This is to certify that the thesis entitled ‘ , ' 7176-é/1LQ. . é 1W6 BaCYlQHOP/L 0% [refie ld\ . frdag/g/fjé fihf€€/r€W/‘7 fV/V‘V‘Jm‘ P )1 W present dby " I fflr/ Swot/fl“— has been accepted towards fulfillment of the requirements for PM who/077 Major professor [)3th >779 0-7 639 TWO BACTERIOPHAGE T4 GENE PRODUCTS WHICH REGULATE GENE EXPRESSION AFFECT DNA STRUCTURE: I. The Alc Gene Product Unfolds the Host Nucleoid II. The PseT Gene Product Can "Shuttle" DNA Phosphates from 3' to 5' Termini By Karl Sirotkin A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1978 ABSTRACT TWO BACTERIOPHAGE T4 GENE PRODUCTS WHICH REGULATE GENE EXPRESSION AFFECT DNA STRUCTURE: I. The Alc Gene Product Unfolds the Host Nucleoid II. The PseT Gene Product Can "Shuttle" DNA Phosphates from 3' to 5' Termini By Karl Sirotkin When T4 polymerizes cytosine (instead of its usual hydroxyl- methylcytosine) into its DNA it does not produce any progeny because it does not express its true-late genes normally. However, previous data indicate that T4's alg_gene can be altered so that T4 with cytosine containing DNA develops normally. Such gig? T4 express combinations of the following pleotrophic defects: 1) defectively unfolding the chromosome of its host, g. ggli; 2) failing to induce a polypeptide that normally co-purifies with the host RNA polymerase; and 3) defectively preventing host transcription. The selection that yields T4 that induce altered alg gene products sometimes yields T4 that induce altered Egg: gene products as well. (The T4 Egg: gene has previously been shown to be necessary for the induction of 3' phosphatase activity and for growth on an .g..ggli strain, CTer). Genetic evidence is presented here that Egg: is the structural gene, not only for 3' phosphatase, but also for 5' polynucleotide kinase. In addition, the pseT gene product Karl Sirotkin is required for normal T4 true-late gene expression during infection 0f.§- coli CTer. Adding efficient amber suppressing ability to .E- coli CTer alters it so that all pseT T4, even deletions, can grow on it. Thus, an amber mutation in an E, coli CTer gene prob- ably prevents the growth of pseT T4. It follows then that the nor- mal functioning of a similar host gene in commonly used laboratory strains is probably required for growth of pseT T4 on those strains. ACKNOWLEDGMENTS I would like to thank Drs. H. Sadoff, R. Costilow, L. Robbins and J. Hanna for serving on my committee. Special thanks to Dr. Costilow for the use of some of his laboratory equipment. Of course, my deepest gratitude to Dr. Larry Snyder, without whose patient tutelage I could not have had such a rewarding appren- ticeship. I also acknowledge a three year National Science Foundation predoctoral fellowship. My gratitude is also expressed to Drs. R. Patterson and H. Struck for helping me improve my writing skills. ii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . v LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . vi INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . 1 LITERATURE REVIEW . . . . . . . . . . . . . . . . . . . . . . . 4 Isolated Bacterial Nucleoids . . . . . . . . . . . . 4 Nucleoid Structure . . . . . . . . . . . . . . 5 Nucleoids: Artifact or Biologically Significant? . 8 Nucleoids as a Probe of Replication Related Events . . . 9 Nucleoids as Transcription and Replication Templates . . ll Enzymes Altering DNA Supercoiling . . . . . . . . . . . . . 12 Their Biological Significance . . . . . . . . . . . . . 13 In vitro Observations . . . . . . . . . . . . . . . . . l4 Intermediates and Models . . . . . . . . . . . . . . 16 The Relationship of Nicking or Capping Enzymes to Gene Expression . . . . . . . . . . . . . . . . . . . . . . . . 17 Requirements for T4 True—late Gene Expression . . . . . . . 18 T4 Induced Modifications of RNA Polymerase . . . . . . . . 19 Is alc the Structural Gene for Polypeptide Number 2? . . 19 The Structure of T4 DNA after Replication has Begun . . . . 23 T4 nucleoids . . . . . . . . . . . . . . . 24 Effects of Cytosine Substitution in T4 DNA . . . . . 25 T4 Mediated Alterations of Host DNA Organization and T4' 3 Effect upon Host Gene Expression . . . . . . . . . . . . . 26 Host Chromosome Disruption and Unfolding . . . . . 26 Does the alc Gene Product Unfold the Host Nucleoid? . . 27 The Shutoff of Host Gene Expression . . . . . . . . . 27 Does the alc Gene Product Prevent Transcription of host RNA . . . . . . . . . . . . . . . . 28 Models for ale and pse TGene Product Function . . . . . . . 29 Recommendations . . . . . . . . . . . . . . . . . . . . . . 31 REFERENCES 0 O O O O O O O O O O O O O O O O O O O O O O O O O C 34 ARTICLE I . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 A ROLE IN TRUE-LATE GENE EXPRESSION FOR THE T4 BACTERIOPHAGE 5' POLYNUCLEOTIDE KINASE 3' PHOSPHATASE iii Page APPENDIX I O O O O O O I O O O O O C O O O O O O O O O O O O O 82 APPENDIX II 0 O O O I O O O O O O O O O O O O O O O O O O O O O 87 APPENDIX III . . . . . . . . . . . . . . . . . . . . . . . . . . 92 iv LIST OF TABLES Table 1 Strains Used . . . . . . . . . . . . . . . . . . . . 2-A 3' Phosphatase Assays . . . . . . . . . . . . . . . . . 2—B 5' Polynucleotide Kinase Assays . . . . . . . . . APPENDIX I 1 Map Position of alc- Mutations . . . . . . . . . . . . 2 Hybridisation of RNA Labelled after Infection by an alc- Mutant to £0 c011 DNA 0 0 O O O O O O O O O O O O O O Page 50 54 55 84 86 LIST OF FIGURES Figure 1 A T4 Induced Polypeptide, Migrating with the Same Molecular Weight as 5' Polynucleotide Kinase is Missing in Extracts from pseTZ Infected Cells . . . 2 Late Proteins are Synthesized at a Lower Rate after pseTAl Infection of E, coli CTer . . . . . . . . . . 3 Complementation of pseTl on E. coli CTer and the Effect of Efficient Amber Suppression . . . . . . . . 4 The Model Categories Depicting Possible pseT Gene Product Functions . . . . . . . . . . . . . . . . . . APPENDIX I 1 An alc- Mutant is Lacking a T4-coded polypeptide on the RNA Palymerase O O O O O O O O O O O O O O O O O 2 Sucrose Gradients of Nucleoids after Infection by an 31C- Mlltant o o o o o o o o o o o o o o o o o o 3 Time Course of Unfolding after Infection by alc- Mutants . . . . . . . . . . . . . . . . APPENDIX II 1 Fine Structure Map of pseT and its Position with' Respect to Other Nearby Genes . . . . . . . . . . . . vi Page 58 61 65 74 83 85 85 87 INTRODUCTION Studies of bacteriophage T4 transcription may be used to ad- vance our knowledge of gene expression and its relationship to DNA structure because T4 drastically affects both the gene expression and DNA structure of its host, E. 991;. After T4 infection, host gene expression ceases, host DNA organization changes, and, even- tually, the host DNA itself largely disappears. Another reason to use T4 to study the relationship between DNA structure and function is that T4 induces many enzymes that are essentially duplications of host enzymes. For example, T4 directs the synthesis of many of the gene products necessary for the replication of its DNA (for example, see Alberts, et al., 1975). These duplications can make it easier to study functions as altered by phage mutants, since host mutants can be more difficult to isolate. In addition, the regulation of T4's own gene expression is com- plex and interesting (for a review see Rabussay and Geiduschek, 1977). At least four classes of T4 genes have been described: immediate early, delayed early, quasi-late, and true-late. The early genes are transcribed almost exclusively from T4's l—strand. Models for their control based upon either initiation at specific promoter se- quences or termination at specific sequences can be used to explain existing data. The true-late genes are transcribed, generally, from T4's r-strand. Less is known about the control of true-late tran- scription. However, it is known that the expression of T4's true- late genes seems to depend on specific DNA structures. The 313 and pggl_genes are examples of T4 genes that can be exploited to study gene expression and its relationship to DNA structure. The 313 gene product normally blocks T4 true-late gene expression when cytosine (instead of the usual hydroxylmethylcytosine) is polymerized into T4 DNA (Snyder et al., 1976). The T4 alg_gene product is also required to "unfold" the host's chromosome and to block transcription of host RNA (Sirotkin et al., 1977 - appendix 1). The alg gene may induce an RNA polymerase subunit (Sirotkin et al., 1977 - appendix 1). The selection that yields T4 212: mutants sometimes yields T4 that are pgglf as well (table 1 of article 1). In addition, when T4 that are only 3127 are propagated, psglf mutants are often re- covered (table 1 of article 1). Studies presented in article 1 demonstrate several properties of the p§32_gene: l) the pggg_gene product can, in vitro, remove 3' phosphates from DNA termini and add gamma phosphates from ATP to 5' hydroxyl termini; thus it can be said to "shuttle" phosphates; 2) a host gene product can substi- tute for the pggI_gene product; 3) the Egg; gene product or its sub- stitute, is required for normal T4 true-late gene expression. Normally T4 true-late gene expression depends upon recently replicated DNA. The putative substitute for the Egg: gene product can therefore be expected to support host gene expression that de- pends upon recently replicated DNA. Further research is currently using pgng mutants in a selection intended to yield host mutants in the gene(s) corresponding to the pseT gene (see appendix 3). It is not clear how either the alg_or p§33_gene function in 2122: although some models to explain their function can be suggested (see literature review and article 1). For example, the alg_gene product might function either as an RNA polymerase subunit that blocks transcription of cytosine containing DNA, or as an activity that removes the supercoiling of cytosine containing DNA. Some models for Egg: function describe the end products of phosphate "shuttling" observed in vitgg_as intermediates appearing only tran- siently in givg. For example, these intermediates may appear as DNA supercoils are created or destroyed. Thus to provide an adequate background for the understanding of the alg_and‘p§gz_gene products, it is necessary to summarize research concerning DNA supercoiling. In addition, research in five areas more obviously related to the alg_and p§32_gene products will be discussed: 1) isolated bacterial nucleoids; 2) the structure of replicating T4 DNA; 3) T4 induced modifications of the host RNA polymerase; 4) T4's effect on the or- ganization of host DNA and host gene expression; and 5) the regula- tion of T4 true-late gene expression. Research exploring the 5' polynucleotide kinase and 3' phosphatase activities has been sum- marized in article 1 rather than in the literature review. LITERATURE REVIEW Isolated Bacterial Nucleoids g, gglifisDNA is about one mm long (Cairns, 1963). Because the dimensions of E, ggli_cells average about one thousandth of this, DNA must be highly convoluted to fit into cells. Transcription, segrega- tion, recombination, and replication all force further constraints on the DNA's organization. Understanding these processes probably re- quires understanding the restrictions they place upon DNA structure. Although it is reasonable to expect that lysing cells would neces- sarily destroy all of this organization, this is apparently not the case. Stonington and Pettijohn (1971) reported the first lysis pro- cedure that retained somqu, ggli_DNA organization. Usually, unless the cells were quite dilute, most lysis procedures yielded viscous lysates. Their lysis procedure, however, produced lysates that were not viscous, because the DNA from the lysed cells did not spread throughout the solution. Being compact, this DNA sedimented more quickly than disperse DNA would have sedimented. All further work (see below) on these compact DNA structures, called "nucleoids," retained certain elements of Stonington and Pettijohn's (1971) three step lysis procedure. Cells were first resuspended on ice in a solu- tion containing 20% sucrose, NaCl, and NaN3. The sucrose and salt gave the solution the correct osmolality to prevent spheroplasts from lysing prematurely. The NaN3 blocked ATP production via 4 oxidative phosphorylation. Interestingly, ATP alone, when added to these lysates after the third solution, turned them viscous (K. Sirotkin, R. Pearson, L. Snyder, unpublished), implying that the nucleoids were "unfolded." Lysozyme then turned the cells into»sphero— plasts. The third solution contained NaCl to a final concentration of at least 1 molar (to inhibit nucleases) and detergent to lyse the spheroplasts. These nonviscous crude lysates were often layered on sucrose gradients to purify or analyze the nucleoids. Nucleoid structure Stonington and Pettijohn (1971) reported that either SDS or heat "unfolded" the nucleoid structure, as shown by increased vis- cosity. Nicks introduced by DNAase, also disrupted the nucleoid structure (Stonington and Pettijohn, 1971; Worcel and Burgi, 1972), with a kinetics that implied that five to forty separate super- coiled loops or domains were involved (Worcel and Burgi, 1972). Surprisingly, both groups reported that RNAase unfolded nucleoid structure. Although in early work Worcel and Burgi, (1972) claimed that one nick, introduced by RNAase, was sufficient to unfold nu- cleoid structure, later workers (Pettijohn and Hecht, 1973; Drlica and Worcel, 1975), found intermediates after RNAase treatment. Thus, more than one nick was necessary to unfold nucleoid structure. The RNA molecules most tightly bound to the nucleoid and hence im— plicated in maintaining its structure, were studied by Hecht and Pettijohn (1975, 1976). They found that: 1) the RNA molecules were bound at more than one site (1975); 2) DNA that was either about to be replicated or freshly replicated did not bind more RNA than did other DNA (1975); 3) about 60-90 molecules were bound per nucleoid (1976); 4) about 300 base pairs per RNA molecules were bound, probably as RNAzDNA hybrids (1976); 5) neither their hybridization kinetics nor their rifampicin mediated disappearance kinetics dis- tinguished these preferentially bound RNA molecules from the bulk of g, ggli's RNA; and 6) both rRNA and mRNA species were represented. Rifampicin treatment, which is capable of blocking all RNA initiation, unfolded the nucleoid in_vigg_so that after the lysis procedure out- lined above, the lysates were viscous (Dworsky and Schaechter, 1973; Pettijohn and Hecht, 1973). Whether the nucleoid structure is dis- rupted before or after lysis, the resulting solution is spoken of as containing "unfolded" nucleoids. All of these data are consistent with the model that the associa- tion of RNA with the nucleoid DNA occurs shortly after the RNA is transcribed, possibly by competing with the non-transcribed strand for hybridization, perhaps at regions that are slightly underwound because of supercoiling (supercoiling is discussed in a following section). Other cellular constituents than RNA were isolated with nu- cleoids. Proteins and cell membrane pieces, and even part of the cell wall can be attached to the nucleoid. Depending on the tempera— ture during the incubation with high salt and detergent, either mem- brane-attached (O°-4°C), membrane-released(22°C)cu'a.mixture of types (10°C) were observed (Worcel and Burgi, 1974). Some of the at- tachment points, as measured by X-ray mediated breakage and later ad- herence to Mg++-Sarkosyl crystals, disappeared after rifampicin treatment, as if they arose during RNA transcription (Dworsky and Schaechter, 1973). Korsch et al. (1976) showed that the membrane attached nucleoids also had cell wall parts attached, and that, be- cause of this, careful attention must be paid to the physiological state of the cells because of physiologically linked variations in lysozyme susceptibility. This varying lysozyme susceptibility was used by Korsch et a1. (1976) to explain differences in sedimenta- tion velocities for nucleoids from, for example, amino acid starved cells. By equilibrium CsCl centrifugation of fixed membrane re- leased nucleoids, Giorno et al. (1975a) showed that the DNA to RNA to protein ratios were constant in spite of the observed heteroge- niety of sedimentation profiles. The proteins included RNA poly- merase (Stonington and Pettijohn, 1971). Both protein and RNA maintain nucleoid structure (Drlica and Worcel, 1975). Different treatments produced partially disrupted membrane released nucleoids that could be distinguished by both their viscosity and sedimentation rate. Treatments that removed all super- coiling, 2ug/ml ethydium bromide or DNAase, yielded nucleoids that made slightly more viscous suspensions or were slower sedimenting than untreated nucleoids. Treatment with RNAase or incubation at 35°C in low (0.1 M NaCl) salt yielded more unfolded nucleoids. Since nucleic acid hybridization stability decreases at low salt (Marmur and Doty, 1962), and lessened ionic strength can be expected to in- crease the binding of proteins to nucleic acid, both the RNAase and low salt probably released the RNA:DNA hybrids mentioned earlier, rather than releasing structures stabilized by proteins. Treatment of these above slightly unfolded (35°C 0.1 M NaCl) nucleoids with trypsin, or treatment of fresh nucleoids with 1% SDS completely un- folded them. These last treatments would be expected to destroy all interactions between proteins and nucleic acids. Thus both RNA and protein stabilize nucleoid structure. It is important to realize that different groups used different gravitational field strengths, and that this alone can lead to dif- ferent apparent sedimentation rates for nucleoids (Hecht, et al., 1977). The higher the field strength, the lower the apparent sedi- mentation rate. Nucleoids: artifact or biologically significant: One concern is that the procedures for isolating nucleoids might produce artifactual protein-DNA associations. Along this line, Silberstein and Inouye (1974) reported that the concentration of lysozyme commonly used in nucleoid isolation can lead to artificial DNA aggregation. However, Korsch et a1. (1976) reported that the high salt of the lysis prevents such artificial aggregations. By flourescence microscopy, the dimensions and gross structure of the isolated nucleoids appeared similar to the appearance of DNA in. intact cells (Hecht et al., 1975), corroborating the preceding re- sult. If nucleoids occur widely in nature, their universality would imply that some essential organization might be leading to their particular structure, even if their isolation gives rise to struc— tures different from those occurring in 3339. Benyajat and Worcel (1976) isolated chromosomes from Drosgphila melanogaster that resemble E, coli nucleoids in having separate supercoiled loops. But because they were too unstable to purify, it was difficult to detail their structure. Pifisn and Salts (1977) isolated supercoiled chromosomes from Saccharomyces cerevisiae. The isolated chromosomes contained at least 60 independent supercoiled domains and protein stabilized their structure. The chromosomes also differed in sedi- mentation rate depending upon the part of the yeast cell cycle from which they were isolated. Griffith (1976) presented evidence that procaryotic and eucaryotic DNA had a similar beaded structure when viewed by electron microscopy, although because lysozyme was used in the presence of low salt to lyse the procaryotic cells, artificial associations between lysozyme and DNA might have occurred (Silberstein and Inouye, 1974). These observations together imply that the structure of iso- lated nucleoids, even if not exactly corresponding to $2.V1V° struc- tures, at least might arise from fairly widespread in_vivo organiza- tion. Nucleoids as a probe of replication related events Since the first claim that nucleoids from E. coli cells were released from their membrane attachments at the end of a round of replication in amino acid starved cells (Worcel and Burgi, 1974), many workers have investigated nucleoid structure and its relation- ship to the replication cycle. Ryder and Smith (1974) found evi- dence that membrane attached nucleoids can be found after lysing amino acid starved cells. They attributed the earlier reported absence to the membrane associated nucleoids being pelleted in a prespin used by Worcel and Burgi (1974). In fact, Ryder and Smith (1974) claimed that envelope associated nucleoids isolated from amino acid starved 10 cells sedimented about 50% faster than controls. These workers also reported (Ryder and Smith, 1975; Ryder et al., 1975) that a change in the sedimentation rate of isolated nucleoids can be observed that cor- relates with the initiation of DNA synthesis. They report that the sedimentation rate of isolated bacterial nucleoids increased about 14% at high temperature in a ts gnag mutant and that the sedimentation rate decreased to normal after shift down to the permissive tempera- ture. Disappointingly, no correlation with DNA content, degree of supercoiling, or number of membrane attachment sites was described by these workers. Korsch et a1. (1976) explained at least some of the preceding with the observation that amino acid starved cells were less affected by lysozyme. They also found that the membrane associated nucleoid contained cell wall material. After adjusting for the different susceptibility to lysozyme, amino acid starved cells gave similar results to control cells. By using cells growing with poor carbon sources so that some cells would not be replicating their DNA (Cooper and Helmstetter, 1968), Korsch et a1. (1976) showed that membrane associated nucleoids isolated from cells sedimented at about the same rate whether or not they were replicating their DNA. The slight dif- ference between the two rates could be accounted for by the greater DNA content of the replicating nucleoids. This, however, did not ex- plain the results Ryder and Smith (1974) obtained with the ts dgag mutant. Thus, in spite of the fact that about 0.5% of the g. ggli_chromo- some will not replicate without protein synthesis (Marunouchi and Messer, 1973) no convincingly different isolated nucleoids have been 11 obtained from amino acid starved cells. After a lysis similar to that used to isolate E. coli nucleoids, DNA near the B, subtilis origin was preferentially associated with mem- brane components after shearing (Yamaguchi and Yoshikawa, 1977; Imada et al., 1975). Since B, subtilis has a mesosome to which DNA seems to be attached, it is hard to compare these results to results obtain- ed using E, coli. Nucleoids as transcription and replication templates Another way to test whether or not nucleoids retain components of their in 3139 organization as they are isolated is to try to use them as templates for transcription and replication. Kornberg, et a1. (1974) isolated membrane associated nucleoids with 10 mM spermidine replacing 1.0 M NaCl and tried to replicate them. About one percent of the genome replicated at about five percent of the in gi!g_rate. Although some increased viscosity after incubation in the replication mixture was observed, no decrease in sedimentation rate was detected. Interestingly, if both polymerase I and ligase were added, small DNA fragments were changed to large fragments, as occurred in vigg (Okazaki, et al., 1968). Giorno, et al. (1975b) studied transcription with nucleoid templates. They found that the size of transcripts and their elon- gation rate were similar on folded and unfolded nucleoids. However, the number of active RNA polymerase molecules and the chain initia- tion rate were both greater on the folded chromosome. As is dis- cussed below, it is possible that supercoiling alone accounted for this effect although this would not be established by these 12 experiments alone. Enzymes AlterinngNA Supercoiling The first report on an enzyme catalytically relazing supercoil- ed DNA but leaving the DNA covalently closed, was made by Wang (1971). Burrington and Morgan (1976) purified this "omega" protein from E. coli and confirmed Wang's report that it only relaxed negative DNA super- coils. This is significant because most models for replication involve the generation of positive supercoiling (see Cairns, 1963, for example). However, Morgan (1970) proposed a model with linked progeny strands that would generate negative supercoiling. Gellert et al. reported (1976a) an ATP dependent reaction capa- ble of catalytically introducing negative supercoils into DNA. This supercoiling enzyme has been named DNA "gyrase." They also reported (1976b) that novobiocin and coumermycin inhibited DNA gyrase purified from 222?: but not 522? E, coli, Egg, a gene determining sensitivity to coumermycin, near 82 minutes on the standard g, 2211 map, is prob- ably then the structural gene for a gyrase component. Very recent reports (Gellert et al., 1977; Sugino et al., 1977) identified a second gyrase component sensitive to nalidixic acid. This component, apparently the product of E, ggli_gene nalA (48 minutes), had a number of interesting characteristics (Sugino et al., 1977). including the ability to relax supercoiled DNA. It is unlike the omega protein im- munologically and in its ability to relax positively supercoiled DNA. Enzymes capable of relaxing DNA supercoils have been purified from many eucaryotic sources (Bauer et al., 1977; Mattoccia et al., 1976; Champoux and McConaughy, 1976; Baase and Wang, 1976; Pullyblank 13 and Morgan, 1975; Keller, 1975; Champoux and Dulbecco, 1972; Vosberg et al., 1976). These enzymes could relax both positively and nega- tively supercoiled DNA and are thus similar to the nalidixic acid sensitive gyrase component. Their biological significance Supercoiling probably plays a role in replication. The previ- ously mentioned antibiotic resistance studies (Gellert et al., 1976b; Gellert et al., 1977) imply that novobiocin and nalidixic acid which block replication disturb DNA supercoiling. Gyrase also plays a role in bacteriOphage ¢xl75 replication. Unless RFIV is supercoiled to RFI, the obligatory gene A_product cleavage will not occur (Marians et al., 1977). Because novobiocin and coumermycin inhibit bacteriophage T7 replication in Egg? but not as severelyin Egg? hosts (Itoh and Tomizawa, 1977), supercoiling also probably plays a role in T7 replication. Interestingly, nucleoid-like structures are also observed in T7 infected cells, but these structures do not have the same RNAase or proteinase sensitivity as do the E. coli nucleoids (Serwer, 1974; Hiebsch and Center, 1977; Paetkan, et al., 1977). As they replicate, plasmids exist briefly, with fewer supercoils before becoming fully supercoiled (Timmis et al., 1976; Crosa et al., 1976). This implies that supercoiling might be an essential part of plasmid "maturation." It is not known, however, which plasmid func- tions require supercoiled DNA. As mentioned in the introduction to this thesis, it is possible that the pseT gene product alters DNA supercoiling. The alc gene 14 product might also affect DNA supercoiling. It might, for example, unfold its g. Ell host's nucleoid (Sirotkin, et al., 1977 - appendix 1) by removing its supercoiling. R. Pearson and L. Snyder (personal communication) in our laboratory have obtained preliminary results indicating that the alg_gene product affects the supercoil- ing of bacteriophage lambda DNA. In vitro observations Three complications impede attributing igzzi!g_biological sig- nificance to the biochemistry of DNA supercoiling observed in Xl££2° The first is common to all biochemistry: the assumption that enzymatic actions observed in vitro also occur ig_!igg. Although this could be argued, it is certainly a reasonable initial assumption. The second complication concerns the meaning of the finding that ex- extracted DNA is supercoiled. As a few moments of experimentation will demonstrate, to coil anything tightly requires rotation. The supercoiling would only stress the coil to the degree that it is extended. However, the DNA would tend to extend as it diffuses away from a coil that concentrates it. Thus, the equilibrium state of a supercoil must balance the energy stored in the coiling and the energy stored in the increased local DNA concentration.. "Supercoil— ing" therefore refers as much to DNA topology as to any internal stress. This point is independent of the number of strands making up the coil. As discussed in Griffith (1975), if a double helix is slightly denatured, in 3119, by proteins or ionic conditions, it will supercoil in the absence of the denaturing proteins or ionic condi- tions. (However the interpretation that RNA:DNA hybrids are 15 maintaining the structure of the isolated bacterial nucleoid (see above) supports the model that supercoiling generates some single stranded regions in_yivg). The third complication involves the use of supercoiled DNA substrates. Not only are such experiments neces- sarily performed igflxitgg, but in addition, the DNA used may be in a form never found in_vivg. Nonetheless, many such experiments have been performed and some information can certainly be gleaned from them. For example, that an endonuclease specific for single stranded 'DNA makes some nicks in supercoiled DNA (Wang, 1974) implies that such DNA contains, at least transiently, single stranded regions. Perhaps this tendency towards unwinding explains why RNA polymerase binds much more tightly to supercoiled DNA (Warner and Schaller, 1977), and why there are more, and more stable, promoters on supercoiled DNA (Richardson, 1975). More difficult to interpret, Bochtan et al., (1973) found that more initiation occurs on lambda DNA if it is super- coiled, but that this transcription of the supercoiled template con- tains a lower proportion of lambda early mRNA and is less sensitive to lambda §l_repressor. They assumed that these later facts implied that this transcription of the supercoiled template was not biological and so they reported no controls to see if it was only transcribed from one strand or experiments to determine whether such transcripts can be translated into late proteins. Some late lambda proteins can be made in the absence of the g_gene product (Sato and Campbell, 1970; Echols, 1971), which positively regulates lambda late gene ex- pression. It is tempting to speculate that supercoiling mimics or substitutes for lambda g_gene product function. l6 12;!1££2 lambda integrative recombination requires closed cir- cular DNA (Mizuuchi and Nash, 1976) and this DNA must be supercoiled (H. A. Nash, personal communication). Holloman and Radding (1976) report that supercoiled DNA can accept DNA fragments 12_Xl££2: that the Eggé_gene product can recombine into the supercoiled DNA in 3139. Perhaps the transient single strandedness of supercoiled DNA leads to both of these phenomena. Intermediates and models The relaxation complexes of some plasmids might represent some type of intermediate in supercoiled DNA processing, although it is not clear what biological significance they might have or even, in fact, if they involve omega proteins or gyrases at all. When the super- coiled plasmid and its associated (not covalently bound) protein are treated with SDS or pronase, one DNA strand is nicked, leaving a 3' hydroxyl (Guiney, 1975). This seems similar to the supercoil relaxing protein, the 331A gene product, that leaves a double strand break in the DNA when the nalidixic acid inhibited gyrase complex is treated with SDS (Sugino et al., 1977). The reported intermediate with a single strand nick for the rat liver omega protein (Champoux, 1976) may also be similar. Sugino et a1. (1977) proposed a simple model for the two compo— nent gyrase. The Egg gene product, according to this model, melts the double helix forcing the same number of turns of the double helix into a shorter area thus causing the nonmelted regions to be more posi— tively supercoiled. The nalé gene product then relaxes these positive supercoils. When the complex leaves the DNA, the DNA is underwound. 17 All of these observations might be related to the phosphate "shuttling" performed by the p§32_gene product ipflvigrg. Although the relaxation complex has the same phosphate orientation as would be left by the Egg: gene product (article 1), no information is available on the mechanics of winding or unwinding of supercoiled DNA. So there is not any information, as yet, on the question of whether any of these other proteins "shuttle" phosphates as they function. The fact that they have not been observed to function ipuvigrg similarly to 5' polynucleotide kinase might simply mean that they assay differently. The Relationship of Nicking or Gapping Enzymes to Gene Expression Besides the examples involving T4 DNA ligase and bacteriophage T5 exonuclease mentioned in article 1 of this thesis, some examples involving eucaryotic RNA polymerase and single stranded DNA have been reported. Mammalian RNA polymerase beta can initiate at single strand- ed DNA breaks if they have 3' hydroxyl termini (Dreyer and Hausen, 1976). ssDNA sequences, obtained using hydroxylappatite, appear disproportionately frequently as chick embryo mRNA (Tapiero et al., 1976), implying that DNA gaps play a role in mammalian gene expression. The phosphate "shuttling" performed by the png_gene product En 31532, if significant 32:3339) could imply that the png_gene product affects termini processing. The p§33_gene product can alter phosphates at DNA termini so that, if those termini are at nicks or gaps, the ability of other enzymes to process those termini could be altered: a 3' phosphate would interfere with gap closing by DNA polymerase and either a 3' phosphate or a 5' hydroxyl would prevent DNA ligase from 18 sealing a nick. Requirements for T4 True-late Gene Expression Because this topic has recently been thoroughly reviewed (Rabussay and Geiduschek, 1977) and because it is discussed as it re— lates to the pggl gene product (article 1), this section will only provide a brief overview. Normally, T4 true-late gene expression re- quires concomitant T4 replication (Riva et al., 1970a), but by in- activating T4 DNA ligase (gene 30) and a T4 exonuclease (gene 46), T4 true-late gene expression can be uncoupled from replication (Riva et al., 1970b). The products of T4 genes 45 (Wu and Geiduschek, 1975), 33 and 55 (Bolle et al., 1969) are also required for T4 true- late gene expression. Unless the alg_gene is altered, the DNA tem— plate must contain hydroxylmethylcytosine (Snyder et al., 1976). In a recent paper appearing after the review, Wu and Geiduschek (1977) point out some additional requirements for T4 true-late gene expression. Protein synthesis early in the late period is required for efficient and abundant transcription of the true-late template DNA strand. Interestingly, a temperature sensitive mutation in gene 55 that is normally thermoreversible is irreversible when T4 true- 1ate gene expression is uncoupled from replication. This is espe- cially interesting because (see the section below discussing T4 in- duced RNA polymerase modifications) the product of gene 55 binds to the host RNA polymerase, (Ratner, 1974b), which is used throughout T4 develOpment (Haselkown et al., 1969; Mizuno and Nitta, 1969; di Mauro et al., 1969). 19 T4 Induced Modifications of RNA Polymerase Because the host RNA polymerase's response to antibiotics is retained by the T4 infected cell throughout infection (Haselkorn et al., 1969; Mizuno and Nitta, 1969; di Mauro et al., 1969), T4 is known to use the host polymerase throughout its development. How- ever, the alpha subunit of the RNA polymerase was altered after in— fection (Goff and Weber, 1970). Two distinct gene products are re- sponsible for this modification (Horvitz, 1974a, b). Both of these attach an adenine nucleotide to an arginine residue of the alpha subunit (Goff, 1974), and are therefore called ADP-ribosyltransferases (Rohrer et al., 1975). These enzymes are nevertheless not essential for T4's development (Horvitz, 1974b). Stevens (1972) found that four T4 induced polypeptides bind to RNA polymerase. Confirming her gene assignments, Horvitz, (1973) proved that gene 33 codes for one peptide and Ratner (1974b) proved that gene 55 codes for another. The other two were simply named poly- peptides numbers 2 and 4, with no known gene assignments. These polypeptides were distinguished by molecular weight. Some evidence (see below) implies that the alg gene may code for polypeptide number 2 (Sirotkin et al., 1977 - appendix 1). The product of gene 45 also binds to the host RNA polymerase, but much more loosely (Ratner, 1974a) than the others. Is alc the structural gene for polypeptide number 2? Although the RNA polymerase purified from cells infected with some alc- mutants has lacked a polypeptide that might be polypeptide number 2 (Sirotkin et al., 1977 - appendix 1), there is some question 20 as to whether al£_is actually the structural gene for polypeptide num— ber 2. Even assuming that the missing polypeptide from 312? infection is polypeptide number 2, genetic uncertainties remain. When algf T4 are originally selected or propagated, mutations in other genes appear (article 1 and unpublished). The T4 used to infect the cells from which the RNA polymerase was purified may have had second mutations in the actual gene coding for polypeptide number 2. Furthermore, when 2127 T4, that originally did not induce a polypeptide number 2, were propagated they sometimes changed so that polypeptide number 2 could be co-purified with the polymerase from infected cells. Although this may seem to be fairly good evidence that 312 is not the struc- tural gene for polypeptide number 2, there were other indications that weaken that interpretation. Not only do these 3127 T4 that changed so that they now induce a polypeptide number 2 that co-purifies with the RNA polymerase usually grow less well, by plaque size, with cytosine containing DNA, they also typically became less defective. in their ability to unfold the host nucleoid (see below). That glgf T4 sometimes tend to become "less 312?" and also to lose some asso- ciated phenotypes could be explained by alg being required for a fully normal burst size, at least when T4 contain hydroxymethyl- cytosine. .Alg_might even be an essential gene. This could explain the selection of the above mutants, because they were all propagated with hydroxymethylcytosine containing DNA. The question then becomes whether the mutation(s) that "brings back" polypeptide number 2 and causes these other changes is in the £12 gene. This question will remain until recombinants from infections with different alc- T4 that 21 originally did not induce polypeptide number 2, but now do induce the polypeptide are found (see the recommendations below in a following section). So far (unpublished), when we have attempted to find 312' recombinants from crosses with gig? parents, they either eluded us or were so infrequent that they may actually have been revertants. Another technique may lead to an answer to this question. C. Goff (personal communication) is using antibody precipitation to purify RNA polymerase from cells infected with progeny from multi- factor genetic crosses. We have sent him some of our mutants to use. There is, however, a potential problem with comparing his results to ours (Sirotkin et al., 1977 - appendix 1). Antibody precipitation may detect a looser association of polypeptide number 2 with the poly- merase than would phosphocellulose chromatography. Until such a chromatographic step is added, conclusions regarding the presence or absence of polypeptide number 2 cannot be compared between the dif- ferent methods. (Of course, the bound antibody would totally distort such a chromatographic step). Thus, difference in methods could account for the difficulty repeating our results that C. Goff has personally communicated to us. Another approach to this question would not be affected by the above problem. A T4 mutant exists which induces two species of poly- peptide number 2 that differ slightly in molecular weight (H. R. Horvitz, D. I. Ratner, and A. R. Poteet, personal communication). C. Goff (personal communication) is using multifactor crosses to lo- cate the responsible mutation. So far, this mutation seems to map between genes 63 and 31, placing it, at least, very close to alc 22 (Sirotkin et al., 1977 - appendix 1). One model for £19 gene product function has it acting as a pro- tease (see the models in a following section). If this is the case, the mutation causing the altered polypeptide number 2 could be in the 313 gene and the alg_gene would still not be the structural gene for polypeptide number 2. In this case, the recombination studies could also yield similarly deceptive "positive" results. .The only way to prove that polypeptide number 2 is coded for by the alg_gene would be similar to that used by Horvitz (1973) to prove that gene 33 codes for one of the other polypeptides co-purifying with the poly- merase. He used multiple amber mutations in gene 33 and showed that the number of tyrosine residues inserted by an amber suppressing RNAtyr increased as the number of amber mutations. However, this method presupposes that it is possible to get amber mutations in ale that have the Halgf" phenotype. Since we do not even know if the 3157 phenotype can result from a deficiency, we cannot know if this is possible. If, as mentioned earlier, alc is an essential gene, an amber alc may not be "alc-" under any conditions. However, if an amber 313 could be found, it would certainly produce an altered polypeptide number 2 when grown on nonsuppressing bacteria, if alg_is the struc- tural gene for polypeptide number 2. We have not yet been successful in selecting an amber alg, This, however, may be due to the neces- sity of forcing such mutants to grow on a nonsuppressor with cytosine containing DNA in order to enrich for them. 23 The Structure of T4 DNA after Replication has Begun Altman and Lerman (1970) analyzed normal hydroxymethylcytosine containing T4 DNA from a low salt lysozyme lysis that had been treated with RNAase and trypsin. Because of the low salt and fairly high lysozyme concentration used, it is possible that some artifactual ag- gregation occurred (see Silberstein and Inouye, 1974). In spite of this problem, their work still merits discussion. They found DNA sedimenting at different rates: 1) "slow," sedimenting at less than lOOs--probably genomic size DNA; 2) "fast," sedimenting at about 170— 6003; 3) "phage," DNA in virions; 4) "bottom," sedimenting at 950 to 3,0003. Interestingly, the DNA in the fast and bottom categories only partially chased to the slow and phage categories; about one-third of it seems "locked into" the faster sedimenting states. Gene 49 product, which is required for the packaging of DNA into phage particles, seems to cut single strand gaps in DNA at about geno- mic intervals (Curtis and Alberts, 1976). Under conditions where T4+ DNA sediments at about 2003, DNA from T4 with an amber mutation in gene 49 sediments at 1,000 to 1,8005 (Kemper and Brown, 1976). Ob— viously, at least some of the differences reported by Altman and Lerman (1970) occurred because of packaging. Another sedimentation difference is observed that correlates with whether or not T4 true-late genes are being expressed (Snyder and Geiduschek, 1968; Cox and Conway, 1975a, b). Both groups found that a slower sedimenting DNA fraction is absent from cells infected with T4 having an amber mutation in gene 55. (See the discussion of 24 T4 true-late gene expression). Although it is possible that some true-late genes (gene 49, for example) might be processing the DNA into a slower sedimenting form, it is tempting to speculate that this slower sedimenting state contains the DNA from which true-late mRNA can be transcribed. Consistent with this are the facts that the tran- scription of the DNA in the slower sedimenting form is sensitive to rifampicin (the faster sedimenting form is not sensitive) and, by immune precipitation, that the RNA transcribed from the slower sedimen— ting form can be translated into true—late gene products (Cox and Conway, 1975a). Not only that, but Cox and Conway (1975b) also report that mutants in gene 46-47 which express a DNA arrested phenotype, but which express true-late genes normally, created none of the faster sedimenting DNA and an increased amount of the slower form. It is difficult to compare these results with those obtained with bacterial nucleoids because the conditions used to isolate the T4 DNA were very different from those used to isolate bacterial nucleoids. T4 nucleoids Hamilton and Pettijohn (1976) used a lysis procedure similar to that used to obtain isolated bacterial nucleoids to prepare corre- sponding particles of replicating T4 DNA. By flourescent microsc0py, the particles had about the same dimensions as the cells from which they came. There was little effect of nicking by DNAase or RNAase treatment, but about 25% to 40% of the DNA leaves the nucleoids with SDS treatment. The change in sedimentation rate observed with dif- ferent ethidium bromide concentrations was consistent with about 25 one-eighth of the nucleoid DNA being supercoiled. Perhaps the DNA that was supercoiled (or that which is removed by SDS treatment) cor- responds to the slower sedimenting DNA mentioned earlier. Another possibility is that the DNA lost during nucleoid purification con- tains this DNA. Mutant studies could answer this point. Effects of cytosine substitution in T4 DNA Normally, T4 DNA contains alpha and beta glucosylated (Lehman and Pratt, 1960) hydroxylmethyl cytosine (Wyatt and Cohen, 1952). Except for it being susceptible to bacterial restriction enzymes, T4 with DNA that is unglucosylated because of mutations in the glucosyl transferase genes grows almost as well as T4 with glucosylated DNA (Hattman, 1969). T4 forced to replicate with cytosine in their DNA cannot express true-late genes, even though, as observed on neutral or alkaline sucrose gradients, there is little size difference between it and the glucosylated, hydroxylmethylcytosine containing DNA (Kutter et al., 1975). Some RNA from the strand (r) that codes for late genes was transcribed from cytosine containing DNA (Wu and Geiduschek, 1975), but it obviously has some unknown defect. Snyder et a1. (1976) reported that the T4 alg_gene product prevents true- late gene expression on cytosine containing DNA. Sirotkin et a1. (1977 — appendix 1) report that other phenotypes sometimes occur when this gene is altered. For convenience, throughout the rest of this review, alc- T4" means T4 with a mutation in the alc gene that allows T4 with cytosine containing DNA to produce progeny. 26 T4 Mediated Alterations of Host DNA Organization and T4's Effect upon Host Gene Expression Although T4 induces a number of nucleases that allow it to break down host DNA and incorporate host nucleotides into its genome (Wiberg, 1966; Warner et al., 1970; Hercules et al., 1971), these will not be discussed further here (see Snyder, 1976 for a review). The changes imposed upon the organization of the host genome will be emphasized rather than interruptions to the host genome's phosphate- sugar backbone. Host chromosome disruption and unfolding T4 infection causes a disruption of the location of the host chromosome visable by light microscopy with proper staining (Luria and Human, 1950; Murray et al., 1950), and T4 infection also causes the host nucleoid to be unfolded upon isolation (Tutas et al., 1974). The gene responsible for nuclear "disruption" has been identified as ng_(Snustad and Conroy, 1974), and the gene product responsible for "unfolding,' 22;, was mapped between genes 63 and 31 by Snustad, et a1. (1976). T4 mutants deficient in nuclear disruption still unfold the host chromosome and vice versa (Snustad et al., 1972, 1976; Snustad and Conroy, 1974), although the "unfolding" is more extreme in terms of the resulting viscosity (see the previous discussion of bacterial nucleoids; Snustad et al., 1976; Sirotkin et al., 1977 - appendix 1) if the infecting T4 are also deficient in nuclear "dis- ruption" (unpublished;zuuil).P. Snustad, personal communication). 27 Does the alc gene product unfold the host nucleoid? Appendix 1 (Sirotkin et al., 1977) contains evidence that 312 (Snyder et al., 1976) and Egg (Snustad et al., 1976) are the same gene. It is clear (Sirotkin et al., 1977 - appendix 1) that many mutants selected for the 3137 phenotype are also gpff. In addition, D. P. Snustad (personal communication) has recently informed us that the one gaff mutant selected directly (Snustad et al., 1976) is also 312?. Although we have not proven that the mutation that leads to the glgf phenotype also leads to the gaff phenotype, this seems likely. Finding alg' progeny from a genetic cross between to 3137, ngf parents could provide additional evidence. As previously men- + tioned, however, we have not had success obtaining alc recombinants. The shutoff of hostggene expression There may be two distinct processes which stop host gene ex- pression after T4 infection. One prevents translation of host mRNA sequences (Kennel, 1968, 1970) the other, requiring T4 gene expres- sion, blocks at least some host transcription. It is not clear whether the lack of host protein translation results from the mRNA being defective or from some T4 induced modification to the transla— tion apparatus that discriminates against host mRNA. Kennel (1970), however, presents data indicating that host mRNA may be excluded from polysomes, supporting the latter possibility. The transcrip- tional shutoff, which is emphasized here, requires T4 gene expres- sion to inhibit host RNA transcription (Duckworth, 1971; Nomura et al., 1962, 1966; Haywood and Green, 1965; Sirotkin et al., 1977 - 28 appendix 1). Horvitz (1974b) reported that T4 that did not modify the alpha subunit of the host still shutoff the transcription of stable host RNA species. However, a role in RNA transcriptional shutoff was at- tributed to the genes responsible for altering the alpha subunit of the host (see the previous discussion) by Mailhammer et a1. (1976), because they found that the T4 adenylated RNA polymerase transcribes both bacteriophage lambda and E, ggli_DNA poorly ipuyitrg, They also found that, 12 21539, the presence of T4 DNA blocked transcrip— tion of other transcripts, possibly by competing for RNA polymerase (see also Khesin et al., 1972). Does the alc gene product prevent transcription of host RNA? Appendix 1 (Sirotkin et al., 1977) contains evidence that at least one 3127 mutant (£191) allows more transcription of host se- quences than T4+. However, we do not know for sure if the algf mu- tation is responsible for this. If a mutation at a second site is responsible for this difference, however, it is likely to be very closely linked to the 113’ mutation, because the mutant used (Sirotkin et al., 1977 — appendix 1) had been backcrossed against its parent. More 3197 mutants and segregants need to be tested for the ability to block host transcription. P. Dennis (personal communication) is testing many of our algf T4 for the ability to prevent transcription of host sequences carried by different transducing phage. Techni- cally, it is easier to tell if host sequences carried by a transduc- ing phage are present as RNA after infection than to examine total host‘RNA sequences. This technique also has the added advantage of 29 detecting if only certain sequences are affected. In addition, E. Kutter (personal communication) has found that two 31g? mutants (E131 and 3122) allow the transcription, 1p_v1yg, of host mRNA that can later be translated, 1pfly1££g. It would be potentially informative to repeat some of Kennel's (1970) 13:3139 work using these mutants. Another dramatic and independent confirmation that the §1gf mutant affects host transcription comes from the work of T. Mattson (personal communication). He found that only infecting phage carry- ing the §1gf (E151) mutation could have their defective late genes complemented by T4 wild type late genes cloned on host plasmids. Models for alc and pseT Gene Product Function If a1g_is the structural gene for polypeptide number 2, then a simple model can explain most of our data: that the E12 gene prod- uct, by binding to and altering RNA polymerase, blocks the initiation of transcription (or causes its early termination) on cytosine con- taining DNA. This primary defect would then cause the other observed phenotypes. The "unfolding" can be explained by way of the cessation of normal transcription, as occurs after rifampicin treatment (Dworsky and Schaechter, 1973; Pettijohn and Hecht, 1973). The beauty of this model is its simplicity. However, it does not explain why p§317 mutations appear along with E197 mutations. Some secondary role for the §1g_gene product's function on normal hydroxylmethylcytosine containing DNA would have to be invoked to account for the selection of these mutations. 30 Another model can explain both "unfolding" and host shutoff. We can simply propose that the §1g gene product has an "omega-like" (see discussion of enzymes altering DNA supercoiling) or endonucleo- lytic activity on cytosine containing DNA. Then the "unfolding" of the host nucleoid would be caused by releasing its supercoiling and this lack of supercoiling would then lead to the transcriptional shut- off. Furthermore, it could be postulated that, on hydroxylmethyl- cytosine containing DNA, the a1g_gene product may also unwind the helix, and this may be required in some way for normal phage develOp- ment. This would make its function similar to the function proposed for the 2311 gene product (Sugino et al., 1977). The interaction with the pggg gene product could occur if the two gene products together functioned similarly to DNA gyrase (Gellert et al., 1976a). Then, §1g_would correspond to p§1A and p§31_to Egg. One prediction of this model would be that E, 2211 CTr5x (see article 1) has a gyrase deficiency and that, by blocking the gyrase of other E, 2911 strains with novobiocin, pggz T4 would not grow as well as psggf T4 in the presence of the antibiotic. This, however, does not seem to occur (L. Snyder, personal communication). To preserve the model, one could presume that novobiocin generates a dominant defect, even in the presence of a functional T4 p§31_gene. A third model would suggest that the E13 gene product alters other gene products, perhaps as a protease. There is a precedent for this: the ESSA gene product, protein X, cleaves bacteriophage lambda's repressor (J. W. Roberts, C. W. Roberts, N. L. Craig, per— sonal communication). This would explain both the pleotrophic 31 defects of E127 T4 and its interaction with the pggl gene product; it could even explain the previously mentioned polypeptide number 2 species observed in cells infected in one mutant (H. R. Horvitz, D. I. Ratner, A. R. Poteet, personal communication). Although this could explain much of the data, this could not be called a simple model. The first two models predict that infections with a12f, but Epgf phage would produce nucleoids that would be partially unfolded. They should differ at least slightly from those produced after §1£f infections. The third model does not necessarily predict this; the different phenotypes could be totally independent. Recommendations Two of the more unsettling and interesting aspects of working with E12? T4 are the various pleotrophic defects different §1gf mu— tants express and the difficulty in obtaining-319;+ recombinants. If a nonpermissive host for §1gf T4, that is permissive for 313' T4, could be found starting with a laboratory strain of E, 3911, not only could §1gf recombinants be directly selected, but by mapping and identifying the responsible E, gg11_gene, an understanding of the Q12 gene's function would most probably be advanced. Two of the models for §1g_gene product function differed in whether they explained "unfolding" in terms of the shutoff of host RNA transcription or vice versa. R. Pearson in our laboratory is currently studying the effect of T4 infection after lambda induction on lambda supercoiling and gene expression hoping to distinguish between these explanations. 32 How T4 unfolded E, £911 nucleoids compare to the four struc- tures described by Drlica and Worcel (1975; see the discussion of nucleoid structure above), is not known. This should be examined after infection with a variety of 212? T4, especially those 31gf T4 that still unfold, at least partially, the host nucleoid. Directly studying the E1g_gene product's effect on cytosine con- taining T4 DNA may provide some clues. To test if the 3127, 231+ phenotype arises from a less severe change to the E15 gene product, than the a1gf, BEE: phenotype, the amount and length of r (lated strand transcripts could be measured: 1) for a1gf,‘gpgf phage; 2) for 2137 but Egg? phage; and 3) for §1gf,.ggff phage. If this is done, we would predict increasing amounts or lengths from (1) to (2) to (3). If the slower sedimenting form of replicating T4 DNA (discussed above) is caused by true-late transcription and not by true—late gene products, its presence or absence from infections might provide use- ful information. The way to determine if some of the true-late gene products themselves create this slower sedimenting DNA or if the slower sedimenting DNA is responsible for their transcription is with a temperature shift up experiment with a mutant expressing a thermal labile gene 55 product. If enough time at the permissive temperature was allowed so that all the true-late gene products were present, a pulse-chase experiment after shift up would determine if gene 55 is necessary for the production of the slower sedimenting DNA. If gene 55 product is necessary, this would imply that the slower sedimenting form is necessary for true-late transcription and not simply created by the action of true-late genes. Similarly, studying the distribu— tion of the various DNA components described by Altman and Lerman 33 (1970) may uncover some defects caused by 313 gene product function when T4 replicates DNA containing cytosine. Two aspects of Hamilton and Pettijohn's (1976) work might bear on the effect of 313 gene product function on cytosine containing DNA. First, it might affect the yield or the very existence of T4 nucleoids. Second, the distinctive ethidium bromide sedimentation profile, that is consistent with about one—eighth of the DNA in the T4 nucleoids being supercoiled, might be affected by the 313 gene product. Of course, any effect of 313 on cytosine containing T4 DNA would have to be separated from the effect of the true-late genes themselves. As described above for the slower sedimenting form of T4 DNA, this is not a major problem, especially since we have some mutants that appear to be 3137 only at high temperature (M. Slocum, our laboratory). Using cells synchronized by membrane binding (C00per and Helmstetter, 1968) or amino acid starvation (Marunochi and Messer, 1973), a more dramatic shutoff of host mRNA synthesis might be ob- tained from cells replicating particular areas of their genome. This, of course, presupposes that a parallel to T4's replication coupled transcription exists in E. 3311 and that these replication coupled transcription units would cluster at certain areas of the genome. Currently, by looking for "nibbled" colonies, a search for E. 3311 K12 mutants nonpermissive for 3137 or p331: T4, but permissive for T4+, is in progress. Of course, if the 313T phenotype is due to a partial deficiency, and there is no host counterpart, this is not likely to work for alc. (See appendix 3). REFERENCES REFERENCES Alberts, B. Morris, C. F., Mace, D., Sinha, N., Bittner, M., and Moran, L. 1975. Reconstruction of the T4 Bacteriophage DNA Replication Apparatus from Purified Components. pp. 241-269 in DNA Synthesis and Its Regplation. Eds. M. Goulin, P. Hanawalt, C. F. Fox. W. A. Benjamin, Inc., Menlo Park, California. Altman, S., and Lerman, L. S. 1970. Kinetics and Intermediates in the Intracellular Synthesis of Bacteriophage T4 Deoxyribo— nucleic Acid. J. Mol. Biol. 39: 235-261. Baase, W. A., and Wang, J. C. 1973. An protein from Drosophila melanogaster. Biochem. 13: 4299-4303. Bauer, W. R., Ressner, E. C., Kates, J., and Patzke, J. V. 1977. A DNA Nicking-closing Enzyme Encapsidated in Vaccinia Virus: Partial Purification and Pr0perties. Proc. Nat'l. Acad. Sci. U. S. A. .13: 1841—1845. Benyajati, C. and Worcel, A. 1976. Isolation, Characterization, and Structure of the Folded Interphase Genome of Drosophila melanogaster. Cell 2: 393-407. Bolle, A., Epstein, R. H., Salser, W., and Geiduschek, E. P. 1968. Transcription During Bacteriophage T4 Development: Require‘ ments for Late Messenger Synthesis. J. Mol. Biol. 33; 339- 362. Botchan, P., Wang, J.(3n,and Echols, H. 1973. Effect of Circularity and Superhelicity on Transcription from Bacteriophage A DNA. Proc. Nat'l. Acad. Sci. U. S. A. ‘19: 3077-3081. Burrington, M. G., and Morgan, A. R. 1976. The Purification from Escherichia coli of a Protein Relaxing Superhelical DNA. Can. J. Biochem. 33: 301-306. Cairns, J. 1963. The Bacterial Chromosome and its Manner of Replica- tion as Seen by Autoradiography. J. Mol. Biol. 3: 208-213. Champoux, J. J. 1976. Evidence for an Intermediate with a Single- strand Break in the Reaction Catalyzed by the DNA Untwisting Enzyme. Proc. Nat'l. Acad. Sci. U. S. A. 13: 3488-3491. Champoux, J. J., and Dulbecco, R. 1972. An Activity from Mammalian Cells that Untwists Superhelical DNA - Possible Swivel for DNA Replication. Proc. Nat'l. Acad. Sci. U. S. A. 32: 143-146. 34 35 Champoux, J. J., and McConaughy, B. L. 1976. Purification and Charac- terization of the DNA Untwisting Enzyme from Rat Liver. Biochem. .12: 4638-4642. Cooper, S., and Helmstetter, C. E. 1968. Chromosome Replication and the Division Cycle of Escherichia coli B/r. J. Mol. Biol. 31: 519-540. Cox, G. 8., and Conway, T. W. 1975a. Template Properties of Bac- teriophage T4 Vegetative DNA: I. Isolation and Characteriza- tion of Two Template Fractions from Gently Lysed T4-Infected Bacteria. J. Biol. Chem. .339: 8963-8972. Cox, G. S., and Conway, T. W. 1975b. Template Properties of Bac- teriophage T4 Vegetative DNA: II. Effect of Maturation and DNA-arrest Mutations. J. Biol. Chem. 250: 8973-8977. Crosa, J. H., Luttropp, L. K., and Falkow, S. 1976. Covalently Closed Circular DNA Molecules Deficient in Superhelical Density as Intermediates in Plasmid Life Cycle, Nature, 261: 516-519. Curtis, M. J., and Alberts, B. 1976. Studies on the Structure of Intracellular Bacteriophage T4 DNA. J. Mol. Biol. 102: 793- 816. - di Mauro, E., Snyder, L., Marino, P., Lamberti, A., Cappo, A., and Tocchini-Valentini, G. P. 1969. Rifampicin Sensitivity of the Components of DNAedependent RNA Polymerase. Nature 222: 533—537. Dreyer, C., and Hausen, P. 1976. In the Initiation of Mammalian RNA Polymerase at Single-strand Breaks in DNA. Eur. J. Biochem. 70: 63-74. Drlica, K., and Worcel, A. 1975. Conformational Transitions in the Escherichia coli Chromosome: Analysis by Viscometry and Sedi- mentation. J. Mol. Biol. 23: 393-411. Duckworth, D. H. 1971. Inhibition of Host Deoxyribonucleic Acid Synthesis by T4 Bacteriophage in the Absence of Protein Syn- thesis. J. Virol. 3; 754-758. Dworsky, P., and Schaechter, M. 1973. Effect of Rifampicin on the Structure and Membrane Attachment of the Nucleoid of Escherichia coli. J. Bact. 116: 1364-1374. Echols, H. 1971. Regulation of Lytic Development, In. Bacteriophage Lambda. pp. 247-270. ed. A. D. Hershey, Cold Spring Harbor, New York. . Gellert, M., Mizuuchi, K., O'Dea, M. H., and Nash, H. A. 1976a. DNA Gyrase: An Enzyme that Introduces Superhelical Turns Into DNA. Proc. Nat'l. Acad. Sci. U. S. A. 13; 3872-3876. 36 Gellert, M., O'Dea, M. H., Itoh, T., and Tomizawa, J. 1976b. Novobiocin and Coumermycin Inhibit DNA Supercoiling Catalyzed by DNA Gyrase. Proc. Nat'l. Acad. Sci. U. S. A. 23: 4474- 4478. Gellert, M., Mizuuchi, K., O'Dea, M. H., Itoh, T., and Tomazawa, J. 1977. Nalidixic Acid Resistance: A Second Genetic Character Involved in DNA Gyrase Activity. Proc. Nat'l. Acad. Sci. U. S. A. .13: 4772-4776. Giorno, R., Hecht, R.Id.,and Pettijohn, D. 1975a. Analysis by Isopycnic Centrifugation of Isolated Nucleoids of Escherichia coli. Nucleic Acids Res. ‘3: 1559-1567. Giorno, R., Stamato, T., Lydersen, B., and Pettijohn, D. 1975b. Transcription 13 vitro of DNA in Isolated Bacterial Nucleoids. J. Mol. Biol. .23: 217-237. Goff, C. C. 1974. Chemical Structure of a Modification of the Escherichia coli Ribonucleic Acid Polymerase a Polypeptides Induced by Bacteriophage T4 Infection. J. Biol. Chem. 249: 6181-6190. '___ Goff, C., and Weber, K. 1970. A T4 Induced RNA Polymerase a Subunit Modification. Cold Spg. Hrb. Symp. Quant. Biol. .33: 101-108. Griffith, J. D. 1975. The Unit Chromosomal Structure: Evidence for its Universal Nature. pp. 201-208. In DNA Synthesis and its Rggulation. Eds. M. Goulin, P. Hanawalt, and C. F. Fox. W. A. Benjamin, Inc., Menlo Park, California. Griffith, J. D. 1976. Visualization of Prokaryotic DNA in a Regularly Condensed Chromatin-like Fiber. Proc. Nat'l. Acad. Sci. U. S. A. _Z3: 563-567. Guiney, D.<3.,and Helinski, D. R. Relaxation Complexes of Plasmid DNA and Protein: III. Association of Protein with the 5' Terminus of the Broken DNA Strand in the Relaxed Complex of Plasmid colEl. J. Biol. Chem. 239: 8796-8803. Hamilton, S.,and Pettijohn, D. E. 1976. Properties of Condensed Bacteriophage T4 DNA Isolated from Escherichia coli Infected with Bacteriophage T4. J. Virol. 12; 1012-1027. Haselkorn, R., Vogel,'M.,and Brown, R. D. 1969. Conservation of Rifamycin Sensitivity During T4 Development. Nature 221: 836- 838. Hattman, S. 1964. The Functioning of T-even Phages with Ungluco- sylated DNA in Restricting Escherichia coli Host Cells. Virology 33: 333-348. 37 Haywood, W, S.,and Green, M. H. 1965. Inhibition of Escherichia coli and Bacteriophage Lambda Messenger RNA Synthesis by T4. Proc. Nat'l. Acad. Sci. U. S. A. .33: 1675—1678. Hecht, R,‘M.,and Pettijohn, D. 1975. RNA Molecules Attached to DNA in Isolated Bacterial Nucleoids: Their Possible Role in Sta- bilizing the Condensed Chromosome. pp. 122-137. In DNA Syn- thesis and its Regulation. Eds. M. Goulin, P. Hanawalt, and C. F. Fox. W. A. Benjamin, Inc., Menlo Park, California. Hecht, R.M;, and Pettijohn, D. E. 1976. Studies of DNA Bound RNA Molecules Isolated from Nucleoids of Escherichia coli. Nucleic Acids Res. 3: 767-788. Hecht, R. M., Taggart, R. T., and Pettijohn, D. E. 1975. Size and DNA Content of Purified E. coli Nucleoids Observed by Fluores— cence Microscopy. Nature 253: 60-62. .Hecht, R. M., Stimpson,]D.,and Pettijohn, D. 1977. Sedimentation PrOperties of the Bacterial Chromosome as an Isolated Nucleoid and as an Unfolded DNA Fiber. J. Mol. Biol. 111: 257-277. Hercules, K., Munro, J. L., Mendelsohn, S., and Wiberg, J. S. 1971. Mutants in a Nonessential Gene of Bacteriophage T4 which are Defective in the Degradation of E, coli Deoxyribonucleic Acid. J. Virol. 2; 95-105. Hiebsch, R., and Center, M. S. 1977. Intracellular Organization of Bacteriophage T7 DNA: Analysis of Parental Bacteriophage T7 DNA-membrane and DNA-protein Complexes. J. Virol.ugg: 540- 547. Holloman, W. K., and Radding, C. M. 1976. Recombination Promoted by Superhelical DNA and the RecA Gene of Escherichia coli. Proc. Nat'l. Acad. Sci. U. S. A. 13: 3910-3914. Horvitz, H. R. 1973. Polypeptide Bound to the Host RNA Polymerase is Specified by T4 Control Gene 33. Nature New Biol. 244: 137—140. Horvitz, H. R. 1974a. Control by Bacteriophage T4 of Two Sequential Phosphorylations of the Alpha Subunit of Escherichia coli RNA Polymerase. J. Mol. Biol._gg: 727-738. Horvitz, H. R. 1974b. Bacteriophage T Mutants Deficient in Altera- tion and Modification of the Escherichia coli RNA Polymerase. J. Mol. Biol. .23: 739-750. Itoh,’T.,and Tomizawa, J. 1977. Involvement of DNA Gyrase in Bacteriophage T7 DNA Replication. Nature 270: 78-80. 38 Imada, 8., Carroll, L. E., and Sueoka, N. 1975. Membrane-DNA Complex in Bacillus Subtilis. pp. 187-200. In DNA Synthesis and its Regulation. Eds. M. Goulian, P. Hanawalt and C. F. Fox. W. A. Benjamin, Inc., Menlo Park, Calif. Keller, W. 1975. Characterization of Purified DNA-relaxing Enzyme from Human Tissue Culture Cells. Proc. Nat'l. Acad. Sci. Us So A. 1;: 2550-2554. Kemper, B., and Brown, D. T. 1976. Function of Gene 49 of Bacterio- phage T4: 11. Analysis of Intracellular Development and the Structure of Very Fast-sedimenting DNA. J. Virol. 13; 1000- 1015. Kennel, D. 1968. Inhibition of Host Protein Synthesis During Infec- tion of Escherichia coli by Bacteriophage T4: 1. Continue Synthesis of Host Ribonucleic Acid. J. Virol. E; 1262-1271. Kennel, D. 1970. Inhibition of Host Protein Synthesis During Infec- tion of Escherichia coli by Bacteriophage T4: 11. Induction of Host Messenger Ribonucleic Acid and its Exclusion from Polysomes. J. Virol. 3; 208-217. Khesin, R. B., Bogdanova, E. S., Goldfarb, A. D. Jr., Zograff, Yu. N. 1972. Competition for the DNA Template between RNA Polymerase Molecules from Normal and Phage-infected E, coli. Molec. Gen. Genet. 112: 299-314. Korch, C., ¢vrebo, S., and Kleppe, K. 1976. Envelope-associated Folded Chromosomes from Escherichia coli: Variations under Different Physiological Conditions. J. Bact. 127: 904-916. Kornberg, T., Lockwood, A., and Worcel, A. 1974. Replication of the Escherichia coli Chromosome with a Soluble Enzyme System. Proc. Nat'l. Acad. Sci. U. S. A. 11; 3189-3193. Kutter, E., Beug, A., Sluss, R., Jensen, L., and Bradley, D. 1975. The Production of Undergraded Cytosine ontaining DNA by Bac- teriophage T4 in the Absence of dCTPase and Endonucleases II and IV and its Effects on T4 Directed Protein Synthesis. J. Mol. Biol. 22; 591-607. Lehman, I. R., and Pratt, E. A. 1960. On the Structure of the Gluco- sylated Hydroxymethylcytosine Nucleotides of Coliphages T2, T4, and T6. J. Biol. Chem. 235: 3254-3259. Luria, SulE.,and Human, M. L. 1950. Chromatin Staining of Bacteria during Bacteriophage Infection. J. Bact. 32; 551-560. 39 Mailhammer, R., Yang, H., Reiness, G.,and Zubay, G. 1975. Effects of Bacteriophage T4 Induced Modification of Escherichia coli RNA Polymerase on Gene Expression 13 vitro. Proc. Nat'l. Acad. Sci. U. S. A. 12; 4928—4932. Marians, K. J., Ikeda, J., Schlagman, S.,and Hurwitz, J. 1977. Role of DNA Gyrase in ¢x Replicative-from Replication 13 vitro. Proc. Nat'l. Acad. Sci. U. S. A. 13; 1965-1968. Marmur, J., and Doty, P. 1962. Determination of the Base Composition of Deoxyribonucleic Acid from its Thermal Denaturation Tempera- ture. J. Mol. Biol. .3: 109-118. Marunouchi, T.,and Messer, W. 1973. Replication of a Specific Terminal Chromosome Segment in Escherichia coli which is Re- quired for Cell Division. J. Mol. Biol. '13: 211-228. Mattoccia, E., Attardi, D. G.,and Tocchini-Valentini, G. P. 1976. DNA-relaxing Activity and Endonuclease Activity in Xenopus 1aevis oocytes. Proc. Nat'l. Acad. Sci. U. S. A. 13; 4351- 4554. Mizuno, S.,and Nitta, K. 1969. Effect of Streptovaricin on RNA Syn- thesis in Phage T4 Infected Escherichia coli. Biochem. Biophys. Res. Comm. .33: 127-130. Mizuuchi, K., and Nash, H. A. 1976. Restriction Assay for Integrative Recombination of Bacteriophage A DNA 13 vitro: Requirement for Closed Circular DNA Substrate. Proc. Nat'l. Acad. Sci. U. S. A. .13: 3524-3528. Morgan, A. R. 1970. A Model for Replication by Kornberg's DNA Poly- merase. Nature 227: 1310-1313. Murray, R. G. E., Gillen, D. H., and Heagy, FL C. 1950. Cytological Changes in Escherichia coli Produced by Infection with Phage Nomura, M., Okamoto, K., and Asano, K. 1962. RNA Metabolism in Escherichia coli Infected with Bacteriophage T4: Inhibition of Host Ribosomal and Soluble RNA Synthesis by Phage and Effect of Chloromycetin. J. Mol. Biol. 3: 376-387. Nomura, M., Witten, C., Mantei, W., and Echols, H. 1966. Inhibition of Host Nucleic Acid Synthesis by Bacteriophage T4: Effect of Chloramphenicol at Various Multiplicities of Infection. J. Mol. Biol. 11: 273-278, Okazaki, R., Okazaki, T., Sakabe, K., Sugimoto, K., and Sugino, A. 1968. Mechanism of DNA Chain Growth, 1. Possible Discontin- uity and Unusual Secondary Structure of Newly Synthesized Chains. Proc. Nat'l. Acad. Sci. U. S. A. 32} 598-605. 40 Paetkau, V., Langman, L., Bradley, R., Scraba, D., and Miller, R. C. Jr. 1977. Folded Concatenated Genomes as Replication Inter- mediates of Bacteriophage T7 DNA. J. Virol. 33: 130-141. Pettijohn, D. E., and Hecht, R. 1973. RNA Molecules Bound to the Folded Bacterial Genome Stabilize DNA Folds and Segregate Domains of Supercoiling. Cold. Spg. Hrb. Symp. Quant. Biol. .33: 31-41. Pifihn, R., and Salts, Y. 1977. Isolation of Folded Chromosomes from the Yeast Sacchoromyces cerevisi33, Proc. Nat'l. Acad. Sci. Pulleyblank, D. E., and Morgan, A. R. 1975. Partial Purification of " " Protein from Calf Thymus. Biochem. .13: 5205-5209. Rabussary, D., and Geiduschek, E. P. 1977. Regulation of Gene Action in the Development of Lytic Bacteriophages. Chapter 1. IN. Comprehensive Virology, Vol. 8. eds. H. Fraenkel-Conval and R. E. Wagner, (Plenum Press). Ratner, D. 1974a. The Interaction of Bacterial and Phage Proteins with Immobilized Escherichia coli RNA Polymerase. J. Mol. Biol. 33; 373-383. Ratner, D. 1974b. Bacteriophage T4 Transcription Control Gene 55 Codes for a Protein Bound to Escherichia coli RNA Polymerase. J. Mol. Biol. 32; 803-807. Richardson, J. P. 1975. Initiation of Transcription by Escherichia coli RNA Polymerase from Supercoiled and Non-supercoiled Bacteriophage PMZ DNA. J. Mol. Biol. 21; 477-487. Riva, S., Cascino, A., and Geiduschek, E. P. 1970a. Coupling of Late Transcription to Viral Replication in Bacteriophage T4 Development. J. Mol. Biol. 33; 85-102. Riva, S., Cascino, A., and Geiduschek, E. P. 1970b; Uncoupling of Late Transcription from DNA Replication in Bacteriophage T4 Development. J. Mol. Biol. 33: 103-119. Rohrer, H., Zillig, W., and Mailhammer, R. 1975. ADP-ribosylation of DNA-dependent RNA Polymerase of Escherichia coli by an NAD+: protein ADP-ribosyl Transferase from Bacteriophage T4. Eur. J. Biochem. .33: 227-238. Ryder, O. A., and Smith, D. W. 1974. Isolation of Membrane- associated Folded Chromosome from Escherichia coli: Effect of Protein Synthesis Inhibition. J. Bact. 120: 1356-1363. 41 Ryder, O. A., and Smith, D. W. 1975. Properties of Membrane- Associated Folded Chromosomes 0f.§' coli Related to Initiation and Termination of DNA Replication. Cell 3: 337-345. Ryder. O. A., Kavenoff, R., and Smith, D. W. 1975. Properties of Membrane-Associated Folded Chromosomes of Escherichia coli During the DNA Replication cycle. pp. 159-186. In DNA Syn- thesis and its Regulation. Eds. M. Goulian, P. Hanawalt, and C. F. Fox. W. A. Benjamin, Inc., Menlo Park, California. Sato, K., and Campbell, A. 1970. Specialized Transduction of Galactose by Lambda Phage. Virology 31; 474-487. Serwer, P. 1974. Fast Sedimenting Bacteriophage T7 DNA from T7 Infected Escherichia coli. Virology 32; 70-88. Silberstein, S., and Inouye, M. 1974. The Effects of Lysozyme on DNA-membrane Association in Escherichia coli. Biochimica et Biophysica Acta. 366: 149-158. ' Sirotkin, K., Wei, J., and Snyder, L. 1977. T4 Bacteriophage-coded RNA Polymerase Subunit Blocks Host Transcription and Unfolds the Host Chromosome. Nature 265: 28-32. Snustad, D. P., and Conroy, L. M. 1974. Mutants of Bacteriophage T4 Deficient in the Ability to Induce Nuclear Disruption: I. Isolation and Genetic Characterization. J. Mol. Biol. 32: 663-673. Snustad, D. P., Warner, H. R., Parson, K. A., and Anderson, D. L. 1972. Nuclear Disruption after Infection of Escherichia coli with a Bacteriophage T4 Mutant Unable to Induce Endonuclease II. J. Virol. .19: 124-133. Snustad, D. P., Tigges, M. A., Parson, K. A., Bursch, C. J. H., Carson, F. M., Koerner, T. F., and Tutas, D. J. 1976. Iden- tification and Preliminary Characterization of a Mutant Defec- tive in the Bacteriophage T4-induced Unfolding of the Escherichia coli Nucleoid. J. Virol. 11: 622-641. Snyder, L., and Geiduschek, E. P. 1968. 13_vitro Synthesis of T4 Late Messenger RNA. Proc. Nat'l. Acad. Sci. U. S. A. 32; 459-466. Snyder, L., Gold, L., and Kutter, E. 1976. A Gene of Bacteriophage T4 whose Product Prevents True Late Transcription on Cytosine Containing T4 DNA. Proc. Nat'l. Acad. Sci. U. S. A. 13: 3098-3102. Stevens, A. 1972. New Smalls Polypeptides Associated with DNA- dependent RNA Polymerase of Escherichia coli after Infection with Bacteriophage T4. Proc. Nat'l. Acad. Sci. U. S. A. 69: 603-607. ‘— 42 Stonington, O. G., and Pettijohn, D. E. 1971. The Folded Genome of Escherichia coli Isolated in a Protein-DNA-RNA Complex. Proc. Nat'l. Acad. Sci. U. S. A. 33: 6-9. Sugino, A., Peebles, C. L., Krenzer, K. N., and Cozzarelli, N. R. 1977. Mechanism of Action of Nalidixic Acid: Purification of Escherichia coli nalA Gene Product and its Relationship to DNA Gyrase and a Novel Nicking-closing Enzyme. Proc. Nat'l. Acad. Sci. U. S. A. 13: 4676-4771. Tapiero, H., Leibowitch, S. A., Shaool, D., Monier, M. N., and Harel, J. 1976. Isolation of Single Stranded DNA Related to the Transcriptional Activity of Animal Cells. Nucleic Acids Res. _3: 953-963. ' Timmis, K., Cabello, F., and Cohen, S. 1976. Covalently Closed Circular DNA Molecules of Low Superhelix Density as Inter- mediate Forms in Plasmid Replication. Nature 261: 512-516. Tutas, D. J., Wehner, J. M., and Koerner, J. F. 1974. Unfolding of the Host Genome after Infection of Escherichia coli with Bacteriophage T4. J. Virol. .13: 548-550. Vosberg, H., Grossman, L. I., Vinograd, J. 1975. Isolation and Par- tial Characterisation of the Relaxation Protein from Nuclei of Cultured Mouse and Human Cells. Eur. J. Biochem. 33: 79- 93. Wang, J. C. 1971. Interaction between DNA and an Escherichia coli Protein w . J. Mol. Biol. .33: 523-533. Wang, J. C. 1974. Interactions between Twisted DNAs and Enzymes: The Effects of Superhelical Turns. J. Mol. Biol. 31: 797-816. Warner, H. R., Snustad, D. P., Jorgensen, S. E., and Koerner, J. F. 1970. Isolation of Bacteriophage T4 Mutants Defective in the Ability to Degrade Host Deoxynucleic Acid. J. Virol. E; 700- 708. Warner, C. K., and Schaller, H. 1977. RNA Polymerase-promoter Com- plex Stability on Supercoiled and Relaxed DNA. Febs Letters 13; 215—219. Wiberg, J. S. 1966. Mutants of Bacteriophage T4 Unable to Cause Breakdown of Host DNA. Proc. Nat'l. Acad. Sci. U. S. A. _33: 614-621. Worcel, A., and Burgi, E. 1972. On the Structure of the Folded Chromosomes of Escherichia coli. J. Mol. Biol._Z1: 127-147. 43 Worcel, A., and Burgi, E., 1974. Properties of a Membrane-attached Form of the Folded Chromosome of Escherichia coli. J. Mol. Biol. 33: 91-105. Wu, R., and Geiduschek, E. P. 1975. The Role of Replication Proteins in the Regulation of Bacteriophage T4 Transcription: 1. Gene 45 and Hydroxymethyl-c-containing DNA. J. Mol. Biol. 23: 513- 538. Wu, R., Geiduschek, E. P. 1977. Distinctive Protein Requirements of Replication-dependent and -uncoupled Bacteriophage T4 Late Gene Expression. J. Virol. .33: 436-443. Wyatt, G.li.,and Cohen, S. S. 1952. A New Pyrimidine Base from Bacteriophage Nucleic Acids. Nature 170: 1072-1073. Yamaguchi, K. and Yoshikawa, H. 1977. Chromosome-membrane Associa- tion in Bacillus Subtilis. J. Mol. Biol. 110: 219-253. ARTICLE I A ROLE IN TRUE—LATE GENE EXPRESSION FOR THE T4 BACTERIOPHAGE 5' POLYNUCLEOTIDE KINASE 3' PHOSPHATASE KARL SIROTKIN AND LARRY R. SNYDER Manuscript submitted for publication to Journal of Molecular Biology with the addition of the material in Appendix II 44 ABSTRACT The two T4 induced nucleic acid modifying activities, 5' poly- nucleotide kinase and 3' phosphatase, are both coded by the p331 gene. Therefore, the product of this gene is an enzyme which can remove phosphates from 3' termini and add them to 5' hydroxyl termini and thus could be said to "shuttle" phosphates on polynucleo- tides. This enzyme is involved in regulating T4 true-late gene ex- pression, probably at the level of transcription,by.establishing the required intracellular DNA structure. Our data suggest that a host gene product normally can substitute for the product of the 2331 gene making it nonessential for phage multiplication on most labo- ratory strains of E. coli. 45 INTRODUCTION The replication coupled true-late gene transcription of T4 coli- phage draws the attention of molecular biologists because it seems to require specific DNA structures. This has been suspected since it was discovered that DNA replication is required for normal T4 true- late gene expression (Epstein et al., 1963). Progress has been made toward the elucidation of these structures (Wu and Geiduschek, 1975; Rabussay and Geiduschek, 1977), but we still need more information concerning 13Hy113 requirements for T4 true-late gene expression. T4 codes for about two hundred gene products of which about seventy are required for development on most laboratory strains. Many of the remaining, so called nonessential, genes are required on other normal T4 hosts. For example, the normally nonessential T4 p331 gene product, which is the subject of this publication, is required for phage production on E. 3311 CTr5x (Depew and Cozzarelli, 1974). Many T4 genes induce enzymes which have been assayed 13 21333, While a number of these enzymes have been matched with their corre- sponding genes and probable biological functions, this has not been accomplished for other enzymes, including 5' polynucleotide kinase (E. C. 2.7.1.78) and 3' phosphatase. The 5' polynucleotide kinase activity, discovered by Richardson (1965) and Novogrodsky and Hurwitz (1966) transfers gamma phosphates of ATP specifically to the 5' hydroxyl termini of nucleic acids. 46 47 This enzyme aids in sequencing nucleic acids (Jay at al., 1974; Maxam and Gilbert, 1977) and in synthesizing specific nucleic acids (Khorana et al., 1972; Walker et al., 1975); because of this, the T4 induced 5' polynucleotide kinase can be purchased from many bio- chemical supply companies. T4 mutants defective in the induction of 5' polynucleotide kinase have been reported (Chan and Ebisuzaki, 1970), but because they expressed no observable phenotype, they were not mapped. No polynucleotide kinase activity has been found in uninfected E. 3911., but similar activities have been found in mammalian nuclei (Novogrodsky et al., 1966; Ichimura and Tsukada, 1971; Teraoka et al., 1975; Levin and Zimmerman, 1976). The T4 induced 3' phosphatase activity, discovered by Becker and Hurwitz (1967), selectively removes 3' phosphate termini from DNA. Depew and Cozzarelli (1974) showed that T4 mutants, in a gene they mapped and named 3331, do not induce 3' phosphatase activity. They also showed that T4 p331_mutations cause no phenotype on normal laboratory strains, but they prevent multiplication on a clinical isolate of E, 3311, they called CTr5x. When 23317 mutants infect E, 3311 CTr5x, T4 DNA metabolism is altered and the DNA that is made is not packaged efficiently (Depew and Cozzarelli, 1974). The pheno- types of 23317 mutants are all suppressed by 3327 mutations (closely linked to rIIB) even though the 3' phosphatase still is not induced. Uninfected cells of at least some E, 3311 strains have 3' phosphatase activity (Becker and Hurwitz, 1967). The first connection between these two seemingly disparate kinase and phosphatase activities was made by Cameron and Uhlenbeck 48 (1977) when they reported that a 3' phosphatase activity copurifies with 5' polynucleotide kinase activity; and, by a number of criteria, both activities seem to be associated with the same polypeptide. Thus this enzyme can remove 3' phosphates from DNA termini and add phosphates to 5' hydroxyls at DNA termini; therefore it can be said to "shuttle" phosphates on DNA. Since the p331 gene induces most or all of T4's 3' phosphatase activity (Depew and Cozzarelli, 1974), we suspected that the activ- ity copurifying with 5' polynucleotide kinase was the same as that discovered by Becker and Hurwitz (1967). Thus, we tested some p331? T4 for 5' polynucleotide kinase activity and some T4 mutants that do not induce 5' polynucleotide kinase for 3331 gene function. This led us to evidence, which we present, that: 1) the p331_gene codes for the 5' polynucleotide kinase as well as the 3' phosphatase; 2) the E331 gene product is required for normal true-late gene ex- pression on E, 3311 CTr5x; and 3) the E. 3311 strain CTr5x is non- permissive for pseT T4 because of an amber mutation in one or very few E, coli CTr5x genes. MATERIALS AND METHODS a) Materials Gamma-32P-ATP was made as described by Schendel and Wells (1973) or purchased from Amersham/Searle at a specific activity of about 3 Ci/mmol. Purified polynucleotide kinase was kindly fur- nished by V. Cameron and 0. Uhlenbeck. Deoxynucleoside 3' and 5' monophosphates were purchased from Sigma. Micrococcal nuclease and bovine spleen phosphodiesterase were purchased from Worthington. 4C-leucine was purchased from Schwarz/Mann or Amersham/Searle. b) Bacteriophage and bacterial strains Table 1 describes the origins and relevant characteristics of the various bacterial and bacteriophage strains used. c) Hydroxyalamine (HA) mutagenesis Hydroxyalamine mutagenesis was performed according to Tessman (1968). d) Infections for 3' phosphatase, 5' polynucleotide kinase, and B-glucosyl transferase assays E, coli was grown to an O.D.625 of 0.4 in tryptone broth at 37°C and infected at a multiplicity of 5. Ten minutes after infec- tion, the cells were harvested, pelleted, resuspended, and concen- trated ten times in 0.05 M tris-HCl (pH 7.5) and 0.1% mercaptoethanol. The resuspended cells were sonicated once or twice for 15 seconds. B-glucosyl transferase was routinely assayed as a control on these infections. 49 50 couumau 90mm ecu mocuuov xnueo «flea .w co nuaouw no» muamuaa oau amonu ucoaoaaaou cu ouaauoms Acofiunuficsaaou Hmcomuonv comm cause an uommmunnam noses ucuaouwuu once a saw vouomaom +mam xnuho “Moo .M mo condom“ HmoHc«Ho a can mcumuum wax coosuon vaunmn .cnaa .Haaoumunou van assoc cw vonwuumon xnuho cfimuum .n .oasm +H=m eqm camuum m on mucmEEou mamz :8 .m S xumam ammvcfig scum vocamuno .Hn meow ca mu + om2 nonwmuno an use + + A+chv mocoesou gonzom mfimocommuaz +m3m xnueo xmuhu cofiumuaz co £u30uw E 2 Dumb mzu 10%) of alc‘ mutants are also measurably unf'. Those alc‘ mutants which are not measurably unf‘ are also those which make reduced amounts of late gene prducts when cytosine-containing T4 DNA is made (data not shown). We have also obtained what seem to be partial revertants of the 0ch mutation (that is, they are less alc‘). These have also become unft to varying degrees. The temperature-sensitive mutation which we have mapped is inseparable from alc by recombination and maps in the same region as the unf mutation of Snustad et 01.“. So far, those alc ‘ mutations which confer the most temperature sensitivity have been the most completely unf‘. The 0ch mutant used was back-crossed against its parent three times at a ratio of 1:10, selecting the alc‘ phenotype each time; it was still unf‘. All of these observations c.p.m.><10“2 15- 10- It “A A ’4' 25 50 75 100 ff, of gradient Fig. 2 Sucrose gradients of nucleoids after infection by an alc- mutant. E. coli BE (5 ml) growing at 30 °C in M95 were labelled for 1 h before infection with 16 uCi ml ‘1 thymidine (5.5 mCi mg “). They were infected at a concentration of 4 ><10‘l rril'1 at a multi- plicity of infection of 5 in 10 ugml‘l tryptophan. At 5_ mm after infection, 1.0 ml was poured into a chilled tube contaInIng 0.1 ml of 1 mg ml ‘1 Chloramphenicol. The lysis procedure of Snustad er al.11 was used except that the solutions were made up as follows. Solution I: 0.1 m Tris, 6.5 mg sodium aznde, 1.0 g sucrose, 0.58 g NaCl, and 9 ml H20. Solution II:_ 1.2 ml Tris, 2.0 ml EDTA, 4 mg lysozyme, and 6.8 ml H20. SolutIon III: 0.1 ml Sarkosyl, 1.16 g NaCl, 0.5 m1 EDTA, and 9.0 ml H.O. Tris was 1 M adjusted to pH 8.15 with HCl. EDTA was 0.25 M adjusted to pH 7.0 with NaOH. After 5 min for lysis, two addi- tional steps were performed before layering on gradients. 0.1 ml of 20.0 mg ml ‘1 lysozyme dissolved in solution 11 were added for 10 min and then the lysate was diluted with 2.0 more ml of solu- tion III. Five minutes later the lysate was chilled and 0.15 ml were layered on 4.7 ml 10—30"/o w/v sucrose gradients made up with 0.5 ml Tris, 0.25 ml EDTA, 0.25 ml Sarkosyl, 5.8 g NaCl, and 10 ul mercaptoethanol per 50 ml. There was a 0.5-ml step at the bottom made from a solution containing 0.1 m1 Tris, 6.0 g sucrose, 1.0 g NaCl, 3.0 m1 saturated C50 and 3.0 m1 H,O. They were centrifuged at 4 CC for 25 min at 17,000 r.p.m. in a Spinco SWSOL rotor. Fractions were collected from the bottom which is at the left. a, Uninfected; b, amESl, DD2, 0ch (ref. 17), an alc‘ mutant induced by hydroxylamine and back crossed three times :gainst its parent at a ratio of 1 : 10; and c, the alc“ parent of c2. The arrows show the peak 13f sedimentation of a T4 phage mar er. support the conclusion that ale and unf are the same gene. However, we have not rigorously excluded two other pos- sibilities. One of these is that ale and turf are neighbouring genes and those alc‘ mutants which are also unf‘ are deletions span- ning these two genes. The existence of partially unf‘ mutants such as alc4, however, which is also intermediate in its tem- perature sensitivity and alc‘ phenotype, argues against the 5.0 r n‘rcl _ l 1 2 4 6 8 Min after infection Fig. 3 Time course of unfolding of host nucleoids after infection by alc‘ mutants. Cells were infected and all 5 m1 lysed as in Fig. 2. 12 min after solution 111 was added, the viscosity of the lysate was measured without previous centrifugation by mea- suring the rate of movement in mg 5‘1 through a 20-gauge needle 4 cm long. The ratio of this number to that obtained from lysates of cells at 2 min after infection is plotted. Lysates of uninfected cells flowed at close to the same rate as cells collected 2 min after infection and both were only slightly slower than H20. The process of measuring the viscosity of all the samples took less than 15 min and the uninfected cell lysates were still not viscous at this time. The closed and open figures of each type represent different experiments. We drew the line through the midpoint when the results were slightly different for the same time after infection. (0), amESl, DD2 (ref. 17); D, amESl, DD2, alc4; >’ , amESl, DD2, alc2. deletion hypothesis. Also, the high frequency of ale ‘,-unf‘ mutants has already been given as evidence for point mutants. The other possibility is that alc‘, unf‘ mutants are double point mutants, alc and unf are adjacent genes, both the ale“ and unf‘ mutations enhance the multiplication of cytosine-contain- ing T4, and the temperature sensitivity is due to the unf‘ mutation. However, we think these alternative explanations are very unlikely and that ale and unf are the same gene. If ale and unf are the same gene why are not all alc ‘ mutations measurably unf deficient? The most likely explanation is that T4 mutants which are only partially defective in air function can still multiply with cytosine in their DNA to make a plaque; but are not as temperature sensitive and are scored as unf’ in our unfolding test. Implications for host shutoff and E. coli nucleoid structure Although alc might be expected to play a part in host shutoff, ale“ and unf‘ mutants do not affect the timing of shutoff of synthesis of the host proteins one sees on polyacrylamide gels (ref. 11 and our unpublished observations). Since there could be a translational shutoff superimposed on the transcriptional one, however, we looked at the shutoff of host RNA synthesis directly by testing to see whether any RNA synthesised after infection by an alc‘ mutant can hybridise to E. coli DNA. The results are shown in Table 2. [t is clear that the alc‘ mutation Table 2 Hybridisation of RNA labelled after infection by an alc‘ mutant to E. coli DNA RNA 3H Labelling time c.p.m. Input Uninfected 2-min pulse 8.397 T4+ 4-6 4,736 8—10 3.033 14—16 2.557 0ch 4-6 5.048 8—10 3,495 14—16 2,840 c.p.m. Hybridised ‘X, of RNA Hybridised °/.. of RPA which is .co I 1061 12.7 (100) 107 2.26 17.8 31 1.04 8.2 103 4.0 31.5 224 4.4 34.6 196 5.5 43.0 261 9.2 72.5 RNA was pulse labelled at the times indicated in M9 medium at 30 6C with 10 uCi ml ’3 aH-uridine (5 Cimmol“).The phagewere ale] which had been crossed against the wild type (T4*) to remove the amESl and DD2 mutations". The multiplicity of infection was 10 and tryptophan was added at 10 ug ml 3 before infection. Surviving bacteria were less than 0.1°o at 2 min after infection. The RNA was extracted three times with phenol at 60 C by the Method I of Bolle et (II. “t ,ethanol precipitated, treated with DNase (5 ug ml 1)In 0.1 M Tris-Cl pH 8. 0.0.01 M MgCl, for 15 min at 37 CC, extracted twice and reprecipitated with ethanol. Hybridisation was with 20 ug ml ‘3 E. coli DNA“0 for 40 h at 60 f‘C in ZSSC. The hybrids were treated with RNase (12 ugml“1 RNase A at 37 ‘C for 15 min), collected on presoaked Millipore filters, and washed with 0.5 M KC1. 0.01 M Tris,pH 7.5. The specific activities of the RNAs were from top down. 10.6, 6.76, 4.0, 3.32, 7.8, 5.85, and 4.27 c.p.m. ng. The percentage of RNA which is E. coli was determined by dividing the percentage of input counts which hybridised by the percentage of the uninfected labelled RNA which hybridised. This number is affected bychanges in the relative concentrations of E. coli RNA species so should only be considered qualitative. It is perhaps surprising that E. coli transcription can continue so late into infection when the host DNA is being degraded. This RNA is presumably being made on fragmented DNA. affects the shutoff of the synthesis of at least some E. coli transcripts. Thus, selecting T4 mutants which could multiply with cytosine in their DNA has led to the discovery of a host shutoff gene. ale. If alc— mutants are defective in the shutoff of host transcrip- tion, why are they normal in the shutoff of the host proteins one sees on gels? One possibility is that there is a translational shutoff which is independent of transcriptional shutoff, and evidence has been presented for such a translational shutoff by T4 (refs 27 and 28). Host RNA is still being made after infection even by alct phage, as also observed by Kennel”. In fact. at late times the percentage of RNA synthesis attributable to the host apparently increases (see Table 2). Either alc is inefficient in its shutoff of host transcription or it is only responsible for shutting off a subset of the host RNA synthesis. Accordingly, we are under- taking experiments to determine whether all types of host transcription are shut off equally by ale. In conclusion, we have presented evidence that an RNA polymerase subunit can cause the host nucleoid to unfold. Since mutants which lack the function of this subunit allow increased transcription of the host genome after infection, perhaps the alc gene product prevents the synthesis of RNA which holds the nucleoid together. In this connection, it has been reported that treating cells with rifampin causes the unfolding of the nucleoid“. Other types of experiment also suggest that RNA is important in maintaining the highly folded state of the host nucleoid since RNase treatment can cause it to unfold”. According to one model. then, the unfolding of the host chromosome is merely a consequence of shutting off host transcription by aIc. However, another possibility should be considered: the unfolding of the host chromosome by aIc may cause host transcription to cease. If so, perhaps T4 does not make RNA on cytosine-containing T4 DNA because alc causes it to “unfold itself“. It should be possible to design ex- periments to test these. and other, hypotheses. Because of this, the T4 alc gene product promises to be an important tool in the study of the modes and mechanism of prokaryotic transcrip- tion. It also promises to be useful in the study of the roles of RNA and RNA polymerase in maintaining the structure of the prokaryotic chromosome. This work was supported by a grant from the NIH. K.S. acknowledges an NSF predoctoral fellowship. L.S. thanks Larry Gold for inspiration and hospitality in the early stages of this work. Received August 23: accepted October 22. 1976. Haselkom. R., Vogel, M., and Brown, R.. Nature. 221. 836—838 (1969). di Mauro. E., Snyder. L., Marino. P.. Lamberti, A.. Coppo. A., and Tocchini- Valentini. 0.. Smart. 222. 533—537 (1969). Stewns A., Prm. Imm. AIIIII. Sci. L’. S. .4 69. 603— 607(1972). R;.Itner D., J. mulec, Biol” 89. 803— 807(1974 ). Horxitz. H. R.. Nururr mm BIIII..244.137—140(l973). R “liner D., J. rim/er. BML. 83. 373—— 383 (I974). Goff. C., and Weber. K., Cold Spring Harb. Symp. quam. BI'III., 35. 101—108 (1970). Horvitz. H. R.. J. "Inlet. Biol.. 90, 739—750 (1974). Rohrer. H.. Zillig, W., and Mailhammer. R.. Eur. J. Biochem.. 60. '0 Snustad. D., and Conroy. L., J. mnlec. BIOL. 89, 663-673 (1974). H Snustad. D. el al., J. Virol.. 17. 622-641 (1976). ‘3 Tutas, D., Wehner. J., and Koerner. .l.. J. Virol.. 13. 548-550 (1974). ‘3 Hercules K., Munro. 5.. Mendelsohn, 5.. and Wiberg. J. J. VirnI.. 7. 95—105 I—Q \JO'AL'J-v DJ 073 227~238 (1975). (1971). Warner. H.. Snustad, D. Jorgensen S., and Koerner, J., J. Viral. 5. 700- 708 (1970 l. 15 Wiberg, J., Prm‘. nam. Acad. Sci. U. SA. 55. 614— 62l (I966). Wyatt. G., and Cohen. S.. H’Vamre 170.1072 1073(1952) Snyder. L., Gold. L. and Kutter, E. Proc. nam. Acad Sci. U S. ...4 73, 3098—— 3102 (1976). Bolle. A., Fpstein. R. Salser, W., and Geiduschek. E. P.. J. mulec. Biol” 31. 325— 338 (1968) Wu. R.. and Geiduschek. E. P.. J. mulec. Biol.. 96. 513-538 (1975). Stonington. 0.. and Pettijohn, D., Proc. nam. Acad. Sci. U.S.A..68.6—9 (1971). Burgess. R. J. bin! Chem” 244, 61—60 6167(1969). Fairbanks (J Steck. T.. and Wallach. D., BIochemitrrI. 10. 2606~2617 (1971) . Montgomery D. and Sriydcr. L., Vim/(IR). 53. 349— 358(1973). Snyder, L. Vim/urn 50, .396— 403(1972 ). - Honitl H R.. J. Inn/cc. BIOL. 90. 727— 738(1974). Pettijohn, D. and Hecht, R.. Cold Spring Harb. Svmp. quant. BIOL. 38. 31—41 (1973). Kennel D. J. Viral. 6. 208— 217(1970). Svcnson, S B. and Karlstrom, 0.. J. Virol., 17. 326.331 (1976). Mailhammer R. Yang. H.. Reiness, G., and Zubay, G. Proc. nam. Acad. Sti U.S.A.. 72. 4928-4932(1975), 30 Marmur, J., J. ntnit’t‘. BI'III.. 3. 208—218(1961). 31 Luria. S. E., and Human. M. L., J. Bact.. 59, 551—560 (1950). . ...—— ~49 7i 9;— ~35 NN'J'v’J'J Qfibald JNDJ 01-4 Printed ll‘l Great Britain by Henry Ling Ltd. at the Dorset Press. Dorchester. Dorset APPENDIX II Fine structure map of pseT and its position relative to other nearby genes The following work was performed by Wendy Cooley, Karl Sirotkin and Larry R. Snyder. APPENDIX II Fine structure map of pggT and its position relative to other nearby genes. To orient the Egg: gene relative to other genes in this region and to locate mutations within the ps3: gene, we have performed three factor crosses. We constructed double mutants consisting of tsA56 (gene 31) and some psng mutations. These double mutants were crossed against other REEIS mutants and pgggf recombinants were tested for temperature sensitivity. We also tried to determine the endpoints of pgglbl and pseTA3 by crossing them against other pgng mutants, but both deletions fail to recombine with all of those that we used. The relative position of another gene in this region, gig_(-unf?) (Snustad et al., 1976; Sirotkin et al., 1977), was determined by crossing multiple mutants containing pggl- mutations with or without the gig? mutation in the necessary genetic background to test for the algf phenotype (Snyder et al., 1976), selecting EEEIf recombi- nants, and scoring for the gig: phenotype. The map positions are shown in figure 1; the amino and carboxyl termini orientations are included, assuming these genes are transcribed with the same orien- tation as other early genes. It is of interest that pgng, which only fails to induce 3' phosphatase activity (table 2) maps close to p§3247 which by a preliminary datum induces 3' phosphatase activity but which certainly fails to induce 5' polynucleotide 87 88 kinase activity (data not shown). This suggests that the active centers for these activities are not located on opposite ends of the polypeptide chain. REFERENCES Sirotkin, K., Wei, J., and Snyder, L. (1977). Nature 265, 28-32. Snustad, D. P., Tigges, M. A., Parson, K. A., Bursch, C. J. H., Caron, F. M., Koerner, J. F., and Tutas, D. J. (1976). J. Virol. ll, 622-641. Snyder, L., Gold, L., Kutter, E. (1976). Proc. Natl. Acad. Sci. U. S. A. 13, 3098—3102. 89 90 Figure 1 Fine structure map of pseT and its position with respect to other nearby genes. All distances are arbitrary. The map orders were obtained with "three factor crosses" using tsA56 (gene 31). PseTlZ could not be mapped because it plates on E. coli CTr5x at low temperature. We do not know the endpoints of pseTAl and pseTAB; this is represented by dotted lines. 91 60 H muswfim _ F m