The Properties of the Sendai Virus Ribonucleoproteins Involved in Genome Transcription in Infected Cells

Transcription

1 J. gen. Virol. 0974), 24, 4o Printed in Great Britain The Properties of the Sendai Virus Ribonucleoproteins nvolved in Genome Transcription in nfected Cells By V. M. ZADES, O. G. NKOLAEVA, L. M. SELMOVA, O. P. ZHRNOV AND A. G. BUKRNSKAYA The D. L vanovsky nstitute of Virology of the USSR Academy of Medical Sciences, Moscow, USSR (Accepted 24 April 974) SUMMARY The properties and function of parental nucleocapsid-like particles (NLP) in the cytoplasm of Sendal virus-infected Ehrlich tumour cells were studied. The formation of NLP stems from partial uncoating of the virus particles leading to the loss of about 60 % of the mol. mass of virus protein, the genome in the NLP being fully conserved. Uncoating proceeds at 37 C, but not at o C, and does not require the synthesis of protein and host RNA. NLP sedimentation rate (~ 2o0 S) and buoyant density in caesium chloride (.34 g/ml) are fairly close to those of nucleocapsids produced in vitro by treatment of the virus with sodium deoxycholate. However, the following differences between the two types of structures are found: (i) NLP are sensitive to ribonuclease; (ii) NLP contain the largest virus particle protein in addition to the nucleocapsid protein; (iii) most molecules of parental RNA in NLP sediment slower than RNA of native virus particles (50 S), this phenomenon apparently not being due to genome RNA degradation. The involvement of NLP in the parental genome transcription is deduced from the following observations: (i) NLP possess the RNA polymerase activity in vitro; (ii) virus-induced RNA synthesis can be detected in the infected cells at a stage when all cytoplasmic parental RNA seems to be in NLP; (iii) a part of newly synthesized virus-specific RNA is associated with NLP; the kinetics of labelling suggest that the RNA is synthesized in NLP and released after synthesis. NTRODUCTON Sendai virus, as well as the other myxo- and rhabdoviruses, belongs to the group of negative strand RNA viruses (Baltimore, r97). t has been shown that Sendai virus particle RNA contains a fraction of RNA which is complementary to the main part (Portner & Kingsbury, 97o; Robinson, 97o), but the complementary RNA is, apparently, not translated. The presence of the virus particle-associated RNA polymerase in this group of viruses (Baltimore, Huang & Stampfer, 97o; Chow & Simpson, 97; Huang, Baltimore & Bratt, 97 ; Robinson, 97T c) is consistent with the lack of infectivity of the virus particle RNA (Ada et al. 959; Huang & Wagner, 1966; Kingsbury, 1966) and with a possible mrna function &transcripts (Bratt & Robinson, 967; Blair & Robinson, 97o ; Huang, Baltimore & Stampfer, 97O; Pons, 1972). n this connection, it is reasonable to assume that the functioning of the virus particle-associated RNA polymerase is an essential step in the initiation of virus infection. This viewpoint agrees well with the results of Robinson (97 b) and Clavell & Bratt (1970 who demonstrated that the 'early' RNA transcription for Sendal and Newcastle disease viruses does not require protein synthesis in the infected cells. Thus,

2 410 V.M. ZADES AND OTHERS the transcription complex of the negative strand viruses is, apparently, a ribonucleoprotein which contains virus particle-associated RNA polymerase as one of its components. A number of data suggests that the nucleocapsids of the negative strand viruses are involved in transcription complexes, and the situation in vitro is quite clear at least for Sendai virus. Robinson (97 e) reported that RNA polymerase activity is associated with the nucleocapsid fraction of the virus particles and Zaides et al. (972) found that the nascent RNA strands are attached to the nucleocapsids and leave the structures after termination of the synthesis. t had been established previously that parental nucleocapsid-like particles (NLP) are found in cytoplasmic extracts of Ehrlich tumour cells soon after infection with Sendai virus (Bukrinskaya, Zhdanov & Vorkunova, 969). Further studies demonstrated that these structures are associated with virus-specific RNA synthesized by the parental RNA polymerase, since this RNA cannot be separated from the NLP either by sedimentation or by buoyant density fractionation (Bukrinskaya, Zaides & Zhdanov, 972; Zaides et al. t972; Bukrinskaya, 973). These results clearly suggest that the transcription of the parental genomes takes place in association with NLP. Although the system used is an abortive one, it allows the accumulation of virus-specific components (Bukrinskaya et al. 968) and therefore it may be considered as adequate for studies of virus genome transcription at least at early stages. Recently a suggestion similar to ours was put forward by Stone, Kingsbury & Darlington 0972) who discovered RNA polymerase activity in nucleocapsid-like structures synthesized in a productive system. This paper shows that Sendai virus genome transcription in the cytoplasm of infected cells is catalysed by virus particle-associated RNA polymerase associated with 200 S particles. These particles are identified as NLP, and their properties and composition are analysed. METHODS Chemicals and radiochemicals. Actinomycin D, pyruvate kinase and components for polyacrylamide gels were obtained from Reanal (Hungary); bovine pancreas ribonuclease (RNase), bovine serum albumin, ribonucleosidetriphosphates (ATP, UTP and GTP) from Calbiochem (U.S.A.); cycloheximide from Koch-Light (England); dimethylsulphoxide (DMSO) from Merck (Germany); EDTA (4 Na salt) from BDH (England); sucrose (free from RNase) and SDS from Serva (Germany); sodium deoxycholate (DOC) from Sigma (U.S.A.); [5-3H]-uridine ( 6 Ci/mmol), carrier free [32PO4]- and [~4C]-algal hydrolysate from Allunion Firm sotope (USSR); [~H]-CTP (20 Ci/mmol) from the Radiochemical Centre, Amersham (England). Cells. Ehrlich ascites tumour cells were used. Cells were obtained on the 7th or 8th day after tumour inoculation in white, random bred mice. Virus. Sendai virus strain 96o was inoculated in 9-day-old embryonated eggs (lo 5 Ds0/ egg) and the infected eggs were incubated at 37 C for 72 h. The labelled samples were obtained by injecting the appropriate precursors into eggs 2 h after infection: either [3H]- uridine (2oo #Ci/egg), or [14C]-algal hydrolysate (2oo/~Ci/egg), or [3zP]-orthophosphate (5oo #Ci/egg). Chorionallantoic fluid containing virus was clarified by low-speed sedimentation and then the virus particles were pelleted in the SS-34 rotor of the Sorvall RC2B centrifuge at 9ooo rev/min for h. The pellet was resuspended in 'virus buffer' (o. M- NaC; o-oi M-tris-HC1, ph 7"5). The labelled virus preparations were further purified by sedimentation in the 3 x 2o rotor of the Superspeed-5o (MSE) centrifuge in a linear sucrose gradient (2o to 60 % sucrose (w/w) virus buffer) at 2oooo rev/min for 15 to 7 h. After

3 Sendai virus ribonucleoproteins sedimentation the visible band of the virus was collected, diluted O times with the virus buffer and pelleted. The pellet was resuspended in virus buffer and stored at -4 C for no longer than day before use ([l~c]-virus for polypeptide analysis was stored at - o C up to 3 weeks). Specific radioactivity in purified preparations was ~5 x o 4 Ds0/ct/min for [32p]_ and ~ 5 x o 5 Dso/ct/min for [14C]- and [3H-preparations. The radiochemical purity of the preparation can be estimated from Fig. 6(f) (virus particles), 8(a), (b) (proteins) and 9 (a) (RNA). solation ofnucleocapsids. Purified virus labelled with [14C]-amino acids was treated with o'5 % DOC for zo rain at room temperature. The suspension was either analysed immediately in a sucrose gradient or centrifuged through 3o % sucrose (in virus buffer) and pelleted nucleocapsids were further analysed by buoyant density fractionation or in polyacrylamide gels. t should be mentioned that only diluted virus preparation containing no more than o l D~0/ml can be used to prepare the homogeneous 2ooS nucleocapsids. When a more concentrated input virus suspension was used, a heterogeneous fast sedimenting material appeared in sucrose gradients. nfection. Ascites cells were suspended in modified Earle's salt solution (MES; original solution but without glucose and sodium bicarbonate, containing o'oi M-tris-HC1, ph 7"7) at a final concentration of os cells/ml. Then the virus suspension was added to give an m.o.i, of oo to 2oo Dso/cell (labelled virus) or ooo to 2o0o Ds0/cell (unlabelled virus). n some experiments, actinomycin D (20 to 5o #g/ml) or cycloheximide (oo #g/ml) were added simultaneously with the virus. Suspensions were incubated at o C for 3o rain and then diluted with warm MES to give the final concentration of o 7 cells/ml (in experiments with antibiotics, MES contained actinomycin D and cycloheximide at corresponding concentrations). n most cases the infected cells were incubated at 37 C for 2 h. n some experiments (with actinomycin D) [ah]-uridine (3 to ~oo #Ci/ml) was added 5 to 30 min before the end of incubation. Cell fractionation. Suspensions of infected cells were diluted with cold MES and cells were pelleted by sedimentation at ooo rev/min for o rain, washed once in hypotonic buffer (o.o M-NaC1; o'oo2 M-Mg-acetate, o'o M-tris-HC1, ph 7"5) and disrupted in the same buffer (to s cells/ml) in a Dounce homogenizer. The cell homogenate was centrifuged at 25oo rev/min for O rain and the supernatant fluid ('S-2') treated with o.oi M-EDTA and centrifuged at 15oo0 g for O rain. The supernatant fluid obtained ('cytoplasmic extract') was used either for sucrose gradient analysis or to obtain 'particulate fraction' preparations. n the latter case the material was again pelleted by sedimentation through 3o % sucrose in hypotonic buffer (without Mg 2+) in the 5o Ti rotor of a Spinco L2 centrifuge at 4oooo g for h. The pellet ('particulate fraction') was suspended in o.oi M-triethanolamine (TEA), ph 7"5, and used for polypeptide analysis of proteins or for buoyant density analysis. n the former case the suspension was stored before use at -lo C, in the latter case it was immediately treated with 4 % formaldehyde (Spirin, Belitsina & Lerman, 965) and stored at 4 C. When the particulate fractions were used for the determination of structures containing RNA polymerase the 14o00o g pellets were analysed in sucrose gradient immediately. No more than o % of u.v.-absorbing material of cytoplasmic extract (/ o nm) was recovered in particulate fraction. Most of this material was represented by EDTA-derived '50 S' ribosomal subunits. More than 9o ~o of the 200 S particles (the main object of this investigation) were recovered in particulate fraction. RNA extraction. RNA was extracted in STE (o. M-NaC1; o.ooi M-EDTA; o.o M-tris- HC, ph 7"5) by the phenol-sds method at o C (Bratt & Robinson, 967). 4 r

4 412 V.M. ZADES AND OTHERS Sedimentation analysis. Cytoplasmic extracts were layered on to 5 to 30 ~o (w/w) sucrose gradients (in hypotonic buffer without Mg 2+) and centrifuged. The u.v.-absorbing material distribution in the gradients was registered in a flow cell of a Uvicord photometer (LKB). RNA samples were layered on to O to 30 % (w/w) sucrose gradients (STE with 0"5 % SDS). For analysis in denaturation conditions (Strauss, Kelley & Sinsheimer, 968) 5 to 20 % sucrose gradients with 99 % DMSO were used. For details of sedimentation analysis see legends to figures. Buoyant density fractionation. Formaldehyde-fixed samples were analysed in CsC1 preformed gradients (p = 1.45 to ~' 5 g/ml), as described previously (Bukrinskaya et al ~969). The o.oi M-TEA, ph 7"5, and 4 ~o formaldehyde containing gradients were centrifuged in SW 5o.x rotor at 35o0o rev/min, 4 C for x6 h. The refractive indices (fft, Voet & Vinograd, ~96) were determined and buoyant density was calculated after correction of formaldehyde refraction. Assay ofrna polymerase activity. The method described by Robinson 097C) as slightly modified by Zaides et al. (972) was used. Particulate fraction in a buffer (o.~ M-NaC; o-o M-Mg-acetate; o'0o5 M-a-mercaptoethanol; o'1 M-tris-HC1, ph 7"9) was centrifuged in 5 ml linear sucrose gradient in the above buffer. After sedimentation 0"25 ml fractions were collected and RNA polymerase activity was determined in o-1 ml samples. Either o'o5 ml of a complete mixture was added to each sample to give final quantities: unlabelled ribonucleosidetriphosphates (ATP, UTP and GTP), 0'05 #mol of each; [3H]-CTP, 0"05 #Ci; NaC1, 5/zmol; Mg-acetate, 1"5 #mol; 2-mercaptoethanol, o'75/~mol; tris-hc1, ph 7"9, 5#mol, or o'o5 ml incomplete mixture (the same without unlabelled triphosphates). The samples were incubated for h at 29 C, acid-insoluble material was precipitated and washed on membrane filters in the presence of o.~ M-sodium pyrophosphate, and counted. Polyacrylamide gel electrophoresis. We used mainly the method described by Laemmli (~97o). 7"5 ~o gels approx. ~o cm long, with a stacking gel of2. 5 ~o acrylamide, were employed. The gels contained o.z % N,N-methylene bisacrylamide in o'375 M-tris-HCl, ph 8.8, with o.r ~o SDS. Prior to electrophoresis, dissociation of polypeptides was performed by suspending samples in 5 M-urea, 2 % SDS, 5 % 2-mercaptoethanol, and heating them in a boiling water bath for 3 min. Electrophoresis was carried out for 3 to 5 h at 5 ma]gel. After electrophoresis gels were either stained by o.2 % amidoblack in water-acetic acid-methanol mixture (5 : : 5) for 2 h, or the radioactivity was counted in the fractions. n the latter case gels were cut into o'75 to 1.z5 mm slices, each slice was put into a tube with o. ml 3o % H202 and incubated overnight at 37 C. Solubilized materials were put on paper discs, which were counted after drying. The recovery of radioactivity from the gels was 70 to OO %. Mol. wt. of polypeptides were determined for each sample independently by the method of Weber & Osborn 0969) using bovine serum albumin, mol. wt , pyruvate kinase, mol. wt. 57ooo and ovalbumin, mol. wt. 43 ooo as markers. Assay of radioactivity. Casein or yeast RNA and then TCA (to 5 %) were added to samples, the acid-insoluble material was collected on millipore filters and washed there with 5 ~o TCA and ethanol. The filters placed in vials with toluene-ppo-popop scintillation fluid were counted in a Packard scintillation counter. RESULTS The identification of cytoplasmic 200 S particles as Sendai virus transcriptive complexes Robinson (97 b) demonstrated that transcription of Sendal virus parental genome does not require protein synthesis in infected cells suggesting the involvement of virus particle-

5 Sendai virus ribonuceloproteins 4 ~ 3 Table x. RNA and protein synthesis in Ehrlich tumour cells TCA-insoluble radioactivity Cycloheximide [3H]-RNA*, [14Cl-proteinst, ( oo/zg/ml) nfection ct/min/mg ct/min/107 cells ooo ND:~ q ND;~ ND * The cells were infected at m.o.i, of ooo to 2000 Ds0/cell, actinomycin D (50/zg/ml) being added to all samples simultaneously. [3Hl-uridine (30/~Ci/ml) was added to samples after 9o rain of incubation at 37 C for 3o min and RNA was isolated from cytoplasmic extracts and TCA-insoluble radioactivity determined. t Uninfected cells were incubated at 37 C for 3o min (one sample with cycloheximide), then [xac]-amino acids mixture (o'5/~ci/ml) was added. Cells were incubated 3o rain additionally and TCA-insoluble radioactivity was determined in the cytoplasmic extracts. Not done. associated RNA polymerase in this process. We repeated these experiments in our system. Table shows that the incorporation of label into virus-induced RNA in the cytoplasm of infected cells is not decreased when protein synthesis is significantly (by more than 9 o %) inhibited by cycloheximide. Sedimentation analysis of this RNA reveals two size-classes, 8 S and 35 S (Fig. ~) which represent the typical products of Sendai virus RNA transcription in infected cells (Blair & Robinson, 1968; Portner & Kingsbury, 972 ). To localize the cytoplasmic structures involved in the transcription process, the infected cells were exposed to [3H]-uridine for 5 rain (to label transcriptive complexes predominantly) and 3o min, and cytoplasmic extracts were fractionated by rate zonal sedimentation. Fig. 2 shows that an essential part of radioactivity after short pulse is found in the 2oo S position, but the relative amount of label in this peak is decreased when the labelling period is prolonged. Thus, the 2oo S structures may be possible sites of virus RNA transcription. Further experiments were performed to reveal the cytoplasmic structures which possess RNA polymerase activity. For this purpose, the particulate fractions from infected and uninfected cells ~'ere fractionated in velocity sucrose gradients and RNA polymerase activity was assayed in each fraction. Fig. 3 shows that the 2oo S structures from infected cells possess RNA polymerase activity. Thus, the 2oo S structures are apparently involved in the process of virus genome transcription in cytoplasm. t will be shown further, that these structures represent parental NLP. dentification of NLP in cytoplasmic extracts The results presented in this section deal with the fate and processing of labelled virus particles inside the cell. n a preliminary experiment we have shown that about a half of the input virus (labelled with ~H, or 32p, or 14C) is firmly attached to the cells under the conditions used. Up to 5o % of the cell-associated radioactivity was recovered in cytoplasmic extracts and analysed further. Sedimentation analysis of parental RNA and proteins in cytoplasmic extracts Fig. 4 shows the results of sedimentation analysis of cytoplasmic extracts of the cells infected with [14C]- and [3zP]-labelled virus particles (a and b, respectively). Nucleocapsids obtained by DOC treatment of Sendai virus in isotonic buffer of [14C]-labelled virus were 27 VR 24

6 414 V. M. ZADES AND OTHERS 50S 35S28S 18S ~ Fig. i, Sedimentation analysis of newly synthesized RNA in cytoplasmic extracts of infected and uninfected cells. Cells were infected with unlabelled virus (m.o.i. of ooo to 2ooo Ds0/cell) and incubated at 37 C with actinomycin D (5o #g/ml) and cycloheximide (oo #g/ml) for 2 h. 3o min before the end of incubation [3H]-uridine (3o #Ci/ml) was added, RNA was extracted from cytoplasmic extracts and analysed. 4oo #g samples of RNA were layered on 27 ml sucrose gradients (o to 30 ~, STE with 0'5 ~ SDS) and centrifuged in the SW27 rotor of a Spinco L3-5o centrifuge for 6 h at 6OOO rev/min, at 25 C. ml fractions were collected from the bottom of the tubes and TCA-precipitable radioactivity was determined. ---, u.v.-absorbing material; O------O, RNA radioactivity from infected cells; O----, the same from uninfected cells also analysed for comparison (c). t may be seen that all these gradients show the presence of zoos particles. t should be noted that [zzp]-radioactivity at the gradient top was not shifted after more extensive sedimentation, and therefore, represented either very short polynucleotides or some material quite different from RNA (phospholipid and/or phosphoprotein). No 2oo S particles or soluble virus particle proteins were found in cytoplasmic extracts of infected cells incubated at o C, all the radioactivity being associated with the particles which sedimented faster than 35o S. The same results were obtained after sedimentation analysis of parental RNA and proteins from infected cells treated with cycloheximide 0oo #g]ml) and actinomycin D (zo #g/ml) (not shown). Buoyant density analysis of particulate fractions from the infected cells Particulate fractions were prepared from the ceils infected with the virus particles labelled either with [l~c]-amino acids (Fig. 5a) or [3H]-uridine (Fig. 5b). t had been determined previously that sedimentation of parental zoo S particles through 3o % sucrose did not lead to clumping and that the treatment of the material with formaldehyde did not change sedimentation rate (Fig. 5a, insert). t is seen from Fig. 5, that the buoyant density of the parental particles is "34 to 1.35 g/ml. Similar results were obtained after incubation of the

7 Sendai virus ribonucleoproteins ~200S 80S?! # tl *6 10 L ~ ~,.O-- O " O" O"O'O"~'r "! Fig. z. The kinetics of labelling of virus-specific structures containing RNA in cytoplasmic extract. Cells were infected and incubated as described in the legend to Fig.. 5 or 30 min before the end of incubation ph]-uridine (ioo #Ci/ml) was added for 5 and 30 min, respectively, cytoplasmic extracts were prepared and analysed in 5 to 30 ~ sucrose gradients in hypotonic buffer. The gradients were centrifuged in the 3 x 2o rotor of the Superspeed-5o centrifuge at 27ooo rev/min for 90 min. Fractions were collected from the bottom of the tubes, acid-insoluble radioactivity of each fraction was determined and expressed as per cent of the total radioactivity of the gradient , incubation with [3H]-uridine for 5 min;..., the same for 3o min. The arrow (8oS) designates the monoribosome position in a separate tube. infected cells with cycloheximide and actinomycin D (not shown). When Sendai virus nucleocapsids obtained by DOC treatment of the virus particles in isotonic buffer were similarly analysed, their buoyant density was also about "34 g/ml (Fig. 5 c). Thus, sedimentation rate and buoyant density of the intracellular parental subviral particles (NLP) are very similar to those of Sendai virus nucleocapsids. The rate of accumulation of the NLP in the cytoplasm of the infected cells Cells were infected with the virus labelled with [l~c]-amino acids and particulate fractions were prepared at intervals and analysed by buoyant density sedimentation. nfected cells were incubated for 15, 25, 5o and 75 rain at 37 C (Fig. 6a, b, c and d, respectively) and for i2o rain at o C (Fig. 6e). As can be seen from Fig. 6, NLP are found as early as 15 min after incubation at 37 C and their number increases further so that in 75 rain the bulk of parental proteins is found in the NLP. On the other hand, the gradual decrease of parental proteins in the zone of the gradient with a buoyant density of about 1-25 g/ml is clearly observed. This finding suggests that the 1.25 material is a precursor of NLP. ts conservation at o C is in agreement with such suggestion. The possibility that the parental structures 27-2

8 4(6 V. M. ZADES AND OTHERS 185S ro 7 o Fig. 3. RNA polymerase activity of particulate fractions from infected and uninfected cells. Cells were infected and incubated as described in legend to Fig. (without [ah]-uridine). o'3 ml samples of particulate fractions (70 optical units at a = 260 nm) were centrifuged in ]5 to 3o ~ (w/w) sucrose gradients in the SW 5o of Spinco L2-5o centrifuge at 25 ooo rev/min, 4 C for h. After sedimentation, RNA polyrnerase activity was determined in the fractions of the gradients as described in Methods. The values obtained with incomplete incubation mixture (without unlabelled triphosphates) were subtracted from those with the complete one. O--O, the material from infected cell; O ---, the same from uninfected cells. Arrow designates the 85S tobacco mosaic virus position after the same sedimentation in the separate tube (TMV was kindly supplied by Dr B. S. Naroditsky). ~200S 80S "2o& s ~200S 80S L, Fig. 4. Sedimentation analysis of cytoplasmic structures containing parental virus protein and RNA. Cytoplasmic extracts were prepared after 2 h incubation at 37 C of the cells infected with the virus labelled either with p4c]-amino acids (a) or with asp (b). Nucleocapsids were prepared by treatment of [l~c]-virus particles with o'5 ~ DOC in virus buffer (c). The gradients were centrifuged as described in legend to Fig. z and acid-insoluble radioactivity was determined in fractions.

9 . Sendai virus ribonucleoproteins ~'A~ A [ "3 (a) 1~, 335~'&" 1"2 ~, ~ 6/-; s'// [ "'~.,~ 1"4 7.~ * ~'~A 1'34 ~2 (b) 1~ Je ~A 1"4 ~, A (c) i.335~a, 1.2 = 6 ~6 4, 5 l0 15 t Fig. 5. Buoyant density analysis of parental zoos particles and Sendai virus nucleocapsids. Cells were infected with [l~c]-amino acids (a) or [ah]-uridine (b) labelled virus particles and incubated for 2 h at 37 C. Particulate fractions were prepared and analysed in CsC gradients, p4c]-nucleocapsids were prepared by 0"5 % DOC treatment of purified virus particles and analysed similarly after pelleting through 30 % sucrose (c). Sedimentation of P4C]-parental particles of particulate fraction is shown on insert in Fig. 5 (a). with buoyant density of 1.25 g/ml are Sendai virus particles is confirmed by the following observations: (i) the P25 material sediments faster than 35oS in sucrose gradient; (ii) the buoyant density of Sendai virus particles treated with formaldehyde is ~.25 g/ml (Fig. 6 f) (the treatment of the virus particles with formaldehyde has been shown to increase their buoyant density from 1.2t to "25 g/ml as in the case of poliovirus (Agol et al. 97O). The rate of virus particle uncoating was not changed appreciably after tenfold increase of the input m.o.i, but was slightly higher after three to fivefold decrease (not shown). Effect of RNase treatment on the NLP As shown in Fig. 7(a), a significant decrease of p4c]-proteins in NLP 2005 position is observed after treatment of the cytoplasmic extract with RNase (by 75 to 9o % in different experiments). Beside the resistant fraction (very small in this experiment), 15o to 7oS particles were revealed which were, apparently, the aggregates of the nucleocapsid protein subunits. n accordance with earlier data (Kingsbury & Darlington, 968; Kingsbury, Portner & Darlington, 97 o) Sendai virus nucleocapsids were resistant to RNase treatment (Fig. 7b). Analysis of NLP parental proteins Polypeptides of NLP were analysed by electrophoresis in polyacrylamide gel. Electrophoretic patterns of virus particle and of nucleocapsid proteins were also obtained for comparison. t is seen from Fig. 8 (a) and (b) that virus particles contain five major proteins and a certain number of minor bands. The minor bands are distinct only in stained samples. The number of major bands and tool. wt. are in good agreement with the results reported earlier (Mountcastle, Compans & Choppin, 97[ ; Hosaka & Shimizu, i972 ). t is possible that some of the minor bands are due to contamination of virus preparations with non-virus particle proteins. The major proteins GP[ and GP2 are probably the envelope proteins of Sendai virus particles (Hosaka & Shimizu, 972 ). Polypeptide of NP class represents

10 418 V.M. ZADES AND OTHERS l (a) 1"335 1"252 (b) (c) 1"340 1'255? x 3 _ A~ A 1'4 ~ "~A ~ ~A J 1'4..~ ~A~ 1.3 ~ ~A 1.3 "~ "* 1.2 ~ 1.2.'~ 1.2 ~= " ? 4 l (d) ~A ~A (e) 1 "255 ~1 "4~"u' "-] 1"4 ~ 1"3 "~ ~ ~'A ~, "]1"3 >, 1'2'~ ~A-' 1"2 '~ (f) ~,t ~" ~'~A "q 1"3~ >" X ~ & 10 5 _L Fig. 6. The rate of NLP accumulation in the cytoplasm of infected cells. Cells infected with p4c]- amino acid labelled virus were incubated at 37 C for 15 min (a), 25 min (b), 5o min (c), 75 rain (d) and at o C for i2o min (e). Particulate fractions were prepared and analysed in CsC gradients. p4c]-virus particles treated with formaldehyde were also analysed (f). apparently nucleocapsid protein (Fig. 8c)with mol. wt. about (Mountcastle et al. 97O). The smallest protein of Sendai virus (M) is probably the membrane or matrix protein but the function of the largest protein (L) remains unknown. As seen in Fig. 8 (d), two distinct classes of parental proteins are found in the NLP: L (mol.wt. 7[ ooo) and NP. These results are in agreement with those of Stone et al. 097a) who analysed polypeptide composition of the nucleocapsid particles synthesized in Sendai virus-infected chicken cells. However, unlike Stone et al we did not find any increase of the

11 Sendai virus ribonucleoproteins 419 (a) ~ 200S 80S (b) ~ 200S 80S x 2-1! _ r Fig. 7. RNase sensitivity of NLP and nucleocapsids. Cytoplasmic extract from cells infected with p4c]-amino acid labelled virus and p4c]-nucleocapsids were prepared as described in legend to Fig. 4. ml samples were treated with RNase O0/zg/ml) at o C for 20 rain. Samples were analysed in 15 to 3o ~ sucrose. (a) p4c]-particles in cytoplasmic extract: sedimentation in SW27 rotor of Spinco L2-65 centrifuge at 22ooo rev/min for 3 h; (b) p4c]-nucleocapsids: sedimentation in 3 x 2o rotor at 27o0o rev/min for 9o rain. O-----O, samples not treated;... G, RNase treated samples. (a) 2O 7k NP (b) (c) (J) NP 7 6 M i x 5.E 4 E ~3 GP Gp v x NP,' Fraction ntmabcr Fig. 8. Electrophoretic patterns of virus particles (a, b), lmcleocapsid (c) and NLP (d) polypeptides in SDS-polyacrilamide gels. Migration is on the left to right. (a) UnlabeUed proteins of the virus particle in a stained gel (~ 2oo #g of protein were applied); (b) p~c]-proteins of the virus particle; (c) p4c]-proteins of nucleocapsids; (d) P4C]-proteins of particulate fraction prepared after 2 h incubation of [14C]-virus-infected cells at 37 C. relative amount of L class protein in NLP as compared with that in virus particles, in both cases the ratio of radioactivity (NP: L) being about 3"5:1. Thus, the molar polypeptide ratio was about 4:. Analysis of NLP parental RNA Fig. 9 (a) demonstrates that a greater part of [32p]-RNA from purified Sendal virus sediments as a sharp peak of 5oS molecules, agreeing with earlier determinations (Blair &

12 420 V.M. ZADES AND OTHERS (a) 50S 28S18S ~ t i (b) 4? 7 15 = c, 3 x x x 3 "5 10 o t r Fig. 9. Sucrose gradient fractionation of P2P]-RNA from virus particles and NLP. RNA were extracted from purified virus particles (..., ct/min x to ~) and from particulate fraction which was prepared after z h incubation of the cells infected with p2p]-virus particles at 37 C ( 0, ct/min x io-1). (a) Sedimentation of the samples in STE. Conditions of sedimentation were the same as in Fig. (a). (b) Sedimentation of the samples in 99 % DMSO. The pellets of RNA were dissolved in o.ooi M-tris-HCl, ph 7"5, containing o-oo M-EDTA, then an equal vol. of 99 ~ DMSO containing the same concentration of EDTA and tris-hc1 was added. RNA was layered on to 5o ml sucrose gradients (5 to zo ~, 99 ~ DMSO with EDTA and tris-hcl) and gradients were centrifugated in SW 25.2 rotor of Spinco L-a centrifuge at zzooo rev/min at 25 C for 4o h. After sedimentation, the 1.8 mlfractions were collected and TCA-insoluble radioactivity was determined. Robinson, 968). A minor part of the RNA sediments slower as a heterogeneous material and may be due to different factors such as contamination of the virus preparation with extraneous labelled polynucleotide; partial hydrolysis of the virus RNA during extraction (this possibility is not very likely since ribosomal RNA extracted concurrently was not degraded); the presence of the incomplete virus particles in the preparation of purified virus (Kingsbury et al. ]97o); and finally partial self-hybridization of complementary chains of virus particle RNA described for Sendai virus preparation (Portner & Kingsbury, 197o; Robinson, 97O). n our experiments, up to 36 % of the parental RNA acquired resistance to RNase after self-annealing at saturating concentration suggesting that the 960 strain of Sendai virus contains up to 8 % of complementary RNA chains. t may be seen from Fig. 9 (a) that only a small part of RNA from NLP sediments as 5o S molecules (from 15 % to 4o % in various experiments) while the main part sediments in a wide zone from 35 S and slower. t seems to be unlikely that the decrease of sedimentation rate is due to breaks of the parental RNA during extraction procedures since ribosomal RNA sedimentation pattern was fairly distinct on the gradient. On the other hand, a significant amount of parental RNA in ~< 35 S position can hardly be explained by selfhybridization. Two possibilities then remain: (i) most molecules of 50 S RNA are broken in the cell; (ii) the decrease of sedimentation rate is due to cell-induced configurational change of the RNA. To choose between these two possibilities we performed, firstly, a heat denaturating experiment. However, heat treatment of the RNA in boiling water for 3 min induced breaks in polynucleotide strands which masked the results. We then used the technique of separation of hydrogen-bonded polynucleotides by sedimentation in 99 % DMSO (Strauss

13 Sendal virus ribonucleoproteins et al. ~968). As seen from Fig. 9(b), the parental RNA from NLP under these conditions sediments as a symmetric peak in the same position as virus particle RNA. Thus, the virus genome in NLP is fully conserved. t is not known at present whether the slower sedimentation rate of parental RNA is due to inter- or intramolecular interactions. 4z DSCUSSON The data presented here have demonstrated uncoating of Sendal virus particles in the infected cells. The products of uncoating NLP and soluble parental proteins were registered as early as 5 min after incubation of the infected cells at 37 C. Though uncoating is a temperature-dependent process, it does not require de novo synthesis of proteins and host RNA. Two hours after infection, when practically all the parental RNA of the infected cell cytoplasm is in NLP, the synthesis of a virus-specific RNA is induced by virus particleassociated RNA polymerase. Some part of the virus-specific RNA which resembles, in the kinetics of labelling, immature transcription products is associated with 2o0 S particles of the cytoplasmic extract. 2oo S particles also possess RNA polymerase activity in vitro. These observations correlate with the earlier suggestions about transcriptive function of the 2oo S particles (NLP) of Sendal virus (Bukrinskaya et al. 197z ; Stone et al ; Zaides et al. 1972; Bukrinskaya, 1973)- Coefficient of sedimentation (~ 2oo S) and buoyant density (-34 g/ml) of NLP are very similar to those for nucleocapsids obtained by DOC-treatment of Sendal virus particles (the nucleocapsids obtained from the virus particles by DOC treatment in isotonic buffer bad a slightly higher buoyant density in CsC1 than those obtained in hypotonic buffer). However, there are some marked differences between these two types of structures. (i) A significant portion of NLP is RNase-sensitive whereas all the nucleocapsids are RNase-resistant. This fact implies certain changes in RNA-protein interaction in NLP. (ii) As well as the NP class, there are the L class polypeptides found in NLP. (iii) Finally, striking differences in parental RNA are revealed. As shown by DMSO experiments, NLP contain the full virus genome. Thus, the slower sedimentation rate of parental RNA might be explained either by its association with immature products of transcription or by an alteration of its conformation. The results obtained in other experiments argue in favour of the first possibility. f so, the NLP may be considered as actively functioning transcriptive complexes. t is possible that the bulk of NLP is involved in transcription. Yet it cannot be ruled out that a part of NLP does not take part in transcription. n fact, in all experiments we found a certain fraction of genomic RNA at its usual position (5o S). f the shift of 5o S RNA to slower sedimenting species is due to transcriptive activity of the structures, then the lack of the shift may indicate a lack of activity. n addition, if RNase-sensitivity of NLP is in any way connected with transcriptive activity, then the RNase-resistant fraction could be considered as inactive. n this connection, we may assume that a certain fraction of parental nucleocapsids for some reasons cannot be involved in transcription. From our point of view, the nucleocapsids which contain the complementary molecules of the parental RNA (up to 18 % from the total population for 96o strain of Sendal virus) are good candidates for these non-functioning particles. At the same time the present results are not enough to make a complete story because only a part of the cell-associated parental genomes (about 5o %) were revealed in cytoplasmic extracts and analysed. t is evident that NLP contain a significant amount of structural proteins of the virus particles, the sum of L and NP polypeptides being about 4 o % of total protein of Sendal

14 422 V.M. ZADES AND OTHERS virus particle. f we assume that the protein tool. mass of the Sendal virus particle is equal to "~5 on, by analogy to SV5 virus particles (Klenk & Choppin, 969) the calculated tool. mass of NLP parental proteins will be ~ 2 o 8 (this rough estimate is valid if there is not loss of whole protein subunits in NLP). Then, each particle contains about 650 molecules of L polypeptides and about 2600 of NP class molecules. t is unlikely that these proteins are RNA polymerases since such an amount of enzymic molecules would hinder their movement along the template so that we would have to suggest some new mechanism of transcription; and it may be that the RNA polymerase activity is associated with some minor parental protein(s) which have not been detected by the method used. At present, we can only guess about the relative functions of L and NP proteins of NLP. n particular, they may protect the template from nuclease hydrolysis in the cell since the transcription proceeds, apparently, for many hours (Clavell & Bratt, 97). t cannot be ruled out that either L or NP proteins can control transcription, for example by opening adequate sites of the genome. The rate of Sendal virus RNA transcription is constantly increasing during the infection reaching a maximum at 8 to 20 h (Blair & Robinson, 968 ). Therefore, one may assume that newly synthesized virus particle RNAs are involved in transcription and it seems unlikely that the transcription mechanisms for parental and newly synthesized genomes would be markedly different. n particular, the properties of the parental and 'late' transcriptive complexes should be similar. This view is in good agreement with many observations. Blair & Robinson 097o) demonstrated that newly synthesized 5oS RNA of Sendal virus does not exist in a free form : even after 5 rain of pulse labelling, all the genomic RNA is found within nucleocapsids. These structures are associated with newly synthesized heterogeneous RNA which is partially RNase resistant (Robinson, i971 a). Finally, these structures display RNA polymerase activity and their polypeptide composition is fairly similar to that described in this paper (Stone et al 972 ). However, it is necessary to take into account the fact that the relative amount of newly synthesized nucleocapsids involved in the transcription is rather small. The RNase-resistant fraction found in the newly synthesized nucleocapsids, in turn, significantly decreases with the increase of time of labelling (Robinson, 97 a). Thus, some mechanism(s) seems to exist in infected cells which provides maximum RNA transcriptive activity of the parental NLP and limits this activity in newly synthesized structures on the later stages. The nature of this mechanism is not yet known. N. Semenova provided skilful technical assistance. We thank Dr N. V. Kaverin and Dr V.. Agol for helpful discussions. REFERENCES AOA, ~. r., LND, P. E., LANKN, r. & BtJRNET, r. M. (959). Failure to recover infective 'ribonucleic acid' from myxovirus preparations. Nature, London x84, 36o-36. AGOL, V.., LPSKAYA, G. YU., TOLSKAYA, E. A., VOROSHLOVA, M. K. & ROMANOVA, L ). Defect in poliovirus maturation under hypotonic conditions. Virology 4x, o. BALTMORE, D. (971). Expression of animal virus genomes. Bacteriological Reviews 35, BALTMORE, D., HUANG, A. S. & STAMPFER, M. (r970). RNA synthesis of vesicular stomatitis virus.. An RNA polymerase in the virion. Proceedings of the National Academy of Sciences of the United States of America 66, BLAR, C. O. & ROBNSON, W. S. (968). Replication of Sendal virus.. Comparison of the viral RNA and virus-specific RNA synthesis with Newcastle disease virus. Virology 35, BLAR, C. D. & ROBNSON, W. S. (970). Replication of Sendal virus.. Steps in virus assembly. Journal of Virology S, o. BRATT, M. A. & ROBNSON, W. S. (967). Ribonucleic acid synthesis in ceils infected with Newcastle disease virus. Journal of Molecular Biology 23, -2.

15 Sendai virus ribonucleoproteins 423 BC<RNS~ZAYA, a. G. (1973). Transcriptive function of Sendai virus nucleocapsid in infected cells. Virology $z, o. BUKRNSKAYA, A. G., KLMENKO, S.M., SMRNOV, Y.A. & n~uschn, B.V. (1968). nfective substructures of Sendai virus from infected Ehrlich ascites tumour cells. Journal of Virology 2, BUKRNSKAYA, A. G., ZADES, V. M. & ZHDANOV, V. M. (1972). ntracellular transcription of genome of paramyxoviruses within the nucleocapsid. Doklady of the A eademy of Sciences of the U.S.S.R. 2o5, BUKRNSKAYA, A. G., ZHDANOV, V. M. & VORKUNOVA, G. K. (1969). Fate of Sendal virus ribonucleoprotein in virus infected cells. Journal of Virology 4, CHOW, N. z. & SMPSON, R. W. (971). RNA-dependent RNA polymerase activity associated with virions and subviral components of myxovirus. Proceedings of the National Academy of Sciences of the United States of America 68, CLAVELL, L.A. & BRATT, M.A. (97). Relationship between the ribonucleic acid synthesizing capacity of ultraviolet-irradiated Newcastle disease virus and its ability to induce interferon. Journal of Virology 8, 50o-508. HOSAKA, Y. & SHMlZU, Y. K. (1972). Artificial assembly of envelope particles of HVJ (Sendai virus).. Assembly of hemolytic and fusion factors from envelopes solubilized by Nonidet P4o. Virology 49, HUANG, A. S., BALTMORE, D. & STAMPFER, M. (197o). Ribonucleic acid synthesis of vesicular stomatitis virus.. Multiple complementary messenger RNA molecules. Virology 42, HUANG, A. S., BALTMORE, D. & BRATT, M.A. (1971). Ribonucleic acid polymerase in virions of Newcastle disease virus: comparison with the vesicular stomatitis virus polymerase. Journal of Virology 7, rluang, A. S. & WAGNER, R. R. (1966). Comparative sedimentation coefficients of RNA extracted from plaqueforming and defective particles of vesicular stomatitis virus. Journal of Molecular Biology 22, 38t-384. FFT, J. B., VOET, O. H. & VNOGRAD, 1. (196). The determination of density distributions and density gradients in binary solutions at equilibrium in the ultracentrifuge. Journal of Physical Chemistry 65, KNGSBURY, D.W. (1966). Newcastle disease virus RNA.. solation and preliminary characterization of RNA from virus particles. Journal of Molecular Biology x8, 195-2o3. KNGSBURY, D. W. & DARHNGTON, R. W. (1968). solation and properties of Newcastle disease virus nucleocapsid. Journal of Virology 2, KNGSBURY, D. W., PORTNER, A. & DARLNGTON, R. W. 097o)- Properties of incomplete Sendal virions and subgenomic viral RNAs. Virology 42, ~:LENK, H.-D. & CHOPPN, t'. W. (969). Chemical composition of the parainfluenza virus SV5. Virology 37, LAEMMLL U. K. 097O). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, London 227, 68o-685. MOUNTCASTLE, W. E., COMPANS, R. W., CALGUR, L. A. & CHOPPN, P. W. (197o). Nucleocapsid protein subunits of Simian virus 5, Newcastle disease virus and Sendal virus. Journal of Virology 6, MOUNTCASTLE, W.E., COMPANS, R.W. & CHOPP1N, P. W. 097D. Proteins and glycoproteins of paramyxoviruses : a comparison of Simian virus 5, Newcastle disease virus and Sendai virus. Journal of Virology 7, 47-5 z. VONS, M. W. (t972). Studies on the replication of influenza virus RNA. Virology 4/, ~ORTNER, A. & K1NGSBURY, D. W. (970). Complementary RNAs in paramyxovirions and paramyxovirus infected cells. Nature, London 228, ~ VORTNER, h. & KNGSBURV, D.W. (1972). dentification of transcriptive and replicative intermediates in Sendai virus infected cells. Virology 47, ROBNSON, W. S. (970). Self-annealing subgroup 2 myxovirus RNAs. Nature, London 225, ROBNSON, W. S. (971 a). ntracellular structure involved in Sendal virus replication. Virology 43, 9o-oo. ROBNSON, W. S. (97 b). Sendal virus RNA synthesis and nucleocapsid formation in the presence of cyeloheximide. Virology 44, 494-5o2. ROBNSON, W. S. (97 C). Ribonucleic acid polymerase activity in Sendal virions and nucleocapsid. Journal of Virology 8, SVlRN, h. S., BELTSNA, ~. V. & LEgmAN, M.. (965). Use of formaldehyde fixation for studies of ribonucleoprotein particles by caesium chloride density gradient centrifugation. Journal of Molecular Biology x4, STONE, H. O., ~~GSBUR, ~. W. & DARLNGTON, R. W. (972). Sendai virus induced transcriptase from infected cells : polypeptides in the transcriptive complex. Journal of Virology xo, O37-1o43. STRAUSS, J. H., KELLEY, R. B. & SNSnEMER, R. L. (968). Denaturation of RNA in dimethylsulfoxide. Biopolymers 6, WEBER, K. & OSBORN, M. (1969). The reliability of molecular weight determinations by dodecyl-sulfate-polyacrylamide gel electrophoresis. Journal of Biological Chemistry 224, 44o ZADES, V. M., NCOLAEVA, O. G., SELMOVA, L. M. & BUKRNSKAYA, A. G. (972). On the role of Sendai virus nucleocapsid in transcription of virat RNA. Voprosi Virusologii U.S.S.R. NS, 6o (Received 5 January 974)