Further characterization of mengo subviral particles: A new hypothesis for picornavirus assembly

VIROLOGY

97, 266-274 (1979)

Further Characterization A New Hypothesis

of Mengo Subviral Particles: for Picornavirus Assembly

PATRICK W. K. LEE’ AND JOHN S. COLTER2 Department of Biochemistry, University of Alberta, Edmonton, Alberta T6G 2H7, Canada Accepted May 12, 1979 The “50s particle” found in Mengo virus-infected L cells (P. W. K. Lee, E. Paucha, and J. S. Colter, 1978, Virology 85, 286-295) has been further characterized. Its sedimentation coefficient has been estimated to be 53 S from centrifugal analysis in sucrose density gradients. When, during the isolation of 53 S particles, the KC1 concentration in the suspending buffer is increased to 150 mM or higher, some of the particles are converted to structures having a significantly larger sedimentation coefficient. A similar conversion of 53 S to more rapidly sedimenting particles occurs when the former are centrifuged to equilibrium in a CsCl density gradient. The sedimentation coefficient of this new particle has been estimated to be 75 S. Molecular weight determinations of the previously described 14 S particles and of the 53 and 75 S particles by means of Sepharose 4B exclusion chromatography suggest that the molecular compositions of these particles are (~a-&, (~ay)*~, and (~ay)~,,,respectively. Based on this information and the previously reported evidence suggesting a precursor role for the 53 S particles, a new hypothesis regarding the mechanism of Mengo virus assembly has been proposed. In this model, the viral RNA interacts with either a 75 S particle or two 53 S particles to form a complex represented by RNA[(eay)&, before assembly is completed by the addition of two 14 S subunits.

precursor-product relationships between subviral particles and mature virions. It is generally accepted that all picornaIn the case of poliovirus, the most extenviruses share a fundamental similarity with sively studied of the picornaviruses, four respect to gross morphology, physical and subviral particles, having sedimentation cohydrodynamic properties, comp&ition, and efficients of 5, 14, 73, and 125 S, have been the mechanisms by which their proteins and described (for a review, see Casjens and RNA are synthesized. However, the eviKing, 1975). Although there is little doubt dence that members of the different subthat the 14 S structure, with the composigroups follow a common pathway of virion tion (VPO, 1, 3),-or (c(~y)~using the carassembly is less convincing. Uncertainty in diovirus subgroup nomenclature -is a prethis respect can be attributed to the followcursor in the assembly pathway of all piing factors. First, with the exception of cornaviruses, the role of the 73 S particle poliovirus, information regarding the pres(empty capsid) as an intermediate has been ence of subviral particles in picornavirusinfected cells is fragmentary. Second, prog- questioned, partly because there is no convincing evidence that comparable strucress has been hampered by the inherent difficulty associated with kinetic pulse- tures are produced in cells infected with chase experiments designed to establish other members of the picornavirus family. In an earlier publication (Lee et al., 1978), 1 Present address: Department of Microbiology and the discovery of a “50s particle,” in addition Immunology, Duke University Medical Center, Dur- to the 14 S structure, in Mengo virusham, N. C. 27710. infected cells was described. Evidence from * To whom requests for reprints should be addressed. pulse-chase experiments and from comINTRODUCTION

0042-6822/79/120266-09$02.00/O Copyright All rights

0 1979 by Academic F’rew, Inc. of reproduction in any form reserved.

266

MENGO

VIRUS

positional analyses which suggested that the “50s particle” may be an intermediate in the assembly of Mengo virions was presented. In this paper, further characterization of Mengo subviral particles is described. The results of these studies, coupled with previous observations, provide the basis for a new hypothesis concerning the morphogenesis of the Mengo virion. MATERIALS

AND METHODS

Cells and virus. Cells of Earle’s L-929 strain of mouse fibroblasts were grown as monolayers in either 100 x G-mm plastic petri dishes or 110 x 425-mm glass roller bottles in Eagle’s basal minimum essential medium (BME) supplemented with 5% horse serum (HS). The virus used was the plaque variant of Mengo virus designated as M-Mengo (Ellem and Colter, 1961). The procedures for the production, radioactive labeling, and purification of the virus have been described previously (O’Callaghan et al., 19’70; Ziola and Scraba, 1974). Infection of cells and isolation of radioactive virus-spec@c particles. The infection of cells with Mengo virus, the labeling of viral polypeptides, the homogenization of cells, the preparation of S,, and P,, fractions, and the isolation of labeled virions, “50s particles,” and 14 S particles from sucrose density gradients have been described previously (Lee et al., 1978). The S,, fraction is defined as the supernatant obtained by centrifuging a cell homogenate in TBS (20 mM Tris-HCl, pH 7.4, 100 mM KCl, 5 mM MgC&, 6 mM /3-mercaptoethanol) for 20 min at 20,000 g (13,000 rpm, JA-20 rotor, Beckman 521 centrifuge). The pellet obtained by centrifuging the Sz, fraction for 5 hr at 45,000 rpm (type 50 rotor; Beckman L5-65 centrifuge) is referred to as the P,, fraction. Sepharose 43 column chromatography. Sepharose 4B, as a suspension in distilled water, was obtained from Pharmacia Fine Chemicals. After repeated washings in TBS the Sepharose 4B was suspended in TBS, poured into a column (80 x 1 cm diameter) and packed slowly at 4” by passing the same buffer throuph the column at a flow rate of

ASSEMBLY

267

lo-12 ml/hr. Following a further packing period of 7 to 10 days, the void volume of the column, determined as the elution volume of calf thymus DNA, was found to vary by less than 0.5 ml over a period of several weeks. Chromatography was performed at 4”. In most cases, the sample whose elution volume (V,) was to be determined was mixed with 4 OD,,, units of calf thymus DNA (for determination of void volume, V,) and approximately 10,000 cpm of [14C]thymidine (for determination of the total volume of the gel bed, VJ to make up a total loading volume of 0.8 ml. The column was eluted with TBS at a flow rate (monitored by means of an LKB peristaltic pump) of 12 ml/hr. Fractions of approximately l-ml volume were collected. Depending on the sample, aliquots of appropriate fractions were taken for optical density measurements at either 260 or 280 nm and for radioactivity measurements in the presence of 10 ml of Aquasol scintillation fluid (New England Nuclear Corp.) in a Beckman liquid scintillation spectrometer (Model LS-230). Molecular weight determinations of subviral particles. Sepharose 4B chromatography data are expressed in terms of K,,, a parameter which is defined as follows:

(1) where V, = elution volume of a solute; V, = void volume of the column; and V, = total volume of the gel bed. The K,, of a molecule is in turn correlated with its Stokes radius according to the following equation (Laurent and Killander, 1964). (-log K,P = (r(p + a), (2) where a = Stokes radius, and (y:and /3 are constants related to the intrinsic properties of the gel matrix. The validity of these theoretically derived relationships has been confirmed experimentally by Siegel and Monty (1966) who demonstrated a linear relationship between elution volume expressed in terms of K,, and Stokes radius for a variety of proteins. With a Stokes radius measured by means of a column that has been nreviouslv cali-

LEE AND COLTER

268

TABLE 1 CALCULATED STOKESRADIIOF MARKERPARTICLES

Mengo virion Mengo virus 13.4 S subunit R17 phage

Molecular weight (X 106)

Sedimentation coefficient (x 10-l” see)

Partial specific volume (ml@

Stokes radius” (-4

8.32’

1516

0.70*

145

0.73d 0.68”

76 139

0.42Se 3.8Se

13.4d 78”

(2Calculated by means of Eq. (3) (see Materials and Methods). b Scraba et al. (1967). e Calculated from the reported composition (c&& (Mak et al., 1971) and molecular weights of individual polypeptides (Ziola and Scraba, 1974). d Mak et al. (1971). e Boedtker and Gesteland (1975).

Chemicals and radioisotopes. Calf thymus DNA, cordycepin, and bovine fibrinogen were purchased from Sigma Chemical Company, St. Louis, Missouri. [3H]Uridinelabeled R17 bacteriophage was a gift from Dr. S. Igarashi of this department. All radioisotopes were purchased from New Eng6+Vas land Nuclear: 3H-amino acids (1.0 mCi/ml), M= (3) 14C-amino acids (0.1 mCi/ml), [3H]uridine (1 - Vp) ’ (24.2 Cilmmol), [3H]uracil (40-50 Wmmol), where M = molecular weight, a = Stokes and [14C]thymidine (54 mCi/mmol). radius, s = sedimentation coefficient, t, = partial specific volume, r) = viscosity of RESULTS medium, p = density of medium, and N Sedimentation Coejjicient of the “50s = Avogadro’s number. Particle” The markers used in the chromatographic studies were fibrinogen, [3H]uridine-labeled The sedimentation coefficient of the “50s R17 bacteriophages, 14C-aminoacid-labeled particle” was estimated from measureMengo virions and 13.4 S subunits prepared ments of its rate of sedimentation in 15-45s therefrom (Mak et al., 1971). The elution linear sucrose density gradients relative to volumes (V,), and hence the K,,, of these those ofEscherichia coli ribosomal subunits markers were thus easily measured by mon- (50 and 30 S). Illustrative data are shown itoring either the OD,,, (in the case of fibrin- in Fig. 1, in which the positions of L-cell ogen) or radioactivity (in the case of R17, ribosomal subunits in identical gradients Mengo virus, and Mengo 13.4 S subunit) of are also indicated. From the results of these appropriate fractions. The Stokes radius of studies, the sedimentation coefficient of this fibrinogen has been calculated to be 107 A subviral particle was calculated to be 53 S. (Siegel and Monty, 1966). The Stokes radii of the Mengo virion, the 13.4 S Mengo viral Effect of Salts (KC1 and C&l) on the Sedisubunit, and of R17 were calculated by mentation Behavior of 55 S Particles means of Eq. (3), using the known molecular weights, sedimentation coefficients, and It has been shown previously that the partial specific volumes of these markers recovery of 53 S particles (and of mature (Table 1). virions) from cell homogenates is strongly brated with markers of known Stokes radii, and a sedimentation coeffici+ determined by sucrose density gradient centrifugation, reasonable estimates of the molecular weight of a macromolecule can be obtained by employing the following equation:

269

MENGO VIRUS ASSEMBLY

FRACTION

NUMBER

FIG. 1. Sedimentation analysis of a Pd5fraction isolated from Mengo virus-infected cells pulse-labeled with “C-amino acids (5 &i/ml) for 20 min at 5 hr postinfection and harvested after a chase period of 30 min. The P,s fraction was resuspended in a low Mg2+ TBS (0.5 mM MgCl& to which purified E. coli ribosomes ([3H]uracil-labeled) were also added. Centrifugation through a 16-ml15-45% linear sucrose gradient was for 13.5 hr at 20,000 rpm (Beckman 27.1 rotor). Gradient fractions were analyzed for acid-insoluble W counts (0) and 3H counts (0). The positions of L-cell ribosomal subunits (40 and 60 S) were determined from a parallel gradient.

dependent upon the concentration of KC1 in the suspending buffer. Only trace amounts are recovered at concentrations of less than 60 mM, while maximum recovery is achieved at a concentration of 100 mM (Lee et al., 1978). The present study revealed that at concentrations of KC1 higher than 100 mM, 100

mM

I50

KCI

mM

the sedimentation behavior of the 53 S particles is altered. Infected cultures were pulse-labeled with 3H-amino acids for 20 min at 5 hr postinfection, after which they were incubated for an additional 75 min in BME-5% HS containing cordycepin (200 pgl ml) before being harvested and homogenized. The homogenate was then divided into three aliquots in which the KC1 concentration was adjusted to 100,150, and 200 mM, respectively, prior to the separation of P,, fractions. Sucrose density gradient analyses gave the results shown in Fig. 2. The data suggest that when the P,, fraction is sedimented from solutions containing a KC1 concentration of 150 mM or higher, some of the 53 S particles are converted to a more rapidly sedimenting species, the fraction so converted being dependent on the KC1 concentration. A similar shift in sedimentation behavior was observed with 53 S particles that had been centrifuged to equilibrium in a CsCl gradient (the particles band sharply at a buoyant density of 1.296 g/cm3). When such particles were recovered from the CsCl gradient, dialyzed extensively against TBS, and reexamined in a 15-45% sucrose density gradient, they were found to band in a position very close to that occupied by marker R17 (78s) virions (Fig. 3). The sedimentation coefficient of these new narticles was estimated to be 75 S.

KCI

200

a 20 .

0

mM

KCI

150s

150s

150 s

IO

20

30

IO FRACTION

20

30

IO

20

30

NUMBER

FIG. 2. Sedimentation analyses of P,, fractions separated from three aliquots of an homogenate of Mengo virus-infeeted, pulse-labeled cells in which the KC1 concentration had been adjusted to 100, 150, and 200 mM. Centrifugation conditions were as described in the legend to Fig. 1.

LEE AND COLTER

270

FRACTION

NUMBER

FIG. 3. Effect of CsCl on the sedimentation behavior of 53 S particles. W-amino acid-labeled 53 S particles recovered from a CsCl gradient were dialyzed for 48 hr against TBS at 4” and then reexamined in a 16-ml 15-458 sucrose density gradient. Centrifugation was for 13.5 hr at 20,000 rpm (Beckman 27.1 rotor). [3H]Uridine-labeled R1’7phage (78 S) was also present in the same gradient as a marker. (0) 14Ccounts, (0) 3H counts. The position of 53 S particles (before CsCl treatment) as determined in a parallel gradient is also indicated.

Determination of the Molecular Weights of the 14, 55, and 75 5’ Particles Data obtained from previous studies (Lee et al., 1973) suggested not only that the 53 S particle is an intermediate in the assembly of Mengo virions but that the 14 S particle is the immediate precursor of the 53 S particle. It became important then to establish the stoichiometric relationship between these two particles, as well as that between 53 and 75 S particles. The conversion of 53 S particles to particles having a sedimentation coefficient of 75 S, either by banding in CsCl or by exposure to elevated concentrations of KC1 during the isolation procedure, could be due either to some sort of aggregation phenomenon or to a marked conformational change from what would be a highly asymmetric 53 S particle to a more compact 75 S particle. Since this information appeared to be important with respect to gaining some understanding of the mechanism of virion assembly, attempts to determine the molecular weights of all three particles were undertaken. It has been shown by Siegel and Monty (1966) that the elution position of a macromolecule subjected to exclusion chromatography through a column of Sephadex G-200 is a function of its Stokes radius, and that

the Stokes radius of a macromolecule can be determined by measuring its elution behavior from a column calibrated with macromolecules of known Stokes radii. If both the sedimentation coefficient and the partial specific volume of the macromolecule are also known, an accurate determination of its molecular weight can be made by means of Eq. (3) (see Materials and Methods). Considering the size of the subviral particles involved, a Sepharose 4B column was employed. The column was calibrated with the following markers of known Stokes radii (in brackets): Mengo virion (145 A), R17 bacteriophage (139 A), fibrinogen (107 A), and the 13.4 S subunit derived from Mengo virus (76 A). The elution positions of these markers, as well as the elution patterns obtained with the 14, 53, and 75 S particles, are illustrated in Fig. 4. The void volume (V,) and the total bed volume (V,) of the column are indicated by the elution positions of calf thymus DNA and [14C]thymidine, respectively. The calibration curve of the column, obtained by plotting (-log K,y)l’z vs Stokes radii of the markers, is shown in ,

1

1

a4 .

.6

E 0.3. Q

.6

t Q2.

-

‘? 0.6

I f 0

E 4y

%

-e N B

45

0.1 .

.,

I . 20

40 FRACTION

60 NUMBER

-2

-2

,

J

00

FIG. 4. Gel exclusion chromatography of 14,53, and 75 S particles. Samples (0.8 ml) containing 4 OD,,, units of calf thymus DNA, approximately 10,000 cpm of [Wlthymidine, and 3H-amino acid-labeled 14, 53, or 75 S peak fractions from sucrose density gradients were applied to a 80 x l-cm column of Sepharose 4B equilibrated with TBS at 4”. The components of the mixture were then eluted from the column and the elution profiles were determined as outlined under Materials and Methods. The arrows indicate the elution positions of markers of known Stokes radii: a, Mengo virion; b, R17 bacteriophage; c, fibrinogen; and d, Mengo 13.4 S subunit.

271

MENGO VIRUS ASSEMBLY

lar weights of 2.23 and 4.45 x 106, respectively. Since both contain equimolar amounts of polypeptides E, (Y, and y, the combined molecular weights of which equal 95.2 x 103, the two particles may be represented, on the basis of these data, by WY)W-24 and (~a~)~~-~~Described in terms of the 14 S structure unit, the 53 and 75 S particles are best represented by [(vary& and [(~a$&,,, respectively.

a70 1. 0.65.

!t o.so. a 4A oas-

DISCUSSION

FIG. 5. Stokes radii of 14, 53, and 75 S particles

as

determined from their elution positions [in terms of (-log K&l’*] on a Sepharose 4B column. The Stokes radii of the markers were determined as outlined under Materials and Methods.

Fig. 5. Having determined the K,, values for the 14, 53, and 75 S particles from their elution positions from the calibrated column, the Stokes radii of all three particles were read from the calibration curve. The molecular weights of the particles were calculated from Eq. (3), using the experimentally determined Stokes radii and sedimentation coefficients, and the partial specific volume calculated from the amino acid composition of the particles. The results, which are shown in Table 2, indicate that the 53 and ‘75 S particles have molecuTABLE 2 MOLECULAR WEIGHTS OF 14,53, AND 75 S PARTICLES

Stokes radius” (A) 14 s

78

535 75 s

100 141

Molecular weight” (XlW 0.46 2.23 4.45

O1 Average of three separate determinations. The range of values for the 14, 53, and 75 S particles were 77-79, 99-102, and 139-143 A, respectively. * Calculated from Eq. (3), using the value 0.73 ml/g for partial specific volume determined by Mak et al. (1971).

Data obtained from pulse-chase experiments and from compositional analyses prompted us to propose that the 53 S particle is a precursor in the assembly of the mature Mengo virion (Lee et al., 19’78). Additional evidence that this may be the case has been obtained from studies reported elsewhere (Lee and Colter, 1979), in which it was shown that 53 S particles accumulate in cells infected with an RNAmutant of Mengo virus (ts 520, Downer et al., 1976) when cultures are shifted from the permissive to the nonpermissive temperature. The earlier data obtained from pulsechase experiments also suggest that the 14 S structure is the immediate precursor of the 53 S particle, a view substantiated by the observation that 53 S particles can be disassembled into 14 S structures by prolonged incubation at room temperature (data not shown). In this respect the 53 S particle resembles the 73 S particle found in the poliovirus system, since the latter has also been shown to be an aggregate of 14 S particles (Phillips, 1969, 1971). However, it seems unlikely, in view of the molecular weight data presented here, that the 53 and 73 S particles are strictly analogous structures. The finding that the 53 S particles can be converted, by what appears to be dimerization, to 75 S particles is not unique to the Mengo system. A similar observation was made by Su and Taylor (1976), who reported that 45 S particles, isolated in buffer containing 10 mM NaCl from bovine enterovirus-l-infected cells, can be converted to 80 S particles by dialysis against buffer containing 150 mM NaCI. The 53 + 75 S con-

LEE

272

AND

version suggests that the latter species may be analogous to the 73 S particles found in the poliovirus system and believed to be complete empty capsids. However, the estimated molecular weight of the 7.5S particle indicates that it may be an incomplete viral capsid [corresponding to (a,&r,i& as opposed to (c.u,P,-y,i$,,, for a complete capsid]. .A proposed scheme for the assembly of Mengo vu-ions, based on the data presented here, is shown in Fig. 6A. It envisions that five 14 S structure, units aggregate, presumably on smooth cytoplasmic membranes, to form 53 S particles, which in turn dimerize to form 75 S particles (incomplete capsids). After insertion of the viral RNA into the 75 S structure, the virion is completed by the addition of two more 14 S structure units and the final, morphogenic cleavage of most of the E polypeptides. An alternative assembly scheme-also compatible with the experimental evidence and shown in Fig. GB-is one in which the 75 S particle is not formed, but in which two 53 particles interact with a molecule of viral RNA to form an intermediate described by RNA [(eay&,. The insertion of two additional 14 S particles and the cleavage of E residues to p and 6 polypeptides would then complete the assembly process. There is some evidence, albeit circumstantial, that there may be something unique about two of the 14 S structure units that comprise the capsids of picornaviruses. Dunker and Rueckert (1971) reported that 2 of the pentameric subunits produced by the A.

(*a~& 14s

[bay) Jl53s

[(ay.)J,,, ISS RNA P(tayl, 1

B.

(‘=Y& 14s

~YI

J,=

~wn[(ay) J,.

53s

may&. 1 .wA[c~.Yl*c~lJ

-

-b-YJ.

ISOS VRKm

FIG. 6. Proposed alternate pathways of Mengo virus morphogenesis.

COLTER

pH dissociation of ME virus are distinguished from the other 10 by their insolubility in the dissociation buffer and the presence therein of an uncleaved E polypeptide. McGregor and Rueckert (1977) found that rhinovirus 1A contains approximately 10 E polypeptides per virion, the number that would be present in two 14 S structure units. In the light of the assembly scheme(s) proposed here, it is tempting to speculate that the uncleaved E residues found in a number of picornaviruses arise from the two 14 S structure units that complete the encapsidation of the viral genome. The assembly scheme shown in Fig. 6A may be criticized on the grounds that while stable 75 S particles can be formed readily in vitro from 53 S particles, they have not been detected in extracts of infected cells as prepared in these investigations. One can rationalize the failure to do so on the grounds that the assembly of virions is a highly concerted process and that as a result, the short-lived 75 S species may not accumulate to levels that make its detection possible. Be that as it may, the 75 S particle described here (i.e., an incomplete viral capsid) is in some ways a more credible putative intermediate in the assembly process than is a complete empty capsid. Certainly it is easier to envision how a viral genome could be packaged (inserted) into an incomplete capsid than into a complete one. The validity of our conclusions regarding the compositions of the 53 and 75 S particles (and hence the credibility of our proposed model for the assembly of the Mengo virion) depends, of course, on the accuracy of the estimates of the molecular weights of these particles. These in turn depend on the accuracy of the estimates of sedimentation coefficients, Stokes radii and partial specific volumes, and it may be worthwhile to consider the reliability of these values. Confidence in the technique of exclusion chromatography on Sepharose 4B as a method for the determination of Stokes radii has been engendered by the reproducibility of the results obtained from three separate estimations (+-2%), and the fact that the plot of (-log KaV)l12vs Stokes radius for marker particles of various compositions and sizes is strictly linear. Fur-

MENGO VIRUS ASSEMBLY

thermore there is very close agreement between the molecular weight of the 14 S particle (which contains the same polypeptides as do the 53 and 75 S particles) estimated by this method and the value calculated from its composition, (cay), (McGregor et al., 1975), and the molecular weights of the individual polypeptides (Ziola and Scraba, 1974). These considerations make it unlikely that the chromatographic behavior of 14, 53, and 75 S particles is influenced significantly by any parameter other than their Stokes radii. The use of sedimentation markers (ribosomes) having a smaller partial specific volume than do the 53 and 75 S particles could lead to underestimates in the sedimentation coefficients of the latter. However, if one assumes that the true sedimentation coefficients of these two particles are 58 and 82.5 S (i.e., that our values are 10% too low), calculations of their molecular weights by Eq. (3) gives values equivalent to molecular compositions of (bay),,-,, and values still consistent with our (vh52, pentamer-decamer model. The partial specific volume assigned to all the subviral particles considered here was calculated from the known amino acid compositions of their constituent polypeptides and is unlikely to be affected by more than *2% by conformational changes in going from one to the other (Kay, 1960). The possibility that the 53 and 75 S particles are hexamers and dodecamers of 14 S particles cannot be ruled out unequivocally, although such a view is difficult to reconcile with the data presented here unless one assumes that our estimates of Stokes radii, sedimentation coefficients, and partial specific volumes of these particles are all too low. Nonetheless it would be highly desireable to obtain estimates of the molecular weights of these particles by a completely independent (e.g., analytical ultracentrifuge) technique. We hope to do so, but to date have been unsuccessful in our efforts to isolate the two species in the quantities required for such a study. ACKNOWLEDGMENTS We wish to express our appreciation to Irene Shostak and Darlene McClure for growing the cells

273

and preparing the virus used in these experiments. Thanks are also due to Dr. V. Paetkau of this department for helpful discussions. These studies were supported by Grant MT 1191 from the Medical Research Council of Canada. REFERENCES BOEDTKER,H., and GESTELAND,R. F. (1975). Physical properties of RNA bacteriophages and their RNA. In “RNA Phages” (N. D. Zinder, ed.), pp. l-28. Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y. CASJENS, S., and KING, J. (1975). Virus assembly. Ann. Rev. Biochem. 44, 555-611. DOWNER,D. N., SUNDERLAND,S., and COLTER,J. S. (1976). Isolation and partial characterization of temperature sensitive mutants of Mengo virus. ViTology 70, 190-194. DUNKER, A. K., and RUECKERT, R. R. (1971). Fragments generated by pH dissociation of ME-virus and their relation to the structure of the virion. J. Mol. Biol. 58, 217-235. ELLEM, K. A. O., and COLTER,J. S. (1961). The isolation of three variants of Mengo virus differing in plaque morphology and hemagglutinating characteristics. Virology 15, 340-347. KAY, C. M. (1960). The partial specific volume of muscle proteins. Biochim. Biophys. Acta 38, 420-426. LAURENT, T. C., and KILLANDER, J. (1964). A theory of gel filtration and its experimental verification. J. Chromatogr. 14, 317-330. LEE, P. W. K., and COLTER, J. S. (1979). Studies of two temperature sensitive mutants of Mengo virus. Canad. J. Bioehem. 57, 902-913. LEE, P. W. K., PAUCHA, E., and COLTER,J. S. (1978). Identification and partial characterization of a new (50s) subviral particle in Mengo virus-infected L cells. Virology 85, 286-295. MAK, T. W., O’CALLAGHAN, D. J., KAY, C. M., and COLTER,J. S. (1971). Studies of the protein subunit of pa-inactivated Mengo virus variants. II. Physicochemical properties. Virology 43, 579-587. MCGREGOR,S., HALL, L., and RUECKERT, R. R. (1975). Evidence for the existence of protomers in the assembly of encephalomyocarditis virus. J. Viral. 15, 1107-1120. MCGREGOR,S., and RUECKERT,R. R. (1977). Picornaviral capsid assembly: Similarity of rhinovirus and enterovirus precursor subunits. J. Vid. 21, 548553. ~'CALLAGHAN, D. J., MAK, T. W., and COLTER,J. S. (1970). The structural proteins of Mengo virus variants. Virology 40, 572-578. PHILLIPS, B. A. (1969). In vitro assembly of polioviruses: I. Kinetics of the assembly of empty capsids and the role of extracts from infected cells. Virology 39, 811-821. PHILLIPS, B. A. (1971). In vitro assembly of poliovi-

274

LEE AND COLTER

ruses. II. Evidence for the self-assembly of 14s particles into empty capsids. Virology 44, 307-316. SCRABA,D. G., KAY, C. M., and COLTER,J. S. (1967). Physico-chemical studies of three variants of Mengo virus and their constituent rlbonucleates. J. Mol. Biol. 26, 6’7-79. SIEGEL, L. M., and MONTY, K. J. (1966). Determination of molecular weights and frictional ratios of proteins in impure systems by use of gel filtration and density gradient centrifugation. Application to crude

preparations of sulfite and hydroxylamine reductases. Biochim. Biophys. Acta 112, 346-362. Su, R. T., and TAYLOR, M. W. (1976). Morphogenesis of picornaviruses: Characterization and assembly of bovine enterovirus subviral particles. J. Gen. Vi&. 30.317-323. ZIOLA, B. R., and SCRABA, D. G. (1974). Structure of the Mengo virion. I. Polypeptide and ribonucleate components of the virus particle. ViTology 57, 531542.