Purification and Characterization of a GTP-Pyrophosphate Exchange Activity from Vaccinia Virions ASSOCIATION OF THE GTP-PYROPHOSPHATE EXCHANGE ACTIVITY WITH VACCINIA mRNA GUANYLYLTRANSFERASE (GUANINE-7-)METHYLTRANSFERASE (CAPPING ENZYME)

A core-associated enzyme, which catalyzes a nucleo- tide-pyrophosphate exchange with GTP, has been purified from vaccinia virions. The enzyme requires MgC12 for activity, has an alkaline pH optimum, and specifically utilizes GTP as the exchanging nucleotide. The enzyme does not catalyze exchange of GMP with GTP. The GTP-PPi exchange enzyme co-purifies with vaccinia capping enzyme (RNA guanylyltransferase and RNA (guanine-7-)methyltransferase) through succes-sive chromatography steps on DEAE-cellulose, DNA- cellulose, and phosphocellulose. GTP-PPi exchange and capping activities remain physically associated during sedimentation in a glycerol gradient. Under high salt conditions (1 M NaCl), GTP-PPI exchange, capping, and methylating activities co-sediment with an RNA triphosphatase activity and a nucleoside triphosphate phosphohydrolase activity as a 6.5 S multifunctional enzyme complex which contains two major polypep- tides of 96,000 and 26,000 molecular weight. The characteristics of the various enzymatic reactions catalyzed by this complex are described.

The enzyme was characterized as a rifampicin-and a-amanitin-resistant polymerase capable of transcribing singlestranded DNAs in the presence of Mn2+ and the four rNTPs (6). During further studies on the RNA polymerase, we observed a novel nucleoside triphosphate-PPi exchange reaction catalyzed by partially purified RNA polymerase preparations. This exchange reaction was unusual in that it was independent of a DNA template, was insensitive to nucleases, and required only GTP for activity. Furthermore, the activity was not evident in more highly purified RNA polymerase preparations, indicating that the exchange reaction was catalyzed by an enzyme other than RNA polymerase. This exchange activity, previously undetected in vaccinia virus, was observed readily in permeabilized virions and in isolated viral cores.
The occurrence of a virion-associated PPi displacement reaction specific for GTP suggested some relationship between this new exchange activity and the vaccinia virus guanylyltransferase reaction (RNA capping), which has been shown to involve a reversible transfer of GMP from GTP to RNA, with elimination of PPi (2). Recently, Mizumoto and Lipmann have reported that a GTP-PPi exchange activity was associated with partially purified capping enzyme from calf thymus (15). In order to address this question in the case of vaccinia, the properties of the GTP-PP, exchange reaction catalyzed by vaccinia virions have been studied and the enzyme has been solubilized and purified from viral cores. The results presented herein indicate that the GTP-PPi exchange reaction is catalyzed by the vaccinia virus mRNA capping enzyme.
with GTP had only a slight effect on 32P incorporation, while the omission of GTP from the reaction containing the other rNTPs resulted in a drastic reduction in activity. Of the guanosine nucleotides, only GTP supported the exchange reaction; GDP and GMP were inert, as was ITP. dGTP did support 32P incorporation, although less efficiently (-10%) than GTP.
Characterization of the Reaction Product-The 32P-labeled product synthesized by vaccinia virions in the presence of GTP and [32P]PPi was eluted from Norit with ethanolic ammonia and analyzed by thin layer chromatography on polyethyleneimine cellulose developed in 1.6 M LiC1. Under these conditions, the radioactive material migrated as two discrete species with mobilities which were identical with those of the GTP and GDP markers (Fig. 1). No radioactivity could be detected in a position corresponding to GMP or PPi (the latter migrates slightly slower than GTP using these chromatographic conditions).
As shown in Fig. 1, 80% of the radioactive products was GTP and 20% was GDP. We propose that the reaction we are monitoring proceeds as follows: G-P-PP + PPi e G-P-PP to generate GTP labeled at the fland y-phosphate residues.
Radioactive GDP presumably arises via the action of the GTPase activity present in the vaccinia virion. The 32Pi generated by the GTPase would not be detected by our assay procedure and would not be expected to appear in the product analysis in Fig. 1.
Properties of the GTP-PP, Exchange Reaction-The incorporation of 32P into Norit-adsorbable material catalyzed by penneabilized virions was linear with time up to 40 min, and continued at a lesser rate up to 60 min. The activity required Mg2+ as a cofactor (Fig. 2b). Omission of MgCl2 completely abolished activity; optimal exchange occurred between 3 to 7 l~l~ MgClz with a gradual decline in activity at concentrations in excess of 10 m. The dependence of the exchange reaction on PPi concentration is shown in Fig. 2c. The reaction velocity was proportional to PPi concentration up to 1 mM PPi, plateaued between 1 to 1.4 m, and decreased at higher PPi concentrations; activity at 5 m PPi was 24% of the optimal rate. The reaction displayed a hyperbolic dependence on GTP concentration, as indicated in Fig. 2a, with optimal activity at  Table I with the following modifications; reaction volume was increased to 0.25 ml and contained 1 mM GTP and 1 A260 unit of purified virus. After 30 min at 37"C, the reaction was halted as described under "Experimental Procedures." The acidsoluble material was bound to Norit and centrifuged to remove [32P]PPi. The charcoal was resuspended in 5 mM HCI and recentrifuged. This step was repeated three times until all nonadsorbed radioactivity was eliminated from the reaction product. The charcoal was washed with distilled water, recentrifuged, and eluted with a solution containing 50% ethanol and 2% NH4OH. The eluted material was lyophilized, resuspended in water, and chromatographed on PEIcellulose plates developed with 1.6 M LiCl. The chromatograms were cut into I-cm slices and assayed for 32P by Cerenkov counting. The positions of UV marker nucleotides are indicated by the arrows. Properties of the GTP-PPi exchange reaction catalyzed by vaccinia virions. Reaction mixtures containing 0.14 A~w unit of vaccinia virus were prepared as described in the legend to Table I  1 m GTP. Enzymatic activity was dependent on alkaline pH, with optimum exchange between pH 8.1 to 8.5 (Fig. 2 4 . Activities at pH 7.1 and pH 8.9 were reduced by 95% and 35%, respectively. To determine whether the GTP-PPi exchange activity was associated with viral cores, the outer envelope was removed from purified vaccinia virions with 0.5% NP-40 and 50 mM dithiothreitol using a modification (13) of the procedure of Easterbrook (19); the viral cores were isolated by centrifugation as described (6). This treatment removed approximately one-third of the virus protein (Table 11), yet all of the GTPdependent PPi exchange activity was recovered in the viral cores. In order to study the exchange reaction in detail, we have purified the enzyme from viral cores. It was of particular interest to establish the relationship of the exchange activity to two previously described vaccinia enzymes: (a) the DNAdependent RNA polymerase, which was noted to contain a GTP-PPi exchange activity during the initial stage of purification, and ( b ) the RNA guanylyltransferase -7-methyltransferase complex (capping enzyme), which shares certain properties with the exchange activity, i.e. a PPi displacement specific for GTP. The purification (along with Figs. 3 to 6) is presented in the miniprint supplement.
Summary of Purification-The results of the purification of the GTP-PPi exchange enzyme from viral cores indicate that the exchange reaction is catalyzed by the vaccinia virus mRNA capping enzyme. As shown in Figs. 4 and 5, co-chromatography of GTP-PPi exchange activity with RNA (guanine-7-)methyltransferase and/or RNA-guanylyltransferase as well as with a newly described nucleic acid-independent ATPase was observed on columns of DNA-cellulose and phosphocellulose that were developed with linear salt gradients. A summary of the enzyme purification through the phosphocellulose stage (Table 111) indicates that the ratios of GTP-PPi exchange, 7-methyltransferase, and nucleic acid-independent ATPase activities were relatively constant, particularly during the latter purification steps. In the case of the ATPase, the lower yield of enzyme at the DNA-cellulose step may be attributed to the elimination of the nucleic acid-independent ATPase activity contributed by phosphohydrolase I1 (see Fig.  4). Physical association of GTP-PPi exchange with capping enzyme and nucleic acid-independent ATPase was maintained even through glycerol gradient sedimentation in 1 M NaC1. The observed sedimentation coefficient of 6.5 S (Fig. 6)  Cleavage of the y-phosphate of [y-32P]triphosphate-terminated poly(A) was assayed by the loss of acid-insoluble 32P (determined by Cerenkov counting) and is expressed as the percentage of triphosphate termini hydrolyzed. Cap formation was assessed by tritium counting of the same filters in Econofluor scintillation fluid.

TABLE I11
Purification of GTP-PP, exchange enzyme and associated activities GTP-PPi exchange was assayed as described in the legend to    RNA Triphosphatase Associated with GTP-PPi Exchange and Capping Enzyme Complex-During assays of guanylyltransferase in the presence of a [y-32P]triphosphate terminated poly(A) acceptor, it was noted that a loss of acid-insoluble 32P radioactivity (ascertained by Cerenkov counting) occurred in large excess over the amount of ends capped, and that this loss of acid-insoluble 32P was independent of GTP. Venkatesan et al. (20) have recently reported that an RNA triphosphatase activity is associated with the vaccinia capping enzyme, and that conversion of triphosphate-terminated RNA to diphosphate-terminated RNA precedes guanylylation. Fig. 6 shows that our preparation of GTP-PPi exchange enzyme does indeed contain an RNA triphosphatase which co-sediments with capping enzyme and nucleic acid-independent ATPase. The results of Fig. 7 support their conclusion that terminal cleavage precedes capping.
Sodium Dodecyl Sulfate Gel Electrophoresis-Due to the minute amounts of protein present in the more purified enzyme fractions, we have assessed qualitatively the purity of the GTP-PPi exchange enzyme preparations by sodium dodecyl sulfate polyacrylamide gel electrophoresis and subsequent visualization of the protein bands using the highly sensitive silver staining procedure (25). Fig. 8 shows the polypeptide composition of virions, cores, DEAE-cellulose I, DEAE-cellulose 11, DNA-cellulose, and phosphocellulose fractions, revealing an extensive purification via this protocol. The phosphocellulose enzyme contains two major polypeptides of molecular weights 95,800 f 1,600 and 26,400 f 800. These values are the average of four separate molecular weight determinations using sodium dodecyl sulfate gels and have average deviations as indicated. The presence of these same two polypeptides paralleled the GTP-PPi exchange activity and associated activities across the glycerol gradient ( Fig. 9). Martin et al. (1) reported that their capping enzyme preparations contained two major polypeptides of 95,000 and 31,400 molecular weight.
Characterization of the PPi-Exchange Reaction Product Synthesized by the Purified Enzyme-The product synthesued by the purified PPi exchange enzyme in the presence of GTP and [32P]PPi was eluted from Norit and analyzed by thin layer chromatography on PEI-cellulose developed with 1.6 M LiCI. As shown in Fig. loa, virtually all the radioactivity comigrated with the GTP marker; in contrast to the results with permeabilized virions, no peak of radioactive GDP was observed, suggesting that GTPase activity may have been eliminated during enzyme purification (see below). Since this chromatography system served only to resolve nucleoside  mono-, di-, and triphosphates, the reaction product was also analyzed using PEI-cellulose developed with 0.75 M (NH4)2S04, a system capable of resolving the four NTPs. Fig.  10b indicates that only GTP was formed in the PPi-exchange reaction; no labeled ATP, CTP, or UTP were detected as contaminants of GTP.
Fate of GTP during PP, exchange-In view of the presence of an ATPase activity in the purified PPi exchange enzyme preparation, it was of interest to determine whether the failure to observe 32P-labeled GDP as a reaction product ( Fig. 1 uersus Fig. 10) was due to an inherent inability of the ATPase to hydrolyze GTP, or to a lack of GTPase activity under the reaction conditions employed for PPi exchange. This issue was addressed by following the fate of [3H]GTP when incubated with purified enzyme. Fig. 11 shows that incubation of [3H]-GTP with enzyme in the absence of PPi results in significant cleavage of GTP to GDP. GDP formation was linear for 30 min, and continued at a lesser rate up to at least 60 min of incubation. No formation of [3H]GMP was detected at any time (not shown). Clearly, the nucleic acid-independent ATPase is not restricted in its NTP specificity, and is more appropriately termed a nucleoside triphosphate phosphohydrolase. When 1 m~ PPi was included in the reaction mixture, a dramatically different result was obtained (Fig. 11); under conditions which supported GTP-PP, exchange, the cleavage of GTP to GDP was negligible. Thus, the absence of a GDP product is explained by an inhibition of phosphohydrolase activity by PPi. It was noteworthy that in the presence of 1 m~ PPi no formation of [3H]GMP could be detected at any time; this was the case even when (as depicted in Fig. 11   Zn" + Mg'+ 122 discussion of possible exchange reaction intermediates is included below. Characteristics of the PPi Exchange Reaction Catalyzed by Purified Capping Enzyme Complex-The ability of vaccinia guanylyltransferase to catalyze a [32P]PPi exchange with GTP would seem to afford a sensitive and convenient assay for capping enzyme function. In the interest of facilitating such an assay, the optimal conditions for GTP-PPi exchange by purified enzyme were determined. Divalent Cation Requirement-PPi exchange required a divalent cation in the form of Mg2+ ( Table V). The dependence of PPi exchange on MgClz concentration was similar to that described for the exchange in permeabilized virions; optimal activity occurred from 3 to 7 m M MgCh. Of the various divalent cations tested, only Mg2+ supported exchange activity, calcium and cobalt did not support PPi exchange but did not significantly inhibit the reaction in the presence of magnesium. Manganese, copper, and zinc also failed to support activity, but did inhibit PP, exchange by approximately 80% when included with magnesium ( Table V).
pH Dependence-Optimal GTP-PPi exchange occurred from pH 7.9 to 8.5 (60 m~ Tris.HC1 buffer). Activity at pH 7.3 was 37% of the activity at pH 8.3.
PP, Dependence-In the presence of 0.2 mM GTP, PPi exchange readily occurred from 0.6 to 2 mM PPi. Use of lower or higher PPi concentrations resulted in a reduction of activity as shown in Fig. 12.
Nucleotide Dependence-The rate of PPi exchange increased with increasing GTP concentration up to 0.3 mM GTP. Activity decreased at higher GTP concentrations as indicated in Fig. 12. The apparent K,,, for GTP of the purified enzyme was lower than that of permeabilized virions (see Fig.  2b); this may be due to the elimination of GTPase activity during enzyme purification.
Of the various ribonucleoside triphosphates, only GTP was able to participate in the PPi exchange reaction catalyzed by purified capping enzyme complex (Table VI). Among the guanine nucleotides, GDP and GMP were inactive while dGTP did support PPi exchange, albeit far less efficiently than GTP. It is noteworthy that 7-methyl-GTP did not participate in PPi exchange; Martin and Moss have previously shown that 7-methyl-GTP cannot serve as a cap donor in RNA guanylylation (21).
Effect of Salts and Inhibitors The GTP-PPi exchange reaction was unaffected by the omission of dithiothreitol, yet was almost completely inhibited by the sulfhydryl antagonist, p-hydroxymercuribenzoate. This inhibition was reversible by the addition of dithiothreitol (Table VII). The effect of ionic strength on enzyme activity is shown in Table VII. GTP-PPi exchange was unaffected at NaCl concentrations up to 0.1 M and decreased gradually as the ionic strength was increased from 0.1 to 0.7 M NaC1. The PPi exchange activity was not inhibited quantitatively by NaCl concentrations as high as 1 M. Table VII, neither Pi nor inorganic sulfate were inhibitory to GTP-PPi exchange up to 40 m~ concentrations; Pi in particular had a stimulatory effect on enzyme activity.

As indicated in
Properties of the RNA Triphosphatase and Nucleic Acidindependent NTP Phosphohydrolase Associated with Cap-  ping Enzyme-RNA triphosphatase, an activity which cleaves only the y-phosphate of triphosphate-terminated RNA was described initially in extracts of Escherichia coli by Maitra and Hurwitz (23). Two RNA triphosphatases were purified from E. coli. One enzyme hydrolyzed the y-phosphate of ATPterminated RNAs as well as the terminal phosphate of ATP and dATP. A second activity hydrolyzed the terminal phosphate of RNAs without regard to the nucleotide at the 5'-end; this enzyme also cleaved the y-phosphate of all four ribonucleoside triphosphates. Tutas and Paoletti (12) have purified an RNA triphosphatase from vaccinia virus cores. The vaccinia enzyme was reported to cleave the y-phosphate of 5'-ATP-and 5'-GTP-terminated polyribonucleotides, but was reported not to cleave the y-phosphates of ATP or GTP. These authors proposed that the function of this enzyme is to generate diphosphate-terminated RNAs which may in turn be capped by the guanylyltransferase -7 methyltransferase complex. Evidence in support of this model was provided by Venkatesan et al. (20). Our own data support their conclusion that an RNA triphosphatase is a component of the capping enzyme complex; however, we found that the cleavage of yphosphate by the capping enzyme was not restricted to RNA termini but occurred with ATP and GTP as well. Studies of hydrolysis of y-phosphate by purified capping enzyme are presented in the miniprint supplement.
Further Studies of RNA Guanylylation and Methylation-The capping enzyme activity obtained using the present purification procedure catalyzed both the guanylylation and methylation of the 5'-ends of triphosphate-terminated poly (A). The transguanylylation reaction showed an absolute requirement for MgC12 and a polynucleotide cap acceptor (ppp-terminated poly(A)). Activity was extremely sensitive to inhibition by PPi (97% inhibition at 100 PM PPi) but was relatively insensitive to Pi (no inhibition up to 20 mM Pi; 55% inhibition at 40 mM Pi). The donor specificity of the capping reaction was examined indirectly, e.g. by assessing the ability of various nucleotides to substitute for GTP in the GTPdependent methylation of triphosphate-terminated poly(A) (see "Experimental Procedures"). Of the nucleotides tested, only GTP, dGTP, and GTPyS supported the methylation reaction. ATP, CTP, and UTP did not support methylation.
Association of GTP-PPi Exchange Activity and Capping Enzyme with DNA-dependent RNA Polymerase-The studies presented above were all performed with the GTP-PPi exchange activity purified from the DEAE-cellulose 11 unbound fraction. As indicated in Fig. 3 and Table 111, a small but significant portion of the GTP-PPi exchange and RNA (guanine-7-)methyltransferase activities were retained on DEAE-cellulose (DEAE 11) and eluted in parallel with DNAdependent RNA polymerase (and with ATPase assayed in the presence of +X174 DNA). For further purification, column Fractions 10 to 15 were combined and the pooled DEAEcellulose 11-bound fraction was applied to a 6-ml column of phosphocellulose which had been equilibrated with 50 mM NaCl in Buffer A. The column was developed with a 60-ml linear gradient of 0.15 to 0.5 M NaCl in Buffer A. Fractions (-1 ml) were collected and assayed as indicated in Fig. 13. RNA polymerase was retained completely on this column and eluted as a peak at 0.35 M NaC1; here too, the GTP-PPi exchange and capping enzyme activities eluted in parallel with the RNA polymerase, as did nucleic acid-independent ATPase activity. ATPase activity in the presence of a DNA cofactor chromatographed as two (perhaps three) components; the major activity did not co-elute with RNA polymerase, GTP-PPi exchange enzyme, or capping enzyme. The relationship of these ATPase activities to the three activities resolved in Fig.  4 has not been determined.
The chromatograms in Figs. 3 and 13 explain the initial observation of GTP-PPi exchange activity in partially purified RNA polymerase preparations as well as the failure to detect incorporation of [y-32P]ATP into RNA synthesized using the phosphocellulose RNA polymerase preparation.2 The data (Figs. 3 and 13) are consistent with the concept that capping enzyme and RNA polymerase may exist as a complex in the vaccinia virion and that this complex may be dissociated during disruption of the virus particle. Preliminary evidence indicates that the RNA polymerase obtained after glycerol gradient centrifugation (6) does not possess GTP-PPi exchange activity although the fate of the GTP-PPi exchange enzyme during centrifugation was not evaluated. Studies of the presumptive complex between capping enzyme and RNA polymerase are in progress.

DISCUSSION
The vaccinia virion-catalyzed NTP-PPi exchange reaction described above is distinguished by its specific requirement for GTP as the nucleoside triphosphate substrate ( Table I); this reaction clearly differs from the PPi exchange reactions mediated by aminoacyl-tRNA synthetases, T4 RNA ligase, and some DNA ligases insofar as these enzymes utilize an ATP substrate. Moreover, the vaccinia virus-catalyzed PPi exchange reaction contrasts with the NTP-PPi exchange activities associated with DNA and RNA polymerase in its independence of a polynucleotide template and its nuclease insensitivity.
Among those vaccinia enzymes which have been studied in detail, the RNA-guanylyltransferase (capping enzyme) shares common features with the GTP-PPi exchange activity. The * S. Shuman, unpublished work. capping enzyme, which catalyzes the reaction Gp;; + (p)ppRNA + GpppRNA + $f'i + (Pi) specifically requires GTP as a cap donor (21). Several groups have shown that the vaccinia guanylylation reaction involves a PPi displacement in GTP, and that the reaction is reversible by pyrophosphorolysis of the GpppX structure to yield GTP (1-3). A GTP-PPi exchange activity in the absence of a cap acceptor has not been reported previously for the vaccinia capping enzyme, but such an activity has been shown to be associated with guanylyltransferase from rat liver nuclei (15). By the criterion of co-purification, we have established that the vaccinia GTP-PP, exchange reaction is indeed catalyzed by vaccinia virus guanylyltransferase; at no point in purification were the exchange activity and the capping enzyme resolved. The exchange reaction thus affords a simple, inexpensive, and sensitive assay for the presence of vaccinia capping enzyme.
The fact that guanylyltransferase catalyzes GTP-PPi exchange in the absence of a cap acceptor has interesting implications for the mechanism of transguanylylation. We propose (by analogy to the mechanism described for the DNA ligase reaction (22)) that the initial step in cap formation is the generation of an enzymeguanylate complex with concomitant release of PPi. The enzyme guanylate complex would then transfer its GMP residue to the 5'-terminus of a diphosphateterminated RNA to generate the G(5')ppp(5')X cap structure.
In the presence of GTP and [32P]PPi, rapid formation of the enzyme. guanylate complex and reversal of this reaction would constitute the basis of the GTP-PPi exchange described herein. Again, by analogy to DNA ligase, the capping enzyme. guanylate complex may involve a covalent nucleotide-protein linkage. This is consistent with our failure to detect free GMP as a reaction intermediate during PPi exchange and with the lack of [3H]GMP-GTP exchange.
The analogy between the transguanylylation and DNA ligation reactions is not factitious (22). E. coli DNA ligase forms a covalent enzyme-adenylate complex which transfers the AMP moiety to the 5'-end of 5'-phosphate-terminated DNA. The resulting intermediate A(5')pp(5')X is quite similar in general structure to the unmethylated RNA cap; Moreover, the capping enzyme, like DNA ligase, catalyzes nucleotide-PPi exchange in the absence of a polynucleotide acceptor. Direct evidence for the proposed transguanylylation scheme, i.e. the demonstration of a covalent enzyme-guanylate intermediate, has indeed been obtained and will be presented el~ewhere.~ In addition to the guanylyltransferase, 7-methyltransferase, and GTP-PPi exchange activities, the capping enzyme contains an RNA triphosphatase and a nucleic acid-independent NTP phosphohydrolase, all of which are associated in a 6.5 S complex containing two major polypeptides of 96,000 and 26,000 molecular weight. Using a similar purification procedure, Venkatesen et al. (20) have demonstrated an association of capping enzyme with RNA triphosphatase in a complex previously reported (1) to contain major polypeptides of 95,000 and 31,400 molecular weight. Tutas and Paoletti (12) have purified an RNA triphosphatase which does not contain guanylyl-or methyltransferase activities, but, interestingly, this enzyme sediments at 6.2 S on sucrose gradients and consists of major polypeptides of 90,000 and 26,000 molecular weight.

S. Shuman, manuscript in preparation.
On the other hand, Monroy et al. (3) have isolated a guanylyltransferasemethyltransferase complex which lacks an RNA triphosphatase component; this enzyme consisted of polypeptides of 95,000,28,000, and 59,000 molecular weight at an analogous stage of purification. At this time, none of these activities (GTP-PPi exchange, guanylyltransferase, 7-methyltransferase, RNA triphosphatase, or nucleic acid-independent nucleoside triphosphatase) have been assigned with certainty to any of the polypeptides found in the enzymes obtained by various investigators. It is conceivable that the observation, or failure to observe association between the activities may stem from selective inactivation or dissociation during enzyme extraction and purification. However, comparison of the various purification protocols has not yet been performed.
The observation that vaccinia capping enzyme is associated with the virus DNA-dependent RNA polymerase through two column chromatography steps is provocative. These data suggest a pre-existing complex between the two activities in the virus particle. Such a complex would facilitate the capping of nascent RNAs soon after initiation. Although it has not yet been demonstrated, the postulation of a complex between guanylyltransferase and RNA polymerase I1 in the cell nucleus offers one explanation of the preponderence of RNA capping of RNA polymerase I1 products (24).  was chnnutcqaphed'm phosph-llulose as described i n the text. GTP-PP.

3.
The GTP-PP. exchange enobtained frm the Du-cellulose step exchange, ATPase ( i n the absence of DX174 MA) and 7 -r t h y l t n n r f c r a r c we& GTP-dewdent n r t h y l a t i m Of trlmosphate-terminated ~oly(A1) was arrayed arrayed as described i n the legend t o Fig. 3. '"Capping en-"