Characterization of bacteriophage lambda excisionase mutants defective in DNA binding

The bacteriophage lambda excisionase (Xis) is a sequence-specific DNA binding protein required for excisive recombination. Xis binds cooperatively to two DNA sites arranged as direct repeats on the phage DNA. Efficient excision is achieved through a cooperative interaction between Xis and the host-encoded factor for inversion stimulation as well as a cooperative interaction between Xis and integrase. The secondary structure of the Xis protein was predicted to contain a typical amphipathic helix that spans residues 18 to 28. Several mutants, defective in promoting excision in vivo, were isolated with mutations at positions encoding polar amino acids in the putative helix (T. E. Numrych, R. I. Gumport, and J. F. Gardner, EMBO J. 11:3797-3806, 1992). We substituted alanines for the polar amino acids in this region. Mutant proteins with substitutions for polar amino acids in the amino-terminal region of the putative helix exhibited decreased excision in vivo and were defective in DNA binding. In addition, an alanine substitution at glutamic acid 40 also resulted in altered DNA binding. This indicates that the hydrophilic face of the alpha-helix and the region containing glutamic acid 40 may form the DNA binding surfaces of the Xis protein.     

Activity of coliphage HK022 excisionase (Xis) in the absence of DNA binding

A mutated excisionase (Xis) protein of coliphage HK022 whose single Cys residue was replaced by Ser does not bind to its two tandem binding sites (X1, X2) on the P arm of attR. Despite its DNA-binding inability the protein showed 30% excision activity of the wild type Xis both in vitro and in vivo. This partial activity is attributed to the interaction of Xis with integrase that is retained in the mutant protein. This protein-protein interaction occurs in the absence of DNA binding.     

Characterization of the DNA binding domain of bacteriophage lambda O protein

The lambda O and P gene products are required for the initiation of lambda DNA replication. In order to study the biochemistry of this process, we have constructed plasmids that carry the lambda O gene, P gene, and half of the O gene coding for the amino-terminal half of the O protein. Each is under the control of the inducible lambda promoter, PL. We have purified these three proteins from induced cells carrying the plasmids. Our results show that the amino-terminal portion of the O protein binds to the lambda origin of replication in a manner similar to the intact lambda O protein, demonstrating that the amino-terminal portion of O protein contains the DNA binding domain. Using chromatographic procedures, we have isolated a complex of lambda O and P proteins with lambda dv DNA. The amino-terminal portion of the O protein does not complex with P protein under the same conditions. This suggests that the specificity of the lambda O protein for P protein resides in the carboxyl-terminal half of the lambda O protein. Our results also show that, while the intact O protein is active in in vitro replication of lambda dv plasmid DNA, the amino-terminal portion of the O protein is inactive and is a competitive inhibitor of the lambda O protein in this reaction. These results confirm previous genetic observations that were interpreted as indicating a bifunctional structure for the lambda O protein with the amino-terminal domain recognizing the lambda origin of replication and the carboxyl-terminal domain interacting with the lambda P protein.     

Regulation of directionality in bacteriophage lambda site-specific recombination: structure of the Xis protein

Upon induction of a bacteriophage lambda lysogen, a site-specific recombination reaction excises the phage genome from the chromosome of its bacterial host. A critical regulator of this process is the phage-encoded excisionase (Xis) protein, which functions both as a DNA architectural factor and by cooperatively recruiting integrase to an adjacent binding site specifically required for excision. Here we present the three-dimensional structure of Xis and the results of a structure-based mutagenesis study to define the molecular basis of its function. Xis adopts an unusual "winged"-helix motif that is modeled to interact with the major- and minor-grooves of its binding site through a single alpha-helix and loop structure ("wing"), respectively. The C-terminal tail of Xis, which is required for cooperative binding with integrase, is unstructured in the absence of DNA. We propose that asymmetric bending of DNA by Xis positions its unstructured C-terminal tail for direct contacts with the N-terminal DNA-binding domain of integrase and that an ensuing disordered to ordered transition of the tail may act to stabilize the formation of the tripartite integrase-Xis-DNA complex required for phage excision.     

Directional control of site-specific recombination by bacteriophage lambda. Evidence that a binding site for Int protein far from the crossover point is required for integrative but not excisive recombination

Phage lambda controls its integration and excision by differential catalysis of the forward and reverse reactions. The lambda Int protein is required for both directions, but Xis for excision only. To investigate the substrate requirements for directional control, we have characterized two mutations of the phage attachment site that are defective in integrative but not excisive recombination. Both of these mutations produce the same base change in the P'3 binding site for Int protein 79 base-pairs from the center of the crossover region for site-specific recombination. We infer that differential utilization of this distant binding site is crucial for directional control of recombination.     

T7 single strand DNA binding protein but not T7 helicase is required for DNA double strand break repair

An in vitro system based on Escherichia coli infected with bacteriophage T7 was used to test for involvement of host and phage recombination proteins in the repair of double strand breaks in the T7 genome. Double strand breaks were placed in a unique XhoI site located approximately 17% from the left end of the T7 genome. In one assay, repair of these breaks was followed by packaging DNA recovered from repair reactions and determining the yield of infective phage. In a second assay, the product of the reactions was visualized after electrophoresis to estimate the extent to which the double strand breaks had been closed. Earlier work demonstrated that in this system double strand break repair takes place via incorporation of a patch of DNA into a gap formed at the break site. In the present study, it was found that extracts prepared from uninfected E. coli were unable to repair broken T7 genomes in this in vitro system, thus implying that phage rather than host enzymes are the primary participants in the predominant repair mechanism. Extracts prepared from an E. coli recA mutant were as capable of double strand break repair as extracts from a wild-type host, arguing that the E. coli recombinase is not essential to the recombinational events required for double strand break repair. In T7 strand exchange during recombination is mediated by the combined action of the helicase encoded by gene 4 and the annealing function of the gene 2.5 single strand binding protein. Although a deficiency in the gene 2.5 protein blocked double strand break repair, a gene 4 deficiency had no effect. This argues that a strand transfer step is not required during recombinational repair of double strand breaks in T7 but that the ability of the gene 2.5 protein to facilitate annealing of complementary single strands of DNA is critical to repair of double strand breaks in T7.     

The form of the DNA substrate required for excisive recombination of bacteriophage lambda.

We have studied the excision reaction of bacteriophage lambda, both in vivo and in vitro, using as a substrate a λatt²(L × R) phage carrying both the right and left-hand prophage attachment sites. Int and Xis are provided by induction of the heat-inducible defective prophage, λc1857 ΔH1. After a brief induction (5 min) of these cells, excisive recombination is blocked in the presence of the DNA gyrase inhibitor, coumermycin. However, after a longer induction (greater than 30 min) excisive recombination occurs efficiently under conditions where λ integrative recombination is inhibited by coumermycin. In such extensively induced coumermycin-treated cells, infecting λatt²(L × R) DNA is not supercoiled, and recombinants are found among the relaxed covalently closed circular DNA.In vitro, starting with a hydrogen-bonded λatt² DNA substrate, excision is insensitive to high concentrations of coumermycin and novobiocin. To study the DNA substrate requirements for excisive recombination in more detail, we have developed a restriction fragment assay for excisive recombination. With this assay, we demonstrate that supercoiled, hydrogen-bonded, and linear λatt² DNA molecules are all efficient substrates in the in vitro excision reaction. Spermidine is required but ATP and Mg²⁺ are not. We conclude that supercoiling is not an absolute requirement for site-specific recombination of λ.     

The cysteine conserved among DNA cytosine methylases is required for methyl transfer, but not for specific DNA binding.

All DNA (cytosine-5)-methyltransferases contain a single conserved cysteine. It has been proposed that this cysteine initiates catalysis by attacking the C6 of cytosine and thereby activating the normally inert C5 position. We show here that substitutions of this cysteine in the E. coli methylase M. EcoRII with either serine or tryptophan results in a complete loss of ability to transfer methyl groups to DNA. Interestingly, mutants with either serine or glycine substitution bind tightly to substrate DNA. These mutants resemble the wild-type enzyme in that their binding to substrate is not eliminated by the presence of non-specific DNA in the reaction, it is sensitive to methylation status of the substrate and is stimulated by an analog of the methyl donor. Hence the conserved cysteine is not essential for the specific stable binding of the enzyme to its substrate. However, substitution of the cysteine with the bulkier tryptophan does reduce DNA binding. We also report here a novel procedure for the synthesis of DNA containing 5-fluorocytosine. Further, we show that a DNA substrate for M. EcoRII in which the target cytosine is replaced by 5-fluorocytosine is a mechanism-based inhibitor of the enzyme and that it forms an irreversible complex with the enzyme. As expected, this modified substrate does not form irreversible complexes with the mutants.     

DNA binding by the Xis protein of the conjugative transposon Tn916.

We purified the Xis protein of the conjugative transposon Tn916 and showed by nuclease protection experiments that Xis bound specifically to sites close to each end of Tn916. These specific binding sites are close to, and in the same relative orientation to, binding sites for the N-terminal domain of Tn916 integrase protein. These results suggest that Xis is involved in the formation of nucleoprotein structures at the ends of Tn916 that help to correctly align the ends so that excision can occur.     

The Bloom's syndrome protein (BLM) interacts with MLH1 but is not required for DNA mismatch repair

Bloom's syndrome (BS) is a rare autosomal recessive disorder characterized by pre- and postnatal growth deficiency, immunodeficiency, and a tremendous predisposition to a wide variety of cancers. Cells from BS individuals are characterized by a high incidence of chromosomal gaps and breaks, elevated sister chromatid exchange, quadriradial formations, and locus-specific mutations. BS is the consequence of mutations that lead to loss of function of BLM, a gene encoding a helicase with homology to the RecQ helicase family. To delineate the role of BLM in DNA replication, recombination, and repair we used a yeast two-hybrid screen to identify potential protein partners of the BLM helicase. The C terminus of BLM interacts directly with MLH1 in the yeast-two hybrid assay; far Western analysis and co-immunoprecipitations confirmed the interaction. Cell extracts deficient in BLM were competent for DNA mismatch repair. These data suggest that the BLM helicase and MLH1 function together in replication, recombination, or DNA repair events independent of single base mismatch repair.     

Domain structure and DNA binding regions of beta protein from bacteriophage lambda

beta protein from bacteriophage lambda promotes a single-strand annealing reaction that is central to Red-mediated recombination at double-strand DNA breaks and chromosomal ends. beta protein binds most tightly to an intermediate of annealing formed by the sequential addition of two complementary oligonucleotides. Here we have characterized the domain structure of beta protein in the presence and absence of DNA using limited proteolysis. Residues 1-130 form an N-terminal "core" domain that is resistant to proteases in the absence of DNA, residues 131-177 form a central region with enhanced resistance to proteases upon DNA complex formation, and the C-terminal residues 178-261 of beta protein are sensitive to proteases in both the presence and absence of DNA. We probed the DNA binding regions of beta protein further using biotinylation of lysine residues and mass spectrometry. Several lysine residues within the first 177 residues of beta protein are protected from biotinylation in the DNA complex, whereas none of the lysine residues in the C-terminal portion are protected. The results lead to a model for the domain structure and DNA binding of beta protein in which a stable N-terminal core and a more flexible central domain come together to bind DNA, whereas a C-terminal tail remains disordered. A fragment consisting of residues 1-177 of beta protein maintains normal binding to sequentially added complementary oligonucleotides and has significantly enhanced binding to single-strand DNA.     

Identification of two functional regions in Fis: the N-terminus is required to promote Hin-mediated DNA inversion but not lambda excision.

The Fis protein of E. coli binds to a recombinational enhancer sequence that is required to stimulate Hin-mediated DNA inversion. Fis is also required for efficient lambda prophase excision in vivo. The properties of mutant Fis proteins were examined in vivo and in vitro with respect to their stimulatory effects on these two different site-specific DNA recombination reactions. Both recombination reactions are dramatically affected by mutations altering a helix-turn-helix DNA binding motif located near the Fis C-terminus (residues 74-93). These mutations invariably decrease DNA binding affinity and some cause reduced DNA bending. Mutations in the Fis N-terminal region reduce or abolish the stimulation of Hin-mediated DNA recombination by Fis, but have little or no effect on DNA binding or lambda excision. We conclude that there are at least two functionally distinct domains in Fis: a C-terminal DNA binding region that is required for promoting both DNA recombination reactions and an N-terminal region that is uniquely required for Hin-mediated inversion.     

DNA binding in the central channel of bacteriophage T7 helicase-primase is a multistep process. Nucleotide hydrolysis is not required

Many helicases assemble into ring-shaped hexamers and bind DNA in their central channel. This raises the question as to how the DNA gets into the central channel to form a topologically linked complex. We have used the presteady-state stopped-flow kinetic method and protein fluorescence changes to investigate the mechanism of single-stranded DNA (ssDNA) binding to the bacteriophage T7 helicase-primase, gp4A'. We have found that the kinetics of 30-mer ssDNA binding to a preformed gp4A' hexamer in the presence of both Mg-dTMP-PCP and Mg-dTTP are similar, indicating that Mg-dTTP binding is sufficient and hydrolysis is not necessary for efficient DNA binding. Multiple transient changes in gp4A' fluorescence revealed a four-step mechanism for DNA binding with Mg-dTTP. These transient changes were analyzed by global fitting and kinetic simulation to determine the intrinsic rate constants of this four-step mechanism. The initial steps, including the bimolecular encounter of the DNA with the helicase and a subsequent conformational change, were fast. We propose that these initial steps of DNA binding occur at a readily accessible site, which is likely to be on the outside of the hexamer ring. The binding of the 30-mer ssDNA at this loading site is followed by slower conformational changes that allow the DNA to transit into the central channel of gp4A' via a ring-opening or threading pathway.     

Calcium is not required for immulectin-2 binding, but protects the protein from proteinase digestion

Mammalian C-type lectins are calcium-dependent carbohydrate-binding proteins. They serve as cell adhesion molecules in cell-cell interactions, or function as pattern-recognition receptors in innate immunity. Calcium is a direct ligand for carbohydrate binding in mammalian C-type lectins such as mannose-binding proteins and macrophage mannose receptor. In the tobacco hornworm Manduca sexta, a group of lectins named immulectins have been discovered. Each immulectin contains dual carbohydrate-recognition domains. Previously, we showed that immulectin-2 (IML-2) binds to a bacterial lipopolysaccharide, and agglutination of Escherichia coli cells by IML-2 is calcium dependent. In this study, we demonstrated that IML-2 bound to bacterial lipid A, smooth and rough mutants of lipopolysaccharide, lipoteichoic acid and peptidoglycan, as well as to fungal mannan and beta-1, 3-glucan (laminarin and curdlan). Binding of IML-2 to microbial components was calcium independent, and was increased by addition of spermine, a polyamine. In addition, plasma IML-2 bound to mannan-agarose independent of calcium. But trypsin digestion of IML-2 was inhibited in the presence of calcium. Our results suggest that calcium is not required for IML-2 binding but protects IML-2 from trypsin digestion.     

Multimerization of the Drosophila zeste protein is required for efficient DNA binding.

The Drosophila zeste protein forms multimeric species in vitro through its C-terminal domain. Multimerization is required for efficient binding to DNA containing multiple recognition sequences and increasing the number of binding sites stimulates binding in a cooperative manner. Mutants that can only form dimers still bind to a dimeric site, but with lower affinity. Mutations or progressive deletions from the C-terminal show that when even dimer formation is prevented, DNA-binding activity is lost. Surprisingly, binding activity is regained with larger deletions that leave only the DNA-binding domain. Additional protein sequences apparently inhibit DNA binding unless they permit multimerization. The DNA-binding domain peptides bind strongly even to isolated recognition sequences and they bind as monomers. The ability of various zeste peptides to stimulate white gene expression in vivo shows that multimeric forms are the functional species of the zeste product in vivo. The DNA-binding domain peptide binds well to DNA in vitro, but it cannot stimulate white gene expression in vivo. This failure may reflect the need for an activation domain or it may be caused by indiscriminate binding of this peptide to non-functional isolated sites. Multimerization increases binding specificity, selecting only sites with multiple recognition sequences.     

Heat shock protein-mediated disassembly of nucleoprotein structures is required for the initiation of bacteriophage lambda DNA replication

Three Escherichia coli heat shock proteins, DnaJ, DnaK, and GrpE, are required for replication of the bacteriophage lambda chromosome in vivo. We show that the GrpE heat shock protein is not required for initiation of lambda DNA replication in vitro when the concentration of DnaK is sufficiently high. GrpE does, however, greatly potentiate the action of DnaK in the initiation process when the DnaK concentration is reduced to a subsaturating level. We demonstrate in the accompanying articles (Alfano, C. and McMacken, R. (1989) J. Biol. Chem. 264, 10699-10708; Dodson, M., McMacken, R., and Echols, H. (1989) J. Biol. Chem. 264, 10719-10725) that DnaJ and DnaK bind to prepriming nucleoprotein structures that are assembled at the lambda replication origin (ori lambda). Binding of DnaJ and DnaK completes the ordered assembly of an ori lambda initiation complex that also contains the lambda O and P initiators and the E. coli DnaB helicase. With the addition of ATP, the DnaJ and DnaK heat shock proteins mediate the partial disassembly of the initiation complex, and the P and DnaJ proteins are largely removed from the template. Concomitantly, on supercoiled ori lambda plasmid templates, the intrinsic helicase activity of DnaB is activated and DnaB initiates localized unwinding of the DNA duplex, thereby preparing the template for priming and DNA chain elongation. We infer from our results that DnaK and DnaJ function in normal E. coli metabolism to promote ATP-dependent protein unfolding and disassembly reactions. We also provide evidence that neither the lambda O and P initiators nor the E. coli DnaJ and DnaK heat shock proteins play a direct role in the propagation of lambda replication forks in vitro.     

Mapping the interactions of the single-stranded DNA binding protein of bacteriophage T4 (gp32) with DNA lattices at single nucleotide resolution: polynucleotide binding and cooperativity

We here use our site-specific base analog mapping approach to study the interactions and binding equilibria of cooperatively-bound clusters of the single-stranded DNA binding protein (gp32) of the T4 DNA replication complex with longer ssDNA (and dsDNA) lattices. We show that in cooperatively bound clusters the binding free energy appears to be equi-partitioned between the gp32 monomers of the cluster, so that all bind to the ssDNA lattice with comparable affinity, but also that the outer domains of the gp32 monomers at the ends of the cluster can fluctuate on and off the lattice and that the clusters of gp32 monomers can slide along the ssDNA. We also show that at very low binding densities gp32 monomers bind to the ssDNA lattice at random, but that cooperatively bound gp32 clusters bind preferentially at the 5′-end of the ssDNA lattice. We use these results and the gp32 monomer-binding results of the companion paper to propose a detailed model for how gp32 might bind to and interact with ssDNA lattices in its various binding modes, and also consider how these clusters might interact with other components of the T4 DNA replication complex.