Volume 16 Number 3 1988
Nucleic Acids Research
Grapevine yellow speckle virold: structural features of a new virold group
Anna M.Koltunow and M.Ali Rezaian
CSIRO Division of Horticultural Research, PO Box 350, Adelaide, SA 5001, Australia
Received December 1, 1987, Accepted January 8, 1988
ABSTRACT
A single stranded circular RNA was isolated from grapevines infected with
yellow speckle disease. The RNA which we have called grapevine yellow
speckle viroid (GYSV), contains 367 nucleotide residues and has the potential to form the rod-like secondary structure characteristic of viroids.
GYSV has 37Z sequence homology with the recently described apple scar skin
viroid (ASSV; 330 residues) and has some sequence homology with the viroids
in the potato spindle tuber viroid (PSTV) group. The sequence of GYSV has
characteristics which fit the structural domains described for the PSTV
group. However, GYSV lacks the PSTV central conserved sequence. Instead,
there is a conserved sequence in the central region of GYSV and ASSV which
has the potential to form a stem loop configuration and a stable palindromic
structure as does the central conserved region of the PSTV group. These
structural features suggest there is a different central conserved region
for GYSV and ASSV. The results support the viroid nature of GYSV and its
inclusion into a separate viroid group which we suggest should be
represented by ASSV.
XMTltODPCTTOB
Yellow speckle (1) is a disease of grapevines which is wide spread in
the irrigated areas of Australia where most of the wine and drying grapes
are grown (2). The symptoms of the disease are small, yellow flecks, scattered over the leaf surface and these symptoms are prevalent in the hotter
months of summer in infected plants (1). Although the disease is virus-like
in nature, no virus particles have been isolated from diseased tissues and
it has been suggested that the agent could be a viroid (3-5).
Viroids, the
smallest known plant pathogens, are single-stranded circular RNA molecules
containing 246 to 375 nucleotides which can form rod—like secondary structures (6,7).
Viroids have been placed into two distinct groups on the basis
of sequence homologies.
These groups are represented by avocado sunblotch
viroid (ASBV) which is the sole member of this group and potato spindle
tuber viroid (PSTV; 6). The PSTV group contains eight viroids which have
been completely sequenced (6).
In this paper we report the characterization of a circular RNA isolated
© IR L Press Limited, Oxford, England.
849
Nucleic Acids Research
from grapevine tissue infected with yellow speckle disease.
In the absence
of infectivity data we have tentatively called this RNA grapevine yellow
speckle viroid (GYSV).
GYSV is related in sequence to the recently reported
apple scar skin viroid (ASSV; 8) and possesses structural and biological
properties characteristic of viroids.
MATERIALS AHD METHODS
Plants
Grapevine (Vitls vinifera L.) cultivars, were maintained in growth
cabinets as described previously (9) or grown in the field.
The grapevine
source. Cabernet franc, infected with yellow speckle disease was grown from
material generated by fragmented shoot apex culture (4).
Purification of GYSV and the other grapevine circular RNAs
Total RNA was extracted from yellow speckle infected tissue as described previously (9) and the circular RNA was separated from other nucleic
acids by electrophoresis in two dimensional (2D) polyacrylamide gels (9).
Circular RNA bands were excised, the RNA was eluted, ethanol precipitated
and subjected to further purification by prolonged electrophoresis in denaturing polyacrylamide gels (9). Analytical gels were stained with silver
(9).
Citrus exocortis viroid (CEV) and Australian grapevine viroid (AGV)
RNAs were purified from tomatoes and cucumbers respectively which had been
inoculated with grapevine RNA 1 (containing RNA la and lb) as described in
Rezaian et^ a_l. (9). RNAs la and lb were purified from grapevines (9) and
separated by prolonged electrophoresis (9).
Fractionation of grapevine extracts using PEG
The virus purification procedure of Lee et^ aj^. (10) was followed except
that IX 2-mercaptoethanol vas included in the homogenization buffer and the
procedure was terminated after the first 41 polyethylene glycol (PEG 6000),
O.BZ NaCl precipitation step.
retained for further analysis.
Both the PEG pellet and the supernatant were
The PEG pellet was resuspended in 10 mM Trie
pH 8, 0.1 mM EDTA, 1Z SDS and extracted with an equal volume of phenol,
twice with an equal volume of chloroform and the nucleic acids were ethanol
precipitated.
tion of 1Z.
SDS was added to the PEG supernatant to a final concentraThe supernatant fraction was then extracted once with an equal
volume of phenol and then twice with an equal volume of chloroform.
Cellu-
lose powder (Whatman CF11) was added to the extracted supernatant (O.lg
CFll/g tissue) and redistilled ethanol was also added to a final concentration of 35Z to bind nucleic acids to the cellulose (11). The mixture was
stirred at room temperature for 30 minutes and then washed 4 times with 35Z
850
Nucleic Acids Research
ethanol/STE (50fflMTris-Cl pH 7, 0.1 M NaCl, 1 mM EDTA).
Centrifugation
(10,000 rpm, Sorvall GSA rotor, 10 minutes, 4°C) was used to pellet the
cellulose at each washing step.
The washed cellulose pellet was dried under
vacuum to remove ethanol and the nucleic acids were eluted by washing the
pellet twice in STE.
The nucleic acids were then collected by ethanol
precipitation.
Direct enzymic sequencing of RNA
Linear RNA fragments were obtained from GYSV molecules by partial RNase
digestion (Tl or D2) under non-denaturing conditions (12). The resultant
fragments were 5'-
P labelled, fractionated on polyacrylamide gels and
sequenced as described by Uaseloff and Symons (12). Sequence data generated
was used for the manufacture of synthetic oligonucleoti.de primers 1,2 and 3
(see below).
Dideoxy RNA sequencing
Some RNA preparations were sequenced by the dideoxynucleotide chain
termination method of Sanger ^t &1^. (13) with the modifications described by
Rezaian et_ al_• (14).
Synthesis and cloning of GYSV cDNA
Double stranded cDNA was prepared from purified circular RNA using two
different methods.
In one method, 250 ng of synthetic oUgodeoxynucleotide
primer 1 (5'-CTCACTCCCCCTCTGCCC-3") was hybridized to 200 ng of the circular
RNA by heating to 100°C for 2 minutes (in a water bath) and allowing cooling
to room temperature.
First and second strand synthesis reactions were then
performed as described by Gubler and Hoffman (15) using RNase H in the synthesis of the second strand.
The second cDNA synthesis method employed two
synthetic oUgodeoxynucleotide primers.
Reactions were performed essential-
ly as described by Visvader and Symons (16) except that 250 ng of RNA and 64
ng of primer 2 (5"-TAGCGGGGGTTCCGGGG-3') were used in the synthesis of the
first strand and 16 ng of primer 3 (5'-AGAGGTCTCCGGATCTT-3') was used for
second strand synthesis.
Por both methods, the cDNA was inserted into the
Sma I site of M13mpl9 and the nucleotide sequence was determined by the
dideoxy nucleotide chain termination method (13).
Dot blot detection of vlrolds
The procedure used for dot blotting was that described by Thomas (17)
except the RNA samples were denatured by heating in 7.5xSSC, 15Z formaldehyde for 15 minutes at 65°C (18). Hybridization probes were made by
transcribing a full length plus sense GYSV clone in M13mpl9 using the Klenow
fragment of DNA polymerase and the universal 17-mer M13 primer (19).
851
Nucleic Acids Research
Northern blot analysis
Nucleic acid samples for northern blotting were electrophoresed through
21 agarose gels containing formaldehyde as described by Maniatis £t_ al.
(19).
The northern transfer and subsequent treatment of the filter was as
described by Thomas (17). Hybridization probes were the same as those used
in dot blots and the hybridization solution contained probe at a concentration of 1x10
cpm/ml of buffer.
Computer analysis
Sequence data was analysed using MTX and MBIS computer packages (20,
21).
These packages were also used for secondary structure analysis.
RESULTS
Electrophoretic properties of GYSV
In a previous study, we determined that the nucleic acids isolated from
a grapevine cultivar infected with yellow speckle disease contained a single
circular RNA after 2D electrophoresis (9). The circular RNA migrated as a
single RNA band after prolonged electrophoresis under denaturing conditions
(fig. 1A; track 1) and was termed for identification purposes, grapevine
yellow speckle viroid (GYSV).
As with the other circular RNAs that we have
I
I I
WUr
! - • • « • . _
A
B
Fig. 1
Conparison of the electrophoretic mobility of GYSV with the other circular
RMAs isolated from grapevines. A shows the electrophoretic mobilities of
grapevine circular RNAs after prolonged electrophoresis under denaturing
conditions (4.81 acrylamide, 8M urea in TBE pH 8.3, 52 C). C and L indicate
circular and linear forms respectively. B shows the relative mobilities of
the circular RNAs after electrophoresis under native conditions (6Z
acrylamide, TBE pH 8.3). Silver staining was used to detect nucleic acids.
852
m
•
os o-a MO-*
a *
T2
Hucleoti.de residues conserved between GYSV and ASSV and their relationship to the viroid domain •odel. The
residues common to GYSV and ASSV are shown relative to their location in the secondary structure of GYSV. The
numbers above the residues show their actual location in GYSV (G) and ASSV (A).
The solid lines represent
regions of no significant sequence homology between the two viroids. Tl and T2 represent terminal domains,
P represents the pathogenicity region, V the variable region and C the central conserved region. The boxed area
in Tl is also shown in figure AA. The area spanned by the arrows indicates sequences which are not conserved
between GYSV and ASSV but in each case contribute to the formation of the stem loop structure depicted in
figure 5A.
FIR.3
T1
Fig. 2
•ucleotlde sequence and proposed secondary structure of GYSV. The sequence shown Is that determined from a full
length cDNA clone of GYSV which was constructed from the circular UNA from grapevine tissue infected with yellow
speckle disease. Solid and broken lines represent regions of sequence which were also obtained by ensymic RNA
sequencing and dideoxy RNA sequencing respectively. Shaded blocks show the location of the synthetic
oligonucleotide primers used for direct RNA sequencing (primer 1) and for cDNA cloning of GYSV (primers 2 and 3).
3 • o*o«*c o» M O * *
^J.^JvJ_4_4J!.^J/X•.IJlXJ_L•I.-J•.•..•J.J.
"TT" ri "'
3"
3
JO
CD
Z
o_
2
o'
Nucleic Acids Research
isolated from grapevines, the yields of GYSV were low, approximately 1 ug of
RNA/kg of leaf tissue (9).
A comparison was made between the electrophoretic mobility of GYSV and
the RNAs we isolated from grapevines in a previous study (9). The RNAs with
which GYSV was compared all have a similar sire of about 370 nucleotides and
include; Australian grapevine viroid (AGV) which replicates in cucumber,
citrus exocortis viroid (CEV) which replicates in tomato and RNAs la and lb
for which herbaceous hosts have not yet been found (9). The five RNAs were
electrophoresed under both native and denaturing conditions.
After pro-
longed electrophoresis under denaturing conditions (fig. 1A), GYSV (track
1), RNA la (track 2), CEV (track 5) and AGV (track 6) had similar electrophoretic mobilities and RNA lb (track 4) migrated faster than all of these
molecules.
Under native conditions (fig. IB), GYSV (track 1), RNA la (track
2) and RNA lb (track 3) had the same electrophoretic mobility, all moving
faster than CEV (track A) and the co-migrating AGV (track 5).
The identical behavior of GYSV and RNA la under both native and
denaturing conditions suggests that the two molecules are similar species
and quite distinct from the other circular RNAs that we have isolated from
grapevines.
The complete nucleotide sequence of GYSV and its proposed secondary
structure
Partial sequencing of purified GYSV by the direct enzymic method (12)
generated three non-overlapping fragments of sequence (fig. 2). However,
the identity of some of the residues in each fragment was not certain
because of regions of cross-banding in the sequencing gels.
Such cross-
banding is often suggestive of sequence heterogeneity in the sample being
analysed.
Data obtained from enzymic sequencing was used to manufacture a primer
(primer 1) for use in dideoxy RNA sequencing.
The sequence information
obtained using this method is shown in figure 2.
To complete the sequence
of GYSV, cDNA clones were constructed from the yellow speckle circular RNA
using primers 2 and 3.
One apparently full-length cDNA clone of GYSV was obtained using the
double-primer method (16). The clone was sequenced and found to contain an
insert of 367 residues.
Overlapping sequence data obtained from the direct
RNA sequencing methods confirmed that the clone was full length.
The
complete nucleotide sequence of GYSV and its proposed secondary structure is
presented in figure 2 which shows that GYSV can potentially form a rod-like
structure with a high degree of base pairing which is characteristic of the
854
Nucleic Acids Research
other known viroids ( 6 ) . In total, 621 of the residues are base paired and
the paired residues consist of 53Z G:C, 26Z A:U and 21Z G:U base pairs ( f i g .
2 ) . The G+C content i s 60Z which i s comparable to that of the other known
viroids ( 6 ) .
The GYSV sequence was analysed for putative translation products in
both the plus and minus strands using AUG as the possible i n i t i a t i o n codon.
No AUG codons are present in the minus strand and the single AUG codon
present in the plus strand (residues 254-256) i s followed by reading frame
potentially encoding a polypeptide of 68 amino acids (not including the
GYSV
ASSV
PSTV
CEV-A
CSV
TASV
TPMV
Consenus
qUUCUGGUUCCUGUGGUU
U3CGGUUCCUGUGGUU
CJUCSUSGUOCCUGUGGUD
CUUGAGGUUCCUGUGGUG
CUUGUGGUUCCUGUGGUG
CUUGAGGUUCCUGUGGUG
17
10
11
11
11
11
13
-
3*
27
28
28
28
28
juliJuCGUUCCUGUGCuU -
31
CUUGNGGUnCCUGUGGUN,
GC
B
GYSV
ASSV
PSTV
CEV-A
CSV
TASV
TPMV
HSV
63
50
51
52
52
50
53
35
GYSV
ASSV
PSTV
TASV
TPMV
299
272
295
296
293
-
AAAGAAGA
AAAGAAAA
AAAGAAAA
AAAGAAAA
AAAGAAAA
CAAGAAAA
AAAGAAAA
AAAGAAAA
-
CUUUUUCU
CUUUUUCU
CUUUUUCU
CUUUUUCU
CUUUUUCU
-
70
57
58
59
59
57
60
42
-
306
279
302
303
300
Fig.*
Residues which are conserved between GYSV, ASSV and member* of the PSTV
group. A Conserved residues located in the Tl domain. The boxed regions
show sequences which are perfectly conserved amongst the listed viroids. N
represents any nucleotide. B Residues conserved in the A-rich region of the
P domain. C Conserved residues in the U-rich region of the P domain. Viroid
abbreviations: PSTV, potato spindle tuber viroid; CEV-A, citrus ezocortis
viroid isolate A; CSV, chrysanthemum stunt viroid; TASV, tomato apical stunt
viroid; TPMV, tomato planta macho viroid.
855
Nucleic Acids Research
© ©
crtv
AC
B
V
-11.1 kcal
© ©
ASSV
AC
- I I I kcal
S G A G U G A G C C U C G U C G U C G A C G A A G G G G U G C A C U C C
C C U C A C G U C G G G A A G C A G C U G C U G C U C C G A O U G A G G
.,
112
I
1
17
A G -II.4 k*»l
Fig. 5
Structures formed In the C region of GTSV «od ASSV. A Stem loop structures
possible for the C region of ASSV and GYSV. Circled residues are perfectly
conserved between ASSV and GYSV. B Pallndromlc duplex structure which could
be formed In multlmeric GYSV molecules. The boxed region shows the central
conserved residues and the arrow shows the centre of the palindrome. (Free
energies were calculated as described in ref. 31).
termination codon).
The significance of this putative translation product
is not known.
GYSV conforms to the viroid domain model
The sequence of GYSV was compared with the known sequences of viroids
and circular RNA satellites.
the satellite RNAs or ASBV.
GYSV has no significant homology with any of
However, GYSV has 37Z sequence homology with
the recently described ASSV (8). The homologous residues in GYSV and ASSV
are not randomly distributed but occur as blocks of base paired residues in
the secondary structures of both viroids (fig. 3). There is also some
homology between GYSV, ASSV and members of the PSTV group which is limited
to three blocks of residues shown in figure A.
The sequence comparisons between GYSV, ASSV and the PSTV group show
that GYSV conforms to the viroid domain model proposed by Keese and Symons
for members of the PSTV group (22). In the model, the five domains are; Tl
and T2 (left and right terminal domains which are considered to be inter-
856
Nucleic Acids Research
changeable between viroids), P (a region which has been correlated with
viroid pathogenicity; 23,24), C (a conserved central region thought to be
important in viroid processing and replication; 25) and V (a region of high
sequence variability).
The domains are indicated in the proposed secondary
structure of GYSV (fig. 3) and the evidence for the existence of each domain
is described below.
Tl and T2 domains
There is a stretch of 17 residues that exists in the Tl
region of the PSTV group which is also found in a similar position in the
proposed secondary structure of GYSV (fig. 4A). No sequence in the proposed
T2 region of GYSV is shared with viroids of the PSTV group but the region is
defined by stretches of sequence ccramon to GYSV and ASSV (fig. 3).
P reRlon
Characteristics of the PSTV group P domain include a long adenine
dominated oligopurine sequence of 15 to 17 residues on one strand and an
oligo (U
) sequence on the opposite strand (6). There is a 19 residue
oligopurine sequence in GYSV located between residues 61 and 90 (fig. 2)
with an adenine rich portion lying between residues 63 and 73.
The homology
between the adenine rich portion of GYSV, ASSV and members of the PSTV group
is shown in figure 4B.
In the proposed secondary structure of GYSV the
adenine rich sequence base pairs strongly with an oligo 0 sequence in the
opposite strand (residues 299-309; fig. 2). This structural feature is also
present in ASSV (8). A portion of the U rich sequence found in the GYSV is
also present in ASSV and members of the PSTV group (fig. 4C).
C region
GYSV, like ASSV (8), does not contain the C region sequence common
to all of the other viroids in the PSTV group.
However, 16 residues are
conserved between GYSV and ASSV in the structural location of the PSTV
central conserved region (fig. 3). These sequences are potentially involved
in the formation of two structures which resemble those in the C region of
the PSTV group (6,22).
The upper portion of the C region of GYSV and ASSV can potentially
assume a stem loop conformation with the 16 perfectly conserved bases
capping the top of the structure (fig. 5A). Residues flanking the perfectly
conserved sequences (long arrows in fig. 3) contribute to the stem in both
ASSV and GYSV.
A remarkably similar structure can also be formed in the C
region of all the viroids in the PSTV group (22). It has been postulated
that a stem loop structure is involved in the transition between the native
viroid structure and structures important in viroid replication (22).
In GYSV, 36 residues are involved in the formation of the stem loop
structure (fig. 5A). These 36 residues also have the potential to form a
stable palindromic duplex (fig 5B) with another GYSV molecule in linear
857
Nucleic Acids Research
GYSV
•
CEV
•
HSV
•
AGV
•
1a
•#
1b
•#
Fig. 6
Dot blot analysis of grapevine circular BMAs. 1 ng quantities of the
different circular RNAs we have isolated from grapevines were spotted onto a
nitrocellulose filter and probed with the GYSV specific probe.
form.
Alternatively, the palindromic duplex structure could be formed
within multimeric GYSV molecules.
The 16 residues perfectly conserved
between GYSV and ASSV form the central core of the duplex (fig. 5B). ASSV
also has the potential to form the palindromic structure (8). Similar
structures have been postulated for oligomers of all the viroids in the PSTV
group (26). Diener has proposed that the stable duplex may be involved in
the processing of oligomeric viroid intermediates into monomeric progeny
(26).
In the light of the above observations, we propose that the centrally
located sequences conserved in GYSV and ASSV also represent a central conserved region which is unique to these two viroids.
V region
In most of the members of the PSTV group, the V region lies
between residues 119 and 151 and is characterized by an oligopurine:
oligopyrimidine helix with a minifi"1" of 3 G:C base pairs (6). In the
proposed secondary structure of GYSV, the region between residues 122 and
145 has short stretches of oligopurineioligopyrimidine helices, however, a
of 2 G:C base pairs is observed (fig. 2 ) . In order to clearly
858
Nucleic Acids Research
r
i
r _
«*
-5S
I
I
25 60
I
I
I
I
I
I
I
I
I
I
I
50 62
I
35
28
25
14
6
11
39
43
31
40
A
(pg)
B
FlK-7
The detection of GYSV in fractionated tisane extracts and different
grapevine cultivars by northern blotting. A. Nucleic acids were extracted
from 10 g of CF/YS using the virus extraction procedure described in the
methods. All of the nucleic acid recovered from both the pellet and the
supernatant was loaded onto a 21 agarose/formaldehyde gel. After northern
blotting, the filter was probed with a
P-DNA probe, specific for the
detection of GYSV. 0-origin, XC-xylene cyanol and 5S-position of 5S RNA.
B. Nucleic acids were extracted from different grapevines of known disease
status and the nucleic acid equivalent of lg of tissue was loaded per track.
After transfer, the filter was probed with the GYSV probe. The grapevine
cultivars assayed, together with their disease status in brackets are listed
at the top of the autoradiograph. Abbreviations: CF-Cabernet franc,
CS-Cabernet sauvignon, H-healthy, YS-yellow speckle, SM-summer mottle,
LR-leafroll, F-fleck, *-currently being indexed. The numbers at the bottom
of the autoradiograph represent the amount of total nucleic acid that was
loaded per track.
define the end of the V region (and hence the beginning of the T2 domain)
other isolates of GYSV need to be sequenced.
GYSV has sequence homology with RNA lb
Dot blot hybridizations were performed to test the degree of relatedness between GYSV and the other circular RNAs we have isolated from grapevine tissues.
The
P-probe, specific for the detection of GYSV did not
hybridize with CEV, HSV or AGV (fig. 6), confirming that these molecules are
distinct from GYSV.
As expected from the electrophorettc mobility data
(fig. 1 ) , the GYSV probe did hybridize to the RNA la sample suggesting RNA
la contains GYSV related species (fig. 6 ) . The GYSV probe also hybridized
to the RNA lb sample but the degree of hybridization was less than that
observed for RNA la.
Thus, RNA lb is also related to GYSV (fig. 6).
859
Nucleic Acids Research
GYSV is not an encapsidated molecule
One characteristic of viroids is that they are infectious molecules
thus, they can initiate replication when inoculated onto suitable plant
hosts.
We have not been able to observe the replication of GYSV in either
cucumber or tomato (results not shown) which we have used to propogate the
other viroids isolated from grapevine tissue (9). We have also inoculated
GYSV onto grapevine seedlings.
In the first trial, GYSV was not present
when the tissue was sampled 4 weeks after inoculation.
However, it is well
known that the detection of some viroids in woody species may require
several months as is the case of CEV in certain citrus species, CCCV in
coconut palm and ASBV in avocado trees (27). Further infectivity studies of
GYSV in grapevine seedlings are currently in progress.
In the absence of infectivity data, we also examined the possibility
that GYSV was an encapsidated viral satellite molecule which was being
released during the extraction procedure.
To test this possibility, GYSV
was purified from yellow speckle infected tissue using a virus purification
procedure which has been successful in isolating elongated and spherical
virus particles from grapevines (10). GYSV could not be detected in the
fraction which would normally contain the viral pellet (PEG pellet, fig.
7A).
GYSV was, however, detected in the supernatant fraction (fig. 7A) .
This observation suggests that GYSV is not a plant virus satellite and
strengthens the case for the viroid nature of GYSV.
The relationship between the presence of GYSV and the Incidence of yellow
speckle disease
Even though GYSV was isolated from tissue infected with yellow speckle
disease, there was no direct evidence to suggest that the presence of GYSV
correlated with the incidence of yellow speckle disease in field grown
tissue.
A number of grapevine cultivars of known disease status were tested
for the presence or absence of GYSV.
shown in figure 7B.
The results of the experiment are
All cultivars which indexed positive for yellow speckle
disease contained GYSV and healthy cultivars were free of GYSV.
Seedlings
of yellow speckle infected mission grapevines did not contain GYSV
suggesting it is not seed transmissible (fig. 7B). However, three sources
which did not index positive for yellow speckle disease hybridised to the
GYSV probe (CF/SM, CF/CS SA125, CF/R.du Lot; fig. 7B). The hybridization
signals in these sources were much weaker than those obtained from cultivars
which indexed positive for yellow speckle disease.
There are at least three explanations for the apparently anomolous
860
Nucleic Acids Research
Table 1
Clone
Sequence heterogeneity in GTSV
cXHA clones.
Clone
Residue
Change*
location
position
in GYSV
in GYSV
1
338-94
75
A-»C
4
314-94
4
C-frG
7
346-94
75
A-frC
13
18-94
20
62
64
66
75
+G
+A
A-*G
62
64
66
75
322
+A
A-K;
G-*A
A-»-C
U-*A
20
322-94
G-*A
A-*C
26
346-94
350
G-HJ
29
366-94
2
3
6
7
16
17
366
367
C-^G
U-^C
G-+-C
G-»-U
G-*C
C-*U
G-MJ
30
290-94
20
41
62
64
66
75
315
G-*A
A-*C
+UCU
15
322
U-^C
U-»A
35
307-94
+G
+C
+A
A-M;
* The position of an insertion follows
the indicated GYSV residue position.
Primer 1 (fig. 2) was used to generate
these cDNA clones.
results.
Firstly, the sources may contain quantities of RHA lb which we
have shown is related to GYSV (fig. 6). Secondly, the cultivars may possess
a variant of GYSV which induces extremely mild symptoms.
Thirdly, the vines
may have contracted yellow speckle disease since the indexing trials were
861
Nucleic Acids Research
conducted.
The first possibility is currently under investigation.
Sequence variants of GYSV
In the initial cDNA cloning of GYSV, three, partial length cDNA clones
were also obtained.
The sequence of one of these partial length clones
containing 178 residues was identical to GYSV.
The other two clones, 1 and
4 differed from the GYSV sequence by single residue changes (table 1) which
suggested that the RNA population was heterogenous.
To further investigate
the degree of sequence diversity, cDNA clones were constructed from an RNA
sample pooled from four of the grapevine sources shown to contain GYSV by
northern analysis (CF/YS, CF/DIH, Kyoho/Dogridge, CF/SM; fig. 7B). Twenty
of the cDNA clones obtained were sequenced.
None of the sequenced clones
were full length and thirteen of the clones, ranging from 83 to 178 residues
in length were identical in sequence to GYSV (fig. 3 ) . The remaining seven
cDNA clones varied in sequence from GYSV and the residue differences are
shown in table 1.
These results confirm that GYSV is representative of a
mixture of closely related molecules.
DISCDSSIOM
Th» viroid nature of GYSV
The structural and biological evidence we have accumulated suggests
that the circular RNA we have isolated from yellow speckle infected grapevines is a viroid.
GYSV is a circular RNA molecule, in the viroid sire range which has the
potential to assume a highly base paired rod-like structure (fig. 2 ) . GYSV
is probably not seed transmissible (fig. 7B) and does not appear to be encapsidated in a viral particle (fig. 7A). GYSV is related in sequence to a
newly characterized viroid, ASSV (fig. 3), and in common with ASSV shares
sequence homologies with members of tha PS TV group (fig. 4 ) . The sequence
of GYSV has characteristics which conform to the viroid domain model (22;
fig. 3 ) . All of these structural features strongly suggest that GYSV is a
viroid.
Yellow speckle disease itself has a viroid-like etiology.
For example,
virolds are known to replicate more rapidly at higher temperatures (28) and
a higher level of yellow speckle symptom expression is routinely observed in
the field at times of elevated temperature (2). We have observed that
higher yields of GYSV can be obtained from tissue sampled in the hotter
months of summer and also from tissue grown in cabinets at 30 C (9).
Further evidence of the heat stable, viroid-like nature of the yellow
speckle agent is that the disease cannot be eliminated either by heat
862
Nucleic Acids Research
therapy (2,29) or fragmented shoot apex culture at 35°C (4) which are known
to free grapevines from viral infection* (A,29).
Thus, the biological properties of yellow speckle disease point to the
involvement of a viroid agent and GYSV strongly correlates with the
expression of yellow speckle disease in grapevines (fig. 7B). However, we
cannot define GYSV as the causative agent of yellow speckle disease until
infectivity studies have been performed on healthy grapevines and Koch's
postulates have been fulfilled.
Sequence variants of GYSV
There is extreme variability in the intensity of yellow speckle symptom
expression which relates to the grapevine cultivar, plant age and environmental factors.
An example of this are cultivars which show distinct yellow
speckle symptoms in Australia, yet are symptomless in California (1). If
GYSV is indeed the causative agent of yellow speckle, the existence of a
number of GYSV sequence variants (table 1) may have some bearing on the
erratic symptom expression of yellow speckle disease.
In a recent study by Semancik et_ al_. (30), a viroid-like RNA was isolated from a variety of grapevine cultivars which had electrophoretic properties similar to GYSV.
The RNA, named GVI, migrated faster than CEV in a
native gel system, yet co-migrated with CEV under denaturing conditions.
In
the wider survey conducted in that study, no definitive correlation could be
made between GVI and any of the well known grapevine diseases.
fact, found in a number of the healthy cultivars.
GVI was, in
It would be of interest
to establish whether GYSV and GVI are related molecules.
If so, the wide
distribution of GVI might be explained by the presence of mild variants
which could be difficult to detect under Californian field conditions.
GYSV a member of a new viroid group
When the sequence of ASSV was reported it was proposed that it should
be placed into a separate viroid group (8). In this paper, the characterization of GYSV with its structural similaritie* to ASSV strengthens the
existence of another viroid group in addition to the PSTV and ASBV groups.
We suggest that this third viroid group should be represented by ASSV.
We
propose that a distinctive feature of the ASSV group is a unique central
conserved region which is unrelated in sequence to the conserved region of
the PSTV group yet has the potential to assume all of the structures which
are thought to be important in viroid replication and processing.
ACPtCMLKlJGEMHiTS
We would like to thank Professor R.H. Symons for useful discussions. Dr. W.
863
Nucleic Acids Research
Bottomley for manufacturing the synthetic primers, Les Krake and Susan
Johnson for excellent technical assistance.
REFHIEHCES
1. Taylor, R.H. and Woodham, R.C. (1972) Aust. J. agric. Res. 23, 447-452.
2. Woodham, R . C , Taylor, R.H. and Krake, L.R. (1973). Australia. Riv.
Patol. Veg., Ser. IV, 9 (Suppl.) 17-22.
3. Bovey, R., Gartel, W. , Hewitt, W.B., Martelli, G. and Vuittenez, A.
(1980). In Virus and Virus-like diseases of grapevines, pp. 54. Payot,
Lausanne.
4. Barlass, M. , Skene, K.G.M., Woodham, R.C. and Krake, L.R. (1982). Ann.
appl. Biol. 101, 291-295.
5. Flores, R., Duran-Vil», N., Pallas, V. and Semancik, J.S. (1985). J.
Gen. Virol. 66, 2095-3102.
6. Keese, P. and Symons, R.H. (1987). In Viroids and viroid-like pathogens.
(Semancik, J.S. Ed.) pp. 1-47. Boca Raton, FLsCRC Press.
7. Sanger, H.L. (1984). In The Microbe: Part I, Viruses. (Many, B.W.J. and
Pattison, J.R., Eds.) Society for General Microbiology Symposium 36,
291-334. Cambridge University Press. Cambridge.
8. Hashimoto, J. and Koganezawa, H. (1987). Nucleic Acids Res. 15, 7045-7052.
9. Rezaian, M.A., Koltunov, A.M. and Krake, L.R. (1987). J. Gen. Virol,
(in preis).
10. Lee, R.F., Garnsey, S.M., Brlansky, R.H. and Goheen, A.C. (1987).
Phytopathology 77, 543-549.
11. Franklin, R.M. (1966). Proc. Natl. Acad. Sci. USA 55, 1504-1511.
12. Haseloff, J. and Symons, R.H. (1981). Nucleic Acids Res. 9, 2741-2752.
13. Sanger, F., Nicklen, S. and Coulson, A.R. (1977). Proc. N»tl. Acad. Sci.
USA 74, 5463-5467.
14. Rezaian, M.A., Williams, R.H.V., Gordon, K.H.J., Gould, A.R. and Symons,
R.H. (1984). Eur. J. Biochem. 143, 277-284.
15. Gubler, U. and Hoffman, B.J. (1983). Gene 25, 263-269.
16. Visvader, J.E. and Symons, R.H. (1985). Nucleic Acids Res. 13, 29072920.
17. Thomas, P.S. (1980). Proc. Natl. Acad. Sci. USA 77, 5201-5205.
18. White, B.A. and Bancroft, F.C. (1982). J. Biol. Chem. 257, 8569-8572.
19. Maniati8, T., Fritsch, E.F. and Sambrook, J. (1982). Molecular Cloning A Laboratory Manual. Cold Spring Harbor Laboratory. New York.
20. Reisner, A.H. and Bucholtz, C.A. (1986). Nucleic Acids Res. 14, 233-238.
21. Bucholtz, C.A. and Reisner, A.H. (1986). Nucleic Acids Res. 14, 265-272.
22. Keese, P. and Symons, R.H. (1985). Proc. Natl. Acad. Sci. USA. 82, 45824586.
23. Visvader, J.E. and Symons, R.H. (1983). Virology 130, 232-237.
24. Sanger, H.L. (1984). Soc. Gen. Microbiol. Symp. 36, 281-334.
25. Symons, R.H., Haseloff, J., Visvader, J.E., Keese, P., Murphy, P.J.,
Gill, D.S. and Bruening, G. (1985) In Subviral pathogens of plants and
animals: viroids and Prion. (Maramorosch, K. and McKelvey, J.J. Eds.).
26. Diener, T.O. (1986). Proc. Natl. Acad. Sci. USA 83, 59-62.
27. Diener, T.O. (1979). Viroids and Viroid Diseases. Wiley and Sons. New
York.
28. Sanger, H.L. and Ramm, K. (1975). In Modification of the information
content of plant cells. (Markham, R., Davies, D.R., Hopwood, D.A. and
Borne, R.W., Eds.) pp. 229-252. Amsterdam. North Holland Publishing Co.
29. Goheen, A.C. and Luhn, C.F. (1973). Riv. Pat. Veg. Series IV:9, 287-289.
30. Semancik, J.S., Rivera-Buatamante, R. and Goheen, A.C. (1987). Am. J.
Enol. Vitic. 38, 35-40.
31. Tinoco, I., Borer, P.N., Dengler, B. and Levine, M.P. (1973). Nature
Nev Biol. 246, 40-41.
864