BACTERIOPHAGE P35 AND nsmmnousmp TO THE PLASMID DNA OF SALMONELLA PULLORUM Thesis for the Degree of Ph. D.. ‘ MICHIGAN STATE UNIVERSITY WELLIAM L OLSEN 19?: [Hkrh.r This is to certify that the thesis entitled BACTERIOPHAGE P35 AND ITS RELATIONSHIP TO THE PLASMID DNA OF SALMONELLA PULLORUM presented by William L. Olsen has been accepted towards fulfillment of the requirements for Ph .D . degree in my Major professor Date September 14, 1971 0-7639 1am}? 1,-LIBMRYB mMp Stan ' University .j\; ABSTRACT BACTERIOPHAGE P35 AND ITS RELATIONSHIP TO THE PLASMID DNA OF SALMONELLA PULLORUM BY William L. Olsen Salmonella pullorum is closely related to the fre- quently studied species Salmonella typhimurium but differs significantly in ease in which genetic analysis can be performed. Recent F-prime mediated recombination experi- ments indicated the possibility of the presence of both prophage and plasmid DNA in §. pullorum. Experiments were, therefore, performed to analyze the prophage and extrachromosomal DNA in g. pullorum and their inter— relationships. g. pullorum was demonstrated to harbor several prophage and other inducible elements. One of these, the temperate phage P35, was characterized and shown to closely resemble phage P22 in its morphology (a hexagonal capsid 601unin diameter with a short tail and spikes), the size of its DNA molecule (28 x lO6 daltons), its immunity properties and growth characteristics. Phage P35 spon- taneously produced virulent mutants, one of which was William L. Olsen isolated and labeled P35c. RNA-DNA hybridization eXperi- ments indicated that phage P35 has some nucleotide sequences similar to P22. It was demonstrated that g. pullorum contains two distinct plasmid molecules. These were isolated by lysis of lysozyme-EDTA treated cells with Brij 58 and deoxycho- late and centrifugation of the lysate in a cesium chloride— ethidium bromide dye buoyant density gradient. The plasmid fractions isolated from this gradient were subse- quently analyzed by neutral and alkaline sucrose gradient centrifugation. The results indicated the presence in the gradient of two species of covalently-closed, circular DNA molecules. Cosedimentation in neutral sucrose with colicin El DNA showed that the smaller molecule, labeled plasmid PO-l, had a molecular weight of 1.5 x 106 daltons and that the larger molecule, labeled plasmid PO—2, had a molecular weight of 45 x 106 daltons. It was calculated that there are 150 COpieS per host chromosome of plasmid PO—l while there is only 1 or 2 copies of the PO—2 plasmid per chromosome. Electron micrographs of the plasmid DNA showed circular molecules with no unusual structures. When g. pullorum M853 was made lysogenic for phage P35 and the plasmid molecules isolated, it was found that plasmid PO—2 had been excluded. The plasmid was not regained upon curing the lysogen of the prophage. To determine if a relationship existed between the phage and William L. Olsen plasmid, 3H-RNA was synthesized in vitro with RNA poly- merase isolated from Pseudomonas putida using P35c DNA as a template. The RNA was then used in RNA-DNA hybridi- zation experiments with isolated plasmid DNA. The results of these experiments indicated the presence of similar nucleotide sequences in the DNAs of phage P35 and plasmid PO-2. It was postulated then that the PO—2 plasmid con- tains phage nucleotide sequences and is excluded from lysogenic cells either by competition with the super- infecting phage for a single membrane attachment site or when the phage complements a defective plasmid function allowing it to integrate into the host chromosome. During the course of this study it was also deter- mined that wild type §. pullorum does not modify nor restrict the DNA of phage prOpagated on S. typhimurium. BACTERIOPHAGE P35 AND ITS RELATIONSHIP TO THE PLASMID DNA OF SALMONELLA PULLORUM BY William L: Olsen A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1971 ACKNOWLEDGMENTS I wish to express my gratitude to Dr. Delbert E. Schoenhard for his guidance throughout the course of this work and for his continuing interest in my scientific education. I would especially like to thank Dr. Loren Snyder for many helpful discussions and comments particularly in regards to in vitro RNA synthesis and hybridization studies. I would also like to acknowledge Dr. John Boezi who generously supplied the RNA polymerase and Dr. Rene Scherrer for instruction in the use of much of the equip- ment and also for his witty observations on science. I wish to express my appreciation to the American Society for Microbiology for a President's Fellowship which enabled me to study for a period of time in the laboratory of Dr. Donald Helinski, University of Cali- fornia, San Diego. During the course of this study I was supported financially in part by a departmental assistantship and intellectually by Penny. ii TABLE OF CONTENTS INTRODUCTION . . . . . . . . . . . . LITERATURE REVIEW . . . . . . . . . . Part I. Plasmids and Extrachromosomal DNA . . Alkaline sucrose gradient sedimentation Electron microscopy of plasmid DNA . Plasmids in B- hemolytic E. coli . . . The minute plasmid of E. —col1 15. . . II. Bacteriophage P22 . . . . . General description . . . . . . . Phage related to P22. . . . . . Superinfection exclusion determined by prophage P22. . . . . . . . . P22 DNA: Molecular structure and replication . . . . . . . . . III. Phage Related Plasmids. . . . . . . Phage Pl. . . . . . . . . . . Pl-like plasmid . . . . . . . . Phage A: Plasmid mutants . . . . . IV. RNA-DNA Hybridization . . . . . . . V. Modification and Restriction. . . . . Modification and restriction in Salmonella . . . . . . . . . MATERIALS AND METHODS . . . . . . . . . Bacteria and growth conditions . . . Bacteriophage . . . . . . . . . Chemicals . . . . . . . . . . Buffers and dialysis. . . . . . . Radioactive labeling and counting . . Preparation of bacterial lysates. . Sucrose density gradient centrifugation Dye- buoyant density gradient equilibrium centrifugation . . . . . . Extraction and purification of bacterial DNA 0 O C O O O O O O O O 0 iii CmeDQONUTUJ [—1 l—’ u) l7 l7 19 21 22 24 25 27 27 28 28 28 3O 30 31 33 34 Preparation of phage DNA . . In vitro synthesis of 3H labeled RNA Determination of the concentration and specific activity of the 3H- -RNA . . . RNA-DNA hybridization . . . . . . . Electron microscopy. . . . . . . . RESULTS . . . . . . . . . . . . Part I. Isolation and Characterization of Phage P35 . . . . . . . . Relationship of P35 to P22 . . . . . Isolation of P350 . . . . . . . Isolation of S. pullorum MSS3(P35). . . Morphology of P35 . . . . . . . The DNA of P35 . . . . . . . II. Isolation and Characterization of Plasmid DNA from S. pullorum . . . . . . . Isolation —of plasmid DNA . . . . . Sedimentation coefficients and molecular weights of the plasmid DNA. . . . . Alkaline sucrose gradient sedimentation of plasmid DNA. . . . . . . . Sedimentation of a cleared lysate on a neutral sucrose gradient . . . . Electron microscopy of plasmid DNA. . III. Exclusion of Plasmid DNA in S. pullorum. by Lysogeny with P35 . . . Plasmid DNA in S. pullorum MS53(P35) . . Isolation of a cured derivative of S. pullorum MSS3(P35) . . Attempt to follow the fate of plasmid PO- 2 following infection with P35 IV. Hybridization Studies with S. pullorum Plasmid DNA . . . . . . . . . . Assay of Pseudomonas putida RNA polymeras using P35c DNA as a template. In vitro synthesis of 3H- -RNA. . . . Hybridization efficiency . . . . . Hybridization of 3H- -RNA (P35c) to phage DNAs . . Hybridization of S. pullorum M853 plasmid DNA to 3H- -RNA (P35c). . . . . . . V. Modification and Restriction in S. pullorum iv Page 35 36 38 39 39 41 41 41 47 49 49 54 57 57 63 66 66 72 73 73 79 80 88 88 89 92 93 96 101 Page DISCUSSION . . . . . . . . . . . . . . 104 Part I. Isolation and Characterization of Phage P35 . 104 II. Isolation and Characterization of Plasmid DNA in S. pullorum. . . . 106 III. Exclusion —of Plasmid PO- 2 Upon Lysogeny of M853 by P35 . . . . 108 IV. Hybridization of S. pullorum Plasmid DNA to 3H— RNA (P35c) . . . . 111 V. Relationship of Plasmid PO- 2 to Phage P35. . 114 VI. Restriction and Modification . . . . . . 116 SUMMARY . . . . . . . . . . . . . . . 118 LITERATURE CITED . . . . . . . . . . . . 120 LIST OF TABLES Chemicals and sources . . . Hybridization efficiency of to other phage DNAs . . . Hybridization with S. pullorum 53 plasmid DNA 0 O O O O O C O Efficiency of plating of P35 and P22 on various Salmonella strains. vi 3H-RNA (P35c) Page 29 96 100 102 Figure 1. 10. 11. 12. l3. 14. 15. LIST OF FIGURES Adsorption of P35 and P22 to S. pullorum M835 and to S. typhimurium LT2. . . . . Single step growth curve of P35 . . . . . Adsorption of P35c to S. pullorum M853 and 53(P35) o o o u o o o o o o o 0 Electron micrographs of P35 phage particles . Electron micrographs of individual phage particles. . . . . . . . . . . . Sucrose gradient of P35 DNA . . . . . . Isolation of plasmid DNA from S. pullorum M853 . . . . . . . . . . . . Neutral sucrose gradient of plasmid DNA from M853 and colEl DNA . . . . . . . Alkaline and neutral sucrose gradients of S. pullorum M853 plasmids . . . . . . Neutral sucrose gradient of a cleared lysate. Electron micrographs of plasmid DNA of S. pullorum M853 . . . . . . . . . Plasmid DNA of 53(P35) . . . . . . . . Neutral sucrose gradients of a cleared lysate of 8192 . . . . . . . . . . DNA synthesis in P35 infected S. pullorum M853 0 O O O O C O O O O O O 0 DNA synthesis in prelabeled M853 infected with P35 . . . . . . . . . . . . vii Page 44 46 51 53 56 59 62 65 68 71 75 78 82 84 87 Figure Page 16. Assay of RNA polymerase . . . . . . . . 91 17. Hybridization efficiency of 3H-RNA (P35c) to P35c DNA . . . . . . . . . . . 95 18. Isolation of plasmid DNA for hybridization. . 99 viii INTRODUCTION Salmonella pullorum is a non-motile, slow-growing bacterium closely related to the extensively studied species Salmonella typhimurium. It has been demonstrated that all wild type strains of S. pullorum possess two genetic blocks inhibiting the use of inorganic sulfate (51,52) and that upon induced reversion to sulfate proto- trOphy, the now functional sulfite reductase enzyme is temperature sensitive (42). There has been established in S. pullorum a conjugation system which uses F' factors originating in S. typhimurium to mobilize chromosomal genes (35). During mating experiments using S. pullorum donor cells and S. typhimurium recipients, a phage was isolated which had arisen by zygotic induction. This phage was labeled P35 since it was believed to have originally existed as a prOphage in S. pullorum strain M835. Since it had been assumed that S. pullorum M835 was non-lysogenic, it was decided that the origin and characteristics of P35 should be investigated and its relationship to Salmonella phage P22 determined. Also, to understand fully the genetic structure of S. pullorum it would be advantageous to know if the cell contained any genetic information existing in an extra- chromosomal state. It was, therefore, decided to under- take a study of the physical nature of the S. pullorum DNA complement, examining the cell for the presence of plasmid DNA and, if present, its relationship to the suspected prOphage P35. LITERATURE REVIEW Part I Plasmids and Extrachromosomal DNA Since the initial genetic determinations of the existence in bacterial cells of extrachromosomal genetic elements, DNA corresponding to these plasmids has been isolated from many organisms by several techniques. Plas- mids are now found so frequently that it may be only the exceptional cell which does not contain one. Since plas- mids were assumed to consist of DNA similar to the chromo- somal DNA as determined by buoyant density in CsCl, it was often necessary to transfer the plasmid first to a differ- ent host where it could be separated from the larger amount of chromosomal DNA by its density difference. This type of procedure, however, does not provide information about the structure and replication of the plasmid in its natural host. A direct procedure for the demonstration and iso- lation of plasmid DNA was developed by Radloff, Bauer and Vinograd (66) using the intercalating dye ethidium bromide in a preparative CsCl density gradient. Each ethidium bromide molecule which intercalates between the base pairs of a duplex DNA molecule causes a 12 degree unwinding of the helical structure (5). As long as the ends of the duplex are free to rotate, the dye molecules can continue to bind until there is a maximum of one dye molecule bound for every four or five base pairs (88). However, if, as in the closed circular, plasmid DNA molecules, the ends are not free to rotate, then the unwinding due to the bind— ing of the dye will be limited by the extent of super— helical structure of the molecule. The buoyant density of the DNA—dye complex is decreased as the amount of bound ethidium bromide is increased. Since the closed circular molecule cannot bind as much dye per unit length as can the open circular or linear molecules, at saturating con- centrations there will be a density difference between the two forms. The closed circular molecule will be more dense than the Open circular and linear forms. If these DNA-dye complexes are centrifuged to equilibrium in a CsCl density gradient, the closed circular plasmid mole- cules will band lower in the tube at a density greater than the Open circular and linear forms. Radloff 2E 21° (66) had deveIOped the dye-buoyant technique to isolate mitochondrial DNA from Hela cells but it was adapted for the isolation of bacterial plasmid DNA by Bazaral and Helinski (9). Using this procedure, they were able to isolate directly from crude lysates of S. ggli_plasmid DNAs which carried the genetic determi— nants for colicins El, E2, and E3. These were subsequently analyzed by sucrose gradient centrifugation and electron microscopy. In contrast to the results obtained when R factors (31,71) and colEl DNA (36,70) were isolated from Proteus mirabilis, colEl DNA isolated from its native host, S. 921i, by the dye-buoyant density gradient pro- cedure does not appear to be present in aberrant forms (dimers, trimers, etc.). Also, there is only a small number of copies per host chromosome rather than the multiple copies found in E. mirabilis. ColEl supercoiled DNA has a molecular weight of 4.6 x 106 and sediments through a neutral sucrose gradient with a sedimentation coefficient of 235. The open circu- lar form sediments at 178 (8). Alkaline sucrose gradient sedimentation. When covalently closed circular plasmid DNA is centrifuged at pH values higher than 12, the molecules sediment through the gradient at a rate much faster than at neutral pH values. Vinograd ES 3;. (86) proposed that this was due to inability of the strands of the denatured covalently- closed molecule to separate, leading, therefore, to a more compact structure of the same molecular weight. This phenomenon was utilized by Freifelder (32) in devising a method to identify the presence of plasmid DNA molecules in hosts whose chromosomal DNA has the same buoyant density as the plasmid. A lysate of S. ggll_containing an F'lac plasmid was vortexed to shear the large chromosomal DNA and then layered directly on an alkaline sucrose gradient. After centrifugation the plasmid DNA was seen as a distinct peak much further down the tube than the bulk of the denatured chromosomal (single stranded) DNA. It was also demonstrated that the fast sedimenting form was converted to the slower sedimenting open circular form by a dose of X-irradiation sufficient to produce only one single strand break per molecule. This method could, therefore, be used both to verify that a DNA molecule is indeed a covalently closed circular structure and to isolate plasmid DNA. Electron microsc0py of plasmid DNA. The use of electron microscopy to visualize the plasmid DNA molecules has proven valuable in determining the tOpographical structure of these molecules. The most popular technique for preparing DNA molecules for electron microscopy is the protein monolayer technique of Kleinschmidt (53). The DNA is spread and adsorbed to a surface layer of denatured cytochrome c which can then be picked up on an electron microscope grid. The grids are then shadowed with heavy metals to give greater width to the DNA molecules. This technique has recently been modified by the development of a microversion (54) in which the protein film is made on the surface of a drOplet containing only 40u1 of solution. With this procedure as little as 4 ng of DNA are required. Using the Kleinschmidt procedure, Hudson and Vino- grad (43) have found unusual DNA structures present in the plasmid band of a CsCl—EtBr gradient. The DNA in the gradient was isolated from Hela cell mitochondria. These molecules are multiples of the monomer form linked to— gether as in a chain. These catenated molecules probably arise by recombination-like events between mature molecules and differ from the linear concatenated molecules which contain multiple genomes linked end to end. The latter have been prOposed to arise during replication of the DNA of phages P22 and T7 (85). Plasmids in B-hemolytic E. coli. Multiple species of plasmid DNA molecules have been found in beta-hemolytic strains of S. 92;; by Goebel and Schrempf (37). Three molecules were found each differing in molecular weight as shown by sucrose gradient centrifugation of the plasmid band from a CsCl-EtBr gradient and by electron micrOSCOpy. The larger transmissible molecule which has a molecular weight of 58 x 106 daltons and a sedimentation coefficient equal to 725 was tentatively associated with the pro- duction of the beta hemolysin. A molecule sedimenting at 238 (4.2 x 106 daltons) was assumed to be related to the colicinogenic character of the strain due to its non- transferability and size similarity to colEl DNA. The third molecule was present in two forms, one sedimenting at 665 and the other at 453. These were demonstrated to 6 be the closed and Open circular forms of a 46 x 10 Mw plasmid. It sedimented at 1595 in alkaline sucrose gradient. No definite biological function could be described for this molecule though it may be part of the hemolytic factor which has dissociated from the transfer determinants. The minute plasmid of E. coli 15. Recently Cozzarelli EE ES. (26) have isolated minute plasmid molecules from S. coli strain 15. They have a molecular 6 daltons and were assumed to be the weight of 1.5 x 10 smallest known, autonomous, replicating DNA Species yet observed. These molecules are extremely interesting since they could carry sufficient information to code for a polypeptide of only about 75,000 molecular weight or only a few cistrons. There were found to be approximately 15 copies of the plasmid per chromosome in exponentially growing cells. This is the greatest amount of redundancy found for any plasmid in its native host. No biological function was identified with the presence of the plasmid. Part II Bacteriophage P22 General description. Phage P22 was originally iso— lated in 1952 by Zinder and Lederberg (96) from S. typhimurium strain LT22. They found that it was active on strain LT2 and that it was identifiable with their filterable agent which mediated genetic recombination or transduction. P22 is a temperate phage which when it enters the prophage state, integrates into the chromosome of S. typhimurium between the pro A and pro C loci (76,79). Cough and Levine (40) first demonstrated the circularity of the vegetative genetic map and ordered the known genes in a single circular linkage group. The latest genetic map of P22, published by Calandar (21) in 1970, includes 27 genes located at known distances around the map. The prOphage map is a linear permutation of the circular vege— tative linkage map. Levine in 1957 (55) made an extensive study of the genes involved in lysogeny by P22. He demonstrated that by increasing the multiplicity of infecting phage, the frequency of lysogenization was increased. At a multi— plicity of infection above 10 there was nearly 100% lysogenization. P22 produced the typical turbid plague of a temperate phage but each stock lysate contained at a frequency of 0.1% spontaneous mutants which produced clear plaques. By complementation analysis Levine grouped these into three classes: Cl, c2 and c3. These all mapped close together. Revertants of the virulent mutants were not seen. Another class of mutants comprising a separate, single complementation group involved in lysogeny has been isolated (74,78). Although the c genes function normally, these 12E mutants (originally called L mutants) cannot complete the lysogenic reSponse due to an inability to integrate into the host chromosome. These mutants produce turbid plaques which resemble the 10 wild type but the phage genome is progressively diluted out among the segregating progeny cells. The $25 gene is also necessary for normal detachment from the chromosome during induction. Additional virulent mutants have recently been isolated by their ability to form plaques on P22 lysogens (20). Several different classes have been defined and the mutant locus of some have been mapped to lie within the c2 gene. According to Calandar (21) whether an infection by P22 leads to a lytic or to a lysogenic response is deter— mined by four genes: c1, c2, c3 and mpg. Cl is responsi- ble for repression of the phage DNA synthesis. C2 pro- duces the immunity repressor analogous to the c1 gene product which exerts a negative control on the trans- cription of the remaining genes. C3 and TEE are necessary for a high frequency of lysogenization but their exact function is not known. However, it has been shown that the TEE locus produces a cytoplasmic product which is necessary for the stable maintenance of lysogeny (38). Phage related to P22. Yamamoto and Anderson (93) isolated from stocks of P22 a serologically and morpho- logically unrelated phage which they labeled P221. The P22 phage particle is approximately 60 nm in diameter with a short non-contractile tail containing a hexagonal base plate with 6 spikes (47,93). P221 was found to have a long flexible tail with no base plate. A genetic ll relationship between P22 and P221 was demonstrated by the formation of recombinants. The origin of P221 was deter- mined when it was found that the strains of S. typhimurium on which the P22 stocks containing P221 were made were lysogenic for two prophages: Felsl and Fels2 (91). Felsl is serologically and morphologically related to P221. But while P221 is homoimmune to P22, Felsl is heteroimmune. FelsZ appeared to be unrelated. UV induction experiments of Felsl during infection with P22 lead to the hypothesis that P221 arose as a consequence of an infrequent recombi- national event between P22 and Felsl or its defective genome. Independent preparations of P221 contain varying amounts of P22 genetic material. Most of the P221 DNA is derived from Felsl with P22 supplying the C gene immunity region plus various amounts of other genes. P221 also has a different chromosomal attachment site than P22 (94). Later results have indicated that P22 can infect FelsZ lysogens and upon vegetative growth can recombine with the FelsZ genetic material and produce another hybrid, F22 (92). Since the isolation of P221 it has been shown that infection of S. typhimurium LT2 by P22 or P221 induces the formation of a large number of different phage types which can be found in the lysates (95). These differ in morphology, serology, immunity, buoyant density, host range and even in the length of the latent period and the 12 number of infective centers produced from a single burst. P22 has a latent period of 33 minutes when infecting S. typhimurium in an aerated nutrient broth and a burst size of about 160. Some of the other phage isolated from LT2 have latent periods up to 40 minutes and produce as few as ten phage per infected cell. Phage L, which is genetically related to P22, has also been induced by UV light from a substrain of S. typhimurium LT2 (10). The phage is morphologically and serologically similar to P22 and shows the same high resistance to heat inactivation. There are significant differences in their immunity characteristics. P22 clear mutants form turbid plaques on strains lysogenic for phage L. Both P22 and L phage can be isolated from these turbid plaques. Also, P22 is found to abortively infect strain 1559 (the cells are killed without phage pro- duction) but this does not occur with phage L. The abortive infection was later found to be due to the presence or absence of alleles of gene H which controls the reestablishment of protein synthesis in the infected cell (2). Allele LH1 in P22 cannot efficiently reestab- lish protein synthesis unless phage L carrying an active H gene is present. A phage which may be related to P22 was isolated from chicken feces in 1944 (7). This phage was found to infect S. pullorum leading to cell lysis. Electron micrographs 13 identified the phage a particle of about 40—45nm in diameter. Superinfection exclusion determined by prpphage P22. When a P22 lysogen is induced by exposure to UV light and then superinfected with another strain of P22, the super- infecting DNA is excluded (67). Even though the immunity has been lifted, the superinfecting DNA is still excluded and complementation will not occur. The superinfection exclusion was demonstrated to be due to the gig locus on the P22 genome and shown not to involve degradation of the superinfecting DNA. The exclusion is effective also against the heteroimmune phage L. The mechanism for this type of exclusion is presently unknown although there are two indications that it may be localized at or near the cell surface. First, the exclusion is dependent upon the method by which the DNA molecule enters the cell. The exclusion does not Operate if the phage DNA enters as a prOphage during conjugation. Second, non—excluding P22 mutants were found not to produce the somatic antigen 1 which is normally produced by lysogenic conversion of the host upon P22 lysogenization (87). P22 DNA: molecular structure and replication. The DNA of P22 is a circularly permuted collection of linear duplex molecules of 13.711in length and with a molecular 6 weight of 27 x 10 (68). A double stranded terminal redundancy accounts for about 3-5% of the molecule (about 14 1500 nucleotide pairs). When isolated from the mature phage particles, the DNA sediments through a neutral sucrose gradient of 1.0 M NaCl at 355. The circular permutation of the P22 DNA molecule agrees with the evi- dence that the vegetative genetic map is circular (40). The fact that P22 DNA is circularly permuted whereas the A DNA molecule is unique (nonpermuted) with a very short single strand terminal redundancy may be related to P22 being able to perform generalized transduction while A can only perform specialized transduction. If, in the maturation process, P22 is capable of incorporating any piece of DNA of the correct size irrespective of nucleo- tide sequence, then any sufficiently large piece of chromosomal DNA could also be incorporated into the phage head (68). Smith and Levine (75) have shown that immediately after infection of S. typhimurium LT2 by P22, there was a depression of the rate of DNA synthesis regardless of whether the infection led to a lytic or a lysogenic response. After 3 minutes there was a sharp rise in the rate of DNA synthesis which was correlated with one round of replication of the phage DNA. In the lysogenic in- fection by wild type P22, DNA synthesis then ceases while the phage molecule is integrated into the bacterial chromosome after which DNA synthesis reinitiates. Synthesis procedes at a high rate until it nearly returns 15 to the level in the uninfected cell. Lytic infections with cl or c2 mutants showed altered patterns of incor— poration of 3H-thymidine leading to extensive replication of the phage DNA and eventually production of mature phage. It has recently been shown that even in lytic infections with P22 there is no breakdown of the host chromosome into oligonucleotides or even into fragments larger than the size of phage DNA (72). After infection, in conditions leading to lysogeny, the parental linear molecules are converted to covalently-closed, supercoiled molecules (69,85). The conversion occurs after the initial phage replication demonstrated by Smith and Levine (75,77). Later in the infection up to 14 copies of the supercoiled monomer were found. Also, catenates of two interlocked, supercoiled molecules were found. These were similar to those observed by Hudson and Vinograd (43) in Hela cell mitochondria. It is possible that, due to the circular permutation of P22 DNA, the end of one molecule could recombine with the middle of another to produce linear concatemers of fractional genome sizes. Botstein and Levine (13,14) have also studied the intracellular replication of P22 DNA. Soon after in- fection or UV induction, they found that the newly labeled phage DNA is in a form called intermediate I. This has a high sedimentation value (10005), is probably associated with some cellular structure (membrane) and appears to be 16 the replicating form of the phage DNA. A phage specific function is required for the formation of intermediate I. The immunity repressor may play a physical role by in— hibiting phage DNA from associating with the membrane (56). When cells undergoing phage replication were pulse labeled with 3H-thymidine early in infection and then chased with unlabeled thymidine for varying lengths of time, it was found that the label originally in intermediate I was con- verted to a slower sedimenting form. This form, inter- mediate II, possessed a sedimentation coefficient 1.3 to 1.7 times greater than that of the mature phage DNA, i.e., 43 to 56s. The DNA in intermediate II did not replicate but appeared to be linear and of lengths greater than mature phage DNA. It probably contained multiple phage genomes. Intermediate II is the immediate precursor of the DNA in the mature phage particles. It has been found that even with a temperature sensitive c1 mutant which does not make phage DNA, the parental DNA was incorporated into intermediate I (15). If intermediate II is a linear concatenate, then maturation of the mature phage particle may involve a mechanism related to the "headfull" hypothe- sis as proposed for phages T7 and T4 (83). This would also provide a means for production of the general transducing particles. 17 Part III Phage Related Plasmids Phage Pl. Genetic studies by Boice and Luria (11) on P1, the general transducing phage of S. 9211! had indi- cated that the prOphage was not linked to the host chromo- some during Hfr(Pl+) x F-(P1-) matings. These results stimulated two subsequent investigations into the physical relationships of the prophage DNA molecule to the host chromosome. Inselburg (46) isolated the total cellular DNA from both Plkc lysogens and non-lysogens and sedimented the purified DNAs in neutral sucrose gradients. Pooled fractions from the gradients were then hybridized to P1 DNA immobilized in agar. He found that the P1 specific- binding DNA was distributed throughout the gradient at a constant proportion relative to the amount of chromosomal DNA. Isolated Pl DNA with a molecular weight of 60 x 106 sedimented in a single peak and when it was hybridized to the P1 DNA, the fraction with the maximum specific hybridi— zation corresponded to the peak of P1 DNA. He concluded from this that Pl prophage DNA does not exist in the cell as a separate entity but, in contrast to the genetic data, is integrated. Contrasting results were obtained by Ikeda and Tomizawa (45) in a more extensive analysis utilizing DNA— DNA hybridizations on membrane filters coupled with cesium 18 chloride-ethidium bromide density gradient centrifugation and electron microscopy. They found that there was an average of one COpy of Plkc per chromosome. The total cellular DNA was extracted, fragmented into pieces with an average size distribution of twice the size of P1 DNA molecules and then centrifuged in a CsSO gradient con— 4 taining HgCl Since P1 DNA has a different GC content 2. than that of the S. 39;; chromosome and if it is not physically associated with the chromosome, it will band at a position of slightly greater density than the chroso— somal DNA. If the P1 DNA were associated with chromosomal DNA, it would be distributed in the gradient between the position of pure chromosomal DNA and pure P1 phage DNA. They then hybridized the DNA in each fraction to P1 DNA on nitrocellulose membrane filters and found that Pl prophage.DNA banded at a position distinct from the bulk of the cellular DNA. The sucrose gradient experiment of Inselburg was also repeated by Ikeda and Tomizawa except that they were not forced to pool fractions but could test the hybridization efficiency in each fraction. By this method they were able to detect a separation of prophage Pl DNA from the bacterial DNA. As added proof of the plasmid nature of P1 prophage DNA, Ikeda and Tomizawa isolated a satellite peak of DNA in a CsCl-EtBr gradient and demonstrated by hybridization that it was composed of 79% Pl-specific DNA. Electron microscopy of the satellite peak showed it contained 19 circular DNA molecules of an average length of 31.5 u which is slightly shorter than the 35.5 u observed for the phage DNA molecule. The terminal redundancy of the phage molecule could account for its extra length. It has recently been shown that the bacterial S92 allele is necessary for plasmid formation by P1 (82). The lgg+ gene function is necessary for plasmid formation although P1 can grow normally on lgp_ strains. Mutants containing Sgp- also have pleiotropic effects on UV sensi- tivity, septum formation and regulation of capsular poly— saccharide formation. This effect on plasmid formation by 122 is restricted to phages P1 and A (N_) since both P and R factors can form plasmids in lpp— strains and is assumed to involve the membrane (plasmid membrane binding sites). The strict replication control which allows only one COpy of P1 prophage per chromosome may also relate to the exclusion observed by Luria SE 3;. (59) when Pl 233+ super- infects a lysogen containing Pldl prophage. After Pl 335+ +0 infection the "quasi—stable lac strains segregate lac- segregants as opposed to those strains in which the lac characteristic had become associated with the chromosomes. Lac-segregants no longer contained the Pldl prophage. Pl-like plasmid. An extremely interesting finding was made by Ideda, Inuzuka and Tomizawa (44) when they studied the plasmid DNA in S. coli strain 15. In addition 20 to the small, supercoiled DNA molecule previously described by Cozzarelli 2E 21- (26), they found two additional 6 6 daltons which plasmid molecules of 63 x 10 and 104 x 10 they labeled P15B and P15C respectively. By DNA—DNA hybridization it was found that nearly all of the P158 DNA was homologous to phage Pl DNA. Induction of strain 15 with mitomycin c produces two phage-like particles neither of which carry DNA homologous to the DNA of plasmid P15B. An explanation for the absence of P15B DNA in the phage particles is the defectiveness of the P15B plasmid. Upon further examination it was found that the presence of plasmid P15B confers upon the cell a different restriction and modification pattern than does the presence of P1 prOphage. P15B DNA also has a buoyant density in CsCl of 1.708 gm cm'3 while that of P1 DNA is 1.706 gm cm’3. Although there appear to be differences in the modifi— cation-restriction system, Pl-specific immunity is expressed by P15B. When a cell containing the plasmid P15B was infected with PlCM (a derivative of P1 carrying the chloramphenicol resistance determinant of an R factor; PlCM can form a stable lysogen) and plated on agar plates containing chloramphenicol, only one cell in 107 survived. These resistant cells were found to have simultaneously acquired the Pl modification and restriction system but were found not to have increased the amount of P1 specifi- cally hybridizable DNA which they contained. Chloram- phenicol sensitive segregants were isolated and these 21 were found to be lacking in the Pl—like DNA. Therefore infection of S. 99;; 15 with PlCM can lead to exclusion of the P15B plasmid. Although P15B is about 5% larger than the Pl prophage molecule, the authors assumed that the ex— clusion may be due to a competition between P15B and P1 for a limited number of intracellular sites. Previous data (45) had indicated a strict replication control for the P1 prophage. Phage A: plasmid mutants. It has recently been shown that mutants of the A prophage can exist as autonomously replicating plasmids. Two reports indicate that mutants in the N gene lead to plasmid formation (57,73). The N gene product may be necessary for the association of supercoiled A molecules with the membrane (41) as well as being required for transcription of the early phage genes involved in DNA replication (28). Signer (73) postulated that without the product of the N gene, phage DNA repli- cation could continue but since most other A genes were repressed, no mature phage particles would be produced and the cells would not be killed. The replicated phage DNA would remain in the cell as a plasmid. The A N_ plasmids were found to be present in the cell at 10-20 COpieS per host chromosome. Matsubara and Kaiser (61) isolated from A yig a deletion mutant which had lost most of the phage genome except that part required for phage DNA replication and 22 its regulation. This molecule of 8.6 x 106 daltons was also shown to exist in the host cell as a plasmid with 20 COpies per bacterial chromosomes. The molecular weight of mature A DNA as determined by electron microscopy is 33 x 106 daltons (23,60). Also, when A infects a hybrid cell containing genetic material of S. coli K12 and S. typhosa, it is adsorbed but cellular lysis does not occur (6). It was proposed that this diploid Salmonella hybrid inhibits the function of the N gene product and was later demonstrated that, like A N— mutants in S. coli, A does exist in the hybrid as a plasmid (30). Part IV RNA—DNA Hybridization DNA-RNA and DNA-DNA hybridizations have been used to study the relatedness of many nucleic acids and several methods have been developed to perform and analyze the products of the hybridization reactions. When it is possible, it is advantageous to immobilize one of the components (usually the unlabeled DNA) in agar (12) or on nitrocellulose filters (27,34). This procedure reduces the probability of the denatured unlabeled strands recom- bining and, therefore, no longer being available for hybridization with the labeled test DNA or RNA. But these procedures do require that the immobilized unlabeled 23 DNA be present in much higher concentrations than the labeled mobile component since any reaction of this com- ponent with itself will inhibit its hybridization with the immobile component. Greatly lowering the relative concentration of the mobile component will, therefore, favor its reaction with the immobile component. If high concentrations of the unlabeled DNA are not available, it is necessary to perform the hybridization reaction with both components in solution. Several methods to detect the resulting DNA-RNA complexes are available. The most convenient method was developed by Nygaard and Hall (63) who collected the complexes on nitrocellulose filters. They found that at moderate salt concentrations nitro- cellulose filters will adsorb denatured DNA and DNA-RNA complexes but that free RNA will pass through the filter. By presoaking the filters before collection of the com- plexes and then extensive washing of the filters, the background can be greatly reduced so that less than 0.5% of the free, labeled RNA will be trapped by the filter. Labeled RNA for hybridization can be isolated from bacteria but it possibly may not have a sufficiently high 3H—RNA of high specific activity can specific activity. be synthesized 12 vitro using DNA and RNA polymerase. RNA polymerase was first identified and isolated in the early 1960's from S. coli (33,81) and from Micrococcus lysodeickticus (89,90) and since has been extracted from 24 several bacterial sources. Burgess (16) has published a procedure for the assay and large scale purification of S. 9213 RNA polymerase. Characterization showed that core enzyme was composed of four subunits: dZBB' (17). A similar enzyme has been isolated and characterized from Pseudomonas putida (48). It was found to be present in two forms; one with the subunit structure dZBB' and the other with the subunit structure of dZBB'o. This enzyme can transcribe DNA from the Pseudomonas phage gh-l as well as bacterial DNA (J. Boezi personal communication). Part V Modification and Restriction Arber and Lin (4) have recently reviewed the litera- ture on the modification and restriction systems of bacteria and phage. They indicate that the DNA molecule of a phage is "strain—specifically marked by the host cell during vegetative phage growth. A particular bacterial host strain can be successfully infected only with bacterio— phage carrying the DNA modification produced by that strain, a prOperty obtained through prior growth on the same strain: the inappropriately or unmodified DNA is broken down upon penetration into the host cells, these being restrictive for this DNA." Modification in the S. 3211 K12 and A phage system has been shown to involve enzymatic methylation of DNA at strain—specific sites (39) while restriction is caused by an endonucleolytic scission of the entering DNA molecule presumably at an unmcthylated (modified) strain-Specific site (29). Restriction is never complete and the unmodified infecting phage DNA does have a low but consistent probability of escaping the re- striction and undergoing vegetative phage growth. During this process the phage DNA will be modified to the specifi- city of its new host and will now be able to reinfect this strain at high efficiency. Modification and restriction capabilities can also be carried by temperate phages such as P1 which will when it lysogenizes a cell, confer this prOperty on the new host (4). Plasmids such as R factors (3) can also confer upon their hosts specific modification and restriction abilities. Modification and restriction in Salmonella. Salmonella typhimurium is an inefficient recipient of DNA transferred to it from S. coli. This was demonstrated by Okada 3E 21' (65) to be due to a host specific modifi- cation and restriction system in S. typhimurium strains LT7 and LT2. They succeeded in isolating from strain LT7 a mutant, LT7 £25! which had an increased ability to act as a recipient of foreign DNA. It was also found that when P22 was propagated on LT7 ESE! it showed a greatly reduced efficiency of plating on other S. typhimurium strains. However, P22 grown on the other strains plated 26 at an equal efficiency on LT7 ISE and on the other non- fertile strains. These results were interpreted as indi- cating that LT7 ESE had lost the ability for host-controlled modification and restriction. Okada and Watanabe (64) subsequently isolated from S. typhimurium LT2 by two steps of mutagenesis mutants which were able to act as high efficiency recipients for R factor DNA from S. 9211' Mutant R£g£1 had a one hundred fold increase in its ability to act as a recipient while mutant R£g£2 had approximately a ten thousand fold in- creased ability. That both mutant strains had alterations in their modification and restriction systems was seen by comparison of the efficiencies of plating of P22 grown on the mutant strain to the efficiencies of phage grown on the wild type parent strain. RS332, the strain with the greatest recipient ability, also apparently had undergone an alteration in the cell wall composition since it was now sensitive to S. 99;; phages T3, T7, and P1, was resistant to P22, and produced a rough colonial morphology. RSEEI, while not quite as good a recipient as RSSSZ, remained sensitive to P22. MATERIALS AND METHODS Bacteria and growth conditions. Salmonella pullorum strains M835, M853 and M86-18 were from our laboratory stock collection. Salmonella typhimurium LT2 was origi— nally received from P. E. Hartman and has been maintained in our laboratory. Escherichia coli JC411 (colE1+) was obtained from D. E. Helinski. The colicinogenic character of S. coli JC411 (colEl+) was tested on Escherichia coli 0 which was ob- tained from R. R. Brubaker. Salmonella typhimurium LT2 Rferl and Rfer2 were received from A. Laskin and Salmonella montevideo from E. Sanders. For routine cultivation, L broth containing 10 g of tryptone (Difco), 5 g of yeast extract and 10 g of NaCl per liter were used. For solid media, Difco agar was added to a final concentration of 1.5%. The purity of the S. pullorum cultures were routinely determined by their reaction with antisera (Difco, Salmonella O antiserum group D factor 9) and in SIM agar. For radioactive labeling, the bacteria were grown in TCG broth which contains 0.1 M tris (hydroxymethyl) aminomethane (Tris)-hydrochloride pH 7.4, 0.4% vitamin—free 27 28 casamino acids, 250 ug/ml of deoxyadenosine. Glucose (40%) was autoclaved separately and added just before use to a final concentration of 0.4%. Bacteriophage. P35 is a temperate phage originally isolated from S. pullorum M835 by zygotic induction. P35c was isolated as a spontaneous, clear plaque mutant of P35. P22 and P22c2 were originally obtained from M. Levine. Phage were propagated and titered as described by Adams (1). Stock lysates were prepared by infecting log phase bacterial cells in aerated L broth with phage at a multiplicity of infection (m.o.i.) of 0.1 and incubating for 4-5 hours at 37C. Chloroform was added to the culture and the bacterial debris was removed by centrifugation for 10 min at 8000xg. The phage in the supernant fluid were stored at 4C over a drop of chloroform. The phage were titered by assay of the number of plaque forming units (pfu) per ml by the soft agar overlay method. Chemicals. All general chemicals were reagent grade and purchased from standard commercial sources. Other chemicals are listed in Table 1. Buffers and dialysis. Two general buffers made in deionized, distilled water were used in many of the experi- ments reported here. TES:0.05 M tris (hydroxymethyl) aminomethane (Tris)--hydrochloride, 0.005 M (ethylenedi— nitrilo) tetraacetic acid (EDTA), 0.05 M NaCl, pH 8.0. SSC (standard saline citrate):lx concentrated SSC contains 29 Table 1. Chemicals and sources Chemicals Source Pronase B grade, ribonuclease (RNase) Calbiochem 5x cryst A grade, ethidium bromide (EtBr), mitomycin c, adenosine tri- phosphate (ATP), guanine triphosphate (GTP), uridine triphosphate (UPP) and cytosine triphosphate (CTP) Lysozyme (egg white) Cesium chloride (CsCl) Bovine serum albumin (BSA) Brij 58 Dithiothreitol Antisera 2,5 diphenyloxazole (PPO) 1,4,-bis 2(4-methyl-5- phenyloxazole)—benzene (POPOP) San Diego, Calif. Armour Pharmaceuti- cal Co., Kankakee, Ill. Mann Research Laboratories, New York Pentex Biochemicals, Kankakee, Ill. Emulsion Engineer- ing Co., Elk Grove, Ill. P & L Biochemical, Milwaukee, Wis. Difco Laboratories, Detroit, Mich. Packard Instrument Co., Downers Grove, Ill. 30 0.15 M NaCl, 0.015 M sodium citrate, pH 7.0. The 0.1x and 2x solutions are one tenth and twice concentrated, respectively. All dialysis was performed using sterile dialysis tubing which had been boiled in 0.5 M EDTA for 10 min and then autoclaved in 0.05 M Tris, pH 8.0. Radioactive labelipg and counting. Radioactive chemicals were purchased from The Radiochemical Centre, Amersham. Cells growing in TCG broth containing deoxyadenosine were labeled with 3H-thymidine (15 Ci/m mole) or l4C-thymidine (60 mCi/m mole) at l uCi/ml of broth for 3H and 0.05 u Ci/ml for 14C. The radiochemicals were added to the culture and were allowed to be incor- porated by the cells for at least five generations. This was necessary because S. pullorumincorporates thymidine very inefficiently from the media. All radioactive counting was done in a scintil- lation mixture containing 1.35 g of 2,5—diphenyloxazole (PPO) and 27 mg of l,4,-bis 2(4-methy1—5-phenyloxazole)- benzene (POPOP) per liter of toluene. Ten ml of this mixture was added to vials containing the dried radio- active samples on filter paper or nitrocellulose filters and these were counted either in a Nuclear Chicago Mark I liquid scintillation counter or a Packard Model 2002 Tricarb liquid scintillation counter. Preparation of bacterial lysates. Lysates for the isolation of plasmid DNA were prepared by a modification 31 of the procedure of Clewell and Helinski (25). Radio- actively-labeled cells were harvested in late log phase from a 10 ml TCG broth culture by centrifugation for 10 min at 10,000xg in the Sorval RC2 centrifuge. The cell pellet was resuspended in 0.3 ml of 25% sucrose in 0.05 M Tris pH 8.0, and 0.1 ml of a lysozyme solution was added (5 mg/ml in 0.25 M Tris pH 8.0). This mixture was incu- bated for 5 min in an ice bath, and then 0.1 ml of 0.25 M EDTA pH 8.0 was added and the mixture incubated on ice for another 5 min. The cells were then lysed by addition of 0.5 m1 of a "lytic—mix" containing: 1% Brij 58, 0.4% sodium deoxycholate, 0.0625 M EDTA, and 0.05 M Tris pH 8.0. The cell preparation was incubated in an ice bath for 15-30 min after which lysis was completed by subject- ing the cells to 4-5 cycles of freeze-thawing. The cell suspension was placed in a 5/8 x 2 1/2 inch polyallomer tube which was alternately placed in an ethanol—dry ice bath and a 45C water bath. The chromosomal DNA was removed by centrifuging the lysate at 48,000xg (20,000 RPM) in the RC-2B centrifuge for 25 min. This step normally pelleted 95% of the chromosomal DNA leaving the plasmid DNA in the supernatent fluid. This fluid henceforth will be referred to as the cleared lysate. Sucrose density gradient centrifugation. The DNA sample in 0.1—0.3 m1 of TES buffer was layered on a linear 20-31% sucrose gradient (5.0 m1) made in 0.05 M Tris, 32 0.005 M EDTA, 0.5 M NaCl, pH 8.0. The gradient was centrifuged in a Beckman Model L or Model L3-50 ultra— centrifuge at 15C. Thirty to thirty-two fractions of 0.17 ml (8 drops) were collected from the bottom of the tube using a Beckman fraction recovery system and dropped directly on to 0.75 inch squares of #1 filter paper. The numbered filter papers were dried under a heat lamp and washed successively in 250 m1 of cold 5% TCA, 70% ethanol and anhydrous ether. The filter papers were then dried, placed in a vial containing 10 ml of toluence scintil— lation fluid and counted in a scintillation counter. Alkaline gradients were made as described by Freifelder (32). Linear gradients of 20-31% sucrose in 0.3 M NaOH, 1.0 M NaCl, 0.01 M EDTA, pH 12.0 were made in 1/2 x 2 inch nitrocellulose tubes which had been soaked with 100 ug/ml denatUred calf thymus DNA and 1.0 mg/ml BSA in TBS for 1 hour before use. Centrifugation and fractionation were carried out as above. Recoveries on all gradients were greater than 90% of the added radioactivity. Sedimentation coefficients (8) were calculated from the distances sedimented (D) during cosedimentation with molecules of known 8 values by the equation of Burgie and Hershey (18). (02/01) = (52/51) 33 Molecular weights of the supercoiled DNA were then deter- mined using the relationship of S to the molecular weight shown by Bazaral and Helinski (8) and Clayton and Vinograd (24). The linearity of the gradients was occasionally checked by determining the refractive index of the fractions in a Bausch and Lomb refractometer. Dye—buoyant density gradient equilibrium centrifu- gation. Isolation of plasmid DNA was performed in cesium chloride—ethidium bromide (CsCl-EtBr) dye-buoyant density gradients following the procedure of Bazaral and Helinski (9). Three m1 of a cleared lysate from a 30 ml TCG cul- ture, 2.7 ml of TES buffer, 0.5 ml of EtBr (5 mg/ml TES) and 6 g anhydrous CsCl (final density 1.54 g/ml) were mixed in a vial and poured into a pretreated polyallomer centrifuge tube. Tubes were pretreated by boiling 15 min in TES buffer and then soaking with 100 ug BSA/ml TES buffer for 1 hour. The contents of the tube were over- laid with light mineral oil. The tube was capped and centrifuged in a Type 50 rotor for 44 hours at 44,000 RPM in the Model L or L3-50 ultracentrifuge at 15C. The gradient was fractionated into approximately 60 fractions of 0.1 ml each by collecting 12 drops from a #24 gage needle punctured into the bottom of the tube. The fractions were collected in autoclaved 12 x 75 mm polypropylene tubes, and 5 ul aliquots of each fraction 34 were then spotted on filter paper squares which were washed, dried, and counted as before. The plasmid peak fractions were pooled and dialyzed in the dark against TES buffer to remove the EtBr and CsCl. Extraction and purification of bacterial DNA. Bacteria were grown overnight to 2 x 109 cells/ml in L broth, harvested by centrifugation and washed once with TES buffer. Cells were resuspended in 3 ml of TES buffer and 0.5 ml of lysozyme (5 mg/ml in TBS buffer) was added plus 0.5 ml of 0.25 M EDTA (pH 8.0). The cell suspension was incubated in an ice bath for 20 min after which 0.3 ml Of 10% SDS plus 0.3 m1 of the lytic mix were added and incubated for an additional 30 min on ice. The lysate was then subjected to 3—4 cycles of freeze—thawing after which 1 ml of self-digested pronase (4 mg/ml TES buffer) was added. The pronase was pre-incubated at 60C for 2 hr. The lysate was incubated with the pronase for 6 hr at 37C and then deproteinized by 3 extractions with TES saturated— redistilled phenol. The aqueous layer containing the DNA. was then mixed with an equal volume of ether, the layers allowed to separate and the lower aqueous layer dialyzed against a buffer containing 1 M NaCl, 0.01 M Tris and 0.01 M EDTA pH 8.0, to remove the remaining phenol and then against TES buffer over night at 4C. The DNA was then treated with 3 ug/ml of RNase which had been 35 pre-incubated at 85C for 10 min. After 10 min of incu- bation with RNase at 37C, the DNA was reextracted with phenol 3 times, dialyzed as above and stored frozen at minus 30C. The purity and concentration of the isolated DNA was determined by analysis of its UV absorbance spectrum. Preparation of phage DNA. BacteriOphage were puri- fied and the DNA extracted using modifications of the pro- cedures described by Thomas (50,68,84). Bacteria were grown in 1 liter of aerated L broth to a concentration of approximately 1 x 108 cells/ml and infected with phage. After lysis the remaining cells were removed by centrifu— gation at 8,000xg using the GSA rotor in a Sorval RC—2 centrifuge at 4C. The phage were then pelleted in a Type 30 rotor which was centrifuged at 30.000 RPM and 4C in the Model L or L3-50 ultracentrifuge. The phage were then resuspended over night at 4C in 1 ml of phage buffer (0.5 M NaCl, 0.001 M MgSO 0.01 M Tris, pH 7.4). The 4, resuspended phage were then recentrifuged at 7,000xg in the Sorval RC-2 centrifuge to remove contaminating cell debris. Two ml of the supernatant fluid were gently layered on a preformed CsCl step gradient (0.5 ml steps of 80%, 70%, 60%, 50%, 40% and 30% saturated CsCl in dis- tilled water) and centrifuged at 35,000 RPM and 15C in a SW39 or SW50L rotor in the Model L or L3-50 ultracentrifuge for 1 hr. The phage, which form a dense white band at 36 about the middle of the tube, were collected by drop col- lecting from the bottom of the tube and immediately dialyzed against a buffer containing 1 M NaCl, 0.01 M Tris, 0.01 M EDTA, pH 7.0, and then successively against 5x and 1x SSC. DNA was extracted from the purified phage by gently shaking the phage suspension with an equal volume of TBS-saturated redistilled phenol. The phenol and aqueous layers were separated by a brief centrifu— gation in an International clinical centrifuge and the upper aqueous phase removed, fresh phenol added and the extraction repeated twice more. The extracted DNA was then dialyzed against a buffer containing 1 M NaCl, 0.01 M Tris, 0.01 M EDTA, pH 8.0 and then against the appropriate buffer for the experiment in which it was to be used. Purity and concentration of the isolated DNA were determined by analysis of its UV absorbance spectrum determined on a Beckman DB—G spectrOphotometer. In vitro synthesis of 3H labeled RNA. 3H-RNA was synthesized using P35c DNA as a template and the RNA poly- merase purified from Pseudomonas pptida (48) which was generously supplied by Dr. John Boezi. A modification of the assay procedure of Burgess (16) was used in the assay and synthesis. The specific activity of the RNA polymerase using P350 DNA as a template was assayed in sterile 0.4 m1 re- action mixtures containing 0.04 M Tris, pH 7.9, 0.01 M 3 MgCl 0.05 M KCl, 0.15 mM ATP, GTP, CTP and H-UTP 2' L ., pr 37 (10 uc/u moles), 50 ug P35c DNA/m1 and RNA polymerase at 0, 25, 50 and 100 ug/ml. The RNA polymerase was diluted -4 in buffer (0.01 M Tris, pH 7.9, 0.01 M MgCl 10 M EDTA, 2, 10-4 M dithiothreitol and 5% glycerol) and added last to start the reaction. The reaction-mixture was incubated at 37C for 10 min. The reaction was then stopped by addition of 0.05 ml BSA (1 mg/ml) plus 1 m1 of 5% TCA and it was then placed in an ice bath. The precipitated 3H-RNA was collected on nitrocellulose filters, washed with 15 m1 cold 5% TCA, dried and counted in a scintil- lation counter. The synthesis was performed in a sterile reaction mixture containing the following: 0.04 M Tris-Cl, pH 7.9, 0.01 M MgClz, 0.09 mM 3H-UTP (81.5 uc/u mole), and 50 ug P35c DNA per 0.05 M KCl, 0.15 mM GTP, ATP and CTP plus m1. This was made up in a 2x concentrated solution and placed at 37C. The reaction was started by addition of RNA polymerase to a final concentration of 41 ug protein/ ml plus sterile distilled water to a final volume of 3 ml. The reaction mixture was incubated for 10 min at 37C. It was stopped by addition of 0.3 m1 of 10% SDS. 1 M sodium acetate pH 5.2 was then added to a final concentration of 0.2 M and the reaction mixture placed in an ice bath. The 3H-RNA was extracted three times with water saturated phenol. An equal volume of phenol plus 0.015 ml of 1.0 M MgCl2 (final concentration of 0.005 M) were added to the reaction mixture which was then heated to 65C 38 with occasional shaking. After chilling in an ice bath, the phases were separated by a short centrifugation at 2,000xg at 4C. The upper aqueous layer was removed care- fully avoiding the interface. The aqueous layer was then extracted twice more. The 3H-RNA was then dialyzed against a buffer con- taining l M NaCl, 0.01 M Tris, pH 7.9 for 8 hours and then against 0.01 M Tris pH 7.9 over night and stored frozen. The extraction was followed at each step by removing 0.05 ml samples which were mixed with BSA and precipitated with TCA. The amount of precipitated 3H—RNA was deter- mined as above. Determination of the concentration and specific activity of the 3H-RNA. The amount of 3H—UTP incorporated into RNA was determined by precipitating duplicate 0.05 ml samples of the RNA plus 0.05 ml of carrier BSA (1 mg/ml) with 1 ml of 5% TCA. This was then collected on a glass fiber filter (Whatman glass fiber paper, GF/A), dried, placed in a vial and solubilized by addition of 1 m1 of Soluene-100 (Packard Instruments) and allowed to stand over night at room temperature. After the RNA had solu- bilized, 10 m1 of toluene-scintillation fluid was added and the radioactivity counted in a scintillation counter. Duplicate samples of the 3H-UTP used were also counted in the same way to ensure equal counting efficiency for each sample. The determined specific activity of the 3H—UTP 39 (as cpm/u mole) was then used to calculate the amount of 3H-RNA which was synthesized. RNA-DNA hybridization. The DNA-RNA hybridization procedure of Nygaard and Hall (63) was used. A sample of the test DNA was denatured by boiling for 10 min in 0.1x SSC followed by rapid cooling in an alcohol ice bath. The denatured DNA was then mixed with the appropriate concentration of 3H—RNA in a 1 ml solution of 2x SSC. Annealing was carried out at 60C for 6 hr after which the DNA—RNA complexes were collected on nitrocellulose mem— brane filters (Type B6 25 mm from Schleicher and Schuell, Inc., Keene, N.H.). Filters were washed with 2x SSC, dried, placed in vials, covered with scintillation fluid and counted in a scintillation counter. Electron microscopy. Plasmid DNA was prepared for electron microscopy by the microversion of Lang and Mitani (54). A droplet of 40 ul was prepared in a sterile, dis- posable polycarbonate petri plate. Thirty-five ul of 0.15 M ammonium acetate, pH 6.0, containing 1 x 10"3 to 1 x 10..4 ug of plasmid DNA was first placed on the hydro- phobic surface and to this was added 4 ul of cytochrome c (0.13 mg/ml Of water) and 1 ul of 8% formaldehyde. The drops were allowed to set for 5 to 60 min which allowed the protein and DNA to diffuse to the surface where the cytochrome c denatured and formed a protein monolayer which absorbed the DNA molecules. The protein film was 40 picked up on a 300 mesh carbon coated collodion grid by touching the surface of the droplet with the grid and lifting it off at a 45 degree angle. The grid was then touched to the surface of a 95% ethanol solution for 1 sec and dried face down on filter paper. The grids were shadowed with platinum-palladium (BO-20%) from two directions 90 degrees apart at an angle of 7 degrees. The grids were viewed in a Phillips 300 electron micro- SCOpe and photographs taken on an instrument magnification of 7000-9000x. Phage were prepared for electron microscopy by placing a drOp of a CsCl step-gradient purified phage preparation on a 300 mesh grid coated with collodion, air-dried and stained with 0.5% phosphotungstic acid (PTA). Photographs were taken at an instrument magnifi- cation of 67,000-108,000x. RESULTS Part I Isolation and Characterization of Phage P35 During experiments in conjunction with the develop- ment of a mating system in S. pullorum, Godfrey (35) occasionally observed plaques which arose in the bacterial lawns produced by S. typhimurium recipient strains after they had been mated with S. pullorum M835 donor cells. One of these plaques was picked and found to have resulted from the presence of a phage. This phage was labeled P35 since it was assumed to have arisen by zygotic induction from S. pullorum M835. Phage P35 produced the turbid plaques which are typical of a temperate phage. The exact origin of this phage remains obscure since it was found that P35 was able to be prOpagated and to produce plaques on F- strains of M835 and M853. Phage P35 was shown to be antigenically similar to the S. typhimurium phage P22 and to be able to carry out generalized transduction. Relationship of P35 to P22. The close relationship of P35 to P22 was demonstrated by finding that both require the same cell surface site for adsorption. Phage P22 is 41 42 known to require antigen 12 for adsorption to S. typhimurium (96). P22 and P35 adsorb to S. pullorum and to S. typhimurium (Figure 1) both of which have antigen 12 on the cell surface (19). S. montevideo has a similar surface antigen structure but does not produce antigen 12 (19). When P35 and P22 were added to a culture of S. montevideo, the phage were not adsorbed to the cells indi— cating that antigen 12 may be required for adsorption of both P22 and P35. The latent period and burst size of P35 were deter— mined by a one step growth experiment (Figure 2). It can be seen that when P35 infects S. pullorum growing in aerated L broth, there is a latent period of approxi- mately 40 min and a burst yielding 300 pfu per infected cell. Although P35 can be propagated on strains of S. pullorum, it was thought advantageous to determine if S. pullorum could be induced to liberate phage or bacterio- cins. Log phase cultures of M835 and M853 were treated with mitomucin c (l ug/ml for 10 min) or irradiated with UV light and subsequently tested for the presence of phage or bacteriocin activity by spotting chloroform treated samples on fresh lawns of S. typhimurium LT2. The response to these treatments was highly variable. On several occasions plaques were observed on the indicator plates. Each plaque was picked and reinoculated into Figure l. 43 Adsorption of P35 and P22 to S. pullorum M835 and to S. typhimurium LT2. Log phase (1 x 108 cells/ml) cultures in aerated L broth were infected with phage at an m.o.i. = 0.1. At intervals 1.0 ml samples were removed and treated with chloroform. After removal of the cells by centrifugation the phage remain- ing in the supernant were titered. 44 100 10 2 cl L x 5 . 3 P 1 venompoc: coup—n. \O \ P35204535 R. Q. Q.. 0’ P22 XMS 35 - 0.1 10 15 20 T IME (min) 5 Figure 2. 45 Single step growth curve of P35. A log phase culture (1 x 108 cells/m1) of S. pullorum M835 growing in aerated L broth was infected with P35 at an m.o.i. = 0.1. At intervals samples were removed and the plaque forming units per ml were immediately titered. 46 1000 100 0 __Ou 10.uo.._c_\:K "IIIIII’I' 4O 60 80 TIME(min) 20 47 3 ml of L broth containing a log phase culture of S. typhimurium LT2. The ability to increase the titer of plaque forming units after several hours incubation was taken as a presumptive indication of the presence of phage particles. The presence of no plaque forming units would indicate the original plaque was possibly produced by a bacteriocin or was an artifact. The assay was per— formed on many plaques but only three indicated the presence of phage. They were derived from mitomycin c induced cultures of M835. The identity of these phage with the originally isolated P35 could not be immediately inferred due to the known variety of phage inducible from Salmonella (95). Comprehensive genetic, serological and morphological comparisons have not been performed on the separate phage isolates. Isolation of P35c. Occasionally a clear, sharp- edged plaque would appear on a plate which contained many plaques of P35. One of these plaques, which was assumed to be caused by spontaneously occurring mutations in the phage, was picked and the phage purified by several plaque isolations on S. pullorum M835. The phage was labeled P35c. The frequency of the occurrence of the mutation(s) leading to the clear plaque phenotype was calculated by using the equation presented by Stent (80) where M is the chance per round of phage DNA replication of the mutation event occurring (mutation frequency) and no and p0 are 48 the proportion of mutants and the total number of phage in the culture at time equal to zero. A and p are then the prOportion of mutants and total phage in the culture after some fixed time interval. TT-TT O M = ln(p/p0) M for the mutation(s) to the clear plaque phenotype was calculated to equal 1.2 x 10-4 when P35 was grown in S. pullorum M835 and 1.4 x 10”4 when P35 is propagated in S. typhimurium LT2. Spontaneous revertants of P35c were not observed. P35c is a virulent mutant which has apparently lost the ability to enter into the prophage state. Super— infecting P35c are adsorbed by lysogenic strains of S. pullorum but are not propagated indicating that the mutant is sensitive to and inhibited from replication by the immunity repressor produced by the resident prophage genome (67). When P35c was simultaneously infected at an equal multiplicity of infection with P22c2, no comple- mentation was observed (i.e., no turbid plaques were pro- duced) indicating that P35c probably has a defect in the gent for production of the repressor protein which is necessary for the establishment of lysogeny (75). P35c grows at the same rate as wild type P35 but produces maximum titers of 2 x 1010 — 3 x 1010 9 pfu/ml in contrast to the 6 x 10 pfu/ml produced in P35 lysates. 49 Isolation of S. pullorum M853(P35). A log phase (1 x 102 cells/ml) culture of S. pullorum M853 in aerated L broth was infected with P35 at an m.o.i = 10.0. After 3 hours growth the surviving cells were streaked for iso— lated colonies on L agar. After overnight incubation at 37C many of the colonies had a rough appearance instead of the typical smooth colony of S. pullorum. Several of these were picked and purified by three more clonal iso- lations. Each of these when tested was found to be lyso- genic as indicated by Spontaneous phage liberation and an inability to produce plaques when superinfected with P35 or P35c. Antisera and reactions in SIM agar were positive for S. pullorum. The lack of plating of the phage was due not to an inability to adsorb the superinfecting phage (Figure 3) but apparently to immunity due to the presence of the prophage. When grown in either L broth or TCG broth, a culture of M853(P35) will contain 1.5 x 10"2 free pfu/cell. This lysogenic strain, MSS3(P35), was used to further investigate the relationship of P35 to P22. It was found that upon superinfection of the lysogen by P22, although the phage were adsorbed, no plaques were pro— duced. This indicated that P22 and P35 have similar immunity systems (i.e., are homoimmune). Morphology of P35. To further characterize P35, electron micrographs were taken of PTA stained phage preparations. Figure 4 shows a field containing many Figure 3. 50 Adsorption of P35c to S. pullorum 53 and 53 (P35). Log phase cultures (1 x 108 cells/ml of S. pullorum 53 and 53 (P35) were infected with P35c at an m.o.i. = 0.1. At intervals 1 m1 samples were removed and treated with chloroform. After removal of the cells by centrifugation, the phage in the supernant fluid were titered on sensitive indicator lawns of M853. Only the clear plaques of P35c were counted. P35 which is spontaneously liberated from 53 (P35) produces a turbid plaque. The titer of P35 remained con— stant at 2 x 106 pfu/ml throughout the time of the experiment. 53(P35) 0 pmbeompocn mmoxm “N0 5 10 TIME (min) Figure 4. 52 Electron micrograph of P35 phage particles. P35 was purified on a CsCl step gradient. The electron micrograph was taken after negative staining with 0.5% PTA. Magnification is 201,000x. The bar represents 100nm. 54 phage particles isolated after propagation on M853. Many of the phage heads appear to be devoid of a core (nucleic acid). This is believed to be an artifact of the prepa- ration since a similarly treated sample of S. 99;; phage T4 also appeared to be empty headed. Figure 5A is an enlarged View of two P35 particles. It can be seen that the phage head is hexagonal in shape and approximately 60-70 nm in diameter. There is a short tail and a base plate with spikes. Very occasionally in the lysates prepared on S. pullorum M853 there were observed phage- 1ike particles with an unusual structure. Figure 5B shows such a particle with a long tail (150 nm) and triangular shaped head. These particles may have been contaminants or have been induced from S. pullorum upon infection with P35. Since different phages have not been detected by plaque morphology or growth character- istics these particles may be defective phage or bacteriocins. The DNA of P35. To determine the size of the P35 DNA molecule labeled phage DNA was analyzed on a sucrose gradient. A phage lysate was prepared on S. pullorum M853 grown in 30 ml of TCG broth containing 3H—thymidine (l uCi/ml). The 3H labeled phage were harvested and purified on a CsCl step gradient. DNA was then extracted from the phage by treatment with phenol. A 0.1 m1 sample 3 of the purified P35 H-DNA was then mixed with 0.05 ml of 14C labeled colEl DNA. This mixture was then layered Figure 5. 55 Electron micrographs of indicidual phage particles. A. PTA stained preparation of phage P35. Magnification is 540,000x. The bar represents 100 nm. B. An example of a particle with unusual morphology found in a PTA stained preparation of P35. Magnification is 540,000x. The bar represents 100 nm. 56 57 on a neutral 20-31% sucrose gradient. The colEl DNA had been extracted from p. ggli JC411 (31+) and purified in a CsCl—EtBr dye buoyant density gradient. When prepared in this manner, colEl DNA is a supervoiled molecule with a sedimentation coefficient of 238 (9). The gradient was centrifuged for 115 min at 50,000 RPM and then fraction- ated by collecting drops from a needle that had been forced through the bottom of the tube. Figure 6 shows that the 3H labeled P35 DNA moved slightly further in the gradient than the 14C labeled colEl DNA. By comparison with the known 5 value of colEl DNA, it was calculated that P35 DNA has a sedimentation coefficient in neutral sucrose of 335. If the phage DNA molecule is a linear structure, as is P22 DNA (68), then from the relationship of Bazaral and Helinski (8) it can be determined that the 6 molecular weight of the P35 DNA molecules is 28 x 10 daltons. Part II Isolation and Characterization of Plasmid DNA from S. pullorum Isolation of plasmid DNA. To determine if S. pullorum contains plasmid DNA a lysate was layered on CsCl EtBr solution and centrifuged to equilibrium. Cells were labeled with 3H-thymidine (l uCi/ml) in 30 ml of TCG and were lyxed with lysozyme-EDTA and Brij 58-deoxycholate treatment followed by freeze thawing. After removal of Figure 6. 58 Sucrose gradient of P35 DNA. An 0.15 ml sample containing 3H labeled P35 DNA and 14C labeled colEl DNA was layered on a 20-31% neutral sucrose gradient. The gradient was centrifuged in an SW50L rotor at 50,000 RPM for 115 min. Fractions were collected directly on to filter paper squares, dried, washed with TCA and counted in a scintillation counter. 3H 14C -o--o-. Sedimentation in this and all subsequent sucrose gradients is from right to left. 400 CIPAA 59 Col ‘1 P35 1‘ \ 0 °/ \0‘9‘0-0-0-0-0-0-0 .I’WO‘o-O‘o-O—Moo- .. 1 10 20 30 Fractions 60 most of the chromosomal DNA by centrifugation in the RC-2B centrifuge, the cleared lysate was layered on a CsCl-EtBr dye-buoyant solution and centrifuged to equilibrium. At equilibrium two fluorescent bands could be observed when the tube was illuminated with a mineral lamp. The gradient was fractionated into approximately 60 fractions by collecting drOps from a needle placed through the bottom of the tube. Aliquots of 10 ul from the fractions containing the fluorescent bands were spotted on filter paper squares, washed with TCA, dried, and the radioactivity counted in a scintillation counter. Figure 7A shows that there were two peaks of radio- activity in the CsCl-EtBr gradient of the cleared lysate from S. pullorum M853. The denser peak in fractions 17-21 represented covalently-closed circular plasmid DNA molecules as demonstrated by Radloff, Bauer and Vinograd (66). The lighter peak was composed of Open—circular plasmid and linear chromosomal DNA. From a comparison of the total amount of radioactivity in the satellite fraction (7.36 x 106 cpm) to the total radioactivity taken up by the cells (4.5 x 107 cpm), it was calculated that about 16% of the total cellular DNA existed as extrachromosomal plasmid DNA. The plasmid fractions were pooled, dialyzed against TES buffer and a 0.1 ml sample was layered on a neutral 20-31% sucrose gradient. After centrifugation the 61 Figure 7. Isolation of plasmid DNA from S. pullorum M853. A. 3.0 m1 of a cleared lysate of S. pullorum M853 labeled with 3H-thymidine were centrifuged in a CsCl-EtBr dye-buoyant density gradient for 30 hours at 44,000 RPM in a type 50 rotor. Fractions were collected and 10 ul samples spotted on filter paper squares, TCA washed and counted for radioactivity. Fractions 16-21 containing plasmid DNA were pooled and dialyzed in the dark against TES buffer. 0.1 ml of the pooled plasmid DNA from the CsCl-EtBr gradient was layered on a neutral 20—31% sucrose gradient. The gradient was centrifuged for 180 min in an SW39 rotor at 38,000 PRM. Fractions were collected directly onto filter paper squares, TCA washed and counted for radioactivity. 62 200 n -2123 .:n 0 O 1 0 2 "6:23-..." 0 . ti 3O 25 20 )5 Fractions 15 0 5 1 90:23-: 20 10 Fractions 63 gradient was fractionated and the results shown in Figure 7B indicate that S. pullorum M853 contains two species of plasmid DNA of greatly different sedimentation values. Identical results were also obtained with S. pullorum M835. Sedimentation coefficients and molecular weights of the plasmid DNA. To determine the size and approximate molecular weights of the supercoiled DNA from S. pullorum M853 the sedimentation coefficients were calculated by cosedimenting the plasmids with colEl DNA. S. pullorum plasmids labeled with 3H and 14C labeled colEl DNA from S. 39;; JC411 (El+) were first isolated on CsCl-EtBr dye- buoyant density gradients and then cosedimented in a neutral sucrose gradient. It has been shown by Bazaral and Helinski (9) that when colEl DNA was isolated from a CsCl—EtBr gradient that it was a supercoiled molecule of 4.6 x 106 daltons with a sedimentation coefficient of 233. From the results shown in Figure 8 it can be calcu- lated that the faster sedimenting plasmid of strain M853 has a sedimentation coefficient of approximately 653 while the smaller plasmid is about 175. From the relation- ship given by Bazaral and Helinski (8) it can be computed that the 17s plasmid corresponds to a supercoiled DNA molecule of approximately 1.5 x 106 daltons and the 655 6 supercoiled plasmid has a molecular weight of 45 x 10 daltons. Figure 8. 64 Neutral sucrose gradient of plasmid DNA from M853 and colEl DNA. A 0.15 ml sample containing 0.1 m1 of 3H labeled M853 plasmid DNA and 0.05 ml of 14C labeled colEl DNA were layered on a 20-31% neutral sucrose gradient. The gradient was centrifuged in an SWSOL rotor for 95 min at 50,000 RPM. 14c -o-o-, 3H 1000 CPM 500 65 Fractions 66 The 175, 1.5 x 106 daltons molecule will hereafter be referred to as the PO-l plasmid and the 45 x 106 daltons molecule as the PO-2 plasmid of S. pullorum Alkaline sucrose gradient sedimentation of plasmid SSS. To verify that these plasmids are indeed supercoiled DNA, samples of the plasmid fraction from a CsCl—EtBr gradient were centrifuged in an alkaline sucrose gradient. It has been shown by Clayton and Vinograd (24) that when covalently-closed circular DNA is denatured and sedimented in an alkaline gradient, the 5 value is increased by approximately 2.5 fold. Figure 9 shows that when the plasmid DNA from S. pullorum M853 is centrifuged in an alkaline sucrose gradient, both peaks move a greater distance than in the neutral gradient. The data of a previous section indi— cated that the smaller PO—l plasmid sedimented at 175 and the larger PO-2 plasmid at 655 in neutral sucrose. By comparison the plasmids have sedimentation values of 355 and 1605 respectively in alkaline sucrose which would be the approximate values expected for supercoiled DNA. These data, therefore, verified the condlusion that the DNA observed in the neutral sucrose gradient was super— coiled as would be eXpected by their isolation from CsCl-EtBr gradients. Sedimentation of a cleared lysate on a neutral sucrose gradient. The isolation of plasmid DNA from a Figure 9. 67 Alkaline and neutral sucrose gradients of S. pullorum 53 plasmids. 0.1 ml samples of 3H labeled plasmid DNA isolated from a CsCl-EtBr gradient were layered on neutral and alkaline 20-3l% sucrose gradients. The gradients were centrifuged simultaneously in an SW50L rotor at 50,000 RPM for 45 min. 68 Neutral 900 600 300 90. Alkaline 3H-c PM 600 300 10 20 30 Fractions 69 Cs-Cl-EtBr gradient excluded the possibility of identifying any species of plasmid DNA that may have been nicked to the open circular form. To determine if any open circular molecular forms are present, a cleared lysate was layered directly on a neutral sucrose gradient and centrifuged. Three peaks of radioactivity were observed instead of two observed in the CsCl—EtBr purified preparation. As shown in Figure 10 this third peak sediments between the 17s and the 655 peaks in a region approximately corresponding to 455. According to Clayton and Vinograd (24), this is the expected sedimentation value of the Open circular form of a 655 supercoiled DNA molecule. Another indication that the 455 peak represents the open circular form of the PO-2 supercoiled molecule is the quantitative conversion of the material isolated from a CsCl-EtBr gradient from 655 to 455 during storage at room temperature or upon repeated thawings of a frozen sample. Probably due to its P- much greater size, the PO—2 plasmid was more likely to be nicked than was the PO-l plasmid. Plasmid PO-l showed no Open-circular or linear forms in any cleared lysate ‘0 produced during this work. From three similar gradients of cleared lysates of S. pullorum 53 it was calculated that the larger PO—2 plasmid molecule (both closed and Open circular forms) accounted for approximately 16% of the total plasmid DNA. Therefore, since there is five times as much DNA in the Figure 10. 70 Neutral sucrose gradient of a cleared lysate. A 0.2 ml sample of a cleared lysate of S. pullorum 53 labeled with 3H-thymidine was layered directly on a neutral 20—3l% sucrose gradient which was centrifuged for 220 min at 39,000 RPM using an SW39 rotor. 3H-c PM 600 300 10 71 Fractions 20 3O 72 PO-l peak as there is in the PO-2 peaks, there is 5 x 45 x 106 daltons = 225 x lo6 daltons of PO-l DNA for each COpy of the PO-2 molecule. There is, therefore, 225 x 106/1.5 x 106 = 150 COpies of the smaller PO-l plasmid per copy of PO-2. At one copy of the large molecule per cell there would be 225 x 45 x 106 = 270 x 106 daltons of plasmid DNA per cell. Assuming a molecular weight of the S. pullorum chromosome close to that of S. coli 9 which is 2 x 10 (22) then the plasmid DNA would account for 14% of the total. This is in close agreement with the amount of plasmid DNA calculated from the original CsCl-EtBr gradient. Electron microscopy of plasmid DNA. The previous sections have demonstrated that S. pullorum contains two distinct species of supercoiled plasmid DNA molecules. To observe the structure of these molecules, samples from the plasmid fraction of a CsCl-EtBr density gradient were pre- hr pared for viewing in the electron microscope. Cytochrome c protein monolayers containing the plasmid DNA were pre— pared by the microversion of Lang and Metani (54) and picked up on carbon-coated collodion grids. These were then viewed in the electron microsc0pe and pictures taken of representative fields containing the circular DNA molecules. Figure 11A shows one of the many small circu- lar structures seen in the preparation. From its size it was assumed that this probably is a molecule of the PO-l plasmid in the Open circular configuration. 73 Figures 118 and 11C show a much larger molecular species which is assumed to be the PO-2 plasmid. Both molecules show some interraveling of the strands and areas in which the strands appear to be discontinuous but these may be due to some supercoiled structure remaining in the open circular forms. No evidence for interlocked catenated molecules was observed in the PO-2 molecules. Part III Exclusion of Plasmid DNA in S. pullorum by Lysogeny with P35 Plasmid DNA in S. pullorum M853(P35). To deter- mine if lysogenization by P35 had affected the plasmids which were already present in S. pullorum M853 the lyso- genic strain M853(P35) was labeled with 3H-thymidine, lysed and the supercoiled DNA isolated in a CsCl-EtBr dye-buoyant density gradient. Centrifugation of the pooled plasmid fractions in a neutral 20-31% sucrose gradient revealed that now, after lysogeny of this strain, the PO-2 plasmid was no longer present (Figure 12). The PO-l plasmid was still present and sedimentation in an alkaline gradient indicated that it remained in the covalently-closed circular configuration. The loss of the plasmid was an unexpected finding since it had already been demonstrated that the DNA isolated from phage P35 has a molecular weight of about 28 x 106 and plasmid PO-2 is 1.6 times larger than this. Figure 11. 74 Electron micrographs of plasmid DNA of S. pullorum M853. DNA from the plasmid fractions of a CsCl—EtBr density gradient were prepared for viewing in the electron microscope by the microdrop technique. The bar in each picture represents 1 u. A. Small Open circular plas- mid, PO-l. Magnification is 34,000x. B. PO-2 plasmid molecule. Magnification is 34,600x. C. PO-2 Open circular molecule. Magnification is 49,500x. yuan... __ _. “A 75 1 1‘ .1 All! tit} 111‘ 11.14. v 76 Figure 12. 77 Plasmid DNA of 53(P35). A. 3.0 ml of a cleared lysate of 53(P35) labeled with 3H—thymidine were centrifuged in a CsCl-EtBr gradient for 36 hours at 44,000 RPM in a type 50 rotor. Fractions were collected, and the radioactivity in 5 ul samples counted. The plasmid DNA in fractions 14-19 was pooled and dialyzed against TES buffer. 0.1 ml samples of the pooled plasmid DNA from the CsCl-EtBr gradient were layered on neutral and alkaline 20-3l% sucrose gradients. The gradients were centrifuged for 50 min at 50,000 RPM in an SWSOL rotor. to l 3H-CPMxlO'3 0 l 3H-CPMX Io'3 CsCI'EtBr .—a O I 78 :L O ”rt-CPMMO‘3 Alkaline U 20 Fractions Neutral A T T I r on O Fra ctions l I l 10 20 3 O 79 Isolation of a cured derivative of S. pullorum M853(P35). Since it has been demonstrated that when S. pullorum M853 is made lysogenic for P35, the PO-2 plasmid is excluded, it was of interest to find out what would happen when the strain was cured of the prophage. Two possibilities existed: the plasmid would not be re~ gained or the plasmid might be regained if it originated from the chromosome and was continuously and randomly being excised. Approximately 50 isolated clones of S. pullorum 53 (P35) were picked and grown in L broth. After 8 hours of incubation the supernatant fluids were tested for the presence of free phage particles by spotting chloroform treated aliquots on to fresh lawns of sensitive cells. Four isolates were found not to liberate phage and were purified by three cycles of clonal isolation and then retested for phage production. These isolates were then tested for their ability to propagate phage P35. All were able to adsorb and propagate the phage. The iso- lates were then tested for their ability to produce phage upon treatment with mitomycin c or UV light. One, 8192, was found to be non-inducible and also to be sensitive to P35, P35c, and P22. To determine if 8192 had regained the PO-2 plasmid a sample of a cleared lysate was analyzed directly on a neutral sucrose gradient. This allowed detection of both 80 the closed and open circular forms of the PO-2 molecule. The results shown in Figure 13 indicated that 8192 like its parent 53(P35) did not contain the larger PO-2 plasmid molecule. Attempt to follow the fate of plasmid PO-2 following infection with P35. Two experiments were performed to determine the fate of the PO-2 plasmid subsequent to infection of S. pullorum M853 with phage P35. In the first experiment log phase cells were infected with P35 at an m.o.i. of 10. 3H-thymidine was then added to the culture which was then incubated at 37C. At intervals after infection 10 ml samples were removed and the DNA replication quickly stOpped by addition of KCN (to a final concentration of 0.002 M) and EDTA (to a final concen— tration of 0.05 M) and by chilling the cells in an ice water bath. The cells were then lysed and the cleared lysates centrifuged in neutral sucrose gradients. The results shown in Figure 14 indicate that by 30 min after infection much of the newly made DNA is in a form which sediments at a rate about equal to phage DNA. At 50 min the bulk of the DNA sediments in an extremely broad peak and at 90 min most of the DNA remaining in the cells is of phage DNA size. It also can be seen that by 90 min there is no indication Of a peak at 655. The PO-l plasmid was present in each gradient and served as a reference marker for sedimentation. Figure 13. 81 Neutral sucrose gradient of cleared lysate of 5192. 0.1 ml of a 3H labeled cleared lysate of 8192 was layered on a 20-31% neutral sucrose gradient centrifugation was in an SW39 rotor for 220 min at 39,000 RPM. 3H-c PM 1200 10 82 F ractions 20 3O Figure 14. 83 DNA synthesis in P35 infected S. pullorum M853. 10 ml cultures of M853 were grown to log phase in aerated TCG broth and labeled with 3H-thymidine. The cultures were infected with P35 (m.o.i. = 10) and at the times indi- cated after infection a culture was chilled in an ice bath and 0.1 ml of KCN (0.2 M) plus 1.0 m1 of EDTA (0.5 M) was added. Cleared lysates were made of each culture and 0.15 ml samples were layered on 20-31% neutral sucrose gradients which were centrifuged at 38,000 RPM for 220 min in an SW39 rotor. 84 15min 30 min l 1 r 10 20 . 30 Fractions 85 In the second experiment cells were first labeled with 3H-thymidine, harvested and resuspended in TCG broth without 3H-thymidine. They were then infected with P35 at an m.o.i. of 10. As before, samples were removed at intervals, treated with KCN and EDTA, chilled and lysed. When these cleared lysates were centrifuged in neutral sucrose gradients, it was seen that soon after infection the cells contained a very broad distribution of labeled DNA (Figure 15). Apparently not all of this label could have come from the amount originally present as plasmid DNA but was probably derived from chromosomal DNA. Again late in infection there does not appear to be 655 peak present. These experiments gave very little information about the process of or the reason for the exclusion of plasmid PO-2. More definite results would have been obtained if if were possible to pulse label the infected cells for short time intervals with 3H-thymidine but the poor uptake ability of S. pullorum for thymidine necessitated the lengthy exposures to the label. To explain the apparent exclusion it was postulated that the plasmid must in some way be related to the phage. They may perhaps share common cellular replication sites or be sensitive to the same immunity system. Either of these would require that there be a degree of nucleotide similarity (homology) between the phage DNA and the plas- mid DNA. 86 .uouou mmam cm CH 2mm ooo.mm um cHE omm ROw ODODMHHDCOO can mucoflcmnm omOHOSm Hauuawa wamlom co Omuomma oumB mmuwmma Umummao on» mo mdeEmm HE ma.o .cmmma can OHOMNQ we deem cam zum nuflz moEHu ODDMOHOCH on» um Omumonu onm3 mmusuaso may Aoa u .H.O.EV mmm zuflz cofluommcfl umuma .oaficflemnulmm usocufl3 nuoun OOH amoum cfl Ooccmdmsmmn can coco cmgmm3 .Omumm>umn mum3 maamo on» cofluommcfl DHOmon Dan ea musmflm CH Omnflnomme mm coaonmfl can CBOHO mnm3 mmmz mo mousuado .mmm sues pouommcfi mmmz Umamnmamud ca mammzucwm mzo .mH musmflm . a... 87 acetone“. acetone“. op Om. on op _ _ _ . loop [009 _ 528. icon _ _ scan 188 h w s n :82 m w 3 room 4 recon w . r002 EEn NH 522. Tooom _ _ a _ loco. 1000 £83 looom 5:2... 1005. 88 Part IV Hybridization Studies with S. pullorum Plasmid DNA The demonstration in Part III that plasmid PO-2 is excluded from cells lysogenized by phage P35 indicated a probable close relationship between the plasmid and phage DNAs. To determine if, in fact, they do contain homologous sequences it was decided to perform hybridization experi- ments between the phage and the plasmid. Initially, DNA- DNA hybridizations were attempted but these were unsuccess— ful due to an inability to isolate 3H-labeled DNA of high specific activity and due to the low DNA yields from P35 phage preparations. It was, therefore, decided to synthe- size 12 XEEEQ 3H—labeled RNA of high specific activity from a phage template and to use this 3H-RNA to hybridize with the plasmid DNA. The DNA extracted from P35c was used as a template rather than DNA from P35 since the yield of purified DNA per liter of lysate was greater with P35c than with P35. Assay of Pseudomonas putida RNA polymerase using P35c DNA as a template. Before using the RNA polymerase to produce a quantity of 3H-RNA sufficient to use in hybridization studies, it was necessary to assay the enzyme's ability to incorporate nucleotide triphosphates into RNA when using P35c DNA as a template. The prepa- ration Of RNA polymerase provided by Dr. J. Boezi 89 contained 14.4 mg of protein per ml which when assayed on phage gh-l DNA had a specific activity of 2000 n moles of 3H-CMP incorporated per hour per mg of protein. The results of an assay using P35c DNA as a template are shown in Figure 16 and the specific activity calcu- lated from the data is 450 n moles of 3H-UTP incorporated per hr per mg of protein. This, therefore, indicated that the S. putida RNA polymerase could transcribe P35c DNA although at a rate one third that using gh-l DNA and that the enzyme could be used for the 12 XAEEQ synthesis of 3H—RNA. In vitro synthesis of 3H-RNA. Since the assay had indicated that the S. putida RNA polymerase could use P35c DNA as a template, a procedure was set up to synthesize sufficient RNA for hybridization studies. To 3 m1 of the reaction mixture were added 41 mg/ml of RNA polymerase, 50 ug/ml of P35c DNA and 3 H-UTP (81.5 uc/u mole) to a final concentration of 0.09 mM. This was incubated for 10 min at 37C. The reaction was stOpped by addition of 0.3 ml of 10% SDS and the RNA extracted three times with water-saturated phenol. After dialysis to remove the phenol and any remaining unincorporated nucleotides, the amount of 3H-RNA present was determined. The concentration of 3H-UTP which was added to the reaction mixture was determined from its UV absorbance at 269 nm to be 4.1 mM (the molar extinction coefficient of UTP at 269 nm is 104). The amount of radioactivity in the 90 Figure 16. Assay of RNA polymerase. Each assay mixture of 0.4 ml contained 50 ug of P35c DNA and 0.15 mM 3H-UTP at 10 uc/u mole (2.09 x 106 Cpm—u mole) and was incubated for 10 min at 37C. 91 20 o 15 ‘1" O X 10 2 O- U | o I ('0 5 o P O 25 50 100 ug protein/ml 92 3H-RNA and 3H-UTP was determined by spotting samples on glass fiber filters and solubilizing them with soluene- 100 for maximum and equal counting efficiency. The 3H-UTP was diluted 1:20 and a 0.1 ml sample of the dilution containing 0.0205 u moles and was spotted and counted. It was found to have 784,795 Cpm. The specific 3H-UTP, therefore, was 3.83 x 107 cpm/u mole. activity of A 0.05 ml sample of the synthesized 3H-RNA was counted and found to have 6436 cpm. Therefore, in a 0.05 ml sample of 3H-RNA there is 6436 Cpm/3.83 x 107 Cpm/u mole 3H-UTP or 0.168 n moles of 3H—UTP as 3H-RNA. Assuming equal molar amounts of the four nucleotide triphosphate incorporated into RNA and the average molecular weight of the nucleotide monOphosphates is 411, then 4 x 0.168 n moles/0.05 ml equals 0.672 n moles of nucleotide mono- phosphate per 0.05 ml. And 411 x 10-9 gm/n mole nucleo- tide monophosphate x 0.672 n moles of nucleotide mono- phosphates as RNA/0.05 ml equals 272 x 10.9 gm per 0.05 ml. There is, therefore, 0.272 ug RNA/0.05 m1 x 20 equals 5.45 ug 3H-RNA/ml. Hybridization efficiency. To test the efficiency 3 of hybridization of the H—RNA to its complementary DNA and to determine the Optimal DNA concentrations to be used in subsequent experiments, varying concentrations of denatured P35c DNA were annealed to the RNA for 6 hours at 60C. After incubation the DNA-3H RNA complexes 93 were collected on nitrocellulose filters. The results shown in Figure 17 indicate a linear increase in the amount of 3H-RNA collected as the amount of test DNA added was increased up to a maximum concentration. At greater DNA concentrations it is likely that some DNA- DNA interaction may interfere with the DNA—RNA hybridi- zation. In all further experiments, where possible, 5 ug of the test DNA per ml were used in the hybridization reactions. It was also noted that the percentage hybridization varied from experiment to eXperiment. Therefore, in all subsequent experiments a control hybridization with P35c DNA was included and the results expressed relative efficiency to the control. Hybridization of 3H-RNA (P35c) to phage DNAs. To test the relatedness between phage P35c and its assumed parent P35 and between P35c and S. typhimurium phage P22, hybridizations were performed using the RNA transcribed from P35c and the DNAs isolated from phages P35, P22, and the unrelated S. 29;; phage T4. It was necessary to definitely determine that P35c and P35 are identical since P35 is the lysogenic phage which excludes the PO-2 plasmid but the RNA was made with P35c DNA as a template. The data shown in Table 2 indicate that P35c and P35 each contain homologous DNA and likely are the same phage. The DNA isolated from P22, however, shows a Figure 17. 94 Hybridization efficiency of 3H-RNA (P350) to P35c DNA. Each hybridization was performed in 1 ml of 2x SSC which contained 0.126 ug of 3H-RNA (9200 cpm/ug) and the indicated concen— tration of denatured P35c DNA. Annealing took place for 6 hours at 60C after which the DNA— H-RNA complexes were collected on nitro- cellulose filters. 95 .~ 7 2 voicing: