Article No. jmbi.1999.3169 available online at http://www.idealibrary.com on
J. Mol. Biol. (1999) 293, 781±793
Satellite Cereal Yellow Dwarf Virus-RPV (satRPV) RNA
Requires a DouXble Hammerhead for Self-cleavage
and an Alternative Structure for Replication
Sang Ik Song, Stanley L. Silver, Michelle A. Aulik, Lada Rasochova,
B. R. Mohan and W. Allen Miller*
Plant Pathology Department
Iowa State University
351 Bessey Hall
Ames, IA 50011, USA
The 110 nt hammerhead ribozyme in the satellite RNA of cereal yellow
dwarf virus-RPV (satRPV RNA) folds into an alternative conformation
that inhibits self-cleavage. This alternative structure comprises a pseudoknot with base-pairing between loop (L1) and a single-stranded bulge
(L2a), which are located in hammerhead stems I and II, respectively.
Mutations that disrupt this base-pairing, or otherwise cause the ribozyme
to more closely resemble a canonical hammerhead, greatly increase selfcleavage. In a more natural multimeric sequence context containing the
full-length satRPV RNA and two copies of the hammerhead, wild-type
RNA cleaves much more ef®ciently than in the 110 nt context. Mutations
in the upstream hammerhead, including a knock-out in the catalytic core,
affect cleavage at the downstream cleavage site, indicating that multimers
of satRPV RNA cleave via a double hammerhead. The double hammerhead includes base-pairing between two copies of the L1 sequence which
extends stem I. Disruption of L1-L1 base-pairing slows cleavage of the
multimer. L1-L2a base-pairing is required for ef®cient replication of
satRPV RNA in oat protoplasts. Mutations that affect self-cleavage of the
multimer do not correlate with replication ef®ciency, indicating that the
ability to self-cleave is not a primary determinant of replication. We present a replication model in which multimeric satRPV RNA folds into
alternative conformations that cannot form in the monomer. One potential metastable intermediate conformation involves L1-L2a base-pairing
that may facilitate formation of the double hammerhead. However, we
conclude that L1-L2a also performs some other essential function in the
satRPV RNA replication cycle, because the L1-L2a base-pairing is more
important than ef®cient self-cleavage for replication.
# 1999 Academic Press
*Corresponding author
Keywords: hammerhead ribozyme; pseudoknot; RNA folding; rolling
circle RNA; viroid-like RNA
Present addresses: L. Rasochova, University of
Wisconsin, Institute for Molecular Virology, 1525 Linden
Dr., Madison WI 53706, USA; B. R. Mohan, University
of Pennsylvania School of Medicine, Department of
Microbiology, 315 Johnson Pavilion, Philadelphia, PA
19104, USA
Abbreviations used: BYDV, barley yellow dwarf
virus; CYDV-RPV, cereal yellow dwarf virus-RPV;
satRPV RNA, satellite RNA of CYDV-RPV; satLTSV,
satellite lucerne transient streak virus; satVTMoV,
satellite velvet tobacco mottle virus; CarSV, carnation
stunt viroid-like RNA; PSTVd, potato spindle tuber
viroid; satTRSV, satellite tobacco ringspot virus; WT,
wild-type; HH, hammerhead; CS, cleavage site.
E-mail address of the corresponding author:
wamiller@iastate.edu
0022-2836/99/440781±13 $30.00/0
Introduction
Circular (D type) satellite RNAs (Mayo et al.,
1995) and viroids (Branch & Robertson, 1984) replicate via a rolling circle mechanism, in which the
infectious circular (‡) RNA is copied into linear,
multimeric (ÿ) strands (Bruening et al., 1991). The
negative strand is used for the synthesis of multimeric (‡) RNAs as either linear multimeric or circular monomeric forms. The multimeric (‡) RNA
cleaves and ligates, yielding circular (‡) monomeric RNAs (reviewed by Symons, 1992, 1997).
The infectious monomeric RNAs range from about
225 to 450 nt and do not encode any open reading
frames. Satellite RNAs differ from viroids in that
# 1999 Academic Press
782
they depend on their helper viruses for replication
and encapsidation. They have no sequence similarity to their helper virus genome. In contrast, viroids are autonomously replicating subviral
pathogens that do not require a helper virus. Thus,
they must be replicated by a pre-existing host
RNA polymerase(s) and they are not encapsidated
(reviewed by Diener, 1987; Riesner & Gross, 1985;
Semancik, 1987; Symons, 1990).
Under in vitro conditions, most viroids form a
very stable rod-like structure, whereas satellite
RNAs tend to form branched structures. Alternative, metastable foldings have been suggested to be
involved in the viroid replication process (Hecker
et al., 1988; Loss et al., 1991; Riesner et al., 1992).
Multi-hairpin viroid structures are formed relatively quickly during the in vitro synthesis of oligomeric RNA (Hecker et al., 1988) and folding
simulations of potato spindle tuber viroid (PSTVd)
RNA suggest that the RNA folding pathway plays
an important role in the viroid replication cycle
(Gultyaev et al., 1998). The functional importance
of RNA folding pathways is not likely con®ned to
viroids. Here we propose such alternatives in a satellite RNA.
The subject of this study is satellite cereal yellow
dwarf virus-RPV (satRPV) RNA which was previously called satellite barley yellow dwarf virus
(sBYDV) RNA (Miller et al., 1991). Its helper virus,
the former RPV serotype of BYDV (BYDV-RPV),
has been reclassi®ed as cereal yellow dwarf virusRPV (CYDV-RPV) in the new genus Polerovirus
(D'Arcy et al., 1999). The serotypes that remain
classi®ed as BYDV do not support any known satellite RNA. SatRPV RNA and nepoviral satellite
RNAs differ from other D type (rolling circle replication mechanism) satellite RNAs in that the
encapsidated form of the RNA is linear (Prody
et al., 1986; Miller et al., 1991). The molecule is circularized only after entering the cell (Chay et al.,
1997). The encapsidated forms of other D type satellite RNAs are predominantly circular (Symons,
1997).
Multimeric forms of D type satellite RNAs and a
few viroids undergo self-cleavage at the plus
strand hammerhead ribozyme site to generate the
monomeric unit (Forster & Symons, 1987; Symons,
1997). The hammerhead structure of (‡) polarity
satRPV RNA differs from canonical hammerheads
(Vaish et al., 1998) in at least three features
(Figure 1; Miller & Silver, 1991). Firstly, the trinucleotide at the 50 side of the cleavage site is AUA,
instead of GUC, which is found in naturally occurring hammerheads except those of cherry small circular viroid-like RNA (AUA in (‡) strand, GUA in
(ÿ) strand) (Di Serio et al., 1997), the minus strands
of satellite lucerne transient streak virus RNA
(satLTSV; Keese et al., 1983), and satellite velvet
tobacco mottle virus RNA (satVTMoV; Haseloff &
Syumons, 1982) which cleave at GUA. Secondly,
an unpaired cytosine base (C24) is present immediately 30 to the conserved CUGANGA sequence,
and an unpaired adenine base (A73) occurs
Roles of Alternative Conformations of satRPV RNA
Figure 1. Alternative structures in the hammerhead
ribozyme in satRPV (‡) strand RNA. Numbering of
nucleotides is based on the full-length satellite RNA
(Miller et al., 1991); the three major stem-loops are numbered as described by Hertel et al. (1992). Single-strand
bulges and loops in the three main stems are pre®xed
with L (Miller & Silver, 1991). A cleavage site is
indicated by arrowhead, the absolutely conserved
nucleotides of the catalytic core are in bold. (a) Twodimensional representation drawn as described by
Miller & Silver (1991). The pseudoknot alternative to the
hammerhead formed by base-pairing between bases in
loops L1 and L2a is indicated by the double-headed
arrow. Mutations used in the study are indicated.
(b) The same structures drawn to resemble the threedimensional crystal structure as described by Pley et al.
(1994) and Scott et al. (1995).
immediately 50 of the conserved GAAA sequence,
opposite C24 (Figure 1). The bases at these positions are normally Watson-Crick base-paired to
each other at the proximal end of stem II
(Bruening, 1989), except in carnation stunt viroidlike (CarSV) RNA (ÿ) strand which has unpaired
C and A bases at the same positions (Hernandez
783
Roles of Alternative Conformations of satRPV RNA
et al., 1992), and satLTSV RNA which has an
unpaired U base at the position homologous to
C24 (Forster & Symons, 1987). The most striking
and, so far, unique feature of the satRPV ribozyme
is that the loop (L1) at the end of stem I can basepair with a ®ve base, single-stranded bulge (L2a)
in the compound stem II (Figure 1). This L1-L2a
base-pairing forms an alternative tertiary structure
consisting of a kind of pseudoknot (or ``kissing
stem-loop'') that greatly inhibits self-cleavage of a
110 nt transcript comprising the (‡) strand satRPV
RNA ribozyme (Miller & Silver, 1991). This led us
to propose a molecular switching model in which
this tertiary structure controls the rate of self-cleavage and possibly other functions of the satellite
RNA.
We now ®nd that self-cleavage occurs much faster in the larger than unit length satellite RNA than
in the 110 nt hammerhead context. Our results
show that the (‡) satRPV RNA cleaves via a
double hammerhead. We also show that base-pairing between L1 and L2a is essential for satRPV
replication in oat protoplasts even though it is not
present in the double hammerhead and it inhibited
self-cleavage of the 110 nt ribozyme. On the basis
of these results, we propose a sliding model for the
possible folding pathway of multimeric (‡) satRPV
RNA.
Results
Self-cleavage of wild-type and mutated
transcripts in the 110 nt hammerhead context
We reported previously that the alternative
tertiary structure, which contains base-pairing
between L1 and L2a, attenuated self-cleavage of
the isolated 110 nt ribozyme of satRPV (‡) strand
(Miller & Silver, 1991). Transcripts containing each
of the single mutations that disrupt L1-L2a basepairing (G8U and C32A, Figure 1), self-cleave over
100-fold more rapidly than wild-type RNA
(Table 1). The double mutant G8U/C32A that
restored base-pairing, cleaved much more slowly
than either of the single mutants. Therefore,
alternative tertiary structure (L1-L2a base-pairing),
attenuates self-cleavage of the ribozyme in the isolated 110 nt context (Miller & Silver, 1991).
We next examined the role of the unpaired
nucleotides (C24 and A73) situated at the proximal
end of stem II. Insertion of unpaired bases at these
sites was shown to greatly reduce cleavage of an
ef®cient, arti®cial hammerhead (Ruffner et al.,
1990; Long & Uhlenbeck, 1994). Bases C24 or A73
were deleted in satRPV RNA hammerhead
mutants C24 and A73, respectively (Figure 2).
The cleavage rates of both single deletion mutants
did not differ signi®cantly from that of wild-type
(WT) (Figure 2). However, the hammerhead containing the double mutation, C24/A73, had at
least a 50-fold shorter half-life (39 minutes) than
either of the single mutants or the wild-type
sequence (Figure 2). Thus deletion of both C24 and
Table 1. Half-lives of wild-type and mutant satRPV
RNA transcripts
Mutant
Wild-type
C24
A73
C24/A73
C24/A73/C32A
C32A
G8U
G8U/C32A
G8U/G19A
110 nt hammerhead
structure alonea
(minutes)
1.5-mer structurea
(minutes)
>2000
>2000
>2000
39
6
5b
11b
180b
NT
45
75
55
90
35
60
>350
75
1
See Figure 1 for positions of altered bases. , deletion. NT,
not tested.
a
The half-lives (t1/2) of uncleaved RNAs are calculated from
cleavage assays as in Figures 3 and 4.
b
Data from Miller & Silver (1991).
A73 together favored formation of hammerhead
over the pseudoknot, or it increased cleavage
chemistry once the hammerhead formed, or both.
Mutant C24/A73 is interesting because it
cleaves rapidly even though it can potentially form
the L1-L2a helix. To test whether disrupting the
L1-L2a helix could further increase cleavage of
C24/A73, we added the mutation C32A which
disrupts L1-L2a base-pairing. The half-life of
C24/A73/C32A (six minutes), is indeed faster
than that of C24/A73 (Figure 2), but it is not
signi®cantly different from the ®ve minute half-life
Figure 2. Self-cleavage of wild-type and mutant transcripts comprising the plus satRPV ribozyme shown in
Figure 1. Cleavage of 32P-labeled, 110-nt transcript was
initiated by addition of MgCl2. Products obtained after
the indicated cleavage times were analyzed by denaturing polyacrylamide gel electrophoresis (see Materials
and Methods). Cleavage fragments were quanti®ed by
phosphorimagery.
784
of the C32A mutant alone (Table 1). Either the
effects of the C24/A73 double deletion and the
C32A mutation are not additive or it is quite possible that this cleavage rate represents the detection
limit of our assay.
Self-cleavage of wild-type and mutated
transcripts in the context of greater than
full-length satellite RNA
All the above self-cleavage studies were conducted in the context of the isolated 110 nt ribozyme. In order to analyze cleavage and
replication rates in a more natural context, 470
nt, ``one-and-a-half-meric'' transcripts (1.5-mer,
designated by the pre®x 1.5-) were constructed
with the same mutations that were studied in
the isolated 110 nt ribozyme. These transcripts
contained two copies of the hammerhead structure (HH1 and HH2; Figure 3(a)) ¯anking the
full-length monomeric unit. In all the mutants,
only the ®rst hammerhead (HH1) was mutated.
The resulting full-length monomer obtained after
cleavage at both hammerheads contains the
desired mutation, because all mutations were 30
of the cleavage site. The cleavage rates at HH1
and HH2 were observed by measuring accumulation of the 50 and 30 fragments (5EF and 3EF),
respectively, that ¯ank the full-length monomer.
The overall result of cleavage at either site was
also observed by monitoring the decay of the
1.5-mer.
The 45 minute half-life of the wild-type 1.5-mer
(1.5-WT), calculated from three separate assays
(Table 1), was at least 45-fold shorter than that of
the 110 nt hammerhead alone (>2000 minutes,
Table 1). Several mutants that greatly increased
self-cleavage of the 110 nt hammerhead (C32A,
G8U/C32A, C24/A73, C24/A73/C32A)
had little or a slight negative effect on self-cleavage
of the 1.5-mer (Table 1 and Figure 3). This is probably due to the much more ef®cient cleavage of
wild-type 1.5-mer compared to the 110 nt transcript. The effects of mutations in L1 and L2a were
nearly the opposite of those observed in the 110 nt
context: mutant 1.5-G8U cleaved much more
slowly than wild-type, cleavage of 1.5-C32A was
slightly reduced, and the double mutant, 1.5-G8U/
C32A cleaved at wild-type rates (Table 1 and
Figure 3).
Another surprise was that the mutations in HH1
affected cleavage mainly at the HH2 cleavage site
(CS2) as indicated by accumulation of fragments
5EF-M (348 nt) and 3EF (122 nt) (Figure 3). One
possible explanation is that multimeric satRPV
RNA self-cleaves via a double hammerhead that
requires base-pairing between the two copies of
L1, as shown in Figure 4. This prevents formation
of the alternative pseudoknot structure (base-pairing between L1 and L2a) and explains why the 1.5G8U mutation, which weakens the L1-L1 basepairing, has such a strong negative effect on the
cleavage rate of the 1.5-mer. It does not explain
Roles of Alternative Conformations of satRPV RNA
why restoration of L1-L2a base-pairing in the
double mutant, 1.5-G8U/C32A, restores the cleavage rate (see Discussion).
Additional support for a double hammerhead
structure comes from the behavior of a single
base mutation that arose fortuitously in a 1.5G8U clone. The essential CUGA sequence in
HH1 was mutated to CUAA (1.5-G8U/G19A,
Figures 1 and 5). Because G19 (G5 in standard
hammerhead nomenclature; Hertel et al., 1992) is
an essential base in the catalytic pocket, the
G19A mutation completely destroys ribozyme
function (Ruffner et al., 1990; Symons, 1992;
Long & Uhlenbeck, 1994). The 1.5-G8U/G19A
RNA cleaved normally at CS1 but even more
slowly (if at all) than G8U at CS2 (Figure 5, see
the accumulation of 5EF (26 nt) and 3EF (122
nt), respectively). The simplest explanation is
that CS1 was cleaved by HH2 via a double hammerhead structure (Figure 4). The lack of
accumulation of the 122 nt fragment (3EF) from
CS2 also indicates that very little, if any, cleavage occurred via a single hammerhead.
Biological activity of mutant satRPV RNAs
Previously, we proposed that the L1-L2a basepairing was required for a role in replication other
than self-cleavage (Miller & Silver, 1991). To determine whether it has a role in the satRPV RNA
replication cycle, plant cells were inoculated with
satRPV RNAs containing mutations that affect L1L2a base-pairing and with other hammerhead
mutants. In separate inoculations, each monomeric
mutant was electroporated along with CYDV-RPV
helper virus genomic RNA into oat protoplasts.
After 36 hours, Northern blot hybridization of total
RNA revealed accumulation of oligomeric series of
both strands of satRPV RNA, indicating rolling circle replication (Figure 6). Mutant G8U accumulated
approximately 50-fold less plus strand than wildtype satRPV. The C32A mutant replicated even
more poorly in both strands. The double mutant
G8U/C32A, which restores L1-L2a base-pairing,
replicated only threefold less ef®ciently than wildtype satRPV RNA. Mutant G8U/G19A, which
abolished plus strand cleavage by hammerhead
structure, replicated poorly, as expected. Interestingly, both G8U and G8U/G19A accumulated signi®cant amounts of minus strand. Mutations
C24, A73 and C24/A73, had little or no
effect on replication ef®ciencies, although A73
was enriched for high molecular mass plus strands
(Figure 6). All the mutants except C32A accumulate some minus strand. It appears that the
mutation in L2a was more harmful to replication
than the mutation in L1, even though the double
mutant restored nearly complete replication
activity. Clearly, base-pairing between L1 and L2a
is much more important than their primary
sequences for replication of satRPV RNA.
785
Roles of Alternative Conformations of satRPV RNA
Discussion
Effects of mutations on self-cleavage rate of
110 nt hammerhead context
The results presented here support our notion
that satRNAs are so compact that any given
sequence may participate in more than one function. These different functions, such as self-cleavage, ligation, origins of replication and assembly
may be determined by different conformations of
the same sequence. According to the rolling circle
model, multimeric copies are formed from a monomeric, circular template. The multimer must then
fold in a way to allow the hammerhead conformation to form, despite potential competing folding possibilities. Uhlenbeck and colleagues (Fedor
& Uhlenbeck, 1992; Stage-Zimmermann &
Uhlenbeck, 1998) described three major steps in the
cleavage process: formation of an active ribozyme
structure, phosphodiester bond cleavage, and dissociation of the cleaved fragments. Often the rate
of formation of the active structure is slower than
the actual cleavage rate and sometimes a portion
of the molecules remain in an inactive conformation and never cleave (Stage-Zimmermann &
Uhlenbeck, 1998). In the 110 nt context, we proposed that the structure exists in an equilibrium
between the inactive pseudoknot (base-pairing
between L1 and L2a) (Figure 1) and the hammerhead, with the majority of the molecules remaining
folded in the inactive pseudoknot conformation
(Miller & Silver, 1991). Alternatively, because the
crystal structure indicates that L1 and L2a may be
in proximity in the hammerhead conformation
(Pley et al., 1994; Scott et al., 1995), base-pairing
between them may not greatly distort the active
site and rapid cleavage may occur in the pseudoknot context, but the reverse reaction (re-ligation)
may be allowed, due to L1-L2a base-pairing holding the cleavage products in place long enough to
allow re-ligation.
The unpaired nucleotides (C24 and A73) situated
at the proximal end of stem II of the hammerhead
may provide suf®cient length and ¯exibility in
stem II to allow L1-L2a base-pairing (Figure 1).
Thus, deletion of both bases may move the folding
equilibrium toward the hammerhead, in the 110 nt
context. Alternatively, this double deletion may
directly increase the activity of the catalytic core
(Ruffner et al., 1990), and thus increase the actual
chemical rate of cleavage. However, mutant 110 nt
transcripts that disrupt base-pairing between L1
and L2a, G8U, C32A and C24/A73/C32A,
cleave much faster than C24/A73. This is consistent with the ®rst possibility, that C24/A73
shifts structural equlibrium toward the hammerhead rather than increasing the cleavage rate per se,
because G8U and C32A still have the C24-A73
mismatch, as does a very rapid cleaving mutant
that has a minimal stem II with no L2a sequence at
all (M5 by Miller & Silver, 1991). Furthermore,
deletion of C24 and A73 had little effect on the rate
of cleavage (or replication) in the 1.5-mer structures (1.5-C24/A73; Figure 3). Thus, C24 and
A73 play a role in the formation of the alternative,
non-cleaving structure in the 110 nt context, but
not in the multimeric context.
Cleavage via double hammerhead
Wild-type satRPV RNA in the 1.5-meric context
cleaved much faster than in the 110 nt hammerhead context (Table 1). Mutations in HH1, including one that inactivates the active site, affected
cleavage at cleavage site 2 (CS2) much more than
at CS1 (Figure 3). This strongly supports cleavage
via a double hammerhead (Figure 4). Attenuation
of self-cleavage by ``intra-monomer'' L1-L2a basepairing seems to occur only in the 110 nt hammerhead context in which it is the strongest uninterrupted helix. In contrast, in the 1.5-mer, basepairing between L1 and L1 (and adjacent bases),
and/or between L2a and L2a are predicted by
MFOLD (Zuker, 1989) to be more stable than L1L2a. Thus, the double hammerhead which requires
L1-L1 base-pairing forms readily (Figure 4). Consequently, the 1.5-G8U mutant, in which the L1-L1
helix is weakened, cleaved more slowly than the
other mutants. The L2a-L2a helix is not necessary
for the double hammerhead, thus altering it had
little effect on the cleavage rate of 1.5-C32A
(Figure 3).
The double hammerhead could not form in the
circular monomer. This would be an advantage
because it would ensure that the circular template
for (ÿ) strand synthesis would remain intact. This
also implies that the linear forms that are encapsidated are derived only from cleaved (‡) multimer
and not from a circular template. Thus, we propose
that the origin of assembly structure is not present
in the circular monomer.
Most satellite RNAs undergo intramolecular
self-cleavage via single hammerhead structures.
However, those with a very short stem III are
known (Forster et al., 1988; Hernandez et al.,
1992), or predicted (Collins et al., 1998) to cleave
via a double hammerhead that extends the
length of stem III. Stability of the single hammerhead stem III was found to be an important
factor in determining whether self-cleavage
occurs by a single or double hammerhead structure (Sheldon & Symons, 1989). However,
satRPV RNA cleaves via a double hammerhead
despite a stable stem III (ten base-pairs with
a one-base bulge) in the single and double
hammerheads. In this case, the weak four-base
stem I in the single hammerhead structure is
strengthened by the adjacent L1-L1 base-pairing
in the double hammerhead (Figure 4). The only
other natural hammerhead with a stem I this
short is in the (ÿ) strand of CarSV RNA. However, it self-cleaves ef®ciently as a single hammerhead (Hernandez et al., 1992). This is also
the only other hammerhead with the C-A mismatch at the proximal end of stem II. To our
786
Roles of Alternative Conformations of satRPV RNA
Figure 3 (legend opposite)
Roles of Alternative Conformations of satRPV RNA
Figure 4. A schematic diagram of proposed double
hammerhead structure for cleavage of multimeric
satRPV (‡) strand RNA. Stems II and III are is as in the
single hammerhead, except that strands of stem III are
from different monomeric units. Stem I is increased by
L1-L1 base-pairing. A broken line indicates the remainder of satRPV RNA monomer not involved in (‡)
strand hammerhead formation.
knowledge satRPV RNA would be the ®rst
satRNA whose ribozyme requires a double
hammerhead for which the double hammerhead
stabilizes stem I rather than stem III.
Replication in oat protoplasts
The effect of covarying mutants G8U, C32A and
G8U/C32A on replication in protoplasts showed
that base-pairing between L1 and L2a is essential
for replication of satRPV RNA, independent of
their primary sequences. The slightly reduced
787
accumulation of (‡) strand on infection with the
double mutant G8U/C32A compared to wild-type
could re¯ect the slightly weaker base-pairing
between L1 and L2a. The wild-type helix has a
G ˆ ÿ 12.0 kcal/mol versus ÿ10.6 kcal/mol for
G8U/C32A. The G8U mutation weakens L1-L1
base-pairing in the double hammerhead and
greatly slows the cleavage rate of the 1.5-mer
(Table 1), yet this mutant replicates signi®cantly,
albeit at a low level (Figure 6). The fact that it produces substantial (ÿ) strand supports the notion
that cleavage is not necessary for a satRNA to act
as a replicase template (Sheldon & Symons, 1993).
In contrast, mutant 1.5-C32A is virtually dead,
even though it has only a slightly reduced cleavage
rate and is not involved in the double hammerhead. Thus, ef®ciency of multimer cleavage in vitro
is not a determinant of replication ef®ciency.
The L1-L2a helix may play a positive role in
some function other than self-cleavage, such as
replicase recognition or ligation. The unpaired
ACAAA sequence, or its complement, in stem II of
the hammerhead could be a replication origin, as it
is identical with the conserved 50 -terminal ®ve
bases of the genomic and subgenomic RNAs of
one of its helper viruses, beet western yellow virus
(Rasochova et al., 1997) and differs by one transition from the 50 end of CYDV-RPV (ACGAA; R.
Beckett, unpublished data). Because it is opposite
L2a in stem II, this ACAAA sequence would likely
have different conformations in the presence or
absence of L1-L2a base-pairing.
L1-L2a base-pairing may be required for circularization. Alternative conformations for cleavage
and ligation have been identi®ed in other RNAs.
In PSTVd, which requires host components for
cleavage, the switch from cleavage to ligation is
driven by a change from a tetraloop to a conformation that resembles the loop E structure found
in 5 S ribosomal RNA (Baumstark et al., 1997).
Chay et al. (1997) showed that linear satellite
tobacco ringspot virus (satTRSV) RNA folds into a
non-hammerhead conformation to facilitate ef®cient ligation to form a circular molecule. The ligation structure is the most stable predicted
structure for monomeric satTRSV RNA. It juxtaposes the 50 and 30 ends of the linear form in close
Figure 3. Synthesis and self-cleavage of wild-type and mutant 1.5-meric satRPV RNAs. (a) Right, Map of 1.5-meric,
uncleaved transcript. Bold vertical bars indicate L1 and L2a regions. CS1 and CS2 indicate the cleavage sites (small
vertical arrowheads) in the upstream hammerhead (HH1) and downstream hammerhead (HH2), respectively, that
delimited the infectious, 322 nt monomer (shaded). Asterisks indicate sites that were mutated (all in HH1, downstream of CS1). Shaded regions at the ends of the open box indicate vector-derived bases (not present in infectious
monomer). Bold lines below map indicate full-length (F, 470 nt) transcript, partial cleavage products (5EF-M, 444 nt;
and M-3EF, 348 nt), monomeric satRPV RNA (M, 322) and terminal cleavage fragments (5EF, 26 nt; and 3EF, 122 nt).
Left, Autoradiographs of cleavage products of wild-type and mutant 1.5-meric (470 nt) transcripts at designated
times after addition of MgCl2 to the gel-puri®ed 1.5-mer. The 32P-labeled products were separated by electrophoresis
on a 6 % polyacrylamide, 7 M urea denaturing gel. (b) Plots of relative counts in bands from gels shown in (a) and in
Figure 5 (WT-b and G8U/G19A) measured by phosphorimagery and quanti®ed with ImageQuantNT4.0 software
(Molecular Dynamics). A decrease in full-length 1.5-mer re¯ects cleavage at CS1 or CS2; an increase in 50 fragment
shows cleavage rate only at CS1, and the increase in the 30 fragment shows cleavage rate only at CS2.
788
Roles of Alternative Conformations of satRPV RNA
Figure 5. Cleavage of wild-type
and G8U/G19A mutant 1.5-meric
satRPV RNAs. Right, Map of M19/
19A mutant transcript and cleavage
products as in Figure 3A, with
G19A mutation indicated (*). Left,
Autoradiographs
of
cleavage
products of 1.5-meric (470 nt) transcripts at the indicated times after
addition of MgCl2. 32P-labeled
products were separated by electrophoresis on a 6 % polyacrylamide,
7 M urea denaturing gel. A plot
of quanti®ed cleavage product
accumulation
is
shown
in
Figure 3(b).
proximity via base-pairing. In satRPV RNA, L1-L2a
base-pairing leads to a non-hammerhead structural
arrangement in the most stable predicted structure
of monomeric satRPV RNA (Figure 7(e)). However,
several bases remain single stranded at the 50 end
of satRPV RNA, unlike satTRSV, and considerable
¯exibility would be predicted between the 50 and
30 ends. Perhaps the pseudoknot structure is the
optimal ligation substrate. The 50 and 30 ends
would be held in close proximity with little ¯exibility, owing to the complex base-pairing
(Figure 1(b)).
Alternative conformations of the very different,
complex ribozyme in the antigenomic RNA of
hepatitis delta virus (HDV) may serve similar biological roles as those in satRPV RNA. A short
duplex, P2a, can inhibit and promote ribozyme
activity depending on salt concentrations, in vitro
(Perrotta & Been, 1998). In this ribozyme, a fournucleotide sequence within the self-cleaving
domain pairs with a sequence just outside the 30 boundary of the ribozyme. Thus, it also appears
that the inhibitory effect is due to an alternative
secondary structure or a distortion of the threedimensional structure (Perrotta & Been, 1998).
Sliding model: possible folding pathway of the
multimeric (‡) satRPV RNA in the
cleavage reaction
We propose that satRPV RNA undergoes a
series of conformational changes, depending on the
stage of replication. The predicted secondary structure of lowest free energy state for multimeric (‡)
satRPV contains L1-L1 base-pairing but does not
form hammerheads owing to the presence of
additional L2a-L2a base-pairing that prevent formation of stem II (stable-conformation; Figure 7(a)).
A slight change in the stable conformation can generate either of two possible intermediate conformations (C-form and R-form; Figure 7(b) & (c))
which contain base-pairing between L1 and L2a.
The C-form contains inter-monomeric base-pairing
between L1 and L2a, whereas the L1 and L2a
sequences that base-pair in the R-form are from the
same monomeric satRPV unit. The R-form is a concatemer of the most stable monomeric satRPV
RNA structure (Figure 7(e)). Because it doesn't
form a hammerhead, it may be the conformation
of the stable, linear multimers that are present in
signi®cant quantities in virions (Miller et al., 1991).
The C-form structures contain the distal end of
stem II, which could serve as a nucleation point for
the transition to the structure that contains the
active double-hammerhead (active conformation;
Figure 7(d)). This requires disruption of L1-L2a
base-pairing and re-formation of L1-L1 helix. Thus,
we call this a sliding model, in which the two
strands of the stacked helices involving L1 and L2a
(in the stable structure) slide relative to each other
to form the L1-L2a helix in the intermediate structure (C-form), and then again to re-form the L1-L1
helix in the active double-hammerhead conformation. After cleavage, folding into the most stable
monomeric structure (Figure 7(e)) would facilitate
separation of the cleavage products, and prevent
the reverse reaction, by disruption of stem I and
the proximal end of stem II. Preliminary structure
probing data support the most stable monomer
structure (data not shown).
In this sliding model, a single hammerhead
never forms. This is supported by the cleavage pattern of the G8U/G19A mutant in which cleavage
occurs only at the 50 site (CS1). Very little, if any
self-cleavage (indicated by cleavage at the 30 site,
CS2) occurs. This is consistent with folding in
789
Roles of Alternative Conformations of satRPV RNA
Figure 6. Northern blot analysis
of total RNA extracted from oat
protoplasts inoculated with wildtype and mutant satRPV RNA. Protoplasts were electroporated with
CYDV-RPV genomic RNA and gelpuri®ed, monomeric satRPV RNA
transcript. Protoplasts were harvested at the indicated times after
electroporation. Total RNA was
fractionated on a denaturing 1.5 %
agarose gel, blotted to nylon membrane, and hybridized with antisense CYDV-RPV RNA probe to
detect viral genomic RNA (RPV
gRNA), antisense satRPV RNA
probe, or plus sense satRPV RNA
probe, as indicated. One blot was
used, stripped, and reprobed with
each of the probes. Exposures were
at ÿ80 C with one intensifying
screen for 16 hours. Mobilities of
monomeric and multimeric forms
(1-5 and X which represents bands
larger than pentamers) of satRPV
RNAs are indicated on the right.
Nearly equal amounts of (‡) and
(ÿ) satRPV RNA were present in
wild-type infections.
an inactive conformation such as the predicted
monomer structure (Figure 7(e)) after cleavage.
Otherwise, cleavage at CS1 by HH2 would allow
refolding into a single hammerhead for cleavage
by the same catalytic domain at CS2. In other selfcleaving satellite and viroid RNAs, the hammerhead is also absent in the most stable, predicted
conformation of the full-length monomer (Keese
et al., 1988; Hernandez et al., 1992; Chay et al., 1997;
Navarro & Flores, 1997; Collins et al., 1998).
This model also explains why restoration of L1L2a base-pairing in the double mutant, 1.5-G8U/
C32A, restores the cleavage rate even though the
L1-L1 base-pairing is weakened. Without L1-L2a
base-pairing, the multimer may remain trapped in
the stable conformation (Figure 7(a)) even if L1-L1
base-pairing is weakened. L2a-L2a (with ¯anking
base-pairs) is the strongest helix in the stable structure. Only in wild-type or G8U/C32A 1.5-mers,
could L1-L2a form transiently to produce the
C-form that we propose is the intermediate
required for formation of the active double
hammerhead conformation (Figure 7(d)).
All of the above models will be dif®cult to
test because of the dif®culty of capturing metastable conformations in vivo and assigning their
biological functions. Progress has been made in
identifying folding intermediates of the Tetrahymena group I intron ribozyme and PSTVd RNA.
The folding pathway has been predicted to be
important in the replication of PSTVd (Gast et al.,
1996; Gultyaev et al., 1998), and alternative con-
formations have been demonstrated to have
different functions (Baumstark et al., 1997). Both
the kinetics of folding and the presence of intermediates of the Tetrahymena ribozyme were
observed by a kinetic oligonucleotide hybridization assay (Zarrinkar & Williamson, 1994;
Treiber et al., 1998) and synchrotron hydroxyl
radical footprinting (Sclavi et al., 1998).
While conformational changes associated with
functional changes have been demonstrated for
other satellite and viroid monomers, here we provide evidence of alternative conformations that can
form only in a multimeric context. The alternative
structures of satRPV multimeric RNA provide a
new example of the complex and subtle means by
which a small, non-coding RNA can provide structural information needed to effect its replication by
host and viral proteins. Any studies of RNA structure and function, or RNA-protein interactions
should bear in mind the multifunctional, structurally ¯exible potential of any given RNA sequence.
Materials and Methods
Synthesis of hammerhead context and
1.5-mer structures
RNA molecules were synthesized by in vitro transcription of satRPV sequences in pGEM3Zf(ÿ) (Promega,
Madison, WI). Plasmid pSS1 contains the wild-type (‡)
strand satRPV hammerhead sequence: bases 310-322 and
1-89 of the satellite sequence (Miller et al., 1991). Cleavage occurs between bases 322 and 1. Plasmids and
790
Roles of Alternative Conformations of satRPV RNA
-
-
A*U
A*U
A*U
A*U
G*C
G*C
C
C
C*G
C*G
A U
A*U
A U
A*U
A G
A
A
C*G
A G
A
A
C*G
A
G
A
G
C*G
A
G
A
G
C*G
G*C
C
A
U*A
G*C
C
A
U*A
A*U
A
U
A*U
A*U
A
U
A*U
U*G
A
A
A
U*G
A
A
A A
A
C
C*G
C
C
C A A
A
A
C
C*G
C
C
C
A
A
UGC UCGUC
AGA
CAG GCGCGU CUG---UCU-----GACGA GUA---UCCGCGCGGA
UGC UCGUC
AGA
CAG GCGCGUACAGA GCGCGU CUG---UCU----GACGA GUA---UCCGCGCGGA
***
***** ***
**********
*** *****
***
*** ****** ***
***
***** ***
**********
*** *****
***
*** ******
AGA
CUGCU CGU
AGGCGCGCCU---AUG AGCAG-----UCU---GUC UGCGCG GAC
AGA
CUGCU CGU
AGGCGCGCCU---AUG AGCAG-----UCU---GUC UGCGCGC
C
C
G*C
C
A
A
A A C
C
C
G*C
C
A
A
A
A
A
G*U
A
A
A
G*U
U*A
U
A
U*A
U*A
U
A
U*A
A*U
A
C
C*G
A*U
A
C
C*G
G*C
G
A
G
A
G*C
G
A
G
A
G*C
A
A
G A
G*C
A
A
G A
U*A
U A
U*A
U A
G*C
G*C
C
C
C*G
C*G
U*A
U*A
U*A
U*A
-
-
G A
G A
U
C
U
C
C*G
C*G
U*A
U*A
G*C
A*U
G*C
A*U
U
G G
A*U
U
G G
A*U
C*G
U
U
G*C
C*G
U
U
G*C
A*U
C*G
C
A*U
C*G
C
U*A
G*C
A U
C*G
U*A
G*C
A U
C*G
U
G*C
G
A A*U
U
G*C
G
A A*U
C
A
C C*G
C
A
C C*G
G AA*U
G AA*U
G*C
C
U
A C*G
G*C
C
U
A C*G
A
A
C*G
UC*G
A
G U*A
C*G
UC*G
A
G U*A
G*C
A*U
A
AAA*U
G*C
A*U
A
AAA*U
C*G
G*C
A
A
C*G
G*C
A
A
-ACAGAG*C-----G*CAAC
ACAGAG*C-----G*CAAC
ACAGAC-
-
-
-
-
-
--
--
--
--
-
-
-
-
-
CAAC*G-----C*GAGACA
CAAC*G-----C*GAGACAA
A
C*G
G*C
A
A
C*G
G*C
U*AAA
A
U*A
C*G
U*AAA
A
U*A
C*G
A*U G
A
G*CU
G*C
A*U G
A
G*CUA
G*C
A
G*C A
U
C
C*G
G*C A
U
C
C*G
G
G*C C
A
U*AAG
C
G*C C
A
U*AA
C
U*A A
G
C*G
U
U*A A
G
C*G
U
G*C
U A
C*G
A*U
G*C
U A
C*G
A*U
C
G*C
U*A
C
G*C
U*A
G*C
U
U
G*C
G*C
U
U
G*C
U*A
G G
U
U*A
G G
U
U*A
C*G
U*A
C*G
A*U
A*U
G*C
G*C
C
U
C
U
A G
A G
-
A*U
A*U
A*U
A*U
G*C
G*C
G G
C
G G
C
U
U
C*G
U
U
C*G
C*G
U A
A*U
C*G
U A
A*U
G*C A
C C*G
G*C A
C C*G
G*C G
A C*G
G*C G
A C*G
G U*A
G U*A
G AA*U A
G AA*U A
C U
AAA*U
C U
AAA*U
A
A
UC*G A
A
UC*G A
A
A*U A
A
A*U A
A
G*C A
C
G*C A
C
U
G C
UC
G*CAAC
A
U
G C
UC
G*CAAC
A
GAGCGCG--UAC GUC-U A GAC-GUA CGCGC
GAGCGCGACAGA GCGCG--UAC GUC-U A GAC-GUA CGCGC
***** *** *** * * *** *** *****
***** *** *** * * *** *** *****
*****
CGCGC AUG-CAG A U-CUG CAU--GCGCGAG
CGCGC AUG-CAG A U-CUG CAU--GCGCGAG
CGCGCCAAC*G
CU
C G
U
A
CAAC*G
CU
C G
U
A
CAAC*G
A C*G
C
A C*G
C
A C*G
A U*A
A
A U*A
A
A U*A
A
A G*CU
A
A
G*CU
A
A
G*CU
A
A
A
U*AAA
U C
U*AAA
U C
U*AAA
U C
G
G
A*U G
A U*AAG
A*U G
A U*AA
A*U G
A U*AA
G*C A
G C*G
G*C A
G C*G
G*C A
G C*G
G*C C
A C*G
G*C C
A C*G
G*C C
A C*G
U*A
A U
G*C
U*A
A U
G*C
U*A
A U
G*C
G*C
U
U
G*C
U
U
G*C
U
U
C
G G
C
G G
C
G G
C*G
C*G
C*G
U*A
U*A
U*A
U*A
U*A
U*A
U U
G G
U
U
C
G G
U
U
G G
U
A
UCC-GUGGAUA-ACAG GCGCGU*** ******* **** ******
AAG CACCUA UGUC UGCGCGA CU A
C
A
G
A
G
C
A
C*G C A G
A*U
U*A
A*U
A G*C
C
AU
G
A
C
A
G
C
A
C
A
C*G
C*G
U*A
G*CU A
C
G
U*AA
C*G
C*G
G*C
U
U
G G
U
C*G
G*C
G*C
G
A
G AA*U
C
U
A
UC*G
A
A*U
A
G*C
A
G*CAAC
C GAC-GUAUCCGCGC
A
*** *** *****
G CUGUCAU--GCGCGAGACA
U
G
A
U*A
C*G
A*U
A GGUGGCCACC*G
U U
C
*** *****
U
C
C CCA-CGGUG A
C*G
A
G*C
UU
A
A*U
A C*G
G*C
C*G
C*G
U*A
C
C
C*G
C
A
A
U*AC
G*C
U
U
A
A*U
U
A
G*CG
G
C
A*U
C*G
A
A*U
U*A
G
A*U
U*A
A
U*A
G*C
UU
A
G*C
U*A
U C
U
AUCCACCGAA
GUGUU
CUGGUG GUCC U
****** ***
*****
****** ****
AUAGGUG-CUU
CACAG
GACUACUCAGG CU
U
A*U
C
A
G*C
U*G
U*A
C*G
U*A
C*G
C*G
U*A
C*G
U*AA C
A*U
C
A
A C
A*UA
G C
G*C
C*G
U*A
A*U
A G*C
A
A
A
A
C U ACUC-UUGC
G **** **** A
C U UGAGCAACG
A*U
G*C
A*U
C
U
ACU
-
C*G
G*C
G*C
G AA*U
C
A
UC*G
A*U
G*C
G*C A
A
C
C
G
A
C
A
A
G
C
AU
C*G
U*A
A*U
U*A
G A C G*C
A
C
G
A
A
G
A
A UC
C
A
UUC-GUGGAUA-ACAG GCGCGU CUGU AUCCAC GAA
*** ****** **** ****** **** ****** ***
AAG CACCUA UGUC UGCGCG GACA-AUAGGUG-CUU
A CU A
C
A
A
G
A
G
C
A
C*GC A G
A*U
U*A
A*U
A G*C
C
AU
G
A
C
A
G
C
A
C
A
C*G
C*G
U*A
G*CU A
C
G
U*AA
C*G
C*G
G*C
U
U
G G
-
C*G
G*C
G*C
AA*U
G
C
A UC*G
A*U
G*C
G*C
AA
C
C
G
A
C
A
G
A
C
AU
C*G
U*A
A*U
U*A
G A C G*C
A
C
G
A
A
G
A UC
C
A
ACAGA GCGCGU CUGU AUCCAC GAA
****** ***
AUAGGUG-CUU
U
Figure 7 (legend shown opposite)
RNAs derived from them are named using the formula
xNy, where x is the wild-type base, N is its position in
satRPV RNA, and y is the new base ( indicates
deletion). Multiple mutations are separated by slashes. A
plasmid and its transcript are named similarly with the
lower case p indicating plasmid. Mutations in pC24,
pA73, pC24/A73, pC24/A73/C32A, pG8U,
pC32A and pG8U/C32A were introduced into pSS1
(Miller & Silver, 1991) by the two-step PCR method
described by Herlitze & Koenen (1990). The following
plasmids were made with indicated primers that are
complementary to the satRPV (‡) strand sequence.
pC24, 50 CGCGGATACTCGTCAGACAG 30 ; pA73,
50 GTGGATTTCGTATCTATTTG 30 ; pG8U, 50 AGACAGTACGAGCTCTGTTATC 30 ; and pC32A, 50
TTCTAGTCCGTCGGATACGTCG 30 . A base change or
791
Roles of Alternative Conformations of satRPV RNA
deletion () is underlined. To construct the 1.5-meric
satRPV cDNA clones (transcripts preceded by 1.5-), the
above wild-type or mutant derivatives of pSS1 were
ampli®ed with the following primers: MA418, 50
CTCAAGCTTATTTCGTGG 30 ; and MA419, 50 CTCAGATCTTTACTTCGGTGG 30 . The PCR products were
digested at their ends with HindIII and BglII (underlined) and cloned into the wild-type dimeric cDNA
clone, pT7sat (Rasochova & Miller, 1996), which had
been digested with the same enzymes. The presence of
the mutations was veri®ed by sequencing.
Self-cleavage assays
The [a-32P]UTP-labeled RNAs were transcribed in vitro
as described (Miller & Silver, 1991) from EcoRI-linearized
plasmids. The RNA transcript (SS1) from XbaI-linearized
pSS1 and the mutant derivatives, contained only three
vector-derived bases at the 50 end and at most six at the
30 end. Uncleaved transcripts were puri®ed by elution
in elution buffer (0.5 % NH4OAC; 1 mM EDTA; 0.1 %
(w/v) SDS) after electrophoresis of transcripts on a 6 %
polyacrylamide, 7 M urea gel. Conditions for self-cleavage of SS1 were 50 mM Tris-HCl (pH 7.5), 10 mM
MgCl2, at 37 C. Self-cleavage reactions contained
approx. 30,000 cpm gel-puri®ed, uncleaved transcription
product in cleavage buffer (50 mM Tris-HCl (pH 7.5),
10 mM Mg2Cl). Reactions were started by the addition
of the Mg2Cl. To stop the reaction, 8 ml aliquots were
removed at designated time points, added to an equal
volume of stop buffer (95 % formamide, 0.5 mM EDTA,
0.025 % SDS), and then immediately frozen at ÿ80 C.
Samples were thawed and denatured by boiling for one
minute and then quenching on ice immediately prior to
electrophoresis on 6 % polyacrylamide, 7 M urea sequencing-type gels. Gels were exposed on a Phosphorimager
and the bands quanti®ed using ImageQuaNT (Molecular
Dynamics, Sunnyvale, CA).
Isolation of satRPV RNA monomer
Unlabeled, 1.5-meric satRPV RNAs were synthesized
by in vitro transcription of EcoRI-linearized templates
using phage T7 RNA polymerase (RiboMax kit, Promega). To induce self-cleavage, the RNA transcripts were
incubated at 37 C in cleavage buffer for approximately
three hours. The 322 nt monomer obtained by self-cleavage was eluted with elution buffer from a 6 % acryl-
amide, 7 M urea gel after electrophoresis and staining
with ethidium bromide.
Electroporation of protoplasts and extraction of
total RNA
Oat (Avena sativa cv. Stout) protoplasts were isolated
from cell suspension culture (cell line S226 obtained from
Howard Rines, USDA/ARS, University of Minnesota) as
described by Dinesh-Kumar & Miller (1993). Protoplasts
were electroporated with 50 ng of viral RNA puri®ed
from a mixture of CYDV-RPV and BYDV-PAV viruses,
and 20 ng of gel-puri®ed monomeric satRPV RNA transcript. At designated times, cells were collected by centrifugation, quick frozen in liquid nitrogen, and stored at
ÿ80 C. Total RNA was isolated from protoplasts using
the RNeasy Plant Mini kit (Qiagen, Valencia, CA). The
RNA was ultimately eluted from the resin by resuspension in DEPC-treated water.
RNA analysis
Denaturing 1.5 % agarose gel electrophoresis and
Northern blot hybridization was as performed as
described by Rasochova & Miller (1996). Each lane was
loaded with equal amounts of total RNA as determined
by spectrophotometry and con®rmed by ethidium bromide staining of ribosomal RNA before Northern blot
hybridization. The 32P-labeled RNA probes were synthesized by in vitro transcription according to the manufacturer's instructions (Promega, Madison, WI) using
[a-32P]CTP label. Antisense satRPV RNA probe was synthesized by SP6 RNA polymerase transcription of HindIII-cut plasmid pT7Sat (Rasochova & Miller, 1996).
Sense satRPV RNA probe was prepared by T7 RNA
polymerase transcription of EcoRI-cut plasmid pT7Sat.
Antisense CYDV-RPV probe, complementary to bases
809-1191 was prepared by T7 RNA polymerase transcription of ApaI-cut plasmid pML5 (Rasochova et al.,
1997). Northern blot hybridization was as described by
Passmore & Bruening (1993) using 5 105 to 10 105
cpm/ml of radioactive probe (32P-labeled transcript) per
hybridization. Membranes hybridized with probes were
exposed to PhosphoroImager screens. Before re-probing
with another probe, blots were stripped by boiling in
0.1SSC and 0.1 % SDS, and exposed to ensure that all
counts had been removed. The MFOLD program is used
to predict secondary structure of lowest free energy state
for RNAs (Zuker, 1989).
Figure 7. Possible conformations of multimeric (‡) satRPV RNA to facilitate the cleavage reaction via the sliding
model. The conformations differ only in the hammerhead regions of multimeric satRPV RNAs; the secondary structure of the remainder of the satRPV RNA sequence (continuous lines) is predicted to be identical with that shown in
the monomeric satRPV RNA structure (e). Cleavage sites are indicated by solid triangles. L1 and L2a sequences are
boxed and shaded. Superscripts indicate the satRPV monomeric unit to which each copy of L1 and L2a belong. For
clarity, each satellite monomeric unit is shown in a different color. Curved broken lines at right of each structure indicate inde®nite number of monomeric repeats. (a) Stable conformation. The predicted secondary structure of lowest
free energy state for multimeric (‡) satRPV RNA (using MFOLD) containing n copies of the monomer. (b) The stable
conformation may rearrange by sliding of the two main strands on the horizontal axis relative to each other to form
the intermediate C-form Metastable conformation. This structure has L1-L2a base-pairing between monomeric units.
(c) Alternatively, the stable structure may fold so that L1-L2a from the same monomeric unit base-pair to each other
(R-form). This is a concatenation of the most stable monomeric structure (e). Both metastable structures include the
distal end of hammerhead stem II (shaded gold). (d) The horizontal strands in C-form metastable structure may slide
again, re-forming the L1-L1 base-pairing, but with L2a single stranded in stem II, to generate functional ``inter-monomeric'' hammerheads (double hammerheads). (e) After cleavage, the monomeric products could fold to the most
stable monomeric conformation (predicted by MFOLD and con®rmed by chemical and enzymatic probing (not
shown)).
792
Acknowledgments
The authors thank Olke Uhlenbeck and Tracy StageZimmermann for helpful discussions, and Randy Beckett
for technical support. This work was funded by USDA/
NRI grant number 94-37303-0469 and the Korean Science
and Engineering Foundation (to S.I.S.). This is paper No.
J-18478 of the Iowa State University Agricultural and
Home Economics Experiment Station Project 3267, and
supported by Hatch Act and State of Iowa funds.
References
Baumstark, T., Schroder, A. R. & Riesner, D. (1997). Viroid processing: switch from cleavage to ligation is
driven by a change from a tetraloop to a loop E
conformation. EMBO J. 16, 599-610.
Branch, A. D. & Robertson, H. D. (1984). A replication
cycle for viroids and other small infectious RNA's.
Science, 223, 450-455.
Bruening, G. (1989). Compilation of self-cleaving
sequences from plant virus satellite RNAs and
other sources. Methods Enzymol. 180, 546-558.
Bruening, G., Passmore, B. K., van Tol, H., Buzayan,
J. M. & Feldstein, P. A. (1991). Replication of plant
virus satellite RNA: evidence favors transcription of
circular templates of both polarities. Mol. PlantMicrobe Interact. 4, 219-225.
Chay, C. A., Guan, X. & Bruening, G. (1997). Formation
of circular satellite tobacco ringspot virus RNA in
protoplasts transiently expressing the linear RNA.
Virology, 239, 413-425.
Collins, R. F., Gellatly, D. L., Sehgal, O. P. &
Abouhaidar, M. G. (1998). Self-cleaving circular
RNA associated with rice yellow mottle virus is the
smallest viroid-like RNA. Virology, 241, 269-275.
D'Arcy, C., Domier, L. & Mayo, M. A. (1999). Luteoviridae. In Seventh Report of the International Committee
on the Taxonomy of Viruses, Elsevier, Oxford.
Diener, T. O. (1987). The Viroids, Plenum, New York.
Dinesh-Kumar, S. P. & Miller, W. A. (1993). Control of
start codon choice on a plant viral RNA encoding
overlapping genes. Plant Cell, 5, 679-692.
Di Serio, F., Daros, J. A., Ragozzino, A. & Flores, R.
(1997). A 451-nucleotide circular RNA from cherry
with hammerhead ribozymes in its strands of both
polarities. J. Virol. 71, 6603-6610.
Fedor, M. J. & Uhlenbeck, O. C. (1992). Kinetics of intermolecular cleavage by hammerhead ribozymes. Biochemistry, 31, 12042-12054.
Forster, A. C. & Symons, R. H. (1987). Self-cleavage of
plus and minus RNAs of a virusoid and a structural
model for the active site. Cell, 49, 211-220.
Forster, A. C., Davies, C., Sheldon, C. C., Jeffries, A. C.
& Symons, R. H. (1988). Self-cleaving viroid and
newt RNAs may only be active as dimers. Nature,
334, 265-267.
Gast, F. U., Kempe, D., Spieker, R. L. & Sanger, H. L.
(1996). Secondary structure probing of potato spindle tuber viroid (PSTVd) and sequence comparison
with other small pathogenic RNA replicons provides evidence for central non-canonical base-pairs,
large A-rich loops, and a terminal branch. J. Mol.
Biol. 262, 652-670.
Gultyaev, A. P., van Batenburg, F. H. & Pleij, C. W.
(1998). Dynamic competition between alternative
structures in viroid RNAs simulated by an RNA
folding algorithm. J. Mol. Biol. 276, 43-55.
Roles of Alternative Conformations of satRPV RNA
Haseloff, J. & Symons, R. H. (1982). Comparative
sequence and structure of viroid-like RNAs of two
plant viruses. Nucl. Acids Res. 10, 3681-3691.
Hecker, R., Wang, Z. M., Steger, G. & Riesner, D. (1988).
Analysis of RNA structures by temperature-gradient gel electrophoresis: viroid replication and processing. Gene, 72, 59-74.
Herlitze, S. & Koenen, M. (1990). A general and rapid
mutagenesis method using polymerase chain reaction. Gene, 91, 143-147.
Hernandez, C., Daros, J. A., Elena, S. F., Moya, A. &
Flores, R. (1992). The strands of both polarities of a
small circular RNA from carnation self-cleave
in vitro through alternative double- and single-hammerhead structures. Nucl. Acids Res. 20, 6323-6329.
Hertel, K. J., Pardi, A., Uhlenbeck, O. C., Koizumi, M.,
Ohtsuka, E., Uesugi, S., Cedergren, R., Eckstein, F.,
Gerlach, W. L. & Hodgson, R., et al. (1992). Numbering system for the hammerhead. Nucl. Acids Res.
20, 3252.
Keese, P., Bruening, G. & Symons, R. H. (1983). Comparative sequence and structure of circular RNAs
from two isolates of lucerne transient streak virus.
FEBS Letters, 159, 185-190.
Keese, P., Visvader, J. E. & Symons, R. H. (1988).
Sequence variability in plant viroid RNAs. In RNA
Genetics (Domingo, E., Holland, J. J. & Ahlquist, P.,
eds), vol. 3, pp. 71-98, CRC Press, Boca Ration, FL.
Long, D. M. & Uhlenbeck, O. C. (1994). Kinetic characterization of intramolecular and intermolecular
hammerhead RNAs with stem II deletions. Proc.
Natl Acad. Sci. USA, 91, 6977-6981.
Loss, P., Schmitz, M., Steger, G. & Riesner, D. (1991).
Formation of a thermodynamically metastable
structure containing hairpin II is critical for infectivity of potato spindle tuber viroid RNA. EMBO J. 10,
719-727.
Mayo, M. A., Berns, K. I., Fritsch, C., Kaper, J. M.,
Jackson, A. O., Leibowitz, M. J. & Taylor, J. M.
(1995). Subviral agents: Satellites. In Virus Taxonomy: Sixth Report of the International Committee on
Taxonomy of Viruses (Murphy, F. A., Fauquet, C. M.,
Bishop, D. H. L., Ghabrial, S. A., Jarvis, A. W.,
Martelli, G. P., Mayo, M. A. & Summers, M. D.,
eds), pp. 487-492, Springer-Verlag, Wien, New
York.
Miller, W. A. & Silver, S. L. (1991). Alternative tertiary
structure attenuates self-cleavage of the ribozyme in
the satellite RNA of barley yellow dwarf virus.
Nucl. Acids Res. 19, 5313-5320.
Miller, W. A., Hercus, T., Waterhouse, P. M. & Gerlach,
W. L. (1991). A satellite RNA of barley yellow
dwarf virus contains a novel hammerhead structure
in the self-cleavage domain. Virology, 183, 711-720.
Navarro, B. & Flores, R. (1997). Chrysanthemum chlorotic mottle viroid: unusual structural properties of a
subgroup of self-cleaving viroids with hammerhead
ribozymes. Proc. Natl Acad Sci. USA, 94, 1126211267.
Passmore, B. K. & Bruening, G. (1993). Similar structure
and reactivity of satellite tobacco ringspot virus
RNA obtained from infected tissue and by in vitro
transcription. Virology, 197, 108-115.
Perrotta, A. T. & Been, M. D. (1998). A toggle duplex in
hepatitis delta virus self-cleaving RNA that stabilizes an inactive and a salt-dependent pro-active
ribozyme conformation. J. Mol. Biol. 279, 361-373.
Roles of Alternative Conformations of satRPV RNA
Pley, H. W., Flaherty, K. M. & McKay, D. B. (1994).
Three-dimensional structure of a hammerhead ribozyme. Nature, 372, 68-74.
Prody, G., Bakos, J. T., Buzayan, J. M., Schneider, I. R. &
Bruening, G. (1986). Autolytic processing of dimeric
plant virus satellite RNA. Science, 231, 1577-1580.
Rasochova, L. & Miller, W. A. (1996). Satellite RNA of
barley yellow dwarf-RPV virus reduces accumulation of RPV helper virus RNA and attenuates
RPV symptoms in oats. Mol. Plant-Microbe Interact.
9, 646-650.
Rasochova, L., Passmore, B. K., Falk, B. W. & Miller,
W. A. (1997). The satellite RNA of barley yellow
dwarf virus-RPV is supported by beet western yellows virus in dicotyledonous protoplasts and
plants. Virology, 231, 182-191.
Riesner, D. & Gross, H. J. (1985). Viroids. Annu. Rev.
Biochem. 54, 531-564.
Riesner, D., Steger, G., Wiese, U., Wulfert, M., Heibey,
M. & Henco, K. (1992). Temperature-gradient gel
electrophoresis for the detection of polymorphic
DNA and for quantitative polymerase chain reaction. Electrophoresis, 13, 632-636.
Ruffner, D. E., Stormo, G. D. & Uhlenbeck, O. C. (1990).
Sequence requirements of the hammerhead RNA
self-cleavage reaction. Biochemistry, 29, 10695-10702.
Sclavi, B., Sullivan, M., Chance, M. R., Brenowitz, M. &
Woodson, S. A. (1998). RNA folding at millisecond
intervals by synchrotron hydroxyl radical footprinting. Science, 279, 1940-1943.
793
Scott, W. G., Finch, J. T. & Klug, A. (1995). The crystal
structure of an all-RNA hammerhead ribozyme: A
proposed mechanism for RNA catalytic cleavage.
Cell, 81, 991-1002.
Semancik, J. S. (1987). Viroids and Viroid-like Pathogens,
CRC Press, Boca Ration.
Sheldon, C. C. & Symons, R. H. (1989). RNA stem stability in the formation of a self-cleaving hammerhead
structure. Nucl. Acids Res. 17, 5665-5677.
Sheldon, C. C. & Symons, R. H. (1993). Is hammerhead
self-cleavage involved in the replication of a virusoid in vivo? Virology, 194, 463-474.
Stage-Zimmermann, T. K. & Uhlenbeck, O. C. (1998).
Hammerhead ribozyme kinetics. RNA, 4, 875-889.
Symons, R. H. (1990). Viroids and related pathogenic
RNAs. Semin. Virol. 1, 117-126.
Symons, R. H. (1992). Small catalytic RNAs. Annu. Rev.
Biochem. 61, 641-671.
Symons, R. H. (1997). Plant pathogenic RNAs and RNA
catalysis. Nucl. Acids Res. 25, 2683-2689.
Treiber, D. K., Rook, M. S., Zarrinkar, P. P. &
Williamson, J. R. (1998). Kinetic intermediates
trapped by native interactions in RNA folding.
Science, 279, 1943-1946.
Vaish, N. K., Kore, A. R. & Eckstein, F. (1998). Recent
developments in the hammerhead ribozyme ®eld.
Nucl. Acids Res. 26, 5237-5242.
Zarrinkar, P. P. & Williamson, J. R. (1994). Kinetic intermediates in RNA folding. Science, 265, 918-924.
Zuker, M. (1989). On ®nding all suboptimal foldings of
an RNA molecule. Science, 244, 48-52.
Edited by D. E. Draper
(Received 15 June 1999; received in revised form 2 September 1999; accepted 3 September 1999)