J. gen. Virol. (I973), x8, I8I-I87
Printed in Great Britain
18I
The Structure of Narcissus Mosaic Virus
By H. R. W I L S O N , P. T O L L I N
AND A. R A H M A N
Carnegie Laboratory of Physics, University of Dundee, Dundee, Scotland
(Accepted 3 October 1972 )
SUMMARY
Analysis of the X-ray diffraction patterns from oriented specimens of narcissus
mosaic virus is described. This shows that there is a marked feature in the virus
particle at a radial position of about 33 A, and it is suggested that this corresponds to the sugar-phosphate backbone of the RNA. Electron microscope studies
of stained transverse sections of the virus particles are not incompatible with this
interpretation. Some aspects of the possible RNA conformation in the virus particle
are discussed.
INTRODUCTION
Narcissus mosaic virus (NMV) is an elongated flexible particle about 55oo A long and
about I3O A in diameter (Tollin et al. I967). X-ray diffraction patterns from oriented
specimens of NMV can be interpreted in terms of a helical arrangement of protein subunits,
with an integral, or near integral number, 5q + 4, in five terms of the helix (Tollin, Wilson &
Young, 1968). The pitch of the helix is 33 A in dry specimens and 36 A at 98 ~ relative
humidity. Here we describe further studies of NMV which suggest the possible location of
the nucleic acid in the virus.
METHODS
X-ray diffraction. Oriented specimens of NMV were prepared by the coverslip method
(Bernal & Fankuchen, 1941; Tollin et al. 1968 ). X-ray diffraction patterns were obtained
with pin-hole cameras of the type described by Langridge et al. (I96O), using nickel-filtered
Cu Kc~ radiation. Humidity was controlled with saturated salt solutions. The specimen-tofilm distance was about 58 mm. Intensity measurements were made with a Joyce-Loebl
recording microdensitometer.
Electron microscopy. The oriented dry specimens used for X-ray diffraction are suitable
for preparing ultrathin sections for electron microscopy. Pieces of the oriented specimens,
about o'5 mm long, were fixed in glutaraldehyde and postfixed with osmium tetroxide. The
material was dehydrated in alcohol, stained with uranyl acetate during dehydration,
embedded and sectioned. Sections were poststained with lead citrate before examination in
the electron microscope.
RESULTS
X-ray diffraction
The earlier X-ray studies (Tollin et al. 1968) showed that there were 5q + 4 protein subunits
in five turns of the helix, but the difficulty of distinguishing between a truly meridional and
a near-meridional reflexion made it impossible to determine the value ofq direction from the
diffraction patterns, although they indicated that 6 ~< q ~< 9. Because the reflexion on the
44th layer-line appeared shorter than the other apparently meridional reflexions, Tollin
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 01:37:48
I82
H . R . W I L S O N ~ P. T O L L I N A N D A. R A H M A N
36
0
50
100
150
r(A)
Fig. I. A cylindrical Patterson synthesis for NMV using data to a resolution of about 8/~ and a
large artificial temperature factor in order to eliminate effects due to the abrupt termination of
data. The dashed line which starts at the origin shows where features would be expected from a
helical arrangement of scattering matter at a radius of 33 ~.
20
19
18
17
16
15
14
I
i
12
ll
,° t
9
8
7
5
43 ~ ~ ~ ~ "
2
1
I
0
I
0-04
R(A 1)
I
0-08
Fig. 2. The squared Fourier transform, cylindrically averaged, of a helical arrangement of spherical
blobs of radius 2"5/~, with centres at a radial position of 33 A. Observed intensity is indicated by
vertical lines.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 01:37:48
The structure o f N M V
183
34
20
15
10
5
0
Layer-line
Fig. 3- X-ray diffraction pattern from an oriented specimen of N M V (at 57 ~ r.h.).
et al. 0968) made the tentative suggestion that the value of q was 8. At that time no information was available about the mol. wt. of the protein subunit. Recently, however, M. Mayo
(personal communication) has determined the subunit mol. wt. using the technique of
polyacrylamide gel electrophoresis, and obtains a value of 3oooo (+ Io ~). Accepting this
value, and using an argument similar to that used by Wilson & Tollin (I969) for potato
virus X (PVX), it is possible to estimate the number of protein subunits/turn of helix in
NMV as 6.8 _+o'7. Since the X-ray diffraction patterns show that there is an integral, or near
integral, number in five turns of the helix, this suggests that the number of protein subunits/
turn is 6.8. This corresponds to a q value of 6, which is within the range suggested by the
X-ray diffraction pattern.
Further analysis of the NMV diffraction patterns can be made by calculating the cylindrical Patterson synthesis (MacGillavry & Bruins, ~948) and by using helical diffraction
theory (Cochran, Crick & Vand, I952 ). The cylindrical Patterson function, calculated using
data to a resolution of about 8 A is shown in Fig. I. A large artificial temperature factor
was applied in order to eliminate effects due to the abrupt termination of data. The main
feature of the Patterson synthesis is a positive region at a position corresponding to a helical
arrangement of scattering matter at a radius of about 33 A. It is reasonable to assume that
this marked feature at a radial position of about 33 A. is the sugar-phosphate backbone of
13
VIR
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 01:37:48
18
I84
H . R . W I L S O N , P. T O L L I N A N D A. R A H M A N
Fig. 4. Electron micrograph of a transverse section of a dried specimen of narcissus mosaic virus.
Micrograph by Mr I. M. Roberts.
the RNA because this has a higher electron density than the protein. This position is close to
the suggested radial position of the RNA in PVX, and from the similarities in the structures
of NMV and PVX (Tollin et al. 1967) it might be expected that the RNA would be similarly
situated in the two viruses.
An alternative interpretation of the diffraction data can be made using the helical diffraction theory. The observed diffraction reflexions in the centre of the NMV pattern (at 57 %
r.h.) fall under the maxima of the transform of a helical arrangement of scattering matter
situated at a radius of about 33 A (Fig. z). This again suggests that there is a marked feature
in the virus at this radial position. The sugar-phosphates of an R N A molecule at low resolution approximates to a continuous helix of scattering matter and this enhances layer-lines
at the centre of the pattern corresponding to the pitch of the helix. The fifth and tenth layerlines are particularly strong compared with adjacent layer-lines as would be expected from
this enhancement (Fig. 3).
Electron microscopy
Electron microscopy of stained transverse sections of tobacco rattle virus (TRV) showed
two dark rings, one of which could possibly correspond to the R N A (Tollin & Wilson, I97I ).
In the hope that similar studies of NMV would give support to the X-ray results, stained
sections were examined in the electron microscope. Fig. 4 shows a stained section cut perpendicular to the virus axis and shows the hexagonal close-packed arrangement of virus
particles. Some cracking of the specimen due to severe drying has occurred and this shows
up as clear regions between the virus domains. To demonstrate more clearly the hexagonal
packing of the particles, Fig. 5 shows a portion of Fig. 4 which has been rotated on itself
six times at angular intervals of 6o °. From these photographs it is possible to measure the
interparticle distance as Io6 A (+ Io A) which is in good agreement with the value of ~o6 A
( -+4 A) for the interparticle distance estimated from the X-ray diffraction pattern from dry
specimens (Tollin et aL I967).
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 01:37:48
The structure of N M V
I8 5
Fig. 5- A n e n l a r g e m e n t o f a p o r t i o n of Fig. 4 which h a s been rotated
o n itself six times at a n g u l a r intervals o f 6o ° .
The main feature of the stained particles are an inner region of diameter 20 to 30 A which is
relatively free from stain, a circular region, of mean diameter ~ 50 A which is heavily stained,
and an outside region with again little staining. This pattern of stain can be interpreted as
being due to positive staining of some features of the virus particles. The particles have an
outside diameter of ~ ~30 A and therefore interlock in the dry state, hence the outer lightly
stained region must represent coat protein. It is interesting that the interparticle distance of
IO6 A is about as close as interlocking helices of diameter 13o A_ can come without having
two helices trying to share the same gap in a third. The lightly stained region in the centre
may represent a hole down the particle but might equally well correspond to protein.
One possible interpretation of the darkly stained ring is that this represents the mean
diameter of the nucleic acid. However, this would not agree with the conclusions from the
X-ray diffraction studies. An alternative interpretation is that both the R N A and part of
the protein may be staining and that the observed ring is due to a combination of these. This
interpretation is suggested by the fact that similar sections of TRV show two dark rings, only
one of which corresponds to the probable R N A position. TRV is a much larger particle
than NMV, with an outer diameter of about 250 A and the dark rings have diameters of
I6O A and ~ 82 A. The outer ring corresponds to the possible R N A position. It may be
that the NMV sections have stained in a similar way, but because of the smaller size of the
virus particles the two rings are not resolved in this case. If this were so then the outer part
of the stained ring in NMV could correspond to the RNA and this would be compatible
with the X-ray diffraction results. However, it cannot be claimed that the electron microscope results support the X-ray results, but merely that they are not incompatible with them.
13-2
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 01:37:48
I86
It. R. W I L S O N ,
P. T O L L I N
A N D A. R A H M A N
DISCUSSION
We can consider whether the radial position of 33 A for the RNA sugar-phosphate is a
reasonable value. If we assume that the phosphate-phosphate separation along the R N A
has the same value of 5"I 5 A in NMV as in TMV, we can estimate the number of nucleotides/
turn of the helix. This gives a value of 4o'8 which corresponds to a 6 ~ R N A content, and,
if there are 6.8 protein subunits/turn of the helix, to six nucleotides associated with each
subunit. A 6 ~ RNA content is similar to that found in most of the PVX group of viruses,
and corresponds to a mol. wt. of 2-2 × io 6. This value compares reasonably well with the
value of 2"5 × IO8 estimated by M. Mayo (personal communication) using the technique of
acrylamide gel electrophoresis.
Only in the case of TMV is there direct evidence about the location of the RNA (Franklin,
I956; Franklin & Holmes, I958), from which information, combined with the value of the
pitch of the helix and the RNA content, the mean value of the phosphate-phosphate distance
along the RNA can be calculated. In the case of TRV, PVX and NMV the evidence concerning the possible radial position of the R N A is indirect, and hence conclusions based on
this are tentative. Nevertheless, it is of interest that the results seem to be consistent with a
similar mean phosphate-phosphate distance in all the viruses, although the number of
nucleotides associated with a protein subunit is not the same in the different viruses. The
simplest interpretation of such a result is that the phosphate-phosphate distance is constant
along each RNA molecule, and that the nucleotide conformation is similar in the different
virus RNAs. We can consider whether this is stereochemically possible.
From symmetry considerations it is to be expected that there is an integral number of
nucleotides associated with each protein subunit, and, apart from differences in the bases,
these groups of nucleotides will be equivalent. Within a group, however, there can be
tertiary structure; that is, all the nucleotides need not be equivalent. This is the most general
kind of model for the virus RNA, and in such a model the tilt of the bases relative to the
virus axis need not be the same for each nucleotide, nor the phosphate-phosphate distance
between neighbouring nucleotides. However, if all the nucleotides were equivalent, then the
R N A would have secondary structure only, as in this case the tilt of the bases relative to the
virus axis would be the same for all nucleotides and the phosphate-phosphate distance along
the chain would be constant. This is the simplest possible structure for the virus RNA.
Single crystal studies of nucleosides and nucleotides have shown that the preferred rotations about bonds are limited to relatively small ranges, and the values for double-helical D N A
and R N A models also lie within these ranges (Arnott & Hukins, 1969; Sundaralingam, I969).
It is reasonable to assume that the same preferred conformations will also occur in
the virus RNAs, so that the single-crystal results can be used as constraints in constructing
possible molecular models of the RNA. For the general kind of virus RNA model mentioned
earlier, with tertiary structure within the group of nucleotides associated with a protein
subunit,rthe single-crystal data are not restrictive enough to decide between the many possible
models. However, in the case of the simpler RNA model, the equivalence of nucleotides
imposes added constraints, and we can consider whether such models are stereochemicaIly
acceptable. In such R N A models the bases would be expected to lie about 3"4 A apart,
and because the rise/nucleotide parallel to the virus axis is much less than this, the bases
would have to be tilted relative to the virus axis. The phosphate-phosphate distance of
5"15 A is less than that in double-helical D N A and R N A models - in B-DNA the phosphatephosphate distance is 6"5 A, in A-DNA it is 5"5 A, and in double-helical R N A it is 5"6 A.
This presents a problem if the same orientation of the C (5')-O (5') bond relative to the sugar
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 01:37:48
The s t r u c t u r e o f N M V
187
ring is retained in the virus R N A model as in the D N A a n d double-helical R N A models.
This orientation, referred to as gauche-gauche, is also the orientation observed in all 5'nucleotides which have so far been studied. However, in nucleosides, other orientations of
the C ( 5 ' ) - O ( 5 ' ) b o n d relative to the sugar ring are observed, a n d if these are allowed in the
virus R N A model the p h o s p h a t e - p h o s p h a t e distance c a n be made acceptable. O u r preliminary model building shows that models with a similar nucleotide c o n f o r m a t i o n can be
constructed to satisfy the helical parameters of TRV, TMV, PVX a n d N M V , although there
are certain features of the models which are n o t completely satisfactory. A more systematic
study of such models may enable a more critical assessment of their acceptability to be
made, b u t it should be emphasized that even if models of this k i n d are completely satisfactory they would n o t necessarily be unique. Other, more complex models, m a y also be
equally satisfactory.
As we stated earlier, the evidence c o n c e r n i n g the possible location of the R N A in TRV,
P V X a n d N M V is indirect. The discussion of possible R N A c o n f o r m a t i o n s will be on firmer
g r o u n d when more direct i n f o r m a t i o n is available a b o u t the R N A location a n d also when
more is k n o w n a b o u t the preferred c o n f o r m a t i o n s of nucleotides.
We t h a n k M r W. P. M o w a t for supplying the virus a n d M r I. M. Roberts for preparing
the sectioned specimens a n d for much of the electron microscopy. H. R. W i l s o n a n d A.
R a h m a n wish to acknowledge the award of a Medical Research Council grant.
REFERENCES
ARNOTT,S. & HUKINS,D. W. L. (1969). Conservation of conformation in mono- and poly-nucleotides. Nature,
London 224, 886-888.
BERNAL,J. D. & FANKUCHEN,L 094I). X-ray and crystallographic studies of plant virus preparations. Journal
of General Physiology 25, I I I-165.
COCHRAN,W., CRICK,F. H. C. & rAND, V. (I952). The structure of synthetic polypeptides. I. The transform of
atoms on a helix. Acta erystallographica 5, 58t-586.
fRANKLIN,R. r. (1956). Location of the ribonucleic acid in the tobacco mosaic virus particle. Nature, London
x77, 928-930.
FRANKLIN,R. E. & HOLMES,K. C. ( I 9 5 8 ) . Tobacco mosaic virus; application of the method of isomorphous
replacement to the determination of the helical parameters and radial density distribution. Acta crystallographica i i , 213-220.
LANGRIDGE, R., WILSON, H. R., HOOPER, C. W., W1LK1NS, M. H. F. & HAMILTON, L. O. (196o). The molecular
configuration of deoxyribonucleic acid. I. X-ray diffraction study of a crystalline form of the lithium salt.
Journal of Molecular Biology 2, 19-37.
MACGILLAVRY,C. H. & BRUINS,E. M. (I948). On the Patterson transforms of fibre diagrams. Acta crystallographica x, I56-I58.
SUNDARALINGAM,M. (1969). Stereochemistry of nucleic acid constituents. IV. Allowed and preferred conformations ofnucleosides, nucleosidemono-,di-, tri-, tetraphosphates, nucleic acids and polynucleotides.
Biopolymers 7, 821-86o.
TOLLIN, P. & WILSON,n. R. 0971). Some observations on the structure of the CAMPINASstrain of tobacco
rattle virus. Journal of General Virology 13, 433-44o.
TOLLIN, P., WILSON, n . R . & YOUNG, O . W . 0968). X-ray diffraction evidence of the helical structure of
narcissus mosaic virus. Journal of Molecular Biology 34, I89-I92.
TOLLIN, P., WILSON, H. R., YOUNG, D. W., CATHRO, J. & MOWAT, W. P. 0967). X-ray diffraction and electron
microscope studies of narcissus mosaic virus and comparison with potato virus X. Journal of Molecular
Biology 26, 353-355.
WILSON,H. R. & TOLLIN,P. (I969). Some observations on the structure of potato virus X. Journal of General
Virology 5, I5I-I54.
(Received 2I August I972 )
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 01:37:48