The Structure of Heated Poliovirus Particles

Transcription

1 J. gen. Virol. (I97I), IX, I Printed in Great Britain The Structure of Heated Poliovirus Particles By M. BREINDL Institut fiir Mikrobiologie, Medizinische Hochschule Hannover, Hannover, W. Germany and Roche Institute of Molecular Biology, Nutley, New Jersey, U.S.A. (Accepted 2I January 1971 ) SUMMARY Purified poliovirus preparations were heated and analysed by sucrose gradient centrifugation. They consisted of virus-like particles containing RNA and sedimenting at about 8os, empty 8o s capsids, 35s virus RNA, and a non-sedimentable capsid polypeptide (VP 4)- By electron microscopy the 8os ribonucleoprotein particles (8os RNP)were similar in appearance to intact poliovirus particles. They contained infectious, RNase-sensitive RNA that could be liberated from the capsid by treatment at room temperature with 1% sodium dodecyl sulphate. One of the virus polypeptides was missing, and they had lost the antigenicity of the mature virus and the ability to adsorb to HeLa cells. Degradation of poliovirus particles probably occurs in two steps: the first is the splitting off of a minor part of the capsid protein (VP 4) followed under certain conditions by a liberation of the RNA from the capsid. The alteration of the physical and biological properties of the virus particle is probably due to the loss of the protein rather than to the liberation of the RNA. INTRODUCTION Heating of poliovirus causes extensive changes in the biological and physical properties of the virus particle. The antigenicity of the virus is changed by heating (Le Bouvier, 1955; Mayer et al. I957). While mature infectious virus particles possess antigenic D (or N) reactivity, heated virus preparations are C (H) reactive. This is accompanied by a loss of infectivity (Mayer et al. I957). Since heat-inactivated poliovirus particles are not able to adsorb to the host cell (Graham, I959; Katagiri, Hinuma & Ishida, I968) although they contain infectious, intact RNA (Koch, I96o), the inactivation by heat of poliovirus must be due to an alteration of the protein moiety or the physical integrity of the virus particle rather than to RNA inactivation. Several authors have established that heat treatment causes a splitting of the poliovirus particle into empty capsids and free RNA (Drees & Borna, I965; Hinuma et al. 1965). Moreover, one of the four virus polypeptides (VP 4) is liberated from the virus particle together with the RNA by heat treatment (Maizel, Phillips & Summers, 1967). Kinetic studies on the degradation of poliovirus by heat (Hinuma et al. I965) and ultraviolet light (Katagiri, Hinuma & Ishida, I967) suggested a relationship between RNA liberation and change of biophysical properties of the virus. I previously reported that 8o s virus-like particles containing intact RNA could be obtained by heat treatment of purified poliovirus preparations (Breindl, I969). In this paper I describe further studies on the properties of these particles and present a two-step mechanism of poliovirus heat degradation. The first step is the liberation of one of the virus capsid

2 i4 8 M. BREINDL polypeptides (VP 4) causing the alteration of the biophysical properties of the virus particle. This is followed, under certain conditions, by the release of virus RNA from the capsid. METHODS Virus. Poliovirus type I (MAHONEY) was used throughout these studies. Cells. HeLa cells strain S 3 were grown in suspension at densities between 2 and 6 x ~o 5 cells/ ml. in Eagle's medium supplemented with 5 ~ horse serum (Grand Island Biological Company, New York). Propagation, labelling and purification of virus. Cells were concentrated and suspended at a density of Io 7 cells/ml, in serum-free medium and infected in the presence of 5 #g./ml. actinomycin D (gift of Bayer AG, Leverkusen, Germany) at a multiplicity of 5 to 3 o p.f.u./ cell (Levintow & Darnell, I96o). Tritiated amino acids or lac-labelled uridine were added after 2½ hr. The following radioactive isotopes were purchased from the Radiochemical Centre, Amersham, Buckinghamshire: L-leucine, 250 to looo mc/m-mole; [ah]dl-isoleucine, 5o to 350 mc/m-mole; L-proline, Ioo to Iooo mc/m-mole; [ah]dl-serine, 50 to 250 mc/mmole; [ah]dl-valine, 250 to 5oo mc/m-mole; pac]uridine, 4o0 mc/m-mole. Six hr after the initiation of infection, cells were frozen and thawed three times, and the virus Was purified by a procedure similar to the one described by Phillips, Summers & Maizel (I968). The suspension was made I ~o with respect to sodium dodecyl sulphate and stirred at room temperature for I5 min. Cell debris was removed by low-speed centrifugation and the virus was sedimented by centrifugation in the rotor 3o of the Spinco L ultracentrifuge (3o,ooo rev./min., IO, 2 hr). The pellets were suspended in 0.02 M-phosphate-buffered o'i5 M-NaC1, ph 7.2 (buffered saline) and homogenized in a Braun homogenizer. Large particles were removed by low-speed centrifugation, and CsC1 was added to a final density of 1.33 g./cm-a. Virus was banded by overnight centrifugation (SW 65 rotor, 45,ooo rev./min., 6 ), and collected with a syringe from the top of the tube. After removing CsC1 by dialysis, virus was further purified by sucrose gradient centrifugation (15 to 30 ~ (w/v) sucrose in buffered saline, SW 4o rotor, 4o,ooo rev./min., 6, 9 min.), and a second equilibrium centrifugation in caesium chloride. Unless otherwise stated, virus was suspended in buffered saline. The virus preparations gave the characteristic u.v. absorption spectra with maxima at 26o nm., minima at 24o nm., and ratios of Ez6o/E28o of I'69 to I"72 (Schwerdt & Schaffer, I955). Titration of infectivity. Titres of virus and/or RNA infectivity were determined by a plaque assay described by Koch, Quintrell & Bishop (I966). RNA dilutions were made in serum-free medium containing DEAE-dextran (2oo #g./ml., mol. wt 2 x io6---pharmacia, Uppsala, Sweden) and Io ~ dimethyl sulphoxide. (G. Koch, personal communication). Sucrose gradient centrifugation. One ml. samples were layered on I2 ml. linear I5 to 30 ~o (w/v) sucrose gradients containing phosphate-buffered saline and I mm-edta unless otherwise stated. Centrifugation was performed in the Spinco SW 4o rotor at 4o,ooo rev./ rain. and I5. After appropriate centrifugation the tubes were punctured at the bottom. If required, extinction at 26o nm. was monitored by a Zeiss PMQ II spectrophotometer, and fractions were collected by a LKB Ultrorac. Measurement of radioactivity. Total or trichloroacetic acid precipitable radioactivity was measured in a Packard Tri Carb scintillation spectrometer as described by Bishop & Koch (I969). Acrylamide gel electrophoresis. Acrylamide gel electrophoresis of poliovirus capsid proteins was done as described by Summers, Maizel & DarneU {I965). After electrophoresis

3 The structure of heated polio virus particles gels were stained with Coomassie brilliant blue (Maizel, 1966), de-stained electropho retically, and the bands were recorded photometrically. Determination ofantigenicity. Antigenicity of virus preparations was determined by double diffusion in agar (Ouchterlony, I948). Rabbit hyperimmune sera were kindly provided by Drs O. Drees and J. Drescher. Monoreactive D antiserum was prepared by the adsorption method of Hummeler & Tumilowicz (I96o). Adsorption of virus to HeLa cells. Cells were washed and concentrated in serum-free medium at densities between 3 and 6 x ro 7 cells/ml. Labelled virus was added, and adsorption was permitted under continuous stirring at 6. After appropriate times cells were sedimented by centrifugation, and unadsorbed virus was determined by measuring radioactivity in the supernatant fluid. Electron microscopy. Drops of virus suspensions were placed on formvar grids and examined after negative staining with 2 % phosphotungstic acid at ph 6"7. The Siemens Elmiskop Ia was used at 80 kv and an electron microscopic magnification of4o,ooo x. The electron microscopical analysis was made by Dr L. Luciano, Institute of Anatomy, Department of Electron Microscopy, Medizinische Hochschule, Hannover, W. Germany. The formation of 8os RNP RESULTS As reported previously heat degradation of poliovirus under appropriate conditions results in the formation of virus-like 8os particles containing RNA (Breindl, i969). In order to get information on the conditions of formation and the properties of these particles, a series of heat degradation experiments was done with purified poliovirus preparations and the resulting degradation products analysed by sucrose gradient centrifugation. A poliovirus preparation labelled with [3H]amino acids and p4c]uridine was heated for Io rain. at 5 in buffered saline and then centrifuged in a sucrose gradient. Three peaks of radioactivity were obtained under these conditions (Fig. I). One sedimented at about 8os and consisted of RNA and protein; this peak represented a mixture of two components, namely, empty poliovirus capsids, and virus-like particles containing RNA and designated as 8os ribonucleoprotein (8os RNP). The second peak sedimented with about 35s and represented free poliovirus RNA, and the third one was a protein not sedimenting under the conditions employed in these experiments. This was one of the virus capsid proteins termed VP 4 and known to be liberated from the particle by heat treatment (Maizel et al. I967). The determination of the sedimentation coefficient of the 8o s RNP was done by simultaneous centrifugation of empty poliovirus capsids in buffered saline obtained by heating poliovirus for 3o rain. at 56 and subsequent incubation with RNase (Ioo #g./ml.). The 35s mark was determined by centrifugation of virus RNA extracted from purified poliovirus by phenol. The relative amounts of 8os RNP and 35s RNA varied somewhat from experiment to experiment and were dependent on the conditions of heat degradation. Low salt concentrations (o.oi M-phosphate buffer ph 7"z) impaired the formation of the 8os RNP. The bulk of the RNA sedimented in the 35 s peak under these conditions. In a sucrose gradient containing o-oi M-phosphate buffer and no NaC1, the sedimentation rates of both the 8os RNP and 35s RNA were somewhat lower than in o'i5 M-NaC1, whereas the sedimentation rate of empty capsids was not affected. Accordingly, the protein label sedimented as an asymmetrical peak in the 8os region, indicating the heterogeneity of the material. High salt concentration (I"5 M-NaC1) stabilized poliovirus against heat degradation. No 8os RNP or 35s RNA was found after io min. heating at 5 o, a result in agreement with the wellknown phenomenon of cationic stabilization of enteroviruses (Wallis & Melnick, I96I). i49

4 15o M. BREINDL Increasing the temperature of heat treatment to 56o had no effect on the nature of the degradation products but altered the relative amounts of 8o s RNP and free RNA. After 2 min. heating at 56, 75 to 9o % of the RNA sedimented in the 8os peak while after I min. at this temperature 60 to 8o ~o of the RNAwas found as 35s RNA. Since the best yields of I I I I i a 80s 35s 1 A._= 200 E g 200.E E g u io Fraction no. I I I I I I I I I I b 80s 35s c 80s 35s ~, oo-!\'. i AA ~ I00 I Fraction no. Fig. 1. Sucrose gradient centrifugation of poll virus preparations labelled with [3H]amino acids and [z4c]uridine; a after I rain. heating at 5o ; b after heating and treatment with ~ % sodium dodecyl sulphate at neutral ph and room temperature; c after heating and digestion with RNase. A--,, 3H (counts/rain.), 0, 14C (counts/rain.). 8os RNP were obtained by heating virus suspended in phosphate-buffered saline for 2 rain. at 56, the further characterization of this component was done with preparations produced in this manner. Prolongation of the heat treatment at 5 o to 30 rain. did not significantly alter the ratio of 8os RNP and 35s RNA, indicating that the 8os RNP was relatively stable under these conditions.

5 _ m The stability of the 80 s RNP The structure of heated poliovirus particles 151 To find out under which conditions the Sos RNP was stable, virus preparations were heated and treated with certain chemicals before sucrose gradient analysis (Table 0. The decreased stability of the 8os RNP as compared to the intact virus particle could be demonstrated by its sensitivity at room temperature to t ~ sodium dodecyl sulphate at ph conditions under which intact poliovirus is stable (Mandel, I964) - and by the susceptibility of its RNA to RNase (too #g./ml., ph 7.2, 20 min. at room temperature). By treatment with sodium dodecyl sulphate, 35s RNA was released from the 8os RNP, indicating that it contained intact RNA. Table I. The stability of the 80 s RNP Stability of the Conditions of, heat treatment Additional treatment 8os RNP 35 s RNA IO rain., 5o in 1% sodium dodecyl sulphate, ph 7"2, - + buffered saline room temperature ioo t~g. RNase, ph 7'2, room temperature ~o% dimethyl sulphoxide + + 2oo #g. DEAE-dextran/ml.* g./cm? CsCI +* + * See text. Additional treatment of the heated poliovirus preparations with to ~ dimethyl sulphoxide or DEAE-dextran (2oo/~g./ml.) did not destroy the 8os RNP, but DEAE-dextran increased the sedimentation rates, both of the 8os RNP and of the 35 s RNA (unpublished observations). The addition of CsC1 to the heated preparation to a final density of 1.33 g-/ cm? did not influence the degradation products. After removal of the CsC1 by dialysis both components sedimented with the normal rates. In contrast, by equilibrium centrifugation in CsC1 the RNA was removed from the 8os RNP, the protein of which banded at t.29 g./cm?- the buoyant density of empty poliovirus capsids (Jamison & Mayor, I966). The addition of 2 ~ formaldehyde known to stabilize ribonucleoproteins (Spirin, Belitsina & Lerman, I965) did not prevent the 8os RNP from breakdown by CsC1 equilibrium centrifugation. Electron microscopy of the 8o s RNP It was of special interest to investigate the electron microscopical appearance of the 8os RNP. This was done by analysis of negatively-stained preparations from the 8os region of a sucrose gradient centrifugation of heated poliovirus.partmes of both empty and fullappearance were found in these preparations (Fig. 2b). The apparently full 8os particles seemed to be more heterogeneous and irregular in shape than the intact virus particles; some of them appeared partially penetrated by phosphotungstate, whereas others were indistinguishable from intact particles by this technique. After RNase treatment RNA-containing particles were no longer present (Fig. 2 c). In some cases strands similar to those described by McGregor & Mayor (t968) were observed in heated preparations before centrifugation in sucrose gradients. However, these seemed to be artificial aggregates dependent on the ionic strength of the virus suspension rather than products of physiological significance (Fig. 2d).

6 I52 M. B R E I N D L Infectivity of the 80 s R N P The release of 35s R N A from the 8os RNP particles after treatment with sodium dodecyl sulphate suggested that they contained intact virus RNA. This was proved by testing the infectivity of this component. The infectivity of the 8o s RNP as well as of the 35 s R N A was below the level of detection under normal conditions, i.e. without addition of a polybasic compound (Fig. 3). All infectivity under these conditions was due to virus that had escaped Fig. 2. Electron micrographs of poliovirus particles; a, unheated poliovirus; b, preparation from the 8us region of a sucrose gradient centrifugation of heated poliovirus; c, empty capsids; d, aggregates of strands and virus particles. beat inactivation as can be seen by its resistance to RNase. The addition of DEAE-dextran and dimethyl sulphoxide caused an increase of infectivity both in the 8us and in the 35s region, showing that the 8us RNP contains infectious virus RNA. Liberating the R N A by treatment with sodium dodecyl sulphate had no influence on the specific infectivity of the 80 s RNP. RNase destroyed the infectivity both of the 80 s RNP and the 35 s RNA. Under the conditions employed in these infectivity titrations, no difference between the specific

7 The structure of heated poliovirus particles I53 I I 80s $ I I I 35s O r~ 1 J 1! 100 z~ ) Fraction no. I 25 Fig. 3. Infectivity of poliovirus heat degradation products. Heated poliovirus preparations were centrifuged in sucrose gradients, and samples of each fraction were analysed for infectivity:.~., without polybasic compound; (3--0, with DEAE-dextran and d/methyl sulphoxide; ±--±, with DEAE-dextran and dimethyl sulphoxide after treatment with sodium dodecyl sulphate; A--A, with DEAE-dextran and dimethyl sulphoxide after treatment with RNase. VP2 VP1J VP3 VP4 o LU ' Fig. 4. Polypeptide composition of the 8os RNP (above) and intact poliovirus (below). A

8 154 M. BREINDL infectivity of the 8o s RNP and 35 s RNA was detected, but under other conditions the protein moiety of the 8o s RNP might influence on the infectivity of the RNA, Polypeptide composition of the 8os RNP Acrylamide gel electrophoresis showed that VP 4 was absent from the 8o s RNP (Fig. 4). Antigenic and adsorbing properties of the 8os RNP. The 8os RNP had the antigenic C reactivity (Fig. 5)- In an adsorption experiment under the conditions applied, 9 2 ~oo of the intact virus particles adsorbed to HeLa cells, but only about 7 ~ of the 8os RNP adsorbed under identical conditions (Fig. 6). 8O -~ 60 o to 4o "'71 l I T I'm A~A-- A.j 20 Fig. 5 Fig Time at 6 (rain,) Fig. 5. Antigenicity of the 8o s RNP. The figure shows a tracing of a photograph taken from an Ouchterlony plate. Holes z and 7, D-antigen; I and 8, C-antigen; 4 and 5, 8os RNP; 3, rabbit hyperimmune serum; 6, monoreactive D-antiserum. Fig. 6. Adsorption of the 8os RNP ((3 (3) and unheated poliovirus (J, A) to HeLa cells at 6. DISCUSSION In a previous publication (Breindl, I969) I reported that one of the poliovirus heat degradation products was a component containing protein and infectious virus RNA and sedimenting in a sucrose gradient at about 80 s, i.e. at nearly the same rate as empty poliovirus capsids. In this paper, a further characterization of this component designated as Sos ribonucleoprotein (8os RNP) is given. The sedimentation coefficient of the 8os RNP has not been exactly determined, but slight differences between the sedimentation rates of the 8os RNP and empty capsids seem to exist at least when they are centrifuged in sucrose gradients containing a low salt concentration. At present there is no explanation for the fact that the RNA containing 8os RNP sediments at almost the same rate as empty capsids and at half the rate of intact poliovirus. This could be due to a change of the virus surface and/or stability caused by the liberation of VP 4. Moreover, it may be that the RNA in the 8os RNP is no longer totally enclosed by the capsid, thus causing a reduction of the sedimentation velocity. This possibility is supported by the findings that alteration of salt concentration as well as addition of DEAE-dextran affect the sedimentation rates of 8os RNP and 35s RNA but not of empty capsids.

9 The structure of heated poliovirus particles 155 Joklik & Darnell (i961) described unstable poliovirus particles obtained by elution of virus adsorbed to HeLa cells. Although these particles resemble the 8os RNP in some properties (e.g. breakdown during CsC1 equilibrium centrifugation, loss of adsorbing properties and infectivity in spite of tile presence of infectious RNA), they sediment at the same rate as intact poliovirus and are relatively insensitive to RNase digestion. It is therefore not possible to decide whether the liberation of one of the capsid proteins also caused this alteration in the properties of the virus, and, accordingly, if this is part of the natural uncoating mechanism. Recent studies on plasma membranes (Chan & Black, 197o) suggested that the virus was more labile after interaction with plasma membranes of susceptible cells. Defective or unstable virus particles have been described for several bacteriophages (Sugiyama, Hebert & Hartmann, 1967; Hohn, 1967; Roberts & Argetsinger Steitz, i967; Rossomando & Zinder, 1968). They show striking similarities to the poliovirus 8os RNP characterized in this paper, in that they have little or no infectivity, a reduced sedimentation coefficient, are sensitive to RNase, and fail to adsorb to the host cell. Since these properties of the phage particles are localized on, or determined by, one minor capsid protein termed maturation protein (A protein) the results presented in this paper suggest that a functionally comparable protein also exists in poliovirus. The idea that the VP 4 in poliovirus could serve as maturation protein is favoured by findings concerning the morphogenesis of the virus, particularly by the fact that it is not present in its final configuration in the procapsid, a precursor of the virus particle (Jacobson & Baltimore, 1968). However, another problem is raised in this connexion; since empty poliovirus capsids and 80 s RNP sediment at nearly the same rate in sucrose gradients, it should be determined whether or not the 8os particle identified as poliovirus precursor (Jacobson & Baltimore, 1968) contains RNA. An elucidation of this question would be significant for our understanding of poliovirus morphogenesis. I wish to thank Dr G. Koch for the support of this work and his helpflfl criticism. Moreover, I am thankful to Mrs M. Kauls for her expert technical assistance, Dr L. Luciano for the electron microscopic examination, and Dr A. Shatkin for his discussion during the preparation of the manuscript. This investigation was supported in part by the Deutsche Forschungsgemeinschaft. REFERENCES msnov, s. M. & KOCH, G. (I969). Infectious replicative intermediate of poliovirus: purification and characterization. Virology 37, 521. BREII~DL, M. (I969). Zur Struktur hitzebehandelter Poliovirus-Partikeln. Hoppe Seyler's Zeitsehrift fiir Physiologisehe Chernie 350, crt~n, v. v. & BLACK, F. L. (1970). Uncoating ofpoliovirus by isolated plasma membranes. Journal of Virology 5, 3o9. DREES, O. & BORNA, CH. (1965). Ueber die Spaltung physikalisch intakter Poliovirus-Teilchen in Nucleinsgure und leere Proteinhiillen durch W~irmebehandlung. Zeitsehrift fiir Natmforschung 2ob, 87o. GRAHAM, A. F. (I959). Symposium on the Biology of cells modified by viruses or antigens. III. Physiological conditions for studies of viral biosynthesis in mammalian cells. Bacteriological Reviews 23, 224. HINUMA, Y., KATAGIRI, S., FUKUDA, M., I~UKUSm, K. & WATANABE, Y. (I965). Kinetic studies on the thermal degradation of purified poliovirus. Biken's Journal 8, 143. I-IOI~N, T. (1967). Selfassembly of defective particles of the bacteriophage ft. European Journal of Biochemistry 2, 152. HUMMELER, K. & TUMILOWICZ, J. J. (I960). Studies on the complement-fixing antigens of poliomyelitis: II. Preparation of type-specific anti-n and anti-h indicator sera. Journal of Immunology 84, 63o. JACOBSON, M. F. & BALTIMORE, D. (1968). Morphogenesis of poliovirus: I. Association of the viral RNA with coat protein, Journal of Molecular Biology 33, 369. JAMISON, R. M. & MAYOR, H. D. (I 966). Comparative study of seven picornaviruses of man. Journal of Bacterio- Iogy 9 x, JOKLtK, W. K. &DARNELL, J. E., JUN. (I96~). The adsorption and early fate of purified polioviruses in HeLa cells. Virology I3, 439.

10 156 M. BREINDL KATAGIRI, S., HINUMA, Y. & ISHIDA, N. (1967). Biophysical properties of poliovirus particles irradiated with ultravio]et light. Virology 32, 337. KATAGIRI, S., HINUMA, Y. & ISHIDA, N. (I968). Relation between the adsorption to cells and antigenic properties in poliovirus particles. Virology 34, 797. KOCH, G. (1960). Influence of assay conditions on infectivity of heated poliovirus. Virology 12, 6Ol. KOCH, G., QUINTR~LL, N. & BISHOP, J. M. (1966) An agar cell-suspension plaque assay for isolated viral RNA. Biochemical and Biophysical Research Communications 24, 3o4. LE ~OUVlER, G. L. (1955). The modification of poliovirus antigens by heat and ultraviolet light. Lancet 2, loi3. LEVlNTOW, L. & DARNELL, J. E., StrN. (I960). A simplified procedure for purification of large amounts of poliovirus: Characterization and amino acid analysis of type I poliovirus. Journal of Biological Chemistry 335, 70. MAmEL, S. V., JUN. (1966). Acrylamide-gel electrophorograms by mechanical fractionation: Radioactive adenovirus proteins. Science, New York I5I, 988. MAIZEL, 1. V., JUN., t'mlln's, B. A. & St~RS, D. F. (I967). Composition of artificially produced and naturally occurring empty capsids of poliovirus type I. Virology 32, 692. MANDEL, n. (1964). The extraction of ribonucleic acid from poliovirus by treatment with sodium dodecyl sulfate. Virology 22, 36o. MAYER, M. M., RAPP, H. J., ROIZMAN, B., KLEIN, S. W., COWAN, K. M., LUKENS, D., SCHV~ERDT, C. E., SCHAFFER, F. L. & CrIARNEY, J. (I957). The purification of poliomyelitis virus as studied by complement fixation. Journal of Immunology 78, 435. MCGREGOR, S. & MAYOR, n. D. (I968). Biophysical studies on rhinovirus and poliovirus: I. Morphology of viral ribonucleoprotein. Journal of Virology 2, I49. OUCHTERLONV, O. (1948). In vitro method for testing the toxin producing capacity of diphtheria bacilli. Acta pathologica et microbiologica scandinavica 25, 186. VmLLrVS, B. A., SUMMERS, O. V. & MAIZEL, 3. V., attn. (I968). In vitro assembly of poliovirus-related particles. Virology 35, 216. ROBERTS, J.W. & ARGETSINGER STEITZ, J.E. (I967). The reconstitution of infective bacteriophage R l7. Proceedings of the National Academy of Sciences of the United States of America 58, 14 ~ 6. ROSSOMANDO, E. V. & ZINDER, N. D. (I 968). Studies on the bacteriophage ft. I: Alkali-induced disassembly of the phage into DNA and protein. Journal of Molecular Biology 36, 387. SCHWERDT, C. E. & SCHAF~ER, F. L. (1955). Some physical and chemical properties of purified poliomyelitis virus preparations. Annals of the New York Academy of Sciences 6x, 74o. SPIRIN, A. S., BELITSlNA, ~. V. & LER~N, M. I. (I965). Use of formaldehyde fixation for studies of ribonucleoprotein particles by caesium chloride density-gradient centrifugation. Journal of Molecular Biology x4, 611. SUGIYAMA, T., HEBERT, R. R. &HARTMANN, K. A. (1967)- Ribonucleoprotein complexes formed between bacteriophage MSz RNA and MS2 protein in vitro. Journal of Molecular Biology 25, 455. SUMMERS, O. V., M~ZEL, J. V., JUN. ~ DARNELL, J. E., JUN. (I965). Evidence for virus-specific noncapsid proteins in poliovirus-infected HeLa cells. Proceedings of the National Academy of Sciences of the United States of America 54, 5o5. WALLIS, C. & MELN~CK, 3. L. (2961). Stabilization of poliovirus by cations. Texas Reports on Biology and Medicine I9, 683. (Received 5 October 197 o)