J. gen. ViroL (1987), 68, 2961-2966. Printedin Great Britain
2961
Key words: human rotavirus/rearranged genome/biophysical characterization
Biophysical Characterization of Rotavirus Particles Containing Rearranged
Genomes
By M. M c l N T Y R E , 1 V. R O S E N B A U M , 2 W. R A P P O L D , 2
M. D E S S E L B E R G E R , z D. W O O D 3 AND U. D E S S E L B E R G E R 1.
~lnstitute of Virology, University of Glasgow, Church Street, Glasgow G l l 5JR, U.K.,
2Institut.[fir Physikalische Biologie, UniversitiJt Diisseldorf Universitfttsstrasse 1, 4 Dftsseldorf 1,
F.R.G. and 3North Manchester Regional Virus Laboratory, Booth Hall Children's Hospital,
Manchester M9 2AA, U.K.
(Accepted 29 July 1987)
SUMMARY
Human rotaviruses containing rearranged genomes were found to package up to
1800 additional base pairs in virus particles. The viruses compared were indistinguishable in respect of their particle diameters and their apparent S values. Particles
containing rearranged genomes were found to be of a higher density than rotavirus
particles containing a standard genome as determined by equilibrium density
ultracentrifugation. The increase in density was directly proportional to the number of
additionally packaged base pairs.
At present, five groups (A to E) of rotaviruses can be distinguished by serology and genome
analysis (Bridger et al., 1986; Pedley et al., 1986). The rotavirus genome consists of 11 segments
of dsRNA. The segments of group A rotaviruses have between 660 and approximately 3400 bp,
resulting in a total genome size of approximately 18 600 bp (Holmes, 1983 ; Rixon et al., 1984).
Recently, group A rotaviruses have been described (Pedley et al., 1984), which were obtained
from immunodeficient, chronically infected children and in which one or several of the RNA
segments were rearranged. Normal R N A segments were missing from the electrophoretic
profile or decreased in concentration, but more slowly migrating additional bands of dsRNA
were observed in the gel and found to represent concatemeric forms of normal RNA segments
(Pedley et al., 1984). Similar forms of genomic rearrangements were also observed in variants of
bovine rotavirus (BRV) isolated after serial passage at a high multiplicity of infection (Hundley
et al., 1985). Direct sequence analysis of one of the concatemeric forms of a BRV isolate
(Pocock, 1987) revealed partial duplication of a normal RNA segment (Scott et al., 1986).
Furthermore it has been found that rotaviruses possessing rearranged genomes reassort with
normal (standard) rotaviruses and that rearranged bands of genomic R N A replace normal RNA
segments structurally and functionally in reassortants (Allen & Desselberger, 1985; Graham et
al., 1987). Recently, we have been able to grow human rotavirus isolates with genome
rearrangements in vitro in BSC-1 cells. These isolates consisted of various subpopulations
differing grossly in genotype and possessing different combinations of rearranged RNA bands
(Hundley et aL, 1987). We calculated that these viruses had between 450 and 1790 additionally
packaged bp, amounting to 2.4 to 9.6~ of the standard genome size. We therefore considered
that the biophysical characteristics of such particles also might have changed. We now report
electron microscopical observations and velocity and equilibrium sedimentation data for
rotavirus particles possessing rearranged genomes and compare these to rotaviruses having a
standard genome.
BRV (UK Compton strain) was grown in MA104 cells and semipurified by differential
centrifugation as previously reported (Follett et aL, 1984). Human rotaviruses (HRV), isolated
0000-7874 © 1987 SGM
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 16:08:37
2962
Short communication
from faeces of a chronically infected immunodeficient child, were adapted to growth in BSC-1
cells. From the yields numerous subpopulations of rotaviruses possessing different forms of
genome rearrangements were obtained after several rounds of plaque-to-plaque purification
(Hundley et al., 1987). The RNA profiles of some of these viruses selected for biophysical study
are shown in Fig. l. Rearrangements of segment 11-, segment 10- and segment 8-specific
sequences into concatemeric forms are indicated. The HRV of genotype 2 has RNA segment 11
replaced by band e; HRV of genotype 3 carries band c instead of segment 11 and band a instead
of segment 8; HRV of genotype 7 carries band c instead of segment 11, band d instead of
segment 10 and band a instead of segment 8; and HRV of genotype 9 carries band e instead of
segment 11, band d instead of segment 10 and band f instead of segment 8 (Hundley et al., 1987).
As a result, rotaviruses of genotypes 2, 3, 7 and 9, respectively, have approximately 450, 1070,
1570 and 1790 additionally packaged bp. The HRVs were semipurified in the same way as
described for BRV, and all viruses were Freon-extracted according to the procedure described
by McCrae (1985). The viruses were then banded using CsC1 equilibrium ultracentrifugation
(density of CsCl-virus suspension before centrifugation, 1.36 to 1.37 g/ml; centrifugation at
150000 g for 18 h at 15 °C). Bands which consisted of single-shelled particles containing the
RNA genome (density 1.37 to 1-38 g/ml; Rodger et al., 1975; Tam et al., 1976) were collected and
diluted with phosphate-buffered saline (PBS). Virus was then pelleted (100000 g, 90 min, 4 °C)
and resuspended in PBS. A second round of CsC1 gradient centrifugation as described above
was performed with mixtures of the different single-shelled virus particles and with the
individual virus suspensions as controls. Whereas the controls yielded single bands again,
mixtures of BRV and of HRVs of genotypes 3, 7 and 9 formed visible double bands (results not
shown). The RNA of gradient fractions (120 ~tl) was extracted and analysed on polyacrylamide
gels (Follett et al., 1984); it was found that the lower band contained the HRVs with rearranged
genomes (data not shown).
Bands of single-shelled particles of HRV genotype 7 and of BRV obtained by
ultracentrifugation were isolated, negatively stained with 3 ~ phosphotungstic acid (pH 6) and
examined by electron microscopy. The diameters of 100 particles of each virus stock were
measured from the electron microscopical image using a calibrated MOP-Videoplan system for
image analysis (Kontron Bildanalyse, Munich, F.R.G.). Single-shelled particles of BRV
measured 54.8 + 5-1 nm (arithmetic mean _+S.D.) compared with 55.4 _+4.2 nm for particles of
HRV genotype 7. The small difference is not significant (t-test, P < 0.05).
It had been observed that after density gradient ultracentrifugation a mixture of HRV
genotype 3 (1070 additional bp) and of B RV could still be resolved as a narrow double band by
the naked eye. This was less distinct than the resolution obtained with the HRV 7/BRV mixture.
However, mixtures of HRV genotype 2 (450 additional bp) and BRV could no longer be resolved
visually after ultracentrifugation nor by analysis of the resulting fractions. The resolution of
double bands of HRV/BRV mixtures was increased and quantifed in terms of density
differences by analytical ultracentrifugation. The experiments were performed using a Spinco
Model E ultracentrifuge equipped with a high intensity ultraviolet illumination system,
photoelectric scanner and electronic multiplexer (Flossdorf, 1980) as well as with fluorescence
detection optics (Rappold, 1986). Velocity sedimentation centrifugation was carried out using
charcoal-filled Epon single-sector centrepieces of 3 mm optical pathlength, at a rotor speed of
10000 r.p.m, and a temperature of 22 °C. The composition of the sample solution was 0.15 MNaC1 (buffered at pH 7.2 with 0.01 M-phosphate), 6 ~ (w/w) sucrose and 10-5 M-ethidium
bromide. Sedimentation profiles were monitored with the fluorescence detection optics
recording the movement of ethidium bromide bound to virus particles. For equilibrium density
gradient centrifugation in CsC1, charcoal-filled Epon single-sector centrepieces of 12 mm
optical pathlength were used. The rotor speed was maintained at 40000 r.p.m, throughout the
experiment with the temperature at a constant 24 °C. The virus particles were suspended in
TNE buffer (100 mM-Tris-HC1 pH 7.2, 100 mM-NaC1, 10 mM-EDTA), and CsC1 was added to
give an overall density of 1.391 g/ml before centrifugation. Several scans were made in the
period between 2 and 12 h after the start of the experiment using the u.v. absorption optics at 265
nm. On the scans the difference in density between the various virus particles is known to be
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 16:08:37
Short communication
BRV
2
2963
H R V variant
of genotype
3
7
9
(a)
BRV
BRV
BRV
,~HRV :
HRV
--8R-f
--8R-a
HRV 9
+4-50
IR-c
~ 110R-d
+ 1070
+ 1790 bp
1,
Centrifugal force in density gradients
11R-e
(b)
!
0.010
!
<
H R V /
0.005
<3
10
11
I
+
450
+
+
+
1070 1570 1790
Fig. 1
I
1000
2000
Number of additional bp packaged
Fig. 2
Fig. 1. RNA profiles of BRV and of HRV with rearranged genomes of genotypes 2, 3, 7 and 9. RNAs
were extracted from semipurified virions, separated by electrophoresis on 2.8~ polyacrylamide-6 Murea slab gels and silver-stained as described (Follett et al., 1984). The segment numbers (1 to 11) are
indicated to the left. Position and origin of rearranged bands of genomic RNA are identified on the
right; bands a and f are derived from RNA segment 8, band d from segment 10, and bands c and e from
segment 11 (see text). The numbers of additionally packaged bp (indicated below the profiles) were
calculated from the relative migration rates of rearranged RNA bands and the RNA segments of BRV,
the sizes of which are known (Rixon et al., 1984).
Fig. 2. (a) Scans after analytical equilibrium centrifugation in CsC1 of mixtures of single-shelled
particles containing RNA of BRV and of HRV of genotypes 2, 3 or 9. Scans of all mixtures were made
in the same experiment after 2 h 15 min. The viruses contained in the peaks are denoted, and the
numbers of additionally packaged bp are indicated below the scans. (b) Plot of difference in density as
determined from data shown in (a) against number of additionally packaged bp.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 16:08:37
2964
Short communication
proportional to the separation in distance of the corresponding peaks. This density difference
was calculated from densities of individual bands according to the equation (Schachman, 1959):
Pr = Pr~ + [(r - re) x e) 2 × re/fl(pe)] where Pr is the density of band at radius r from the centre of
rotation; pr~ is the density of original mixture before ultracentrifugation and reached at radius re
after ultracentrifugation; re = X/[(r2m+ rb2)/2]where r m denotes the radius at the meniscus and r b
the radius at the base of the cell; ~ois the angular velocity in radians/second,/~(Pe) = 1.388 x 109
cm5/g, s~'.
Results of equilibrium centrifugation experiments are shown in Fig. 2 (a). HRV of genotypes
2, 3 and 9 differed in density from BRV by 0.0031, 0-0062 and 0.0096 g/ml, respectively. When
the difference in density was plotted against the number of additionally packaged bp, an
approximately linear relationship was obtained (Fig. 2b) for the three sample mixtures (HRV
2/BRV, HRV 3/BRV, HRV 9/BRV) analysed in the same run (Fig. 2a).
During ultracentrifugation the absolute particle density was found to increase at a constant
rate in time (2 to 12 h); this rate differed from one virus preparation to another. Consequently,
the density differences in BRV/HRV mixtures were also a function of time. This is illustrated by
two examples in Fig. 3 (a). The increase in absolute density of all particles must be due to a
physical change, possibly the influx of CsC1 or the loss of capsid protein. Therefore, in order to
obtain data about the particles in their original state at the start of the experiment, the data
shown in Fig. 3 (a) were extrapolated to time zero (Fig. 3a) and plotted against the differences in
additionally packaged bp (Fig. 3b). A straight line by definition passing through the origin
(Ap = 0, Abp = 0) was obtained. These results indicate that the earlier the recordings are taken,
the more representative they are of the viruses in their original state. The data shown in Fig. 2
(t = 2 h 15 min) give a result for the differences in density per number of additional bp which is
very close to that obtained from the extrapolated values (t = 0) in Fig. 3. Therefore the data of
Fig. 2 correctly describe the density differences for the viruses in their original state.
In velocity sedimentation experiments, no differences in the apparent S value of BRV and the
HRVs was observed; BRV and HRV of genotype 7 both sedimented with apparent S values of
390 + 8. This corresponds to an S value under standard conditions s°0,w (0~ sucrose, 20 °C) of
440 Svedberg units (results not shown).
The biophysical data reported here further characterize rotaviruses with genome rearrangements. They show that up to 1800 additional bp, found as concatemeric forms in segmentspecific, rearranged bands of genomic dsRNA, are packaged into virus particles without
evident physical constraint. The apparent particle size does not change measurably. The value
of approximately 55 nm for the diameter of single-shelled particles is in agreement with data
reported by Mc N ulty (1979). The apparent S value of 390 agrees well with that measured by Tam
et al. (1976) for single-shelled particles containing RNA of density 1.38 g/ml. As the RNA
content of rotavirus particles is approximately 15~ by weight (for reovirus particles 14-6~
RNA content was found; Gomatos & Tamm, 1963), an additional 2 to 10~ of RNA packaged
will amount to a differential RNA content in particles of 0.3 to 1.5~ changing the apparent S
value by that amount. The S.D. in the determination of the S value under the conditions used was
2~o, explaining why the subtle change was not registered.
The change in density, however, could be resolved on CsC1 gradients by equilibrium
centrifugation and could be accounted for by the extra amount of RNA packaged. Tam et al.
(1976) saw four different bands of rotaviruses in CsC1 density gradients after equilibrium
ultracentrifugation. These had specific densities of 1.28, 1.30, 1.36 and 1.38 g/ml corresponding
to double-shelled empty, single-shelled empty, double-shelled and single-shelled particles both
containing RNA, respectively, as determined by electron microscopy. Similar data were
obtained by Rodger et al. (1975). Thus, packaging of the total standard rotavirus genome of
18 600 bp increases the specific density of particles by 0.08 g/ml. By simple scaling, this predicts
density increases of 0.0019 to 0.0077 g/ml (0.024 x 0.08 to 0-096 x 0.08) for packaging of 450 to
1800 additional bp. These values are in close agreement with the experimental data obtained.
The results show that rotaviruses have considerable capacity to package additional genomic
RNA, and one wonders what the upper limit of this capacity might be. Whereas a total of 11
RNA segments/rearranged bands are invariantly packaged, there seems to be much less
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 16:08:37
2965
Short communication
I
0.010
I
I
I
I
I
I
(a)
0.010 [ (b)
'
'
/
,.0072
HRV 7
-~ 0-005
~.0048
•
I
I
200
I
"
"
I
400
Time (min)
i
I
600
/
0 0051
'n-V--R3
I
~
V ,
7y
I
1000
2000
Number of additional base pairs
packaged
Fig. 3. (a) Plot of difference in density for mixtures of BRV with HRV of genotypes 3 or 7 as a function
of time during ultracentrifugation. Differences in density are extrapolated for t = 0 as indicated by
arrows. (b) Plot of extrapolated values of density differences (t = 0; a) against number of additionally
packaged bp.
constraint on the length of individual RNA segments/bands assembled into the maturing virus
particle.
As genome rearrangements have recently also been found in rotaviruses isolated from
immunocompetent hosts, both human (Besselaar et al., 1986) and animal (calves: Pocock, 1987;
rabbit: Thouless et al., 1986; pigs: Bellinzoni et al., 1987), and also in orbiviruses (Eaton &
Gould, 1987), it is suggested that genome rearrangements are an important mechanism of the
evolution of rotaviruses and possibly of other double-stranded RNA viruses.
We thank Drs D. Riesner and J. Szilagyi and Mr N. Meyer for discussions and advice, Billy Clark and Christa
Meiering for excellent technical assistance and Fiona Conway for typing of the manuscript. We are grateful to Mr
Steve Willett, Carl Zeiss (Oberkochen) Ltd., Welwyn Garden City, U.K., for the loan of the MOP-Videoplan
system, and to Professor B. Marsden, Department of Histopathology, Royal Manchester Children's Hospital,
Pendlebury, U.K., for the use of the Zeiss EMI09. M. Mcl. was supported by a Wellcome Trust Vacation
Scholarship. Part of this work was supported by grants from the Medical Research Funds, University of Glasgow,
and the Scottish Home and Health Department to U . D , and from the Deutsche Forschungsgemeinschaft and the
Fonds der Chemischen Industrie.
REFERENCES
ALLEN,A. M. & DESSELBERGER,U. (1985). Reassortment of human rotaviruses carrying rearranged genomes with
bovine rotaviruses. Journal of General Virology 66, 2703 2714.
BELLINZONI, R. C., MATTION, N., BURRONE, O., GONZALEZ, A., LA TORRE, J. & SCODELLER, E. A. (1987). Isolation of
group A swine rotaviruses displaying atypical electropherotypes. Journal of Clinical Microbiology 25, 952-954.
BESSELAAR, T. G., ROSENBLATT,A. & KIDD, A. H. (1986). Atypical rotavirus from South African neonates. Archives of
Virology 87, 327-330.
BRIDGER, J. C., PEDLEY, S. & McCRAE, M. A. (1986). Group C rotaviruses in humans. Journal of Clinical Microbiology
23, 760-763.
EATON, B. T. & GOULD, A. R. (1987). Isolation and characterization of orbivirus genotypic variants. Virus Research 6,
363-382.
FLOSSOORF,J. (1980). Erweiterte M6glichkeiten in der analytischen Ultrazentrifugation durch Verwendung eines
neuartigen Kollimators. Makromolekulare Chemic 181, 715-724.
FOLLETT,E. A. C., SANDERS,R. C., BEARDS,G. M., HUNDLEY,F. &DESSELBERGER,U. (1984). Molecular epidemiology of
human rotaviruses. Analysis of outbreaks of acute gastroenteritis in Glasgow and the west of Scotland
1981/82 and 1982/83. Journal of Hygiene 92, 209-222.
GOMATOS,P. J. & TAMM,I. (1963). The secondary structure of reovirus RNA. Proceedingsof the National Academy of
Sciences, U.S.A. 49, 707-714.
GRAHAM, A., KUDESIA, G., ALLEN, A. M. & DESSELBERGER, U. (1987). Reassortment of human rotavirus possessing
genome rearrangements with bovine rotavirus: evidence for host cell selection. JournalofGeneral Virology68,
115-122.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 16:08:37
2966
Short communication
HOLMES,I, H. (1983). Rotaviruses. In The Reoviridae, pp. 359 423. Edited by W. K. Joklik. New York and London:
Plenum Press.
HUNDLEY, F., BIRYAHWAHO,B., GOW, M. & DESSELBERGER,U. (1985). G e n o m e rearrangements of bovine rotavirus
after serial passage at high multiplicity of infection. Virology 143, 88-103.
HUNDLEY, F., MclNTYRE, M., CLARK, B., BEARDS, G., WOOD, D., CHRYSTIE, I. & DESSELBERGER, U. (1987).
Heterogeneity of genome rearrangements in rotaviruses isolated from a chronically infected immunodeficient
child. Journal of Virology (in press).
McCRAE, ~. A. (1985). Double-stranded R N A viruses. In Virology - A Practical Approach, pp. 151-168. Edited by
B. W. J. Malay. Oxford and Washington, D.C.: I R L Press.
McNULTY, M. S. (1979). Morphology and chemical composition of rotaviruses. In Viral Enteritis in Humans and
Animals, pp. 111-140. Edited by F. Bricout & R. Scherrer. Paris: Editions I N S E R M .
PEDLEY, S., HUNDLEY, t:., CHRYSTIE, I., McCRAE, M. A. & DESSELBERGER, U. (1984). The genomes of rotaviruses
isolated from chronically infected immunodeficient children. Journal of General Virology 65, 1141-1150.
PEDLEY, S., BRIDGER, J. C., CHASEY,D. & Mc-CRAE,M. A. (1986). Definition of two new groups of atypical rotaviruses.
Journal of General Virology 67, 131-137.
POCOCK, D. H. (1987). Isolation and characterization of two group A rotaviruses with unusual genome profiles.
Journal of General Virology 68, 653-660.
RAPPOLD, W. (1986). Entwicklung einer Fluoreszenzoptik far die analytische Ultrazentrifuge und ihre Anwendung auf
biophysikalische Fragestellungen. Ph.D. thesis, University of Diisseldorf.
RIXON, F., TAYLOR, P. & DESSELBERGER,U. (1984). Rotavirus R N A segments sized by electron microscopy. Journal
of General Virology 65, 233-239.
RODGER, S. M., SCHNAGL, R. D. & HOLMES, I. H. (1975). Biochemical and biophysical characteristics of diarrhea
viruses of h u m a n and calf origin. Journal of Virology 16, 1229-1235.
SCttACHMAN, H. K. (1959). Ultracentrifugation in Biochemistry. New York & London: Academic Press.
SCOTT, G. S., McCRAE, M. A. & POCOCK, D. (1986). Molecular analysis of a bovine rotavirus with a gross genomic
abnormality. Abstracts, International Symposium on Double-Stranded RNA Viruses, Ox3'brd.
TAM,J. S., SZYMANSKI,M. T., MIDDLETON,P. J. & PETRIC, M. (1976). Studies on the particles of infantile gastroenteritis
virus (orbivirus group), lntervirology 7, 181-191.
THOULESS, M. E., DIGIACOMO,R. F. & NEUMAN, D. S. (1986). Isolation of two lapine rotaviruses: characterization of
their subgroup, serotype and R N A electropherotypes. Archives of Virology 89, 161-170.
(Received 27 M a y 1987)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 16:08:37