out the potential functional roles played by these protein-protein interactions in presynapsis, I have examined the ability of the uvsX protein to form stable filaments along ssDNA in the presence of these proteins. It is shown that the uvsY protein relieves the inhibition to filament formation due to the presence of the gene 32 protein, but experiments with the E. coli SSB protein (the bacterial analogue of gp32) suggest that this effect does not involve a direct interaction between UvsY and gp32.

Isolation and genetic characterization of new uvsW alleles of bacteriophage T4

The uvsW gene of bacteriophage T4 is required for wild-type levels of recombination, for normal survival and mutagenesis after UV irradiation, and for wild-type resistance to hydroxyurea. Additionally, uvsW mutations restore the arrested DNA synthesis caused by mutations in any of several genes that block secondary initiation (recombination-primed replication, the major mode of initiation at late times), but only partially restore the reduced burst size. A uvsW deletion mutation was constructed to establish the null-allele phenotype, which is similar but not identical to the phenotype of the canonical uvsW mutation, and to demonstrate convincingly that the uvsW gene is nonessential (although uvsW mutations severely compromise phage production). In an attempt to uncouple the diverse effects of uvsW mutations, temperature-sensitive uvsWts mutants were isolated. Recombination and replication effects were partially uncoupled in these mutants, suggesting distinct and separable roles for uvsW in the two processes. Furthermore, the restoration of DNA synthesis but not recombination in the double mutants uvsW uvsX and uvsW uvsY prompts the hypothesis that the restored DNA synthesis is not recombinationally initiated.

Frameshift and double-amber mutations in the bacteriophage T4 uvsX gene: analysis of mutant UvsX proteins from infected cells

The bacteriophage T4 uvsX gene encodes a 43 kDa, single-stranded DNA-dependent ATPase, double-stranded DNA-binding protein involved in DNA recombination, repair and mutagenesis. Mutants of uvsX have a DNA-arrest phenotype and reduced burst size. Western blot immunoassay of UvsX peptides made by a number of amber mutants revealed amber peptides ranging from 25-32 kDa. Wild-type UvsX protein was also detected in lysates of cells infected with uvsX amber mutants, suggesting that their mutations are suppressed by translational ambiguity. We investigated the effects of mutations near the 5' end of uvsX. A frameshift mutation was engineered at codon 33. Western immunoblots for UvsX protein demonstrated that the frameshift mutant expresses no detectable wild-type UvsX; instead, a 37 kDa reactive peptide was detected. In order to determine if this peptide represents truncated UvsX protein, the mutation was regenerated in the cloned uvsX gene and expressed in transformed Escherichia coli. Endopeptidase digestion of the 37 kDa protein from the cloned gene generated peptide fragments indistinguishable from those obtained from wild-type UvsX. A double-amber mutant of uvsX was also generated by oligonucleotide site-directed mutagenesis. No UvsX protein was detected in lysates of cells infected with the uvsXam64am67 double mutant. Plaque size and sensitivity to UV inactivation for both the double-amber and the frameshift mutants were indistinguishable from those of other uvsX mutants. Mutations in uvsY had no demonstrable effect on efficiency of plating or UV sensitivity of uvsX mutants. Thus, null mutants of uvsX are viable.

Homologous pairing in vitro initiated by DNA synthesis

A number of models have been proposed for the initiation of general genetic recombination. One of these, originally proposed by Meselson and Radding, imagines that the single-stranded 5' tail that results from strand displacement DNA repair synthesis can initiate homologous recombination by invading a homologous duplex. The resultant D-loop intermediate is then processed into mature products. We demonstrate here that an in vitro system composed of the bacteriophage T4 uvsX protein (a RecA-like "strand transferase") and part of the T4 DNA polymerase holoenzyme efficiently mediates pairing between nicked double-stranded circular and linear duplex DNAs, thereby demonstrating the feasibility of a key step in the Meselson-Radding model.

Membrane-associated assembly of a phage T4 DNA entrance vertex structure studied with expression vectors

The DNA entrance vertex of the phage head is critical for prohead assembly and DNA packaging. A single structural protein comprises this dodecameric ring substructure of the prohead. Assembly of the phage T4 prohead occurs on the cytoplasmic membrane through a specific attachment at or near the gp20 DNA entrance vertex. An auxiliary head assembly gene product, gp40, was hypothesized to be involved in assembling the gp20 substructure. T4 genes 20, 40 and 20 + 40 were cloned into expression vectors under lambda pL promoter control. The corresponding T4 gene products were synthesized in high yield and were active as judged by their ability to complement the corresponding infecting T4 mutants in vivo. The cloned T4 gene 20 and gene 40 products were inserted into the cytoplasmic membrane as integral membrane proteins; however, gp20 was inserted into the membrane only when gp40 was also synthesized, whereas gp40 was inserted in the presence or absence of gp20. The gp20 insertion required a membrane potential, was not dependent upon the Escherichia coli groE gene, and assumed a defined membrane-spanning conformation, as judged by specific protease fragments protected by the membrane. The inserted gp20 structure could be probed by antibody binding and protein A-gold immunoelectron microscopy. The data suggest that a specific gp20-gp40-membrane insertion structure constitutes the T4 prohead assembly initiation complex.

Impaired expression of certain prereplicative bacteriophage T4 genes explains impaired T4 DNA synthesis in Escherichia coli rho (nusD) mutants

The Escherichia coli rho 026 mutation that alters the transcription termination protein Rho prevents growth of wild-type bacteriophage T4. Among the consequences of this mutation are delayed and reduced T4 DNA replication. We show that these defects can be explained by defective synthesis of certain T4 replication-recombination proteins. Expression of T4 gene 41 (DNA helicase/primase) is drastically reduced, and expression of T4 genes 43 (DNA polymerase), 30 (DNA ligase), 46 (recombination nuclease), and probably 44 (DNA polymerase-associated ATPase) is reduced to a lesser extent. The compensating T4 mutation goF1 partially restores the synthesis of these proteins and, concomitantly, the synthesis of T4 DNA in the E. coli rho mutant. From analyzing DNA synthesis in wild-type and various multiply mutant T4 strains, we infer that defective or reduced synthesis of these proteins in rho 026-infected cells has several major effects on DNA replication. It impairs lagging-strand synthesis during the primary mode of DNA replication; it delays and depresses recombination-dependent (secondary mode) initiation; and it inhibits the use of tertiary origins. All three T4 genes whose expression is reduced in rho 026 cells and whose upstream sequences are known have a palindrome containing a CUUCGG sequence between the promoter(s) and ribosome-binding site. We speculate that these palindromes might be important for factor-dependent transcription termination-antitermination during normal T4 development. Our results are consistent with previous proposals that the altered Rho factor of rho 026 may cause excessive termination because the transcription complex does not interact normally with a T4 antiterminator encoded by the wild-type goF gene and that the T4 goF1 mutation restores this interaction.

UvsY protein of bacteriophage T4 is an accessory protein for in vitro catalysis of strand exchange

The uvsX and uvsY genes are essential to genetic recombination, recombination-dependent DNA synthesis and to the repair of DNA damage in bacteriophage T4. Purified UvsX protein has been shown to catalyze strand exchange and D-loop formation in vitro, but the role of UvsY protein has been unclear. We report that UvsY protein enhances strand exchange by UvsX protein by interacting specifically with UvsX protein: gene 32 protein (gp32) is not necessary for this effect and UvsY protein has no similar effect on the RecA protein of E. coli. UvsY protein, like UvsX protein, protects single-stranded DNA from digestion by nucleases, but, unlike UvsX protein, shows no ability to protect double-stranded DNA. UvsY protein enhances the rate of single-stranded-DNA-dependent ATP hydrolysis by UvsX protein, particularly in the presence of gp32 or high concentrations of salt, factors that otherwise reduce the ATPase activity of UvsX protein. The enhancement of ATP hydrolysis by UvsY protein is shown to result from the ability of UvsY protein to increase the affinity of UvsX protein for single-stranded DNA.