‘ , THE CONTROLDOF RNA TRANSCRIPTIDNIIN : . , ' VBACTERIOPHAGEigh :1‘,-~-INFECTED PSEUDOMONAS if” ‘ ' , ‘.Dissert'atIOn far the-Degreeot—Phl. D. .i ; : ~ ~ ' . «Mncmcmrmum f ‘ . _',JAMESD'FREDERICKJOLLY?fifgffff’5 31;“ - .V‘t 111‘:1:4a H "to N ,0}- . I . .__ . . ‘\,..': 3:3 .l-I «a. .‘u 21.: A“; fi'l'lfi A. t . 31$ \‘:f~'l.; —.',_;-.‘-'.',~‘. _ a . "dzahwh'am *~._- -Z-.»L}.hri(~fitn\ - t 1: LI .V V. {43' _.‘ . .'. ' .-‘ . J '. -._'l. .'; I, ~ . .’ A“ ~ .- . . ‘. .'.-; , a ‘ ..V‘t‘5\{i{'§fi ‘\)N J} H: , ," This is to certifyvthet the thesis entitled THE CONTROL OF RNA TRANSCRIPTION IN BACTERIOPHAGE gh-l-INFECTED PSEUDOMONAS PUTIDA presented by James Frederick Jolly has been accepted towards fulfillment of the requirements for Ph . D . degree in Biochemistry yMflGW Major professor Date May 21, 1976 0-7 639 d d pR F ABSTRACT THE CONTROL OF RNA TRANSCRIPTION IN BACTERIOPHAGE gh’I-INFECTED PSEUDOMONAS PUTIDA BY James Frederick Jolly The process of RNA transcription was examined in bacteriophage gh-l-infected Pseudomonas putida. The purpose was to define the mechanisms of transcriptional control in gh—l-infected cells. The approach to this problem was to determine the specificity of gh-l RNA transcription at different time intervals of the infec- tious cycle and to define the conditions necessary to duplicate the specificity of gh—l RNA transcription in XEEEQ by utilizing highly purified RNA polymerase and gh-l-DNA. The specificity of RNA transcription during three different time intervals of gh-l phage development was determined. During the early time interval (0-5 minutes post infection) the host RNA polymerase transcribes early RNA sequences from the L strand of gh—l DNA. The host RNA polymerase continues to transcribe early RNA sequences 51.4. .M t .l P .l. .1 James Frederick Jolly from the L strand of gh-l DNA during the intermediate time interval (5-12 minutes post infection) but about 90% of the gh-l RNA sequences synthesized during this period of infection are late RNA sequences transcribed by the gh-l phage-induced RNA polymerase1 from the L strand of gh—l DNA. Host RNA synthesis is completely inhibited during the intermediate time interval of gh-l phage development. The host RNA polymerase continues to transcribe early RNA sequences from the L strand of gh-l DNA and the gh—l RNA polymerase continues to transcribe late RNA sequences from the L strand of gh—l DNA during the late time interval (12 minutes post infection to lysis), but the relative amount of early RNA synthesis is much greater (50-60%) than during the intermediate time interval. In addition, during the late time interval complementary RNA is transcribed by host RNA polymerase from the H strand of gh—l DNA and represents about 20% of the gh—l RNA synthesized in gh-l-infected cells during the late time interval. The transcriptional specificity of highly puri— fied RNA polymerase was examined in yitrg to determine if the in yiyg transcriptional specificity could be duplicated. It was determined that the host RNA poly- merase could be limited essentially to transcription 1Howard C. Towle; James F. Jolly; and J. A. Boezi (1975), J. Biol. Chem., 250, 1723. James Frederick Jolly of early RNA sequences from the L strand of gh—l DNA in zitrg if the molar ratio of host RNA polymerase to gh-l DNA was below 1. The sigma factor was also required for the limitation of transcription to early RNA sequences. At higher molar ratios of enzyme to gh—l DNA the host RNA polymerase was no longer limited to transcription of early RNA sequences but was capable of transcribing the entire gh-l genome. Highly purified gh-l RNA polymerase was limited to transcription of the L strand of gh-l DNA in yitrg as it is during gh—l phage development but the enzyme was not limited to transcription of late RNA sequences in yitrg. This result was not affected by the molar ratio of gh—l RNA polymerase to gh-l DNA. The following model is suggested for the control of gh—l RNA transcription during gh-l phage development. jDuring the early time interval of gh-l phage development fthe host RNA polymerase is limited to transcription of ithe early region from the L strand of gh-l DNA. This ‘limitation is probably due to a low molar ratio of host lRNA polymerase to gh—l DNA in yiyg since host DNA is in ilarge excess and most host RNA polymerase molecules are iprobably associated with the host genome. The gh-l RNA Fpolymerase is produced and begins to transcribe late RNA isequences from the L strand of gh-l DNA during the inter- mediate time interval of phage development while the host l James Frederick Jolly RNA polymerase is still limited to the early region. During the late time interval of gh-l phage development host DNA is completely degraded. Thus, all host RNA polymerase molecules should be available for gh-l RNA synthesis. If all host RNA polymerase molecules in an infected cell are available for gh-l RNA synthesis the molar ratio of host RNA polymerase to gh-l DNA would be about 50:1. Under these conditions host RNA polymerase transcribes complementary RNA in vitro. THE CONTROL OF RNA TRANSCRIPTION IN BACTERIOPHAGE gh-l-INFECTED PSEUDOMONAS PUTIDA BY James Frederick Jolly A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1976 DEDICATION To Setsuko ii ACKNOWLEDGMENTS I am grateful to Professor John A. Boezi for giving me advice and encouragement. I would also like to thank Dr. Howard Towle who helped guide my research during my first two years of graduate school. I wish to express my thanks to my thesis com- mittee: Professors Rottman, Deal, Filner and Snyder. I especially would like to thank my wife Setsuko for all her encouragement and advice throughout my graduate carrier. This work was supported by a NIH trainee grant and in part by a grant to Dr. Boezi. TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . LIST OF ABBREVIATIONS . . . . . . . . . QITERATURE REVIEW . . . . . . . . . . Introduction . . . . . . . . . Structure and Function of Bacterial RNA Polymerase . . . . . . Organization of the T7 Genome . . . . Protein Synthesis in T7- Infected Cells . . . Pleiotropic Effect of the T7 Gene 1 Product . RNA Synthesis in T7- Infected Cells . . . The Shutoff of Host RNA Synthesis in T7 Phage- Infected Cells . . . . Binding of Host RNA Polymerase to T7 DNA . . Transcription of T7 Phage Early RNA . . . . Characterization of the T7—Specific RNA Polymerase . . Late Phage RNA Synthesis in Infected Cells . Control of RNA Synthesis in T3— Infected Cells NTRODUCTION . . . . . . . . . . . ATERIALS AND METHODS . . . . . . . . . Materials . Growth of Pseudomonas Putida and Infection with Bacteriophage gh-l . . Plaque Assay of Bacteriophage and One Step Growth Experiment . Purification of Bacteriophage and Isolation of DNA RNA Polymerase Assay . . . . . . . . Purification of RNA Polymerase . . . . ‘Preparation of 3H- labeled RNA in vitro . . . Preparation of Pulse-Labeled RNA . . . . . Preparation of Competitor RNA . . . . . . DNA-RNA Hybridization . . . . . . . . Hybridization-Competition . . . . . . . iv o Page vi Page Degradation of Host DNA During Infection . . . . 58 Separation of gh- 1 DNA Strands . . 59 Detection of RNA Duplex Formation by $1 Nuclease Digestion . . . . . 59 Polyacrylamide Gel Electrophoresis . . . . . . 61 Other Procedures . . . . . . . . . . . . 63 RESULTS . . . . . . . . . . . . . . . 64 Time Course of Nucleic Acid Synthesis In gh-l- Infected Cells . . . 64 Time Course of Early and Late gh— 1- Specific RNA. Synthesis . . . 71 RNA Polymerase Activity in Late Infected Cell Extracts . . . . 79 Asymmetry of gh- 1- -Specific RNA Transcription . . 90 Characterization of the H Strand- Specific RNA Transcribed by the Host RNA Polymerase . . . . 97 Size of the 12 vivo gh- 1- -Specific RNA . . . . 106 Sequence Homology Between gh- 1 DNA and T3 and T7 DNA . . . 107 Region Specificity of gh— 1 RNA Synthesis 1n vitro . 110 Strand Specificity of gh- 1 RNA Synthesis 1n vitro . 115 Size of gh-l—Specific RNA Transcribed 1n vitro . . 130 139 DISCUSSION . . . . . . . . . . Model for the Control of gh-l RNA Synthesis in Infected Cells . . . 139 Synthesis of Complementary RNA in Other Systems . 150 Speculation Concerning the Function and Appearance of Complementary RNA in gh-l—Infected Cells . . 152 BIBLIOGRAPHY . . . . . . . . . . . . 160 APPENDIX . . . . . . . . . . . . . . 169 LIST OF TABLES Table l. Asymmetry of gh-l-specific RNA Transcription 1n 3129 . . . . . . . . . . . . 2. Effect of Rifampicin on Burst Size . . . 3. Hybridization of gh- 1- specific RNA to T3 and T7 DNA . . c o c o c . Asymmetry of gh- 1- specific RNA Transcription in vitro . o o o o I o v I vi Page , 111 Figure 1. 10. 11 LI ST OF FIGURES Time course of RNA synthesis in gh-l- infected cells . Degradation of host DNA during infection Hybridization-competition of in vivo 3H- labeled RNA synthesized 0-5_minutes post infection in the presence of chloramphenicol . o o o o u u o Hybridization-competition of in vivo 3H- labeled RNA synthesized 8-10 minutes post infection . . Hybridization-competition of 13 vivo RNA synthesized 16-18 minutes post infection Hybridization—competition of 3H—labeled RNA synthesized in vitro by gh-l RNA polymerase activity in IKfected-cell extracts Hybridization-competition of 3H- labeled RNA Page 0 synthesized in vitro by host RNA polymerase activity in infected-cell extracts Hybridization-competition of in vivo 3H- labeled RNA synthesized 16-18 minutes post infection in the presence of rifampicin . CsCl density gradient centrifugation of denatured gh— 1 DNA complexed with poly (U, G) . . Hybridization-competition of 13 vivo 3H- labeled RNA synthesized 16—18 minutes post infection to the H strand of gh—l DNA Detection of RNA duplex formation by 81 nuclease digestion of Ln va0 3H— labeled RNA . 65 69 73 77 80 83 86 88 91 98 101 Figure 12. 13. 14 15. 16. 17. 18. 19. 20. 21. 22. Page Polyacrylamide gel electrophoresis of Ln vivo 3H— labeled RNA synthesized 16- -18 minutes post infection . . . . . . . . . . 108 Hybridization-competition of 3H-labeled RNA synthesized Ln vitro by host RNA polymerase holoenzyme . . . . . . . . . . 113 Hybridization-competition of 3H-labeled RNA synthesized Ln vitro by host RNA polymerase core enzyme at different molar ratios of enzyme to DNA template . . . . . . . 116 Hybridization-competition of 3H-labeled RNA synthesized 1E vitro with E. coli RNA polymerase . . . . . . . . . . . 118 Hybridization-competition of 3H-labeled RNA synthesized 1n vitro by gh-l RNA polymerase 120 Detection of RNA duplex formation by 81 nuclease digestion of 3H— labeled RNA synthesized 1n vitro . . . . . . . Hybridization-competition of 3H— —labeled RNA synthesized in vitro by holoenzyme to the H strand of gh- 1 DNA . . . . . . . 128 Hybridization-competition of 3H—labeled RNA synthesized 1E vitro by holoenzyme to the L of gh-l DNA . . . . 131 Polyacrylamide gel electrophoresis of 3H- labeled RNA synthesized Ln vitro by holoenzyme . . . 134 Polyacrylamide gel electrophoresis of 3H- labeled RNA synthesized Ln vitro by the gh— 1 RNA polymerase . . . . 137 Model for the control of gh-l RNA synthesis in infected cells . . . . . . 142 BSA DEAE DTT EDTA GSC NTP PFU SDS SSC TCA Tris LIST OF ABBREVIATIONS bovine serum albumin chloramphenicol diethylaminoethyl dithiothreitol ethylenedinitrilo-tetraacetic acid glucose—salts—casamino acid media nucleoside triphosphate plaque forming unit sodium dodecyl sulfate standard saline citrate (0.15g NaCl, 0.015M sodium citrate) trichloroacetic acid tris-(hydroxymethyl)-methy1amine ix LITERATURE REVIEW Introduction How genetic expression is controlled in the cell has been a fundamental question of molecular biology since the development of molecular biology as a science. Since the science of molecular biology developed histori- cally from the study of bacteriophages (1), it is not surprising that the study of bacteriophages has provided much of our current knowledge of how control of genetic expression occurs. It was clear from the beginning of such studies that a fundamental shift in genetic expres- sion occurs during phage development. The control of ‘host biosynthetic mechanisms and the utilization of 1host molecules after phage infection resulting in the production of hundreds of phage particles within a period of twenty minutes was dramatic evidence of a shift of genetic expression in infected cells. The study of DNA phage development during the past thirty years has indicated that the control of gene expression in phage systems is primarily at the level of transcription. Transcription of a phage genome is carried out initially by the host RNA polymerase, but transcription of later classes of phage genes is carried 1 out by either a modified host RNA polymerase or a new, phage-specific RNA polymerase encoded by the phage genome. In both cases transcription of the host genome is in- hibited by phage infection. The second case, that involving the synthesis of a new, phage-encoded RNA polymerase, will be reviewed here in detail because of its relevance to the research presented in this dissertation. As will be described below, the Pseudomonas putida'bacteriophage gh-l induces the appearance of a new RNA polymerase in gh-l-infected cells. The most extensively studied phages which encode for their own specific RNA polymerases are the Escherichia 991; bacteriophage T7 and the closely related T3 phage. The process of T7 phage infection will first be dis- cussed in detail and then the transcriptional control of T3 phage development will be compared to the T7 phage system. Several reviews on transcription in T7-infected cells are available (2,3). Before the process of T7 phage infection is described, a short discussion of the structure and func- tion of bacterial RNA polymerases will be presented. Structure and Function of Bacterial RNA Polymerase The following is a short discussion of the structure and function of bacterial RNA polymerases. This discussion is based on the reviews of bacterlal RNA polymerase literature by Chamberlin (4) and Burgess (5). RNA polymerase from bacterial sources is composed of four types of subunits: a, B, B' and o. The molecular weights of the subunits vary somewhat for each source of RNA polymerase. The subunit molecular weights for the E. putida RNA polymerase subunits are: a, 44,000; 8, 155,000; 8', 165,000 and 0, 98,000. Bacterial RNA polymerases exist in two forms: holoenzyme and core enzyme. The subunit stoichiometry of the holoenzyme form of RNA polymerase is aZBB'o. The subunit stoichiometry of the core enzyme form of RNA polymerase is azBB'. Chromatography of E. 9911 holoenzyme on phosphocellulose dissociates the enzyme into core enzyme and the 0 subunit while both holoenzyme and core enzyme can be obtained from phosphocellulose chromatography of B. putida holoenzyme. Both forms of bacterial RNA polymerase are capable of transcribing DNA. Both holoenzyme and core enzyme utilize poly [d(A-T)] and calf thymus DNA efficiently. The holoenzyme, however, is much more efficient with most phage DNA templates, such as T4 and T7 DNA. The holoenzyme also transcribes phage DNA asymmetrically; only the strand that is transcribed 11 3139 is transcribed by holoenzyme in vitro. The core enzyme transcribes both strands of phage DNA. For this reason 0 factor is believed to be necessary for correct initiation of RNA synthesis. The 0 factor acts catalytically during RNA synthesis. After initiation of RNA synthesis by holoenzyme, the 0 factor is released from the DNA-RNA polymerase complex and the core polymerase continues the process of RNA chain elongation. The 0 subunit can bind to another molecule of core enzyme to allow correct initiation. The antibiotic rifampicin inhibits bacterial RNA polymerase activity by binding to the B subunit of RNA polymerase. Inhibition of RNA polymerase by rifampicin occurs prior to RNA chain initiation. The process of RNA chain elongation is not inhibited by rifampicin. Organization of the T7 Genome The advantage of using the T7 phage system for experimentation is that it is genetically well defined. The size of the T7 genome is small enough (25X106 daltons [6]) that many gene functions are known, as well as the Size of each RNA transcript and the size of the corre- sponding polypeptide product. Twenty-five genes have been identified on the T7 genome: 19 essential genes have been identified by the isolation and characterization of amber mutants essential for T7 growth (7-9) and 6 non-essential genes have been identified by studying deletion mutants (7,10). The genes have been ordered on a linear genetic map and are numbered from left to right (7,8). The essential genes are numbered by integral numbers, from 1 to 19, while the non-essential genes are numbered by fractional numbers depending on their map position (eg, gene 1.3 is located three-tenths of the distance between gene 1 and gene 2). Approximately thirty T7-specific polypeptides with molecular weights ranging from 7,000 to 150,000 daltons have been identified in T7-infected cells by SDS-gel electrophoresis. These proteins account for virtually all the coding capacity of T7 DNA (7). Three temporal classes of proteins appear in a T7 time course of infection (7): (a) class I proteins are synthesized between 4 and 8 minutes after infection and are synthesized at a normal rate during infection with T7 gene 1 mutants grown under non-permissive condi- tions. Class I proteins are the products of T7 genes 0.3 to 1.3; these genes are referred to as early genes and comprise the early region of T7 DNA. Early genes are transcribed 13 3139 and 13 31112 by the host RNA polymerase. Ligase (gene 1.3) is classified as a class I protein because it is synthesized at a normal rate during infection with gene 1 mutants even though it usually continues to be synthesized until class II protein synthesis ends. (b) Class II proteins are synthesized between 6 and 15 minutes post infection and are specified by genes 1.7 to 6. (c) Class III proteins are specified by genes 7 to 19 and are synthe- sized from about 6 minutes after infection until lysis (about 25 to 30 minutes at 30°). The genes specifying Class II and III proteins are referred to as late genes and are expressed only if functional T7 gene 1 product is present. As with other phage genomes, genes with related functions tend to cluster along the genetic map of T7 DNA: genes 2 to 6 affect DNA synthesis (9), genes 7 to 17 specify proteins that are found in mature T7 phage particles (8), and genes 18 and 19 are involved in maturation of T7 DNA from DNA replicating intermediates (11,12). Protein §ynthesis in T7-Infected Cells The order of appearance of T7 proteins in infected cells is the same as the order of the corresponding genes on the T7 genetic map (7); T7 genes are expressed accord- ing to their genetic map position. It has been found that the stop point for the transcription of the early region by host RNA polymerase (to the right of the ligase gene) is not 100% effective-- small amounts of class II and class III proteins are made after infection with T7 gene 1 mutant phage in which the stop signal is present. The reason for this may be that a small amount of gene 1 product is present, which regulates late genes, or host RNA polymerase post termination signals are present to prevent widespread transcription of late genes by host RNA polymerase. The latter appears more likely because a double mutation in gene 1 produces the same amount of late proteins. It is not known why the synthesis of class II proteins stop midway during phage infection; the synthesis of class III proteins slows at the same time. Regulation of class II and III protein synthesis could be controlled at either the level of transcription or translation, or possibly both. The mechanism responsible for deter- mining the relative rates of synthesis of different phage proteins is also unknown. At least 20 major T7 polypeptides have been syn- thesized 19 31999 in a protein synthesizing system derived from uninfected 9. 9911 and programed by RNA extracted from T7-infected cells (13). Identification of poly- peptides as T7 gene-specific products was made in experi- ments using T7 amber and deletion mutants. The pattern of polypeptide synthesis 19 31119 programed with RNA extracted from infected cells was compared to the pattern observed 19 3139 throughout the time course of T7 infec- tion. The results indicated that nearly all polypeptides displayed a similar time course of synthesis 19 31919 and 19 vivo--the fidelity of the 19 vitro pattern extended even to relative differences in the levels of synthesis among most of the various T7 polypeptides. These results make it unlikely that alterations in ribosomes, or initiation factors play a major role in the regulation of T7 gene expression. Not only is phage protein synthesis controlled during T7 infection, but host protein synthesis is inhibited. This inhibition may be required to insure that a sufficient supply of amino acid precursors is available for phage protein synthesis and that the host protein synthesis machinery is available for the synthesis of phage proteins. T7 phage infection induces two nega- tive control mechanisms of protein synthesis (14): host protein synthesis is "repressed" by a "T7 repressor" and early T7 phage protein synthesis is inhibited by a late phage protein. The "repressor" for host protein synthesis is an early T7 protein, but is not one of the known early genes. The "repressor" must be an early T7 gene product because mutations in gene 1 are capable of inhibiting host protein synthesis. These mechanisms do not dis- tinguish between a translational or transcriptional control mechanism for synthesis of host and early phage proteins. It has been suggested that the expression of early phage genes is inhibited by a late protein (14). Pleiotropic Effect of the T7 Gene 1 Product T7 amber mutants in genes 2 to 19 delete only one of the 25 to 30 proteins normally found in wild- type T7-infected cells. In contrast, gene 1 has a highly significant effect on T7 protein synthesis. During infection with gene 1 amber mutants only two or three of the T7 early proteins are synthesized while none of the T7 late genes are expressed (15). Thus, it is clear that unlike the other T7 genes, gene 1 is capable of exerting a pleiotropic effect on the expres- sion of T7 late genes. The mechanism by which gene 1 exerts its pleiotropic effect on late gene functions was determined by Chamberlin and co-workers who found that the gene 1 product is a T7-specific RNA polymerase (16). It was found that the T7 RNA polymerase tran- scribes only the late region of T7 DNA 19 31999 which explains why the gene 1 product has a pleiotropic effect on the expression of late phage genes. The T7 RNA polymerase was found to be insensitive to the antibiotics rifampicin and streptolydigin. Two derivatives of rifamycin were found to be effective inhibitors of T7 RNA polymerase when present at high concentrations (100 ug per ml); these derivatives inhibit bacterial RNA polymerase when present at a much lower concentration (10 ug per ml). 10 The results of Chamberlin and co-workers on T7 RNA polymerase were extended by Gelfand and Hayashi (17) using a coupled transcription-translation system from uninfected 9. 9911 cells dependent on T7 DNA and host RNA polymerase. When rifampicin was added to this system after 20 minutes of RNA synthesis, RNA synthesis became rifampicin resistant which is consistent with the presence of a new rifampicin—resistant RNA poly- merase activity. However, when T7 DNA with a gene 1 mutation was used as template, RNA synthesis stopped immediately when rifampicin was added, indicating that the rifampicin—resistant RNA polymerase activity was not present. RNA Synthesis in T7-Infected Cells The pattern of RNA synthesis in T7-infected cells was found to closely resemble the pattern of T7-specific protein synthesis (8,15). After infection of 9. 9911 with phage T7, 12 or 13 phage-specific RNA species are synthesized while only 5 of the normal T7 RNA species are synthesized if protein synthesis is blocked by chloramphenicol. If cells are infected with T7 phage containing an amber mutant in gene 1 then only four T7-specific RNA species are detected. These results indicate that gene 1 is transcribed in the absence of protein synthesis and is required for transcription of 11 the remainder of the T7 genome (late region). The control of T7 late protein synthesis is therefore mediated at the level of transcription by means of positive regulation-- the appearance of the T7 RNA polymerase controls the transcription of T7 late genes and therefore the syn— thesis of late proteins. The RNA transcripts and proteins specified by the 5 early T7 genes have been identified by sodium dodecyl sulfate—polyacrylamide gel electrophoresis (18, 19). The five early transcripts are from genes 0.3, 0.7, l, 1.1, and 1.3. Several small RNA species of unknown function were found which probably originate from the left of the 0.3 gene. The early T7 RNA species can be mapped by heteroduplex mapping of deletion mutants in the early region, which together with genetic and electrophoretic analysis of T7 RNA transcripts and proteins has produced a detailed physical map of T7 DNA (20). The position of genes on T7 DNA is given as a percentage of the total length of wild type T7 DNA measured from the left end of the T7 genome. The 5 early genes occupy the region between 1.8 and 20.2, while the host RNA polymerase stop signal is mapped at 20.2. The size of the five early RNA species found 19 3139 are, from the left of the genome, 0.24, 0.70, 1.15, 0.24 and 0.47X106 daltons corresponding to genes 0.3, 0.7, l, 1.1 and 1.3. The 12 molar amount of each RNA is the same except for the gene 0.3 transcript which is present in a 10—fold excess (20). Using the electron microscope to observe DNA-RNA hybrids constructed 19 31919 with 19 3139 RNA synthesized in T7-infected cells and the H strand of T7 DNA, Hyman (21) found that after infection essentially the entire 1 T7 genome is expressed as RNA. The early RNA species mapped contiguously on the genome starting at map position 0.010 and ending at 0.198 (the repetitive ends on T7 DNA are terminally redundant and extend to 0.007 [22]). The size of the early region measured in this way is in ‘ good agreement with the size of the early region deter- mined by heteroduplex mapping (20). Unlike bacterial messenger RNA T7 RNA is meta- ‘ bolically stable. Actinomycin D and rifampicin decay ‘ experiments show that T7 RNA is long lived, with a half life of more than 20 minutes (23,24). Marrs and Yanofsky (25) have shown that messenger RNA transcribed from the 9139-operon of the host genome is degraded normally in T7-infected cells with a half—life of 2.5 minutes. Marrs and Yanofsky also confirm that T7 RNA is stable when measured by hybridization to T7 DNA. With the use of T7 amber mutants in gene, which produce only early RNA 19 3139, Yamada 99‘91 (26,27) has found that the functional activity of T7 early RNA-- tested in an 19 vitro protein synthesizing system--decayed 13 at a rapid rate (half-life of 6.6 minutes). The ability of the RNA to direct f-met tRNA binding also decayed rapidly. This functional decay was related to a loss of structural integrity of the RNA, as detected by polyacrylamide-agarose gel electrophoresis, which indi— cated that the size of the RNA became smaller with loss of functional activity. Thus, although T7 RNA is meta- bolically stable, the ability of the RNA to direct 19 31919 protein synthesis decayed at a rapid rate. Why T7 RNA is stable and 9. 9911 RNA is not stable is not clear. Perhaps T7 RNA has some structural features that make it resistant to degradation. Summers and Szybalski (28,29) were the first to find that poly (G) binds to only one of the two strands of T7 DNA. This allowed separation of the two T7 DNA strands by CsCl gradient centrifugation in the presence of poly (G). The DNA strand that bound poly (G) banded at a region of high density in CsCl equilibrium gradients and was referred to as the "heavy" or H strand, while the DNA strand that did not bind poly (G) banded at a lower density and was referred to as the "light" or L strand of T7 DNA. The poly (G) binding sites on T7 DNA consist of dC-rich clusters of 15 to 40 nucleotides, as deter- mined by the melting temperature and sedimentation rates 3 of H-labeled poly (G) fragments recovered from RNase T1 14 treated T7 DNA-poly (G) hybrids (28,29). Summers and Szybalski determined that 30 to 75 poly (G) binding sites exist evenly distributed on each T7 DNA molecule. It was further determined (30) that only the poly (G) binding strand (H strand) of T7 DNA was transcribed during all phases of T7 phage development. Pulse- labeled phage RNA isolated from T7-infected cells during different phases of phage development hybridized only to the H strand of T7 DNA. The Shutoff of Host RNA Synthesis in T7 Phage-Infected Cells Summers and Siegel have found (31) that after infection with wild type T7 phage, host RNA synthesis gradually ceases until 5 to 7 minutes post infection when it is undetectable by hybridization. The shutoff of host RNA synthesis may be due to the degradation of host DNA or to a more direct mechanism such as the modification of host RNA polymerase. The breakdown of host DNA is a late gene function, while host RNA shut- off occurs in infection with gene 1 mutants. Therefore, an early phage protein is probably responsible for the shutoff of host RNA synthesis, not the degradation of host DNA. In support of this idea, a T7 mutant which fails to synthesize one of the early T7 proteins of 40,000 daltons (gene 0.7) has been shown by Brunovskis .v I :1fo 11". 15 and Summers to be deficient in the ability to shutoff host RNA synthesis (32,33). A possible mechanism of host RNA synthesis shut- off has been recently elucidated by Hesselback 99_91 (34,35) who find that a phagefinduced protein is associated with host RNA polymerase and is responsible for inactivation of host RNA polymerase activity in infected cell extracts. This inhibitor protein can be removed from host RNA polymerase by phosphocellulose column chromatography; removal of the inhibitor protein resulted in the reactivation of host RNA polymerase. The inhibitor protein has been purified to homogeneity and consists of a 14,000 dalton polypeptide. The purified protein specifically inhibits the holoenzyme form of RNA polymerase. It does not inhibit core enzyme or the T7 RNA polymerase. The inhibitor protein was found to in- hibit RNA synthesis at the level of initiation by binding to the holoenzyme and preventing it from binding to DNA. The inhibition was not specific for T7 DNA but occurred when other DNA templates were used. These results have been confirmed by Ponta 99 91_ (36). Although it is reported that gene 0.7 is the host RNA shutoff gene (32,33), extracts of gene 0.7 mutant T7—infected cells have completely inactivated RNA poly- merase activity (35); it is more likely that the host RNA shutoff gene is gene 0.3 which has been shown by 16 Ratner (37) to be the only DNA binding protein among the T7 early phage proteins and to have a molecular weight of 14,000 daltons. This possibility, however, has not yet been tested directly. Bacteriophage T7 codes for a protein kinase (38). From a study of RNA and protein synthesis after T7 infection it has been suggested that the phage specific kinase may be responsible for the early transcriptional control observed during T7 infection, including the shutoff of host RNA synthesis. It was shown that in kinase—deficient mutants early RNA species are over- produced and the shutoff of host RNA synthesis is delayed (39). Zillig and co-workers (40) have recently proposed a mechanism for the action of the T7 protein kinase in early transcriptional control. Their work indicates that the 8' subunit, and to a lesser extent the 8 subunit, of 9. 9911 RNA polymerase are phosphorylated by the T7 protein kinase along with about 40 other host polypeptides during T7 infection. Phosphorylation of the B' and 8 subunits of 9. 9911 RNA polymerase may modify the transcriptional specificity of 9. 9911 RNA polymerase in T7-infected cells resulting in the shutoff of host RNA and early phage RNA synthesis. It is also possible that some of the other phosphorylated proteins may function in turning off genes. It has yet to be 17 shown that phosphorylated E. coli RNA polymerase has altered transcriptional specificity 19 vitro. Binding of Host RNA Polymerase to T7 DNA Hinkle and Chamberlin (41-43) have utilized the fact that RNA polymerase is quantitatively retained on nitrocellulose filters to study the binding of RNA polymerase to T7 DNA. Complexes between RNA polymerase and labeled T7 DNA are retained on nitrocellulose filters with an efficiency of 70% while T7 DNA itself is not retained. The attachment of a single RNA polymerase molecule to a molecule of T7 DNA is sufficient to cause retention of the complex. Both RNA polymerase holoenzyme and core enzyme bind to T7 DNA. Holoenzyme binds to two classes of binding sites on T7 DNA: class A sites which result in the formation of a stable complex and class B sites which result in formation of an unstable complex. Eight RNA polymerase (holoenzyme) molecules can bind to a T7 DNA molecule to form highly stable complexes. These stable complexes are thought to be located at specific promotor regions (44). There are many sites on the T7 DNA molecule for unstable complex formation, where the holoenzyme binds weakly and reversibly. It was found by Chamberlin and co—workers that holoenzyme can locate and bind tightly to class A sites with a half-time of 18 20 seconds after being added to T7 DNA. The rate limit- ing step in site selection by holoenzyme is the binding and release from the many class B sites on T7 DNA which occurs before the enzyme forms a stable complex at a class A site. The relatively low affinity of holoenzyme for class B sites is essential for rapid location class A sites. In contrast, core enzyme binds to one class of binding site but there are many such On T7 DNA. of only sites of It is apparent that the sigma subunit RNA polymerase is not required for binding of RNA polymerase to DNA-~both enzymes bind with high affinity to T7 DNA. It is possible to suggest that RNA poly- merase may exist in two conformational states--one designed for promotor binding and initiation (holoenzyme) and the other for RNA chain elongation (core enzyme). The sigma subunit may function in site selection by switching the enzyme between the two conformational states. Single-stranded breaks in T7 DNA also serve as binding sites for RNA polymerase. Single-stranded breaks in T7 DNA enhance transcription by core enzyme by pro- viding sites for RNA chain initiation (45). Single- stranded breaks in T7 DNA inhibit holoenzyme activity. Inhibition of holoenzyme activity is due to a decrease in the amount of enzyme that can initiate RNA synthesis when DNA containing single—stranded breaks serves as l9 template (single-stranded breaks are tight binding sites for holoenzyme but few such sites can serve as RNA chain initiation sites). Thus, the structure requirements for [a tight binding site for RNA polymerase on T7 DNA are less stringent than the requirements for an RNA chain initiation site; fidelity of initiation of RNA synthesis depends not only on recognition of binding sites by RNA polymerase but also on rigid structural requirements for RNA chain initiation once the enzyme is bound. Transcription of T7 Phage Early RNA The exact mechanism by which early RNA is tran- scribed by host RNA polymerase has not been completely defined. A number of models have been proposed, includ- ing: (a) post—transcriptional cleavage of a precursor RNA transcript of the entire early region by a sizing factor, RNase III, into five early RNA species. (b) punctuation of transcription of the early region by the transcription termination factor rho. (c) existence of independent promotor and terminator sequences for each early phage gene. The experimental basis for these different models will be discussed below. (a) Transcription of T7 DNA 19 31919 by purified E. coli RNA polymerase without added factors produces 20 large RNA molecules which correspond to the entire early region of T7 DNA (46-48). The early region is tran- scribed starting at three closely spaced independent initiators located near the left end of T7 DNA and ending at a terminator at 20.2; RNA chains started from two of the initiators begin with ATP, those from the third promotor begins with GTP. Endonuclease activity from uninfected cells can cleave the large RNA molecules at specific sites to generate RNA species essentially the same size as the early T7 RNA species found 19 3139. The large precursor RNA species can also be cut by purified RNase III to yield similar results: 5 large and 3 small RNA molecules; the small 3 RNA species come from the left of gene 0.3. In agreement with this (49), the early region of T7 DNA is transcribed as a single, large RNA molecule in 1 a RNase III—deficient mutant. This RNA precursor is essentially identical to the RNA molecules produced 19 21919 by host RNA polymerase. These RNA molecules are cleaved by RNase III in the same manner as are the 19 31919 precursor RNA molecules. It should be noted that the RNase III mechanism predicts equal amounts of each early RNA, but, as described above, this is not seen 19 3139. (b) The action of rho termination factor during transcription as originally described by Roberts (50) 21 can be summarized as follows: rho factor depresses over- all synthesis of RNA by decreasing the length of RNA chains and restricts transcription of host RNA polymerase to particular regions of DNA. Rho factor interacts with RNA polymerase at specific DNA sites leading to formation of RNA species of well-defined length. Rho factor acts catalytically affecting many enzyme molecules. The following model was suggested by Dunn 99_91 (51) in order to explain early T7 RNA transcription 19 31919: all host RNA polymerase molecules with sigma factor start transcription on T7 DNA at a single unique initiation site at the left end of the T7 genome and propagate towards the right. There are several rho- dependent RNA termination signals, one for each early RNA, where rho—mediated release of RNA occurs. The RNA polymerase is not released and continues to propagate along the H strand of T7 DNA. After transcription of the early region, a second kind of termination site releases the polymerase. Goldberg and Hurwitz (52) have found that rho factor can interact with DNA to form a protein-DNA complex, but it is doubtful that this is the mechanism of action. The binding of rho to DNA is not stoichio— metrically related to its effects on transcription, nor is the effect of rho factor altered by a lO—fold excess of DNA. The ratio of rho factor to RNA polymerase 22 is the most critical factor. Rho factor appears to interact with host RNA polymerase at specific DNA sites leading to formation of RNA chains of well defined length. Even though rho factor can terminate RNA synthesis at specific sites it does not bind tightly to RNA polymerase, in fact its presence is not needed until RNA polymerase approaches the termination site. The mechanism of rho termination is not yet known. Richardson 99H91 (53) have recently found that rho factor has RNA-dependent ATPase activity and may be the product of the host SuA gene. Poly (C) is particu— larly effective in stimulating the ATPase activity of rho factor (54). If this is true then rho factor may play a significant role in causing polar effects (55). Richardson 99 91_ propose that rho factor recognizes some structural features in nascent RNA as it emerges from the RNA polymerase molecule. It is proposed that the binding of rho factor to this nascent RNA site affects RNA polymerase in such a way that RNA polymerase is released from DNA template resulting in termination of RNA synthesis. If, however, the rho binding site on the nascent RNA is translated by a ribosome before rho factor binds then termination of RNA synthesis would be prevented. If translation is prematurely terminated at a nonsense codon, then the ribosome would leave the nascent RNA and expose the emerging nascent RNA to rho 23 factor. Thus, a mutation in one gene of an operon could reduce the level of expression of the genes in the operon that lie on the operator—distal side of the mutant gene producing the polar effect. In summary, in the absence of rho factor one large RNA transcript is synthesized, while transcription in the presence of rho factor produces five different RNA species. In addition, transcription in the presence of rho factor is totally asymmetrical and restricted to the early region of the T7 genome (56). The role of rho factor 19 3139, however, is not clear since only two of the RNA species transcribed in the presence of rho factor 19 31919 appear to be identical to 19 3139 early RNA species. An alternative explanation for the function of rho factor has been proposed (56). Since it has been shown that only some of the RNA products of 19 31919 RNA transcription in the presence of rho factor match RNA species present in T7—infected cells 19 3139, rho factor may be a defense mechanism of the host cell to disrupt the transcription of the early region by termi- nating transcription at incorrect positions; this would be particularly fatal if transcription of gene 1 were interrupted. However, there is no evidence that rho factor acts 19 vitro in the same manner that it does 24 19 3139; it may in fact recognize only correct termi- nation points 19 3139. (0) Another model for 19 3139 early RNA syn- thesis has been proposed by Minkley (57,58). He sug- gests that each early gene is controlled by its own promotor and terminator. He has identified lZ-specific RNA species syn- thesized at early times during T7 infection in ultraviolet—irradiated host cells. Ultraviolet- irradiation of host cells prevents host protein and RNA synthesis but not phage macromolecular synthesis (59). The molecular weights of these RNA species range 6 daltons. The sum of the molecular from 0.02-1.05X10 weights can fit within the early region on the T7 genome without overlaps. The early RNA species identified by Minkley were not present in equimolar amounts. Thus, they could not originate by post-transcriptional cleavage of a single precursor molecule of the entire early region as suggested by the RNase III model of transcription. The ability to correlate a specific T7 protein with a monocistronic transcript of its gene by gel electro- phoresis indicated to Minkley that the rate at which proteins are synthesized during phage T7 infection may be dependent on the number of copies of each RNA tran— script present; in otherwords, on the efficiency of 25 initiation sites for each RNA transcript. The location of the four largest transcripts have been mapped by Minkley within the early region of T7 DNA. Three ATP initiation sites, one GTP initiation site and four termination sites within the early region were found 19 31919 by labeling RNA molecules at their 5' termini with Y —32P-labeled nucleoside triphosphates (58); an extremely short GTP initiation transcript from the incorrect L strand of T7 DNA was also found. A possible reason for the failure of other investigators (46-48) to obtain these results is that highly specific initiation of 19 31919 RNA synthesis occurs only at low I molar ratios of RNA polymerase to DNA template. RNA polymerase is capable of initiating RNA synthesis at random nonspecific sites at high molar ratios of enzyme to DNA template. Minkley and Pribnow have found that dinucleotides can be used to stimulate transcription 19 31919 from selected initiation sites on T7 DNA (60). Using this method seven initiation sites (three very close together in the early promotor region) and five termination sites were mapped relative to known deletions in the early region of the T7 genome. Over twenty H strand-specific RNA products were formed; most resulted, however, from readthrough of termination signals. Despite this there was a strong correlation between the size of the products 26 synthesized 19 31919 by this method and the size of 19 3139 early RNA species. The ratio of products initiated with ATP to GTP by this method varied from about one to four, showing only a modest preference for ATP. ATP was used exclusively at three initiation sites, GTP was used at two sites, and either seem to be used at one site. In addition, Minkley and Pribnow found that holoenzyme was capable of transcribing 19 31919 a product initiated with GTP with a molecular weight of 80,000 daltons from the L strand of T7 DNA. Initiation sites may consist of a strong RNA polymerase binding sequence adjacent to or including the DNA sequence of the 5' end of the corresponding RNA (61-63). Thus, the dinucleotide-DNA complex may be stabilized by the presence of RNA polymerase. RNA polymerase should be capable of binding to the RNA polymerase binding sequence on DNA and also bind the dinucleotide, if the dinucleotide contains the sequence of the 5' end of the corresponding RNA. Minkley and Pribnow found that several different dinucleotides could stimulate synthesis from the same initiation site (60). Thus, there may be a short region (5-6 nucleotides) where RNA polymerase can initiate with a paired dinucleotide. CpA was ten times more efficient in priming RNA synthesis than was ATP, probably because of the extra 27 stability provided by the dinucleotide. The most effec- tive dinucleotide at most initiation sites contained a purine nucleoside in the 3' position of the dinucleotide which corresponds to the 5' triphosphate of the RNA chain initiated at that site. Less efficient dinucleo- tides contained the purine nucleoside triphosphates in the 5' position and presumably have the RNA initiation base as the 3' end group. CpC was very efficient in initiating RNA synthesis, but it is not known if it is contained in the sequence before or after the 5' end of the corresponding RNA. One possible fault of this procedure is that it may increase the relative efficiency of weak initiation sites (60). In summary, three models for early RNA tran— scription have been proposed: 1. The host RNA polymerase synthesizes one continuous transcript from the entire early region. The precursor RNA transcript is cleaved by RNase III into the five RNA species observed 19 3139. 2. Only one major RNA polymerase binding site exists and transcripts are synthesized by a rho mediated termination-reinitiation mechanism without release of the polymerase at several rho termination sites through- out the early region. 3. Autonomous RNA initiation and termination sites exist throughout the early region, each of which 28 controls the synthesis of RNA from an independent tran- scriptional unit. I The first two models are not capable of completely describing early RNA synthesis as it exists 19 3139. The first model cannot adequately explain the 10 fold excess amount of gene 0.3 RNA observed 19 3139. The second model is inadequate because rho factor fails to produce RNA molecules 19 31919 that are the same size as those found 19 3139. The third model predicts that proteins are synthesized at a rate dependent on the efficiency of the initiation sites for each RNA tran- script. If this is true, there should be a correlation between the molar amount of a given monocistronic tran- script and the molar amount of the corresponding protein. This is found to be true: the gene 0.3 product is the most plentiful early protein in T7 infected cells and the gene 0.3 RNA transcript is the most prominent RNA transcript. The other early proteins are present in more nearly equal molar amounts as are their RNA tran— scripts. However, the specificity of 19 31919 tran- scription observed by Minkley is not in total agreement with the 19 3139 specificity of transcription. Characterization of the T7-Specific RNA Polymerase The T7 RNA polymerase has been extensively studied by Chamberlin and co-workers (64—68). The T7 29 RNA polymerase is distinguishable from host RNA poly- merase by its stringent template specificity and resistance to the bacterial RNA polymerase inhibitors rifampicin and steptolydigin. Only T7, T3 and salmon sperm DNA have appreci- able template activity. The activity of T7 RNA poly- merase with T3 DNA as template is about one-half that with T7 DNA as template. The enzyme shows some activity with salmon sperm DNA but large concentrations of salmon sperm DNA are needed for maximum activity. Denaturation of T7 DNA decreases its template activity and also abolishes the absolute requirement of T7 RNA polymerase for its specific promotor. Denaturation of T7 DNA results in a loss of strand specificity during transcription. Thus, poly (dT) is an active template for T7 RNA polymerase while poly (dA) ' poly (dT) is not; but the loss of specificity is not complete since T7 RNA polymerase will utilize only poly (dT) and poly (dC) as templates. The DNA templates poly (dC) ° poly (dG) and poly (dI) ' poly (dC) are active templates with T7 RNA polymerase, but only the dC-containing strand is tran— scribed. Promotors recognized by holoenzyme are called Class I promotors while those recognized by T7 RNA poly- merase are called class II promotors. All phage and bacterial DNA templates tested contain class I promotors, 30 while only T7, T3, and possibly M13 DNA contain class II promotors. Class II promotors may also be present on salmon sperm DNA, but at a much lower concentration than on T7 DNA. The specificity of host RNA polymerase for class I sites is lost when single-stranded DNA templates are transcribed or when DNA templates containing single- stranded breaks are transcribed. The T7 RNA polymerase also loses specificity when some single-stranded DNA templates are transcribed but does not lose specificity when DNA templates containing single-stranded breaks are transcribed. Chamberlin and co-workers have also studied the inhibition of T7 RNA polymerase activity. Synthesis of RNA by T7 RNA polymerase is inhibited by actinomycin D, heparin, high salt concentrations and poly r(U); all these inhibitors block transcription immediately, even after initiation of RNA synthesis, indicating that RNA chain growth is specifically inhibited. RNA chain elongation by holoenzyme is not sensitive to these inhibitors (except actinomycin D). RNA synthesis catalyzed by T7 RNA poly— merase is not affected by the bacterial RNA polymerase inhibitors rifampicin and streptolydigin (except those derivatives noted above) or by preincubation of the enzyme with poly r(C), poly r(I) or a variety of heterologous DNA molecules. Host RNA polymerase is stimulated by KCl up to a concentration of 0.15M, then 31 activity is inhibited sharply between 0.15-0.3M KCl while the T7 RNA polymerase is inhibited between 0-0.15M KCl with no activity observed above 0.2M KCl. When poly r(U), or heparin are added before RNA synthesis is initiated the host RNA polymerase is in- hibited; however, when added 15 seconds after initiation of RNA synthesis they have no effect on RNA synthesis. In contrast, poly r(U) and heparin inhibit T7 RNA poly- merase activity equally well if added before or after RNA synthesis has begun. Ultraviolet-irradiated T7 DNA inhibits initiation of RNA synthesis by T7 RNA polymerase (ultraviolet—irradiation of T7 DNA blocks its template activity). The following mechanisms of inhibition by heparin and ultraviolet-irradiated T7 DNA have been suggested by Chamberlin and co—workers (65). Heparin inhibits chain elongation catalyzed by T7 RNA polymerase presumably because the DNA binding site of T7 RNA polymerase is partly exposed during RNA synthesis and heparin interacts with it. It is also possible that it interacts with another site on the enzyme. Ultraviolet-irradiated T7 DNA acts as an inhibitor of T7 RNA polymerase activity by providing a promotor for the binding of T7 RNA poly- merase but inhibiting RNA chain elongation by the bound T7 RNA polymerase molecules. When ultraviolet-irradiated T7 DNA is added after RNA synthesis has begun with native '32 T7 DNA as template it has little effect on the reaction (for a short reaction period). Ultraviolet-irradiated DNA is therefore an inhibitor of free T7 RNA polymerase and does not affect RNA chain growth catalyzed by T7 RNA polymerase once the enzyme has already initiated RNA synthesis. Late Phage RNA Synthesis in Infected Cells Golomb and Chamberlin (67-68) have found that 19 31919 transcription of T7 DNA by T7 RNA polymerase produces six discrete size classes of T7 RNA species with a broad range of molecular weights (from 2x105-5x106 daltons). The six major T7 RNA species are synthesized in approximately equimolar amounts, with the exception of species III, which is present in twice the amount of the other species and may actually consist of two RNA spices transcribed from separate regions of the T7 genome. The RNA species are initiated independently with GTP at the 5' end and elongated at a rate of 230 nucleotides per second under standard 19 31919 conditions. When artificially shortened T7 DNA templates are tran- scribed, four of the seven RNA spices are truncated or deleted. This indicates that the four RNA spices are terminated near the right end of T7 DNA, probably at a common termination site near 98.5%. Since the approximate lengths of the transcripts are known, the promotor sites 33 for the T7 RNA spices can be mapped at 56, 64, 83, and 97% on T7 DNA. Only one major T3 RNA species is tran- scribed by T7 RNA polymerase with a promotor at 83% and a terminator at 98.5%. Niles 99 91 (69) found that the RNA products synthesized 19 31919 by the T7 RNA polymerase are larger than the late RNA transcripts found in infected cells. It has been concluded therefore by Dunn and Studier (47) that late RNA transcripts may also be cleaved by RNase III. Niles 99 91 (70,71) have recently used frac- tionated and unfractionated late T7 RNA species syn- thesized 19 31919 by the T7 RNA polymerase to program an 19 31919 protein synthesizing system derived from uninfected 9. 9911. They conclude from observing polyacrylamide gels of both the RNA transcripts and the translation products that virtually the entire late region of T7 DNA is transcribed 19 31919 by T7 RNA polymerase. The initiation sites found agree with those described by Golomb and Chamberlin, but in addition they found an initiation site between gene 1 and 1.3 for T7 RNA polymerase, which, if recognized 19 3139, explains why gene 1.3 continues to be transcribed after the other early genes are shutoff. They have also found a termination site to the right of gene 10 not described by Chamberlin. Band VI, as described by Chanberlin, 34 does not appear to stimulate the synthesis of any T7 protein 19 vitro. In addition, they found that the majority of the RNA transcripts transcribed by T7 RNA polymerase 19 31919 translate into class III proteins i_E_V_i_tr_°- The individual T7 late RNA spices are transcribed by a pattern of overlapping transcription and are tran- slated with different efficiencies (70,7l). Genes l7 and 19 are transcribed from three initiation sites, but both proteins are translated in small amounts 19 3139 and 19 31919. However, gene 10 is also transcribed from two or three initiation sites but is very effectively translated both 19 31919 and 19 3139. More dramatically, gene 11 is transcribed from only one initiation site but is in greater molar yield than either gene 17 or 19. Control of RNA Synthesis in T3-Infected Cells The control of RNA synthesis in T3-infected cells is similar to that described above for T7-infected cells. Dunn 99 91 (51) have found that five early T3—specific RNA species are synthesized from the H strand of T3 DNA 19 31919. In addition, he found that 19 31919 tran- scription of T3 DNA by 9. 9911 RNA polymerase can closely approximate the 19 3139 pattern of RNA products only if the rho termination factor is present under optimal ionic conditions. In the absence of rho factor host RNA 35 polymerase transcribes both strands of T3 DNA but syn- thesizes only anti—late RNA species. 19 31919 tran- scription from the H strand of T3 DNA is limited to the early region of the T3 genome in the presence of rho factor at low ionic strength (0.05M KCl). The RNA products were similar in size to those found 19 3139. Transcription from the L strand of T3 DNA is strongly depressed in the presence of rho factor. L strand termination occurs even at high ionic strength (0.2M KCl) indicating that there may be strong preferential termination of L strand transcripts. It was concluded (51) that the termination site between the early and late regions of T3 DNA operates at approximately 50% efficiency 19 31919 in the absence of rho factor, while the same site operates at 95% effi- ciency on T7 DNA. Thus, rho factor is needed to prevent readthrough to the late region of T3 DNA but not T7 DNA. On T7 DNA there is incorrect L strand initiation but also rho-independent termination. With T3 DNA the corresponding termination site has a low efficiency and rho factor is needed to improve asymmetry of tran- scription. The T3 specific RNA polymerase transcribes only the H strand of T3 DNA 19 31919 (51). The products are discrete RNA species similar in size to those isolated late in infection from T3—infected cells. Rho factor 36 has no effect on transcription by T3 RNA polymerase. Unlike the T7 RNA polymerase, the T3 RNA polymerase transcribes both early and late regions of the T3 genome 1239:2- Dunn 99 91 have confirmed these results using another experimental technique (72). The size distri- bution of the RNA species coding for three T3-specific enzymes—-the SAM cleaving enzyme, T3 RNA polymerase and T3 1ysozyme--were determined by sedimenting the RNA isolated from T3-infected cells in a sucrose gradient. A cell-free protein synthesizing system was utilized, followed by enzymatic analysis of the translation products, to determine the position of the corresponding RNA transcript in the sucrose gradient. This procedure was followed with 19 3139 RNA and with RNA synthesized 19 31919 by host RNA polymerase from T3 DNA. In the presence of rho factor, synthesis of early RNA 19 31919 (coding for the SAM cleaving enzyme and the T3 RNA polymerase) was identical in size and coding capacity to early RNA isolated from infected cells. It was also found that RNA transcribed by T3 RNA polymerase consisted mainly of late RNA (coding for T3 lysozyme) but also contained functional early RNA. The T3-specific RNA polymerase has been charac— terized by Maitra and by Bautz and their co-workers (73- 77) and is similar to the T7-specific RNA polymerase 37 described by Chamberlin 99 91 (16). It consists of a single polypeptide of 105,000 daltons and requires the presence of Mg+2, nucleoside triphosphates and T3 DNA for optimal activity. T7 DNA is utilized as a template, but only at 2—5% of the optimal activity with T3 DNA as template. It was found that rho factor has no effect on the rate, yield or size of RNA formed 19 31919 with T3 DNA as template. T3 RNA polymerase was found to be highly sensitive to salt concentrations above 0.03M KCl and to sulfhydryl group reagents. The enzyme catalyzes a T3 DNA dependent 32 P-labeled PPi exchange into nucleoside triphosphate, but the reaction is limited to GTP alone. No pyrophosphorolysis of free RNA was found. Maitra and co-workers have also studied the transcription of denatured DNA templates with the T3 RNA polymerase (78). The T3 RNA polymerase transcribes a variety of denatured DNA templates including denatured T3, T7, calf thymus, flXl74 and fd DNA templates, but at markedly reduced rates. Thus, as with the T7 RNA poly- merase, it is evident that the high degree of template specificity of the T3 RNA polymerase resides both in the enzyme and in the secondary structure of T3 DNA template. The kinetics of 19 31919 RNA synthesis catalyzed by T3 RNA polymerase show that activity is critically dependent on GTP concentration (79). The nature of the 38 data obtained indicates that T3 RNA polymerase initiates chains with pppGpG, while with T7 DNA as template T3 RNA polymerase initiates RNA chains with pppGpA. A high level of pyrophosphate exchange is obtained in the presence of GTP alone as substrate which confirms this interpretation. T7 RNA polymerase, unlike T3 RNA polymerase, shows a high rate of PPi exchange with GTP alone as substrate on both T3 and T7 DNA. This suggests that there are differences in initiation regions for T3 and T7 RNA polymerases—-T3 RNA polymerase may be more stringent in its site speci— ficity. A further indication that the two phage poly- merases may differ in structure is the fact that RNA synthesis by T3 RNA polymerase is not inhibited by heparin, unless it is added at the same time as substrate, when it inactivates both unbound RNA polymerase and RNA poly- merase bound to DNA. Heparin will inhibit the T7 RNA polymerase regardless of when it is added. Maitra and co-workers have recently studied 19 31919 transcription by both the T3 and host RNA poly— merase (80). RNA synthesized by T3 RNA polymerase 19 31919 hybridizes only to the H strand of T3 DNA, as does all 19 3139 phage RNA. Host RNA polymerase transcribes RNA that hybridizes with both the H and L strands of T3 DNA, although greater than 70% hybridizes with the H strand. In contrast to other studies summarized above, the ratio of RNA hybridization to the H or L strand was v 39 unaltered by the presence of excess sigma factor, or the presence of rho factor, or by the molar ratio of RNA polymerase to DNA present in the reaction mixture. It was shown by Maitra 99 91 that the 19 vitro products of T3 RNA polymerase contain all late RNA sequences and a11 early RNA sequences. Thus, T3 RNA polymerase transcribes the entire early and late region of T3 DNA 19 31919. The RNA products of host RNA poly- merase transcription of T3 DNA 19 31919 contain all the sequences present in early 19 3139 RNA, but also contain 50% of the sequences present in late 19 3139 RNA. In the presence of rho factor all early sequences were present, but few late RNA sequences were present. Thus, rho factor restricts host RNA polymerase transcription to the early ‘ region of T3 DNA 19 31919. Dunn and Bautz have found different template specificities with the T3 and T7 RNA polymerases (81).. The two RNA polymerases transcribe T3 and T7 DNA with greatly different efficiencies. T7 RNA polymerase tran— scribes T7 DNA about twice as effectively as T3 DNA, while T3 RNA polymerase has very low activity on T7 DNA but is highly active with T3 DNA. Hercules 99 91 have found (82) that the processing of early T3 and T7 RNA by RNase III is necessary for their efficient translation both 19 3139 and 19 31919. Uncleaved T3 and T7 RNA species are poor templates for 19 vitro 40 synthesis of the phage enzymes SAMase, DNA ligase and lysozyme. Only small amounts of phage enzymes are syn— thesized after T3 or T7 infection of 9. 9911 RNase III- deficient strains (83). This low rate of phage—specific enzyme synthesis 19 3139 is not due to a lack of phage- specific RNA, since equal amounts of phage RNA are found in wild-type and RNase III-deficient hosts. There is a low rate of translation of phage RNA from RNase III- deficient host cells in an 19 31919 RNase III-deficient protein synthesizing system; this activity is probably a result of residual RNase III activity (83). If sizing of RNA were required for translation of bacterial mRNA, then the RNase III—deficient strain should be inviable. However, the strain used retains some RNase III activity, which may be enough to maintain the slow growth rate observed. Srinivasan and co-workers have presented evidence that the host RNA polymerase is modified after T3 phage infection (84). The modified host RNA polymerase behaves differently in its elution properties from DEAE-cellulose, phosphocellulose and DNA-cellulose columns. The specific activity of the modified enzyme is about one-fourth that of the unmodified host RNA polymerase which Srinivasen indicates may be due to an alteration in the 8' subunit of host RNA polymerase. Modification of the host RNA polymerase also occurs in infection with T3 gene 1 mutant 41 phage. However, an analogous modification was not found by Srinivasan in 9. 9911 infected with T7 phage. Thus, it appears that a T3 early gene controls the modification of host RNA polymerase but a similar gene does not exist on the T7 genome. Srinivasen 99 91 also found evidence for an inhibitor protein capable of inhibiting 9, 9911 RNA polymerase (85,86). This protein has no effect on core polymerase, indicating a requirement for sigma factor. The protein did not bind directly to DNA or RNA poly- merase or to DNA-RNA polymerase complex, instead the presence of ATP and GTP are required for the binding of the inhibitor protein to RNA polymerase-DNA complex. The inhibitor protein is not an early protein, thus it must be a late gene under the control of the T3 RNA polymerase. This is in contrast to the T7 inhibitor protein which appears to be an early.T7 protein capable of directly binding to host RNA polymerase (34,35). INTRODUCTION The studies in this dissertation were initiated largely with the discovery by Chamberlin and co—workers (16) that bacteriophage T7 codes for a new RNA poly- merase upon infection of 9. 9911. It was hoped that the bacteriophage gh—l which was previously characterized by Lee and Boezi (87) would also induce a novel RNA poly— merase upon infection of Pseudomonas putida. The search for a new RNA polymerase in gh—l-infected 1. putida cells was then undertaken. The results of the purification and characterization of the gh-l-induced RNA polymerase have been recently published and a reprint is presented in the appendix. The gh-l-induced RNA polymerase proved to be similar to the T7 RNA polymerase. Despite this it was hoped that the control of transcription in gh-l-infected cells could be defined on a molecular level and that the control mechanisms found in gh—l-infected cells would be significantly different from the mechanism of transcriptional control in T7-infected 9. 9911 which was being elucidated at the same time as work was pro- gressing on the gh-l phage system. It was also hoped that the mechanisms of transcriptional control operating 42 43 in gh-l-infected cells could be duplicated 19 31919 with highly purified RNA polymerase and gh—l DNA. The results described in this dissertation clearly show that the control of transcription in gh-l— infected cells is lost during late periods of infection. A model describing how this loss of control may take place at the molecular level is presented in the Dis- cussion. The loss of control during late time inter- vals of gh-l infection is indicated by the appearance of complementary RNA. Complementary RNA is not syn- thesized during T7 phage infection and is known to be synthesized in very few systems either procaryotic or eucaryotic. This dissertation provides evidence that the gh-l phage system is unlike other phage systems described in that transcription of gh-l phage RNA is not stringently controlled during infection. This dissertation also indicates that the molar ratio of RNA polymerase to phage DNA may be important for the control of RNA transcription 19 3139. It has been shown by other investigators (57, 58) that the molar ratio of RNA polymerase to DNA is important in limiting transcription to certain regions of DNA 19 31919 but no evidence has been provided that it is important to the control of RNA transcription 19 3139. Finally, the gh-l phage system is the only phage system so far described in which large amounts of 44 complementary RNA are synthesized 19 3139. While the function of complementary RNA, if any, is not known the gh-l phage system provides an excellent opportunity to study the significance of complementary RNA. Before the results of this dissertation are discussed previous work on the gh-l phage system will be quickly reviewed. The gh-l phage was originally isolated by Lee and Boezi (87). The size and shape of the gh-l phage particle is similar to the T7 phage particle. In addition, the size of the gh-l genome (23X106 daltons) is also similar to the size of the T7 genome. The latent period of the gh—l phage lytic cycle is about 21 minutes with a burst size of 100 plaque forming units (PFU) per infected cell. The RNA polymerase from uninfected Pseudomonas putida has been purified to near homogeneity and charac- terized by Johnson 99 91_ (42). The enzyme has a sub- unit composition analogous to the 9. 9911 RNA polymerase and like the 9. 9911 enzyme transcribes a large variety of DNA templates 19 31919 and is sensitive to the drug rifampicin. The gh-l-induced RNA polymerase has also been purified to near homogeneity and characterized as described in the appendix. The reader should read the appendix first since these results are not discussed in the Results section. MATERIALS AND METHODS Materials Whatman DEAE-cellulose (DE 52) and phosphocellulose (PCl and PC 11) were purchased from Reeve Angel. Lysozyme, chloramphenicol, bovine serum albumin and B-mercaptoethanol were obtained from Sigma. Calf thymus DNA, unlabeled ribonucleoside triphosphates and dithiothreitol were from P-L Biochemicals. 3 3 H-labeled thymidine (6.7 Ci/mmol), H-labeled CTP (26.2 Ci/mmol), 3H-labeled uridine (26.2 Ci/mmol), and Aquasol were purchased from New England Nuclear. Bio—Gel P-150, Bio-Gel A-1.5m, and agarose were obtained from Bio—Rad Laboratories. Poly (U,G), poly d(A—T) and poly d(C)° poly d(G) were from Miles Labo- ratories. 9. 9911 RNA polymerase was purchased from Boehringer Mannheim, ribonuclease from Cal Biochem and deoxyribonuclease I from Worthington Biochemical Corpo- ration. Nitrocellulose membrane filters were obtained from Schleicher and Schuell. "Stains-all" dye was from the Eastman Kodak Company. The rifamycin derivatives were gifts from Dr. Luigi Silvestri, Gruppo Lepetit, Inc., Milan, Italy. Cordycepin triphosphate and 9. 9911 ribosomal RNA were the kind gifts of Dr. Fritz Rottman of this department. T3, T4 and T7 phage were generous 46 gifts of Dr. Loren Snyder, Department of Microbiology and Public Health, Michigan State University. Growth of Pseudomonas Putida and Infection with Bacteriophage gh-l Pseudomonas pgtida (ATCC 12633) was grown at 33°C in a medium containing, in grams per liter: yeast extract, 5; tryptone, 5; glucose, 5; NaCl 8; NaZHPO4, 6; KH2P04I logarithmic phase, gh—l phage was added to a multi- 3. After cell growth had reached the mid- plicity of 5-10 plaque forming units (PFU) per bacterial cell. The infected cells were allowed to incubate for 10 minutes and were then poured onto a half volume of crushed, cooled (-20°C) ice and centrifuged. The infected cell pellets were quick-frozen in an acetone-dry ice bath and stored at -20°C. Cells infected by this procedure were used for the preparation of gh—l RNA polymerase and competitor RNA. In all other cases, where complete and synchronous infection was required, another procedure was used. To insure complete and synchronous infection, after 9. putida cells had reached the mid-logarithmic stage of growth they were concentrated ten-fold by centrifugation and chilled to 4°C. Gh-l phage was then added at a multiplicity of 5-10 PFU per cell. At 4°C adsorption of gh—l phage to 9. putida cells takes place. After 10 minutes at 4°C, the culture was diluted into 47 pre-warmed media (33°C). The time of dilution into pre- warmed media was taken as the 0 time of infection. Plaque Assay of Bacteriop99ge and One Step Growth Experiment Bacteriophage gh-l was assayed by the plaque assay technique described by Lee and Boezi (87). The one step growth experiments were as described by Lee (90) with the exception that rifampicin (100 ug per ml) was added to the media at the indicated times. The one step growth experiments with T3 phage were done in a similar way except that T3 antiserum was not available. The multiplicity of infection was 1 PFU per 9. 9911 cell. Plaque assays of the infected cell culture during the one step growth experiments were done at 10 differ- ent time intervals during a 60 minute incubation period. The dilution of the infected culture was such that at least 100 plaques were formed for each plaque assay. The one step growth experiments were repeated at least twice with identical results. Purification of Bacteriophage and Isolation of DNA Bacteriophage gh-l was purified from 9. putida lysate by the method of Lee and Boezi (87). The phage was purified from cell debris by differential centri- fugation and further purified by chromatography on 48 DEAE-cellulose. Bacteriophage T3 was purified from 9. 9911 lysate by differential centrifugation and by banding in a preformed CsCl density gradient. Bactri- ophage DNA was isolated by the method of Thomas and Abelson (88) while 9. putida DNA was isolated by the procedure of Thomas 99_91 (89). Commercial calf thymus DNA was extracted twice with phenol (in the presence of 0.1% SDS) and extensively dialyzed before use. RNA Polymerase Assay The RNA polymerase assay of 9. putida and gh-l RNA polymerase measured the incorporation of 3H—labeled CMP into a TCA-insoluble product. The volume of the standard reaction mixture for host RNA polymerase was 0.125 ml and contained: 50mM Tris-HCl (pH 8), lOmM MgCl 50mM KCl, lmM DTT, 0.4 mM in each 2’ 2’ of the four NTP's, 500 ug per ml BSA, 50uCi per ml 3H- 2.5mM MnCl labeled CTP and 100 ug per ml calf thymus DNA. The final specific activity of the 3H-labeled CTP was 2X104 CPM per nmol. The assay was started by the addition of enzyme. The reaction was incubated at 30°C for 10 minutes and stopped by the addition of 0.2ml of a 0.1% SDS solution followed by 5ml of cold 10% TCA containing 1% sodium pyrophosphate. A few drops of 2.5mg per ml herring sperm DNA as carrier was added. The mixture was allowed to stand for about five minutes at 4°C and then 49 the precipitate was collected on nitrocellulose membrane filters. After washing the filters with 15ml of cold 10% TCA, they were dried at 80°C for ten minutes and counted by liquid scintillation spectroscopy using a toluene-based fluor (5ml). The procedure was the same for the assay of gh-l RNA polymerase except the standard reaction mixture contained lOOug per ml gh-l DNA and the MnCl2 and KCl were omitted. Various components were either added or omitted from the standard reaction mixtures as indicated in the legends to the appropriate tables and figures. Purification of RNA Polymerase The gh-l-induced RNA polymerase was purified by the method described by Towle, Jolly and Boezi (91). The 9. putida RNA polymerase holoenzyme and core enzyme were purified by the method of Johnson 99_91_ (92) with the following modifications: (a) all buffers contained 15% glycerol (V/V) with the exception of the final storage buffer which contained 50% glycerol (V/V). (b) Cell 1 disruption was accomplished by use of a French pressure call at 7000 lb per in. (c) The 30% ammonium sulfate fractionation step and the second DEAE-cellulose column chromatography step were omitted; instead, the final step in the enzyme purification was gel filtration using a Bio-Gel A-l.5m column. The final concentration of 50 enzyme was 22 mg per ml and the specific activity was 900 units per mg. By the procedures outlined above all RNA poly- merase preparations were greater than 90% pure as deter- mined by SDS-polyacrylamide gel electrophoresis. Both enzymes were judged to be free of RNase activity by the failure of the enzymes to alter the profile of 3H-labeled ribosomal RNA as analyzed by polyacrylamide-agarose gel electrophoresis after incubation of 0.1mg per ml enzyme with 3H-labeled ribosomal RNA (150,000 CPM per ml) for 30 minutes at 30°C. Similarly, the enzymes were judged to be free of DNase activity by their failure to change the sucrose density gradient sedimentation profile of 32P-labeled gh-l DNA. RNase III activity was also found to be absent when assayed by the procedure of Robertson e19 (93)- Preparation of 3H-labeled RNA in vitro l 3H—labeled RNA synthesis was carried out 19 31919 by the 9. putida holoenzyme and core enzyme and the 9. 9911 RNA polymerase in a reaction mixture containing 50mM Tris-HCl (pH 8), lOmM MgClZ, 2.5 mM MnCl 50mM KCl, lmM DTT, 0.4mM in each of the four 2' NTPs, 500 ug per m1 BSA, and 0.2 mCi per ml 3H-labeled CTP. The final specific activity of CTP was about 8X104 CPM per nmol. The amount of DNA and enzyme in fl'i’s”¥‘z—.fira‘7u—;'~_ ~.—eir~_.—e__.__r.-._==__ 1..._ _ i L i i, 51 each reaction mixture was varied as indicated for each figure. The reaction was begun by the addition of enzyme. The reaction was incubated for 45 minutes at 30°C and then chilled to 4°C. A small amount of solid DNase was added and the mixture was incubated at 37°C for 15 minutes. The 3H-labeled RNA was extracted twice with an equal volume of phenol and the aqueous phase was dialyzed extensively against 2XSSC. The procedure for 3 H-labeled RNA synthesis 19 31919 by the gh-l RNA polymerase was the same except the reaction mixture contained gh-l DNA and the MnCl2 and KCl were omitted. The conditions of 19 31919 3H-labeled RNA synthesis were such that the rate of reaction was proportional to enzyme concentration. The procedure for 3H-labeled RNA synthesis using infected cell extracts was similar to that above except various RNA polymerase inhibitors were added to inhibit either the host RNA polymerase or the gh-l RNA polymerase. To inhibit the host RNA polymerase rifampicin was added to a final concentration of 10 ug per m1 and streptolydigin to 100 ug per ml. The gh-l RNA polymerase was inhibited in the crude extract mixture by the addition of cordycepin triphosphate (3'—dATP) to a final concentration of 50uM (91). All components of the standard reaction mixture for RNA synthesis were present except DNA. The inhibitors were added 5 minutes prior to the addition of 3H—labeled 52 CTP. The crude extract of infected cells was prepared by lysing infected cells with lysozyme buffer as described. The final protein concentration in the reaction mixture was about 0.5 mg per ml. The 3H- labeled RNA was extracted by the procedure of B¢vre and Szybalski (94). Preparation of Pulse-Labeled RNA 9. putida cells were infected as described with the exception that the cells were grown in GSC media. The RNA was pulse-labeled with 3H—labeled uridine (25 uCi per m1) at appropriate time intervals. The pulses were terminated by rapidly chilling the cells in an ice-water bath and collecting by centrifugation. Fol- lowing centrifugation, the cell pellets were frozen in an acetone—dry ice bath and stored at -20°C until use. 3H-labeled RNA was extracted by the procedure outlined below for preparation of competitor RNA. Pulse-labeled chloramphenicol RNA was prepared in the same manner except chloramphenicol (400 ug per ml) was added to the media 5 minutes before infection and the RNA was pulse-labeled from 0-5 minutes post infection. Preparation of Competitor RNA 9. putida cells were grown and infected as described. At appropriate times the infection was terminated by pouring the culture onto a half volume 53 of crushed, cooled (-20°C) ice. The infected cells were collected by centrifugation and the infected cell pellets were quick-frozen in an acetone-dry ice bath and stored at —20°C. Chloramphenicol competitor RNA (CAM RNA) was prepared from cells infected for 5 minutes in the presence of 400 ug per ml chloramphenicol (added five minutes before infection). Total RNA was extracted according to the hot phenol method of B¢vre and Szybalski (94). Infected cells were lysed with lysozyme (200 ug per ml in a buffer containing 50mM tris-HCl (pH 8), lOmM MgCl and 0.1M KCl) and alternately freezing and thawing 2 the infected cell suspension in the presence of 0.1% SDS. After lysis, an equal volume of 0.1M sodium acetate buffer (pH 5.4) was added, followed by two volumes of freshly distilled phenol. The mixture was then vigorously swirled in a 65°C water bath for three minutes followed by rapid chilling in an ice—water bath. The liquid phases were separated by centrifugation and the aqueous phase was extracted with phenol two more times. The RNA in the final aqueous phase was precipitated with 2 volumes of cold ethanol (-20°C), and then collected by centrifugation at 12,000xg for 30 minutes. The RNA pellet was dissolved in a minimal amount of 2 SSC and dialyzed overnight against 2XSSC at 4°C. The final con- centration of RNA was at least 10 mg per ml. 54 9. c011 competitor RNA was prepared as described by B¢vre and Szybalski (94). DNA—RNA Hybridization 3H-labeled RNA was hybridized to DNA by two methods. The procedure of Gillespie and Spiegelman (95) involved the hybridization of labeled RNA to denatured DNA immobilized on nitrocellulose membrane filters. Gh-l DNA or host DNA (100 ug) in buffer containing lOmM Tris—HCl (pH 8) and lOmM NaCl was denatured by heating for ten minutes in a boiling water bath, followed by quick chilling in an ice-water bath. Three m1 of cold 10XSSC was added and the denatured DNA was slowly filtered onto nitrocellulose membrane filters that had been pre-soaked for 6 hours in 6xSSC. Each filter was then washed extensively with 6xSSC (100ml per filter) and allowed to air-dry overnight. The DNA containing filters were then baked at 80°C for 3 hours. Labeled RNA was incubated with DNA filters in 2XSSC (enough volume to cover the surface of the filter, about 2m1) for 24 hours at 65°C. The filters were then treated with RNase, dried and assayed for radioactivity as described below. The presence of gh-l DNA sequences in DNA extracted from infected cells was determined by hybridization of 3H-labeled gh-l RNA synthesized 19 vitro by 9. putida holoenzyme to extracted DNA. DNA 55 filters were prepared from DNA extracted from infected cells at different time intervals post infection (100 ug per filter) and 3H-labeled gh-l RNA was hybridized to the DNA. The time course of gh-l DNA synthesis was followed by the amount of 3H-labeled gh-l RNA that hybridized to the DNA extracted from infected cells at different time intervals. ' Labeled RNA was hybridized to the separated strands of gh—l DNA by the method of Nygaard and Hall (96). 3H-labeled RNA was hybridized to 2ug of either of the separated strands of gh-l DNA in a volume of 0.5 ml and a final salt concentration of 2XSSC. The hybrid— izationndxtureswere incubated at 65°C for 6 hours, chilled to 4°C, and then treated with RNase (5ug per ml in 2xSSC) for 15 minutes at 37°C. The RNase solution was pre-heated at 98°C for 10 minutes to destroy con- taminating DNase activity. The hybridization mixtures were then diluted with 2xSSC to a total volume of 10ml and filtered slowly through nitrocellulose membrane filters which had been soaked in 2xSSC for at least 6 hours. Each filter was washed with 100ml of 2xSSC and dried and counted as described. Since only denatured DNA and DNA-RNA hybrids but not RNA, duplex RNA or duplex DNA bind to nitrocellulose membrane filters under these conditions, the radioactivity of the filter reflects the amount of hybrid formed in each hybridization mixture. 56 To determine if DNA was in excess during hybrid- ization increasing amounts of input 3H-labeled RNA were hybridized to a fixed amount of DNA. If DNA is in excess, the relationship between input 3H-labeled RNA and amount of hybrid formed should be linear. This was found to be true for hybridization involving DNA immobilized on a nitrocellulose filter. Under the conditions of hybrid- ization employed for this method 3H—labeled gh-l or host RNA synthesized 19 31919 hybridized with nearly 100% efficiency. The hybridization efficiency (to both gh—l and host DNA) of 3H-labeled 19 3139 RNA under conditions of excess DNA by this method of hybridization was about 50%. Hybridization with the separated strands of gh—l DNA was not done under conditions of excess DNA due to the difficulty in preparing appreciable amounts of separated strands. Hybridization of 19 31919 3H-labeled RNA to the separated strands of gh—l DNA was accomplished with a 40% hybridization efficiency. The hybridization efficiency of 19 3139 3H—labeled early RNA isolated 0—5 minutes post infection to the separated strands of gh-l DNA was less than 1% because of the presence of excess 3H-labeled host RNA. The hybridization efficiency of 19 3139 3H-labeled late RNA (isolated at 8 minutes or later post infection) to the separated strands of gh-l DNA was about 20%. 57 All hybridization experiments were repeated several times. Hybridization blanks (minus DNA) were less than 100 CPM. Hybridization-Competition Hybridization-competition was carried out by the method of Nygaard and Hall (96) as described by B¢vre and Szybalski (94). The reagents were mixed at 4°C and contained in a total volume of 0.5 ml lOug of denatured gh-l DNA (or 2 ug of either separated strand of gh-l DNA), varying amounts of competitor RNA, 3H—labeled RNA (at least 10,000 CPM per hybridization mixture), and a final salt concentration of 2xSSC. Hybridization was carried out at 65°C for 6 hours. Under these conditions the hybridization efficiency was about 65% for 19 31919 3H-labeled RNA, 35% for 19 3139 RNA pulsed at late times post infection and less than 1% for 19 3139 RNA pulsed at early times of infection. Hybridization efficiency with separated strands of gh-l DNA was the same as that described above. After hybridization the hybridization mixtures were rapidly chilled to 4°C and then treated as described for the Nygaard and Hall method. Each point on the hybridization—competition curves represents an average of 2-4 determinations. Blanks (minus DNA) were less than 100 CPM. 58 Degradation of Host DNA During Infection A 10ml culture of 9. putida was grown in glucose- salts-casamino acid (GSC) media, which contained in 4, 6; KH2P04, 4 2; casamino acids, 5; glucose, 20; and trace amounts of grams per liter: N HPO 3; NaCl, 8; NH Cl, FeCl3, manganese acetate, and Na SO . The GSC media 2 4 contained in addition 2.5 mg per 10ml adenine and 3H- labeled thymidine (750uCi per 10ml). The cells were allowed to grow for three generations at which time unlabeled thymidine (0.6 mg per 10ml of growth medium) was added. After incubation for an additional 10 minutes the cells were chilled and collected by centrifugation. The cells were then suspended in lml of GSC media containing 0.6mg per 10ml thymidine and 2.5 mg per 10ml adenine, chilled and infected. Infection was begun by adding the cells to 9ml of pre-warmed GSC media (33°C) containing 0.6mg per 10ml thymidine and 2.5mg per 10ml adenine. Aliquots (1 ml) of the infected-cell culture were precipitated in 5ml of cold 10% TCA at the indi- cated time intervals and filtered onto nitrocellulose filters. The filters were dried and counted by liquid scintillation spectroscopy. The radioactivity of each filter was taken to be a measure of the amount of 3H— labeled DNA present at each time interval. 59 Separation of gh-l DNA Strands Samples (1 m1) of a solution containing 160 ug gh-l DNA, 300 ug poly (U,G), 0.1% SDS and 0.35mM EDTA were heated in a boiling water bath for 10 minutes and quickly cooled in an ice-water bath. CsCl solution (2ml, o=l.4080 [25°C]) was added to silicone-treated nitrocellulose centrifuge tubes. Sample (0.4ml) was layered into each tube and the tubes were filled with mineral oil to within 5mm of the top of the centrifuge tube. The tubes were centrifuged for 60 hours at 114,000xg in a SW 50.1 rotor at 10°C. After puncturing the bottom of the centrifuge tube 5-drop fractions were collected. The absorbance at 260nm and refractive index (25°C) were determined for alternate fractions. 1 The fractions containing the separated strands were pooled and dialyzed against 2xSSC for 12 hours (at 4°C). The pooled fractions were then dialyzed against 0.1N NaOH for 8 hours at 37°C followed by extensive dialysis against 2xSSC at 4°C. The separated strands were self- hybridized for 3 hours at 65°C and then chilled in an ice—water bath. Detection of RNA Duplex Formation by 81 Nuclease Digestion Samples of 3H—labeled RNA were heated in a boiling water bath for 10 minutes and then rapidly chilled in an ice-water bath. Aliquots of the RNA were 60 incubated under annealing conditions at 65°C for 6 hours and then chilled to 4°C. The "self-hybridized“ RNA and aliquots of unhybridized, denatured RNA (to serve as controls) were digested with 51 nuclease at 45°C in an assay mixture (0.200 ml) containing: 30mM sodium acetate (pH 4.5), lmM ZnSO 5% glycerol and 50 ug Sl nuclease. 4, 81 nuclease was purified by the procedure of Vogt (97) through the DEAE—cellulose chromatography step. Samples of the digestion mixture were precipitated in cold 10% TCA at the indicated time intervals and filtered onto nitrocellulose filters which were assayed for radio- activity. Digestion was continued until the control digestion mixtures no longer contained TCA insoluble radioactivity and there was no further digestion by $1 nuclease in the experimental mixture. Because the SI nuclease digests only single-stranded nucleic acid the radioactivity found in the "self-hybridized" RNA digestion mixture which is not digested by 81 nuclease is a result of RNA duplex formation during annealing of the RNA sample. The amount of radioactivity resistant to $1 nuclease digestion is therefore a measure of the amount of complementary RNA present in the 3H-labeled RNA sample. The amount of 3H-labeled RNA self—hybridized was at least 100,000 CPM. The amount of radioactivity remaining after digestion of the unhybridized, 3H-labeled 61 RNA controls was less than 500 CPM. Digestion was usually complete in 2 hours. Polyacry1amide Gel Electrophoresis 3H—labeled RNA was synthesized 19 31919 for electrophoresis as described except BSA was not added to the reaction mixtures (because of possible RNase contamination). The reaction mixtures were incubated for 30 minutes at 30°C. Rifampicin (lOug per ml) or the rifamycin derivative DNPI (97,98) (100 ug per ml) were then added to the reaction mixtures to inhibit initiation of RNA synthesis by the 9. putida or gh—l RNA polymerase respecively. After addition of the inhibitors reaction mixtures were incubated for an additional 10 minutes to insure that synthesis of all RNA molecules was complete. The reactions were terminated by addition of SDS to a final concentration of 0.1% and an additional 5 minutes incubation at 30°C to dissociate the RNA polymerase and RNA product from DNA template. The reaction mixtures were then chilled to 4°C and centrifuged to remove precipitated SDS and used directly for gel electro- phoresis. RNA was pulse labeled 16—18 minutes post infection as described above except the pulse was terminated by addition of rifampicin DNPI (lOOug per ml). The culture was incubated for an additional 2 minutes and cells were 62 collected as described above. 3H-labeled RNA was extracted by rapid freezing and thawing in lysozyme buffer followed by phenol extraction as described above. The 3H—labeled RNA was used directly for polyacrylamide gel electrophoresis. Electrophoresis was carried out on 1.75% poly- acrylamide-0.5% agarose cylindrical gels according to the procedure of Peacock and Dingman (99). Electro- phoresis was done at 5mA per gel for about 2.5 hours (30 minutes after the bromophenol blue marker had completely run off the end of the gels). The gels were sliced into 2mm slices using a gel slicer and each gel slice was placed in a scintillation vial and incubated with 0.5ml of 0.5N NaOH for 1 hour at 92-98°C. The vials were then cooled and 5ml of Aquasol containing 5ml of glacial acetic acid per liter was added. The vials were then vigorously shaken and radioactivity was determined by liquid scintillation spectroscopy. Polyacrylamide gel electrophoresis of each 3H- labeled RNA sample described in Results was repeated at least twice with essentially the same results in all cases. The yield of radioactivity from the gels was greater than 75%. 9. 9911 ribosomal RNA provided standard RNA markers for estimating the molecular weights of the labeled RNA species. The gel containing 9. coli 63 ribosomal RNA (40ug) was stained in a 100ml solution of 0.005% stains-all dye in 50% formamide for at least 24 hours. The gel was destained with distilled water in the dark and the positions of the ribosomal RNA species were measured directly. The molecular weight of RNA was estimated from a plot of log molecular weight vs fraction number. It was assumed that the molecular weight of 9. coli 23s ribosomal RNA was 1.1x106 and 165 ribosomal RNA was 0.55x106. Other Procedures Protein concentration was determined by the method of Lowry (100), using BSA as a standard protein. The concentration of nucleic acid was determined by assuming that the extinction coefficient of pure nucleic acid is absorbance at 260nm (1% solution) = 200. RESULTS Time Course of Nucleic Acid Synthesis In gh-l-Infected Cells The time course of RNA and DNA synthesis in gh-l- infected cells was examined to determine when phage— specific RNA and DNA are synthesized during the gh-l infectious cycle. As seen in Figure l, gh-l RNA synthesis for the first 6 minutes of the infectious cycle is very limited, less than 1% of the total RNA synthesized during this period hybridized to gh-l DNA. After this initial period, however, the relative amount of RNA hybridizing to gh-l DNA increases greatly. The RNA pulse-labeled 12-14 minutes post infection hybridized almost exclusively to gh-l DNA. Figure 1 also indicates a corresponding decrease in the relative amount of host RNA synthesis during the infectious cycle. RNA synthesized during the initial 6 minutes of the infectious cycle hybridized almost exclu- sively to host DNA. However, after 6 minutes of the infectious cycle the relative amount of RNA hybridizing to host DNA decreases drastically. Little or none of the RNA pulse-labeled 12-14 minutes post infection hybridized to host DNA. 65 Figure l.--Time course of RNA synthesis in gh-l—infected cells. 3H-labeled RNA synthesized in gh-l- infected cells during 2 minute pulse labeling periods was hybridized to both gh-l and 9. putida DNA filters under conditions of excess DNA. The percent input 3H-labeled RNA hybridized to either the gh—l DNA filter (0) or the 9. putida DNA filter (a) was determined for each pulse (designated by the midpoint of the 2 minute pulse interval). The amount of input 3H-1abe1ed RNA varied from 10,000 CPM to 100,000 CPM. The procedure for pulse— labeling infected cells and the conditions for hybridization are described-in Materials and Methods. 66 TIME POST INFECTION (minuteS) l 1 L0 8 N (°/o) OEZIGIHSAH VNH-Hg 1(1le 67 The hybridization data in Figure 1 provides information about the relative amount of RNA synthesis that is gh-l-specific or host-specific. It does not yield quantitative information about the absolute amount or rate of RNA synthesis in infected cells. Thus, the conclusion from Figure l is that the relative amount of gh-l RNA synthesis increases greatly and host RNA synthesis is shutoff after 6 minutes of the infectious cycle. The large increase in the relative amount of gh— 1-specific RNA can be explained by assuming that the appearance of the gh-l RNA polymerase results in a large increase in the transcription of gh-l DNA in infected Cells. This appears likely because the gh-l RNA poly- merase is first detected by 5 minutes post infection (98) and the large increase in the relative amount of gh-l- specific RNA synthesis starts shortly afterward at 6 minutes post infection (Figure l). The appearance of the gh—l RNA polymerase, how- ever, cannot account for the shutoff in host RNA syn— thesis observed in Figure l. The two most plausible reasons for the shutoff of host RNA synthesis are that the host RNA polymerase is modified or inhibited so that it no longer is capable of transcribing host DNA, or host DNA is degraded during infection and is no longer available to support RNA synthesis. 68 To determine if host DNA is degraded during gh-l infection host DNA was labeled by incorporation of 3H-labeled thymidine before infection. The labeled cells were then infected with gh—l phage and the degra- dation of host DNA, determined by the amount of 3H- labeled thymidine remaining TCA insoluble, was followed throughout the infectious cycle. Figure 2 indicates that host DNA was degraded but not until 8—10 minutes post infection. By 12 minutes post infection 65% of the labeled host DNA was degraded. The complete degradation of host DNA is not observed in Figure 2. Instead, the amount of TCA insoluble 3H-labeled DNA increases after 12 minutes post infection and reaches 75% of the amount of TCA insoluble 3H-labeled DNA found in uninfected cells by 16 minutes post infection. 3H-labeled gh-l RNA synthesized 19 31919 was hybridized to DNA extracted from infected cells at the indicated times in Figure 2. The 3H-labeled gh-l RNA served as a qualitative probe for the appearance of gh—l DNA in the DNA extracted from infected cells. The results in Figure 2 indicate that gh—l DNA synthesis begins after 12 minutes post infection and thus un- doubtedly accounts for the increase in TCA insoluble 3H-labeled DNA in infected cells after 12 minutes post infection. 69 Figure 2.—-Degradation of host DNA during infection. DNA in uninfected cells was labeled by incorpo- ration of 3H-labeled thymidine and the cells were infected with gh-l phage as described in Materials and Methods. At different time intervals after infection the radioactivity of 1 ml aliquots of the infected culture was determined by TCA precipitation (C)) as described in Materials and Methods. The ability of DNA isolated from infected cells at different times post infection to hybridize 3H—labeled gh—l RNA (0 ) was also determined. The amount of 3H—labeled thymidine incorpo- rated into DNA in a 1 ml aliquot of culture before infection (control) was 21,000 CPM. The amount of 3H-labeled gh-l RNA in each hybridization mixture was 50,000 CPM. (%) NOIIVZIOIHBAH BALLVWBH 70 O O .0. ID I fi _0 N o -9 —§ -00 o 0 "V 1» I I O O o no (°/o) EWBO‘IOSNI V31 VNG -H 9 TIME POST INFECTION (minutes) 71 It is apparent that host DNA degradation does occur in gh-l-infected cells, and that the process begins between 8 and 10 minutes post infection. The nucleotides resulting from the degradation of host DNA are then used for the synthesis of gh-l DNA beginning at 12 minutes post infection, with the result that about 75% of the nucleotides originally present in host DNA become indorporated into gh-l DNA. It is concluded from Figures 1 and 2 that shut— off of host RNA synthesis is probably a result of the degradation of host DNA in infected cells. This con- clusion is reasonable because the time course of the shutoff of host RNA synthesis and the degradation of host DNA are similar. Time Course of Early and Late gh-l- Specific RNA Synthesis The gh-l-specific RNA that is synthesized before the gh—l RNA polymerase is present in infected cells will be defined as "early" gh-l-specific RNA, or simply early RNA. Early RNA can be labeled with 3H-labeled uridine during the first 5 minutes of the infectious cycle in the presence of chloramphenicol. Since protein synthesis is inhibited by chloramphenicol, the synthesis of the gh-l-induced RNA polymerase will be inhibited. Thus, under these conditions, the transcription of gh-l- specific RNA will be limited to transcription of early 72 RNA by the host RNA polymerase. RNA synthesized by this procedure will be referred to as chloramphenicol RNA or CAM RNA. CAM RNA is equivalent to early RNA. Early RNA sequences by definition are those gh—l sequences syn- thesized by host RNA polymerase during the initial period of gh-l infection. This does not mean, however, that early RNA sequences are not synthesized at other periods of infection nor does it mean that they must be transcribed by the host RNA polymerase. The time course of early RNA synthesis was followed by determining the ability of RNA isolated from infected cells at different times during infection to compete with 3H—labeled CAM RNA in hybridization- competition experiments. It is clear that if all early RNA sequences or their complements are present then full competition of 3H-labeled CAM RNA will be observed. 3H-labeled early RNA can be prevented from hybridizing to gh-l DNA by competition from an excess amount of unlabeled early RNA for the same DNA site or by formation of RNA duplexes with an excess of RNA complementary to early RNA. RNA duplexes form under the same conditions that DNA-RNA hybrids form, but RNA duplexes are not retained by nitrocellulose filters. The results in Figure 3 indicate that by 2 minutes post infection the majority of early RNA sequences or their complements are present. By 4 minutes post 73 Figure 3.--Hybridization-competition of 19 vivo 3H- 1abeled RNA synthesized 0—5 minutes post infection in the presence of chloramphenicol. 3H-labeled gh-l RNA isolated from cells in- fected for 5 minutes in the presence of 3H- labeled uridine and chloramphenicol (400 ug/ ml) was competed with the following unlabeled RNA competitors: 2 minute RNA (A ), 4 minute RNA (A), CAM RNA (0), and 18 minute RNA (0). The procedure for pulse labeling in- fected cells and the conditions for hybrid- ization-competition are described in Materials and Methods. 74 :E\oEv aumawu ou Hmwwm wwumummwm ou QOprNHpflugma How macapflpco . . p mum mpcmuu Immasm Mow muscmooum wnB .H> mfl coflpmfluomcmua fizm URMAommmIHIQm mo th®EENm IRHOQ «am HIrm man Hmsnhm II as as .mmmmmm .m II me me mmmmmm .m oee memos eIem teen: .m II em msoz mmmmmm .M e.m Nae es HInm H.e mom ms HIam ooa nmom filam HIsm II mm . mcoz HInm .anwwamm meHWMmWWHnmm coaumwwmflmmmm cH mmmnmemaom mzm .NZQ B a n .c 8 me 09 42m cemeommmIHIQm mo coaumNflpaHnmmII.m mqmde 112 (holoenzyme) may transcribe only early RNA sequences 19 31919 was explored (Figure 13). The extent of early RNA synthesis was determined by hybridization-competition experiments in which 3H-labeled RNA synthesized 19 31919 was competed with unlabeled CAM RNA. If competition is complete, then only early RNA sequences or their complements are synthesized 19 31919. It was found that transcription by holoenzyme could be limited to the early region of gh-l DNA 19 31919 if the molar ratio of RNA polymerase to gh-l DNA was below 1.0. Apparently an excess of holoenzyme allows initiation at other promotors (late region) which are not normally transcribed 19 3139. This experiment however does not distinguish between early RNA transcripts from the L strand of gh-l DNA and complementary RNA transcripts from the H strand of gh-l DNA since both species are competed by CAM RNA competitor. The host core enzyme transcribed predominately early RNA sequences or their complements 19 31919 (Figure 14), but was not limited to transcription of early RNA sequences by decreasing the molar ratio of core enzyme to gh—l DNA template. Apparently the presence of sigma factor is necessary to limit tran- scription to the early region of gh—l DNA. 113 Figure l3.--Hybridization-competition of 3H—labeled RNA synthesized in vitro by host RNA polymerase holoenzyme. —3H-labeled RNA synthesized 19 vitro by host RNA polymerase holoenzyme at different molar ratios (R) of enzyme to gh-l DNA template was competed with un- labeled CAM RNA. The procedure for 19 vitro RNA synthesis and the conditions for hybrid- ization—competition are described in Materials and Methods. In the following cases the concentration of gh-l DNA was 250 ug/ml and the concentration of enzyme was: 2.7 ug/ml (R=0.5) (A), 13.7 ug/ml (R=2.5) (A), and 27.5 ug/ml (R=5.0) (O). In one case the concentration of gh—l DNA was 100 ug/ml and the concentration of enzyme was 120 ug/ml (R=50) (O). The data is replotted in the inset to yield straight lines. The intercept of each line is the value of the relative hybridization at infinite RNA competitor concentration. 114 2533 42m «.95.... _ N _ Roam « .. I/‘. .28 amnmmfiz: \_\ haum II I IOI w. on 00.. (°/o) NOIlVZICIItIBAI-I BAIiVl—IEIEI 115 A commercial preparation of 9. 9911 RNA poly- merase was found to be limited to transcription of gh-l early RNA sequences or their complements 19 31919 (Figure 15). This enzyme fraction contained both holo— and core enzyme. The molar ratio of RNA polymerase to gh-l DNA template was estimated to be about 3. Thus, 9. 9911 RNA polymerase, which transcribes only early RNA sequences from T3 and T7 DNA (8,51), surprisingly tran- scribed only early RNA sequences or their complements from gh-l DNA as well. The gh—l RNA polymerase transcribes only late RNA sequences 19 3139, but is capable of transcribing both early and late RNA sequences or their complements 19 31919 (Figure 16). The enzyme synthesized about equal amounts of early and late RNA 19 31919. This result was not dependent on the molar ratio of enzyme to DNA template. Perhaps an additional factor is necessary to limit the transcription of gh—l RNA poly- merase to the late region of gh-l DNA. Strand Specificity of gh—l RNA Synthesis in vitro To demonstrate that an 19 31919 transcriptional system can duplicate RNA transcription 19 3139, the strand specificity of transcription must also be It was determined that early RNA is tran- duplicated. scribed only from the L strand of gh-l DNA by host RNA 116 . . 3 Figure l4.--Hybridization-compet1t1on of H-labeled RNA synthesized 19 31919 by host RNA polymerase core enzyme at different molar ratios of enzyme to DNA template. 3H—labeled gh-l _ RNA was synthesized in vitro by the 9._p§£1§§ core enzyme at two dIfferent molar ratios (R) of enzyme to gh-l DNA template. 3H—labeled RNA synthesized at a molar ratio of R950 (45 ug/ml enzyme and 100 ug/ml gh-l DNA) 18 completed by CAM RNA competitor (0 ) and 18 minute RNA (0 ). RNA synthesized at a molar ratio of R=0.5 (2.7 ug/ml enzyme and 250 ug/ml gh—l DNA) is competed with CAM RNA (A) and 18 minute competitor RNA (A) ~ The procedure for in vitro RNA synthesis and the conditions for—hybridization—competition are described in Materials and Methods. 117 UNLABELED COMPETITOR RNA (mg/ml) (°/o) NOIlVZICIIHBAH BALLV'IBtI 118 Figure 15.--Hybridization-competition of 3H-labeled RNA synthesized 19 vitro with 9. coli RNA poly- merase. 3H-labeled RNA was synthesized 19 vitro with 9. coli RNA polymerase (67 ug/ml) and gh—l DNA (100 ug/ml) as template. The molar ratio was estimated to be R=3 assuming that the enzyme preparation is 10% pure. The 3H-labeled RNA was competed with unlabeled CAM RNA (0) and 18 minute RNA (0 ). The procedure for 19 vitro RNA synthesis and the conditions for hybridization-competition are described in Materials and Methods. 119 1.5351 .p m mzo HIsm mo mpcmsum pwumummwm was 0» cORNNNRcHHbm: ch pmnwuomwp mm B .Nza HIsm mo mpcmupm pwumummwm may ou thRpHHQWS MOM wGOHUHUQOU mflmnmalm mSE .mH UGWI.mH .wH .mH mwhsmflw Ou mpcmmwfl swap mmwswzmmwswmooum mcu an mmmmw ca pmuflmwaucmm we mzm pwawanImm map cs ted. .coprucmwhm m >Huom oamaommm map Eonm pmumagoamo mm muse moH usonm we coflpcummmnm map mo pH . . . a: pmmmn COHmeHumm an we m mo msac> mafia coHmeSwmm mph co mfihnsm may umnu . . . eInm Awe m1 mm m N ihe.emc ammm em HIem m “Nooav mowm Amaznsm wuoo paw mEmNcmoHOSP H00 .m Ahmmc mmmm mm a Awmvv mmhw Awemucw muooP . Nee . ihb.emc meow m.e ateuam .a Awe mvv m Amfiwncw whooP . mess Ahmec see on acepam .a Admvv AwEmNcwoaonv 1 «Ram ihemi mesa m.o othusm .m Awom Awemwcmoaonw iheec seem “News emee em aceuam .m @599 wmmsmE>Hom mzm ER: vacuum m 3: 33 ohmzmwwpmmmmwmnmm on coflmuegnnmm on dawnsm mo Chumm HMHOS .mmmmm.mm coeumenomcmue 42m oemeomdmIeIcm no snudsssmaII.e mamas 124 enzyme is capable of transcribing only the biologically correct L strand of gh—l DNA 19 31919. The gh-l RNA polymerase is found to be capable of transcribing only the L strand of gh—l DNA 19 31919 (Table 4), even though it is not limited to late RNA transcription 19 31919 (Figure 16). When RNA synthesized by holoenzyme 19 31919 was examined for the presence of complementary RNA by the S1 nuclease assay it was found that nearly all H strand- specific RNA sequences were complementary sequences (Figure 17). For example, when the molar ratio of enzyme to DNA template was 50, about 40% of the RNA transcribed 19 31919 was H strand-specific (Table 4). If all the H strand-specific RNA were complementary RNA then after self-hybridization 80% of the RNA should be in the form of $1 nuclease-resistant RNA-RNA duplexes. This was found to be the case as seen in Figure 17 (a). When the molar ratio of enzyme to template was lowered to 0.5, about 20% of the RNA transcribed 19 31919 by (holoenzyme hybridized to the H strand of gh—l DNA (Table 4). If all H strand-specific RNA were complementary RNA then 40% of the RNA following self-hybridization should be $1 nuclease-resistant. This also was found to be true as indicated in Figure 17(a). In contrast, RNA synthesized by the gh-l RNA polymerase 19 31919 is all L strand-specific (Table 4) and therefore should not 125 Figure l7.--Detection of RNA duplex formation by S1 nuclease digestion of 3H—labeled RNA syn- . thesized in vitro. 3H-labeled RNA synthe51zed 19 vitro 5? 9. putida holoenzyme (a) is tested for the presence of complementary RNA by self—hybridization and 81 nuclease digestion (0) at molar ratios of enzyme to gh—l DNA template of R=50 and R=0.5 as indicated in the figure. An unhybridized RNA control was also digested with 81 nuclease (O). 3H-labeled RNA synthesized 19 VitrO by the g -1 RNA polymerase (R=50) (b) was also tested for the presence of complementary RNA as described in (a). The procedure for RNA synthesis 19 vitro by P. putida RNA polymerase is described in—the legend to Figure 14. The procedure for RNA synthesis in the legend to Figure 17. The conditions of self-hybridization and $1 nuclease diges- tion are described in Materials and Methods- 126 TIME OF INCUBATION (h) 127 form 81 nuclease-resistant RNA-RNA duplexes after self— hybridization. Figure 17(b) shows that this was observed to be true. It is apparent that nearly all RNA transcribed by holoenzyme from the H strand of gh-l DNA is comple- mentary RNA. The amount of the H strand transcribed was determined by hybridization-competition of labeled 19 31919 RNA (synthesized by holoenzyme) by CAM RNA and 18 minute late RNA for sequences on the H strand of gh—l DNA (Figure 18). This experiment was complicated by the fact that competition of labeled 19 31919 RNA can occur in two ways: (a) direct competition for H strand sequences by unlabeled complementary RNA present in 19 3139 competitor RNA (due to transcription of comple- mentary late RNA by host RNA polymerase 19 3139) and (b) indirect competition due to RNA-RNA duplex formation between labeled 19 31919 RNA and complementary 19 3139 competitor RNA. Formation of such duplexes would prevent labeled RNA from hybridizing to the H strand of gh—l DNA. The data in Figure 18 indicates that CAM RNA competitor competed 75—80% of the 3H—labeled 19 31919 RNA. Since there is no complementary RNA present in CAM RNA (Figure 11) the competition observed must have been indirect, indicating that 75-80% of the RNA syn- thesized 19 31919 by holoenzyme from the H strand of gh-l DNA was complementary early RNA. The remaining 128 Figure l8.--Hybridization-competition of 3H-labeled RNA synthesized in vitro by holoenzyme to the H strand of ghfi DNA. 3H-1abeled RNA was syn— thesized by 9. putida holoenzyme 19 vitro at a molar ratio of R=50 as described in the legend to Figure 14. The 3H-labeled RNA was competed by unlabeled CAM RNA (0 ) and 18 minute competitor RNA (0) for sequences on theIIstrand of gh-l DNA. The H strand was self-hybridized before use in hybridization- competition as described in Materials and Methods. The conditions of hybridization- competition are described in Materials and Methods. The data is replotted in the inset to yield straight lines. The intercept Of each line is the value of the relative hybridization at infinite RNA competitor concentration. 129 I N _ _I e N _ _,,, :2 E 0 g 2: I- .— N V E ' <1 -— CD 2 I. E l o: .9 IA _ \ H " —' < n- 0 2 at m I I o: E 8 E _(°/o)NOIlVZIOI88AH _, _ I: Lu 0. 2 I I o f o I. a DJ _J LIJ _ _ (I) <1 0 .J . S J O O 9 In (°/o) NOIl‘C/ZIGIHBAH BAIIV'IEIH 7130 20-25% must have been complementary late RNA as indi- cated by the fact that 18 minute late RNA completely competed the 3H-labeled RNA, although competition must have been both direct and indirect. It is not possible to conclude that the complementary late messenger RNA synthesized 19 31919 by holoenzyme is the same as the complementary late RNA synthesized 19 3139 by host RNA polymerase. The same experiment was also done with the L strand of gh-l DNA (Figure 19). This data indicates that 30% of the L strand-specific RNA transcribed by holoenzyme 19 31919 is early RNA while the remaining 70% must be late RNA. Size of gh—l-gpecific RNA Transcribed in vitro The size of RNA transcribed by holoenzyme 19 31919 was determined by polyacrylamide gel electro- phoresis. The conditions of RNA synthesis were such (as outlined in Materials and Methods) that only complete 'transcripts were synthesized. The RNA product was determined to be entirely from the early region of gh-l DNA and at least 75% L strand-specific. The results (Figure 20) indicate that five major RNA transcripts are synthesized 19 31919 with molecular weights of approximately 1.2, 1.5, 1.8, 2.2 and 2.5x106 daltons. In addition, a heterogenous group of smaller LA. ‘_.-‘- I» \ “A 131 Figure l9.—-Hybridization-competition of 3H—labeled RNA synthesized in vitro by holoenzyme to the L strand of Shel DNA. 3H-labeled RNA syn- thesized in vitro by the 9. putida holoenzyme (as descrIbed in the legend to Figure 18) was competed by unlabeled CAM RNA (<>) and 18 ndnute RNA competitor RNA (0) for sequences on the L strand of gh-l DNA. The conditions for hybridization-competition are described in Materials and Methods. We was 95 IS 132 P 1 o o o In (°/o) NOIlVZlCIItIBAH BALLV'IEIH UNLABELED COMPETITOR RNA (mg/ml) 133 RNA species are present on the gel. The total molecular weight of the five major RNA species, assuming each species is unique in sequence, is 9.2x106 daltons, which corresponds to about 80% of the gh—l genome. 6 daltons) is If the largest RNA species (2.5x10 assumed to be a precursor molecule of the entire early region then the early region would comprise 21.7% of the gh-l genome, which is in agreement with the size of the early region of T7 DNA. The smallest major RNA peak observed (1.2xlo6 daltons) is the correct size to code for the gh-l RNA polymerase, which almost certainly is the product of a ghel phage early gene. An attempt was made to elute 3H—labeled RNA from the polyacrylamide gel and characterize the RNA as early or late by hybridization-competition and as H or L strand-specific by hybridization to the separated strands of gh—l DNA. This, with the exception of one case, was not possible because the eluted 3H—labeled RNA did not hybridize efficiently to gh-l DNA. The 'inability to hybridize RNA eluted from polyacrylamide- agarose gels with high efficiency has been observed by other investigators (57) and is probably due to an inhibitory effect of agarose in the hybridization mix— ture. It was possible, however, to show that the pooled, eluted group of small heterogenious RNA species seen in Figure 20 hybridized to the H strand of gh-l DNA. These 134 Figure 20.--Polyacrylamide gel electrophoresis of 3H- labeled RNA synthesized 19 vitro by holoenzyme: 3H-labeled RNA was synthesized 19 vitro by B- putida holoenzyme for gel electrophoreSis as described in Materials and Methods. The molar ratio of enzyme to DNA was R=0.5 (250 ug/ml gh-l DNA and 2.7 ug/ml enzyme). The procedure for gel electrophoresis is described in Materials and Methods. The estimated mole- cular weight of each peak is indicated in daltons xlo-6. enIyII by f. s as IIIOIII lml :eduIe I r I o In (D 9* _o q- o “to U) m 01—. N _0 L0, (\I 00. N. _o N _. '0. N L I J I9 o In (Q_OIXWcI3) All/\IlOVOICIVtI FRACTION NUMBER LA “‘ 136 species may represent all the H strand—specific RNA in the sample since they comprise about 20% of the RNA on the gel. If this is true it indicates that only small RNA species are transcribed from the H strand of gh-l DNA 23 2&- It is seen in Figure 12 that in yigg RNA contains a RNA species of l.2xlO6 daltons, which may be the same RNA species as that detected in yitrg. It is also noted that no RNA species larger than 1.2x106 daltons is present in yiyg. This indicates that the large RNA species syn- thesized in yitrg are probably large precursor molecules of early region of the gh-l genome. It is not known if such precursor molecules are synthesized in Kiyg. The size of RNA synthesized in yitrg by the gh—l RNA polymerase was also determined by polyacrylamide gel electrophoresis (Figure 21). The RNA species present, if unique in sequence, would represent about 100% of the gh-l genome. The 1.8, 1.6 and 1.4xlO6 dalton RNA species are probably large precursor molecules since RNA species ‘of this size are not detected in vivo (Figure 12). 137 Figure 21.--Polyacrylamide gel electrophoresis of 3H- labeled RNA synthesized in vitro by the gh-l RNA polymerase. H:labeled RNA was synthesized in vitro-by gh—l RNA polymerase at a molar ratio of R=30 (12.5 ug/ml enzyme and 109 ug/ml gh-l DNA) for gel electro- phoresis is described in Materials and Methods. The estimated molecular weight 0f each Peak is indicated in daltons x10'6- 138 mm. Omd .O._ _._N_ mmm. v._ m m._ r0 (€_Ol x was) All/\uOVOIOW 1A J‘ DISCUSSION Model for the Control of gh-l RNA Synthesis in Infected Cells A model for the control of gh-l RNA synthesis in infected cells is presented in Figure 22. The gh—l DNA is shown in the model to contain a L strand and a complementary H strand. The evidence that gh-l DNA contains H and L strands is that the strands of gh-l DNA can be separated by CsCl density gradient centrifugation when complexed with poly (U,G) (Figure 9). The gh-l genome is also shown to contain an early region and a late region. For simplicity the early region is placed at the left end of the gh-l genome. The early region is shown to be 20% of the gh—l genome because the size of the early region was estimated to be about 20% of the gh—l genome by deter- mining the size of early RNA species transcribed in yitrg (Figure 20). The evidence for the size of the early region of gh—l DNA is not strong evidence but it is consistent with the size of the early region of T7 DNA (20,21). In the model the time course of gh—l infection is divided into three time intervals. The RNA species synthesized during each time interval and the RNA 139 140 polymerase responsible for the transcription of the RNA species is illustrated in the model. During the early time interval of infection (0-5 minutes post infection) early RNA is transcribed from the early region of the gh-l genome by the host RNA polymerase. As indicated in the model, the early RNA (early[L]) is complementary to the L strand of gh-l DNA. The gh-l RNA synthesized before the gh-l RNA polymerase is present is early RNA by definition. The duration of the early time interval of infection is not known pre- cisely, but it was shown (Figure 3) that all early DNA sequences are not transcribed in infected cells until about 4 minutes post infection. The gh-l RNA polymerase is first detectable in infected-cell extracts at about 5 minutes post infection (98). Thus, the early time inter- val of infection probably lasts until about 5 minutes post infection when all early DNA sequences are tran- scribed and the gh-l RNA polymerase is present. The early RNA must be transcribed by host RNA pOlymerase because early RNA is transcribed in the presence of chloramphenicol, which prevents the synthesis of the gh-l RNA polymerase in infected cells (98). The evidence that early RNA is complementary to the L strand of gh-l DNA is that early RNA synthesized in infected cells hybridized to only the L strand of gh-l DNA (Table l). 141 During the intermediate time interval of infection (5-12 minutes post infection) early RNA complementary to the L strand of gh—l DNA (early RNA[L]) continues to be transcribed by the host RNA polymerase. In addition, as indicated in the model, late RNA is transcribed from the L strand of gh-l DNA (late RNA ]) by the gh-l RNA poly- [L merase. The evidence that early RNA continues to be synthesized during the intermediate time interval of infection is that about 10% of the RNA pulse—labeled 8-10 minutes post infection that hybridized to gh-l DNA was competed by CAM RNA (Figure 4). The remaining RNA that hybridized to gh-l DNA was competed by 18 minute competitor RNA indicating that most of the RNA synthesized during the intermediate time interval of infection is late RNA. The fact that the RNA pulse-labeled from 8-10 minutes post infection hybridized almost exclusively to the L strand of gh-l DNA (Table 1) indicates that both the early RNA and late RNA synthesized during this time interval are complementary to the L strand of gh-l 'DNA . It was not proven that host RNA polymerase tran- scribes early RNA complementary to the L strand of gh-l DNA during the intermediate time interval of infection as indicated in the model. However, it was shown that host RNA polymerase transcribes early RNA complementary to the L strand of gh-l DNA during the early time interval 142 Figure 22.--Mode1 for the control of gh—l RNA synthesis in infected cells. Details of this model are presented in the text. Early RNA and late RNA are represented as single lines for simplicity. It is not suggested that early and late RNA are transcribed as large pre- cursor molecules. The length of the line representing late RNA complementary to the H strand of gh-l DNA (late RNA(H ) indicates that about 20% of the RNA synthesized during the late time interval of infection capable of hybridizing to gh-l DNA is late RNA(H). It does not indicate that only a limited amount of the H strand late region is trans scribed. The extent to which the late region of the H strand of gh-l DNA is tran- scribed is not known. 143 :1sz ME} e $05838 aim «we; 3IEf Clii. :8 (Hit? iatedi; 30:. i”. l n- I re s sequi: . rifle: is dear: tumil‘ii 31h]? of p-‘Ji" rem; 31h; okra :1 cc 03' i- antes?- » debt: 3 (iii??- H i: :2 aunt?- igttfi" - 12 hours. Destaining was performed on a difiusion destainer in 10% trichloroacetic acid and 33% methanol for 6 hours. Gels were removed from the diffusion destainer and were incubated in 10% trichloroacetic acid at 30° until the background was clear (approximately 4 hours). Gels were stored at 4° in 10% trichloro- acetic acid. Other Methods—Extracts of either gh-l-infected or uninfected P. putida, which were used to assay RNA polymerase activity directly, were made by suspending cells in 2 volumes of Buffer A. These suspensions were sonicated for 11% min (in 30-s bursts) at a setting of 70 on a Biosonik sonicator and then were centrifuged at 16,000 X g for 20 min to remove cellular debris. RNA polym- erase assays were performed with varying amounts of extract to ensure the enzyme activity was linearly proportional to the protein concentration. The purification of P. putida RNA polymerase was performed by the method of Johnson et al. (20). The preparation of RNA po- lymerase used in these studies was more than 95% pure, as deter- mined by SDS-polyacrylamide gel electrophoresis. RESULTS A Novel RNA Polymerase Activity in Bacteriophage gh-I-in- fected P. putida—The first evidence that a novel RNA polym- erase is synthesized after gh-l infection of P. putida was obtained from measurements of the RNA polymerase activity in extracts of uninfected and gh-l-infected cells. In extracts from unin- fected cells, RNA polymerase activity was inhibited 97% by the addition to the reaction mixture of the antibiotics, rifampicin and streptolydigin (Table I). This activity is largely, if not entirely, due to the P. putida RNA polymerase, which is known to be sensitive to these antibiotics (20). In extracts from gh-l- infected cells, the specific activity of RNA polymerase was 11 times greater than the specific activity in extracts from uninfected cells. Furthermore, this activity from infected cells was in- hibited only 4% by the addition to the reaction mixture of the two bacterial RNA polymerase inhibitors. Addition to the reac— tion mixture of actinomycin D and nogalamycin, which inhibit RNA synthesis by binding to DNA, almost completely inhibited the activity from extracts of both uninfected and gh-l-infected TABLE I Specific activity of RNA polymerase in extracts of uninfected and bacteriophage gh-I -infected Pseudomonas putida Components of the standard reaction mixture and preparation of cell extracts were as described under “Experimental Proce- dure.” Rifampicin and streptolydigin, when added to the reaction mixture, were at concentrations of 5 ,ug per ml and 100 #g per ml, respectively. Actinomycin and nogalamycin, when present, were both at a concentration of 10 pg per ml. Reactions were initiated by the addition of extract to a final protein concentration between 50 and 400 pg per ml. 1725 cells (32, 33). These activities were, therefore, due to DNA- directed processes. In extracts from cells infected with gh-l in the presence of chloramphenicol, the specific activity of RNA polymerase was essentially the same as that in uninfected cells. This activity also was sensitive to rifampicin and streptolydigin. Thus, protein synthesis was necessary for the appearance of the rifampicin- and streptolydigin~resistant RNA polymerase ac- tivity. Although other interpretations are possible, these results can be explained most readily by the synthesis of a novel DNA- dependent RNA polymerase after gh-l infection of P. putida. This explanation was verified by the purification of the gh-l- induced RNA polymerase and by a study of its structure and catalytic properties. Purification of the gh-I-induced RNA Polymerase—The results of the purification of the gh-l-induced RNA polymerase, per- formed as described under “Experimental Procedure,” are shown in Table II. The Bio-Gel fraction which was used for many of the catalytic studies reported below had a specific enzyme ac- tivity of 42,000 units per mg. This represents a 280-fold puri— fication from the initial extract fraction. An accurate determina- tion of the specific enzyme activity of the glycerol gradient frac- tion could not be made due to the difficulty of determining pro- tein concentration at the relatively low level present in this frac- tion. An estimate of the protein concentration of the glycerol gradient fraction, however, could be made from the SDS-poly- acrylamide gel electrophoresis of this fraction (Fig. 10). By measuring the area under the peaks of the scan at 550 nm of the SDS-polyacrylamide gel and comparing with the area under the peaks of known amounts of the reference proteins, the amount of protein present in the gel could be determined. This determina- tion is dependent on the demonstration that the amount of stain absorbed by the SDS-polyacrylamide gel is linearly related to the amount of protein present (20). From this estimate of protein concentration, a specific enzyme activity of 86,000 units per mg was calculated for the glycerol gradient fraction. Analysis of the Bio-Gel fraction for RNase and DNase activi- ties, contaminants of RNA polymerase preparations which can alter the observed RNA polymerase activity, were negative. The Bio-Gel fraction also did not contain any RN ase III activity, the enzyme involved in the “sizing” of T7 early mRNA in E. coli (34, 35). The Bio-Gel fraction also is free of any host RNA polymerase activity. The slowest migrating polypeptides on SDS-polyacrylamide gel electrophoresis of the Bio-Gel fraction TABLE II Summary of purification Summary of purification of gh-l-induced RNA polymerase from 30 g (wet weight) of gh-Linfected Pseudomonas putida as described under “Experimental Procedure.” Specific activity of RNA polymerase Fthractdof Com t l . . , .puti a ponen s o the reaction mixture Extract of [Ectra‘ctgdof infected with uninfected . , ~cg’d w‘fth gh-l in the P. putida m e E 1 presence of g ' chloram- phenicol“ units/mg Standard ..................... 17 193 15 Standard plus rifampicin and streptolydigin. . . . ........... 0.6 186 0.5 Standard plus actinomycin and nogalamycin ............ 0.3 1.7 0.4 Total Recofvery S ifi Fraction Total ot ° zym ° pee C Dr em :ftivit; 225$; acuvxty mg “131-1: % units/mg Initial extract fraction ..... 2,900 44 100 150 NH401 wash fraction. . . . . . 1,300 39 89 300 DEAE fraction ............ 330 ' 21 48 640 Phosphocellulose fraction. . 7.2 6.3 14 8,700 Bio-Gel fraction ........... 0.62 2.6 6 42,000 Glycerol gradient fraction. (0.23)“ 2.0 5 (86,000)° ° Chloramphenicol was added to the growth medium to a final cancentration of 100 pg per ml 1 min before the addition of gh-l p age. “ Based on the protein concentration determination made from SDS—polyacrylamide gel electrophoresis of the sample as described under “Results.” l _O 0.9 _ B .0 6"" ABSORBANCE AT 550 nm .0 on I °—o 0.4 — — 0.2 ~— O— 1 l J | 8 IO 2 4 6 DISTANCE MIGRATED (Cm) FIG. 1. SDS-polyacrylamide gel scans of fractions from the purification of gh—l—induced RNA polymerase. Samples of the phosphocellulose fraction (A, 12.7 pg), Bio-Gel fraction (B, 9 pg), and glycerol gradient fraction (0, approximately 5 pg) of gh-l polymerase were subjected to SDS-polyacrylamide gel electro- phoresis as described under “Experimental Procedure.” Elec- trophoresis was performed at 4 volts per cm of gel length for 6.25 hours at 25°. After staining and destaining, the gels were scanned at 550 nm on a Gilford linear transport. The direction of migra- tion was from left to right. The arrows indicate the peak positions of the reference proteins: phosphorylase a (a), bovine serum albumin (b), and catalase (c). migrated significantly ahead of the 6 and [3’ subunits of purified P. putida RNA polymerase. The phosphocellulose and Bio-Gel fractions could be stored at ~20° in buffer containing 50% glyc- erol for several months with little less of activity, if the protein concentration was equal to or greater than 0.5 mg per ml. Analysis of the phosphocellulose fraction, the Bio-Gel fraction, and the glycerol gradient fraction was performed by SDS-poly— acrylamide gel electrophoresis (Fig. 1). The glycerol gradient fraction contained one major polypeptide which comprised ap— proximately 80 %, by weight, of the total protein present (Fig. 10). No other polypeptide present comprised more than 8% of the total protein. The molecular weight of the major poly- peptide was determined by comparison of its mobility to the mobility of the reference proteins, phosphorylase a (subunit molecular weight of 94,000), catalase (68,000), and bovine serum albumin (60,000). Using a standard curve of the logarithms of the molecular weights of the reference proteins to the distances of migration, a molecular weight of 97,000 was estimated for the major polypeptide of the glycerol gradient fraction. This major polypeptide is thought to be the only polypeptide comprising the gh—I polymerase. It is the only polypeptide which increased in relative purity in the last two steps of the purification procedure. Its increase in purity parallels the increase in specific enzyme activity of gh-l polymerase in these last two steps. Finally, the .5 a: 9‘ a) .N b CMP lNCORPORATION (units) 33 0 IS 20 25 FRACTION NUMBER FIG. 2. Glycerol gradient centrifugation of gh-l-induced RNA polymerase. gh-l polymerase (phosphocellulose fraction, 18 ug) was mixed with 150 pg of bovine serum albumin and diluted to 0.15 ml in a bufi'er containing 20 mM Tris-H01, pH 8.0, 0.5 mM dithio- threitol, and 0.2 M KC]. After dialysis for 6 hours against the same buffer, a 0.1-ml sample was layered on a 4.8-ml 10 to 30% linear glycerol gradient prepared in the above buffer and contain ing 0.5 mg per ml of bovine serum albumin. On three parallel gradients, 0.1-m1 samples of the reference proteins (Pseudomonas putida RNA polymerase holoenzyme (80 pg), Escherichia coli alkaline phosphatase (70 pg), and beef heart lactate dehydrogen- ase (7 #3)) Were layered. All gradients were centrifuged for it hours at 4° in a Spinco SW 50.1 rotor at 45,000 rpm. Aftercar- trifugation, 32 fractions of 0.16 ml were collected from each gradient and enzyme assays were performed on the fraction“: described under “Experimental Procedure.” The arrows indicate the peak positions of the reference proteins, P. putida RNA P01)“ erase (a2fifl'a), alkaline phosphatase (AP), and lactate dehydm' genase (LDH). The recovery of gh-l polymerase activity was approximately 90%. molecular weight of the gh-l polymerase, as determined by SDS: polyacrylamide gel electrophoresis, is consistent with a determl- nation of 98,000 made by glycerol gradient centrifugation and gel filtration (see below). The gh-l polymerase polypeptide was 10 to 15%, by weight, of the total protein in the phosphocell‘llO_se fraction (Fig. 1A) and 50 to 55% of the Bio—Gel fraction (Flt- 13). Molecular Weight and Structure of the gh-I induced RN A POW erase—The molecular Weight of the gh-l polymerase was calm- lated using experimentally obtained values for its sedimentation coefficient and molecular Stokes radius. A molecular weight value calculated in this manner is not dependent on assumpllonS concerning the shape of the macromolecule (27)- The sedimentation coefficient of gh-l polymerase was deli" mined by sedimentation velocity centrifugation in a 10 to 30% glycerol gradient (Fig. 2). The reference proteins (alkaline phos- phatase, lactate dehydrogenase, and P. putida RNA P01ymemse.) were centrifuged under identical conditions. Based on the sedl- mentation coefficients of the reference proteins, the EM polym' erase exhibited a sedimentation coefficient of 6.1 i 042 S- The molecular Stokes radius of the gh—l-induced RNA POIYm' erase was obtained by gel filtration on a Bio‘Gel l’-200 colum11 (Fig. 3). The reference proteins (alkaline phosphatase’ Jamie dehydrogeuase, glucose-6—phosphate dehydrogenase’ and hemo- globin) were chromatographed under identical conditions 10 standardize the column. Using the relationship, derived by ACkeTS (36), between molecular radius and distribution COefii' cient, the molecular radius of gh—l polymerase was calculated to be 38 A. By combining the molecular radius determined by? filtration and a sedimentation coefficient from sedimenl‘az‘wn l l l l l I; LDH GDH AP Hb ’3 .2 4.8- .— c 3 Z 9 l- 36.— d I T 0 O. m 8 2.4— - 3 CL 2 U I.2—- — l I l l l 30 so 70 90 ELUTION VOLUME (ml) FIG. 3. Gel filtration of gh-l-induced RNA polymerase on Bio- Gel P-200. gh-l polymerase (phosphocellulose fraction, 60 pg) was diluted to 1 ml in a buffer containing 20 mM potassium phos- phate, pH 7.5, 0.5 mM dithiothreitol, 0.2 M KC], and 7.5% (v/v) glycerol and was dialyzed against the same buffer for 6 hours. The l-ml sample was layered on the top of a. Bio-Gel P-200 (100- to 200- mesh) column (1.5 X 77 cm) which had been equilibrated previ- ously with a buffer containing 20 mM potassium phosphate, pH 7.5, 0.5 mM dithiothreitol, 0.2 M KC], 5% (v/v) glycerol, and 0.5 mg per ml of bovine serum albumin at 4°. The column was de- veloped in the same buffer at a. flow rate of 2.1 ml per hour and fractions of 1.5 ml were collected. To standardize the column, two samples containing markers were chromatographed under exactly the same conditions in subsequent runs. One sample contained the markers 0.2% (w/v) blue dextran 2000, Escherichia coli alkaline phosphatase (1 mg), and beef heart lactate dehydrogenase (100 pg). The markers in the second sample were 0.2% (W/v) blue dex- tran 2000, yeast glucose-6-phosphate dehydrogenase (500 pg), and bovine hemoglobin (1.5 mg). Fractions from the column were analyzed for various enzyme activities as described under “Ex- perimental Procedure.” Blue dextran 2000 and hemoglobin were assayed spectrophotometrically at 650 nm and 410 nm, respec— tively. The peak position of the markers are shown by the arrows: blue dextran 2000, V0; lactate dehydrogenase, LDH; glucose-6- phOSphate dehydrogenase, GDH; alkaline phosphatase, AP; and hemoglobin, H b. velocity centrifugation and assuming a 17 of 0.73 ml per g, a molecular weight estimate of 98,000 can be calculated for the gh-l polymerase. This value for the molecular weight is in good agreement with the value of 97,000 obtained using SDS-poly- acrylamide gel electrophoresis. Together these results indicate that the gh-l-induced RNA polymerase is composed of a single polypeptide with a molecular weight of 95,000 to 100,000. Characterization of RNA Synthesis by gh-I -induced RNA Polym- erase Using gh-I DNA as Template—The general requirements for in vitro RNA synthesis by the purified gh-l polymerase were examined by varying the components of the standard reaction mixture (Table 111). When the enzyme, the gh-l DNA, one of the four ribonucleoside triphosphates, or the Mg2+ was removed from the reaction mixture, little or no RNA synthesis occurred. Near maximal enzyme activity was maintained over a broad concentration range of 5 to 20 mM Mg” with the optimal ac- tivity occurring at approximately 10 mM (data not shown). No detectable RNA synthesis occurred when the MgH was replaced in the standard reaction mixture by the divalent metal ions (Mnfi, Znfi, or Ca2+) at concentrations between 0.5 and 8 mM (Table III). In fact, the addition of any of these divalent metal ions at 2 mM to the reaction mixture containing Mg2+ inhibited the enzyme activity 93 to 100%. The activity of the gh-l po- lymerase was also inhibited quite markedly by relatively low 1727 TABLE III Characteristics of RNA synthesis by gh-I -induced RNA polymerase The components of the standard reaction mixture were as de- scribed under “Experimental Procedure.” Where indicated, the appropriate component was removed from or added to the stand- ard reaction mixture. Reactions were initiated by the addition of 2.4 pg per ml of gh-l polymerase (Bio-Gel fraction). Components of the reaction mixture CMP incorporated nmol/hr Standard. . ........................... 9.30 Minus enzyme. .................... . . . 0 Minus gh-l DNA. . ................... 0 Minus ATP, GTP, or UTP ............ 0—0.04 Minus MgC12 .......................... 0 Minus MgClz; Plus MnClz, CaClz, or ZnClg (0.5—8 mM) .................... 0 Plus 2 mM MnClz or CaClz. ........... 0.84 Plus 2 mM ZnCla ..................... 0 Plus 85 mM KCl. . .................... 4.45 Plus 200 mM KCl ...................... 0.18 concentrations of monovalent ions. At a concentration of 85 mM KCl, the gh-l polymerase activity was inhibited 50 %, whereas at 200 mM, the reaction was essentially completely inhibited. An almost identical inhibition of enzyme activity was observed with either NaCl or NH4Cl (data not shown). Apparent Km values for each of the four ribonucleoside triphos- phates which are substrates for RNA synthesis were determined. For these studies, the concentration of three of the ribonucleoside triphosphates was fixed at a high level, greater than 5 times the Km value for any ribonucleoside triphosphate. The concentra- tion of the fourth ribonucleoside triphosphate was varied and the initial reaction rates were measured at each concentration. To analyze the results, Michaelis-Menten kinetics was assumed ap— plicable to this complex reaction, and the results were plotted in Lineweaver-Burk double reciprocal plots (1 /v versus 1/[S]). All data were analyzed by a computer program to determine the highest correlation to a least squares straight line for the equa- tion: Km"(v/ [NTPl") as n was varied in increments of 0.05 unit (37). An n value so determined is equivalent to the Hill coefficient, n, and should equal 1.0 if the double reciprocal plot is linear. For the purine ribonucleoside triphosphate ATP, the double reciprocal plot was linear (Fig. 4A). The apparent Km value for ATP was 3.5 x 10‘5 M. Likewise, the pyrimidine ribonucleo- side triphosphates, CTP and UTP, yielded linear double recip— rocal plots (data not shown). The apparent Km value for both of these substrates in the RNA polymerase reaction was 4.0 X 10‘5 M. For the purine ribonucleoside triphosphate GTP, how- ever, the double reciprocal plot was curvilinear (Fig. 4A). An n value of 1.2 for GTP was determined by the computer analysis. Thus, the best fit to a straight line was obtained when l/v was plotted versus 1 /[GTP]1'2 (Fig. 4B). The kinetics of RNA syn- thesis at the lowest GTP concentration used in the Km study was linear for at least 5 min and showed no appreciable lag in initia- tion (data not shown). Thus, the higher order n value is not due to nonlinear reaction rates at the lower substrate concen- trations. The apparent Km value for GTP, using the higher order value of substrate concentration in the Michaelis—Menten equation, was 8.0 X 10‘5 M or twice that seen for the other three ribonucleoside triphosphates. v = Vmax — .0 40 so I/[NTP] (mM") l/V mus“) IOO I50 I/[GTP]"z (mM") FIG. 4 (left). The effect of varying the concentration of a single nucleoside triphosphate on the activity of gh—l-induced RNA polymerase. Reaction mixtures for gh-l polymerase were pre- pared as described under “Experimental Procedure,” except that the concentration of one nucleoside triphosphate was varied whereas the concentrations of the other three nucleoside triphos- phates were kept constant at 0.4 mM. The reaction mixtures Were prewarmed to 30° and RNA synthesis initiated by the addition of gh-l polymerase (Bio-Gel fraction) to a final concentration of 2.4 pg per ml. After 5 min of incubation, the reactions were ter- minated and the incorporation of [3H]CTP into acid-insoluble ma- terial was determined as described under “Experimental Pro- cedure.” A, double reciprocal plot of 1/1) versus l/concentration of nucleoside triphosphate for ATP (0) and GTP (D). B, the The initiation process of RNA synthesis by gh-l-induced RNA polymerase with gh—l DNA as template was measured using 'y-“P—labeled purine ribonucleoside triphosphates. As shown in Fig. 5, gh-l polymerase incorporated [7-32P1GTP into acid-in- soluble material and this incorporation continued during the entire period of incubation. On the other hand, ['y-“PlATP was not incorporated significantly under the same conditions. The incorporation of [’y-“P]GTP into acid—insoluble material did not occur in the absence of either enzyme or gh-l DNA. The 321’- labeled product, isolated from the reaction mixture after 20 min of incorporation, was rendered completely acid-soluble by treat- ment with either pancreatic RNase (1 pg per m1 at 37° for 1/2 hour) or alkali (1 N NaOH at 37° for 6 hours). Thus, the [7-32P]GTP is being incorporated into RNA by the gh-l polym- erase in a DNA—dependent process. For the purine ribonucleo— side triphosphates, gh-l polymerase initiates RNA synthesis on gh-l DNA exclusively with GTP. Several antibiotics and antibiotic derivatives, which are inhibi- tors of host RNA polymerase activity, were added to the stand— ard reaction mixture for gh—l polymerase to test their effect on in vitro RNA synthesis catalyzed by the phage enzyme (Table IV). The antibiotics, actinomycin D and nogalamycin, inhibit RNA synthesis by intercalating into the DNA structure at G-C— rich and A—T-rich regions, respectively (32, 33). These two antibiotics are effective inhibitors of the host and phage polym- erases, as expected, because both catalyze DNA—dependent proc~ "a. 'o [y-3ZP]—NTP INCORPORATION (pmoI) O "u. IO 20 30 40 TIME (min.) data of A for GTP replotted as l/v versus l/concentration of GTP raised to the 1.2 power. FIG. 5 (right). The kinetics of incorporation of [y-“PlGTP and [y-“PlATP into RNA by the gh—l—induced RNA polymerase. Re- action mixtures for gh-l polymerase were prepared 88 described under “Experimental Procedure,” except that the final concentra- tions of ATP and GTP were lowered to 0.2 mM. Either the ATP (III) or GTP (O) was labeled with y-“P to a final specific activity of 2100 to 2500 cpm per pmol. In all of the assays in which [7‘31” ATP was the labeled substrate, 0.1 mM ADP was included in the reaction mixture to inhibit any trace amounts of polyphosphflte kinase which might be present (38). Reactions were initiated by the addition of 0.3 pg of gh-l polymerase (Bio-Gel fraction) and were incubated at 30° for the times indicated. The reactions were terminated and processed as described by Maitra et al. (39% TABLE IV Effect of various RNA synthesis inhibitors on Pseudomonas putidfl and gh-I -induced RNA polymerase activities The components of the standard reaction mixture were as de- scribed under “Experimental Procedure.” All inhibitors excepi Streptolydigin were added from stock solutions in 10% M") d" methyl sulfoxide. The final concentrations of dimethylsulfoxide in the reaction mixtures (0.4 to 0.8%) did not alter the over-til1 incorporation of [3H]CTP by either enzyme. Reactions W?” Initiated by the addition of 2.4 pg per ml of gh-l p01ymerase (310‘ Gel fraction) or 12.8 pg per ml of P. putida RNA polymerase. In the standard reaction mixture with no additions, 1.85 anl 0f CMP were incorporated by the gh-l polymerase in 10 min and 0'95 ”“01 by “1% Relative 85‘5“” Addition to the standard reaction mixture Concentration gh-l P. WM" Polymerase Wham“ #g/ml % None ................. 100 100 Actinomycin D. . . . .. . . 4 10 7 Nogalamycin ......... 4 5 13 Streptolydigin ........ 100 96 5 Rifampicin. . . . . . . . . . .. 10 95 0'5 Rifamycin AF/013 or ....... 35 50 0'5 AF/DNFI . . . __\J 7: ,3: —. -» __—-._.-_—r—._‘ __ _. *_..4 1,1: esses. The activity of the gh-l polymerase is highly resistant to the antibiotics, rifampicin and streptolydigin, present at con- centrations which markedly inhibit the host RNA polymerase. The effect of 13 other derivatives of rifamycin on gh-l polym— erase activity was tested. The derivatives examined, using the nomenclature of Gruppo Lepetit, were: AF /AOP, Rifamycin AG, AF/APR, AF/DEI, AF/DA-AMP, Rifamide, 4-Dessosi SV, PR/ 14, AF/013, AF/ABDP-cis, AF/AP, AF/BO, AF/DNFI, and PR/ 19 (for review of structures see Ref. 40). All 13 of these derivatives were effective inhibitors (>95 %) of RNA synthesis by the host RNA polymerase when present at a concentration of 10 pg per ml. When added to the gh-l polymerase reaction mixture at 100 pg per ml, seven of the derivatives (AF/013, AF/DNFI, AF/BO, AF/AOP, AF/ABDP, PR/l9, and AF/ DEI) were found to inhibit polymerase activity to a significant degree (>20 %) (data not shown). The most effective inhibitors were AF/013 and AF /DN FI, which inhibited RNA synthesis by 50% at concentrations of 35 pg per ml and almost completely at concentrations of 80 pg per ml or more. The relative order of effectiveness of the rifamycin derivatives in inhibiting gh-l po- lymerase activity was virtually the same as that observed for T7-induced RNA polymerase (41). Inhibitors of the phage polymerase activi ' s, however, were far more effective. Even those rifamycin derivatives which were more effective against the activity of the host RNA polymerase. 3’-Deoxyadenosine 5’-triphosphate, the triphosphate deriva- tive of the antibiotic cordycepin, has been shown to be an in vitro inhibitor of RNA synthesis "by certain bacterial RNA polym- erases (42, 43). This ATP analog presumably inhibits RNA 80— - 70— - : . in S E _ 60» ~ 5 - 3 < ’— 50" -' Z Z _ 8 40— — El ° I (1 so— — 20~ — I I o \ . i \n,_l L e 7 6 5 4 3 2 — LOG [3’— (MT P] _ FIG. 6 (left). The effect of 3’-deoxyadenosine 5’-triphosphate on in vitro RNA synthesis by Pseudomonas putida and gh-l-induced RNA polymerases. Standard reaction mixtures were prepared as described under “Experimental Procedure,” except that 3’-dATP was added to some reactions as indicated. Reactions were ini- tlated by the addition of either 0.3 pg of gh-l polymerase (Bio-Gel fraction) (0) or 1.6 pg of P. putida RNA polymerase (I). After 10 min of incubation, the reactions were terminated and the in- corporation of [3H]CTP into acid-insoluble material determined as described under “Experimental Procedure.” Incorporation in reactions containing various concentrations of 3’-dATP were com- pared to control reactions containing no 3’~dATP. For gh-l polymerase, 100% activity (no 3’-dATP) was equal to 1.95 nmol of CMP incorporated in 10 min and for P. putida RNA polymerase, 0.95 nmol. 1 729 synthesis by being enzymatically incorporated into an RNA chain at a position normally occupied by an AMP residue. If incorporated, the 3’-dAMP would act as a chain terminator in RNA synthesis, because it does not contain a 3’-hydroxyl group necessary for the formation of the next phosphodiester bond. As shown in Fig. 6, 3’-dATP, when added to the standard reaction mixture, inhibited RNA synthesis by both the gh-l-induced and P. putida RNA polymerases. It was a much more effective inhibitor, however, of the gh-l polymerase. The 3’-dATP con- centration required to produce a given level of inhibition with the host RNA polymerase was approximately 80 times greater than that required to inhibit the gh-l polymerase to the same extent. Thus, at the concentration of ATP present in the standard reac- tion mixture, 0.4 mM, 50% inhibition of the host polymerase oc- curred at an ATP :3’-dATP molar ratio of 20, whereas the same degree of inhibition of the phage enzyme occurred at an ATP :3’- dATP molar ratio of 1600. By selecting the appropriate concen- tration of 3’-dATP, the gh-l polymerase activity can be essen- tially completely inhibited, whereas the host polymerase activity is almost completely unafiected. Neither the nucleoside, 3’- deoxyadenosine (cordycepin), nor the diphosphate derivative, 3’—deoxyadenosine 5’-diphosphate, had any effect on either en- zyme activity at concentrations up to 1 mM (data not shown). Double reciprocal plots of U2) versus 1/ [ATP] in the absence and presence of 3’-dATP were experimentally determined to study further this interesting inhibitory effect (Fig. 7). Within experimental error, 3’-dATP acted as a competitive inhibitor of ATP for both enzymes. The apparent Km values for ATP for both enzymes were similar: 6 X 10‘5 M for the host enzyme 30- 20— IO- 20- —— - l 5— -_ ' d I A/ IO" —- / / “ I ‘/ 05— «a / r/ .. A/ I m_ 1 1 4— '20 O 20 40 60 80 I/[ATPl (mt/1") FIG. 7 (right). The effect of varying the concentration of ATP in the absence and presence of 3’-deoxyadenosine 5’-triphosphate on in vitro RNA synthesis by gh-l-induced and Pseudomonas putida RNA polymerases. Reaction mixtures were prepared as described in the legend to Fig. 4, except that some reaction mix~ tures included 3’-dATP at the concentrations listed below. Reac- tions were initiated by the addition of either 1.6 pg of P. putida RNA polymerase (A) or 0.3 pg of gh-l polymerase (B). After 10 min of incubation, the reactions were terminated and the incor- poration of [3H]CTP into acid-insoluble material was determined as described under “Experimental Procedure.” Final concen- trations of 3’-dATP in the reaction mixtures were: 0, (O); 0.12 pM, (A); 0.4 pM, (I); 8 uM, (A); 01‘ 40 pM, (Cl). 1730 and 3.5 X 10"5 M for the gh—l polymerase. The apparent K .~ values for 3’—dAT P were, however, quite different: 2 X 10’6 M for the host enzyme and 2 X 10“ M for the phage enzyme. Thus, the difference in sensitivity of the two enzymes toward 3’—dAT P, as seen in Fig. 7, was reflected in the relative difference of the apparent K .- values. These results indicate that 3’—dATP inhibited the polymerase by competing for a common binding site with ATP. This conclusion was substantiated by the find— ing that the poly(dC) -poly(dG)-primed polymerization of GTP by the gh-l polymerase (see below) was not affected by the presence of 3’-dATl> at levels which completely inhibit the gh-l DNA-primed reaction (data not shown). Another structural analog of ATP is 3’~0-methyladenosine 5’-triphosphate. 3’—AmTP is similar to 3’—dATP in that it differs from ATP only at the 3’ position of the ribose moiety. 3’-AmTP was an inhibitor of RNA synthesis by both the P. putida and the gh-l—induced RNA polymerases (data not shown). The large differential inhibitory effect seen for these two RNA polymerases with 3’-dATP was not observed for 3’—AmTP. The apparent K .- value for 3’-AmTP calculated for the P. putida RNA polymerase was 4.1 X 10‘5 M, or approximately 20 times higher than the apparent K,- value for 3’-dATP. For the gh—l polym— erase, the apparent K ,- value of 3’-AmTP was 1.3 X 10“ M 3 over 3 orders of magnitude greater than that of 3’—dATP. Thus, the 3’-0—methylated analog of ATP is not as efficient as an inhibitor of in vitro RNA synthesis as the 3’—H analog for these RNA po- lymerases. Template Specificity of yh—I -induced RNA Polymerase—One of the most striking characteristics of the gh-l—induced RNA po— lymerase—catalyzed reaction is the stringent template specificity. When DNA from many sources was tested, only the homologous phage gh-l DNA was found to be an efficient template for in vitro RNA synthesis (Table V). The gh—l polymerase would not utilize DNA from coliphages T3, T4, or T7, nor would it utilize calf thymus or P. putida DNA. When the gh—l DNA TABLE V Template specificity of gk-I -induced and Pseudomonas putida RNA polymerase toward DNA from various sources RNA polymerase reactions were prepared and run as described under “Experimental Procedure,” except that the gh—l DNA was replaced by DNA from various sources as indicated. The final concentration of DNA in all cases was 50 pg per ml. The assays contained 4.3 pg per ml of gh-l polymerase (Bio~Gel fraction) or 12.8 pg per ml of P. putida RNA polymerase. With gh-l DNA as template, the gh-l polymerase incorporated 3.3 nmol of CMP in 10 min, and the P. putida RNA polymerase incorporated 1.4 nmol. Where indicated, DNA solutions were denatured immediately before use by heating for 10 min at 100°, folloWed by rapid chilling at 0°. Relative activity DNA template \ gh-f polymerase P. putida polymerase % gh-l ........................ 100 100 T3 ......................... 0.8 99 T7 ......................... <0.5 137 T4 ......................... <0.5 28 Calf thymus ................ 0.6 61 P. putida ................... <0.5 29 Denatured gh-l ............. 1,8 33 Denatured T3 .............. 1.2 22 Denatured calf thymus ..... <0.5 25 \___ was denatured, it became an inefficient template for the gh-l polymerase. Thus, some feature inherent in the doublestranded structure of the gh-l phage DNA is necessary for its function as an efficient template. Likewise, denatured DNA from either coliphage T3 or calf thymus supported little or no RNA synthe- sis. By contrast, the host RNA polymerase can utilize all of the above templates, although at varying efficiencies. Several synthetic polydeoxyribonucleotides were tested as templates for RNA synthesis by the gh-l polymerase (Table VI). The alternating copolymer, poly[d(A—T)], which was an efficient template for the host RNA polymerase, was not utilized effec- tively by the gh—l polymerase. The gh-l polymerase utilized the homopolymer duplex, poly(dC) -poly(dG), to direct the polymerization of GTP at a rate 7 times higher than the polynt erization of CTP from this template. Several single-stranded polydeoxyribonucleotide homopolymers also were tested as tem- plates for gh-l polymerase. Either of the pyrimidine—containing polymers, poly(dT) or poly(dC), would support the synthesis of the corresponding ribohomopolymers. Little or no template activity, however, could be detected with the purine-containing homopolymers, poly(dA) or poly(dl). Thus, with either single- stranded or double—stranded deoxyribonucleotide homopolymers, the gh—l polymerase markedly prefers to utilize the pyrimidine- containing templates as compared to the purine-containing ones. It should be noted that the highest enzyme activity on any tem- plate other than native gh-l DNA, namely that for poly(dT), was less than 5% of the enzyme activity on native gh-l DNA, in terms of total nanomoles of nucleotide incorporated per hour per mg of protein. DISCUSSION The infection of P. putida by the bacteriophage gh-l induces the synthesis of a novel DNA-dependent RNA polymerflS€~ This gh-l-induced RNA polymerase has been purified to near homogeneity. It is composed of a single polypeptide chain with a molecular weight of approximately 98,000. The structure of TABLE VI Template specificity of gh-I -induced and Pseudomonas Putida RNA polymerase toward synthetic polydeozyribonucleotides RNA polymerase reactions were prepared and run as C198“ ibed under ”Experimental Procedure,” except that the template and nucleoside triphosphates were changed as indicated» Each nucleoside triphosphate was present at a final concentration 0f 0.4 mM. The following concentrations of template were em- ployed: poly[d(A-T)], 3 Am units per ml; Poly(dC)'P01Y(dG>’ 2.5 Am units per ml; poly(dA), poly(dC), poly(dI), and P°1y(dT)’ 50 pM (expressed in terms of nucelotide phosphate)- The assays contained either 4.3 pg per ml of gh-l polymerase (BiO‘Gel (”0‘ tlon) or 12.8 pg per ml of P. putida RNA polymerase. _ _ gh-1 iP pit/iii Template Nucleoslde triphosphate substrates P01)“ p0 y - merase 1113”“ \ nmol PHINMP interpolated/hour gal ............... Paloma ATP, GTP UTP 19.8 8-72 Poly[d(A-T)] ....... [3H1ATP, UTP] 0.22 21.3 Poly-poly