DNase I digestion of chromatin from avian liver nuclei liberates DNA-dependent RNA polymerase II.

In an effort to develop mild conditions for the isolation of DNA-dependent RNA polymerase II, we have used DNase I covalently coupled to Sepharose 4B to digest chromatin from hypotonically lysed nuclei from rooster liver. The RNA polymerase II released was at least 2 times more active in in vitro transcription than was RNA polymerase II prepared by the conventional method of sonication of chromatin in buffers containing high salt. The numbers of RNA polymerase II molecules in RNA polymerase preparations prepared both by DNase I-Sepharose digestion and by sonication were determined by titration with [3H]amanitin and were similar in both preparations, indicating that the two methods were equally efficient at liberating RNA polymerase from the chromatin but that treatment with DNase I-Sepharose yielded higher levels of enzymatic activity. In order to identify RNA polymerase II in complex mixtures of proteins, we have utilized the technique of binding r3H]amanitin to RNA polymerase II followed by electrophoresis on gradient polyacrylamide gels under nondenaturing conditions. We have identified two forms of RNA polymerase II, having molecular weights of 640,000 and 550,000. The release of highly active eukaryotic RNA polymerases by the very gentle treatment of chromatin from lysed nuclei with DNase I-Sepharose may facilitate the reconstitution of an in vitro transcription system using RNA polymerase II.

In an effort to develop mild conditions for the isolation of DNA conditions. We have identified two forms of RNA polymerase II, having molecular weights of 640,000 and 550,000.
The release of highly active eukaryotic RNA polymerases by the very gentle treatment of chromatin from lysed nuclei with DNase I-Sepharose may facilitate the reconstitution of an in vitro transcription system using RNA polymerase II.
The mechanisms involved in DNA-dependent RNA synthesis in eukaryotes are largely unknown. Three major classes of DNA-dependent RNA polymerases have been isolated and purified from various eukaryotic organisms (see reviews by Chambon, 1975;Roeder, 1976), but very little is known about their interaction with DNA and various control elements to initiate specific and accurate transcription of eukaryotic genes. In order to study eukaryotic gene regulation it would be advantageous to develop an in vitro DNA-dependent RNA transcription system capable of proper initiation and faithful RNA synthesis. An essential element in such an in vitro system is an intact RNA polymerase enzyme.
For transcription of the 5 S ribosomal genes, several laboratories have succeeded in reconstituting an in vitro transcription system utilizing eukaryotic RNA polymerase. Recent evidence with Xenopus germinal vesicle extracts has demonstrated correct initiation and termination of cloned 5 S rDNA * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 USC!. Section 1734 solely to indicate this fact. $ Recipient of a postdoctoral fellowship from the American Cancer Society. transcripts as well as the almost complete asymmetry of transcription (Birkenmeier et al., 1978). This indicates that the nuclear extract of Xenopus contained all the components necessary for authentic 5 S ribosomal RNA transcription with RNA polymerase III. However, in studies using RNA polymerase II, the enzyme responsible for messenger RNA synthesis in eukaryotes, attempts to demonstrate specificity of initiation and selective transcription have, thus far, been unsuccessful. An important area for reconsideration is the manner in which the RNA polymerase II has been isolated and purified. Nuclear RNA polymerases actively engaged in DNA-dependent transcription are bound very tightly to the DNA template (see review by Chambon et al., 1974). This tight binding necessitates extreme measures to fragment the chromatin and release the RNA polymerase. The usual method has been a brief sonication of whole cell homogenates or purified nuclei in the presence of high salt (usually 0.3 to 0.5 M ammonium sulfate or potassium chloride) (Roeder and Rutter, 1969;Kedinger et al., 1972; see review by Chambon et al., 1974). However, with enzymes as complex and fragile as the eukaryotic RNA polymerases, such harsh conditions as high salt and sonication may remove loosely bound subunits or cofactors which could have an important role in the template specificity and activity of the polymerase.
Another important consideration in the development of an in vitro transcription system is whether to utilize the most rigorously purified polymerase and run the risk of having lost some aspect of its physiological function or to utilize crude polymerase and be prepared to cope with ambiguities that may result from working with complex mixtures of proteins. Faced with this dilemma, we have chosen to eschew the drastic manipulations required for complete purification of the enzyme because we believe that this would inevitably result in loss of transcriptional fidelity. Instead, we have endeavored to develop the gentlest possible procedures for preparing RNA polymerase that is dependent on exogenous DNA for enzymatic activity. As part of a series of experiments designed to isolate RNA polymerase that is dependent on exogenous DNA for activity and that has not been subjected to sonication or high salts, we prepared DNase I coupled to Sepharose 4B by cyanogen bromide activation of the resin. Part of this work has been presented in preliminary form (Kastern et al., 1978 To increase the sensitivity of the assay, unlabeled DNA was not added. Zsolation of Nuclei-Nuclei were isolated from the livers of 4week-old cockerels by a modification of the technique of Marzluff and Huang (1975). Approximately 15 g of tissue were minced in cold homogenization buffer (0.01 M Tris-HCl (pH 8.0), 0.3 M sucrose, 5 mM magnesium acetate, and 1 mM dithiothreitol). The tissue then was homogenized in 10 volumes (w/v) of homogenization buffer using a Dounce homogenizer with the "A" pestle. After centrifugation at 500 x g for 10 min at O"C, the pellet was resuspended in 4 volumes of homogenization buffer containing 0.1% Triton X-100 and centrifuged at 500 x g for 10 min. The nuclear pellet then was resuspended in 3 volumes of homogenization buffer containing 0.1% Triton X-100 and made 1.9 M in sucrose using 2.3 M sucrose in homogenization buffer. This mixture was overlayed onto 15.ml cushions of 2.0 M sucrose in homogenization buffer in polyallomer tubes (39 ml). After centrifugation for 1 h at 20,000 X g in an SW 27 rotor (Beckman Instruments), the nuclear pellet was resuspended in nuclei buffer (0.05 M Tris-HCl (pH 8.0). 5 mM magnesium acetate, 25% glycerol, and 2 tIIM dithiothreitol).
This nuclear suspension was stored in liquid N, until use. Isolation of RNA Polymerase by DNase-Sepharose Treatment-Frozen nuclei were thawed and subjected to centrifugation for 10 min at 4,000 X g. The nuclear pellet then was resuspended in lysis buffer (0.05 Tris-HCl (pH 8.0), 2 mM dithiothreitol, and 20 ag/ml of PMSF') after which DNase I-Sepharose in lysis buffer was added. Upon completion of the dialysis, the nuclear lysate was stored under liquid NZ. When prepared in this manner, the lysate was stable for at least 6 months under these conditions. Alternatively, acid DNase II-Sepharose was employed to digest the nuclear lysate. This method was essentially the same as for DNase I-Sepharose except that 0.05 M sodium acetate, pH 6.0, was substituted for the Tris in the lysis buffer, and the pH of the lysate was raised to 8.0 with 1.0 M Tris-HCl, pH 8.0, immediately after the DNase II digestion, in order to stabilize the RNA polymerase. Isolation of RNA Polymerase by Sonication in High Salt-The method used was essentially the same as previously reported (Krebs and Chambon, 1976 and Schuell) and rinsed for 30 min in 0.5 M KC1 followed by two rinses in cold 10% trichloroacetic acid containing 20 mM pyrophosphate.
After drying, the radioactivity on the falters was determined by liquid scintillation spectrometry. One unit of enzyme activity incorporated 1 pmol of UTP in 30 min of incubation using the above conditions. Measurements of the Concentrations of DNA and Protein-The concentration of DNA in nuclear suspensions or nuclear lysates was measured using the diphenylamine reaction according to Burton (1956) with modifications as described previously (Leyva and Kelley, 1974) and using commercial calf thymus DNA as a standard. The concentration of protein in various RNA polymerase preparations was determined using a micromodification of the biuret method as described by Zamenhoff (1957 3 mM disodium EDTA. The buffer was continuously recirculated and maintained at 5°C with a Lauda K-2/RD circulating water bath. After electrophoresis, the gels were stained in a solution of 10% methanol, 10% acetic acid, and 0.1% Coomassie brilliant blue R-250.

RESULTS
Pancreatic DNase I was coupled to cyanogen bromide-activated Sepharose 4B according to the method of Cuatrecasas and Anfinsen (1971) and activity was measured by incubation of the immobilized enzyme with "H-labeled DNA. The activity of the DNase Sepharose is shown in Fig. 1. Since it is essential that all of the DNase be removed by pelleting the Sepharose beads, we tested the supernatant fluid for residual DNase activity by incubating it with nick-translated sea urchin ["HIDNA.
Since there was no substantial decrease in acidprecipitable radioactivity (Fig. l), we concluded that there was no appreciable DNase activity remaining in the supernatant fluid after the DNase I-Sepharose was removed.
We then tested the capacity of DNase I-Sepharose to liberate RNA polymerase II from chicken liver nuclei. These preparations were assayed for RNA polymerase II activity using excess denatured calf thymus DNA as template and ["HIUTP as the labeled precursor in an in vitro transcription assay. From the assay of a typical preparation of DNase I-Sepharose-treated nuclear lysate, it is apparent that a transcriptionally active enzyme activity could be obtained (Table  I). To demonstrate that it is the DNase I-Sepharose digestion of chromatin that liberated this RNA polymerase activity, an identical nuclear lysate that had been incubated in the presence of magnesium ions alone (without DNase I-Sepharose)  was tested. The treatment of the nuclear lysate with DNase I-Sepharose resulted in at least a 2-fold increase in the RNA polymerase activity obtained (Table I). Since incubation of the nuclear lysate with EDTA in the absence of DNase I-Sepharose (Table I) did not alter the basal activity, we conclude that this activity is not a function of endogenous nucleases and may represent a pool of unbound, or loosely bound, RNA polymerases. Moreover, it is unlikely that incubation with DNase I-Sepharose yielded higher RNA polymerase activity due to the adsorption of inhibitory substances to the Sepharose resin since incubation with Sepharose 4B alone resulted in no increase in activity over the basal level seen without DNase I-Sepharose (data not shown). A comparison of the kinetics of RNA polymerase release from lysed nuclei revealed that the time of incubation with DNase I-Sepharose was an important factor in the amount of RNA polymerase activity obtained (Table I). There was a basal level of activity obtained without incubation of the nuclear lysate. This activity was relatively constant regardless of the presence or absence of DNase I-Sepharose or other treatments. In the presence of DNase Sepharose, the release of RNA polymerase increased with time and was essentially complete by 30 min of incubation.
We also characterized the salt requirements of the RNA polymerase in our preparations of DNase I-Sepharose-treated, crude, nuclear lysate. The enzyme displayed an optimum of activity at 7 mM manganous ions and an optimum for ammonium sulfate concentration at 80 mM (data not shown). These optima were typical of RNA polymerase II isolated from other organisms (see review by Roeder, 1976).
Eukaryotic RNA polymerase of the three major classes display differing sensitivities or resistance to the fungal toxin a-amanitin (see reviews by Chambon, 1975;Roeder, 1976). We used these criteria, along with their optimal monovalent cation requirements, to assay for each of the three eukaryotic RNA polymerases in our crude nuclear extracts. The amount of RNA polymerase of each of the three types was compared with the amount of polymerase of each type resulting from the conventional method of sonication of an identical amount of nuclei in buffers containing high concentrations of ammonium sulfate.
The results from DNase-treated nuclear lysate (Table II) indicated that greater than 90% of the in vitro transcription activity was due to RNA polymerase II, as seen by its sensitivity to very low amounts of cy-amanitin. There was a low level of polymerase I activity as seen by the insensitivity of RNA synthesis under polymerase I conditions to high concentrations (i.e. 100 yg/ml) of a-amanitin.
In contrast, at least 30% of the in vitro transcriptional activity in the sonicated preparation was accounted for by RNA polymerase I. In addition, the RNA polymerase II activity/mg of nuclear DNA in the DNase I-Sepharose-treated preparation was almost 3fold greater than that obtained by sonication. The low level of incorporation in the presence of high levels (100 pg/ml) of LYamanitin under polymerase III conditions was also insensitive to actinomycin D and continued in the absence of one of the four nucleotide triphosphates (data not shown). Therefore, it was due to an enzyme activity other than an RNA polymerase and has not been further characterized.
Thus, there was no detectable RNA polymerase III in our preparations. From these results, we conclude that the DNase I-Sepharose method of RNA polymerase isolation was almost completely specific for RNA polymerase II from avian liver nuclei.
At the completion of the incubation with DNase I-Sepharose, the remaining chromatin was pelleted along with the DNase resin and other nuclear debris during the low speed centrifugation.
We tested these chromatin fragments for the presence of residual RNA polymerase II activity by removing the DNase I-Sepharose beads and then subjecting the debris to sonication in buffer containing 0.3 M (NH&S04. An RNA polymerase II assay of the partially fractionated sonicate revealed that not more than 10 to 15% of the total RNA polymerase II activity remained in the residual chromatin (data not shown).
A particular difficulty with most conventional methods of RNA polymerase isolation concerns the removal of the fragments of endogenous DNA which may be present in the preparation.
Since most studies of RNA polymerase require the enzyme to be completely dependent upon exogenous DNA for activity, various methods such as ammonium sulfate frac- RNA polymerase was isolated from purified rooster liver nuclei by either the DNase I-Sepharose or sonication methods as described under "Experimental Procedures." The unpurified enzyme preparations were assayed for each of the three RNA polymerase activities using an in vitro transcription assay as outlined under "Experimental Procedures," in the presence of high or low concentration of 01amanitin as indicated. Activity is expressed as the total amount of RNA polymerase activity obtained frbm each preparation. RNA  tionation (e.g. Roeder, 1974) or protamine sulfate precipitation (see Kedinger et al., 1972) are necessary to eliminate the residual DNA. The presence of endogenous DNA was assayed in the DNase I-Sepharose-treated crude nuclear lysate by two independent methods. Both an in vitro transcription assay without exogenous DNA (Table III) and a sensitive assay for the presence of DNA using diphenylamine (data not shown) revealed that there was no endogenous DNA present in the RNA polymerase preparations.
Thus, it was not necessary to include purification methods designed specifically to remove DNA from the RNA polymerase II preparation.
We also tested the ability of another readily available DNase (which we coupled to Sepharose), acid DNase II, to release RNA polymerase II from lysed nuclei. When incubated with ["HIDNA in an assay of DNase activity, DNase II-Sepharose solubilized the ["H]DNA much faster than did DNase I-Sepharose (Fig. l), suggesting that a more efficient release of RNA polymerase II from the nuclear lysate might be possible.
The in vitro transcription activity of the RNA polymerase prepared with DNase II-Sepharose was compared with an identical preparation that had been treated with DNase I-Sepharose. The RNA polymerase activity was essentially the same in both preparations except that DNase I-Sepharose yielded more activity (Table III) even though the chromatin appeared to be more completely digested with the DNase II-Sepharose treatment. The low RNA polymerase activity obtained t.hrough the use of DNase II-Sepharose may reflect a relative instability of RNA polymerase in the acidic conditions necessary for DNase II activity rather than differential release from the chromatin. In all other respects, both DNase I and DNase II yielded identical products. The RNA polymerase preparations had the same template activities with regard to denatured and native DNA, they had the same degree of DNA dependence for transcriptional activity, and they had the same low level of RNA polymerase I activity (Table III). Thus, a different DNase yielded an RNA polymerase preparation which was by several criteria the same as that derived from DNase I treatment.
To further characterize the RNA polymerase II obtained by DNase I-Sepharose digestion of hypotonically lysed chicken liver nuclei, we have subjected it to various procedures including gel filtration and ion exchange chromatography in order to effect some purification.
For our purposes, a useful first step in purification was the precipitation of the enzyme by bringing the solution to 70% saturation with ammonium sulfate. The major advantage of this step was to concentrate Equal amounts of purified liver nuclei were lysed and incubated with either DNase I-Sepharose or DNase II-Sepharose as described under "Experimental Procedures." The RNA polymerase activity in each lysate was assayed using an in vitro transcription assay under either RNA polymerase I or II conditions as under "Experimental Procedures," using either native or denatured calf thymus DNA. RNA polymerase activity is expressed as the total amount of RNA polvmerase activity obtained from each preparation. II eluted in a broad peak within the included volume of the column (Fig. 2). The large amount of UV-absorbing material that eluted in the void volume was mostly RNA, and RNA was undetectable in the RNA polymerase fraction. There was a substantial loss of total protein during this step of purification, and this was reflected in the approximately 95% increase in specific activity over the unpurified nuclear lysate (  l ---0. The fractions between the arrows were pooled for further analysis. Either wheat &rm RNA polymerase II or E. coli RNA polymerase (holoenzyme) were incubated at 37'C with an excess of r"Hlamanitin. The mixtures then were applied to a 2 to 16% gradient polyacrylamide slab gel and subjected to electrophoresis under nondenaturing conditions as described under "Experimental Procedures." One track of each enzyme type was stained for protein with Coomassie brilliant blue, and the other track of each enzyme was cut into 2-mm slices. The slices were solubilized and the radioactivity in each slice was determined by liquid scintillation. The arrows indicate the positions of the stained protein bands. JH radioactivity migrated with wheat germ RNA polymerase II (C--O) or with E. coli RNA polymerase (U).
the addition of the ["Hlamanitin, the unlabeled a-amanitin effectively competed with the ["Hlamanitin, and no radioactivity was recovered in the void volume of the column (data not shown).
We have used the technique of ["Hlamanitin binding to follow the recovery of RNA polymerase II through the steps of purification described above. A comparison of the results obtained by in vitro transcription assays with those obtained through ["Hlamanitin titration (Table IV) demonstrated that the values obtained through these two independent techniques were in close agreement. In the case of DEAE-Sephadex chromatography, there was an apparent 4.8-fold purification with an 18% yield as seen by transcriptional activity and a 4.7-fold purification with an 18% yield calculated using ["Hlamanitin binding. Indeed, for all stages of purification, the results of ["Hlamanitin binding were in close agreement with those obtained by in vitro transcription, thus verifying and supporting the transcriptional results. However, a discrepancy between the two assays was observed when material prepared from an equal number of nuclei (as determined by the DNA content) by sonication in buffer containing high salt (Krebs and Chambon, 1976) was compared with that obtained through the DNase I-Sepharose method. The preparation of enzyme isolated by sonication had an equal or slightly greater capacity for binding [:'H]amanitin than did the crude extract prepared by the DNase I-Sepharose method, but it had a significantly decreased ability to transcribe in the in vitro assay (Table IV). One possible explanation for this result is that sonication damages a large number of polymerase molecules in such a way as to inactivate their enzymatic activity but not their ability to bind amanitin.
Many studies have examined the multiple forms of eukaryotic RNA polymerase II following elect,rophoresis of the polymerase on polyacrylamide gels under nondenaturing conditions (see review by Roeder, 1976). However, since it is difficult to recover and assay RNA polymerase activity from acrylamide gels Roeder, 1976) With the ability to form stable complexes with ["Hlamanitin, the RNA polymerase II could be subjected to electrophoresis on 2 to 16% gradient polyacrylamide gels under nondenaturing conditions and subsequently located by the radioactive label. When commercially prepared purified wheat germ RNA polymerase II was treated in this manner, the profile of radioactivity shown in Fig. 4 was obtained. Since wheat germ RNA polymerase can form a dimer under the conditions of nondenaturing electrophoresis, two peaks of radioactivity are seen. RNA polymerase from E. coli, which is not affected by and does not bind cu-amanitin, did not bind radioactive amanitin (Fig. 4).
The 2 to 16% gradient polyacrylamide slab gels had several advantages for the study of RNA polymerase II, including the ability to visualize very small amounts of protein after electrophoresis since the proteins in the mixture formed discrete protein bands as a function of molecular weight when the time of electrophoresis took place over extended periods of time (e.g. 15 h). Thus, we could detect very small amounts of protein (approximately 10 to 50 rig/band). Moreover, the extended time of electrophoresis combined with the shallow gradient of acrylamide concentrations in the gel had the advantage of eliminating all proteins with a molecular weight less than approximately 100,000. Thus, it was possible to detect RNA polymerase in even the most crude preparations with a very low protein background in the gel.
When ["Hlamanitin was bound to avian liver RNA polymerase II and the resulting complex was subjected to electrophoresis under nondenaturing conditions, the profile of radioactivity from the sliced gel revealed two peaks of radioactivity (Fig. 5). These peaks of radioactivity corresponded to two stained protein bands with molecular weights of 640,000 and 550,000, respectively. When a similar procedure was followed utilizing a 4 to 30% gradient acrylamide slab gel instead of a 2 to 16% gradient as described above, the peaks of radioactivity still migrated with the same two protein bands even though their positions had changed substantially (data not shown). Although analysis of the nuclear lysates on a 4 to 30% gradient acrylamide gel resulted in the retention of several low molecular weight proteins (i.e. 50,000 to lOO,OOO), there were no new ["Hlamanitin-binding bands. Complexes between ["HIamanitin and unpurified RNA polymerase II from avian liver were formed by incubation at 37°C. The mixture was applied to a 2 to 16% gradient polyacrylamide slab gel and subjected to electrophoresis under nondenaturing conditions as described under "Experimental Procedures." One track was removed and cut into 2-mm slices. The slices were solubilized and the radioactivity in each was determined by liquid scintillation counting. Parallel track from the gel was stained for protein with Coomassie brilliant blue and a photograph of the stained gel appears above the profile of radioactivity properly oriented. The arrows indicate the putative RNA polymerase II bands with their respective molecular weights as determined by comparison of their migration with molecular weight standards as described under "Experimental Procedures." Isolation of RNA Polymer-use II by DNase I-Sepharose Digestion Avian RNA polymerase II, in a total volume of 10 ml in DC buffer was applied to a column (0.2 X 2.0 cm) of DEAE-cellulose at a flow rate of 4 ml/h at 4°C. The enzyme activity was eluted from the column with a 1% solution of CM-dextran in DC buffer. Fractions of 0.06 ml were collected and the RNA polymerase II activity in each fraction was determined using the LJZ vitro transcription assay. The activity represents the total acid-precipitable incorporation of [3H]UMP in a lo-4 aliquot of each fraction. The CM-dextran eluted in Fraction 6.
In some of the isolation and purification steps such as Sepharose GB-CL gel filtration, the RNA polymerase II enzyme was present in a rather dilute mixture of proteins. This made gel electrophoretic analysis of the polymerase under nondenaturing conditions difficult. Moreover, conventional concentration procedures, such as ammonium sulfate precipitation or ion exchange chromatography, may have resulted in inactivation or loss of the enzyme, or both. To avoid this difficulty, we have concentrated dilute mixtures of RNA polymerase II by displacement chromatography.L The technique permitted stabilization of the RNA polymerase activity by increasing the protein concentration without the addition of exogenous protein.
A lOO-~1 column of DEAE-cellulose was loaded with a dilute mixture of RNA polymerase II obtained by Sepharose GB-CL gel filtration. The protein was eluted with a 1% solution of carboxymethylated dextran (CM-dextran). CM-dextran had a suffkient charge density to displace all proteins from the DEAE-cellulose, resulting in their eluting in a very small volume just ahead of the CM-dextran front. By collecting fractions of small volume, RNA polymerase preparations that were free from the CM-dextran could be obtained. A typical profile of transcriptional activity obtained by this procedure (Fig. 6)  Samples of avian RNA polymerase II were either concentrated using displacement chromatography or were applied directly to a 2 to 16% gradient polyacrylamide slab gel and subjected to electrophoresis under nondenaturing conditions as described under "Experimental Procedures." The gel then was stained for protein with Coomassie brilliant blue as described. The molecular weight markers consist of: thyroglobulin, 669,000; ferritin, 44O,ooO, catalase, 232,ooO, and lactate dehydrogenase, 140,000. The arrows indicate the putative bands of avian liver RNA polymerase II. The material applied to each track was A, dialyzed supernatant fluid resulting from DNase I-Sepharose treatment of hypotonically lysed nuclei; B, dialyzed supernatant fluid resulting from DNase II-Sepharose 4B digestion of hypotonically lysed nuclei; C, RNA polymerase preparation obtained by 70% saturated ammonium sulfate precipitation of sonicated nuclei; D, same as in A; E, 70% saturated ammonium sulfate precipitate of the material in D; F, RNA polymerase II concentrated by displacement chromatography; G, pooled fractions containing the peak of RNA polymerase activity eluted from the Sepharose GB-CL column; and H, molecular weight marker proteins.
ious stages of purification were applied to the displacement column (data not shown).
Using CM-dextran displacement chromatography to concentrate dilute RNA polymerase II preparations from Sepharose GB-CL and DEAE-Sephadex chromatography, we have been able to obtain samples of RNA polymerase II from each step of purification that are sufficiently concentrated for analysis on gradient acrylamide slab gel electrophoresis under nondenaturing conditions. A typical profile of the stained 2 to 16% gradient acrylamide slab gel (Fig. 7) demonstrated that there were no gross differences in the molecular weights of the two forms of RNA polymerase II throughout all stages of purification.
Thus, at least at this level of resolution, the RNA polymerase II molecule appeared to remain intact throughout the purification steps we have employed. DISCUSSION We have developed a procedure that utilizes the observation that DNase I preferentially digests actively transcribing regions of chromatin (Weintraub and Groudine, 1976). Since RNA polymerase is expected to be associated with actively transcribed genes, digestion of the chromatin with DNase I appeared to be a method for specifically releasing actively transcribing RNA polymerase. Treatment with soluble DNase I had been employed as a means for isolating DNA-dependent RNA polymerase from E. coli (Berg et al., 1971). However,