Journal of General Virology (2006), 87, 705714 DOI 10.1099/vir.0.

81325-0

Cytological analysis of Saccharomyces cerevisiae

cells supporting cymbidium ringspot virus defective
interfering RNA replication
Beatriz Navarro, Marcello Russo, Vitantonio Pantaleo and Luisa Rubino
Correspondence Istituto di Virologia Vegetale del CNR, Sezione di Bari, c/o Dipartimento di Protezione delle
Luisa Rubino Piante e Microbiologia Applicata, Universita degli Studi, Bari, Italy
l.rubino@ba.ivv.cnr.it

Received 8 July 2005
Accepted 11 November 2005
The replicase proteins p33 and p92 of Cymbidium ringspot virus (CymRSV) were found to support
the replication of defective interfering (DI) RNA in Saccharomyces cerevisiae cells. Two yeast
strains were used, differing in the biogenesis of peroxisomes, the organelles supplying the
membranous vesicular environment in which CymRSV RNA replication takes place in infected plant
cells. Double-labelled immunofluorescence showed that both p33 and p92 replicase proteins
localized to peroxisomes, independently of one another and of the presence of the replication
template. It is suggested that these proteins are sorted initially from the cytosol to the endoplasmic
reticulum and then to peroxisomes. However, only the expression of p33, but not p92, increased the
number of peroxisomes and induced membrane proliferation. DI RNA replication occurred in
yeast cells, as demonstrated by the presence of monomers and dimers of positive and negative
polarities. Labelling with BrUTP showed that peroxisomes were the sites of nascent viral synthesis,
whereas in situ hybridization indicated that DI RNA progeny were diffused throughout the
cytoplasm. DI RNA replication also took place in yeast cells devoid of peroxisomes. It is suggested
that replication in these cells was targeted to the endoplasmic reticulum.

INTRODUCTION and a readthrough product of 9295 kDa (ORF2). These

Replication of positive-strand RNA viruses takes place in
association with host-cell membranes, but different viruses
recruit diverse intracellular membranes for the assembly
of their replication complexes. Study of the targeting signals
of viral proteins interacting with host-cell membranes, as
well as details of the formation, maintenance and func-
tioning of the virus replication complex, is of importance for
comprehension of the virus life cycle and, therefore, for
envisaging a possible antiviral therapy. Positive-strand RNA
viruses constitute a large group of viruses, including agents
of important plant, animal and human diseases.
The basic similarity in the replication strategy of positive-
strand RNA viruses has encouraged the use of any simple
virus of this group as a model to analyse the factors involved
in replication. Members of the plant virus genus Tombusvirus
(family Tombusviridae) have proved to be good models
for studying virushost interactions (Russo et al., 1994;
White & Nagy, 2004). Tombusviruses contain a positive-
sense, single-stranded RNA genome of ~4?8 kb with five
open reading frames (ORFs). The genomic RNA acts as
mRNA for translation of a protein of 3336 kDa from ORF1
Supplementary figures showing immunofluorescent analysis of UTL-7A
yeast cells transformed with plasmids expressing CymRSV proteins are
available in JGV Online.
two proteins constitute the viral replicase and are the only
viral products essential for replication. The 41 kDa protein
(ORF3) is translated from a subgenomic RNA. A second
subgenomic RNA is synthesized for translation of two nested
ORFs (4 and 5) encoding proteins of 22 and 19 kDa (p19),
required for cell-to-cell movement and systemic invasion,
respectively. In addition, p19 has been shown to act as a
suppressor of RNA silencing (Voinnet et al., 1999). Infections
caused by tombusviruses are normally associated with small
defective interfering (DI) RNAs that are deletion mutants of
the viral genome. DI RNAs do not encode any proteins and
depend for replication on the viral replicase supplied in trans
by the viral genome. However, as DI RNA molecules contain
all cis-acting elements required for replication, they repre-
sent a useful model for studying tombusvirus replication
(Russo et al., 1994; White & Nagy, 2004).
Cymbidium ringspot virus (CymRSV) and Carnation Italian
ringspot virus (CIRV) are two well-characterized tombus-
viruses. Their differential trait resides in the targeting signal
present in the 33 kDa (p33; CymRSV) or 36 kDa (p36;
CIRV) protein, which addresses the replication complex
to the peroxisomal membrane (CymRSV) or to the outer
mitochondrial membrane (CIRV) (Burgyan et al., 1996).
Both targeting signals are characterized by two transmem-
brane domains and accompanying short sequences respon-
sible for the specificity of the target (Rubino & Russo, 1998;

0008-1325 G 2006 SGM Printed in Great Britain 705

B. Navarro and others
Weber-Lotfi et al., 2002; Navarro et al., 2004). Targeted orga-
nelles are transformed in conspicuous structures (multi-
vesicular bodies) where the limiting membrane is stimulated
to abnormal growth, accounting for the formation of flask-
shaped vesicles open to the cytoplasm in which the viral
RNA replication takes place (Russo et al., 1983, 1987; Di
Franco et al., 1984). Interestingly, by exchanging only
the genome segments containing the targeting signals, the
site of the replication complex becomes peroxisomal for
CIRV and mitochondrial for CymRSV (Burgyan et al., 1996;
Rubino & Russo, 1998).
Replication of tombusviruses can be studied in Saccharomyces
cerevisiae cells by using DI RNA molecules as the replication
template and replicase proteins derived from CIRV (Pantaleo
et al., 2003) or Cucumber necrosis virus (CNV; Panavas &
Nagy, 2003). Analysis of yeast cells in which the replication of
DI RNA is supported by CIRV replicase proteins has shown
that the membrane-associated complex, in which the viral
replicase proteins accumulate and progeny DI RNA is synthe-
sized, derives from mitochondria (Pantaleo et al., 2003, 2004).
Following previous analysis of the expression of CymRSV
p33 in S. cerevisiae (Navarro et al., 2004), the present study
has identified and characterized the replication complex in
yeast cells where DI RNA is synthesized following expres-
sion of CymRSV p33 and p92 replicase proteins. The way in
which the replicase proteins may be sorted to the replication
site, whether directly from the cytosol or indirectly via the
endoplasmic reticulum (ER), is also discussed.
METHODS
Plasmids. To express the CymRSV p33 protein, the coding sequence
was amplified by PCR from a CymRSV full-length infectious clone
and cloned between the oleate-inducible catalase A gene (CTA1)
promoter and the terminator of vector pEL26, a 2 mm plasmid
containing URA3 as a marker (Elgersma et al., 1996; Navarro et al.,
2004) to obtain clone pEL26-33K.
The p92-coding sequence was amplified by PCR from the same
full-length infectious clone or from a clone in which the stop codon
of p33 was mutated to a tyrosine codon and inserted between the
alcohol dehydrogenase gene (ADH1) promoter and the terminator of
vector YE, a 2 mm plasmid containing LEU2 as a selectable marker
(Pantaleo et al., 2003). This process yielded clones YE92Kwt and
YE92Ktyr.
Plasmids pEL26-Myc33K and YEHA92K contained the same CymRSV
p33 and p92 proteins, tagged at their 59 termini with Myc (EQKLISEEDL;
Evan et al., 1985) and haemagglutinin (HA) (MYPYDVPDYAG;
Kolodziej & Young, 1991), respectively. Plasmids pEL26-Myc92Kwt
and pEL26-Myc92Ktyr contained the wild-type or tyrosine Myc-tagged
version of p92 cloned in the vector pEL26.
The CymRSV DI RNA sequence (DI-3, 481 nt; Burgyan et al., 1992)
was used as replication template. It was cloned between the galactose-
inducible (GAL1) promoter and the ADH1 terminator into the low-
copy-number centromeric plasmid vector pBMI3S containing TRP1 as
a selectable marker (Ishikawa et al., 1997; Pantaleo et al., 2003). As a
negative control of replication, a mutated version of DI-3 was used, in
which the G residue at position 24 from the 39 end was mutated to A
(DI RNADG) (Pantaleo et al., 2003).
706
Yeast strains and growth conditions. The yeast strains used
were UTL-7A (MATa leu2-3,112 ura3-52 trp1) (Erdmann et al., 1989)
and UTL-7Apex19D (same as UTL-7A but pex19D : : LEU2 (Gotte
et al., 1998). Yeasts were transformed by using the lithium acetate/
polyethylene glycol method (Ito et al., 1983), plated on selective
medium (SM) containing 0?67 % yeast nitrogen base supplemented
with appropriate amino acids and 2 % glucose, and incubated at
30 uC for 23 days. UTL-7A and UTL-7Apex19D transformants were
initially grown at 30 uC in an oleate-based medium (SYO) as de-
scribed previously (Navarro et al., 2004), then at 26 uC in SYO con-
taining 1 % galactose to induce DI RNA transcription and finally in
the same medium without galactose.
Immunofluorescence. Rabbit polyclonal antisera to yeast peroxi-
somal Fox3p, ER Kar2p and Golgi Emp47p protein markers
were kindly provided by B. Distel (University of Amsterdam, The
Netherlands), M. Rose (Princeton University, USA) and H. D. Schmitt
(Max-Planck-Institut fur biophysikalische Chemie, Germany), respec-
tively. Mouse monoclonal antibody (mAb) against the mitochondrial
marker CoxIII was from Molecular Probes. mAbs against the Myc
epitope, and mAbs and rabbit polyclonal antibodies against the HA
epitope and the mitochondrial outer-membrane marker Tom40p
were from Santa Cruz Biotechnology. Rhodamine-conjugated goat
anti-rabbit and Alexa 488-conjugated anti-mouse secondary antibodies
were from Molecular Probes. Fixation of yeast cells and immuno-
fluorescent staining were done according to Redding et al. (1991).
Detection of nascent and progeny RNA. Semi-intact cells were
prepared according to Schlenstedt et al. (1993). Labelling of newly
synthesized viral RNA by BrUTP incorporation and in situ hybridi-
zation were done as described by Restrepo-Hartwig & Ahlquist
(1999). Cells were then processed for immunofluorescence as above.
Antibodies against BrUTP and digoxigenin (DIG) were from Sigma
and Roche, respectively.
Protein and RNA analysis. Proteins were extracted and examined
by Western blotting as described previously (Pantaleo et al., 2003;
Navarro et al., 2004). Total RNA was extracted by using acid phenol
(Leeds et al., 1991). One microgram of RNA was denatured with
formamide and formaldehyde, run on a formaldehyde-permeated
agarose gel, transferred to a nylon membrane and hybridized with
DIG-labelled probes consisting of the last 300 nt of DI-3 in positive
and negative orientations.
Microscopy. Fluorescent cells were viewed and photographed by
using an Axioplan 2 imaging fluorescence microscope (Zeiss). Image
acquisition was done with an ApoTome imaging system, allowing
the generation of optical sections through fluorescent samples.
Images were taken with a CCD camera (AxioCam) and the informa-
tion was visualized with the software package AxioVision (Zeiss).
For electron microscopy, cells were fixed with 2 % glutaraldehyde in
0?1 M cacodylate buffer (pH 7?2), washed extensively with the same
buffer, treated briefly with lyticase and post-fixed with 4 % potas-
sium permanganate. Cells were then bulk stained with uranyl acet-
ate, dehydrated with ethanol and embedded in Spurrs resin.
RESULTS
Replication of CymRSV DI RNA in yeast strain
UTL-7A
Replicase protein p33 was expressed in yeast strain UTL-7A
with the plasmid pEL26, under the control of the oleate-
inducible CTA1 promoter (pEL26-33K; Navarro et al., 2004).
Using this plasmid, expression and activity of p33 could
then be investigated under growth conditions favourable to
Journal of General Virology 87

biogenesis of peroxisomes (Elgersma et al., 1996). The repli-
case protein p92 was expressed with the plasmid YE92K under
the control of the constitutive promoter ADH1. Cells were
transformed with plasmids expressing the replicase proteins
and DI-3 RNA (clone pBDI-3) in different combinations.
Transcription of DI-3 RNA was induced first by growing
yeasts in SYO containing galactose. An aliquot of this culture
was then subinoculated into SYO without galactose and the
rest was pelleted and frozen. After 24 h further growth, RNA
was extracted from all cultures. Northern blot analysis of
RNA extracts showed the presence of positive-strand DI-3
RNA only in samples in which both p33 and p92 replicase
proteins were expressed (Fig. 1a, lanes 8 and 16). No RNA
band was detected if one or both replicase proteins were
absent (Fig. 1a, lanes 27 and 1015) or if the putative repli-
cation template was DI-3 RNADG with the GRA muta-
tion at position 24 from the 39 end (Fig. 1a, lanes 1 and
9). Above the DI-3 RNA monomer, a minor band corres-
ponding to the size of the DI-3 dimer was detected (Fig. 1a,
lane 16).
http://vir.sgmjournals.org
CymRSV DI RNA replication in yeasts
Clones containing both replicase proteins and non-replicable
DI RNADG or replicable DI RNA were analysed for the pres-
ence of negative-strand DI RNA. Two bands corresponding
to DI-3 RNA monomers and dimers were detected only for
the clone expressing p33, p92 and DI RNA (Fig. 1b, lanes 1
and 2).
To test for the possibility of expressing both p33 and p92
proteins from a single clone, the wild-type p92 sequence
containing the p33 amber stop codon was cloned into
plasmid YE (YE92Kwt) and transformed into yeast cells
together with plasmid pBDI-3. DI-3 RNA replicated in these
cells, indicating that the stop codon of p33 was read through,
as in infected plants, producing a functional p92 (Fig. 1c,
lane 1). However, the level of replication was lower than
that in cells in which p33 and p92 were supplied from sepa-
rate plasmids (Fig. 1c, lane 3). These results were in agree-
ment with those of Panavas & Nagy (2003), who showed that
a readthrough mechanism exists in yeast, but contrasted with
those reported for CIRV replicase proteins in strain YPH499
(Pantaleo et al., 2003). Therefore, the experiments with CIRV
Fig. 1. Replication of DI-3 RNA in yeast
strain UTL-7A expressing the CymRSV repli-
case proteins p33 and p92. (a, b) Northern
blot analysis of RNA extracts from cells trans-
formed with plasmids expressing replicable
(+) or non-replicable (+*) DI RNA or empty
vector () and expressing (+) or not () p33
and p92 replicase proteins. Cells were grown
in the presence (a, lanes 18; b, lanes 1 and
3) or absence (a, lanes 916; b, lanes 2 and
4) of galactose. Blots in (a) and (b) were
probed to detect positive- and negative-
strand DI RNA sequences, respectively. (c)
Northern blot analysis of positive-strand RNA
extracted from yeast cells expressing replic-
able DI RNA and wild-type p92 (p92wt; lane
1), the p92 mutant with the tyrosine codon
substituted for the p33 stop codon (p92tyr;
lane 2) or both p33 and p92tyr (lane 3). (d)
Northern blot analysis of RNA extracts from
yeast cells expressing (+) (lane 1) or not
expressing () (lane 2) Myc-tagged p33 and
HA-tagged p92 together with DI RNA. Blots
in (c) and (d) were probed to detect positive-
strand DI RNA sequences. The positions of
dimeric (dim) and monomeric (mon) DI RNA
molecules in (ad) are indicated. (e, f)
Western blot analysis of extracts from cells
expressing (+) or not () p33 and p92 native
(e) or epitope-tagged (f) proteins. (e) was
probed with an anti-p33 antiserum and (f)
with mixed antibodies to the Myc and HA
epitopes. The positions of p33 and p92 are
indicated. Molecular size markers (kDa) are
shown on the left.
707
B. Navarro and others

were repeated to test the possibility that DI RNA replication
could also be sustained by the CIRV replicase proteins p36
and p95 expressed by a plasmid carrying the wild-type p36/
p95 gene alone. It was found that the readthrough product
p95 was formed correctly along with p36, although below
the limit of detection in Western blots, as it supported the
replication of DI RNA (M. Russo & L. Rubino, unpublished
results).
p33 and p92 co-localize to peroxisomes
Expression and localization of p33 in yeast strain UTL-7A,
grown in oleate-containing medium, were extensively
investigated in a previous work, where it was found that
p33 accumulated in one or two large bodies consisting of
aggregates of peroxisomes mixed with membranes and
mitochondria (Navarro et al., 2004). These observations
were repeated and extended here to compare the cytological
effect of p33 with that of the sole protein p92 or of both
proteins together, in the presence or absence of DI RNA.
Peroxisomes, mitochondria, ER and Golgi were identified
by using the immunological markers Fox3p, CoxIIIp, Kar2p
and Emp47p, respectively. The distribution of markers of
mitochondria, ER and Golgi in cells of all transformants was
indistinguishable from that of control cells (see Supple-
mentary Fig. S1a, available in JGV Online) and is shown
only for the transformant competent for DI RNA replication
(see Supplementary Fig. S1b, available in JGV Online).
Mitochondria were visualized as a branched tubular network
near the cortex of the cell. The ER appeared mainly as the
perinuclear membrane with some extensions into the cyto-
plasm up to the periphery of the cell. Finally, the Golgi
apparatus had a punctate cytoplasmic pattern. As for per-
oxisomes, the cytology of transformants showed essentially
two different patterns. Cells expressing p33 alone (Fig. 2c, e)
or together with p92 (Fig. 2a, b, h) showed structures
corresponding to peroxisomes consisting of one or two
prominent bodies occupying a large part of the cell, whereas
the appearance and distribution of these organelles in cells
expressing p92 only (Fig. 2d, f) had the typical punctate
pattern made up of a few small scattered bodies, as in cells
expressing DI RNA only (Fig. 2g) or in control cells trans-
formed with empty vectors (see Supplementary Fig. S1a).
Western blot analysis showed that the expression level of p92
was somewhat lower than that of p33 (Fig. 1e); however, we
were not able to say whether this was sufficient to explain
the different structural changes in peroxisomes in yeast
expressing p33 or p92. As, in these experiments, p33 and p92
were expressed by two different plasmids with different
promoters (CTA1 and ADH1, respectively), p92 was also
expressed by using plasmid pEL26. In addition, p33- and
p92-coding sequences were tagged with the epitope Myc at
the N terminus to analyse their subcellular distribution.
Yeast cells were transformed with these clones separately and
analysed by immunofluorescence using anti-Myc in con-
junction with each organelle marker as primary antibodies.
It was confirmed that CymRSV p33 localized largely to
peroxisomal aggregates (Fig. 3a, upper row), in line with
previous findings using this protein fused to the green fluor-
escent protein (Navarro et al., 2004). Expression of p92
resulted in punctate structures, most of which coincided
with scattered peroxisomes (Fig. 3a, lower row). These
results indicated that p92 is targeted to peroxisomes as p33,
but, although entirely containing the p33 sequence, does not
alter the distribution of these organelles.
To analyse the expression and distribution of CymRSV
replicase proteins when expressed together, the Myc and HA
tags were fused in frame to the N terminus of p33 and p92
and expressed with plasmids pEL26 and YE, respectively.
Yeast transformants were prepared that expressed these
proteins together with DI RNA. Western blot analysis
showed that the tagged proteins could be detected by using
the respective anti-Myc and anti-HA mAbs (Fig. 1f, lane 1)
and Northern blot analysis demonstrated that they were
fully competent for the replication of DI-3 RNA (Fig. 1d).
Concurrent with biochemical analysis, yeasts were fixed and
processed for double-labelled immunofluorescence using
anti-Myc and anti-HA antibodies. Signals corresponding
to Myc-tagged p33 and HA-tagged p92 largely coincided

Fig. 2. Expression of p33 modifies the size and distribution of peroxisomes. One or two large bodies co-localized with the
peroxisomal marker Fox3p in transformants expressing p33 alone (c and e) or together with p92 (a, b and h). Cells not
expressing p33 (d, f and g) displayed punctate structures identical to cells transformed with empty vectors (see
Supplementary Fig. S1a, available in JGV Online). The distribution of mitochondrial (CoxIIIp), ER (Kar2p) and Golgi (Emp47p)
protein markers was unaffected in all transformants and is shown only for the transformant expressing p33, p92 and DI RNA
(see Supplementary Fig. S1b, available in JGV Online). Bars, 1 mm.

708 Journal of General Virology 87


Fig. 3. p33 and p92 localize to peroxisomes. (a) Upper row:
p33 localizes to and modifies the structure of peroxisomes;
lower row: p92 localizes to peroxisomes without altering their
size and distribution. Both proteins were tagged with the Myc
epitope. Cells were immunolabelled by using anti-Myc and anti-
Fox3p antibodies. (b) p33 and p92 co-localize in the presence
(upper row) or absence (lower row) of DI RNA. Cells were
immunolabelled with anti-Myc and anti-HA. (c) p33 and p92
co-localize to peroxisomes. Cells were immunolabelled by using
anti-Myc and anti-Fox3p (upper row) or anti-HA and anti-Fox3p
(lower row) antibodies, respectively. Images of each label and
their superimposition are shown. Bars, 1 mm.
(Fig. 3b, upper row). These observations strongly indicated
the co-localization of replicase proteins p33 and p92. To
investigate whether co-localization of p33 and p92 was
http://vir.sgmjournals.org
CymRSV DI RNA replication in yeasts
dependent on DI RNA replication, cells transformed with
plasmids expressing the replicase proteins and plasmid
pBMI3S, i.e. without the DI RNA sequence, were analysed.
The distribution of p33 and p92 was independent of tran-
scription and replication of DI RNA (Fig. 3b, lower row).
The site of co-localization of p33 and p92 appeared
morphologically similar to the aggregates of peroxisomes
described above in cells expressing p33 (Fig. 2ac, e and h;
Fig. 3a, upper row). To verify these results, samples were
double-labelled with antibodies to either Myc or HA and
Fox3p. As shown in Fig. 3(c), the localization of both p33
(upper row) and p92 (lower row) correlated with the
peroxisomal marker.
Nascent RNA and progeny RNA do not
co-localize
To investigate the site of DI RNA replication, spheroplasts
were prepared from cells expressing the replicase proteins
and DI RNA grown in the absence of galactose, in which
the presence of DI RNA was dependent only on replication
and not on DNA-driven transcription. Spheroplasts were
incubated with BrUTP for 10 min and then fixed and
double-labelled with antibodies to BrUTP and Fox3p.
Incorporated BrUTP was detected in cytoplasmic sites,
which largely coincided with peroxisome aggregates also
labelled with anti-Fox3p antibodies (Fig. 4a, upper row).
Nascent viral RNA was never found associated with the ER
or mitochondria (see Supplementary Fig. S2a, available in
JGV Online). No incorporation of BrUTP was detected in
yeast cells expressing the replicase proteins and the non-
replicable form of DI RNA (Fig. 4a, middle row) or in cells
expressing DI RNA, but not the replicase proteins (Fig. 4a,
lower row). To visualize the distribution of DI-3 RNA
replication progeny, in situ hybridization was performed
with DIG-labelled oligonucleotides as described previously
(Restrepo-Hartwig & Ahlquist, 1999; Pantaleo et al., 2004).
As shown in Fig. 4(b, upper row), DI RNA progeny
accumulated at the site of peroxisomal aggregates and was
diffused throughout the cytoplasm, but not in the ER or
mitochondria (see Supplementary Fig. S2b, available in JGV
Online). No DI RNA-specific signal was found in control
cells (Fig. 4b, middle and lower rows). In conclusion, these
data suggest that replication of DI RNA occurs at the site of
peroxisome aggregations and that progeny RNA is then
released into the surrounding cytoplasm.
Electron microscopy fails to identify features
associated with DI RNA replication
Thin-sectioned yeast transformants were examined with
the aim of identifying fine-structural details of DI RNA
replication sites. Cells expressing only DI RNA showed the
presence of five to ten round to ovoid peroxisomal profiles
intermingled with mitochondria (Fig. 5a). By contrast, sec-
tions of cells also expressing the replicase proteins contained
a much higher number of peroxisomes (2030) of differ-
ent sizes, which were tightly appressed to the point of no
longer displaying a defined morphology (Fig. 5b). Several
709
B. Navarro and others

Fig. 4. Distribution of DI RNA progeny in UTL-7A cells. Newly synthesized DI RNA localized predominantly in association with
peroxisomes (a, upper row), whereas DI RNA progeny accumulated in the cytoplasm (b, upper row). Cells were double-
labelled by using anti-BrUTP (a) or anti-DIG (b) and anti-Fox3p antibodies. No significant BrUTP signal was detected in
control cells expressing non-replicable DI RNA (middle rows) or replicable DI RNA in the absence of replicase proteins (lower
rows). Images of each label and their superimposition are shown. Bars, 1 mm.

Fig. 5. Expression of p33 in UTL-7A cells

significantly increases the size and number
of peroxisomes. (a) A cell transformed with
empty vectors. (bd) Cells expressing p33
together with p92 and DI RNA transcripts
displaying clusters of peroxisomes (b, c),
many partially encircled by electron-dense
tubular structures (c, arrows) and accumula-
tions of membranes (d) free in the cytoplasm
or encircling lipid droplets. P, Peroxisomes;
M, mitochondria; N, nucleus; L, lipid drop-
lets. Bar, 0?25 mm.

710 Journal of General Virology 87


such peroxisomes were partially encircled by membranous
elements resembling ER strands, but differing because of
the higher electron density (Fig. 5c). These electron-dense
membranes were also occasionally apposed to mitochon-
dria, but only when these organelles were intermingled with
peroxisome clusters. A few cells also displayed stacks of
membranes often massively surrounding lipid droplets
(Fig. 5d). Despite extensive analysis, in no cases were
vesiculated structures observed resembling the multivesi-
cular bodies typical of tombusvirus infections in plant cells
(Russo et al., 1987). It was concluded that the major cyto-
logical features, i.e. peroxisomal aggregates and membrane
proliferation, were due to the expression of the replicase
proteins, probably p33 alone, as suggested by the light-
microscope observations reported above.
DI-3 RNA replication in the absence of
peroxisomes
Pex19p is an oleic-inducible, farnesylated protein of ~40 kDa
required for peroxisome biogenesis (Gotte et al., 1998).
It interacts with the majority of peroxisomal membrane
proteins (PMPs), escorting them from the cytosol to the
peroxisomal membrane (Jones et al., 2004). The S. cerevisiae
mutant UTL-7Apex19D (in which PEX19 is deleted; Gotte
et al., 1998) is unable to synthesize Pex19p and lacks
peroxisomes, so that both peroxisomal matrix and mem-
brane proteins are mislocalized in the cytosol or, occasion-
ally, in other organelles (Hettema et al., 2000; Sacksteder
et al., 2000; Rottensteiner et al., 2003). The competence
of UTL-7Apex19D cells to support CymRSV replicase-
mediated DI RNA replication was tested. As cells of this
mutant cannot be selected for leucine auxotrophy, the
plasmid YE92Ktyr (carrying the LEU2 selectable marker)
could not be used. Therefore, cells were co-transformed with
plasmid pBDI-3 and plasmid pEL26-Myc92Kwt, which
contained the wild-type p92 sequence including the amber
stop codon of p33 and the Myc tag fused in frame at the N
terminus. As a control, UTL-7A cells were transformed with
the same plasmids. Northern blot analysis showed that
replication of DI-3 RNA also took place in UTL-7Apex19D
cells, although at a level at least five times lower than that
of the control wild-type UTL-7A cells (Fig. 6a). Immuno-
fluorescence analysis showed that p33 (and presumably the
readthrough product, p92) accumulated in elongated struc-
tures identified as ER strands as a result of their reactivity
with the anti-Kar2p antibody (Fig. 6b, upper row). No co-
localization with mitochondria was observed (Fig. 6b, lower
row). These results were in line with a previous analysis
where it was shown that a CymRSV p33GFP fusion protein
expressed in strain UTL-7Apex19D cells mislocalized to the
ER instead of to the peroxisomal membrane (Navarro et al.,
2004). Given the low level of DI RNA replication, no attempt
was made to establish the site of replication by BrUTP
incorporation. However, this is likely to coincide with the
accumulation of the replicase proteins, i.e. ER strands in
this case.
http://vir.sgmjournals.org
CymRSV DI RNA replication in yeasts
Fig. 6. DI RNA replication in UTL-7Apex19D cells. (a) Northern
blot analysis of RNA extracted from UTL-7Apex19D (lane 1)
and UTL-7A (lane 2) cells expressing DI RNA and p92wt.
Lanes 1 and 2 were loaded with 5 and 1 mg, respectively. The
positions of dimeric (dim) and monomeric (mon) DI RNA mole-
cules are indicated. (b) UTL-7Apex19D cells expressing DI
RNA and p92wt were immunolabelled by using anti-Myc and
anti-Kar2p antibodies (upper row) and anti-Myc and anti-
Tom40p (lower row) antibodies. Images of each label and their
superimposition are shown. Bars, 1 mm.
DISCUSSION
In the present work, we employed a yeast system to study
the replication of CymRSV-derived DI RNA template sup-
ported by CymRSV replicase proteins. Two yeast strains
were used: UTL-7A, which is particularly favourable for the
analysis of peroxisome biogenesis, and UTL-7Apex19D,
which is not, as it is unable to synthesize Pex19p, a farnesyl-
ated protein required for peroxisome biogenesis. Evidence
that DI RNA replicated in UTL-7A cells was given by the
presence of dimeric molecules in addition to unit-length
monomers of positive and negative polarities, and by BrUTP
incorporation into nascent RNA progeny only when both
replicase proteins p33 and p92 were expressed. Expression of
p33 and p92 could be achieved from two separate plasmids
(provided that the p33 stop codon was mutated to a tyrosine
codon) or from only one plasmid encoding the entire p92-
coding sequence, including the p33 termination codon.
Since in S. cerevisiae, as in higher eukaryotes, several anti-
codon mutations of tRNA genes produce nonsense suppres-
sors capable of reading stop codons as sense codons (Beier &
Grimm, 2001), the complete set of CymRSV replicase
proteins was synthesized, leading to DI RNA replication.
Visual inspection of RNA blots indicated that the DI RNA
progeny derived from the expression of CymRSV replicase
from a single plasmid accumulated less than in transfor-
mants in which both p33 and p92 were overexpressed from
separate plasmids. As suggested by Panavas & Nagy (2003),
suppression of the amber stop codon may not be as effi-
cient as the readthrough mechanism operating in plant cells
infected with tombusviruses.
Fluorescence microscopy was used to analyse the cytological
modifications occurring in transformants expressing either
711
B. Navarro and others
p33 or p92, or both, co-expressed with replicable or non-
replicable DI RNA. Aggregates of peroxisomes constituted
the major cytological alteration of UTL-7A cells expressing
p33, regardless of whether this protein was expressed alone
or with p92 and DI RNA template. There was a clear co-
localization of p33 and the peroxisomal marker. By con-
trast, expression of p92 only did not perturb the size and
distribution of peroxisomes, which remained the only site of
localization of this protein.
When expressed together, p33 and p92 always co-localized
on peroxisomal aggregates in the presence or absence of
DI RNA, thus indicating that template RNA is not required
for accumulation of p33 and p92 on peroxisomal mem-
branes. However, co-expression of CymRSV replicase pro-
teins and template RNA may be required for the formation
of an active replication complex, as shown for the assembly
of the CNV replication complex, which was stimulated
by co-expression of template RNA and replicase proteins
(Panaviene et al., 2004).
BrUTP incorporation into the nascent DI RNA strand
occurred at the site of accumulation of p33 and p92 in
UTL-7A cells, providing evidence of the involvement of
peroxisomes in the replication of DI RNA sustained by the
CymRSV replicase. A similar association of CNV replicase
proteins p33 and p92 with peroxisomal membranes was
shown to take place in yeast cells together with template
RNA (Panavas et al., 2005a). DI RNA progeny do not
accumulate at the sites of replication, instead being diffused
throughout the cytoplasm, similar to the accumulation of
RNA progeny in yeast cells expressing CIRV (Pantaleo et al.,
2004) and Brome mosaic virus (BMV) (Restrepo-Hartwig &
Ahlquist, 1999) replicase proteins. Time-course experiments
with CNV replicase-directed DI RNA replication in yeast
have confirmed that the majority of RNA progeny does
not accumulate at the site of synthesis and is distributed
throughout the cytoplasm (Panavas et al., 2005a). Inciden-
tally, it is also worth noting that this indicates that the
organelles involved in tombusvirus replication depend only
on the replicase rather than on the RNA replication
template. In fact, the same DI RNA (DI-3 RNA) derived
from CymRSV genome is replicated at the level of either
mitochondrial aggregates in the case of CIRV (Pantaleo et al.,
2004) or peroxisomes in the case of CymRSV (this paper).
Electron-microscopy observations of yeast cells in which
there was active DI RNA replication did not show cyto-
logical alterations comparable to the typical multivesicular
bodies of CymRSV-infected plant cells. Similar negative
results were obtained with BMV (Schwartz et al., 2002) and
Flock house virus (FHV) (Miller et al., 2003). In particular,
the typical vesiculated mitochondria found in FHV-infected
insect cells (Miller et al., 2001) were absent in the corres-
ponding yeast cells, where very few vesicles were found on
severely damaged and hardly recognizable mitochondria.
For BMV, better visualization of vesicle accumulation in the
perinuclear space was obtained when nuclei were extracted,
thus permitting the observation of many such organelles
712
(Schwartz et al., 2002). In plant cells infected with CymRSV
and other tombusviruses (Russo et al., 1983), the peroxi-
somal membrane undergoes profound modifications con-
sisting of dilations that close on themselves, thus yielding
enclaves where the bulk of vesicles accumulate. The scarcity
of membranous structures related to virus replication in
yeasts, as compared with the plant host, may simply be
due to the short life (2448 h) of individual yeast cells,
which may not permit the production and accumulation of
replication-related vesicles. Alternatively, the space between
closely appressed peroxisomes might constitute the pro-
tected environment for CymRSV DI RNA replication.
As peroxisomes were shown to be the sites of DI RNA repli-
cation directed by CNV (Panavas et al., 2005a) and CymRSV
(this paper) replicase proteins, the low but consistent DI
RNA replication in strain UTL-7Apex19D is difficult to
reconcile with the absence of peroxisomes (Gotte et al.,
1998). However, it was shown that a mutant strain of the
yeast Pichia pastoris, unable to express Pex19p, contained
tiny vesicular and tubular structures, possibly derived from
the ER, identified as precursors of peroxisomes (Snyder et al.,
1999). If also present in S. cerevisiae strain UTL-7Apex19D,
these structures may be candidates for DI RNA replication
sites when biogenesis of peroxisomes is blocked at a very
early stage because of the absence of Pex19p.
It is an open question where CymRSV p33 and p92 pro-
teins are synthesized, whether on free or membrane-bound
ribosomes, and how they are targeted to the peroxisomal
membrane, whether directly or indirectly. Most PMPs are
synthesized in the cytosol and are targeted directly to peroxi-
somes; others are synthesized on ER-bound polysomes and
localized in the ER before being targeted to peroxisomes
(Baerends et al., 2000; Titorenko & Rachubinski, 2001).
Direct insertion of PMPs is mediated by the cytosolic
chaperone Pex19p, itself an integral PMP (Hettema et al.,
2000; Jones et al., 2004), which recognizes and binds specific
signals in the PMPs (Rottensteiner et al., 2004). Another
peroxisomal protein (Pex3p) is required for correct sorting
of this class of PMP (Hettema et al., 2000), acting as a
docking factor in the peroxisomal membrane (Fang et al.,
2004). Other PMPs are targeted indirectly to the peroxi-
somal membrane, i.e. they are first sorted to the ER and then
to peroxisomes. For instance, overexpression of Pex3p
(Baerends et al., 1996; Kammerer et al., 1998) and Pex15p
(Elgersma et al., 1997) results in profound proliferation of
ER membrane, which is taken as an indication that these
proteins are sorted first to the ER and then to peroxi-
somes. A similar route is suggested for Pex2p and Pex16p
(Titorenko & Rachubinski, 1998). It was shown previously
that CymRSV p33 and Pex19p failed to interact in a yeast
two-hybrid assay system (Navarro et al., 2004). Further-
more, definite targeting of p33 to the ER and DI RNA
replication were shown to take place in the mutant UTL-
7Apex19D, which does not synthesize Pex19p and lacks
peroxisomes (Navarro et al., 2004; this paper). These results
may be taken as an indication that the p33 protein does not
Journal of General Virology 87

follow the direct sorting to peroxisomes via the chaperone
activity of Pex19p, but is targeted first to the ER. A similar
route was suggested for CNV p33 (Panavas et al., 2005a).
Interestingly, PEX19 is not among the 96 identified yeast
genes whose absence affects the replication of tomato bushy
stunt virus DI RNA (Panavas et al., 2005b). However, by
using an algorithm developed for yeast PMPs (Rottensteiner
et al., 2004), three putative binding sites of p33 for Pex19p
have been predicted (S. Lorenzen & H. Rottensteiner, per-
sonal communication). Clearly, further experimental data
are required to establish the route of p33 to peroxisomes.
ACKNOWLEDGEMENTS
The authors wish to thank Professor G. P. Martelli for advice and
critical reading of the manuscript, Mrs A. Antonacci for valuable
technical help and the Istituto di Tecnologie Biomediche del CNR,
Sezione di Bari, Italy, for use of the fluorescence microscope. This
research was partially supported by MIUR, Project Cluster CO3, Legge
488/92, Studi di geni di interesse biomedico e agroalimentare.
REFERENCES
Baerends, R. J. S., Rasmussen, S. W., Hilbrands, R. E. & 7 other
authors (1996). The Hansenula polymorpha PER9 gene encodes a
peroxisomal membrane protein essential for peroxisome assembly
and integrity. J Biol Chem 271, 88878894.
Baerends, R. J. S., Faber, K. N., Kiel, J. A. K. W., van der Klei, I. J.,
Harder, W. & Veenhuis, M. (2000). Sorting and function of
peroxisomal membrane proteins. FEMS Microbiol Rev 24, 291301.
Beier, H. & Grimm, M. (2001). Misreading of termination codons in
eukaryotes by natural nonsense suppressor tRNAs. Nucleic Acids Res
29, 47674782.
Burgyan, J., Dalmay, T., Rubino, L. & Russo, M. (1992). The
replication of cymbidium ringspot tombusvirus defective interfering-
satellite RNA hybrid molecules. Virology 190, 579586.
Burgyan, J., Rubino, L. & Russo, M. (1996). The 59-terminal region
of a tombusvirus genome determines the origin of multivesicular
bodies. J Gen Virol 77, 19671974.
Di Franco, A., Russo, M. & Martelli, G. P. (1984). Ultrastructure and
origin of cytoplasmic multivesicular bodies induced by carnation
Italian ringspot virus. J Gen Virol 65, 12331237.
Elgersma, Y., Vos, A., van der Berg, M., van Roermund, C. W. T.,
van der Sluijs, P., Distel, B. & Tabak, H. F. (1996). Analysis of the
carboxyl-terminal peroxisomal targeting signal 1 in a homologous
context in Saccharomyces cerevisiae. J Biol Chem 271, 2637526382.
Elgersma, Y., Kwast, L., van den Berg, M., Snyder, W. B.,
Distel, B., Subramani, S. & Tabak, H. F. (1997). Overexpression
of Pex15p, a phosphorylated peroxisomal integral membrane pro-
tein required for peroxisome assembly in S. cerevisiae, causes
proliferation of the endoplasmic reticulum membrane. EMBO J 16,
73267341.
Erdmann, R., Veenhuis, D., Mertens, D. & Kunau, W.-H. (1989).
Isolation of peroxisome-deficient mutants of Saccharomyces cerevi-
siae. Proc Natl Acad Sci U S A 86, 54195423.
Evan, G. I., Lewis, G. K., Ramsay, G. & Bishop, J. M. (1985). Isolation
of monoclonal antibodies specific for human c-myc proto-oncogene
product. Mol Cell Biol 5, 36103616.
http://vir.sgmjournals.org
CymRSV DI RNA replication in yeasts
Fang, Y., Morrell, J. C., Jones, J. M. & Gould, S. J. (2004). PEX3
functions as a PEX19 docking factor in the import of class I
peroxisomal membrane proteins. J Cell Biol 164, 863875.
Gotte, K., Girzalsky, W., Linkert, M., Baumgart, E., Kammerer, S.,
Kunau, W.-H. & Erdmann, R. (1998). Pex19p, a farnesylated protein
essential for peroxisome biogenesis. Mol Cell Biol 18, 616628.
Hettema, E. H., Girzalsky, W., van den Berg, M., Erdmann, R. &
Distel, B. (2000). Saccharomyces cerevisiae Pex3p and Pex19p are
required for proper localization and stability of peroxisomal mem-
brane proteins. EMBO J 19, 223233.
Ishikawa, M., Janda, M., Krol, M. A. & Ahlquist, P. (1997). In vivo
DNA expression of functional brome mosaic virus RNA replicons in
Saccharomyces cerevisiae. J Virol 71, 77817790.
Ito, H., Fukuda, Y., Murata, K. & Kimura, A. (1983). Transformation of
intact yeast cells treated with alkali cations. J Bacteriol 153, 163168.
Jones, J. M., Morrell, J. C. & Gould, S. J. (2004). PEX19 is a
predominantly cytosolic chaperone and import receptor for class 1
peroxisomal membrane proteins. J Cell Biol 164, 5767.
Kammerer, S., Holzinger, A., Welsch, U. & Roscher, A. A. (1998).
Cloning and characterization of the gene encoding the human
peroxisomal assembly protein Pex3p. FEBS Lett 429, 5360.
Kolodziej, P. A. & Young, R. A. (1991). Epitope tagging and protein
surveillance. Methods Enzymol 194, 508519.
Leeds, P., Peltz, S. W., Jacobson, A. & Culbertson, M. R. (1991). The
product of the yeast UPF1 gene is required for rapid turnover of
mRNAs containing a premature translational termination codon.
Genes Dev 5, 23032314.
Miller, D. J., Schwartz, M. D. & Ahlquist, P. (2001). Flock house virus
RNA replicates on outer mitochondrial membranes in Drosophila
cells. J Virol 75, 1166411676.
Miller, D. J., Schwartz, M. D., Dye, B. T. & Ahlquist, P. (2003).
Engineered retargeting of viral RNA replication complexes to an
alternative intracellular membrane. J Virol 77, 1219312202.
Navarro, B., Rubino, L. & Russo, M. (2004). Expression of the
Cymbidium ringspot virus 33-kilodalton protein in Saccharomyces
cerevisiae and molecular dissection of the peroxisomal targeting
signal. J Virol 78, 47444752.
Panavas, T. & Nagy, P. D. (2003). Yeast as a model host to study
replication and recombination of defective interfering RNA of
Tomato bushy stunt virus. Virology 314, 315325.
Panavas, T., Hawkins, C. M., Panaviene, Z. & Nagy, P. D. (2005a).
The role of the p33 : p33/p92 interaction domain in RNA replication
and intracellular localization of p33 and p92 proteins of Cucumber
necrosis tombusvirus. Virology 338, 8195.
Panavas, T., Serviene, E., Brasher, J. & Nagy, P. D. (2005b).
Yeast genome-wide screen reveals dissimilar sets of host genes
affecting replication of RNA viruses. Proc Natl Acad Sci U S A 102,
73267331.
Panaviene, Z., Panavas, T., Serva, S. & Nagy, P. D. (2004). Purifica-
tion of the Cucumber necrosis virus replicase from yeast cells: role of
coexpressed viral RNA in stimulation of replicase activity. J Virol 78,
82548263.
Pantaleo, V., Rubino, L. & Russo, M. (2003). Replication of
Carnation Italian ringspot virus defective interfering RNA in
Saccharomyces cerevisiae. J Virol 77, 21162123.
Pantaleo, V., Rubino, L. & Russo, M. (2004). The p36 and p95
replicase proteins of Carnation Italian ringspot virus cooperate in
stabilizing defective interfering RNA. J Gen Virol 85, 24292433.
Redding, K., Holcomb, C. & Fuller, R. S. (1991). Immunolocalization
of Kex2 protease identifies a putative late Golgi compartment in the
yeast Saccharomyces cerevisiae. J Cell Biol 113, 527538.
713
B. Navarro and others
Restrepo-Hartwig, M. & Ahlquist, P. (1999). Brome mosaic virus
RNA replication proteins 1a and 2a colocalize and 1a indepen-
dently localizes on the yeast endoplasmic reticulum. J Virol 73,
1030310309.
Rottensteiner, H., Stein, K., Sonnenhol, E. & Erdmann, R. (2003).
Conserved function of Pex11p and the novel Pex25p and Pex27p in
peroxisome biogenesis. Mol Biol Cell 14, 43164328.
Rottensteiner, H., Kramer, A., Lorenzen, S., Stein, K., Landgraf, C.,
Volkmer-Engert, R. & Erdmann, R. (2004). Peroxisomal membrane
proteins contain common Pex19p-binding sites that are an integral
part of their targeting signals. Mol Biol Cell 15, 34063417.
Rubino, L. & Russo, M. (1998). Membrane targeting sequences in
tombusvirus infections. Virology 252, 431437.
Russo, M., Di Franco, A. & Martelli, G. P. (1983). The fine structure
of Cymbidium ringspot virus infections in host tissues. III. Role of
peroxisomes in the genesis of multivesicular bodies. J Ultrastruct Res
82, 5263.
Russo, M., Di Franco, A. & Martelli, G. P. (1987). Cytopathology in
the identification and classification of tombusviruses. Intervirology
28, 134143.
Russo, M., Burgyan, J. & Martelli, G. P. (1994). Molecular biology of
Tombusviridae. Adv Virus Res 44, 381428.
Sacksteder, K., Jones, J. M., South, S. T., Li, X., Liu, Y. & Gould, S. J.
(2000). PEX19 binds multiple peroxisomal membrane proteins, is
predominantly cytoplasmic, and is required for peroxisome mem-
brane synthesis. J Cell Biol 148, 931944.
714
Schlenstedt, G., Hurt, E., Doyle, V. & Silver, P. A. (1993).
Reconstitution of nuclear protein transport with semi-intact yeast
cells. J Cell Biol 123, 785798.
Schwartz, M., Chen, J., Janda, M., Sullivan, M., den Boon, J. &
Ahlquist, P. (2002). A positive-strand RNA virus replication com-
plex parallels form and function of retrovirus capsids. Mol Cell 9,
505514.
Snyder, W. B., Faber, K. N., Wenzel, T. J., Koller, A., Luers, G. H.,
Rangell, L., Keller, G. A. & Subramani, S. (1999). Pex19p interacts
with Pex3p and Pex10p and is essential for peroxisome biogenesis in
Pichia pastoris. Mol Biol Cell 10, 17451761.
Titorenko, V. I. & Rachubinski, R. A. (1998). Mutants of the yeast
Yarrowia lipolytica defective in protein exit from the endoplasmic
reticulum are also defective in peroxisome biogenesis. Mol Cell Biol
18, 27892803.
Titorenko, V. I. & Rachubinski, R. A. (2001). The life cycle of the
peroxisome. Nat Rev Mol Cell Biol 2, 357368.
Voinnet, O., Pinto, Y. M. & Baulcombe, D. C. (1999). Suppression of
gene silencing: a general strategy used by diverse DNA and RNA
viruses of plants. Proc Natl Acad Sci U S A 96, 1414714152.
Weber-Lotfi, F., Dietrich, A., Russo, M. & Rubino, L. (2002). Mito-
chondrial targeting and membrane anchoring of a viral replicase in
plant and yeast cells. J Virol 76, 1048510496.
White, K. A. & Nagy, P. D. (2004). Advances in the molecular biology
of tombusviruses: gene expression, genome replication, and recom-
bination. Prog Nucleic Acid Res Mol Biol 78, 187226.
Journal of General Virology 87