Journal of General Virology (1994), 75, 2133-2138. Printed in Great Britain
2133
Selection versus recombination: what is maintaining identity in the
3' termini of blueberry leaf mottle nepovirus RNA1 and RNA2?
J. W. B a c h e r , l * t D. Warkentin, 1,2 D. Ramsdell 2 and J. F. H a n c o c k I
1 Department o f Horticulture and ~Department o f Botany and Plant Pathology, Michigan State University,
East Lansing, Michigan 48824, U.S.A.
The 3' non-coding regions (NCR) of RNA1 and RNA2
of blueberry leaf mottle nepovirus (BBLMV) are nearly
identical with differences occurring at only four
positions. The presence of this 1-4kb duplication
indicates that recombination has occurred at least once
in the evolutionary history of BBLMV. Since high
mutation rates are common in RNA viruses, strong
selection pressure and/or high frequency of recombination must be operating in order to maintain identity
in this duplicated region. The possible involvement of
high frequency RNA recombination in maintaining
identity was investigated. The four conserved differences
between the 3' NCR of RNA1 and RNA2 were used as
markers to detect recombinants in a viral population.
Nucleotide sequences of BBLMV cDNA clones were
compared to the 3' consensus sequence and deviations
were examined to determine whether they were due to
single base mutations or recombinational events. No
evidence of recombination was found in any of the
eDNA clones sequenced and all differences were
attributed to mutations. If recombination occurred in
the 3' NCR of BBLMV, the frequency was below 1.1%
between markers. The data indicate that identity in the
3' NCR of RNA1 and RNA2 of BBLMV was maintained
without high levels of recombination. The high number
of mutations observed in a BBLMV population and lack
of observable recombination indicate that other
mechanisms, such as selection, play an important role in
the conservation of identity in the 3' NCR.
The 3' non-coding regions (NCRs) of blueberry leaf
mottle nepovirus (BBLMV) RNA1 and RNA2 are
nearly identical (Bacher et al., 1994). The duplicated
region consists of 1390 nucleotides, with differences
occurring at only four positions. How this duplication
originated and how it is maintained is not known.
However, conservation of the 3' termini of genomic
RNAs is not unique to BBLMV. Highly conserved 3'
termini have also been found in all the other nepoviruses
sequenced (Greif et al., 1988; Brault et al., 1989; Meyer
et al., 1986; Serghini et al., 1990; Bertioli et al., 1991;
Rott et al., 1991a, b; Scott et al., 1992). In fact,
conservation of the 3' termini is a common feature of
most multipartite RNA viruses (Matthews, 1991).
High mutation rates in RNA viruses (10 -3 to 10 ~ per
nucleotide per round of replication) means strong
selection pressure must constantly counteract these
changes if a specific sequence is to be conserved
(Steinhauer & Holland, 1987; Domingo & Holland,
1988). Recombination might be working in conjunction
with selection to create and maintain specific viral
sequences (Angenent et al., 1989). For example, Rott et
al. (1991a, b) hypothesized that high frequency recombination between the two genomic RNA components of
tomato ringspot nepovirus (TomRSV) was responsible
for conservation of a 1533 nucleotide region of identity
between the 3' termini of RNA1 and RNA2. The 3'terminal region of only one of the genomic RNAs might
serve as a template for both 3' NCRs, thereby maintaining identity. Scott et al. (1992) found a 1"5 kb region
of sequence identity in the T-termini of RNA1 and
RNA2 of cherry leaf roll nepovirus. The authors
questioned why this duplication is retained when mutation rates are so high in RNA viruses and suggested
that the duplicated sequence is the site for high levels of
recombination between RNA1 and RNA2.
What role recombination plays in the conservation of
the 3' terminus of BBLMV is unknown, but the presence
of a duplicated region indicates that recombination may
have occurred at least once in the evolution of this virus.
In this study, we searched for evidence of recombination
between RNA1 and RNA2 as a means of determining
the importance of recombination in maintenance of the
duplicated 3' region.
"~Present address: Universityof Wisconsin, Department of Horticulture, 1575 LindenDrive, Madison,Wisconsin53706, U.S.A.
0001-2090 © 1994 SGM
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:39:41
2134
Short communication
To determine whether recombination was occurring in
BBLMV, we examined cDNA clones at the positions
where nucleotide differences were found between the 3'
NCRs of RNA1 and RNA2 (Bacher et al., 1994). If
recombination had occurred during the synthesis of a
particular viral RNA molecule, then a recombinant
clone would contain sequences unique to RNA1 on one
side of the crossover point, and sequences unique to
RNA2 on the other side. The ability to use these
differences between RNA1 and RNA2 as markers for
identifying recombinants depended upon their conservation and uniqueness to either RNA1 or RNA2. A
series of 3' cDNA clones were sequenced around the
marker sites. Consensus sequences for both RNA1 and
RNA2 were developed and deviations from these
sequences were examined to determine whether they
were due to single base mutations or recombinational
events.
The BBLMV population used to generate the c D N A
library was probably a heterogeneous mixture of genotypes, since the virus was maintained by repeated passage
in Chenopodium quinoa plants over a period of many
years. Virion purification was as described by Ramsdell
& Stace-Smith (1981), using pooled leaf tissue from 200
BBLMV-infected C. quinoa plants. The initial source of
BBLMV inoculum was purified from infected shoot
termini of Vaccinium corymbosum L. cv. Rubel collected
in southern Michigan.
Total virus RNA was extracted from purified virus
particles according to Ramsdell & Stace-Smith (1981)
and fractionated from an ultracentrifuged sucrose density gradient made with RNase-free sucrose in 1 x SSC
buffer pH 7.0, containing 6 gg/ml purified bentonite
(Fraenkel-Conrat et al., 1961). Complete separation of
the two genomic RNAs was not possible because of the
similarity in their Mr values. Therefore, a mixture
containing both RNA1 and RNA2 molecules was used
as templates for oligo(dT)-primed cDNA synthesis
(Gubler & Hoffman, 1983). Size-fractionated cDNA was
blunt-end ligated into a SmaI-cut pBluescript K S +
vector, then used to transform XL1-Blue Escherichia coli
cells (Sambrook et al., 1991).
One-hundred and sixty cDNA clones were sized and
those longer than 1 kb (about 30%) were selected for
sequencing by the dideoxynucleotide chain termination
method (Sanger, 1981) using single-stranded template~
prepared according to Vieira & Mesing (1987). In order
to avoid sequencing entire clones, subclones were made
by restriction enzyme digestion to remove intervening
sequences that were not informative (i.e., the same on
both RNAs). Sequence data were analysed using the
Genetics Computer Group sequence analysis software
package, version 7.1 (Devereux et al., 1984).
The marker genotypes of 45 cDNA clones were
determined and aligned with the previously identified
sequences from the 3' regions of RNA1 and RNA2 of
BBLMV (Bacher et al., 1994) (Fig. 1). Consensus
sequences were determined for 3' NCRs of both RNAs.
Three of the seven nucleotide differences found between
the original RNA1 (from clone 24) and RNA2 sequences
(from clone 34) were not present in any of the other
clones examined (i.e., positions 2, 3 and 4). However, the
remaining four differences (positions 1, 5, 6 and 7) did
consistently distinguish between RNA1 and RNA2.
Positive identification of RNA1 and RNA2 was possible
in clones longer than 1.4 kb, since they contained coding
regions of known sequence (Bacher et al., 1994). These
coding regions were used as an additional source of
markers to identify potential recombinant clones. In
order to confirm that each clone sequenced was unique,
clones of differing sizes were chosen and the sequence at
the junctions between cDNA and vector were examined.
No evidence of recombination was found in any of the
cDNA clones sequenced; there were no clones that
contained unique makers from both RNAs (Fig. 1). All
differences from the consensus sequences of either RNA 1
or RNA2 were at single sites and did not occur at any of
the four marker positions used to distinguish the RNAs.
This indicates that identity in the 3' NCRs of RNA1 and
RNA2 of BBLMV is being maintained without high
levels of recombination.
Identity is being conserved even though a high number
of mutations were found. In the cDNA clones sequenced,
a total of 11 mutations were identified out of about
22000 nucleotides (approximately 5 x 10 ~), and these
appeared to be randomly distributed throughout the 3'
terminus (Fig. 1). The origin of these mutations cannot
be clearly identified as they may have occurred during
viral replication or during the cDNA synthesis by the
reverse transcriptase enzyme (estimated error rate of
reverse transcriptase is 1 x 10-4; Lewin, 1990). Therefore,
a reliable estimate of the mutation frequency of BBLMV
was not possible. The heterogeneity found in the BBLMV
population was expected as a consequence of the
reported high error rates of RNA polymerases and is
consistent with the heterogeneous nature reported for
other viral populations (Holland et al., 1982; Domingo
et al., 1985).
There was a total of 93 regions between markers where
recombination could have been detected, as all clones
examined contained multiple markers. These markers
span a distance of about 1360 nucleotides. Based on
recombination studies in other RNA viruses, this
distance should be sufficient for recombination to occur.
For example in poliovirus recombination was detected
between markers only 190 bases apart (Kirkegaard &
Baltimore, 1986) and in brome mosaic virus recombination occurred in a 3' N C R about 300 bases long
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:39:41
Short communication
Nucleotides
15
3'NCR 5"...11
110
I
1
2
3
4
5
r
r
e--
T
c
- -
c--
T- r
- -
c - -
RNAI
[l]...r--o--e--
RNA2
[2]...e--G--
(14)
[1]...
(19)
[1]...
(24)
[1]...r--*
(26)
(28)
372 514 560
I
I
I
G
1075
k
7
T
[11...
(30)
--
c -- r - T
- * A * - - T
(32)
-
(33)
-
(38)
[I]...T
(54)
[1]...r--
(55)
(67)
(68)
T
--
T
A
-
A
-
T
-
T
A
-
-
T
A
-
-
T
[l]...r
A
-
-
T
[1]...
A
-
-
T
[1]...T
A
-
-
T
G
(74)
-- G -
(80)
--
Q
-
-
-
(83)
(99)
1302
r - T
A
-
-
T
r
A
-
-
T
A
-
-
T
A
-
-
T
A
-
-
T
- r
T
[1]...
(104)
- C - - T - T
(146)
[1]...
A
-
-
T
(152)
[11...
A
-
-
T
G
-
-
C
-
C
-
C
- - G - -
C
(1)
(3)
-
(8) [ 2 ] . . . c
(9)
(11) [2l...e (12)
G
-
-
G
G
-
-
(23)
G
-
-
*
-
-
e
-
-
c
G
-
-
-
C
G- -
c - -
[2]...e -- G . . . . .
(45)
- T
C
C
G
-
-
C
12
G
-
-
C
G
-
-
C
G
-
-
e
G
-
-
c
(31)
(66)
-
e
[2]...
(49)
-
G
(20)
(34)
-
G
- T
(29)
*
G
(17)
(25)
-
C
c
G
e
G
-
-
C
e
G
-
-
C
G
-
-
C
[2]*..
C
(87)
- T - c
G
-
-
C
(103)
c
G
-
-
C
G
-
-
C
G
-
-
C
G
-
-
C
G
-
(107)
G - - c - - r - c
(122)
-
(129)
- T
(130)
(131)
*
-
-
G
C
c
[21...
(139)
(149)
-
c
-
[2]...
-
*
-
-
G
G
-
-
C
C
-
C
3"
2135
(Bujarski & Kaesberg, 1986). We found no evidence of
recombination in the 3' NCR of BBLMV, indicating that
recombinational events were too infrequent to be
detected in the sample population. Our results serve to
set the upper limits for recombination frequency in
BBLMV at 2"2 % (0 recombinants out of 45 clones) in
the population sampled and 1.1% (0 recombinants
between 93 markers) between markers within the 3'
NCR.
These estimates are consistent with recombination
frequencies found in other RNA viruses. In closely
related strains of poliovirus, the frequency of recombination between genetic markers 190 bases apart was
estimated at 0.13% (Kirkegaard & Baltimore, 1986).
This would be equivalent to approximately 1% recombination frequency over a distance equal to the
length of the 3' NCR ofBBLMV. Recombinants between
isogenic strains of another picornavirus, foot-and-mouth
disease virus, were detected at a frequency of 0"92 % in
vitro (McCahon et al., 1977).
Recombination has been reported in RNA plant
viruses, but recombination frequencies have not been
determined (Bujarski & Kaesberg, 1986; Allison et al.,
1990). In most studies of recombination in plant RNA
viruses, modified virus RNAs were placed under strong
selection for rescue of functional recombinants (Bujarski
& Kaesberg, 1986; Allison et al., 1990). For example, in
brome mosaic virus, a 20 nucleotide deletion was made in
the tRNA-like structure of the 3' NCR that is involved
in initiation of BMV RNA replication (Bujarski &
Kaesberg, 1986). Co-infection of the mutant RNA
component with the remaining wild-type components
resulted in the restoration of wild-type sequences of the
mutant RNA component through recombination. This
research indicates that under conditions of strong
selection pressure, recombination can play an important
role in restoring optimum (wild-type) sequences in
defective or poorly adapted genotypes. However, rescue
of the wild-type genome could have resulted from
extremely low levels of recombination, due to the high
rates of virus replication.
The conservation of duplicated regions on the 3'
terminus of BBLMV and other nepoviruses has caused
Fig. 1. C o n s e n s u s c D N A sequences for the 3' N C R s o f R N A 1 a n d
R N A 2 o f B B L M V for the seven p o s i t i o n s where differences were f o u n d
between clones 24 a n d 34. These clones were used o r i g i n a l l y to
d e t e r m i n e the sequence of the 3' termini o f R N A 1 a n d R N A 2 . The
m a r k e r g e n o t y p e s o f 45 B B L M V c D N A clones are aligned w i t h
consensus sequences ( m a r k e r p o s i t i o n s l a c k i n g a n u c l e o t i d e s y m b o l
were n o t sequenced in t h a t region). B r a c k e t e d n u m b e r s indicate 200 to
400 nucleotides o f c o d i n g sequence f r o m either R N A 1 [1] or R N A 2 [2];
the c D N A clone n u m b e r is given in p a r e n t h e s e s ; m u t a t i o n s (nucleotides
different from the consensus sequence) are m a r k e d w i t h a n asterisk. All
sequences end w i t h a poly(A) sequence at the 3' end.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:39:41
2136
Short communication
speculation that recombination may be involved in
preserving this identity. The exchange of RNA sequences
between genomic RNAs would transfer unique mutations from one RNA to the other, equalizing any changes
that have occurred during replication and preventing the
accumulation of neutral mutations. However, in order
for recombination to contribute significantly to the
maintenance of identity it would have to occur at
relatively high frequencies. The results described herein
indicate that recombination is relatively rare.
Rott et al. (1991a, b) have suggested that only the 3'terminal region of one of the genomic RNAs may serve
as a template for both 3' NCRs in TomRSV. The authors
proposed that the replication of the negative strand
always begins on the same genomic RNA and that
template switching occurs at or near the junction between
the non-coding and coding regions, thus maintaining
identical sequences in the 3' NCRs of both RNAs. A
similar mechanism has been proposed for coronaviruses,
where a leader RNA begins transcription from one end
of the RNA template, dissociates, then rejoins the
template at a downstream site to serve as a primer for
transcription of a subgenomic RNA (Lai, 1990). However, the results of this study clearly indicate that the 3'
NCRs of both RNA1 and RNA2 of BBLMV serve as
templates for replication. Nineteen of the Y cDNA
clones sequenced were sufficiently long to contain
portions of coding regions from either RNA1 or RNA2.
In all these clones, the unique nucleotides found in the 3'
NCR always corresponded to the same genomic RNA
(Fig. 1).
Since the history of the BBLMV gene pool involved is
unknown, it is possible that a recent recombinational
event is responsible for the near identity between the 3'
ends of BBLMV. The recombinant genotype would have
to confer sufficient selective advantage for the variant to
predominate in the viral population. However, this type
of event would be rare and unlikely to be responsible for
the conservation of nearly identical 3' ends observed in
all nepoviruses sequenced to date.
If recombination is not occurring at high frequencies
in BBLMV, what is maintaining identity in the 3' NCR?
Since the exact mutation frequency of the BBLMV RNA
polymerase is not known, it is unclear what level of
recombination is necessary to maintain the observed
sequence conservation. Therefore, we cannot exclude the
possibility that low levels of recombination may be
occurring and aiding in the maintenance of identity.
However, the high number of mutations observed (well
above expected reverse transcriptase background error)
in the BBLMV population and the lack of observed
recombination indicate that other mechanisms, such as
selection, play an important role in the conservation of
specific sequences.
References
ALLISON, R., THOMPSON, C. 8¢ AHLQUIST, P. (1990). Regeneration of
functional RNA virus genome by recombination between deletion
mutants for cowpea chlorotic mottle virus 3a and coat genes for
systemic infection. Proceedings" of the National Academy of Sciences,
U.S.A. 87, 1821~1824.
ANGENENT, G.C., POSTHUMUS, E., BREDERODE, F.T. & BOL, J.F.
(1989). Genome structure of tobacco rattle virus strain PLB: further
evidence on the occurrence of RNA recombination among tobraviruses. Virology 171, 271-274.
BACHER, J.W., WARKENTIN, D., RAMSDELL, D. • HANCOCK, J.F.
(1994). Sequence analysis of the 3' termini of blueberry leaf mottle
nepovirus. Virus Research (in press).
BERTIOLI, D.J., HARRIS, R.D., EDWARDS, M.L., COOPER, J.I. &
HAWES, W. S. (1991). Transgenic plants and insect cells expressing
the coat protein of arabis mosaic virus produce empty virus-like
particles. Journal of General Virology 72, 1801-1809.
BRAULT, V., HIBRAND, L., CANDRESSE, T., LE GALL, O. & DUNEZ, J.
(1989). Nucleotide sequence and genetic organization of Hungarian
grapevine chrome mosaic nepovirus RNA 2. Nucleic Acids Research
17, 7809-7819.
BUJARSKI,J. J. ~/. KAESI~ERG,P. (1986). Genetic recombination between
RNA components of a multipartite plant virus. Nature, London 321,
528-531.
DEVEREUX,J., HAEBERLI,P. & SMITHIES,O. (1984). A comprehensive set
of sequence analysis programs for the VAX. Nucleic Acids Research
12, 387-395.
DOMINC,O, E. & HOLLAND, J. J. (1988). High error rates, population
equilibrium and evolution of RNA replication systems. In RNA
Genetics, vol. 3, pp. 3 36 Edited by E. Domingo, J. J. Holland & P.
Ahlquist. Boca Raton: CRC Press.
DOMINGO, E., MARTINEZ-SALAS,E., SOBRINO, F., DE LA TORRE, J. C. &
PORTELA, A. (1985). The quasispecies (extremely heterogenous)
nature of viral RNA genome populations : biological relevance - a
review. Gene 40, 1-8.
FRAENKEL-CONRAT,H., SINGER,B. & TSUGITA, A. (1961). Purification
of viral RNA by means of bentonite. Virology 14, 51 58.
GREIF, C., HEMMER, O. & FRITSCH, C. (1988). Nucleotide sequence of
tomato black ring virus RNA-1. Journal of General Virology 69,
1517 1529.
GUBLER, U. & HOFFMAN, B.J. (1983). A simple and very efficient
method for generating cDNA libraries. Gene 25, 263 269.
HOLLAND, J. J., SPINDLER, K., HORODYSKI, F., GRABAU,E., NICHOL, S.
& VANDEPOL,S. (1982). Rapid evolution of RNA genomes. Science
215, 1577-1585.
KIRKEGAARD, K. & BALTIMORE, D. (1986). The mechanisms of RNA
recombination in poliovirus. Cell 47, 433.
LAI, M.M. (1990). Coronaviruses: organization, replication and
expression of genome. Annual Review of Microbiology 44, 303-333.
LEWlN, B. (1990). Genes IV, p. 349. Oxford: Oxford University Press.
MCCAHON, D., SLADE, W. R., PRISTON, R. A. J. & LAKE, J. R. (1977).
An extended genetic recombination map for foot-and-mouth disease
virus. Journal of General Virology 35, 555-565.
MA~rTHEWS, R. E. F. (1991). Plant Virology, 3rd edn, pp. 17~176.
London & New York: Academic Press.
MEYER, M., HEMMER, O., MAYO, M.A. & FRITSCH, C. (1986). The
nucleotide sequence of tomato black ring virus RNA-2. Journal of
General Virology 67, 1257-1271.
RAMSDELL, D. C. & STACE-SMITH, R. (1981). Physical and chemical
properties of the particles of ribonucleic acid of blueberry leaf mottle
virus. Phytopathology 71,468-472.
ROTT, M. E., TREMAINE,J. H. ~¢ ROCHON, D. M. (1991 a). Comparison
of the 5' and 3' termini of tomato ringspot virus RNA 1 and RNA
2: evidence for RNA recombination. Virology 185, 468472.
ROTT, M. E., TREMAINE, J. H. & ROCHON, D. M. (1991b). Nucleotide
sequence of tomato ringspot virus RNA-2. Journal of General
Virology 72, 1505 1514.
SAMBROOIK, J., FRITSCH, E.F. & MANIATIS, T. (1991). Molecular
Cloning: A Laboratory Manual, 2nd edn. New York: Cold Spring
Harbor Laboratory.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:39:41
Short communication
SANGER,F. (1981). Determination of nucleotide sequence in DNA.
Science 214, 1205-1210.
SCOTT, N. W., COOPER, J. I., LIu, Y. Y. & HELLEN, C. U. T. (1992). A
1"5 kb sequence homology in T-terminal regions of RNA-I and
RNA-2 of a birch isolate of cherry leaf roll nepovirus is also present,
in part, in a rhubarb isolate. JournalofGeneral Virology73, 481-485.
SERGHINI,M. A., FUCHS, M., PINCK, M., REINBOLT, J., WALTER, B. &
PINCK, L. (1990). RNA2 of grapevine fanleaf virus: sequence
2137
analysis and coat protein cistron location. Journal of General
Virology 71, 1433-1441.
STEINHAUER, D. A. & HOLLAND, J. J. (1987). Rapid evolution of RNA
viruses. Annual Review of Microbiology 41, 40%433.
VIEIRA, J. &MESSlNG, J. (1987). Production of single stranded plasmid
DNA. Methods in Enzymology 153, 3-11.
(Received 15 September 1993; Accepted 1 March 1994)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:39:41