Proc. Nat. Acad. Scu. USA Vol. 72, No. 7, pp. 2611-2615, July 1975
Biochemistry
Methylation of messenger RNA of Newcastle disease virus in vitro by a virion-associated enzyme (RNA nucleotidyltransferase/RNA transcriptase/inhibition of RNA methylase/5' terminus of messenger RNA)
RICHARD J. COLONNO AND HENRY 0. STONE Department of Microbiology, The University of Kansas, Lawrence, Kans. 66045
Communicated by Charles D. Michener, April 25,1975
Purified Newcastle disease virus contains an ABSTRACT enzyme that incorporates the methyl group from S-adenosylL-methionine into RNA synthesized in vitro by the virion-associated RNA polymerase (RNA nucleotidyltransferase). Incorporation of radioactivity from S-adenosyl-L-[methyl3HJmethionine was totally dependent upon RNA synthesis. The methylation reaction was completely inhibited by S-adenosyl-L-homocysteine, suggesting the transfer of only the methyl group of S-adenosy7-methionine to RNA products. Velocity sedimentation and hybridization of the in vitro product RNA indicated that both [3H]methyl and [32PJGMP labels resided in single-stranded 18S RNA molecules which were virus specific. Approximately 1 to 2 methyl groups were incorporated per RNA molecule. DEAE-cellulose chromatography of product RNA after alkaline hydrolysis suggested that the 5' terminus was the site of methylation.
Methylated nucleotides have recently been detected in messenger RNAs from mouse L cells (1) and from Novikoff hepatoma cells (2), as well as in a number of viral messenger RNAs synthesized in vitro by virion-associated RNA polymerases (RNA nucleotidyltransferases) (3-10). The methylated nucleotides in viral messenger RNAs synthesized in vitro by virions of cytoplasmic polyhedrosis virus (3), reovirus (4, 5), vaccinia virus (6), and vesicular stomatitis virus (7) are located in an unusual structure at the 5' terminus of the RNA molecules. The structure consists of a terminal base-methylated nucleotide linked to a ribose-methylated nucleotide through a 5' to 5' pyrophosphate bond, such as m7G(5')ppp(5')AmpNp or m7G(5')ppp(5')GmpNp (8-11). Methylation is required for proper processing of ribosomal RNA (12), for activation of the virion-associated RNA polymerase of cytoplasmic polyhedrosis virus (3), and for translation of viral messenger RNAs in a protein-synthesizing system in vitro (42). Newcastle disease virus (NDV), an avian paramyxovirus, has a single-stranded RNA genome with a sedimentation coefficient of 50 S (13, 14) and a molecular weight of 6 million (15). Messenger RNAs from polysomes of NDV-infected cells sediment primarily at 18 S, are complementary in base sequence to the viral genome, and contain polyadenylate sequences (16, 17). Purified NDV contains a virion-associated RNA polymerase (18) that transcribes the 50S RNA genome in vitro into 18S RNA molecules which are complementary in base sequence to the genome, and which also contain polyadenylate sequences (16). Since the paramyxovirus genome does not serve as messenger RNA (13, 19), the virionassociated transcriptase appears to be responsible for synthesis of messenger RNA in the infected cell (20, 21). Krauter Abbreviations: NDV, Newcastle disease virus; AdoHcy, S-adenosylL-homocysteine; AdoMet, S-adenosyl-L-methionine. 2611
and Consigli (22) have recently shown that methionine is essential for the synthesis of Newcastle disease virus in chicken embryo cells. In addition to its role in protein synthesis, methionine was found to be essential in methylation, and the RNA in purified virions of Newcastle disease virus was found to be methylated (22). In this report we demonstrate the presence of methyl transferase activity in purified NDV, which catalyzes the incorporation of the methyl group from S-adenosyl-L-methionine (AdoMet) into the messenger RNA synthesized in vitro by the virion-associated transcriptase. MATERIALS AND METHODS A strain of Newcastle disease virus, plaque purified from the Beaudette "C" strain described by Kingsbury (13, 23) and Granoff (24), was grown in embryonated chicken eggs (24) at 370 for 36 hr. Virions were purified from the allantoic fluid by differential centrifugation (13), rate zonal centrifugation (25), and isopynic banding in glycerol tartrate gradients (26). The standard reaction mixture in vitro (0.1 ml) for assaying both [32P]GMP and [3H]methyl incorporation contained 50 mM Tris-HCl (pH 7.8) (final reaction pH was 7.1), 120 mM NaCI, 0.4 mM MnCI2, 0.015% (v/v) Triton N101, 0.7 mM each of ATP, CTP, and UTP, 0.28 mM GTP, 3 mM dithiothreitol, 1 4uCi of [a-32P]GTP (15 Ci/mmol), and 0.78 MM (1 /,Ci) S-adenosyl-L-[methyl-3H]methionine ([3H]AdoMet) (12.6 Ci/mmol). Reaction mixtures were incubated with 35-45 gig of purified virions at 320 for 6 hr, the reactions were terminated by the addition of 0.5% sodium dodecyl sulfate, and the in vitro product RNA was isolated by sequential phenol extraction at pH 7 and 9 (27). Extracted RNA in the aqueous layers was pooled, precipitated with ethanol, and assayed for trichloroacetic acid-precipitable radioactivity (28). For further analysis the labeled RNA in vitro was purified free of unreacted triphosphates and anyremaining S-adenosyl-L-methionine by chromatography on a Sephadex G-50 (Pharmacia Fine Chemicals) column (1.5 X 30 cm) equilibrated with 0.01 M sodium acetate, 0.05 M NaCl, 0.5% sodium dodecyl sulfate (pH 5.0). The same buffer was used to elute 70 fractions (1 ml) of which 25-Ml aliquots were assayed directly for radioactivity in 9 ml of PCS (Amersham/ Searle). The fractions of the void volume that contained radioactivity were pooled, and the RNA pellet was resuspended in 0.5 ml of 5 mM Tris-HC1, 1 mM EDTA (pH 7.4). The RNA was layered over a 10-ml 15-30% (w/w) linear sucrose gradient in 5 mM Tris-HCl (pH 7.4), 1 mM EDTA, 0.5% (w/v) lithium dodecyl sulfate, and 0.1 M lithium chloride
2612
Biochemistry: Colonno and Stone
Proc. Nat. Acad. Sci. USA 72 (1975)
Table 1. Requirements for methylation and RNA synthesis Reaction mixture
[3H]Methyl cpm incorporation
[32P]GMP cpm incorporation
15,300 3,600 Complete* 15,700 0 -[3H]AdoMett 379 0 -NDV 531 67 -NaCl 239 10 -Triton N°'0 422 7 -CTP 568 4 -MnCl2 1,000 210 -Dithiothreitol * Standard reaction mixtures contained 1 ,ACi of [a -32P]GTP (15 Ci/mmol), 0.78 IAM S-adenosyl-L-[methyl-3H]methionine, and 44 jtg of Newcastle disease virus, as described in Materials and Methods. t When S-adenosyl-L-[methyl-3Hjmethionine was omitted, 50 mM Tris.HCl (pH 7.3) was used in place of 50 mM Tris.HCl (pH 7.8).
which had been formed over a 0.6-ml 60% (w/w) sucrose cushion. Gradients were centrifuged in a Beckman SW 41 rotor (200, 12 hr, 34,000 rev/min), and [14C]uridine-labeled 18 and 28S ribosomal RNAs from chicken embryo fibroblasts were centrifuged in a parallel gradient as markers. RNA sedimenting in the 18S region of the gradient was pooled and reprecipitated with ethanol. A portion of the precipitated RNA was used in hybridization experiments with 50S genomic RNA (23), and the remainder was hydrolyzed in 0.3 ml of 0.3 M KOH for 16 hr at 37'. After neutralization with 75 Ml of 1.2 M perchloric acid, the hydrolyzed RNA was chromatographed on a DEAE-cellulose (Whatman DE 32) column (29) using a pancreatic ribonuclease (29) digest of rat liver RNA as unlabeled oligonucleotide markers. Radioisotopes were purchased from New England Nuclear Corp.
RESULTS Dependence of Methylation upon Transcription. Incubation of purified Newcastle disease virus particles under optimal conditions for RNA synthesis in vitro resulted in the incorporation of radioactivity from S-adenosyl-L-[methyl3H]methionine into a trichloroacetic acid-precipitable product (Table 1). In contrast to cytoplasmic polyhedrosis virus (3), transcription of NDV was neither stimulated nor dependent upon a methyl donor such as S-adenosyl-L-methionine (Table 1, line 2). Methylation was dependent upon transcription, since methylation exhibited an absolute requirement for all components in the RNA polymerase reaction. In the absence of dithiothreitol, very low levels of methylase (5.6%) and transcriptase (6.5%) activity were detectable (Table 1, line 8). The optimum pH (7.1) and the optimum concentrations of sodium chloride (0.12 M) and of manganese (0.4 mM) were the same for [3H]methyl and for [32P]GMP incorporation (data not shown). Methylation and transcription proceed with either manganese or magnesium, but RNA synthesis is 2-fold higher in the presence of manganese (data not shown). The rates of incorporation of [3H]methyl and [32P]GMP into acid-insoluble products were identical (Fig. 1). Methylation and transcription proceeded at a linear rate for approximately 6 hr (Fig. 1), suggesting that the two processes may be coupled. Inhibition of Methylation by S-Adenosyl-L-homocysteine. Both RNA (30) and DNA (31) methylases of Esche-
10'p
3 0
x
2
E Q
5 3
4)
01
2
~0
1
I
4
8 12 16 20 TIME (HOURS)
24
FIG. 1. Time course of [32P]GMP and [3H]methyl incorporation. A 9-fold standard reaction mixture (0.9 ml), containing 9 'Ci of [a-32P]GTP (15 Ci/mmol), 0.78 1AM S-adenosyl-L-[methyl3H]methionine (12.6 Ci/mmol), and 0.38 mg of purified Newcastle disease virions, was incubated at 320. At the indicated time intervals, 0.1 ml was removed and the RNA was extracted with phenol/ dodecyl sulfate, precipitated with ethanol, and assayed for trichloroacetic acid-precipitable radioactivity (28). The counting efficiencies of 32p and 3H under our conditions were 91% and 41%, respectively. The specific activities of the isotopes were 71 cpm/pmol of [a-32P]GTP and 10,800 cpm/pmol of S-adenosyl-L-[methyl3H]methionine. Incorporation of [3H]methyl group (0) and [32P]GMP (0) into acid-insoluble products.
richia coli are subject to inhibition by the product of the reaction, S-adenosyl-L-homocysteine (AdoHcy). To test for an inhibition of methylation, varying amounts of S-adenosyl-L-homocysteine were added to incubation mixtures containing S-adenosyl-L-[methyl-3H]methionine and [a32P]GTP to determine the extent of methylation and RNA synthesis, respectively. RNA synthesis was not affected by S-adenosyl-L-homocysteine at concentrations that ranged from 1 MM to 0.5 mM (Fig. 2). However, [3H]methyl incorporation was reduced by increasing levels of S-adenosyl-L-
15
I
ptQ
o 2 x
1
E
10o0
C.' x
4
0)
0
0 1 5
-
-0
I
0.1
0.3
0.5
AdoHcy CONCENTRATION (mM). FIG. 2. Effect of S-adenosyl-L-homocysteine (AdoHcy) on methylation and RNA synthesis. Varying amounts of S-adenosylL-homocysteine were added to duplicate standard reactions containing 1 ;Ci of [a-32P]GTP (15 Ci/mmol), 0.78 pM S-adenosyl-L[methyl-3Hjmethionine (12.6 Ci/mmol), and 44 pg of purified Newcastle disease virus. After a 6-hr incubation at 320, the product RNA was extracted with phenol/dodecyl sulfate, precipitated with ethanol, and assayed for trichloroacetic acid-precipitable radioactivity (27, 28). Incorporation of [3H]methyl group (0) and [32P]GMP (0) into acid-insoluble products.
Proc. Nat. Acad. Sci. USA 72 (1975)
Biochemistry: Colonno and Stone
2613
Table 2. Hybridization of NDV [3H]methyl/ [32PJGMP-labeled product RNA 8
cpm Treated RNase-resistant cpm (%)
[3H]- [32p] F 6
:0C)
0
x
3
~0 3
E
4 x
0)
0
1 0
i
I
2 1
5
20 15 10 FRACTION NUMBER
Treatment*
methyl GMP
[
3H]methyl
[32P]GMP
"
25
FIG. 3. Sedimentation of labeled product RNA in vitro. A 20fold standard reaction mixture (2 ml) containing 20 gCi of [a32P]GTP (15 Ci/mmol), 0.78 MM S-adenosyl-L-[methyl-3H]methionine (12.6 Ci/mmol), and 0.72 mg of Newcastle disease virus was incubated for 6 hr at 320. After sequential phenol/dodecyl sulfate extraction at pH 7 and 9, gel filtration on Sephadex G-50, and ethanol precipitation, the product RNA was sedimented on a linear 15-30% sucrose gradient as described in Materials and Methods. Gradients were fractionated (0.4 ml) from the top with an ISCO gradient fraction collector, and 25-Ml aliquots of each fraction were diluted with 0.5 ml of H20 and 9 ml of PCS (Amersham/Searle), and assayed for [3Hlmethyl (0) and [32P]GMP (0) radioactivity. Positions of [14C]uridine-labeled 18 and 28S ribosomal RNA sedimented in a parallel gradient are indicated by arrows.
homocysteine, and at an adenosylhomocysteine to adenosylmethionine ratio of 26, inhibition was 50%. These results suggest that S-adenosyl-L-homocysteine is an end product of the reaction and that only the [3H]methyl group of S-adenosyl-L-[methyl-3H]methionine is incorporated into the RNA product. Velocity Sedimentation of Methyl-Labeled RNA. The products synthesized in the presence of S-adenosyl-L[methyl-3H]methionine and [a-32P]GTP were analyzed by rate-zonal sedimentation on sucrose gradients (Fig. 3). The RNA synthesized in vitro by NDV sedimented primarily at 18 S, the same size as the predominant mRNA from NDVinfected cells (17, 21). The radioactive profiles of [3H]methyl- and [32P]GMP-labeled RNAs coincided, but the ratio of 3H to 32P decreased with increasing sedimentation rate. These results indicate that the methyl groups were not distributed randomly, but that each polynucleotide chain might have the same number of methylated sites. When the molar ratio of methyl groups to GMP residues was calculated for the 18S in vitro product (Fig. 3, fraction no. 13), one methyl group was incorporated per 1315 nucleotides. Estimating the chain length of an 18S RNA molecule at 2100 (32), an average of 1.6 methyl residues were incorporated into each 18S molecule. Methyl-Labeled RNAs are Virus Specific. When the in vitro product synthesized in the presence of S-adenosyl-L[methyl-3H]methionine and [a-32P]GTP was subjected to ribonuclease digestion, both radioactive labels were rendered
Unannealedt
682
Self-annealedt
639
1284 1348
29 (4) 41 (6)
68 (5) 210 (16)
Annealed to 50S 1500 (100) * RNA synthesized in vitro sedimenting in the 14-25S region of sucrose gradients (Fig. 3, fractions 10-15) was annealed as described by Kingsbury et al. (28). t Product RNA was treated directly with 5 ,g/ml of RNase A for 30 min at 240. Product RNA was incubated under annealing conditions before RNase treatment. § Product RNA was annealed in 30 al to 0.32 jsg of purified 50S genomic RNA extracted from purified Newcastle disease virus.
RNAgenomes§ 705
1505
710 (100)
acid-soluble (Table 2), indicating that the radioactivity resided in single-stranded RNA and that methylation did not occur on the polyadenylate segments. After self-annealing, a small amount of the product was converted to a ribonuclease-resistant structure. When annealed with 50S RNA genomes, all of the radioactivity became ribonuclease-resistant, indicating that the [3H]methyl and [32P]GMP labels reside in RNA that is complementary in base sequence to the genome. Location of Methyl Groups in the RNA Molecule. The site of methylation catalyzed by specific methylases is often on the base moiety, and can be identified after alkaline hydrolysis of the RNA and chromatography of the methyl-substituted mononucleotides. Methylation can also occur on the ribose moiety of the nucleotide as 2'-O-methyl substitution. This modification renders the phosphodiester bond with the 3'-adjacent base resistant to alkaline hydrolysis, resulting in a methylated dinucleotide. DEAE-cellulose chromatography of alkaline hydrolysates was used to determine the location of methyl groups in the RNA molecules in vitro. The double-labeled RNA sedimenting in the 14-25S region of sucrose gradients (Fig. 3, fractions 10-18) was precipitated with ethanol and hydrolyzed with KOH. The alkaline hydrolysate was. neutralized and chromatographed on a DEAE-cellulose column, with a pancreatic ribonuclease digest of rat liver ribosomal RNA as oligonucleotide marker. This procedure separates oligonucleotides primarily on the basis of the negatively charged phosphate residues. Unreacted S-adenosyl-L-methionine alone does not bind to the column, while GTP elutes between diand trinucleotides (-3 and -4 charge). Results (Fig. 4) show that after alkaline hydrolysis, the 32P radioactivity eluted coincident with mononucleotide markers, indicating the completeness of the alkali digestion. Although the initial 5'terminal nucleotide has not been determined for NDV messenger RNA, results from studies on reovirus (33, 34, 4), vesicular stomatitis virus (35, 36), and vaccinia virus (10, 41) strongly suggest that the undigested NDV messenger RNA would contain ATP or GTP. The relative proportions of [a32P]GMP label in internal and terminal residues for an 18S RNA molecule would be approximately 500:1. If there are two GMP residues at the 5'-termini (as in m7GapppaG), the relative proportions would be 250:1. Sufficient radioactivity to detect 5'-terminal incorporation (>105 cpm) was not chromatographed on this column; thus all the 32P radioactiv-
2614
Biochemistry: Colonno and Stone
Proc. Nat. Acad. Sci. USA 72 (1975)
-I2 0
4
I
:c~) 0
0()
3
x
E
1
x
2
0
a) 5
I
20
100 80 60 40 FRACT ION NUMBER
120
FIG. 4. DEAE-cellulose chromatography of an alkaline digest of RNA synthesized in vitro in the presence of manganese. Double-labeled RNA, sedimenting in the 14-25S region of sucrose gradients (Fig. 3, fractions 10-15), was pooled and precipitated with ethanol. The RNA was then hydrolyzed in 0.3 ml of 3 M KOH for 16 hr at 370, and neutralized by the addition of 75 ul of 1.2 M perchloric acid. Neutralized hydrolysates were diluted with 9.7 ml of 7 M urea, 0.01 M Tris-HCl, 3 mM EDTA (pH 8), and 3 mg of RNasedigested rat liver ribosomal RNA was added as oligonucleotide marker. The digest was applied to a 0.6 X 17 cm column of DEAEcellulose and eluted with a 160-ml linear gradient from 0 to 0.25 M NaCl in 7 M urea, 0.01 M Tris-HCl, 3 mM EDTA (pH 8) and followed with 10 ml of 2 M NaCl in the same buffer. Fractions (1.4 ml) were collected at a flow rate of 5 ml/min. The absorbance at 260 nm (-----) of each fraction and conductivity of every fifth ) fraction was measured. The sodium chloride concentration ( was obtained from a standard curve. Each sample was placed into 15 ml of counting solution containing Triton X-100 (35), and the radioactivity was determined in a liquid scintillation counter. AdoMet (arrow), GTP (arrow), [3H]methyl label (0), and [32P]GMP label (0) are indicated.
ity detected (3 X 104 cpm) came from internal residues (nearest neighbors of GMP). All of the methyl-labeled material that bound to the column eluted between tri- and tetranucleotides (-4 and -5 charge) (0.144 M NaCl). Since no
methyl label was associated with the mononucleotide peak, base methylation of a single internal nucleotide can be excluded. Ribose methylation of a single internal nucleotide can also be excluded since no methyl label was associated with the dinucleotide peak. At most, only two methyl groups were detected per RNA molecule (Fig. 3). If these methyl groups occur as internal 2'-O-methylribose residues, one trinucleotide or two dinucleotides would be expected. Since the methyl-labeled material did not coincide with the trinucleotide marker, sequential methylation of internal ribose moieties can also be excluded. These results strongly suggest that the [3H]methyl groups reside in the 5' termini of NDV messenger
of S-adenosyl-L-[methyl-3H]methionine was purified, digested with alkali, and chromatographed on DEAE-cellulose. The results (not shown) indicated that methyl-labeled structures synthesized in the presence of magnesium or manganese (Fig. 4) elute in identical positions between triand tetranucleotides. Thus, under the various conditions used, the [3H]methyl label resided in virus-specific RNA, presumably at the 5' terminus after alkaline hydrolysis.
RNAs.
The methyl-labeled nucleotides synthesized in vitro by via structure that appears to differ by one negative charge from the 5' termini of other viral messenger RNAS (4-7, 10). Since the other viral messenger RNAs were synthesized in the presence of magnesium and our experiments used manganese, an additional experiment was performed to determine the influence of magnesium on the elution position of the methylated nucleosides synthesized by NDV. RNA synthesized in vitro by NDV in the presence of magnesium and a 10-fold higher concentration
rions of NDV reside in
DISCUSSION The data presented in this paper demonstrate that purified Newcastle disease virus contains a methyl transferase activity that transfers the methyl group from S-adenosyl-L-methionine to 18S RNA synthesized in vitro by the virion-associated transcriptase. Transfer of only the [3H]methyl group from S-adenosyl-L-[methyl-3H]methionine was inferred from the complete inhibition of [3H]methyl incorporation by S-adenosyl-L-homocysteine. Several lines of evidence presented in this paper demonstrate that the 8H radioactivity was incorporated into RNA. The [3H]methyl-labeled product, which was extractable with phenol and precipitable with trichloroacetic acid, sedimented along with [32P]GMPlabeled RNA and was sensitive to both ribonuclease A and alkaline hydrolysis. Conclusive proof that [3H]methyl groups resided in virus specific RNA was obtained by hybridization of the in vitro product to viral genomes. The extent of methylation was very low,. 1 to 2 methyl groups incorporated per 18S molecule. Initial studies on RNA synthesized in vitro by virions of cytoplasmic polyhedrosis virus (3), reovirus (4, 5), and vesicular stomatitis virus (7) also found 1 to 2 methyl groups per molecule. More detailed experiments have confirmed 2 methyl groups per.molecule for cytoplasmic polyhedrosis virus (8), reovirus (9), and vaccinia virus (10). The [3H]methyl radioactivity eluted from the DEAE-cellulose column between tri- and tetranucleotides. This elution position indicates that the [3H]methyl label is contained in a structure that may differ by one negative charge from the structure reported for other viral messenger RNAs (4-7, 10). A minor peak at this position from vaccinia virus RNA (6, 10) had a structure, m7G(5')pppNp, and was attributed to undermethylated RNA. Undermethylation does not appear to be responsible for the results presented in this paper. The amounts of S-adenosyl-L-[methyl-3H]methionine required to saturate the manganese reaction were determined, and twice the saturation concentration was used. The rates of methylation and transcription were identical over a 6-hr period (Fig. 1), testifying to the adequacy of the S-adenosylL-methionine concentration. It is not presently known if the NDV product is methylated in both ribose and base moieties. Structures that should elute in the position described include: pGmpNp, m7G(5')ppp(5')CGp or
m7G(5')ppGmpNp.
The role of methylation in eukaryotic messenger RNAs is not understood. The presence of S-adenosyl-L-methionine in vitro had no detectable influence on the rate of RNA synthesis by Newcastle disease virus, vaccinia virus (6), or reovirus (4). Price and Rottman (37) demonstrated that methylation does not interfere with the ability of oligonucleotides to direct the binding of aminoacyl-tRNA to ribosomes or with the ability of a partially methylated polynucleotide to direct the incorporation of amino acids into protein (38). Both, Banerjee, and Shatkin (42) obtained results that indicate that methylation of viral messenger RNA is required for translation in vitro.
Biochemistry: Colonno and Stone Methylation could be involved in the control of paramyxovirus replication. Genome transcripts in paramyxovirus-infected cells occur as 18S messenger RNAs (17, 39) and as 50S transcripts which are involved in genome replication (40). The mechanism that regulates synthesis of 18S RNA for translation and 50S RNA for genome replication is not known. It will be of interest to determine if the 50S transcripts are methylated. Transcription of the entire genome into single-stranded methylated viral RNA followed by postsynthetic processing into appropriate segments could occur. We thank Dr. Ron Borchardt for his advice and for the gifts of S-adenosyl-L-homocysteine, S-adenosyl-L-methionine, and S-adenosyl-L-[methyl-14C]methionine. We thank Dr. Aaron Shatkin, Dr. Amiya Banerjee, and Dr. D. H. L. Bishop for helpful discussions. We express our gratitude to Dr. Amiya Banerjee for a manuscript of ref. 7 and to Dr. Aaron Shatkin for a manuscript of ref. 42 prior to publication. The excellent technical assistance of Barbara Stone and of the Kansas University Enzyme Laboratory (Phyllis Shaffer) is gratefully acknowledged. This work was supported by Public Health Service Research Grant AI-11127 from the National Institute of Allergy and Infectious Diseases. 1. Perry, R. P. & Kelley, D. E. (1974) Cell 1, 37-42. 2. Desrosiers, R., Friderici, K. & Rottman, F. (1974) Proc. Nat. Acad. Sci. USA 71, 3971-3975. 3. Furuichi, Y. (1974) Nucleic Acids Res. 1, 809-822. 4. Shatkin, A. J. (1974) Proc. Nat. Acad. Sci. USA 71, 32043207. 5. Faust, M. & Millward, S. (1974) Nucleic Acids Res. 1, 17391752. 6. Wei, C. M. & Moss, B. (1974) Proc. Nat. Acad. Sci. USA 71, 3014-3018. 7. Rhodes, D. P., Moyer, S. A. & Banerjee, A. K. (1974) Cell 3,
327-33. 8. Furuichi, Y. & Miura, K. (1975) Nature 253,374-375. 9. Furuichi, Y., Morgan, M., Muthukrishnan, S. & Shatkin, A. J. (1975) Proc. Nat. Acad. Sci. USA 72,362-366. 10. Wei, C. M. & Moss, B. (1975) Proc. Nat. Acad. Sci. USA 72,
318-322. 11. Rottman, F., Shatkin, A. J. & Perry R. P. (1974) Cell 3, 197199. 12. Vaughan, M. H., Soeiro, R., Warner, J. R. & Darnell, J. E. (1967) Proc. Nat. Acad. Sci. USA 58,1527-1534. 13. Kingsbury, D. W. (1966) J. Mol. Biol. 18, 195-203.
Proc. Nat. Acad. Sci. USA 72 (1975)
2615
14. Duesberg, P. H. (1968) Proc. Nat. Acad. Sci. USA 60, 15111518. 15. Chi, Y. Y. & Bassel, A. R. (1974) J. Virol. 13, 1194-1199. 16. Weiss, S. R. & Bratt, M. A. (1974) J. Virol. 13, 1220-1230. 17. Bratt, M. A. & Robinson, W. S. (1967) J. Mol. Biol. 23, 1-21. 18. Huang, A. S., Baltimore, D. & Bratt, M. A. (1971) J. Virol. 7, 389-394. 19. Kingsbury, D. W. (1973) J. Virol. 12, 1020-1027. 20. Robinson, W. S. (1971) Virology 44,494-502. 21. Collins, B. S. & Bratt, M. A. (1973) Proc. Nat. Acad. Sci. USA 70,2544-2548. 22. Krauter, K. M. & Consigli, R. A. (1974) Am. J. Vet. Res. 35, 985-992. 23. Kingsbury, D. W. (1966) J. Mol. Biol. 18, 204-214. 24. Granoff, A. (1959) Virology 9,636-648. 25. Stone, H. O., Portner, A. & Kingsbury, D. W. (1971) J. Virol. 8, 174-180. 26. Obijeski, J. F., Marchenko, A. T., Bishop, D. H. L., Cann, B. W. & Murphy, F. A. (1974) J. Gen. Virol. 22,21-33. 27. Lee, S. Y., Mendecki, J. & Brawerman, G. (1971) Proc. Nat. Acad. Sci. USA 68,1331-1335. 28. Kingsbury, D. W., Portner, A. & Darlington, D. W. (1970) Virology 42,857-871. 29. Bishop, D. H. L., Mills, D. R. & Spiegelman, S. (1968) Biochemistry 7,3744-3753. 30. Hurwitz, J., Gold, M. & Anders, M. (1964) J. Biol. Chem. 239, 3474-3482. 31. Gold, M. & Hurwitz, J. (1964] J. Biol. Chem. 239,3858-3865. 32. Attardi, G. & Amaldi, F. (1970) Annu. Rev. Biochem. 39, 183-226. 33. Banerjee, A. K. & Shatkin, A. J. (1971] J. Mol. Biol. 61, 643653. 34. Miura, K., Watanabe, K., Sugiura, M. & Shatkin, A. J. (1974) Proc. Nat. Acad. Sci. USA 71, 3979-3983. 35. Roy, P. & Bishop, D. H. L. (1973) J. Virol. 11, 487-501. 36. Chang, S. H., Hefti, E., Obijeski, J. F. & Bishop, D. H. L. (1974) J. Virol. 13,652-661. 37. Price, A. R. & Rottman, F. (1970) Biochemistry 9,4524-4529. 38. Dunlap, B., Friderici, K. H. & Rottman, F. (1971) Biochemistry 10, 2581-2587. 39. Blair, C. D. & Robinson, W. S. (1968) Virology 35, 537-549. 40. Portner, A. & Kingsbury, D. W. (1972) Virology 47,711-725. 41. Urushibara, T., Furuichi, Y., Nishimura, C. & Miura, K. (1975) FEBS Lett. 49, 385-389. 42. Both, G. W., Banerjee, A. K. & Shatkin, A. J. (1975) Proc. Nat. Acad. Sci. USA 72, 1189-1193.
Biochemistry
Methylation of messenger RNA of Newcastle disease virus in vitro by a virion-associated enzyme (RNA nucleotidyltransferase/RNA transcriptase/inhibition of RNA methylase/5' terminus of messenger RNA)
RICHARD J. COLONNO AND HENRY 0. STONE Department of Microbiology, The University of Kansas, Lawrence, Kans. 66045
Communicated by Charles D. Michener, April 25,1975
Purified Newcastle disease virus contains an ABSTRACT enzyme that incorporates the methyl group from S-adenosylL-methionine into RNA synthesized in vitro by the virion-associated RNA polymerase (RNA nucleotidyltransferase). Incorporation of radioactivity from S-adenosyl-L-[methyl3HJmethionine was totally dependent upon RNA synthesis. The methylation reaction was completely inhibited by S-adenosyl-L-homocysteine, suggesting the transfer of only the methyl group of S-adenosy7-methionine to RNA products. Velocity sedimentation and hybridization of the in vitro product RNA indicated that both [3H]methyl and [32PJGMP labels resided in single-stranded 18S RNA molecules which were virus specific. Approximately 1 to 2 methyl groups were incorporated per RNA molecule. DEAE-cellulose chromatography of product RNA after alkaline hydrolysis suggested that the 5' terminus was the site of methylation.
Methylated nucleotides have recently been detected in messenger RNAs from mouse L cells (1) and from Novikoff hepatoma cells (2), as well as in a number of viral messenger RNAs synthesized in vitro by virion-associated RNA polymerases (RNA nucleotidyltransferases) (3-10). The methylated nucleotides in viral messenger RNAs synthesized in vitro by virions of cytoplasmic polyhedrosis virus (3), reovirus (4, 5), vaccinia virus (6), and vesicular stomatitis virus (7) are located in an unusual structure at the 5' terminus of the RNA molecules. The structure consists of a terminal base-methylated nucleotide linked to a ribose-methylated nucleotide through a 5' to 5' pyrophosphate bond, such as m7G(5')ppp(5')AmpNp or m7G(5')ppp(5')GmpNp (8-11). Methylation is required for proper processing of ribosomal RNA (12), for activation of the virion-associated RNA polymerase of cytoplasmic polyhedrosis virus (3), and for translation of viral messenger RNAs in a protein-synthesizing system in vitro (42). Newcastle disease virus (NDV), an avian paramyxovirus, has a single-stranded RNA genome with a sedimentation coefficient of 50 S (13, 14) and a molecular weight of 6 million (15). Messenger RNAs from polysomes of NDV-infected cells sediment primarily at 18 S, are complementary in base sequence to the viral genome, and contain polyadenylate sequences (16, 17). Purified NDV contains a virion-associated RNA polymerase (18) that transcribes the 50S RNA genome in vitro into 18S RNA molecules which are complementary in base sequence to the genome, and which also contain polyadenylate sequences (16). Since the paramyxovirus genome does not serve as messenger RNA (13, 19), the virionassociated transcriptase appears to be responsible for synthesis of messenger RNA in the infected cell (20, 21). Krauter Abbreviations: NDV, Newcastle disease virus; AdoHcy, S-adenosylL-homocysteine; AdoMet, S-adenosyl-L-methionine. 2611
and Consigli (22) have recently shown that methionine is essential for the synthesis of Newcastle disease virus in chicken embryo cells. In addition to its role in protein synthesis, methionine was found to be essential in methylation, and the RNA in purified virions of Newcastle disease virus was found to be methylated (22). In this report we demonstrate the presence of methyl transferase activity in purified NDV, which catalyzes the incorporation of the methyl group from S-adenosyl-L-methionine (AdoMet) into the messenger RNA synthesized in vitro by the virion-associated transcriptase. MATERIALS AND METHODS A strain of Newcastle disease virus, plaque purified from the Beaudette "C" strain described by Kingsbury (13, 23) and Granoff (24), was grown in embryonated chicken eggs (24) at 370 for 36 hr. Virions were purified from the allantoic fluid by differential centrifugation (13), rate zonal centrifugation (25), and isopynic banding in glycerol tartrate gradients (26). The standard reaction mixture in vitro (0.1 ml) for assaying both [32P]GMP and [3H]methyl incorporation contained 50 mM Tris-HCl (pH 7.8) (final reaction pH was 7.1), 120 mM NaCI, 0.4 mM MnCI2, 0.015% (v/v) Triton N101, 0.7 mM each of ATP, CTP, and UTP, 0.28 mM GTP, 3 mM dithiothreitol, 1 4uCi of [a-32P]GTP (15 Ci/mmol), and 0.78 MM (1 /,Ci) S-adenosyl-L-[methyl-3H]methionine ([3H]AdoMet) (12.6 Ci/mmol). Reaction mixtures were incubated with 35-45 gig of purified virions at 320 for 6 hr, the reactions were terminated by the addition of 0.5% sodium dodecyl sulfate, and the in vitro product RNA was isolated by sequential phenol extraction at pH 7 and 9 (27). Extracted RNA in the aqueous layers was pooled, precipitated with ethanol, and assayed for trichloroacetic acid-precipitable radioactivity (28). For further analysis the labeled RNA in vitro was purified free of unreacted triphosphates and anyremaining S-adenosyl-L-methionine by chromatography on a Sephadex G-50 (Pharmacia Fine Chemicals) column (1.5 X 30 cm) equilibrated with 0.01 M sodium acetate, 0.05 M NaCl, 0.5% sodium dodecyl sulfate (pH 5.0). The same buffer was used to elute 70 fractions (1 ml) of which 25-Ml aliquots were assayed directly for radioactivity in 9 ml of PCS (Amersham/ Searle). The fractions of the void volume that contained radioactivity were pooled, and the RNA pellet was resuspended in 0.5 ml of 5 mM Tris-HC1, 1 mM EDTA (pH 7.4). The RNA was layered over a 10-ml 15-30% (w/w) linear sucrose gradient in 5 mM Tris-HCl (pH 7.4), 1 mM EDTA, 0.5% (w/v) lithium dodecyl sulfate, and 0.1 M lithium chloride
2612
Biochemistry: Colonno and Stone
Proc. Nat. Acad. Sci. USA 72 (1975)
Table 1. Requirements for methylation and RNA synthesis Reaction mixture
[3H]Methyl cpm incorporation
[32P]GMP cpm incorporation
15,300 3,600 Complete* 15,700 0 -[3H]AdoMett 379 0 -NDV 531 67 -NaCl 239 10 -Triton N°'0 422 7 -CTP 568 4 -MnCl2 1,000 210 -Dithiothreitol * Standard reaction mixtures contained 1 ,ACi of [a -32P]GTP (15 Ci/mmol), 0.78 IAM S-adenosyl-L-[methyl-3H]methionine, and 44 jtg of Newcastle disease virus, as described in Materials and Methods. t When S-adenosyl-L-[methyl-3Hjmethionine was omitted, 50 mM Tris.HCl (pH 7.3) was used in place of 50 mM Tris.HCl (pH 7.8).
which had been formed over a 0.6-ml 60% (w/w) sucrose cushion. Gradients were centrifuged in a Beckman SW 41 rotor (200, 12 hr, 34,000 rev/min), and [14C]uridine-labeled 18 and 28S ribosomal RNAs from chicken embryo fibroblasts were centrifuged in a parallel gradient as markers. RNA sedimenting in the 18S region of the gradient was pooled and reprecipitated with ethanol. A portion of the precipitated RNA was used in hybridization experiments with 50S genomic RNA (23), and the remainder was hydrolyzed in 0.3 ml of 0.3 M KOH for 16 hr at 37'. After neutralization with 75 Ml of 1.2 M perchloric acid, the hydrolyzed RNA was chromatographed on a DEAE-cellulose (Whatman DE 32) column (29) using a pancreatic ribonuclease (29) digest of rat liver RNA as unlabeled oligonucleotide markers. Radioisotopes were purchased from New England Nuclear Corp.
RESULTS Dependence of Methylation upon Transcription. Incubation of purified Newcastle disease virus particles under optimal conditions for RNA synthesis in vitro resulted in the incorporation of radioactivity from S-adenosyl-L-[methyl3H]methionine into a trichloroacetic acid-precipitable product (Table 1). In contrast to cytoplasmic polyhedrosis virus (3), transcription of NDV was neither stimulated nor dependent upon a methyl donor such as S-adenosyl-L-methionine (Table 1, line 2). Methylation was dependent upon transcription, since methylation exhibited an absolute requirement for all components in the RNA polymerase reaction. In the absence of dithiothreitol, very low levels of methylase (5.6%) and transcriptase (6.5%) activity were detectable (Table 1, line 8). The optimum pH (7.1) and the optimum concentrations of sodium chloride (0.12 M) and of manganese (0.4 mM) were the same for [3H]methyl and for [32P]GMP incorporation (data not shown). Methylation and transcription proceed with either manganese or magnesium, but RNA synthesis is 2-fold higher in the presence of manganese (data not shown). The rates of incorporation of [3H]methyl and [32P]GMP into acid-insoluble products were identical (Fig. 1). Methylation and transcription proceeded at a linear rate for approximately 6 hr (Fig. 1), suggesting that the two processes may be coupled. Inhibition of Methylation by S-Adenosyl-L-homocysteine. Both RNA (30) and DNA (31) methylases of Esche-
10'p
3 0
x
2
E Q
5 3
4)
01
2
~0
1
I
4
8 12 16 20 TIME (HOURS)
24
FIG. 1. Time course of [32P]GMP and [3H]methyl incorporation. A 9-fold standard reaction mixture (0.9 ml), containing 9 'Ci of [a-32P]GTP (15 Ci/mmol), 0.78 1AM S-adenosyl-L-[methyl3H]methionine (12.6 Ci/mmol), and 0.38 mg of purified Newcastle disease virions, was incubated at 320. At the indicated time intervals, 0.1 ml was removed and the RNA was extracted with phenol/ dodecyl sulfate, precipitated with ethanol, and assayed for trichloroacetic acid-precipitable radioactivity (28). The counting efficiencies of 32p and 3H under our conditions were 91% and 41%, respectively. The specific activities of the isotopes were 71 cpm/pmol of [a-32P]GTP and 10,800 cpm/pmol of S-adenosyl-L-[methyl3H]methionine. Incorporation of [3H]methyl group (0) and [32P]GMP (0) into acid-insoluble products.
richia coli are subject to inhibition by the product of the reaction, S-adenosyl-L-homocysteine (AdoHcy). To test for an inhibition of methylation, varying amounts of S-adenosyl-L-homocysteine were added to incubation mixtures containing S-adenosyl-L-[methyl-3H]methionine and [a32P]GTP to determine the extent of methylation and RNA synthesis, respectively. RNA synthesis was not affected by S-adenosyl-L-homocysteine at concentrations that ranged from 1 MM to 0.5 mM (Fig. 2). However, [3H]methyl incorporation was reduced by increasing levels of S-adenosyl-L-
15
I
ptQ
o 2 x
1
E
10o0
C.' x
4
0)
0
0 1 5
-
-0
I
0.1
0.3
0.5
AdoHcy CONCENTRATION (mM). FIG. 2. Effect of S-adenosyl-L-homocysteine (AdoHcy) on methylation and RNA synthesis. Varying amounts of S-adenosylL-homocysteine were added to duplicate standard reactions containing 1 ;Ci of [a-32P]GTP (15 Ci/mmol), 0.78 pM S-adenosyl-L[methyl-3Hjmethionine (12.6 Ci/mmol), and 44 pg of purified Newcastle disease virus. After a 6-hr incubation at 320, the product RNA was extracted with phenol/dodecyl sulfate, precipitated with ethanol, and assayed for trichloroacetic acid-precipitable radioactivity (27, 28). Incorporation of [3H]methyl group (0) and [32P]GMP (0) into acid-insoluble products.
Proc. Nat. Acad. Sci. USA 72 (1975)
Biochemistry: Colonno and Stone
2613
Table 2. Hybridization of NDV [3H]methyl/ [32PJGMP-labeled product RNA 8
cpm Treated RNase-resistant cpm (%)
[3H]- [32p] F 6
:0C)
0
x
3
~0 3
E
4 x
0)
0
1 0
i
I
2 1
5
20 15 10 FRACTION NUMBER
Treatment*
methyl GMP
[
3H]methyl
[32P]GMP
"
25
FIG. 3. Sedimentation of labeled product RNA in vitro. A 20fold standard reaction mixture (2 ml) containing 20 gCi of [a32P]GTP (15 Ci/mmol), 0.78 MM S-adenosyl-L-[methyl-3H]methionine (12.6 Ci/mmol), and 0.72 mg of Newcastle disease virus was incubated for 6 hr at 320. After sequential phenol/dodecyl sulfate extraction at pH 7 and 9, gel filtration on Sephadex G-50, and ethanol precipitation, the product RNA was sedimented on a linear 15-30% sucrose gradient as described in Materials and Methods. Gradients were fractionated (0.4 ml) from the top with an ISCO gradient fraction collector, and 25-Ml aliquots of each fraction were diluted with 0.5 ml of H20 and 9 ml of PCS (Amersham/Searle), and assayed for [3Hlmethyl (0) and [32P]GMP (0) radioactivity. Positions of [14C]uridine-labeled 18 and 28S ribosomal RNA sedimented in a parallel gradient are indicated by arrows.
homocysteine, and at an adenosylhomocysteine to adenosylmethionine ratio of 26, inhibition was 50%. These results suggest that S-adenosyl-L-homocysteine is an end product of the reaction and that only the [3H]methyl group of S-adenosyl-L-[methyl-3H]methionine is incorporated into the RNA product. Velocity Sedimentation of Methyl-Labeled RNA. The products synthesized in the presence of S-adenosyl-L[methyl-3H]methionine and [a-32P]GTP were analyzed by rate-zonal sedimentation on sucrose gradients (Fig. 3). The RNA synthesized in vitro by NDV sedimented primarily at 18 S, the same size as the predominant mRNA from NDVinfected cells (17, 21). The radioactive profiles of [3H]methyl- and [32P]GMP-labeled RNAs coincided, but the ratio of 3H to 32P decreased with increasing sedimentation rate. These results indicate that the methyl groups were not distributed randomly, but that each polynucleotide chain might have the same number of methylated sites. When the molar ratio of methyl groups to GMP residues was calculated for the 18S in vitro product (Fig. 3, fraction no. 13), one methyl group was incorporated per 1315 nucleotides. Estimating the chain length of an 18S RNA molecule at 2100 (32), an average of 1.6 methyl residues were incorporated into each 18S molecule. Methyl-Labeled RNAs are Virus Specific. When the in vitro product synthesized in the presence of S-adenosyl-L[methyl-3H]methionine and [a-32P]GTP was subjected to ribonuclease digestion, both radioactive labels were rendered
Unannealedt
682
Self-annealedt
639
1284 1348
29 (4) 41 (6)
68 (5) 210 (16)
Annealed to 50S 1500 (100) * RNA synthesized in vitro sedimenting in the 14-25S region of sucrose gradients (Fig. 3, fractions 10-15) was annealed as described by Kingsbury et al. (28). t Product RNA was treated directly with 5 ,g/ml of RNase A for 30 min at 240. Product RNA was incubated under annealing conditions before RNase treatment. § Product RNA was annealed in 30 al to 0.32 jsg of purified 50S genomic RNA extracted from purified Newcastle disease virus.
RNAgenomes§ 705
1505
710 (100)
acid-soluble (Table 2), indicating that the radioactivity resided in single-stranded RNA and that methylation did not occur on the polyadenylate segments. After self-annealing, a small amount of the product was converted to a ribonuclease-resistant structure. When annealed with 50S RNA genomes, all of the radioactivity became ribonuclease-resistant, indicating that the [3H]methyl and [32P]GMP labels reside in RNA that is complementary in base sequence to the genome. Location of Methyl Groups in the RNA Molecule. The site of methylation catalyzed by specific methylases is often on the base moiety, and can be identified after alkaline hydrolysis of the RNA and chromatography of the methyl-substituted mononucleotides. Methylation can also occur on the ribose moiety of the nucleotide as 2'-O-methyl substitution. This modification renders the phosphodiester bond with the 3'-adjacent base resistant to alkaline hydrolysis, resulting in a methylated dinucleotide. DEAE-cellulose chromatography of alkaline hydrolysates was used to determine the location of methyl groups in the RNA molecules in vitro. The double-labeled RNA sedimenting in the 14-25S region of sucrose gradients (Fig. 3, fractions 10-18) was precipitated with ethanol and hydrolyzed with KOH. The alkaline hydrolysate was. neutralized and chromatographed on a DEAE-cellulose column, with a pancreatic ribonuclease digest of rat liver ribosomal RNA as oligonucleotide marker. This procedure separates oligonucleotides primarily on the basis of the negatively charged phosphate residues. Unreacted S-adenosyl-L-methionine alone does not bind to the column, while GTP elutes between diand trinucleotides (-3 and -4 charge). Results (Fig. 4) show that after alkaline hydrolysis, the 32P radioactivity eluted coincident with mononucleotide markers, indicating the completeness of the alkali digestion. Although the initial 5'terminal nucleotide has not been determined for NDV messenger RNA, results from studies on reovirus (33, 34, 4), vesicular stomatitis virus (35, 36), and vaccinia virus (10, 41) strongly suggest that the undigested NDV messenger RNA would contain ATP or GTP. The relative proportions of [a32P]GMP label in internal and terminal residues for an 18S RNA molecule would be approximately 500:1. If there are two GMP residues at the 5'-termini (as in m7GapppaG), the relative proportions would be 250:1. Sufficient radioactivity to detect 5'-terminal incorporation (>105 cpm) was not chromatographed on this column; thus all the 32P radioactiv-
2614
Biochemistry: Colonno and Stone
Proc. Nat. Acad. Sci. USA 72 (1975)
-I2 0
4
I
:c~) 0
0()
3
x
E
1
x
2
0
a) 5
I
20
100 80 60 40 FRACT ION NUMBER
120
FIG. 4. DEAE-cellulose chromatography of an alkaline digest of RNA synthesized in vitro in the presence of manganese. Double-labeled RNA, sedimenting in the 14-25S region of sucrose gradients (Fig. 3, fractions 10-15), was pooled and precipitated with ethanol. The RNA was then hydrolyzed in 0.3 ml of 3 M KOH for 16 hr at 370, and neutralized by the addition of 75 ul of 1.2 M perchloric acid. Neutralized hydrolysates were diluted with 9.7 ml of 7 M urea, 0.01 M Tris-HCl, 3 mM EDTA (pH 8), and 3 mg of RNasedigested rat liver ribosomal RNA was added as oligonucleotide marker. The digest was applied to a 0.6 X 17 cm column of DEAEcellulose and eluted with a 160-ml linear gradient from 0 to 0.25 M NaCl in 7 M urea, 0.01 M Tris-HCl, 3 mM EDTA (pH 8) and followed with 10 ml of 2 M NaCl in the same buffer. Fractions (1.4 ml) were collected at a flow rate of 5 ml/min. The absorbance at 260 nm (-----) of each fraction and conductivity of every fifth ) fraction was measured. The sodium chloride concentration ( was obtained from a standard curve. Each sample was placed into 15 ml of counting solution containing Triton X-100 (35), and the radioactivity was determined in a liquid scintillation counter. AdoMet (arrow), GTP (arrow), [3H]methyl label (0), and [32P]GMP label (0) are indicated.
ity detected (3 X 104 cpm) came from internal residues (nearest neighbors of GMP). All of the methyl-labeled material that bound to the column eluted between tri- and tetranucleotides (-4 and -5 charge) (0.144 M NaCl). Since no
methyl label was associated with the mononucleotide peak, base methylation of a single internal nucleotide can be excluded. Ribose methylation of a single internal nucleotide can also be excluded since no methyl label was associated with the dinucleotide peak. At most, only two methyl groups were detected per RNA molecule (Fig. 3). If these methyl groups occur as internal 2'-O-methylribose residues, one trinucleotide or two dinucleotides would be expected. Since the methyl-labeled material did not coincide with the trinucleotide marker, sequential methylation of internal ribose moieties can also be excluded. These results strongly suggest that the [3H]methyl groups reside in the 5' termini of NDV messenger
of S-adenosyl-L-[methyl-3H]methionine was purified, digested with alkali, and chromatographed on DEAE-cellulose. The results (not shown) indicated that methyl-labeled structures synthesized in the presence of magnesium or manganese (Fig. 4) elute in identical positions between triand tetranucleotides. Thus, under the various conditions used, the [3H]methyl label resided in virus-specific RNA, presumably at the 5' terminus after alkaline hydrolysis.
RNAs.
The methyl-labeled nucleotides synthesized in vitro by via structure that appears to differ by one negative charge from the 5' termini of other viral messenger RNAS (4-7, 10). Since the other viral messenger RNAs were synthesized in the presence of magnesium and our experiments used manganese, an additional experiment was performed to determine the influence of magnesium on the elution position of the methylated nucleosides synthesized by NDV. RNA synthesized in vitro by NDV in the presence of magnesium and a 10-fold higher concentration
rions of NDV reside in
DISCUSSION The data presented in this paper demonstrate that purified Newcastle disease virus contains a methyl transferase activity that transfers the methyl group from S-adenosyl-L-methionine to 18S RNA synthesized in vitro by the virion-associated transcriptase. Transfer of only the [3H]methyl group from S-adenosyl-L-[methyl-3H]methionine was inferred from the complete inhibition of [3H]methyl incorporation by S-adenosyl-L-homocysteine. Several lines of evidence presented in this paper demonstrate that the 8H radioactivity was incorporated into RNA. The [3H]methyl-labeled product, which was extractable with phenol and precipitable with trichloroacetic acid, sedimented along with [32P]GMPlabeled RNA and was sensitive to both ribonuclease A and alkaline hydrolysis. Conclusive proof that [3H]methyl groups resided in virus specific RNA was obtained by hybridization of the in vitro product to viral genomes. The extent of methylation was very low,. 1 to 2 methyl groups incorporated per 18S molecule. Initial studies on RNA synthesized in vitro by virions of cytoplasmic polyhedrosis virus (3), reovirus (4, 5), and vesicular stomatitis virus (7) also found 1 to 2 methyl groups per molecule. More detailed experiments have confirmed 2 methyl groups per.molecule for cytoplasmic polyhedrosis virus (8), reovirus (9), and vaccinia virus (10). The [3H]methyl radioactivity eluted from the DEAE-cellulose column between tri- and tetranucleotides. This elution position indicates that the [3H]methyl label is contained in a structure that may differ by one negative charge from the structure reported for other viral messenger RNAs (4-7, 10). A minor peak at this position from vaccinia virus RNA (6, 10) had a structure, m7G(5')pppNp, and was attributed to undermethylated RNA. Undermethylation does not appear to be responsible for the results presented in this paper. The amounts of S-adenosyl-L-[methyl-3H]methionine required to saturate the manganese reaction were determined, and twice the saturation concentration was used. The rates of methylation and transcription were identical over a 6-hr period (Fig. 1), testifying to the adequacy of the S-adenosylL-methionine concentration. It is not presently known if the NDV product is methylated in both ribose and base moieties. Structures that should elute in the position described include: pGmpNp, m7G(5')ppp(5')CGp or
m7G(5')ppGmpNp.
The role of methylation in eukaryotic messenger RNAs is not understood. The presence of S-adenosyl-L-methionine in vitro had no detectable influence on the rate of RNA synthesis by Newcastle disease virus, vaccinia virus (6), or reovirus (4). Price and Rottman (37) demonstrated that methylation does not interfere with the ability of oligonucleotides to direct the binding of aminoacyl-tRNA to ribosomes or with the ability of a partially methylated polynucleotide to direct the incorporation of amino acids into protein (38). Both, Banerjee, and Shatkin (42) obtained results that indicate that methylation of viral messenger RNA is required for translation in vitro.
Biochemistry: Colonno and Stone Methylation could be involved in the control of paramyxovirus replication. Genome transcripts in paramyxovirus-infected cells occur as 18S messenger RNAs (17, 39) and as 50S transcripts which are involved in genome replication (40). The mechanism that regulates synthesis of 18S RNA for translation and 50S RNA for genome replication is not known. It will be of interest to determine if the 50S transcripts are methylated. Transcription of the entire genome into single-stranded methylated viral RNA followed by postsynthetic processing into appropriate segments could occur. We thank Dr. Ron Borchardt for his advice and for the gifts of S-adenosyl-L-homocysteine, S-adenosyl-L-methionine, and S-adenosyl-L-[methyl-14C]methionine. We thank Dr. Aaron Shatkin, Dr. Amiya Banerjee, and Dr. D. H. L. Bishop for helpful discussions. We express our gratitude to Dr. Amiya Banerjee for a manuscript of ref. 7 and to Dr. Aaron Shatkin for a manuscript of ref. 42 prior to publication. The excellent technical assistance of Barbara Stone and of the Kansas University Enzyme Laboratory (Phyllis Shaffer) is gratefully acknowledged. This work was supported by Public Health Service Research Grant AI-11127 from the National Institute of Allergy and Infectious Diseases. 1. Perry, R. P. & Kelley, D. E. (1974) Cell 1, 37-42. 2. Desrosiers, R., Friderici, K. & Rottman, F. (1974) Proc. Nat. Acad. Sci. USA 71, 3971-3975. 3. Furuichi, Y. (1974) Nucleic Acids Res. 1, 809-822. 4. Shatkin, A. J. (1974) Proc. Nat. Acad. Sci. USA 71, 32043207. 5. Faust, M. & Millward, S. (1974) Nucleic Acids Res. 1, 17391752. 6. Wei, C. M. & Moss, B. (1974) Proc. Nat. Acad. Sci. USA 71, 3014-3018. 7. Rhodes, D. P., Moyer, S. A. & Banerjee, A. K. (1974) Cell 3,
327-33. 8. Furuichi, Y. & Miura, K. (1975) Nature 253,374-375. 9. Furuichi, Y., Morgan, M., Muthukrishnan, S. & Shatkin, A. J. (1975) Proc. Nat. Acad. Sci. USA 72,362-366. 10. Wei, C. M. & Moss, B. (1975) Proc. Nat. Acad. Sci. USA 72,
318-322. 11. Rottman, F., Shatkin, A. J. & Perry R. P. (1974) Cell 3, 197199. 12. Vaughan, M. H., Soeiro, R., Warner, J. R. & Darnell, J. E. (1967) Proc. Nat. Acad. Sci. USA 58,1527-1534. 13. Kingsbury, D. W. (1966) J. Mol. Biol. 18, 195-203.
Proc. Nat. Acad. Sci. USA 72 (1975)
2615
14. Duesberg, P. H. (1968) Proc. Nat. Acad. Sci. USA 60, 15111518. 15. Chi, Y. Y. & Bassel, A. R. (1974) J. Virol. 13, 1194-1199. 16. Weiss, S. R. & Bratt, M. A. (1974) J. Virol. 13, 1220-1230. 17. Bratt, M. A. & Robinson, W. S. (1967) J. Mol. Biol. 23, 1-21. 18. Huang, A. S., Baltimore, D. & Bratt, M. A. (1971) J. Virol. 7, 389-394. 19. Kingsbury, D. W. (1973) J. Virol. 12, 1020-1027. 20. Robinson, W. S. (1971) Virology 44,494-502. 21. Collins, B. S. & Bratt, M. A. (1973) Proc. Nat. Acad. Sci. USA 70,2544-2548. 22. Krauter, K. M. & Consigli, R. A. (1974) Am. J. Vet. Res. 35, 985-992. 23. Kingsbury, D. W. (1966) J. Mol. Biol. 18, 204-214. 24. Granoff, A. (1959) Virology 9,636-648. 25. Stone, H. O., Portner, A. & Kingsbury, D. W. (1971) J. Virol. 8, 174-180. 26. Obijeski, J. F., Marchenko, A. T., Bishop, D. H. L., Cann, B. W. & Murphy, F. A. (1974) J. Gen. Virol. 22,21-33. 27. Lee, S. Y., Mendecki, J. & Brawerman, G. (1971) Proc. Nat. Acad. Sci. USA 68,1331-1335. 28. Kingsbury, D. W., Portner, A. & Darlington, D. W. (1970) Virology 42,857-871. 29. Bishop, D. H. L., Mills, D. R. & Spiegelman, S. (1968) Biochemistry 7,3744-3753. 30. Hurwitz, J., Gold, M. & Anders, M. (1964) J. Biol. Chem. 239, 3474-3482. 31. Gold, M. & Hurwitz, J. (1964] J. Biol. Chem. 239,3858-3865. 32. Attardi, G. & Amaldi, F. (1970) Annu. Rev. Biochem. 39, 183-226. 33. Banerjee, A. K. & Shatkin, A. J. (1971] J. Mol. Biol. 61, 643653. 34. Miura, K., Watanabe, K., Sugiura, M. & Shatkin, A. J. (1974) Proc. Nat. Acad. Sci. USA 71, 3979-3983. 35. Roy, P. & Bishop, D. H. L. (1973) J. Virol. 11, 487-501. 36. Chang, S. H., Hefti, E., Obijeski, J. F. & Bishop, D. H. L. (1974) J. Virol. 13,652-661. 37. Price, A. R. & Rottman, F. (1970) Biochemistry 9,4524-4529. 38. Dunlap, B., Friderici, K. H. & Rottman, F. (1971) Biochemistry 10, 2581-2587. 39. Blair, C. D. & Robinson, W. S. (1968) Virology 35, 537-549. 40. Portner, A. & Kingsbury, D. W. (1972) Virology 47,711-725. 41. Urushibara, T., Furuichi, Y., Nishimura, C. & Miura, K. (1975) FEBS Lett. 49, 385-389. 42. Both, G. W., Banerjee, A. K. & Shatkin, A. J. (1975) Proc. Nat. Acad. Sci. USA 72, 1189-1193.
