A.D. Kaiser

The transformation of Escherichia coli with deoxyribonucleic acid isolated from bacteriophage λdg

Deoxyribonucleic acid of λ dg , isolated by phenol extraction, will transform galactose-negative strains of Escherichia coli K12. That the transforming agent is indeed DNA is shown by (1) its sensitivity to pancreatic DNase, (2) its resistance to anti-λ antibody, (3) its resistance to heat up to the characteristic melting temperature of λ dg DNA, and (4) its buoyant density of 1·71 g cm −3 . Transformation is found to occur only if the bacteria exposed to λ dg DNA are simultaneously infected with ordinary λ. The transforming agent appears to be the entire λ dg chromosome since the phage genes c and mi are also present in the galactose-positive transformants.

Enzymatic end-to-end joining of DNA molecules☆

Abstract A way to join naturally occurring DNA molecules, independent of their base sequence, is proposed, based upon the presumed ability of the calf thymus enzyme terminal deoxynucleotidyltransferase to add homopolymer blocks to the ends of double-stranded DNA. To test the proposal, covalently closed dimer circles of the DNA of bacteriophage P22 were produced from linear monomers. It is found that P22 DNA as isolated will prime the terminal transferase reaction, but not in a satisfactory manner. Pre-treatment of the DNA with λ exonuclease, however, improves its priming ability. Terminal transferase can then be used to add oligo(dA) blocks to the ends of one population of P22 DNA molecules and oligo(dT) blocks to the ends of a second population, which enables the two DNAs to anneal to one another to form dimer circles. Subsequent treatment with a system of DNA repair enzymes converts the circles to covalently closed molecules at high efficiency. It is demonstrated that the success of the joining system does not depend upon any obvious unique property of the P22 DNA. The joining system yields several classes of by-products, among them closed circular molecules with branches. Their creation can be explained on the basis of the properties of terminal transferase and the DNA repair enzymes.

Changes in the structure and activity of λ DNA in a superinfected immune bacterium

Labeled intracellular λ DNA isolated from superinfected lysogenic Escherichia coli K12 exhibits a very low level of infectivity. Sedimentation studies revealed that the residual infectivity was due to a few fully infectious molecules (species III), but that a majority of the labeled phage DNA was present as two new and non-infectious species (I and II). At neutral pH, intracellular species I sedimented 1·9 times faster and species II 1·14 times faster than λ DNA isolated from mature phage. In alkali, species II co-sedimented with denatured mature DNA molecules while species I moved 3·8 times faster. During storage, tritiated species I is converted slowly into II and also loses its distinctive capacity to renature readily after alkali treatment. It is suggested that intracellular species I is a condensed form of λ DNA, species II is a circular molecule and species III is a normal linear duplex molecule similar to those obtained from mature phage.

Structure and base sequence in the cohesive ends of bacteriophage lambda DNA

At the ends of molecules of bacteriophage λ DNA, the 5′-terminated strands are 20 nucleotides longer than the 3′-terminated strands. Part of the sequence of bases in the protruding single strands has been determined by the Escherichia coli DNA polymerase-catalyzed addition of nucleotides, lengthening the 3′-terminated strands. In this way 13 residues of dG, 13 of dC, 7 of dA and 7 of dT were incorporated. The equivalence of dG to dC and of dA to dT implies that the protruding single strands have complementary base sequences. The first residue incorporated into the right cohesive end was the same as the 5′-terminal residue on the left end and the first residue incorporated into the left end was the same as the 5′-terminal residue on the right end, showing that the two cohesive ends are the same length. Sequence data is summarized in Figure 7.

The production of phage chromosome fragments and their capacity for genetic transfer.

Molecules of DNA isolated from bacteriophage λ break in half, but no further, when subjected to the appropriate amount of hydrodynamic shear. The half-molecules can be separated from whole molecules by zone sedimentation in a sucrose gradient. The phage particles produced by bacteria exposed to λ DNA and active “helper” phage reveal the genes carried into the bacterium by the DNA. Whole molecules of λ DNA appear to be the entire phage chromosome since they carry genes from all parts of the recombination map: m 6 , m 5 , h, s, i and mi . One of the two half-molecules produced by shear breakage of λ DNA is capable of transferring s, i and mi , but not m 6 , m 5 or h . Since this incomplete complement is a continuous segment of the linkage map of λ , it is argued that the sequence of genes on the recombination map is the same as the sequence of the nucleotides or blocks of nucleotides which correspond to each of the genes in a molecule of λ DNA.

Morphological proteins of phage lambda: Identification of the major head protein as the product of gene E☆

Abstract Phage λ dissociates in boiling SDS into at least three major and three minor proteins, which separate from each other in gel electrophoresis. By joining, in vitro, radioactive heads to nonradioactive tails, and vice versa, and by purifying the resulting complete phage particles it was possible to identify three of the proteins as constituents of the head and three of the tail. The major head protein has a molecular weight of 38,000 and accounts for 60% of the total protein mass of the phage particle. This protein appears to be the product of gene E because it is absent from extracts of induced lysogenic bacteria carrying a susE− prophage, but is present in extracts of lysogens for sus (amber) mutants in any of the other known head genes: A, W, B, C, D, or F. Petit λ, a form of the λ head which lacks a tail and contains no DNA, is also an assembly of the major head protein. However, purified petit λ lacks the second most abundant protein of complete heads which has a molecular weight of 12,000.

Cohesion and the biological activity of bacteriophage lambda DNA.

The two ends of a molecule of gl DNA cohere to each other specifically and reversibly. A closed molecule of λ DNA, one in which the two ends have cohered, has a much lowered biological activity toward helper-infected bacteria. The molecule regains full activity when the cohered ends are separated. Restoration of activity has been used to measure the opening of closed molecules as a function of temperature. Broken molecules of λ DNA also lose most, if not all, of their activity when they cohere. Again the lost activity is regained when the cohered ends are separated. Thus, while the low activity of closed molecules might be explained by the absence of ends of any sort, the low activity of cohered broken molecules must be explained by the absence of free cohesive ends, since cohered broken molecules have ends. If a free cohesive end is required for biological activity, then fragments arising from the interior of molecules which have suffered two or more breaks would not be active. Activity would be an indication of linkage to a cohesive end. When molecules of λ DNA are broken into six pieces, activity of the marker i λ can no longer be detected but the marker pairs sus A sus B and sus Q sus R retain their activity. This is taken to mean that sus A and sus B are linked to one cohesive end, that sus Q and sus R are linked to the other, but that i λ is more than one-sixth the length of the molecule away from either one. Since the cohesive sites are at the ends of the molecule, the observed pattern of activity is taken to mean that the sequence of the DNA of a mature λ particle is congruent (or colinear) with the vegetative, rather than the prophage, recombination map.

On the structure of the ends of lambda DNA

The structure of the ends of a molecule of λ DNA has been deduced from the effect of DNA polymerase and exonuclease III on the infectivity of λ DNA and from the ability of the two ends of the molecule to cohere. The infectivity of λ DNA is destroyed when it primes DNA synthesis catalyzed by DNA polymerase. Inactivation depends on the presence of deoxynucleoside triphosphates and on the amount of enzyme added. The inactivating synthesis behaves like a repair of single-stranded regions, that is, their conversion to double strands, because it occurs at 15°C. Richardson, Inman & Kornberg (1964) found that DNA polymerase catalyzes repair but not net synthesis at temperatures lower than 20°C. Infectivity returns if the polymerase product is subsequently exposed to exonuclease III. The specificity of exonuclease III is such that it would be expected to remove nucleotides added by polymerase. This result shows that inactivation by polymerase is not due to degradation. In a complementary fashion the infectivity of native λ DNA is destroyed by exonuclease III and restored by subsequent synthesis catalyzed by DNA polymerase. The two ends of a molecule of λ DNA can cohere, specifically, to each other ( Hershey, Burgi & Ingraham, 1963 ; Ris & Chandler, 1963 ). To account for cohesiveness and for inactivation by DNA polymerase, it is proposed that 5′-hydroxyl (or 5′-phosphate)-terminated single strands protrude from both ends of the otherwise double-stranded molecule and that the two protruding single strands have complementary base sequences. Cohesion would result from base pairing between the protruding single strands. Synthesis primed by native λ DNA and catalyzed by DNA polymerase at 15°C would be expected to repair partially single-stranded ends and to destroy their cohesiveness. As the infectivity of λ DNA for helper-infected bacteria requires cohesiveness, polymerase would thus inactivate λ DNA. Treating the polymerase product with exonuclease III restores infectivity, according to this model, because removal of the added nucleotides restores cohesiveness.

Isolation and structure of phage λ head-mutant DNA☆

Abstract High molecular weight DNA accumulates in bacteria in which λ is multiplying but cannot complete the formation of new phage particles due to a defect in head assembly. Accumulated λ DNA has been isolated from induced mitomycin C-treated lysogens by means of a shift in buoyant density labels from heavy to light and fractionation by density-gradient sedimentation for completely light DNA. Head formation was blocked in these lysogens by amber mutations in genes D or E , which specify the two major head proteins. The purified DNA is at least 80% λ by DNA-DNA hybridization and some preparations are close to 100% λ by this test. Electron microscopy and sedimentation velocity measurements on these preparations showed that they were a population, of linear molecules (about 5% were circular) with lengths ranging from one to more than four monomer units. After partial denaturation, the long molecules showed a repeating λ-like pattern of denatured sites. The length of the repeating unit equalled the length of DNA molecules isolated from phage particles. The distribution of denatured sites within the repeating unit was the same as that for DNA isolated from phage particles. Thus, the long molecules are head-to-tail polymers of λ DNA. These structures may be head precursor DNA, which accumulates because one of the structural proteins of the head is missing. Perhaps λ DNA is normally cut into monomer units at the same time and as an integral part of head protein assembly and DNA encapsulation.

Head assembly steps controlled by genes F and W in bacteriophage λ

Abstract Bacteria infected with λ mutants defective in gene F or W produce unjoined tails and heads. The heads have the same size, shape, buoyant density and sedimentation velocity as λ heads isolated from tail-defective lysates. DNA isolated from these heads has the same length as that isolated from complete phage particles and has normal cohesive ends. In spite of their apparently normal morphology, F− and W− heads are unable to join tails, but this defect can be complemented, in vitro, by extracts of induced lysogenic bacteria. From such extracts a protein has been purified which renders F− heads able to join tails. Thus F− heads have not reacted with a protein required for tail attachment. Reaction with W protein appears to precede reaction with F protein: purified F− heads can be activated by a W-deficient extract, but purified W− heads cannot be complemented by an F-deficient extract. Thus genes W and F seem to control two ordered terminal steps in head assembly which prepare a DNA-filled head for tail attachment.