Properties of T4D Bacteriophage Grown in Synthetic Media Containing Zn2+, Co2+, or Ni2+*

Earlier work showed that the tail baseplate of bacteriophage T4D contained 5 to 6 atoms of zinc. T4D bacteriophage now has been grown on Escherichia coli B in a glucose/salts medium containing added 2 x lo-” M ZrP, Co’+, or Ni’+ ions. Phage particles have also been grown on E. coli in media containing 2 x lo-” M WoCl,. The added zinc and cobalt stimulated the formation of infectious phage particles 2- to g-fold, while the added nickel depressed the phage yield. The T4D particles formed in the presence of 5HC~ were highly purified, osmotically shocked to remove their DNA, and the cobalt content of the ghost particles determined. It was found that there were 4 f 1 atoms of cobalt/ghost particle, which is in reasonable agreement with the previously determined zinc content. The nonradio-active phage particles produced in the presence of these three metals have also been purified and their physical, chemical, and biological properties examined. Properties of these phage particles including heat lability, sensitivity to metal chelating agents such as o-phenanthroline and dimethylglyoxime, and rate of inactivation by Cd(CN);-treatment varied greatly with the metal ions added to the growth medium. Further, while T4D phage grown

Earlier work showed that the tail baseplate of bacteriophage T4D contained 5 to 6 atoms of zinc. T4D bacteriophage now has been grown on Escherichia coli B in a glucose/salts medium containing added 2 x lo-" M ZrP, Co'+, or Ni'+ ions. Phage particles have also been grown on E. coli in media containing 2 x lo-" M WoCl,.
The added zinc and cobalt stimulated the formation of infectious phage particles 2-to g-fold, while the added nickel depressed the phage yield. The T4D particles formed in the presence of 5HC~ were highly purified, osmotically shocked to remove their DNA, and the cobalt content of the ghost particles determined.
It was found that there were 4 f 1 atoms of cobalt/ghost particle, which is in reasonable agreement with the previously determined zinc content. The nonradioactive phage particles produced in the presence of these three metals have also been purified and their physical, chemical, and biological properties examined. Properties of these phage particles including heat lability, sensitivity to metal chelating agents such as o-phenanthroline and dimethylglyoxime, and rate of inactivation by Cd(CN);treatment varied greatly with the metal ions added to the growth medium. Further, while T4D phage grown in the presence of the added Zn'+ ions was readily inactivated by partially purified hog kidney folic acid conjugase (a yglutamyl peptidase), phage particles grown in the presence of Co'+ or Ni2+ ions were completely resistant to this enzyme. The direct analytical data for cobalt, as well as the virion properties, indicate that Co2+ and Ni'+ can be substituted in vivo for Zn2+ in the phage baseplate and suggest that the baseplate metalloprotein is physically close to the glutamyl residues of the phage baseplate dihydropteroylhexaglutamate.

Investigations
of the properties of the Escherichia coli Teven phage particles continue to be of fundamental interest since they bear on such fundamental questions as (a) the nature and expression of the genetic information needed for the morphogenesis of a complex biological structure and (b) the process by which the viral DNA is efficiently introduced into the host cell. While considerable progress has been made * This work was supported in part by Grant AI 06336 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. on the morphological features of the virus particles and the changes that take place during infection by Simon and Anderson (1,2) using refined electron microscopic techniques and by the analysis of the genetic control of morphogenesis in several laboratories (3-Q our understanding of the chemical steps in morphogenesis and the chemical events during injection has developed quite slowly. For example, a critical feature of the distinctly shaped virus particle is the hexagonal baseplate near the terminal end of the phage particle. During interaction with the host cell, this hexagonal structure is known to change its shape from a hexagonal into a star-like structure (1,2). This morphological rearrangement of a structure whose molecular mass is of the order of 8 x 10" daltons (6)(7)(8), results in an opening in the center through which the DNA apparently passes into the host cell (1,2,9,10). The forces and reactants controlling this morphological change, including the nature of the energy needed for the movement of the DNA and its precise introduction into the host cell, are still obscure. Furthermore, the role of the many viral gene products needed to assemble this baseplate structure are only now being appreciated. For example, it is known at least 18 T4 genes are needed to specify proteins to make up the complex baseplate (6)(7)(8). Additional features of this complex structure are only now being defined. One unusual feature, which was first reported from this laboratory in 1965 (11-131, was the presence in the baseplate of 6 molecules of an unusual form of folic acid, namely, a dihydropteroylhexaglutamate. Subsequent work revealed that the baseplate also contained two phage-induced enzymes, dihydrofolate reductase (14-17) and thymidylate synthetase (18,19). The role that these enzymes play either in forming the baseplate or in aiding injection, as well as the presence of an unusual small enzyme cofactor such as folate in the baseplate, has not been linked precisely to the mechanism of the morphogenesis of the baseplate or to its physiological operation. The understanding of the biochemical features of the assembly of this dynamic phage organelle requires not only the determination of which gene products are involved and the sequence of their interaction, but also a knowledge of the bonds holding the structure together and the forces involved in altering its configuration upon interaction with the host cell.
Recently, and somewhat surprisingly, we found that the hexagonally shaped tail baseplates of E. coli T-even bacteriophage particles each contain about 5 to 6 atoms of zinc (20). This conclusion was based on the analytical determination of zinc in T4D, T4B, and T2H phage ghost particles and the Zinc, Cobalt, and Nickel in  and enough carrier CoCl, added so that-the final concentration was 1.9 x 10-l; M. Two liters of this medium was inoculated with Escherichia coli and the bacteria grown to 2 x lo"/ ml at 33". These cells were infected with T4D so that there was one T4D per three bacteria and the culture shaken for 16 h until lysis was complete.
The phage particles were purified in the usual manner, osmotically shocked to release their DNA, and purified further by two cycles of density gradient centrifugation.   Zinc,Cobalt,and Nickel in T4D Phage phages grown in the presence of zinc, while phages grown in the presence of nickel were even more heat resistant than the cobalt grown phages. These patterns of baseplate sensitivity support the conclusion that these different metal ions are actually incorporated into the tail plate and that zinc can be replaced by cobalt or nickel. Experiment IV shows the effect of another chelating agent, dimethylglyoxime, under somewhat different conditions on the heat stability of these three phage preparations.
Dimethylglyoxime is highly specific for binding nickel ions and is used analytically for this purpose. T4D grown in the presence of Ni'+ ions was highly labilized by this reagent as compared to the effect of this reagent on phage grown in the presence of cobalt or zinc ions.
Martell and Calvin (36) have pointed out that the affinity of almost all chelating agents for these 3 metal ions is nickel > cobalt > zinc, i.e. these agents bind Ni'+ more tightly than Co'+ and in turn bind Co'+ more tightly than Zn'+ ions. However, in the experiments given in Table II, the metal ions are not free in solution, but are part of the protein structure which itself is chelating the metal ion with affinities determined again by the nature of the protein binding groups and by the specific metal. In these experiments, there is competition for the metal ion between the ligand bonds of the protein and ligand bonds of the added agent. In Experiments II and III, it appears that the affinity of the viral proteins for Co'+ and Ni*+ is much higher than their affinity for zinc and, therefore, both o-phenanthroline and Snitro-&hydroxyquinoline inactivated phage containing zinc more readily than phage containing cobalt or nickel. In Experiment IV, the unique and very high affinity of dimethylglyoxime for NiY+ ions is apparent and this agent rapidly inactivated T4D grown '*+ in the presence of Ni ions while the other two phage preparations were less sensitive to this reagent.
Affinity constants for o-phenanthroline and dimethylglyoxime for Zn'+, Co*+, and NY'+ at 20-25" in various solutions (dioxane or alcohol is used for measurements with dimethylglyoxime) have been gathered by Martell (37). (No affinity constants are available for 5-nitro-8-hydroxyquinoline.) Fig. 1 shows the relationship of these aftinity constants and the rate of phage inactivation from Table II. The data in this figure give an indication of the stability constants that this phage substructure has for these 3 different metal ions as compared to the affinity constants for the metal ions in free solution. It is apparent that these metal ions are extremely tightly bound to the phage structure and the affinity is related to the specific metal ion bound in the phage structure. For example, even when heated in the presence of o-phenanthroline, T4D(Ni) is more than twice as stable as T4D(Zn), even though the stability constant of the o-phenanthroline. nickel complex is two logs higher than that of the o-phenanthroline.zinc complex. However, dimethylglyoxime's unique ability to react with protein-bound nickel does compete effectively with the affinity of the nickel ions for the phage proteins.
Inactivation of Various Phage Preparations by Cd(CN),and Hog Kidney Conjugase -T4D grown in the presence of zine, cobalt, or nickel was treated with two reagents known to attack the phage baseplate. Cd(CN),-was shown in 1955 (38) to specifically alter phage tail structure. Recently, it has been pointed out that this reagent probably competes for the same protein binding sites as the normal phage baseplate zinc ions (20). Fig. 2 shows the effect of this reagent on purified phage grown in the presence of 2 x lo-" M zinc, cobalt, or nickel ions. The assay error in this experiment with Cd(CN),-and the experiment described below is about 10%. Zinc-grown T4D was more sensitive to inactivation by this reagent than was cobalt-grown T4D, which in turn was more sensitive than nickel-grown T4D. These results again indicate that these ions can substitute for each other in the phage baseplate; presumably, the added Cd(CN),-competed most effectively with the native zinc and least effectively with the incorporated nickel.
The action of hog kidney folic acid conjugase (a y-glutamyl peptidase) on three different phage preparations is shown in Fig. 3. It has been shown previously that T-even phage  Table I) is compared to the log of the stability constant (37) for the complex formed between the chelating agents and the metal(U) ion present in the medium when the phages were grown. The horizontal arrows for the dimethylglyoxime data reflect the range of stability constants given by Martell (3'7) for cobalt and nickel with this chelating agent.  baseplates contain dihydropteroylhexaglutamate as a structural component (11)(12)(13). Phages grown in the presence of cobalt or nickel ions were completely resistant to this enzyme, while T4D grown in the presence of added zinc ions was sensitive. (The survival of 10% of the T4D(Zn) in this type of treatment has been attributed by Dawes and Goldberg (9) to inactivation of the enzyme during incubation.) Not only does this sensitivity of only T4D(Zn) to hog kidney conjugase support the conclusion that cobalt and nickel are incorporated into the different T4D preparations, but it offers the first indication of the structural relationship of the phage baseplate folic acid to the baseplate metal ion. Typically T-even phages are grown in media presumed or known to contain zinc and all have been shown to be sensitive to this enzyme (11-131. The observation that substitution of cobalt or nickel for zinc alters the baseplate structure so that such phages are resistant to attack by this enzyme not only would be in accord with the ability of cobalt and nickel to tightly bind to baseplate proteins and possibly alter their configuration, but it also suggests that the baseplate metalloprotein is physically close to the hexaglutamyl residues in the baseplate. Direct Demonstration of Incorporation of Cobalt into T4D Particles -The results described above indicated that cobalt and nickel can substitute for zinc. The incorporation of cobalt into T4D particles was determined directly by an isotope dilution method using synthetic medium supplemented with 1.9 x lo-" M 5HCoC1,. E. coli bacteria were grown in this medium, infected with T4D, and the resulting phage particles purified and then osmotically shocked to remove the DNA. These ghost particles were purified further by two density gradient centrifugations. The association of "Vo radioactivity with the ghost particle is shown in Fig. 4. The amount of ghost protein was determined by its fluorescence and this was related to a phage preparation of known particle number (20). In Fig. 4, 6.2 x 10" particles/ml give a fluorescent value of 0.5. From the known specific activity of the added 5RC~, it could be readily calculated that these T4D particles had about 3.7 atoms of cobalt/particle. This value is probably slightly low since some of the phage ghost particles had damaged tail baseplates as observed in the electron microscope. It seems reasonable to conclude that T4D ghosts contain 4 2 1 atoms of cobalt/particle.
It should be noted that this value is quite similar to the value of 5 f 1 zinc atomslT4D ghost particle reported earlier (20).

DISCUSSION
All the observations in this report are consistent with the conclusion that Co'+ and Ni2+ can substitute in vivo for Zn'+ and be incorporated into the phage baseplate. Together with the fact that cobalt and nickel can efficiently reconstitute phages treated with o-phenanthroline in vitro (201, the evidence here is so specific that it would exclude other interpretations of the properties of phages grown in these different media. These observations include: (a) that the baseplate is structurally different, depending on the metal ion added to the growth medium and that there is a correlation of these structural changes with the chemical properties of the metal ion; (b) that these changes in properties indicate a highly specific location of the metal in the baseplate; and (c) that the amount of cobalt actually incorporated into phage particles is about the same as the amount of zinc found in phages grown in zinc-containing media. Direct analysis of phage for NiS+ as a replacement for Zn'+ depresses the phage yield. Spectral evidence for the presence of nickel in urease has been reported (391, but this optical procedure cannot be used with phage preparations since the observed spectral peaks for nickel in urease at 316, 425, and 725 nm would require phage preparations for a comparable optical reading to have concentrations of lO"'/ml, which is about 4 g/ml. Similarly, baseplate preparations would be needed at about 2 to 300 mg/ml for spectral examination, which is also not currently possible. However, the unique effect of dimethylglyoxime in destabilizing nickel-grown phages, but not zinc-or cobalt-grown T4D phages, is strong independent evidence for incorporation of nickel into the phage baseplate.
Some of the biochemical features of this system should be emphasized. Apparently, raising the concentration of cobalt or nickel to 2 x lo-" M in the M-9 medium is sufficient to favor almost the complete exclusion of the normal zinc ions from the virus particle.
However, the properties of the phage preparations reported here do not exclude the possibility that a small fraction of the phage particles, possibly lo%, still contains only zinc or even that individual particles may be hybrids and contain various proportions of zinc to cobalt or nickel. The ability to substitute cobalt or nickel for zinc either in uiuo or in vitro should aid in determining the physiological role of this metal both in maintaining the phage structure and in facilitating the introduction of the phage DNA into the host cell.