Proc. Not. Ad. Sd. USA Vol. 72. No. 10, pp. 38633867. October 1975 Biochemistry DNA replication in vitro starting with an intact 6x174 phage (membranes/phage infection) s. MICHAL JAZWINSKI* AND ARTHUR KORNBERG Department of Biochemistry. Stanford University School of Medicine, Stanford, California 94305 Contributed by Arthur Kmberg, July 10,1975 AWlXACX Conversion of the singlestranded DNA in the intact 4x174 phage particle to the du lex replicative form (RF) has been demonstrated in lysates tom phage-sen- sitive cells. The conversion is resistant to rifampicin and requires participation of both a "membrane'* fraction of the lysate and a multienzyme re licative system. The lip polysaccharide hage receptor, w%ile essential, does not re- place the memgrane fraction. Clear, nonsedimentable ex- tract fractions prepared with a certain nonionic detergent can replace the membrane fraction. Purification of the activ- ity in these extracts by adsorption to polypropylene film ields a fraction with a Cfold increase in activity relative to E polysaccharide and !%-fold increase relative to rotein. TE low buoyant density (1.03 g om3) suggests a hi$ phos- pholipid or detergent content in t h is fraction. 4x174 is an icosahedral coliphage with spikes at the 12 ver- tices and contains a single-stranded (SS), circular DNA (1). Infection proceeds through stages of phage binding, eclipse, and DNA penetration (2) dependent on the presence of the correct lipopolysaccharide (LPS) phage receptor at the cell surface (3). The initial event in viral DNA replication, syn- thesis of the double-stranded, circular, parental-replicative form (RF), is tightly coupled to DNA penetration (4), and requires no new protein synthesis (1). The parental gene H spike protein, the phage adsorption protein (5). is in some way involved in the conversion of the DNA of intact 4x174 phage particles to RF (4X - RF) tn ofvo (6). Progress in the dissection and understanding of the host enzymes directing the conversion of naked 4x174 SS to RF in vitro (7) has led us to inquire about the host and viral re- quirements for uncoating and replicating the DNA of intact 4x174 phage particles. To this extent we have sought to de- vise an tn oitro system for studying RF formation starting with intact phage particles (4X - RF). We have found and we report an two such systems and their properties. 4X - RF conversion in cell extracts prepared with non- ionic detergents requires a "membrane" fraction and solu- ble, replicative enzymes. One of the membrane components required is the LPS phage receptor. Adsorption to polypre pylene film from crude cell extracts leads to a partial purifi- cation of the 4X --t RF activity. MATERIALS AND METHODS Phonic P103 was a gift from BASF Wyandotte (Wyan- dotte, Mich.); Celgard 2400W polypropylene film was a gift Abbreviations: SS. (phage) singlestranded DNA; RF, (phage) dou- ble-stranded circular replicative form; RF I, (phage) covalently closed RF; RF 11, (phage) RF with a discontinuity in at least one strand; 6X - RF, conversion of DNA of intact 4x174 phage parti- cles to RF; LPS. lipopolysaccharide; Buffer A, 10% sucrose-50 mM Tris-HCI (pH 7.5); Buffer B, 10% sucrose-50 mM Tris-HCI (pH 7.5)-50 mM NaCI-50 mM dithiothreitol-0.4 mg/ml of bovine serum albumin; PFU, plaque forming units. *Present address: The Rockefeller University, New York, N.Y. 10021 from Celanese Plastics Co. (Greer, S.C.). Bacterial and phage strains and sources, as well as preparation of radioactively la- beled phage, have been described (3, 8). 32P-labeled 4x174 (ad) preparations contained less than 6% eclipsed particles. Extracts and Fractionation. Escherichia coli cells were grown in M medium (3) to 0.5 Ass at 37" on a shaker. Cells were sedimented for 10 min at 20" and resuspended in 0.01 volume of Buffer A (10% sucrose-50 mM Tris-HC1, pH 7.5). The suspension was frozen in liquid Nz and thawed at 0" in EiO-pl aliquots. CelL were converted to spheroplasts by in- cubation with lysozyme (200 pg/ml) and 0.1 M NaCl for 30 min at 0". Triton X-100 (10%) was added to a concentration of 0.5% and the spheroplasts were incubated for 70 sec at 30". Over 95% of the cells were lysed by this treatment, as determined in a phase-contrast microscope. The suspension (lysate) was kept at 0" or sedimented at 30,000 X g for 30 min (0") and the pellet ("membranes") and supernatant kept at 0". Phonic P103 replaced Triton X-100 in some in- stances. Extracts were fractionated by adsorption to polypro- pylene film. Strips (7 X 4 mm) were added (15 strips per ml of Pluron- ic extract) and the extract was kept at 0" for 12 hr. The ex- tract was then withdrawn (supernatant) and the strips were eluted with an equal volume of 1% Phonic P103 in Buffer A for 8 hr at 0" (eluate). Replicative fractions I and 11 were prepared from E. colt H560 (9). Velocity Sedimentation. Samples were applied to %20% linear sucrose gradients, either neutral (10 mM Tris-HC1, pH 8-1 mM EDTA-1 M NaCl) or alkaline (0.2 M NaOH-1 mM EDTA-0.8 M NaCl). Gradients were formed over a 6.5 M CsCl shelf. Neutral gradients were centrifuged for 16 hr at 24.000 rpm (5") in a Spinco SW27 rotor. Alkaline gradients were centrifuged for 12 hr at 36,000 rpm (5") in a Spinco SW41 rotor. Other Methods. LPS receptor from E. coli HF4704 was prepared as described (3). Protein was determined accord- ing to Lowry et al. (10). LPS was determined according to Janda and Work (11) with E. colt HF4704 LE as standard. RESULTS DNA of intact 9x174 phage particles is converted to RF in cell-free extracts Cells (E. colt HF4704, sensitive to 4x174) were converted to spheroplasts and lysed with the nonionic detergent Triton X-100. Such lysates, fortified with Mg2+ and rib and de- oxynucleoside triphosphates, converted the DNA of in- tact 4x174 phage particles to RF (Table 1). Neither the sedimentable ("membrane") fraction of the lysate by itself nor the soluble replicative enzyme system in the supernatant which converts SS to RF (12) was capable of carrying out this conversion (Table 1). When the "membrane" fraction was recombined with the soluble replicative enzymes. full activity was restored (Table 1, Exp. 1). The soluble enzymes could be supplied as the supernatant fraction of the deter- 3863 3864 Biochemistry: Jazwinski and Kornberg Proc. Nat. Ad. Sci. USA 72 (1975) Table 1. @X + RF conversion in Triton X-100 lysates RF Mem- Enzyme (%of Exp. Lysate branes fraction Phage total) - - - + - Super- + Super- natant natant + + + + + + + + + LPS LPS complex complex + + + +b + + 57 <1 <1 46 41 42 67 41 <2 <1 <10 28 36 36 22 <1 37 0.1 02 ml FIG. 1. +X - FW conversion in Pluronic extracts. Pluronic ex- tract was prepared from E. coli HF4704 cells as described in Mate- rials and Methods. The extract was then centrifuged for 4 hr at 40,000 rpm (0') in a Spinco type 40 rotor in a polypropylene tube. The supernatant and the pellet, resuspended in an equal volume of Buffer A, were kept on ice. Activity was determined by incubating a sample of the fraction with: 25 pl of MgClz; 50 pl of rNTPs; 50 p1 of dNTPs; 10 pl of Fraction I1 (a saturating amount); and 80 pl of 32P-labeled 4x174 am3 phage (3.7 x 10" PFU/ml), aa in Table 1 made up to 475 pl with Buffer B (added first), for 20 min at 30°. The samples were then treated and analyzed by velocity dimen- tation in neutral gradients as described in Table 1. % RF formed was based on the recovered radioactivity (86-10296 pf that applied to gradients). Triton-lysate and the membrane and supernatant fractions from this lysate were prepared from E. coli HF4704 cells as described in Materials and Methods. Membranes were resuspended either in Buffer A (for incubation without Fraction I or Fraction II), in Fraction I, in supernatant, or in Buffer B (10% sucrose-50 mM Tris-HC1, pH 7.5-50 mM NaC1-50 mM dithiothreitol-0.4 mg/ml of bovine serum albumin) for incubation with Fraction II. Phage- E. coli HF4704 LPS complex was prepared by incubating 32P- labeled qjX174 am3 (1.5 x lo1' PFU) with 100 pg of LPS in 300 pl of 10 mM Tris.HC1, pH 7.5-1 mM EDTA-3 mM CaClz for 15 min at 37". To the lysate, supernatant, or membranes were added: 25 pl of 100 mM MgC12; 50 pl of a mixture of rNTPs (10 mM ATP and 1 mM each of CTP, GTP, and UTP); 50 pl of a mixture of dNTPs (1 mM each dATP, dCTP. dGTP, and dTl'P); 50 pl of Fraction II (where indicated); and 100 pl of 32P-labeled 4x174 am3 phage or phage-LPS complex. Phage was added at 5 PFU per cell equivalent and phage-LPS complex was added at 4 PFU per cell equivalent. The final volume of the mixture was 520 pl. Incubation was per- formed for 20 min at 30". The tubes were placed on ice, and 0.5 M EDTA and 10% sodium dodecyl sulfate were added to a concentra- tion of 20 mM and 1%. respectively. Incubation was then con- tinued for 15 min at 37`. The samples were chilled to 0'; [3H]thy- mine-labeled M13 phage was added as marker; samples were ap- plied to neutral sucrose gradients. Gradients were fractionated directly into vials in 6-drop fractions, and radioactivity was mea- sured in Triton-toluene scintillation fluid. Results are expressed as % RF in the gradients based on the recovered 32P radioactivity (81-97% of the 32P label applied to gradients). a Rifampicin was present at 10 pg/ml. "Fifty plaque forming units (PFU) per cell equivalent added in- c Prepared from a 4Xl74-resistant strain (E. coli 5274) d Membranes were washed in Buffer A at 0". gent-lysate, or another particle-free extract (Fraction I), or an ammonium sulfate fraction of the extract (Fraction 11) (Exp. 2). The activity present in the "membrane" fraction wa.~ not removed by washing with buffered solutions (Exp. 5). This 4x174 conversion to RF was also unaffected by rif- ampicin (Exp. 4). a specific inhibitor of .RNA polymerase, just as is the in vitro conversion of naked 4x174 SS (12). Lysates with similar properties were obtained when Brij-58 replaced Triton X-100 for spheroplast lysis (data not shown). stead of 5. It is unlikely that this conversion can be attributed to in- tact cells that survived lysis. In otoo, 20 to 30 phage particles per cell are uncoated and their DNA converted to RF (13). Inasmuch as at least 95% of the cells weie lysed by the Tri- ton treatment, a conversion of 22% of a phage input of 50 per cell equivalent (Table 1, Exp. 5) would have required that each intact cell remaining convert over 200 phage par- ticles. This conclusion is further supported by the fact that the sedimentable fraction of the lysate was inert in 4X -+ RF conversion. Lysates prepared from cells lacking the receptor for 4x174 (E. coli 5274) (3). were not capable of converting 6x174 to RF (Table 1, Exp. 5). More than the LPS receptor component of the "membrane" fraction appears to be in- volved, since a complex of phage with purified LPS receptor was not converted to RF by the soluble replicative enzymes (Exp. 3). However, such a complex was converted poorly, if at all, by the lysate. suggesting that exogenous LPS inhibits conversion. The product of 4X - RF conversion was characterized as RF I1 on the basis of its sedimentation in neutral sucrose gra- dients, alkaline sucrose gradients, and buoyant density in neutral CsCl gradients (data not shown). Comparable andy- ses performed on the product obtained with another form of lysate are presented below. "Soluble" extract for +X -+ RF conversion The Triton-lysate described in Table 1 is viscous and diffi- cult to manipulate. Another lysis procedure was therefore sought. When spheroplasts were treated with the nonionic detergent Pluronic P103, the cell DNA and the bulk of the cell membranes could be sedimented. The clear, nonviscous supernatant (Pluronic extract), fortified with soluble replica- tive enzymes, was capable of converting the DNA of intact 4x174 phage particles to RF (Fig. 1). 4X - RF conversion was dependent on Pluronic extract and soluble replicative enzymes (Fraction 11); conversion was not detected when a Biochemistry: Jazwinski and Kornberg Proc. Nut. Ad. Sci. USA 72 (1975) 3865 #X+RF I ACTIVITY PLUROWIC EXTRACT -e $ 3 w FRACTION NUMBER FIG. 2. Equilibrium sedimentation of gX - FW activity. Plu- ronic extract and polypropylene eluate were prepared from E. coli HF4704 cells (Materials and Methods) and 0.25 and 1.5 ml, re- spectively, applied to gradients composed of the following layers: 6.5 M CsCl, 70, 60, 50, and 25% sucrose (w/v) in 50 mM Tris-HCI (pH 7.5). The samples were overlayered with the same buffer. &a- dients were centrifuged for 44 hr at 54,000 rpm (5') in a Spinco SW56 rotor. Gradients were fractionated in 8-drop fractions. Phage-eclipsing activity was determined in 5- and 30-fil aliquots in the gradients in the top and bottom panels, respectively. +X -. RF activity was determined as in the legend of Fig. 1 on the fractions indicated by the shaded zone. Another zone indicates the position at which purified E. coli HF4704 LPS banded in a parallel gradi- ent. Pluronic extract from 4X-resistant cells (E. coli 5274) was used or when phage were incubated with purified LPS from E. colt HF4704 together with the detergent, Fraction 11, M$+, and rib and deoxynucleotides (data not shown). Upon high-speed centrifugation, 21% of the activity was recovered in the pellet, and 23% was assayed in the superna- tant (Fig. l). When the supernatant and pellet fraction were combined, the activities were only additive (data not shown). As will be seen, this loss of activity was due largely to adsorption to the polypropylene centrifuge tube. The activity of the extracts in eleven successive prepara- tions was 38, 26,45, 150, 12, 32, 22, 15,20,25, and 33 units/ ml. A unit is defined as the conversipn of 1O'O particles (4X - RF) in 20 min at 30". The activities of these Pluronic ex- tracts ranged from the conversion of 2 to 30 phage particles per cell equivalent. Fractionation and characterization of Pluronic extract The loss of 4X + RF conversion activity upon centrifuga- tion of Pluronic extracts in polypropylene tubes (Fig. l) sug- gested adsorption to polypropylene film as a possible frac- tionation procedure. Strips of polypropylene film (Celgard 2400 W) adsorbed nearly half of the 4X - RF stimulating activity from extracts, and most of this was eluted with 1% Pluronic P103 (Table 2). In six successive preparations, the percent of the activity adsorbed and eluted from polypro- pylene film was 15, 29, 39, 40, 26, and 30. Exposure of the activity initially unadsorbed to fresh polypropylene strips re- sulted in improved recoveries. Very little protein was adsorbed to polypropylene film (Table 2). Only 8% of the LPS was adsorbed (Table 2); for two other preparations, the values were 4 and 7%. Neverthe- Table 2. Polypropylene fractionation of Pluronic extract Specific activity Units/mg Units/ Activity Protein LPS of mg of (units) (mg) (mg) protein LPS Pluronic Polypropylene : extract 45 3.4 0.12 13 375 Supernatant 20 3.2 0.12 6 170 Eluate 17 0.025 0.01 680 1700 Pluronic extract, and polypropylene supernatant and eluate were prepared from E. coli HF4704 cells as described in Materkals and Methods. Activity in the fractions was determined as in the legend of Fig. 1. less, these fractions were, respectively, 60.21, and 50% as ac- tive as the Pluronic extract in promoting eclipse of phage particles (determined by nuclease sensitivity after Sarkosyl treatment) (3). (The total phageeclipsing activity recovered in the polypropylene supernatant and eluate fractions was 120-150% that of the Pluronic extract.) These data indicate that the LPS is heterogenwus in terms of its capacity to SUP port phage eclipse. Whether this is due to physial hetere geneity, such as the state of aggregation, or to assqciation with other cell components, is not clear. The fraction of LPS that is more efficient in promoting eclipse may be the same as that which participates in dg - RF conversion. The Pluronic extract and the polypropylene eluate were centrifuged to equilibrium in sucrose gradients and assayed for phage-eclipsing activity. The bulk of the eclipsing activi- ty of the Pluronic extract banded at the density of outer cell membrane (Fig. 2) (3, 14), whereas that of the polypropyl- ene eluate banded near the top of the gradient at a much lower density. A fraction from the less dense band (see shad- ed zone in Fig. 2) contained the 4X - RF conversion activi- ty, at 5 units/ml. This represents a recovery of 74% of the activity applied to the gradient, assuming the 4X - RF ac tivity to be proportional to eclipsing activity in this band. There was no 4X - RF activity associated with the small amount of eclipse activity in the band at the density of outer membrane. Similarly, 4X - RF activity was found only near the top of the gradient of the Pluronic extract, corre- sponding to the small peak of eclipse activity at that posi- tion. The low buoyant density (approximately 1.03 g/cm3) suggests that the material has either a high phospholipid or detergent content. The activity is not homogeneous in terms of buoyant density (Fig. 2). When viewed in the electron microscope, the polypro yl- size (data not shown); no such particles were seen in the re- agents themselves. Product of $X - RF conversion The products of 4X - RF conversion by the Pluronic ex- tract and polypropylene eluate were analyzed by velocity sedimentation, with the synthetic strand labeled with [3H]dTTP (Fig. 3). The Pluronic extract product contained the 3H label in material sedimenting at neutral pH as RF I and RF I1 (Fig. 3A) and coincident with peaks of =P label representing the phage strand. The SS appearing in these gradients (Fig. 3A and B) was derived from eclipsed phage disrupted by sodium dodecyl sulfate treatment during sam- ple preparation. When the peak fraction sedimenting as ene eluate contained'membrane fragments of 500-2000 1 in 3866 Biochemistry: Jazwinski and Kornberg Proc. Nut. Ad. Sci. USA 72 (1975) I NEUTRAL FRACTION NUMBER FIG. 3. Product of gX - RF in vitro. Pluronic extract and polypropylene eluate were prepared from E. coli HF4704 cells (Materials and Methods). [3H]dlTP (47 Ci/mmol, 100 pl) was added to a tube and dried under nitrogen. To this were added: 100 pl of Pluronic extract or 200 pl of polypropylene eluate; 50 pl of MgC12; 100 pl of rNTPs; 100 pl of dNTPs; 25 pl of Fraction II; and 200 pl of 32P-labeled 4x174 am3 phage (as in Fig. 1). The samples were made up to 0.96 ml with Buffer B (added fuat). Incubation was for 20 min at 30°. The samples were then treated and applied to neutral sucrose gradients as in Fig. 1 but without the CsCl shelf. The gradients we10 analyzed in 15-drop fractions. DNA was precipitated in 2504 aliquots of the fractions and collected on Whatman GF/C filters. Radioactivity was measured in toluene scintillation fluid. To 250 pl of the RF I and RF I1 peak fractions in A (Pluronic extract) and the RF peak fraction in B (polypropyl- ene eluate) were added 250 pl of 10 mM Tris-HC1 (pH 7.5)-1 mM EDTA, 20 pl of 5 M NaOH and 1 pl of 31P-labeled 6x174 am3 phage; after 10 min at 37O the samples were applied to alkaline sucrose gradients. Gradients were fractionated in 5-drop fractions; the DNA was precipi- tated and collected; and radioactivity was measured as above. The sedimentation is from right to left. RF I was analyzed in an alkaline sucrose gradient, only 45% sedimented as a closed-circular supercoil (Fig. 3C). Thus ap proximately 20% of the product was RF I. The remaining 3H-labeled material, as well as the material from the RF I1 peak, gave the pattern expected: full-length, linear comple- mentary strand derived from RF I1 (Fig. 3C and D). We at- tribute the composition of the RF I peak when analyzed on alkaline gradients to overlap of the RF I material with the broader peak of RF I1 in neutral gradients in the experiment shown. The product of conversion by the polypropylene el- uate was exclusively RF 11, as judged by sedimentation in neutral (Fig. 3B) and alkaline (Fig. 3E) gradients. The 3H label was incorporated into nearly full-length, linear com- plementary strands (Fig. 3E). In another experiment, how- ever, approximately 14% of the product was RF I; also in other experiments variable amounts of label were incorpo- rated into linear strands that were shorter than unit length. The 4X - RF results, based on the behavior of the 32P- labeled viral strand, were verified by analyses of the com- plementary synthetic strand labeled with 3H. The 4X - RF conversion by the Pluronic extract (Fig. 3A) was 29 and 4496, as judged by 32P and 3H label, respectively. The corre- sponding values for the polypropylene eluate (Fig. 3B) were 13 and 1995, respectively. Since the number of phage parti- cles was measured by plaque formation. a decrease in effi- ciency of plating (i.e., 0.66) could explain the discrepancy between the 32P and 3H results. Another possibility, how- ever, is that some uninfectious particles are present in the phage preparation and can be converted to RF in this sys- tem. DISCUSSION The tn oitro conversion of naked 4x174 SS to RF (7) differs from in oivo parental RF synthesis in at least three impor- tant respects: (a) naked SS are not the template in duo, i.e.. phage DNA penetration is tightly coupled to parental RF formation (4), (b) the gene H protein, a component of the phage coat, is involved in parental RF synthesis in duo (6). and (c) recovery of RF in association with the outer mem- brane (3) suggests that replication occurs at a membrane site (13). In order to explore the early events of phage replication in duo, an in oitro system capable of converting the DNA of intact 4x174 phage particles to RF was devised. The first such system was a cell lysate prepared with the nonionic de- tergent Triton X-100 (Table l). "Membranes" sedimented from the lysate were inactive, but the system could be re- constituted by adding back the soluble replicative enzymes (Table 1). The 4X - RF conversion was resistant to rifam- picin (Table I), an inhibitor of E. coli RNA polymerase, just as the fn dtro replication of naked 4x174 SS is (12), and the product of the reaction was RF I1 (data not shown). Since lysates prepared from cells resistant to 4x174 were inactive in 4X - RF conversion (Table l), we conclude that the correct LPS receptor (3) is a required component of the "membrane" fraction. However, the LPS appears not to be the only factor involved, since a complex of phage with LPS receptor was not replicated by the soluble enzymes alone (Table 1). A distinction should be made between phage complexed with the LPS receptor in oitro and phage released from the complex by the ionic detergent, Sarkosyl. The DNA of the phage-LPS complex is not utilized as a template by the solu- ble replicative enzymes; after Sarkosyl treatment. the DNA is as readily replicated by the soluble replicative enzymes as naked SS (data not shown). The phage-LPS complex forma- tion in dtro leads to a release of the gene H protein (S. L. Biochemistry: Jazwinski and Komberg Proc. Nut. Acud. Sci. USA 72 (1975) 3867 Rowen and A. Kornberg, unpublished observations), unlike the in vivo transfer of this "pilot" protein (15) into the cell with the infecting DNA (6). Thus, the action of the mem- brane fraction should include an involvement of the gene H protein in the C#JX + RF conversion. An inhibitory effect of antibody directed against the gene H spike protein of the phage on 4X - RF conversion in the lysate (data not shown) suggests that the gene H protein may be involved in RF formation in vitro, since the antibody was added after allowing time for phage binding and eclipse. The results do not distinguish between a direct and indirect effect of antibody binding to H protein on RFformation, but they are consistent with the notion that H protein is in- volved in $X - RF conversion in oitro. as it is in vivo (6). A Triton-lysate is viscous and unsuitable for the purifica- tion of the dX -. RF conversion activity. A clear, "soluble" cell extract can be obtained by using the nonionic detergent Pluronic P103. Fortified with the soluble replicative en- zymes (7), such extracts can carry out the 4X - RF conver- sion (Fig. 1). Both the Pluronic extract and the soluble repli- cative enzymes are required. A full-length linear comple- mentary DNA strand is synthesized and the product is main- ly RF 11, although some molecules are sealed, yielding RF I (Fig. 3). A fractionation procedure, starting with the Pluronic ex- tract, depends on adsorption to, and elution from, polypro- pylene (Table 2). A !%-fold purification of q5X - RF activi- ty with respect to protein was achieved (Table 2). Electron microxope analysis revealed that the polypropylene eluate contains membrane fragments (data not shown). The polypropylene eluate retained only 4-8% of the LPS of the Pluronic extract (Table 2). Yet, this fraction was from 21 to 60% as active as the extract in promoting phage eclipse. The polypropylene eluate was enriched for phage- eclipsing activity, possessing all the 4X - RF converting ac- tivity of the Pluronic extract (Fig. 2). and distinct from the bulk of the LPS. The low buoyant density (approximately 1.03 g/cm3) of this fraction (Fig. 2) suggests that it has ei- ther a high phospholipid or detergent content. Further fractionation and resolution of the activity in the Pluronic extract should improve our understanding of its components and elucidate the host function in the 4X - RF conversion as well as the role of the gene H (pilot) protein in the process. This work was supported in part by grants from the National In- stitutes of Health and the National Science Foundation. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Sinsheimer, R. L. (1968) Prog. Nucleic Acid Res. Mol. Bbl. 8, 115-169. Newbold, J. E. & Sinsheimer, R. L. (1970) J. Mol. Bbl. 49, 49-66. Jazwinski. S. M., Lindberg, A. A. & Kornberg, A. (1975) Virol- ogy 66,268-282. Francke. B. & Ray, D. S. (1971) Virology 44,168-187. Jazwinski, S. M., Lindberg, A. A. & Kornberg, A. (1975) Virol- Jazwinski, S. M., Marco, R. & Kornberg, A. (1975) Virology, 66,294405. Schekman, R., Weiner. A. & Kornherg, A. (1974) Science 188, Jazwinski, S. M., Marm, R. & Kornberg, A. (1973) Proc. Nat. Ad. Sd. USA 70,205-209. Wickner, W. & Kornberg, A. (1974) J. Bfol. Chem. 249, 624443249. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Bbl. Chem. 193,265-275. Janda, J. & Work, E. (1971) FEBS Lett. 18,343445. Schekman, R., Wickner, W., Westergaard. 0.. Brutlag, D.. Ceider, K., Bertsch, L. L. & Kornberg, A. (1972) Proc. NQ~. Ad. Sei. USA 69,2691-2695. Knippers, R. & Sinsheimer, R. L. (1968) J. Mol. Bwl. 34, 17- 29. Osborn, M. J., Gander, J. E., Parisi, E. & Carson, J. (1972) J. Bwl. Chem. 247,3962-3972. Kornberg, A. (1974) DNA Synthesis (W. H. Freeman and Co., San Francisco, Calif.). ogy 66,283-293. 987-993. 13. 14. 15.