2P]dGTP (410 Ci/mmol) were purchased from Amersham. Nucleoside triphosphates, sarkosyl (N-laurylsarcosine, sodium

Transcription

1 Proc. Nati. Acad. Sci. USA Vol. 81, pp , October 1984 Biochemistry Isolation of an active transcription initiation complex from HeLa cell-free extract (RNA polymerase 11/adenovirus 2 major late promoter/initiation of transcription/elongation) H. ESER TOLUNAY, LINDA YANG, W. FRENCH ANDERSON, AND BRIAN SAFER Section on RNA and Protein Biosynthesis, Laboratory of Molecular Hematology, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD Communicated by Bernhard Witkop, May 21, 1984 ABSTRACT A two-step procedure has been developed for the formation of RNA polymerase H transcription initiation and elongation complexes. Initiation complexes are rapidly formed in HeLa cell-free extract supplemented with a DNA template containing the adenovirus 2 major late promoter and ATP. Assembly of transcription components required for correct initiation is absolutely dependent on specific eukaryotic promoter sequences. Sarkosyl-sensitive transcription initiation complexes are rapidly converted to Sarkosyl-resistant elongation complexes when supplemented with the remaining nucleoside triphosphates. The 60S initiation complex can be extensively purified by glycerol gradient centrifugation and is easily separated from free RNA polymerase II and free DNA template. Recovery of this stable complex is >90%. Specific transcription cannot be detected if the DNA template is subsequently added to gradient fractions containing HeLa cell-free extract components alone. This suggests that the DNA templates promote the specific assembly of RNA polymerase II and transcription factors required for accurate initiation. Since conversion of purified initiation complexes to elongation complexes can occur without additional HeLa cell components, the presence of transcription components required for initiation and elongation in a single complex is indicated. Initiation of mrna biosynthesis is thought to be mediated by a sequence-specific macromolecular assembly of transcription initiation factors and RNA polymerase II on the DNA template to form an initiation complex capable of specifying correctly the first phosphodiester bond. Several laboratories have attempted to purify these factors by classical chromatographic fractionation procedures to identify their specific function(s) (1-5). This method depends on reconstitution of the original transcriptional activity upon recombination of the purified factors. A major problem with this approach has been the substantial loss of transcriptional activity with increasing fractionation. In addition, there is a lack of specific assays to measure the partial functions of the transcriptional components. A second approach to identify the transcription factors involved in initiation might be to isolate active transcription complexes at different stages of complex assembly. The polypeptide components distinct from RNA polymerase II would then be identified by polyacrylamide gel electrophoresis, and their possible functions would be determined. This would permit their initial purification on the basis of either their physical characteristics or their specific partial functions rather than the formation of runoff RNA, which would require the presence of all transcription initiation factors. In this study we report the development of procedures for the formation and isolation of distinct transcription initiation and elongation complexes. This approach utilizes the differential requirement for ribonucleoside tri- The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C solely to indicate this fact. phosphates during initiation complex assembly and the subsequent conversion of these complexes to elongation complexes. Separation of these macromolecular transcription complexes from free HeLa cell-free extract components can then be achieved by glycerol density gradient centrifugation. EXPERIMENTAL PROCEDURES ta-32p]gtp (410 Ci/mmol; 1 Ci = 3.7 x 1010 Bq) and [a- 2P]dGTP (410 Ci/mmol) were purchased from Amersham. Nucleoside triphosphates, sarkosyl (N-laurylsarcosine, sodium salt), and a-amanitin were from Sigma, and 5'-adenylyl imidodiphosphate was obtained from P-L Biochemicals. Plasmid psmaf, containing adenovirus 2 (Ad2) major late promoter (map coordinates 11-18) at the Sma I site of pbr313, has been previously described (6). HeLa cell-free extracts, prepared as described by Manley et al. (7) were obtained from Bethes'da Research Laboratories. Extracts were dialyzed against 20 mm Tris HCl, ph 7.9/6 mm MgCl2/40 mm (NH4)2SO4/0.2 mm EDTA/2 mm dithiothreitol (buffer A) prior to use. Transcription initiation complex formation was performed at 30 C for 10 min in 25,l of incubation mixture containing 12,l of HeLa cell-free extract (155,ug of protein), 40 mm (NH4)2SO4, 10 mm Tris (ph 7.9), 7.5 mm MgCl2, 2 mm dithiothreitol, 5 mm creatine phosphate, 0.4 units of creatine phosphokinase, 9% (vol/vol) glycerol, 100,uM ATP, and 2 gg of Sma I/Xho I-digested psmaf. Preformed complexes were elongated by the addition of 50,uM ATP, CTP, and UTP; 10,M GTP; and 5-10,Ci of [a-32p]gtp. To prevent new initiation, elongation was performed in the presence of 1.7 mm Mn2' at a final (NH4)2SO4 concentration of 100 mm at 30 C for the indicated length of time. When coupled transcription initiation and elongation activities were monitored, incubation was done under initiation complex formation conditions for 45 min, except that all of the nucleotides were present from the beginning. This allowed both initiation and elongation to proceed simultaneously. Specific RNA transcription was monitored by autoradiography after PAGE of the products as described (6). Sedimentation of transcription complexes was performed on linear 15-40% glycerol density gradients prepared with buffer A. Incubations (200 Al) were loaded on 3.6-ml SW-60 gradients. Centrifugation was for 90 min at 59,000 rpm at 4 C. Gradients were fractionated from the top by displacement using an ISCO apparatus. Gradient fractions were assayed under elongation conditions. Transcription initiation complexes isolated in this way were stable for up to 24 hr at O C. The distribution of RNA polymerase II was determined by the procedure of Hodo and Blatti (8) with calf thymus DNA in the presence and absence of 1 pg of a-amanitin per ml. Autoradiographs were quantitated by densitometry at OD595 using a Beckman DU-8B Spectrophotometer. Abbreviations: Ad2, adenovirus 2; nt, nucleotide; kb, kilobase. 5916

2 RESULTS Formation of Transcription Initiation and Elongation Complexes. Optimization of the monovalent cation concentration in cell-free HeLa extracts for the correct initiation at the Ad2 major late promoter yields a value significantly lower (40 mm) than was found using sequence-independent transcription with denatured calf thymus DNA (100 mm). Since this difference could reflect the lack of specific initiation complex assembly on the latter, the effect of (NH4)2SO4 concentration on transcriptional initiation and elongation of the Ad2 major late promoter was studied. To separate initiation from elongation, exogenous nucleotides required for elongation were omitted from the initial incubation. Formation of the specific runoff transcript from the DNA template psmaf/ Sma I/Xho I [536 nucleotides (nt)] was optimal at 40 mm (NH4)2SO4 and was severely inhibited at concentrations greater than 50 mm (Fig. 1A). When initiation was first performed at 40 mm (NH4)2SO4, elongation of the complex was maximal at 60 mm (NH4)2SO4 (Fig. 1B). The densitometric scans ofa and B show that elongation of preformed initiation complexes could occur at salt concentrations that totally prevent formation of new initiation complexes (Fig. 1C). This allowed the relative amounts of transcription initiation and elongation complexes formed at 40 mm (NH4)2SO4 to be evaluated because, at concentrations greater than 60 mm, initiation did not occur. To determine whether transcription complexes formed at 40 mm (NH4)2SO4 are primarily initiation or elongation complexes, the sensitivity of specific transcription to the ionic detergent sarkosyl was examined. It has been shown previously that elongation of preinitiated RNA chains is insensitive to inhibition by sarkosyl, while transcription complexes formed prior to elongation are disrupted (9, 10). The effect of sarkosyl addition to transcription complexes formed at 40 mm (NH4)2SO4 in the presence of ATP before (designated A '"rn Biochemistry: Tolunay et al B C z Proc. NatL Acad. Sci. USA 81 (1984) 5917 initiation) and after the addition of nucleotides and the adjustment of (NH4)2SO4 to 100 mm (designated elongation) is shown in Fig. 2. For periods of incubation up to 10 min at 40 mm (NH4)2SO4, formation of the specific 536-nt transcript was totally inhibited by sarkosyl (initiation panel). Once elongation was permitted by the addition of nucleotides under salt conditions where new initiation was prevented, sensitivity to sarkosyl was greatly reduced (elongation panel). It appears, therefore, that the complexes formed at 40 mm (NH4)2SO4 are initiation complexes. The relative rates of initiation and elongation complex formation at 40 mm (NH4)2SO4 were examined (Fig. 3). HeLa cell extracts were incubated with Sma I/Xho I-digested psmaf and ATP at 30TC for the indicated times and then were supplemented with the remaining nucleotides and (NH4)2SO4 in control buffer or buffer containing 2% sarkosyl. Both initiation and elongation complexes formed prior to the addition of elongation buffer were allowed to produce the 536-nt runoff transcript (Fig. 3 Left). Runoff transcription of initiation complexes was prevented by sarkosyl, and only runoff transcription of previously converted elongation complexes was seen (Fig. 3 Right). These data indicate that, during incubation at 40 mm (NH4)2SO4, the rate of initiation complex formation is much greater than the rate of its conversion to the sarkosyl-resistant elongation complexes. As the (NH4)2SO4 concentration is raised, initiation complex assembly becomes inhibited, while its conversion to elongation complexes is affected much less. Therefore, by choosing the appropriate ionic conditions and length of incubation and controlling the availability of nucleotides, the transcription complexes formed in cell-free extracts can be primarily initiation or elongation complexes. Isolation of Transcription Initiation Complexes by Glycerol Gradient Centrifugation. Previous attempts to characterize the composition of transcription complexes after isolation of stable complexes by agarose gel electrophoresis have been Initiation + -t I", Elongation + t ""Vw c: '9f SSSN"Wo ill* (ft i.:. ~*,0 X CI INH,)2S0,, FIG. 1. Effect of (NH4)2SO4 concentration on transcription complex formation and elongation. (A) HeLa cell-free transcription extract and psmaf digested with Sma I/Xho I were incubated under the conditions for transcription initiation complex formation for 10 min at the indicated concentrations of (NH4)2SO4. The concentration of (NH4)2SO4 was then raised to 100 mm; ATP, UTP, and CTP at 50,uM, GTP at 10,uM, and 10,uCi of [a-32p]gtp were added, and incubation at 30 C was continued for an additional 30 min. Autoradiograms following 4% PAGE are shown. The lower band corresponds to the correct 536-nt transcript; the upper band is the end-toend transcripts of the Ad2 insert and pbr313. (B) The incubation conditions were the same as in A except that the initial incubation was done at constant (NH4)2SO4 concentration (40 mm); 25-,1 aliquots were then adjusted to the indicated concentration of (NH4)2SO4 and supplemented with UTP, ATP, and CTP at 50 AM, GTP at 10 1LM, and 10,uCi of [a-32p]gtp. Incubation at 30 C was continued for an additional 30 min. (C) A densitometric scan of specific transcript in A (0) and B (o) were performed at OD595. mm TIME OF ADDITION (Minutes) FIG. 2. Effect of 2% Sarkosyl on initiation complex formation and elongation. (A) Aliquots (50 p1) of the mixture of HeLa cell-free extract and Sma I/Xho I-digested psmaf were transferred from a common incubation to tubes containing 2% sarkosyl (lanes +) or H20 (lanes -) after the indicated number of minutes at 30 C. The initial incubation was performed with 40 mm (NH4)2SO4 and 100,uM ATP. The total incubation time under these conditions was 30 min. At 30 min, the transcription assays were supplemented with ATP, UTP, and CTP at 50 p.m, GTP at 10,uM, 10,uCi of [a- 32P]GTP, and (NH4)2SO4 to a final concentration of 100 mm for elongation, and incubation was continued for 30 min. (B) HeLa cellfree extract and Sma I/Xho I double-digested psmaf, as in A, were incubated for 30 min under the initiation complex formation conditions. Sarkosyl (2%) (+) or water (-) were added at indicated times after the reaction conditions were optimized for elongation. Total incubation under elongation conditions was 30 min. The lower band is the 536-nt transcript.

3 5918 Biochemistry: Tolunay et al Control ::. ".!. 2% Sarkosyl. IL.9.SS *R* MINUTES 536 FIG. 3. Kinetics of initiation complex formation and conversion to sarkosyl-resistant elongation complexes. HeLa cell-free extract and psmaf/sma I/Xho I, incubated under conditions optimal for transcription initiation for the indicated times, were then supplemented with the remaining nucleoside triphosphates and (NH4)2SO4 (final concentration 100 mm) either in the absence (Left) or presence (Right) of 2% sarkosyl, and the incubation was continued at 300C for 30 min. The 536-nt transcript is indicated. limited by the very small amounts of complexes obtained by this procedure (11). One approach more amenable to largescale preparative isolation might be to isolate such transcription complexes by sedimentation on glycerol gradients. The reported sedimentation coefficient (S20,w) of RNA polymerase II, 15S (12), would favor sedimentation away from the majority of HeLa cell extract components, particularly when associated with template and transcription factors. To evaluate the functional integrity of complexes isolated in this manner, transcription initiation complexes prepared as described above were sedimented on linear 15-40% glycerol gradients. Gradients were prepared at ionic conditions optimal for transcription initiation so that aliquots could be assayed directly. Transcription initiation complexes isolated by glycerol gradient centrifugation were stable and subsequently could be elongated to produce the correct 536-nt transcript (Fig. 4); 90-95% of the activity loaded on the gradient could be recovered. Most HeLa cell-free extract components sedimented at <30S. A single peak was seen, however, sedimenting at 45S (Fig. 4A). When gradient fractions were incubated with nucleotides at 100 mm (NH4)2SO4, the specific 536-nt runoff transcript was produced by fractions sedimenting at -60S (Fig. 4B). Formation of the correct transcript was totally inhibited by a-amanitin at 1 gg/ml (data not shown). An initiation complex containing all components required for its subsequent conversion to an elongation complex could be resolved from the majority of HeLa transcription extract components, DNA template, and free RNA polymerase II (Fig. 4 A-C); >95% of RNA polymerase II was free and not associated with the transcription initiation complex (Fig. 4C). This may indicate the presence of another limiting component for initiation complex assembly and reflect the generally low efficiency of transcriptional activity found in these extracts in comparison to in vivo rates of transcription. The high sedimentation value of this complex (60S) was unexpected because the protein and nucleic acid components in an elongation complex containing only RNA polymerase II (1SS) and the 0.78-kilobase (kb) (lls) template should be less than 30S. It is likely, therefore, that other transcriptional components may be present in these initiation complexes. Formation of the transcription initiation complexes isolat- Proc. NatL Acad Sci. USA 81 (1984) ed by glycerol gradient centrifugation required ATP and DNA template to be present during complex formation (data not shown). A requirement for ATP to transcribe class II genes correctly in HeLa cell extracts has been shown previously (13). In this study a role for ATP during initiation complex assembly prior to its conversion to a complex capable of elongation was indicated. Similar to the earlier results (13), the nonhydrolyzable analog 5'-adenylyl imidodiphosphate would not support initiation complex assembly (data not shown). No specific transcriptional activity was observed in gradient fractions when the DNA template was left out during incubation prior to glycerol gradient centrifugation. Addition of exogenous DNA template to gradient fractions after centrifugation did not result in specific transcription of the template (data not shown). This indicates that transcription initiation complex formation requires the assembly of DNA template, RNA polymerase II, and other transcription initiation factors. The RNA polymerase II and transcription initiation factors do not exist as an active complex prior to correct template insertion. The relative distribution of free and complexed DNA before and after incubation with HeLa cell extract was examined. The 0.78-kb fragment of psmaf, containing the specific initiation sequence for the 536-nt runoff transcript, sedimented at -11S (Fig. 4A, lower trace) in comparison with a value of 60S when part of the transcription initiation complex. When incubated with extract, however, a direct evaluation of its distribution in extract by its A254 was prevented by the large absorbance peak at S, which was present in extract independent of DNA addition. Distribution of the DNA template could be evaluated, however, by [a- 32P]dGTP labeling of the 0.78-kb fragment (and other DNA fragments of the psmaf/sma I/Xho I double digest) by components of the HeLa cell extract (shown in Fig. 4D Right). While the mechanism of labeling is presently unknown, this allowed the distribution of DNA to be determined. After incubation under transcription initiation conditions, the distribution of the 0.78-kb template was shifted towards the bottom of the gradient (Fig. 4D; the transcription initiation complex was located in fractions 10-15). The amount of larger DNA fragments loaded on the gradients was reduced roughly in proportion to their size by low-speed centrifugation prior to glycerol gradient analysis. In agreement with the low efficiency of in vitro eukaryotic transcription systems, only a small fraction of the DNA template that sedimented faster was engaged in the formation of active transcription initiation complexes. Resolution of the transcription initiation complex on SW-60 glycerol gradients was partially obscured by the large polypeptide complex in the 40-60S region of the gradient, and specific transcriptional components present only in the complex were difficult to identify, but a substantial purification was achieved (Fig. 4E). Further purification of this transcription initiation complex will be reported elsewhere. The apparent stimulation of transcription by sarkosyl reported by Ackerman et al. (11), attributed to release of the ternary complex from the visible precipitate commonly observed during in vitro transcription, appears to be the result of using large DNA templates rather than the shorter 0.78-kb fragment used in these studies. In general, large differences in transcription efficiency have not been noted when transcription is performed with all DNA fragments of the restricted plasmid template present or with the purified DNA template alone. This would indicate that plasmid DNA not containing specific initiation sequences does not deplete the cell-free extract of limiting transcription components. While trapping of the ternary elongation complex in the precipitate may occur under the lower ionic conditions required for agarose gel electrophoresis, this does not seem to occur under the conditions used in this study.

4 t 0 t t - ~~~~~~~~~~~~~536 Biochemistry: Tolunay et at A. A254 B. TRANSCRIPTION ASSAY C. RNA POLYMERASE II ACTIVITY S 60S 80S T 600X kb 4 Proc. Natl. Acadt Sci. USA 81 (1984) 5919 D. DNA DISTRIBUTION IN THE COMPLEX E. PROTEIN DISTRIBUTION FIG.4.Isolation;of atrn i n c T int wr fmd a s t24xf~~~~~~ _ Zp-l- 24o 97t ant l > 290- a c e Ah iv29.0 ace ~~~~~~~~~24.0- O I In_~~~~~~~~~~~~~~ _ FIG. 4. Isolation of a transcription initiation complex by glycerol gradient centrifugation. Transcription initiation complexes were formed in HeLa cell-free extract on the Ad2 major late promoter under conditions optimal for transcription initiation complex formation; 200 A.l was centrifuged on 15-40o glycerol gradient for 90 min at 59,000 rpm in a Beckman SW-60 rotor at 40C. Gradient fractions were assayed for elongation of preformed complexes to generate the specific 536-nt transcript. Incubations were for 60 min at 300C. Transcription complexes were not detected when the DNA template was omitted during the initial incubation, but later was included during the transcription assay. (A) The A254 profile of HeLa cell-free extract on the 15-40%o glycerol gradient (upper trace) and Sma I/Xho I-digested psmaf alone (lower trace). Ribosomal particles were used in a separate gradient to determine the S20, of the initiation complex. (B) Transcription of gradient fractions identifies an active, a-amanitin-sensitive, transcription initiation complex in the 60S region of the gradient. (C) Each gradient fraction (50 Al) was assayed for RNA polymerase activity in the presence and absence of a-amanitin at 1 pg/ml. The difference was plotted as RNA polymerase II activity. (D) Distribution of the DNA on the gradient after incubation with HeLa cell-free extract. (Right) The single lane is an autoradiogram before the centrifugation of psmaf DNA fragments labeled by [a-32p]dgtp. The 2.4-kb band results from incomplete digestion of the Sma I insert by Xho I during initiation complex formation. DNA distribution was monitored by autoradiography after PAGE on 10% gel. Transcription initiation complex was found between fractions 10 and 15. (E) Protein distribution of the transcription initiation complex. Gradient fractions from A were analyzed by NaDodSO4/PAGE. The Coomassie blue stain of polypeptides present in the glycerol gradient fractions is shown. Transcription initiation complexes sediment between fractions 6 and 10. DISCUSSION and DNA template containing specific sequence and/or structural information (1-5). Several laboratories have re- Transcription initiation complex formation requires at least ported progress in the fractionation of cell-free extracts cathree transcription factors in addition to RNA polymerase II pable of specific class II gene transcription (1-5). In general,

5 5920 Biochemistry: Tolunay et al however, difficulties resulting from the low initial activities of these cell-free extracts and the further loss of activity during fractionation, have limited progress in identifying the specific polypeptides involved or their functions. An alternate approach is to isolate specific intermediate steps of the initiation process, with the goal of identifying the specific proteins found in the complexes and determining their functions. This would allow their initial identification and purification to be made without having to maintain transcriptional activity of the entire system. Similar approaches have been used previously to isolate protein factors required for translation (14). A general method of obtaining such complexes is to employ inhibitors that result in the specific accumulation of an intermediate complex. Alternatively, omission of a component necessary for transcription might allow assembly up to, but not beyond, the point where the component is required. The approach we have used to isolate transcription initiation complexes is to arrest the transcription process by not providing the nucleoside triphosphates required for the formation of the first phosphodiester bond. Incubation of the DNA template with HeLa cell-free extract and ATP only allows transcription initiation complexes to form but prevents their conversion to elongation complexes. Although ATP is the first nucleotide specified by the Ad2 major late promoter, an additional requirement for a hydrolyzable 83-y bond has been identified previously (13). Although the function of such ATP hydrolysis is still unknown, our studies show that it is required during assembly of the initiation complex prior to elongation complex conversion. Transcription initiation complexes can be converted rapidly to elongation complexes containing nascent RNA transcripts by providing the remaining nucleoside triphosphates. Transcription initiation and elongation complexes were distinguished by their differential sensitivity to the ionic detergent sarkosyl. Sarkosyl has previously been shown to prevent initiation of RNA polymerase II transcription, while allowing elongation of nascent RNA transcripts (9, 10). Initiation complexes formed in vitro were totally sensitive to sarkosyl. Upon their conversion to elongation complexes, the correct runoff transcript was formed in the presence of sarkosyl (Figs. 2 and 3). The relative abundance of initiation and elongation complexes also could be manipulated by adjusting the (NH4)2SO4 concentration (Fig. 1): 40 mm (NH4)2SO4 was determined to be optimal for the formation of sarkosyl-sensitive transcription initiation complexes, whereas sarkosyl-resistant elongation was maximal in the presence of Mn2+ and at higher (NH4)2SO4 concentrations, where reinitiation was completely inhibited (Fig. 1C). Transcription initiation complexes formed in the presence of ATP at 40 mm (NH4)2SO4 with Ad2 major late promoter, RNA polymerase II, and the transcription factors present in HeLa cell-free extracts can be purified on 15-40% glycerol gradients. The recovery of the initiation complex from glycerol gradient is 90-95% as determined by the formation of correct 536-nt runoff transcript upon incubation of the isolated complex with nucleoside triphosphates. No specific transcription can be detected in gradient fractions assayed by Proc. NatL Acad. Sd USA 81 (1984) subsequent addition of DNA template under either initiation [40 mm (NH4)2SO4] or elongation [100 mm (NH4)2SO4] conditions, if DNA template is left out during incubation prior to glycerol gradient centrifugation. This indicates that in HeLa cell-free extracts, RNA polymerase II and the transcription initiation factors do not preexist as a complex. Therefore, the presence of DNA template is required for the assembly of the transcription initiation complex. Recently, Ackerman et al. have reported the formation and isolation of DNA containing the Ad2 major late promoter in a stable complex with RNA polymerase II and nascent RNA transcripts (11). Several features of these ternary complexes indicate that they are elongation complexes that are distinct from the transcription initiation complexes isolated by glycerol gradient centrifugation. In the presence of all nucleoside triphosphates, complexes isolated by glycerol gradient centrifugation are rapidly converted to sarkosyl-resistant elongation complexes. The ternary complexes isolated by Ackerman et al. in the presence of sarkosyl are, therefore, more closely related to the latter. The mobility of the ternary elongation complexes during agarose gel electrophoresis are identical to that of the free DNA template (11). However, transcription initiation complexes isolated by glycerol gradient centrifugation sediment as 60S particles that can be separated easily from the 11S DNA template and free 15S RNA polymerase II. This suggests that other transcription factor components may be present in the transcription initiation complex. Although the function of these factors during initiation complex assembly is presently unknown, the formation of such a stable complex may indicate a multienzyme transcriptional complex that participates not only in initiation but also in elongation and processing of the nascent transcripts. 1. Matsui, T., Segall, J., Weil, P. A. & Roeder, R. G. (1980) J. Biol. Chem. 255, Samuels, M., Fire, A. & Sharp, P. A. (1982) J. Biol. Chem. 257, Dynan, W. S. & Tijan, R. (1983) Cell 32, Davison, B. L., Egly, J.-M., Mulvihill, E. R. & Chambon, P. (1983) Nature (London) 301, Dynan, W. S. & Tijan, R. (1983) Cell 35, Tolunay, H. E., Yang, L., Kemper, W. M., Safer, B. & Anderson, W. F. (1984) Mol. Cell. Biol. 4, Manley, J. L., Fire, A., Sharp, P. A. & Gefter, M. L. (1980) Proc. Natl. Acad. Sci. USA 77, Hodo, H. G. & Blatti, S. P. (1977) Biochemistry 16, Gariglio, P., Buss, J. & Green, M. H. (1974) FEBS Lett. 44, Shmookler, R. J., Buss, J. & Green, M. H. (1974) Virology 57, Ackerman, S., Bunick, D., Zandomeni, R. & Weinmann, R. (1983) Nucleic Acids Res. 11, Kedinger, C., Gissinger, F. & Chambon, P. (1974) Eur. J. Biochem. 44, Bunick, D., Zandomeni, R., Ackerman, S. & Weinmann, R. (1982) Cell 29, Safer, B., Jagus, R. & Kemper, W. M. (1979) Methods Enzymol. 60,