Structuur en eigenschappen van cowpea-mozaiekvirus

Purified infectious preparations of cowpea mosaic virus (CPMV) consist of three centrifugal components with sedimentation coefficients of 58, 95 and 118 S. These are referred to as top (T), middle (M) and bottom (B) component. M and B are nucleoproteins containing 25 and 36% RNA respectively. All three components are isometric particles with a diameter of 20 nm and have identical capsids. Both nucleoproteins (M and B) are necessary for infectivity (Van Kamen, 1968) and some direct evidence for the distribution of the genetic properties has been given by Bruening (1969) and De Jager and Van Kammen (1970).

The molecular weights of the RNAs have been determined by three emperical methods which differ in the remaining secondary structure of the RNAs. From the sedimentation velocities after formylation of the RNAs (5% secondary structure remaining; Boedtker, 1968) values of 1.4 x 10 ⁶ D (M-RNA) and 2.1 x 10 ⁶ D (B-RNA) could be calculated. From the mobilities of the CPMV-RNAs compared to the mobilities of the Escherichia coli ribosomal RNAs upon electrophoresis in polyacrylamide gels under non denaturing conditions molecular weights of 1.55 x 10 ⁶D (M-RNA) and 2.55 x 10 ⁶D (B-RNA) could be calculated. If the molecular weights were determined by electrophoresis on polyacrylamide gels under denaturing conditions (8 M urea, 60°C) values of 1.37 x 10 ⁶D (M-RNA) and 2.02 x 10 ⁶D (B-RNA) were found. In 8 M urea at 60°C the secondary structure of the Escherichia coli ribosomal RNAs is completely eliminated (Reynders et al., 1973). The CPMV-RNAs are also completely denatured under these conditions as has been shown by hyperchromicity measurements (Fig. 3.3.). As the values determined by gel electrophoresis under denaturing conditions are not affected by differences in secondary structure between the marker-RNAs and the CPMV-RNAs, these molecular weights are the most reliable.

The capsid of CPMV is constructed from two different proteins. CPMV was degraded with 1% sodium dodecyl sulfate in the presence of 1% 2-mercaptoethanol. To prevent aggregation of the proteins it was necessary to protect the SH- groups of the protein with mercaptoethanol or to block these groups by carboxymethylation or performic acid oxidation. The molecular weights of the proteins were estimated by SDS-polyacrylamide gelelectrophoresis according to Weber and Osborn (1969). Molecular weights of 44,000 D (I) and 25,000 D (II) were found for the carboxymethylated proteins and 49,000 D (I) and 27,500 D (II) for the performic acid oxidized proteins. The SDS bound to the protein apparently does not eliminate differences in the conformation of the proteins. The two proteins are found in all three centrifugal components in equimolar ratio, indicating an identical capsid of T, M and B.

The particle weights of the centrifugal components have been determined by light-scattering and sedimentation-equilibrium centrifugation. By light-scattering particle weights of 4.5 x 10 ⁶D, 5.6 x 10 ⁶D and 6.1 x 10 ⁶D, and by sedimentation-equilibrium particle weights of 3.80 x 10 ⁶D, 5.15 x 10 ⁶D and 5.87 x 10 ⁶D were found for T, M and B respectively. The differences between the particle weights determined by light-scattering and by sedimentation-equilibrium are within the experimental error (± 5%) for M and B. The difference is larger for T. This is probably caused by aggregation or contamination with large particles. This will affect the particle weight obtained by light-scattering, but not the particle weight obtained by sedimentation-equilibrium if the contaminant or the aggregates precipitated. Although the differences between the values obtained by the two methods are not very large for M and B, the values of sedimentationequilibrium are probably the best, since the agreement of these particle weights with the molecular weights of the RNAs and the RNA- content of M and B is very good (M : 25 % RNA; Mw-RNA : 1 .37 x 10 ⁶D and B : 36 % RNA ; Mw-RNA : 2.02 x 10 ⁶D ). It can not be excluded, that some aggregation or contamination with large particles affects also the particle weights obtained by light-scattering of M and B, as all light-scattering values are consistently higher than the sedimentation-equilibrium values.

When purified preparations of CPMV are analysed by electrophoresis, two electrophoretic forms are seen : a slow (S) moving and a fast (F) moving electrophoretic form. Each of the electrophoretic forms consists of all three centrifugal components. S and F can be separated by electrophoresis in a sucrose gradient. If the proteins of a CPMV preparation consisting of S and F, are analysed by electrophoresis on polyacrylamide gels, three proteins are seen: one large protein (I) and two small proteins (II and III) (Fig. 4.2.). Protein II is specific for S, and protein III for F. Molecular weights of 25,000 D for protein II and 22,000 for protein III can be estimated by SDS-polyacrylamide gelelectrophoresis of the carboxymethylated proteins. The distance migrated by protein I of S (I _S ) and the distance migrated by protein I of F (I _F ) is the same. To check if there are differences between these two large proteins, the proteins of the separated electrophoretic forms were separated by gel filtration on Sephadex G 200 with 5 M urea as eluent, and the amino acid compositions determined (Table 4.3.). From this table it can be seen that there are not only differences between protein I and protein II and III, but that also protein I of S (I _S ) and protein I of F (I _F ) are clearly distinct. Both proteins of an eletrophoretic form are therefore characteristic for that form.

By incubation with proteolytic enzymes it is possible to increase the electrophoretic mobility of S. This increase in mobility of S is correlated with an increase in mobility of protein II (Fig. 5.1. and Fig. 5.2.). The differences between the molecular weights of protein II and III indicate a larger difference between these two proteins than 7 amino acids. When, as can be seen from Table 4.3., moreover protein I changes upon the conversion of S into F, it is clear, that the conversion of S into F is much more complicated than was supposed by Niblett and Semancik (1969). The properties of CM and the relationship between the centrifugal components, the electrophoretic forms and the proteins are summarized in Figure 10.1.

No differences can be seen in the electron microscope between the particles of S and F. This can be expected, as the conversion of S into F apparently does not lead to a reorganisation of the protein coat. A difference in molecular weight of approximately 2,500 D (estimated by SDS-polyacrylamide gelelectrophoresis) between protein II and III is not detectable in the electron microscope.

By three-dimensional image reconstruction of electron micrographs it was shown that the capsid of CPMV possesses 5, 3, 2 symmetry. Based on the properties of the proteins and the reconstructed image a model is proposed consisting of twelve pentamers of the large protein at the 5- fold positions and twenty trimers of the small protein at the 3-fold positions (Fig. 7.3., Crowther et al., in press). The geometrical figure traced out by the ridges of proteins joining 5-fold and 3-fold positions is approximately a rhombic triacontahedron.

Niblett and Semancik (1969, 1970) showed that S was less infectious than F, even though the RNA of S was more infectious and less degraded than the RNA of F. This is only true, if old preparations are tested. The infectivities of S and F are equal, provided fresh preparations are used. The infectivities of the RNAs of S and F are equal, and there is no appreciable sign of a difference in degradation, as judged by gel electrophoresis (Fig. 6.1.). After in vitro aging of S a loss in infectivity is observed; the initial infectivity can be restored by incubation with trypsin or chymotrypsin (Table 6.3.). Upon aging the mobility of S is not changed, but the mobility increases upon incubation with trypsin or chymotrypsin. Aging in vitro and incubation with trypsin or chymotrypsin affect neither the infectivity nor the mobility of F. Since no increase in mobility of S was observed upon aging, the charge is not decisive for the difference in infectivity between S and F, as has been suggested by Niblett and Semancik (1970). But it is likely that the amino acids which determine the difference between S and F, do play an important role in the loss of infectivity of S upon aging. Removal of these amino acids by incubation with proteolytic enzymes results in an restoration of the initial infectivity.

During storage partial conversion of S into F is sometimes observed. A fast conversion is seen with virus purified by the PEG/NaCl method from fresh leaves and essentially no conversion with virus purified by the butanol-chloroform method from frozen tissue (compare Fig. 5.1. and Fig. 5.3.). Essentially no increase in the mobility of S is observed also with preparations which have undergone an extra purification step, e.g. separation of the electrophoretic forms or separation of the centrifugal components. This partial conversion probably results from proteolytic enzymes contaminating the preparations, as incubation of CM in the soluble protein fraction or in the homogenate of infected Vigna leaves leads to the same conversion (Fig. 5.4.).

If the centrifugal components are centrifuged to equilibrium in CsCl gradients at neutral pH, eight components are found. These components are S and F of top component, S and F of middle component, S and F of a bottom component with a low buoyant density (B _u ) and S and F of a bottom component with a high buoyant density (B _l ). The difference in buoyant density between S and F of a centrifugal component (ca. 0.004 g/cm ³) is probably caused by a difference in the amount of CS ⁺-ions bound to S and F, as the same difference in buoyant density is also found between S and F of top component (which consists of only protein). The buoyant density of all the components is dependent upon the pH. The density increases with increasing pH. The difference in density between pH 6.5 and pH 8 5 is approximately 0,005 g/cm ³. This density-shift is reversible and probably the result of an increased Cs ⁺-ions binding with increasing pH.

Besides these small differences in buoyant density observed at different pH's and between the electrophoretic forms, there is a large difference in buoyant density between the two bottom components (about 0,040 g/cm ³). It is not clear what causes this difference. B _u and B _l have the same RNA-content, the same protein composition, the same sedimentation coefficient and the same contribution to the infectivity. When bottom component is centrifuged in CsCl solutions at pH 8.5 almost only B _u is found and at pH 6.5 almost only B _u (Fig. 8.3.). By increasing the pH from 6.5 to 8.5 or by increasing the length of the run at pH 6.5 (Fig. 8.4.) B _u can be converted in B _l .

The conversion from B _u into B _l is irreversible. A possible explanation for this phenomenon is as follows: the particles swell due to the high ionic strenght. The swelling causes an irreversible change in the conformation of the particles, resulting in an increased binding of Cs ⁺- ions (B _u changes to B _l by increasing the length of the run at pH 6.5)Increasing the pH from 6.5 to 8.5 promotes the swelling. The process of the swelling to B _u and the conversion of B _u to B _l is then so greatly accelerated that after about 20 hours centrifugation (the standard time in these experiments) almost no B _u is left.

Parts of this thesis have been published earlier (Geelen et al., 1972; Geelen et al., 1973; Crowther et al., in press).