Hotel Transylvania | Kosmiczna mistyfikacja / Alien Autopsy (2006) [Lektor] | Handa-kun English Subbed

Sequence studies of poliovirus RNA

Sequence studies of poliovirus RNA

J. Mol. Biol. (1975) 92, 467-477 Sequence Studies of Poliovirus RNA III?. Polyuridylic Acid and Polyadenylic Acid as Components of the Purified Polio...

861KB Sizes 1 Downloads 12 Views

J. Mol. Biol. (1975) 92, 467-477

Sequence Studies of Poliovirus RNA III?. Polyuridylic Acid and Polyadenylic Acid as Components of the Purified Poliovirus Rep&alive Intermediate YOSEUAKI Yoao$ AND ECKARD WIMMER$ Department of Microbiology St. Louis University School of Medicine St. Louis, MO., U.S.A. (Received 3 June 1974, and in revised form 29 November 1974) Minus strands and RNA with plus strand polarity

of the poliovirus replioative intermediate were exhaustively digested with RNAaae T, or RNAase A and the products analyzed for homopolyribonucleotides by affinity and ion-exchange chromatography. Minus strands were found to contain poly(U) covalently linked to the RNA; plus strand RNA of different chain length was found to contain poly(A). The homopolyribonucleotides do not originate from contamixmting free single-stranded polio RNA or polio replicative form, as shown by sucrose gradient centrifugetion and gel electrophoresis. The chain length of the poly(U) (60 to 120 nucleotides) is heterogeneous and similar to that of poly(A) of virion RNA, suggesting a mechsnism of transcription for poly(A). Whether or not the poly(A)-linked plus strands in the replicative intermediate are a structural component of this intermediate or an artifact of isolation procedures remains unknown. In this respect the “single-stranded” and “double-stranded” models for the replicative intermediate ere discussed.

1. Introduction Poliovirus double-stranded RNA (the replicative form, or RFII) has recently been shown by us to contain poly(A) in the plus (genome-like) strand and poly(U) in the minus strand (Yogo t Wimmer, 1973). We have interpreted this finding to indicate that the 3’-terminal poly(A) found in virion RNA (Yoga & Wjmmer, 1972) might be genetically coded and not synthesized by post-transcriptional addition. Since polio RF is product rather than an intermediate of polio RNA replication (Yogo & Wimmer, 1973, and literature cited therein), poly(U) in the mjnus strand might have originated from a modification of the double-stranded RNA molecule in the cytoplasm, for example, by the transcription of the poly(A) in the plus strand after RF synthesis. We have, therefore, analyzed the true intermediate of polio RNA synthesis, which is a multistranded RNA molecule called replicative intermediate (for a review see: t Paper II in this series is Yogo & Wimmer, 1972. $ Present address: Department of Viral Infeotion, Institute of Medical Science, Tokyo, 108 Japan. 8 To whom correspondence should be addressed: Department of Microbiology, School of Basic Health Sciences, State University of New York, Stony Brook, N.Y. 11794, U.S.A. 11Abbreviations used: RF, replioative form (virus-speoifio, double-stranded RNA); RI, replioative intermediate; ssRNA, single-stranded RNA; &ion-poly(A), RI-poly(A), RF-poly(A) : poly(A) isolated from virion RNA, RI or RF, respectively. 467

468

Y.

YOGO

AND

E. WIMMER

Baltimore, 1969; Bishop & Levintow, 1971), the structure of which is not fully We report here the presence of poly(U) in the minus strands of polio understood. RI, supporting a mechanism of transcription of poly(A) from poly(U) in viral RNA replication. Furthermore, we have found poly(A) in plus strand RNA of the RI. This finding is of interest in respect to the structure of the replicative intermediate.

2. Materials and Methods (a) holation

of polio

RI

Polio RI was isolated from HeLa cells infected with poliovirus, type 1 (Mahoney) which were suspended 1.5 h after infection in phosphate-free medium containing 5 rg actinomycin D/ml and 05 mCi carrier-free ““P/ml (Wimmer, 1972). A cytoplasmic extract was prepared 3.5 h after infection and treated with phenol/chloroform as previously described (Yogo & Wimmer, 1973). Total cytoplasmic RNA of 4 x 1Oe infected cells was separated into single-stranded and double-stranded RNA by incubating it for 12 h at 0°C in 2 ml of 2 M-LiCl, 0.05 M-NaCl, 5 mu-Tris*HCl (pH 7*5), 1 m-EDTA (Baltimore, 1968). The RNA which precipitated under these conditions was dissolved in 1 ml of 0.1 M-NaCI, 0.01 M-Tris*HCl (pH 7*5), 1 mM-EDTA, containing 0.5% sodium dodecyl sulfate and fractionated in the same buffer in a column (I.5 cm x 70 cm) of Sepharose 2B (Pharmacia Fine Chemicals); polio RI elutes under these oonditions in the void volume (Baltimore, 1968). Polio RI was precipitated in the presenoe of 8 A,,, units yeast transfer RNA and further purified either by a second gel filtration step on Sepharose 2B (column dimensions: 0.8 cm x 70 cm) or by sucrose density-gradient centrifugation (15 to 30% sucrose, 0.1 M-NaCl, 0.01 M-Tris*HCI (pH 7.5), 1 m&f-EDTA, 0.5% sodium dodeoyl sulfate) in a Spinco SW41 rotor for 45 h at 41,000 revs/min and 24%. (b) Isolation

of poly (U)

fTom minus

strands

Polio [32P]RI, 0.5 to 1 x lo6 cts/min (speo. act. ranging from 0.2 to 1 x lo8 cts/min per pg) was denatured in 2 ml of 1 ma-EDTA (pH 7.5), containing 1 rg poly(U) (Miles Laboratories) for 2 min at 100°C. The solution was rapidly cooled to 0°C and mixed with O-2 ml of 0.2 M-Tris.HCl (pH 7.5), 0.01 M-EDTA, 5 Azso units tRNA and 200 units RNAase T, (Calbiochem). After incubation for 30 min at 37°C the solution wsa cooled to room temperature, mixed with 125 ~1 of 10% sodium dodecyl sulfate, 125 ~1 of 20 x binding buffer (binding buffer is 0.12 M-NaCl, 0.01 M-Tris.HCl, pH 7.4) and 10 Azeo units tRNA and applied to a column of poly(A)-Sepharose 4B (bed volume : 1 to 2 ml). Poly(A)-Sepharose 4B w&9 prepared from CNBr-activated Sepharose 4B (Pharmacia Fine Chemicals) and poly(A) (Miles Laboratories) as previously described (Yogo & Wimmer, 1973). The digest of viral RNA was step-wise eluted from poly(A)-Sepharose as follows : wash I, 9 ml binding buffer containing 0.5% sodium dodecyl sulfate and tRNA (1 A,,, unit/ml) at 24°C; wash II, 8 ml of 1 mM-EDTA, pH 7.5 (degassed) at 24°C; wash III, 9 ml of 85% dimethyl sulfoxide, 0.15 mM-EDTA at 38°C; wash IV, 18 ml of 85% dimethyl sulfoxide, O-15 mMEDTA, 0.1 M-NaCl at 80°C. In contrast to previous experiments (Yogo & Wimmer, 1973) we found that it was necessary to add detergent to wash I and to include NaCl in wash IV in order to prevent the unspecific binding of polynucleotides and the irreversible binding of poly(U), respectively. The change of the properties of the poly(A)-Sepharose 4B might have been caused by the fact that the content of CNBr groups in activated Sepharose 4B has recently been increased by the manufacturer (communication from Pharmacia Fine Chemicals, Inc.). Our recent batches of poly(A)-Sepharose may therefore contain an increased amount of weak positively charged groups (produced during inactivation of CNBr groups) and a larger content of coupled poly(A). Fractions (approx. 1.8 ml) were collected into small glass vials and monitored for Cerenkov radiation (radiation arising from fast p-particles in aqueous solutions, counted in the tritium channel of a Beckman scintillation counter). Those fractions in wash IV which contained radioactive material were combined and mixed with 1/20th vol. 4 M-NaCl and 5 Azao units tRNA. The polynucleotides were precipitated with 2 vol. ethanol and pelleted in a Polyallomer tube in the

POLY(U)

AND

POLY(A)

IN

POLIOVIRUS

RI

469

Spinco SW27 rotor by centrifugation at 16,000 revs/min for 12 h. Uridine-rich fragments were further purified and characterized by ion-exchange chromatography in 7 M-Urea (Yogo & Wimmer, 1972) and by gel electrophoresis (see below). The base ratio of the U-rich polynucleotides was determined by high-voltage eleotrophoresis of alkaline digests on Whatman 3MM paper in 5% acetic acid, 0.5% pyridine (v/v) (pH 3*5), 1 mbn-EDTA. Minus strands of polio [3aP]RI were separated from RNA of plus strand polarity by annealing an excess unlabeled virion RNA to denatured RI under conditions previously described (Yogo & Wimmer, 1973). The RNA species were then separated on Sepharose 2B (column dimensions: O-8 cm x 70 cm) and the double-stranded RNA and singlestranded RNA analyzed for [3aP]poly(U) &8 described above.

(c) Isolation

of poly(A)

from

polio [32P]RI

Polio [32P]RI was heat-denatured in the presence of 0.2 pg poly(A)/ml and digested with RNAase T, under conditions previously described for polio [32P]RF (Yogo & Wimmer, 1973). The digest was passed through a poly(U) filter in binding buffer (Sheldon et al., 1972); the filter was washed and the bound polynucleotides eluted with 2 ml of 0.5% sodium dodecyl sulfate, 0.02 M-TriseHCl (pH 7*4), 1 mm-EDTA, containing 0.25 AQ6e units poly(A)/ml for 5 min at 65°C. The extraction was repeated twice. The material in the combined extracts was precipitated in the presence of carrier tRNA and further analyzed by ion-exchange chromatography in 7 M-urea or by gel electrophoresis. RNA from RI with plus strand polarity was isolated as described in section (b) above and analyzed for the presence of poly(A) as outlined above. A quantitative determination of the content of poly(A) in polio RI was carried out as follows: 4x lo5 cts/min [32P]RI together with in 2 ml of 1 mM-EDTA containing 0.5 rg poly(U), at 5 Am units tRNA were denatured 100°C for 2 min. The solution was rapidly cooled to O”C, adjusted to 0.2 M in NaCl, 0.02 M in Tris.HCl, pH 7.5 and incubated with RNAase A (10 pg/ml) and RNAase T, (100 units/ml) for 45 min at 37°C. The mixture was then mixed with urea to a concentration of 7 M, applied to a column of DEAE-Sephadex A25 (bed volume: 2 ml) and the column washed with 15 ml of 7 M-urea, 0.3 M-Nacl, 0.02 M-Tris.HCl (pH 7.5), 2 mM-EDTA. Polynucleotides bound to the column were eluted with 7 M-UT0&, 1.2 M-NaCI, 0.02 M-Tris . HCl (pH 7*5), 2 mm-EDTA. They were then precipitated from elution solvent (usually 3.2 ml) with 2 Azeo units tRNA and 6.5 ml ethanol and sedimented at 4°C for 15 h in a Polyallomer tube at 15,000 revs/min using a Spinco SW41 rotor. The whole procedure--heat denaturation, enzymatic digestion and column separation-was repeated. [32P]poly(A) thus purified was identified by digesting it together with 3 A,,, unit,s of carrier tRNA in 20 ~1 of 0.04 M-ammonium acetate (pH 4.5), 1 mM-EDTA, with 20 units RNAaso Tz (Calbiochem) for 6.5 h at 37°C and analyzing the products .by high-voltage paper electrophoresis. For the determination of the size of the RNA which contains poly(A), 6 x lo4 cts/min (approx. 0.1 rg) of polio [32P]RI was denatured in 2 ml of 1 mM-EDTA at 100°C; the solution was then mixed with 0.18 ml of 5 M-NaCl and incubated at 70°C for 15 min. The [32P]RNA was precipitated in the presence of 10 A,,, units tRNA, dissolved in 0.1 ml of 0.1 M-NaCl, 0.01 M-Tris*HCl (pH 7.5), 1 mM-EDTA, 0.5% sodium dodecyl sulfate and centrifuged through a sucrose gradient aa described above. 50 ~1 of each fraction of the gradient were analyzed for trichloroacetic acid-precipitable counts before and after incubation with 0.5 ml of 0.3 M-NaCl, 0.03 M-sodium citrate (pH 7.5) containing 55 ~1 RNAase A/ml, for 1 h at 25°C. Gradient fractions were combined and samples thereof mixed with 2 vol. binding buffer. The amount of poly(A)-containing RNA was determined by poly(U) filter binding as described by Sheldon et al. (1972). Under the conditions used (in 0.12 nl-NaCl) an average of 25% of different preparations of virion RNA but less t,han 0.5% of 28 S [3H]rRNA or 18 S [3H]rRNA are retained by poly(U) filters. (d) Polyacrylamide

electrophoresie

High molecular weight RNAs were analyzed in 0.5% Agarose, 1.8% polyacrylamide gels that were prepared by a modified procedure of Peacock & Dingman (1968) as follows: 0.25 g Agarose (Seakem, Sargent) in 25 ml water was heated in a steam bath for 30 min and then kept at 56°C (solution A) ; 3.0 ml of 30% acrylamide, 0.055 ml ethylene diacrylate,

470

Y.

YOGO

AND

E. WIMMER

6 ml glycerol, 10 ml of 5 x buffer E (0.18 nr-Tris.HCl, 0.15 al-phosphate, 0.005 I~~-EDTA. pH 7.6) and 6.7 ml water were degassed and mixed with 0.04 ml n;,N,N’,N’-tetramethS-I ethylene diamine (solution B). Solution B was poured into solution A, 0.25 ml of loo0 ammonium persulfate was added and the mixture filled into plastic eloctrophoresis tubes (0.6 cm x 9 cm) as quickly as possible. One h later the gels were topped with buffer E containing 10% glycerol, and stored at 5°C. For electrophoresis t,he gels were invsrt,ctL to yield a perfectly even surface, and prerun for 30 min with buffer E conbaining 0.20/, sodium dodecyl sulfate and 10% glycerol. Electrophoresis was carried out for the indicated times at 5 mA/gel. Poly(A) or poly(U) were analyzed in 10% polyacrylamide/0.25:b ethylene diacrylate gels in buffer E containing 0.2% sodium dodecyl sulfate and 10% glycerol. Gels were sliced with a Joyce-Loebl gel slicer, incubated with 0.2 ml concentrat,ed ammonia overnight and counted in 2 ml scintilletor, or scanned for absorbance (Yogo CE: Wimmer, 1972).

3. Results and Discussion (a) Properties of preparations of the relplicative intermediate The intermediate in the replication of poliovirus RNA has been shown to be a multistranded structure which, after isolation from the cytoplasm of infected cells, is partially double-stranded and partially single-stranded (for a review, see Bishop & Levintow, 1971). For our study it was important to establish that the preparations of RI were free of polio RF and poliovirus-specific ssRNA since ssRNA contains poly(A) and RF contains poly(A) and poly(U) (Yogo & Wimmer, 1972,1973). Gel filtration separates labeled viral RNA which is insoluble in 2 Br-LiCl into a high molecular weight RNA and RNA with lower molecular weight (Fig. l(a)). Based on their resistance to pancreatic RNAase, their sedimentation values and their migration in polyacrylamide gels (see below) the excluded material (fractions 19 to 24) has been identified as polio RI while the included material (fraction 27 to

600 -

20

40

30

50

60

Fraction no. (b)

FIU. 1. Gel filtration of virus-specifio polio [3aP]RNA, isolated from a Sepharose 2B. (a) Separation of RNA which is insoluble in 2 M-LiCl, into ssRNA (fractions 25 to 60) and low molecular weight material (fractions fractions 19 to 24 of (a) on Sepharose 2B. For column dimensions, see Shaded areas were used in analyses for poly(A) and poly(U).

cytoplasmio extract, on RI (fractions 19 to 24), 61 to 72). (b) Rerun of Materials and Methods.

POLY(U)

AND

POLY(A)

IN

POLIOVIRUS

RI

471

50) represents polio ssRNA. Rerun of the fractions 19 to 24 (Fig. l(a)) indicated that the RI is essentially free of ssRNA (Fig. 1(b)). Most of our experiments were carried out with the three front fractions of the excluded RNA (corresponding to fractions 19 to 21 in Fig. l(a)). This material sediments through sucrose gradients as heterogeneous material and is partially sensitive to RNAase A (Fig. 2) characteristic for polio RI (Baltimore, 1968; Bishop et al., 1969; Savage et al., 1971). The sedimentation profle does not change on repeated sucrose centrifugation (results not shown) indicating that the heterogeneity of the material is not due to aggregation occurring during precipitations under our conditions.

5

IO

15 Fraction

20

25

no

FIG. 2. Zoned centrifugation of [3aP]R1 through a 16% to 30% ~uorose gradient (for aonditions and Methods). Cerenkov radiation represents the total oounta in freotions (approx. material (prior 460 pl) of the gradient, the 3aP oounts represent triohloroaaetio a&d-preoipitable to and after RNAA treatment) of a sample thereof (4 4). All labeled material was acidpreoipitable. Polio [aaP]ssRNA (fractions 28 to 32 of Fig. l(m)), which was found to sediment through suarose and migrete in agarose/polyaorylamide gela like virion RNA, and 18 S rasP]RF (Yogo & Wimmer, 1973) were used aa markers. -o-o-, cerenkov (cts/min x lo-‘); -•--•--, aaP (cts/min x lo-“); --A-A--, RNAase-resistant.

seeMateriels

When fractions of polio RI larger than 25 S are analyzed by electrophoresis in agarose/polyacrylamide gels, all of the radioactive material migrates slower than polio RF (Fig. 3(a)). Thus the preparation does not contain RF or free ssRNA. While most of the RI migrates as a rather homogeneous peak some material does not enter the gel, which might indicate aggregation occurring at the onset of the electrophoresis. The migration of the single-stranded RNAs in these gels is a linear function of the log of their respective molecular weights; as expected, polio RF or polio RI do not follow this function (see insert to Fig. 3(a)) as the electrophoretic behavior of polynucleotides of high molecular weight depends on their secondary structure (compare Fisher & Dingman, 1971). Polio RI as well as virus-specific ssRNA are sensitive to very low concentrations of RNA&se A (0.1 pg/ml at 0.3 u-Nail) while polio RF is stable (Fig. 3(b)). Not all SSRNA of the RI, however, appears to have been removed under these conditions as the double-stranded core of the RI migrates slower than RF.

472

Y. YOGO

AND

E. WIMMER

(a) RI

RF

600 20 40 60 8, M~qrotfon (mm)

000 600

10

44 h&A L 30 20

=0-~~61BBW688888eb-

40 50 60 70 Dlstonce of migration (mm)

80

FIG. 3. Electrophoresis of 3aP-labeled virus-specific RNAs in 0.6% agarose/1.8% polyacrylamide gels. (a) Polio RI (shaded area of Fig. 2), ssRNA (Fig. l(a)) and RF were separated under conditions described in Materials and Methods. sH-labeled ribosomal RNA from HeLa cells was run as marker together with the ssRNA. The graph is a composition of three gels. The insert demonstrates the relation between the log of the moleoular weights and the distance of migration of the RNAs. (The heterogeneity of the RI is indicated by the plane outlined by the little arrows, the peak of RI being located at the origin of the plane.) (b) Separation of the RN,45 as in (a) but after treatment with RNAase A (0.1 pg/ml) in 0.3 M-NaCl, 0.01 iu-Tris . HCl (pH 7.6), 1 mM-EDTA for 30 min at room temperature.

(b) Polyuridylic

acid in minus strands of polio

replica&e intermediate

Polio [32P]RI, after heat denaturation in the presence of unlabeled poly(U), was digested with RNAase T, and the digest fractionated by aflinity chromatography on poly(A)-Sepharose as previously described (Yogo & Wimmer, 1973). Some material (0.95% of the total radioactivity) was bound to the column, which after further purification by ion-exchange chromatography consisted of more than 90% uridine nucleotides. The origin of the U-rich fragment was then determined through an analysis of minus and plus strands of polio RI using the same strategy applied to polio [32P]RF (Yogo t Wimmer, 1973): polio RI was heat-denatured, reannealed in the presence of an excess of unlabeled virion RNA, and the resulting hybrid double-stranded RNA separated by gel filtration from labeled material of plus strand polarity. Minus [3aP]RNA of the hybrid, [32P]RNA with plus strand polarity

POLY(U)

AND

POLY(A)

IN

POLIOVIRUS

473

RI

from the RI, and poliovirus [32P]ssRNA which was isolated from the cytoplasm of HeLa cells 3.5 hours after infection were then analyzed for a labeled U-rich sequence under identical conditions. As can be seen in Figure 4(a)? 2.2% of the total RNAase T, digest of hybrid double-stranded RNA binds to poly(A)-Sepharose, while only 0.32 and 0*22:/, of RI plus strand and of viral RNA, respectively (Fig. 4(b) and (c)), are retarded. The material with affinity to poly(A)-Sepharose (Fig. 4(a)), when

Fraction no. (a)

(b)

(cl

FIG. 4. Separation of uridine-riah polynucleotides from digests of [3aP]RNA with RNAase T, by affinity ohromatography on poly(A)-Sepharose. For the conditions and the components of wash I to IV see Materials and Methods. (a) Digest of minus [32P]RNA of RI; (b) digest of [3zP]RNA of plus polarity of RI; (a) digest of [3aP]ssRNA (compare legend to Fig. 2).

I 0.9

0.6

-0.3

E s -0 p” -0.9

G 5 i ‘0 s

- 0.6

0

~“-oOsAomk”.-.mb..” 20 IO

30

Fraction no.

Fm. 6. Ion exah&nge chromatography of 3aP-labeled poly(U) (a) and poly(A) (b) from polio RI 7 ~-urea. Open circles represent elution of marker tRNA. For details, 888 text. -O-O-, ) NeCl concentration (M). saP (cts/min); --O--O--, Azao nm; ( in

474

Y.

YOGO

AND

E. WIMMER

chromatographed on DEAE-Sephadex (Yogo & Wimmer, 1972) as shown in Figure 5(a), eluted at a higher salt concentration than tRNA; its base composition (Up =poly 90.4x, Ap = 5.6%, Cp = 2*4%, and Gp = l-S:/,) reveals it as uridine-rich nucleotide. The content of bases other than Up in the fragment, particularly that, of Ap, is probably due to traces of contaminating fragments such as internal A-rich oligonucleotides. We have observed similar contaminations in poly(U) of polio RF which could be removed only through two cycles of heat denaturation, enzymic digestion and affinity chromatography (Yogo et al., 1974). Similarly, the small amount of material bound to poly(A)-Sepharose in Figure 4(b) and (c) might be due to aggregates of [32P]poly(A) fragments to excess unlabeled poly(U), which was added prior to the enzymatic hydrolysis, although base analyses of the materials were not perf’ormed. The sequence of the U-rich fragment in RI minus strands is currently under investigation.

(b)

tRNA

30 A

0

IO 20

30 40

,J 50 60

70 80

Distance of migrotion(mml

Fm. 6. Eleotrophoretio migration of homopolymers in 10% polyaorylamide gels. (a) [3aP]poly(U. from polio RI; (b) [3aP]poly(A) from plus strrtnd material of polio RI; (c) [3aP]poly(A) from ssRNA Arrow indioates the position of marker tRNA which wzw determined by scanning the gels at 280 nm prior to cutting them into l-mm slioes. The large amount of material at the origin, which we often observe, is attributed to aggregations between the labeled fragment and the added homopolyribonuGleotidea.

POLY(U)

AND

POLY(A)

IN

POLIOVIRUS

RI

476

The presence of [32P]poly(U) in the hybrid double-stranded RNA indicates that the homopolyribonucleotide is covalently bound to the minus strand. The size of the poly(U) of polio RI is heterogeneous and larger than that of tRNA, as shown by ion exchange chromatography (Fig. 5(a)) and by polyacrylamide gel electrophoresis (Fig. 6(a)). We found in some preparations, however, that 60% of the poly(U) fiagments elute as a sharp peak slightly ahead of tRNA from the ion-exchange column while 40% elute aa a broad band similarly to the poly(U) in Figure 5(a). The shorter poly(U) resembles in base composition (Up = 93*8%, Ap = O*S%, Cp = 3-8%, Gp = 1.6%) and size (average chain length 60 Up residues) the poly(U) as isolated from polio RF (Yogo et al., 1974). We have no data as yet which would explain the varying sizes of the homopolymers poly(A) or poly(U) in all forms of polio-virusspecific RNAs (RI, RF, or ssRNA) isolated from infected oells. The existence of poly(U) in minus strands of polio RI with an average chain length similar to that of &ion-poly(A) suggests that transcription is the mechanism of poly(A) synthesis in polio RNA replication. In support of this mechanism is our failure to detect adenylase polymerase or terminal adenylate transferase activities in a partially purified polio RNA replication complex which synthesizes poly(A)containing viral RNA in the presence of magnesium ions (Hiisler et al., 1973). It should be pointed out that the partially purified replication complex does not synthesize free poly(A) even when manganese ions and oligo(A) primer are added (Dorsch-HLsler t Wimmer, unpublished results). The position of the poly(U) in RI minus strands remains to be determined. Base composition of the poly(U) fragment of polio RF obtained after digestion with RNAase T, [(Uz, C,)Gp] and end-group labeling with polynucleotide kinase or by the borohydrate method have indicated that the homopolymer is 5’-terminal in RF minus strands (Yogo et a-!., 1974) and not 3’-terminal, as originally concluded from studies with venom exonuclease (Yogo & Wimmer, 1973). It is, therefore, likely that poly(U) is the 5’-end of the minus strands also in polio RI which makes unnecessary speculations of cyclic structures involved in polio RNA replication (Yogo & Wimmer, 1973). The long segment of poly(U) in polio minus strands which has been detected also by D. H. Spector & D. Baltimore (personal communication) offers itself as a structural marker in studies of the biosynthesis and turnover of minus strands in vivo or in vitro. (c) Polyadenylic acid in polio replicative intermediate Polio [32P]RI when analyzed for poly(A) under conditions described for polio RF (Yogo & Wimmer, 1973) yields a polynucleotide which is resistant to RNAase T, and binds to a poly(U)-fiberglass filter (Sheldon et al., 1972). We have therefore isolated RNA with plus strand polarity from [32P]RI (see Materials and Methods) and found after digestion with RNAase T, and poly(U)-filter binding, poly(A) that elutes from DEAE-Sephadex at higher salt concentration than tRNA (Fig. 5(b)) and has a base composition of: Ap = 90*4%, Up = 5-l%, Cp = 4-l%, Gp = 0.4%. A hybrid of labeled minus strands and unlabeled virion RNA (see section (b) above) did not yield this polynucleotide (data not shown). Thus, poly(A) is absent in minus strands of RI as in the case of polio RF (Yogo & Wimmer, 1973). When electrophoresed in a polyacrylamide gel, RI-poly(A) is heterogeneous in size and migrates slower than tRNA (Fig. 6(b)). Its size distribution is similar to poly(A) isolated from virion RNA (virion-poly(A); Yogo 6 Wimmer, 1972). This is interesting inview of our 51

476

Y.

YOGO

AND

E. WIMMER

previous finding that poly(A) of polio RF is twice as long as virion-poly(d) (Yogo & Wimmer, 1973). As will be discussed below, poly(A) present in polio RI could be the result of an aggregation of virus-specific ssRNA to RI. We therefore determined the content of poly(A) in RI by a method previously described (Yogo & Wimmer, 1972). [32P]RI, with a sedimentation of 25 S and higher, was heat-denatured and digested with RNAase A and RNAase T, ; the homopolymer was then separated from the oligonucleotides by ion-exchange chromatography in 7 M-urea (experimental details are described in Materials and Methods). After two cycles of enzymatic digestion and of chromatography, pure poly(A) was obtained (base composition: Ap = 99.53%, Up = 0.35%, Cp = O.ll%, Gp = 0%) which represented 0.4% of the total digest (two experiments). A different value (0.56%) was obtained when poly(A) was isolated by poly(U)-filter binding followed by ion-exchange chromatography. This value, however, is probably too high as the poly(A) contained nucleotides (14%) other than AP (86%). If polio RNA is synthesized from 5’ to 3’, the poly(A)-containing RNA can be expected to be of the same size as genome RNA, since poly(A) is 3’-terminal. However, when polio [32P]RI was heat-denatured at low concentration (to prevent reannealing) and separated through a sucrose gradient, poly(A)-linked RNA (as defined by its capability to bind to poly(U)-filters (see Materials and Methods)) was found to be among all size classes in the gradient. Half of the poly(A)-containing RNA sedimented at 35 S, the other half between 18 and 35 S. Since the labeling was carried out in the presence of actinomycin, all poly(A)-containing RNAs must be virusspecific. The presence of poly(A)-linked RNA smaller than 35 S might indicate either degradation of genome-size strands in RI or aggregation of fragmented ssRNA to RI occurring during isolation procedures of the multistranded RNA. The origin of the poly(A)-containing RNA in polio RI is not unknown. If the nascent strands of the RI were bound to the template by long stretches of base-pairs and were replaced by the growing points of the following strand in a semiconservative manner (Baltimore 1969; Bishop & Levintow, 1971) each RI molecule would contain one complete template strand and one complete complementary strand, in addition model is consistent with the to several incomplete strands. This “double-stranded” identification of poly(A) and poly(U) in polio RI. Evidence has been presented, on the other hand, which favors a “single-stranded” structure for picornavirus RI in vivo. This model describes the RI as a ribonucleoprotein which is essentially single-stranded, the RNA strands being held together by RNA polymerase molecules and very short regions of base-pairs (i)bexg t Philipson, 1971; Thach et al., 1974). The single-stranded model, however, neither explains the high efficiency with which polio RI is infectious (Koch, 1973) nor the large amount of poly(A) found in RI. We estimate that the content of poly(A) in single-stranded RI would not exceed 0.05% assuming (i) three to six growing points randomly distributed along the template; (ii) identical rate of RNA synthesis and (iii) 10% “negative RI” in the preparation (Baltimore, 1969; Bishop $ Levintow, 1971). In contrast, the amount of poly(A) in RI which we observed is much higher (0.4%). The number of nascent strands in RI can be calculated from the ribonuclease resistance of the partially double-stranded RNA (Baltimore, 1969). We found that the average resistance to ribonuclease of several of our RI preparations is 56%, on which basis 3.2 nascent strands per RI molecule can be calculated. Using this value,

POLY(U)

AND

POLY(A)

IN

POLIOVIRUS

RI

477

together with the content of poly(A) in RI, we conclude that there must be at least one poly(A)-segment per RI molecule. Contaminations of our RI preparations with free ssRNA or RF can be excluded (compare Fig. l(b) and Fig. 3(a)). If polio RNA were synthesized via a single-stranded RI, the RI-poly(A) which we identified, would indicate an artifact of the isolation procedure, namely that during phenolation of the cytoplasmic extract, free ssRNA is bound to the replicative intermediate. The long sequence of 60 to 120 uridylic acid residues in minus strand RNA represents a site for a rapid association with the poly(A) segment of plus strand RNA during deproteinization. Thus, apart from its suggested role as template for poly(A) synthesis, poly(U) might explain the presence of poly(A)-linked viral RNA in purified molecules of polio RI. We thank Robert E. Thach and David Baltimore for suggestions in preparing the manuscript and Ruth Ann Miller for excellent technical assistance. This work was aided by grants CA14151 from the U.S. Public Health Service and DRG1164 from the Damon Runyon Memorial Fund for Cancer Research. REFERENCES Baltimore, D. (1968). J. Mol. Bid. 32, 359-368. Baltimore, I). (1969). In The Biochemistry of Viruses (Levy, H. B., ed.), pp. 103-176, Marcel1 Dekker, New York. Bishop, J. M. & Levintow, L. (1971). Progr. Med. V’irol. 13, I-82. Bishop, J. M., Koch, G., Evans, B. & Merriman, M. (1969). J. Mol. Biol. 46, 235-249. Fisher, M. P. 6: Dingman, C. W. (1971). Biochemktry, 10, 18951899. Hasler. K., Yogo, Y. & Wimmer, E. (1973). Bacterial. Proc. p. 202. Koch, G. (1973). Current Topics in Microbiology and Immunology, 62, 89-138. oberg, B. & Philipson, L. (1971). J. Mol. Biol. 58, 725737. Peacock, A. C. & Dingman, C. W. (1968). Biochemistry, 7, 668-674. Savage, T., Granboulan, N. & Girard, M. (1971). Biochemie, 53, 533-543. Sheldon, R., Jurale, C. & Kates, J. (1972). Proc. Nat. Acud. Sci., U.S.A. 69, 417-421. Thach, S. S., Dobbertin, D., Lawrence, D., Golini, F. & Thach, R. E. (1974). Proc. Nat. Acad.

Sci.,

U.S.A.

71, 2549-2553.

Wimmer, E. (1972). J. Mol. Biol. 68, 537-540. Yogo, Y. & Wimmer, E. (1972). Proc. Nat. Ad. Sk., U.S.A. 69, 1877-1882. Yogo, Y. & Wimmer, E. (1973). Nature New Bid. 242, 171-174. Yogo, Y., Teng, M. H. & Wimmer, E. (1974). Biochem. Biophys. Commun. 64,1101-1110.