The structure of isometric capsids of bacteriophage T4.

The three-dimensional structure of DNA-filled, bacteriophage T4 isometric capsids has been determined by means of cryoelectron microscopy and image reconstruction techniques. The packing geometry of protein subunits on the capsid surface was confirmed to be that of the triangulation class T = 13. The reconstruction clearly shows pentamers, attributed to capsid protein gp24*, surrounded by hexamers of the major capsid protein, gp23*. Positions of the accessory proteins, Hoc and Soc, are also clearly delineated in the surface lattice. The Hoc protein is the most prominent surface feature and appears as an extended molecule with a rounded base from which a thin neck and a globular head protrude. One Hoc molecule associates with each hexamer. Nearly continuous "ridges" are formed at the periphery of the gp23* hexamers by an association of 12 Soc molecules; however, Soc is absent along the boundaries between the hexamers and the pentamers. The duplex DNA genome forms a highly condensed series of concentric layers, spaced about 2.36 nm apart, that follow the general contour of the inner wall of the protein capsid.

Bacteriophage T4 self-assembly: localization of gp3 and its role in determining tail length.

Gene 3 of bacteriophage T4 participates at a late stage in the T4 tail assembly pathway, but the hypothetical protein product, gp3, has never been identified in extracts of infected cells or in any tail assembly intermediate. In order to overcome this difficulty, we expressed gp3 in a high-efficiency plasmid expression vector and subsequently purified it for further analysis. The N-terminal sequence of the purified protein showed that the initial methionine had been removed. Variant C-terminal amino acid sequences were resolved by determining the cysteine content of the protein. The molecular mass of 20.6 kDa for the pure protein was confirmed by Western blotting, using a specific anti-gp3 serum for which the purified protein was the immunogen. We also demonstrated, for the first time, the physical presence of gp3 in the mature T4 phage particle and localized it to the tail tube. By finding a nonleaky, nonpermissive host for a gene 3 mutant, we could clearly demonstrate a new phenotype: the slow, aberrant elongation of the tail tube in the absence of gp3.

Modulation of bacteriophage T4 capsid size.

Bacteriophage T4 capsid assembly requires the vertex protein (gp24). Mutations that bypass this requirement are found in gene 23, which produces the major capsid protein (gp23). The latter were used to study the role of gp24 in head length control. We found that gp24 is no longer present in the capsids of several gp24 bypass mutants. We measured the capsid lengths of several of these bypass mutants, because gp24 had been reported to be implicated in head length control. One bypass mutant (reported in 1977) produced 40-60% short headed ("petite") phage in the presence of wild-type amounts of gp24. The bypass mutations, when combined with amber mutations in gene 24, produced normal size heads in either suppressor or nonsuppressor host bacteria. When several known bypass mutations were back-crossed with wild-type phage, one-third of the byp/24wt mutants isolated produced large amounts of petite phage, indicating that the ability to produce petite phage is a general property of the bypass mutations. Sequencing several of these bypass mutants showed that those that produced petite phage contained at least one additional missense mutation in gene 23. This suggests that gp24 itself has no direct role in head length regulation, but that in the presence of bypass 24 mutations and certain easily acquired gene 23 mutations (called trb) the gp23-gp24 interactions can modulate head length.

DNA enzymology above 100 degrees C. Topoisomerase V unlinks circular DNA at 80-122 degrees C.

The widespread application of polymerase chain reaction and related techniques in biology and medicine has led to a heightened interest in thermophilic enzymes of DNA metabolism. Some of these enzymes are stable for hours at 100 degrees C, but no enzymatic activity on duplex DNA at temperatures above 100 degrees C has so far been demonstrated. Recently, we isolated topoisomerase V from the hyperthermophile Methanopyrus kandleri, which grows up to 110 degrees C. This novel enzyme is similar to eukaryotic topoisomerase I and acts on duplex DNA regions. We now show that topoisomerase V catalyzes the unlinking of double-stranded circular DNA at temperatures up to 122 degrees C. In this in vitro system, maximal DNA unlinking occurs at 108 degrees C and corresponds to complementary strands being linked at most once. These results further imply that in the presence of sufficient positive supercoiling DNA can exist as a double helix even at 122 degrees C.

Tail length determination in bacteriophage T4.

This report identifies a protein that regulates tail length in bacteriophage T4. Earlier work (Duda et al., 1990) suggested that the gene 29 protein could be involved in T4 tail length determination as a "template" or "tape-measure", similar to that proposed for the gene H protein in bacteriophage lambda. We have altered the length of a recombinant gene 29 by constructing deletions and duplications in different parts of the gene. Each of these constructs was incorporated into the high-level expression vector, pET-11d. Seven plasmids with different lengths of gene 29 were made and used in complementation studies. We have found that the length of the tail can be decreased by deleting the C-terminal part of gene 29 or increased by forming duplications in different parts of this gene, and that the length of the tail can be proportional to the size of the engineered protein. Unlike phage lambda, plasmids with deletions in the middle of gene 29 or from the N-terminal end produced correspondingly smaller but inactive gene 29 protein and no viable phage were formed. Our results show that alterations in the length of gene 29 protein proportionately alters tail length, and argue strongly for a scheme in which 29 protein is a ruler or template that determines tail length during tail assembly.

gp160, the envelope glycoprotein of human immunodeficiency virus type 1, is a dimer of 125-kilodalton subunits stabilized through interactions between their gp41 domains.

The molecular masses, carbohydrate contents, oligomeric status, and overall molecular structure of the env glycoproteins of human immunodeficiency virus type 1--gp120, gp160, and gp41--have been determined by quantitative electron microscopy. Using purified gp160s, a water-soluble form of env purified from a recombinant vaccinia virus expression system, we have measured the masses of several hundred individual molecules by dark-field scanning transmission electron microscopy. When combined with sequence-based information, these mass measurements establish that gp160s is a dimer of subunits with an average monomer mass of 123 kDa, of which approximately 32 kDa is carbohydrate and 91 kDa is protein. Similarly, gp120 was found to be a monomer of 89 kDa and to contain virtually all of env's glycosylation. gp41 is glycosylated only slightly, if at all, and is responsible for the interactions that stabilize the gp160s dimer. A molecular mass map of gp160s derived by image processing depicts an asymmetric dumbbell whose two domains have masses of approximately 173 and approximately 73 kDa, corresponding to a gp120 dimer and a gp41 dimer, respectively. We infer that the average monomer mass of native gp160 is 125 kDa and that in situ, env is either a dimer or a tetramer but is most unlikely to be a trimer.

Expression of plasmid-encoded structural proteins permits engineering of bacteriophage T4 assembly.

A complementation system for studying bacteriophage T4 tail assembly has been developed and used to test the effects of nonviable mutations on the function of a specific T4 tail protein, gp48. The complementation system assays the assembly function of gp48 without requiring that viable phage be produced, circumventing the operational problems of maintaining nonviable mutants of this lytic bacteriophage. The protein to be tested was preexpressed from cloned genes in a host cell prior to infection with the challenge phage. Assembly activity was assayed by monitoring the conversion of one tail assembly intermediate, the baseplate lacking gp48, into baseplates containing gp48 or into tube baseplates (or sheathed tails) assembled from such baseplates. Specific incorporation of gp48 into these structures was confirmed using gp48-specific antiserum, and the same serum was used in direct immunoelectron microscopy experiments to localize gp48 to the baseplate-proximal end of the T4 tail tube, at the site where the tube and sheath bind to the baseplate. The protein gp48 has been previously shown to be a baseplate protein, as well as a tail-tube-associated protein, and was tested for a possible role as a tail-length tape-measure protein. Tests with a deleted variant of gp48 were inconclusive because the protein was inactive. A variant of gp48, 20% longer than wild-type protein due to an internal duplication, was found to be partly functional in our assembly complementation system. This abnormally elongated protein allows several assembly steps to proceed, including the assembly of normal length T4 tails, implying that it does not specify tail length. The insertion-duplication variant of gp48 appears to have a defect in its interaction with the tail sheath protein, leading to abnormal sheath contraction.

Expression and regulation of genes coding for three bacteriophage T4 tail tube-associated proteins.

The assembly and length regulation of the tail tube of bacteriophage T4 requires the function of three proteins: gp29, 48, and 54. Six copies of each protein are found in the completed tail, and the genes for these proteins are adjacent on the T4 genome. Evidence is presented here that gp54 is also a tail tube-associated protein that remains bound to the tail tube after the baseplate is removed by guanidine hydrochloride, suggesting that all three proteins interact structurally. There is a strong polar effect of translation termination mutants in gene 48 upon the expression of the adjacent gene 54, in cis-trans tests. Gene dosage experiments that assay the in vivo expression of these genes show that only gene 48 is expressed at slightly higher than stoichiometric levels during T4 infection. Genes 48 and 54 were placed under the control of a T7 promoter and the corresponding proteins identified. When a frameshift mutation was introduced into gene 48, neither gp48 nor gp54 was made. Transcriptional termination was not the explanation of this result because genes distal to 48 and 54 in the plasmid were expressed. These data suggest that expression of genes 48 and 54 is translationally coupled.

The structure of three bacteriophage T4 genes required for tail-tube assembly.

Three different protein molecules copurify with T4 tail tubes after the tubes are released from the baseplate by guanidine hydrochloride treatment. These tube-associated proteins (TAPs) are the products of genes 29, 48, and 54. To further investigate the structural roles that these proteins may play in T4 tail assembly we have cloned and sequenced the genes coding for these proteins and have deduced their predicted amino acid sequences. The sequence data reveal a region of amino acid sequence similarity between gp54 and the T4 tail-tube structural protein, gp19. We believe that this region of similarity is significant and consistent with the role gp54 may play in initiating T4 tail-tube polymerization.

Mutations that affect structure and assembly of light-harvesting proteins in the cyanobacterium Synechocystis sp. strain 6701.

The unicellular cyanobacterium Synechocystis sp. strain 6701 was mutagenized with UV irradiation and screened for pigment changes that indicated genetic lesions involving the light-harvesting proteins of the phycobilisome. A previous examination of the pigment mutant UV16 showed an assembly defect in the phycocyanin component of the phycobilisome. Mutagenesis of UV16 produced an additional double mutant, UV16-40, with decreased phycoerythrin content. Phycocyanin and phycoerythrin were isolated from UV16-40 and compared with normal biliproteins. The results suggested that the UV16 mutation affected the alpha subunit of phycocyanin, while the phycoerythrin beta subunit from UV16-40 had lost one of its three chromophores. Characterization of the unassembled phycobilisome components in these mutants suggests that these strains will be useful for probing in vivo the regulated expression and assembly of phycobilisomes.

Asymmetrical core structure in phycobilisomes of the cyanobacterium Synechocystis 6701.

The light-harvesting complex of cyanobacteria and red algae, the phycobilisome, has two structural domains, the core and the rods. Both contain biliproteins and linker peptides. The core contains the site of attachment to the thylakoid membrane and the energy transfer link between the phycobilisome and chlorophyll. There are also six rod-binding sites in the membrane-distal periphery of the core. The structure of phycobilisomes in the cyanobacterium Synechococcus 6301 was studied by Glazer, who proposed a model for the internal organization of the bicylindrical core. In the construction of that model, it was necessary to make arbitrary decisions between two possible locations for one of the trimeric protein complexes within a core cylinder and between two possible orientations of the basal core cylinders relative to one another. We isolated the tricylindrical cores from an ultraviolet-light-induced mutant of the cyanobacterium Synechocystis 6701 and obtained, by partial dissociation, a unique core substructure that maintained some contacts between the two basal cylinders. From its structure and spectral properties, we conclude that this particle is a central core substructure that resulted from dissociation of the two layers of peripheral trimers in the intact core. The compositions of this particle and the dissociated trimers were inconsistent with the proposed location of one of the trimers in the 6301 core model, but supported the placement of that trimer in the alternative position within the basal core cylinder. Rod-binding sites within the central core substructure were studied by partial dissociation of the short-rod phycobilisomes from another mutant of 6701. This dissociation generated particles that were interpreted as being central core substructures with the two basal rods attached. The appearance of these particles in the electron microscope suggested that both basal rods would be localized towards the same side of the intact core. Such an asymmetrical arrangement of basal rods is supported by previously published edge-views of intact cores with basal rods from strain 6701. These observations suggest a parallel arrangement of the basal cylinders with respect to each other, creating an asymmetrical core. A phycobilisome model was constructed that incorporated core asymmetry. This model predicts the energy transfer pathways from the basal and upper rods to specific trimers in the core.

Potential length determiner and DNA injection protein is extruded from bacteriophage T4 tail tubes in vitro.

Bacteriophage T4 tails contain a set of extended protein molecules in the central channel of the tail tube through which the DNA must exit during infection. Treatment of tails with guanidine hydrochloride separates the baseplates, leaving the tail tube and several specific tube-associated proteins. Methods were developed to purify these structures. Using specific antisera, immunoblotting, and electrophoretic analysis, these structures were shown to contain proteins gp19, 29, and 48. Electron microscopy showed specifically defined stain penetration into the tail tube, a bulge at one end, and a short fiber extruded from the tube. These structures could be removed by proteases but the gp19 tube itself was resistant. Structural studies of tails and intact phage show that the bulge and fiber are at the end of the tube that interacts with the cell membrane during infection. Since the fiber did not protrude from baseplates or from incomplete (short) tube-baseplates, we propose that it is first assembled as a compact structure formed of six copies of a tube-associated protein, which elongates during tail tube formation to fill the central channel, span the length of the tube, and regulate its length. We suggest that the exit of this fiber during infection signals DNA ejection.

Mass distribution of a probable tail-length-determining protein in bacteriophage T4.

Analysis of dark-field scanning transmission electron micrographs of unstained freeze-dried specimens established that the interior of the intact bacteriophage T4 tail tube contains extra density that is missing in tubes artificially emptied by treatment with 3 M guanidine hydrochloride. The mass of the tail tube is 3.1 X 10(6) daltons, and the central channel is 3.2 nm in diameter. Quantitative analysis of the density data is consistent with the presence of up to six strands of a protein molecule in the central channel that could serve as the template or ruler structure that determines the length of the bacteriophage tail and that could be injected into the cell with the phage DNA.

Intracellular morphogenesis of bacteriophage T4. I. Gene dosage effects on early functions and tail fiber assembly.

Mixed infection of nonpermissive bacteria by amber mutant and wild-type T4 phage reduces the burst size in a manner that depends on the nature of the gene product eliminated by the mutation. While the genotypic proportions produced by such mixedly infected cells were directly proportional to those in the starting mixtures, the fraction of the burst size compared to wild-type-infected cells was dependent on the kind of mutant used. For example, mutants in genes 56, 42, and 43, which control catalytic functions for DNA synthesis, produced normal burst sizes even at ratios of four mutant to one wild-type phage. The multiplicity of infection in these experiments was near 20, and most cells received one wild-type phage. By contrast, infection with mutants in tail fiber structural genes gave a normal burst size with 1:1 mixtures of wild-type phage, but showed a decrease in burst size below this ratio. The conclusions drawn are that enzymes of DNA replication are present in large excess over the amount needed for phage DNA synthesis, and that tail fiber structural proteins are made in a three- to fourfold excess over the amount needed for phage maturation under normal laboratory conditions.

Intracellular morphogenesis of bacteriophage T4. II. Head morphogenesis.

The relative phage yields of cells of Escherichia coli infected with both wild-type and amber mutant phages deficient in head morphogenesis were determined. The decrease in burst size as a function of the ratio of mutant:wild-type-infecting phage was linear and proportional for mutants in genes 20, 22, and 23, while for mutants in genes 21, 31, and 24 the results suggest an excess of intracellular gene product. The initiation of assembly of phage particles was not delayed at reduced gene product levels; only a reduction in the rate of phage assembly was observed. The effects on burst size of pairs of mutations in genes 20 and 23, 22 and 23, and 22 and 24, in both cis and trans arrangements, were identical. Experiments using the mutant E920g in gene 23 show that varying the kind and intracellular amounts of the major capsid protein (gp23) with respect to the major core or scaffold protein (gp22) had a profound effect on the length of the T4 head. Head length determination must therefore depend on the proper intracellular balance between these two proteins.

Regulation of Nostoc sp. phycobilisome structure by light and temperature.

Nostoc sp. strain MAC cyanobacteria were green in color when grown in white light at 30 degrees C and contained phycobilisomes that had phycoerythrin and phycocyanin in a molar ratio of 1:1. Cells grown for 4 to 5 days in green light at 30 degrees C or white light at 39 degrees C turned brown and contained phycoerythrin and phycocyanin in a molar ratio of greater than 2:1. In addition to the change in pigment composition, phycobilisomes from brown cells were missing a 34.5-kilodalton, rod-associated peptide that was present in green cells. The green light-induced changes were typical of the chromatic adaptation response in cyanobacteria, but the induction of a similar response by growth at 39 degrees C was a new observation. Phycobilisomes isolated in 0.65 M phosphate buffer (pH 7) dissociate when the ionic strength or pH is decreased. Analysis of the dissociation products from Nostoc sp. phycobilisomes suggested that the cells contained two types of rod structures: a phycocyanin-rich structure that contained the 34.5-kilodalton peptide and a larger phycoerythrin-rich complex. Brown Nostoc sp. cells that lacked the 34.5-kilodalton peptide also lacked the phycocyanin-rich rod structures in their phycobilisomes. These changes in phycobilisome structure were indistinguishable between cells cultured at 39 degrees C in white light and those cultured at 30 degrees C in green light. A potential role is discussed for rod heterogeneity in the chromatic adaptation response.

Bacteriophage SPO1 structure and morphogenesis. I. Tail structure and length regulation.

Bacteriophage SPO1, a structually complex phage with hydroxymethyl uracil replacing thymine, has been studied by structural and chemical methods with the aim of defining the virion organization. The contractile tail of SPO1 consists of a complex baseplate, a tail tube, and a 140-nm-long sheath composed of stacked disks (4.1 nm repeat), each containing six subunits of molecular weight 60,300. The subunits are arranged in six parallel helices, each with a helical screw angle (omega 0) of 22.5 degrees. The baseplate was shown to undergo a structural rearrangement during tail contraction into a hexameric pinwheel. A mutation in gene 8 which produced unattached heads and tails also produced tails of different lengths. The tail length distribution suggests that the smallest integral length increment is a single disk of subunits. The structural arrangement of subunits in long tails is identical to that of normal tails, and the tails can contract. Many of the long tails showed partial stain penetration within the tail tube to a point which coincides with the top of a unit-length tail. The implications of these findings with respect to tail length regulation are discussed.

Bacteriophage SPO1 structure and morphogenesis. II. Head structure and DNA size.

The capsid of bacteriophage SPO1 is icosahedral, and the subunit arrangement on the 87-nm-diameter head suggests the triangulation number T = 16. The major capsid protein (45,700 daltons) is cleaved from a 47,700-dalton precursor. Tubular heads (polyheads) are produced by mutations in genes 5 and 8 and contain cores as well as capped ends. The lattice constant of these structures is 13.4 nm; diameter is 109.5 nm. The size of the double-stranded SPO1 DNA (containing 5' hydroxymethyl uracil in place of thymine) was measured by sedimentation analysis and electron microscopy and has a molecular weight of 86 X 10(6) (about 140 kilobase pairs), which is smaller than several previously reported values.

Bacteriophage SPO1 structure and morphogenesis. III. SPO1 proteins and synthesis.

The virion proteins of SPO1 have been determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis methods on purified phage components and on phage lysates. The phage head contains 16 proteins, and the connector or neck structure has an additional 3 proteins not found in the head. The proximal part of the tail, composed of sheath, tube and connecting components, contains six proteins. The distal baseplate is the most complex structure, with 28 proteins identifiable on sodium dodecyl sulfate gels. The maximum number of proteins found in phage subassemblies is 53, which would account for nearly half the coding capacity of the SPO1 genome.

Evidence for an internal component of the bacteriophage T4D tail core: a possible length-determining template.

The length of the T4 tail is precisely regulated in vivo at the time of polymerization of the tail core protein onto the baseplate. Since no mutations which alter tail length have been identified, a study of in vivo-assembled tail cores was begun to determine whether the structural properties of assembled cores would reveal the mechanism of length regulation. An assembly intermediate consisting of a core attached to a baseplate (core-baseplate) was purified from cells infected with a T4 mutant in gene 15. When core-base plates were treated with guanidine hydrochloride, cores were released from baseplates. The released cores had the same mean length as cores attached to baseplates. Electron micrographs of these cores showed partial penetration of negative stain into one end, and, at the opposite end, a modified tip which often appeared as a short fiber projecting from the core. When cores were purified and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, two minor proteins and the major core protein were detected. One minor protein, the product of gene 48 (gp48), was present in at least 72% of the amount found in core-baseplates, relative to the amount of the major core protein. These findings suggest that cores contain a fibrous structure, possibly composed of gp48, which may form a "ruler" that specifies the length of the T4 tail.