Materials & Methods

Molecular Plant Pathology On-Line []

The top component of citrus psorosis virus contains two ssRNAs, the smaller encodes the coat protein

Sánchez de la Torre, E 1 Riva, O.1, Zandomeni, R.2, Grau, O.1,2
and M. L. García1

1Instituto de Bioquímica y Biología Molecular (IBBM), Facultad de Ciencias Exactas, UNLP, - 2Centro de Investigación en Ciencias Agropecuarias, Instituto Nacional de Tecnología Agropecuaria.

Sánchez de la Torre, María Eugenia. E-mail:
Zandomeni, Rubén. E-mail:
Grau, Oscar. E-mail:


García, María Laura. Instituto de Bioquímica y Biología Molecular (IBBM), Facultad de Ciencias Exactas, UNLP. Fax number: +54-21-259223 E-mail:

The nucleotide sequence in this paper has been submitted to the GeneBank under the accession number AF060855


Citrus psorosis virus (CPsV) is associated with a widespread and damaging disease of citrus in many parts of the world. Here we show that one isolate of CPsV, called citrus ringspot virus (CtRSV-4), contains at least three genomic RNAs, two of them (RNA 2 and RNA 3) are found in the top component and the largest (RNA 1) in the bottom component. RNA 3, the smallest RNA, was cloned and a partial sequence of 1454 nucleotides was determined. One open reading frame (ORF) was found, which encodes a 439 amino acid protein with a MW of 48,654. This was identified as the coat protein by expressing part of the ORF in E. coli and detecting the product in western blots using an antiserum specific for the virions. Analyses by Northern blot show that both positive and negative strands of CPsV RNA 3 are encapsidated, the negative strand much more abundantly, thus defining the negative strand as the viral (v) RNA. No subgenomic RNAs were found.


Citrus psorosis (CPsV) is associated with a serious disease causing the death of citrus trees in many countries ( Roistacher, 1993). In Argentina the disease seems to be spread by an unknown vector (Beñatena and Portillo, 1984), and continues to cause serious losses (Danos, 1990). However, due to low concentration of the virus and poor stability of the particles, its genome has not been fully characterized.

Partial characterization of several isolates in the USA (Derrick et al., 1988, 1991), Spain (Navas-Castillo & Moreno, 1993, 1995; Navas-Castillo et al., 1993), Israel (da Graça et al., 1993), and Argentina (García et al., 1991a, b) indicates that at least two sedimenting components (top or T and bottom or B) are required for infectivity. In both components, circular filamentous particles are observed, with similar morphology although the B particles are about five times larger than the T particles (García et al., 1994). Both particles share a single coat protein (CP) (Derrick et al., 1988, García et al., 1991b).

Derrick et al. (1988, 1991) and Barthe et al. (1998) have reported that the citrus ringspot virus (CtRSV-4) isolate of CPsV from the USA (Garnsey & Timmer, 1980) contains one small RNA in the top (T) component coding for the CP gene and two larger RNAs in the bottom (B) component, one single stranded, the other double stranded.

In the present study we show that the T component of CPsV contains two genomic RNAs, the smaller one, RNA 3, encodes the CP in the negative sense.


Purification of the T component and viral RNA extraction

The CtRSV-4 isolate (Garnsey & Timmer, 1980) provided by L.W. Timmer and K.S. Derrick, University of Florida, Lake Alfred, was used in this study. The T component was obtained by the method of Derrick et al. (1988). The virus was partially purified from Chenopodium quinoa leaves bearing local lesions, using differential centrifugation; the product (P54 extract) was centrifuged into a sucrose gradient and the resulting T component was collected and electrophoresed in 0.5% agarose. Viral RNAs were extracted from the agarose using an RNAid kit (Bio 101) according to the manufacturer's instructions. Further purified preparations were obtained when the P54 extract was centrifuged for 19 h at 100,000 xg at 4C in 30% Cesium sulfate. The RNAs from this fraction were extracted with phenol-chloroform and after precipitation, the pellets were resuspended in RNase-free water. Total nucleic acids were also phenol-extracted from C. quinoa lesions after grinding with liquid nitrogen, precipitated and resuspended as before. These samples, T fraction and Cesium sulfate fraction were used for Northern blot hybridizations.

cDNA cloning and DNA sequence analysis

Complementary DNA (cDNA) from the RNAid preparations was obtained using the Librarian II cloning kit (Invitrogen) and random primers [pd(N)6 5'-PO4, Na salt (Pharmacia)]. The cDNA was made blunt-ended with T4 DNA polymerase, ligated to BstXI adapters, and inserted into BstXI-digested pcDNAII vector according to the manufacturer's instructions. The recombinant plasmids were cloned in E. coli DH5-a by electroporation using a gene pulser TM apparatus (Bio-Rad). The library was screened by colony blotting using 32P-labeled cDNA as probe prepared from RNA extracted of the T fraction by reverse transcription with random primers (Sambrook et al., 1989). To discard clones containing cDNA inserts derived from host RNAs, a second screening was done with a probe made from a healthy plant preparation corresponding to the T fraction. The clones were sequenced using the Taq Dye DeoxyTM Terminator Cycler Sequencing Kit (Applied Biosystems) and the Applied Biosystems Model 373A DNA Sequencing System equipment. The Wisconsin Genetic Computer Group (GCG) software Package Version 9.0, was used for sequence analysis.

To extend the sequences to the 3' and 5' ends of RNA 3, another strategy was applied using the sequences obtained before. Total RNA extracted from P54 containing partially purified negative and positive CPsV RNA 3 and the parallel healthy preparation were polyadenylated for 20 min at 37C using yeast poly(A) polymerase (Amersham) according to the manufacturer's instructions. After phenol-chloroform extraction and ethanol precipitation, the poly(A)-tailed RNAs were denatured at 95C for 5 min and reverse transcribed by incubation at 42C for 60 min in a reaction mixture (20 µl) using SuperScriptTMII RNAse H- Reverse Transcriptase (Gibco) according to the manufacturer's instructions. The primer PM-1 used for the first strand cDNA (kindly provided by Pedro Moreno, IVIA, Spain) was 5' CCGGATCCTCTAGAGCGGCCGC(t) 17v3', where V is A, C or G. In other cases the cDNA were obtained using the MarathonTM cDNA Amplification Kit (Clontech). In both strategies the primers designed for the second-strand cDNA were the left primer (5' GAAGGCAATTGCAATTAGGG 3') located between nucleotides 1046-1065 binding to the negative strand, and the right primer (5' CCAGTCTCTCTCAGAACTCC 3') located between nucleotides 171-190 binding to the positive strand. For synthesis of the second-strand cDNA an aliquot (1/4) of first-strand cDNA preparation was PCR-amplified in a reaction mixture (25 µl) containing 1x PCR buffer from Promega, 1 U of Taq DNA polymerase (Promega) and 5uM of each primer. The PM-1 primer and the left primer were used to amplify the 3' end of the positive strand RNA, and PM-1 and right primer to amplify the 3'end of the negative strand RNA. The PCR cycling profile (35 cycles) was 94C for 45 s, 45C for 45 s and 72C for 2 min, with a final extension step at 72C for 10 min, in a Perkin-Elmer 2400 thermocycler. The PCR products were separated by agarose gel electrophoresis, and DNA fragments of the expected size were eluted and cloned in the linearized pGEM-T plasmid (Promega) according to the manufacturer's instructions. The clones were sequenced and analyzed as above.

Northern hybridization

Total nucleic acids extracted from local lesions, the T fraction and the Cesium sulfate fraction were used for Northern blot analysis. The same procedures were applied in parallel to the uninfected tissue. For denatured Northern blots the samples were incubated for 1 h at 55C in a mix containing 10% glyoxal, 50% DMSO and 15mM sodium phosphate pH 6.5 and then electrophoresed in 1% agarose gel in 15 mM sodium phosphate pH 6.5 running buffer. The nucleic acids were transferred to a Zeta-Probe membrane (Bio-Rad) by capillarity using 25 mM sodium phosphate pH 6.5. Membranes were probed with [32P]-labeled insert of clone 194 from the RNA 2 and clone 55 from the RNA 3, using the random priming procedure (Sambrook et al., 1989). Riboprobes were [a -32P]-rATP-labeled transcripts prepared with T7 RNA polymerase (New England Biolabs) using 55/BamHI and 218/BamHI clones probing the positive and the negative sense of the RNA 3 respectively. Hybridizations were performed in 6 x SSC, 5 x Denhardt, 0.1% SDS and 3 mg/ml denatured yeast tRNA (SIGMA) for 20 h at 65C when DNA probes were used. For riboprobes 50% formamide, 5 x Denhardt, 0.5% SDS and 10 µg/ml denatured herring sperm DNA (Boerhinger) were used at 68C. After 3 washing steps in 2 x, 1 x and 0.5 x SSC plus 0.1% SDS at 65C, the membranes were exposed for 1.5 h or longer to X-ray film.

Cloning, sequencing and expression of the CP gene in E. coli BL21

Using two primers designed at the 5' end region of the ORF a PCR fragment was obtained from clone 47 from the RNA 3 library. The primers were LP55: 5' CAACTAAATACTAAGAACTGAccatggCGATTCC 3' (containing the first ATG of the ORF) and RP55: 5' CAGTATTTTTGTTCCAggatccACAAGCATATTGAGC 3'. The NcoI and BamHI restriction sites (indicated in lower case) were created substituting some nucleotides (indicated in bold), and ligated in one orientation in pET19b/NcoI-BamHI. The change in the LP55 primer introduced an alanine instead of a serine as the second amino acid of the ORF. In the RP55 primer the three changes did not affect the expression of the polypeptide since they were lost in the BamHI digestion. The recombinant plasmid was electroporated in E. coli BL21(DE). After confirmation by complete sequencing, the pET19b-1CP clone was used to express a 19.5 kDa polypeptide corresponding to the N-terminal portion of the CPsV-CP. A 1/10 dilution from an overnight culture was used to induce the expression of the polypeptide by IPTG (SIGMA) at 37C for 3-4 h. A culture without induction as negative control and a culture of E. coli BL21(DE) without recombinant plasmid were prepared in parallel.

SDS-PAGE and western blot analysis

Aliquots of each culture were centrifuged for 1 min and the pellets were resuspended in sample buffer for SDS-PAGE, boiled for 1 min and electrophoresed in 15% SDS-PAGE in order to test for the expression of the polypeptide. A similar gel was transferred to Inmobilon PVDF membrane (Millipore) for western blot analysis. CPsV-CP antiserum (provided by Dr. E. Luisoni from Istituto di Fitovirologia Applicata, Torino, Italy) diluted 1/500 was used to identify the expressed 19.5 kDa N-terminal polypeptide. Aliquots of P54 prepared from infected and uninfected citrus were taken as positive and negative controls, respectively. Goat anti rabbit IgG and alkaline phosphatase (Gibco) were used for color development.


Analysis of the CPsV RNAs

Two large clones from the top component that did not hybridize with each other were used as probes for Northern analysis of the viral top RNAs ( Figure 1). Each clone produced a single band when hybridized with purified virion RNA. However, the bands had different mobilities indicating the presence of two different RNAs. By comparison with appropriate markers the size of the larger RNA was estimated to be about 1,650 nucleotides and the smaller one 1,500 nucleotides. We have named them RNA 2 and RNA 3, respectively. Neither clone hybridized with RNA from healthy plants or with RNA 1.

Figure 1

cDNA cloning and sequence analysis of RNA 3

From the cDNA library obtained from the top RNA twenty clones were grouped by their hybridization with clone 55 ( Figure 2), and all were sequenced from both ends. A clearly defined consensus sequence was constructed, that included all twenty clones. Several attempts to identify the 5' and 3' ends resulted in many clones close to the ends, fourteen clones from near the 3' end of the positive strand, and seven clones that are probably near the 3' end of the negative strand, thus over-representing these regions. Each region of the RNA was represented by a minimum of six clones and a maximum of twenty seven. From them, seven to thirty seven sequences were obtained ( Figure 3). The sequence obtained was 1,454 bases long ( Figure 4) and contained a single large ORF of 1,317 nt. The first AUG codon was considered the true initiation codon since, as shown below, the molecular weight calculated for the 439 amino acid-long translation product (i.e. 48.654 Da) is similar to the value estimated by gel electrophoresis of the CP. No significant regions of similarity were found between the RNA 3 sequence and any others in all sequence databases.

Figure 2

Figure 3

Figure 4

Hybridization analysis of RNA 3

The polarity and the possible existence of subgenomic RNAs in purified virions and total nucleic acids from infected tissue was analysed by hybridization with riboprobes of RNA 3. These probes were prepared from the T7 promoter up to the BamHI site of the vector containing clone 55 or clone 218 (Figure 2). Riboprobe 55 detected coding (positive) RNA and probe 218 detected negative strand RNA.

A single intense band of negative polarity with the expected size was observed in total nucleic acid extracted from infected tissue but was not present in the control ( Figure 5, panel A). A band of the same size, but ten times less intense and with positive polarity, was detected (panel B). Another band migrating more rapidly was present in both infected and uninfected material, and was not considered specific. These results indicate the presence of a single viral species of RNA 3 in total nucleic acid extracts, therefore no subgenomic RNAs were detected.

Figure 5

When RNA 3 was analyzed from purified top component particles, again, an intense band was generated with the probe detecting negative strands and an aproximately 50 times lower signal was observed with the probe detecting positive strands.

Taken together these data show that the majority of virions encapsidate RNA 3 of negative sense because almost all RNA found in virions is of negative sense and the small amount of positive sense RNA detected decreased when virions were purified. This small amount of positive sense RNA found in the top component particle preparation of purified virions may also be packaged, since it is unlikely that naked RNA would sediment with the T fraction in sucrose gradients, and resist attack by nucleases during purification.

Identification of the product of the RNA 3 ORF

To identify this product, a fragment containing the first 157 codons plus 21 codons belonging to the pET19b plasmid was cloned and expressed in E. coli. Expression of this clone (pET19b-1CP) produced the expected polypeptide of 19.5 kDa ( Figure 6, panel A, lane 4), which was not present in uninduced bacteria. This polypeptide reacted in western blots with a CPsV specific CP antiserum as well as the CP presents in the P54 extract, the positive control (Figure 6, panel B, lanes 4 and 6). Another band of lower molecular weight was present in both infected and uninfected material but it was not considered specific.

Figure 6

The 19.5 kDa polypeptide was also used to obtain an antiserum, which was able to detect the CP extracted from infected tissue (data not shown).


The particles of CPsV are filamentous, circular and of at least two different sizes as seen by electron microscopy and by their sedimentation as "top" and "bottom" fractions in sucrose density gradients. Derrick et al (1991) found that there was dsRNA as well as an ssRNA in the B component, and one ssRNA in the T component, after electrophoresis in non-denaturing conditions. Our Northern blots show that in fact two RNAs slightly different in size are present in the T fraction. If we assume that the B fraction contains only a single species of RNA, then the virus would possess three genomic RNAs.

The morphology of the nucleoprotein suggests that a panhandle structure may be present, formed by complementarity of the 5' and 3' ends of the RNAs. Different strategies were used in our laboratory, to clone and sequence both 3' and 5' ends, however due to the low concentration of virions in the infected tissue we have not succeeded in obtaining a reliable result. As such complementary sequences have not been found yet, we think we have not reached the ends of the RNA or, alternatively, that a different mechanism to generate the circular structures observed by electron microscopy will have to be proposed.

After this manuscript was submitted, Barthe et al. (1998) published a partial sequence of one of the RNAs of CPsV also indicating that it codes for CP in the negative strand. Our results largely agree with those of Barthe et al. (1998). In fact, our sequence obtained from the consensus of a large number of clones, differs only in 8 nucleotides in the region published by Barthe et al., these changes introduce 2 amino acid substitutions of similar structures. This very low drift in the sequence of this isolate is also evidenced in the almost identical sequences obtained by us in all clones from each region of RNA 3.

In order to compare the relative abundance of (+) and (-) RNA encapsidated, positive and negative riboprobes used in Northern blots were prepared with equivalent specific activity.

The small proportion of RNA 3 of positive polarity detected in total nucleic acid preparations could represent transcripts for expression of the CP ORF, replicative forms, or encapsidated positive strands. We would not expect to detect naked RNAs after the sucrose gradient step because they would easily be degraded and in any case would not sediment together with the virus particles. We therefore suppose that a small proportion of the (+) strand of RNA 3 is encapsidated along the (-) strand.

Thus, we conclude that the predominantly encapsidated (-) RNA is the viral (v) RNA, and the other (+) strand is the viral complementary (vc) RNA. However, we did not detect a double-stranded form of RNA 3 in non-denaturing Northern blots of total nucleic acids or of nucleic acids from purified virions (data not shown).

We also did not detect subgenomic derivatives of RNA 3 in these fractions, indicating that the ORF is expressed from the intact positive strand.

The polypeptide derived from the N-terminal portion of the ORF was detected using an antiserum against the virions, confirming that it was part of the CP.

Citrus psorosis virus, a member of the ophiovirus genus, has a multipartite genome, at least one of the three RNAs is negative, and the morphology of the nucleocapside resembles the circular virions of the Tenuiviruses and internal nucleocapside of the Bunyaviridae. However, no similarities where found with the sequences of the coat proteins of those viruses and the coding strategy is not ambisense in RNA 3.


We thank L.W. Timmer and K.S. Derrick for the isolate CtRSV-4; P. Moreno for the PM-1 primer; N. Costa for citrus seedlings and G. Chiarrone for greenhouse work; and L. Nieto for helping in screening the library. The authors are indebted to R.G. Milne for his invaluable help in the critical reading of the manuscript. O.G. and M.L.G. belong to the staff of Facultad de Ciencias Exactas, UNLP.; O.G. is recipient of a research career award from CIC BA and a member of CICA-INTA. M.E.S.T. is a fellow of CONICET and belongs to the staff of Facultad de Ciencias Agrarias y Forestales, UNLP. R.Z. and M.L.G. are recipients of research career awards from CONICET. This work was supported by a grants BID802 OC-AR PID 332, SECyT-CONICET, UE contract number ERBIC18CT960044, and CIC BA no 1454/97.


Barthe GA, Ceccardi TL, Manjunath KL, Derrick KS, 1998. Citrus psorosis virus: nucleotide sequencing of the coat protein gene and detection by hybridization and RT-PCR. Journal of General Virology 79,1531-1537.

Beñatena HN and Portillo MM, 1984. Natural spread of psorosis in sweet orange seedlings. In: Garnsey SM, Timmer LW and Doods JA eds. Proceedings of the Ninth Conference of the International Organization of Citrus Virologists. Riverside, CA, USA: IOCV, University of California, 159-64

da Graça JV, Bar-Joseph M, Derrick KS, 1993. Immunoblot detection of citrus psorosis in Israel using citrus ringspot antiserum. In: Moreno P, da Graça JV, Timmer LW, eds. Proceedings of the Twelfth Conference of the International Organization of Citrus Virologists, 1992. Riverside, CA, USA: IOCV, University of California, 432-34.

Danos E, 1990. La psorosis de los cítricos: la epidemia en curso en Argentina y el desafio de su control. Revista de Investigaciones Agropecuarias 22, 265-77. Ed. International Foundation for Science (IFS) e Instituto Nacional de Tecnologia Agropecuaria (INTA).

Derrick KS, Brlansky RH, da Graça JV, Lee RF, Timmer LW, Nguyen TK, 1988. Partial characterization of a virus associated with citrus ringspot. Phytopathology 78, 1298-301.

Derrick KS, Lee RF, Hewitt BG, Barthe GA, da Graça JV, 1991. Characterization of citrus ringspot virus. In: Timmer LW, Garnsey SM, Navarro L, eds. Proceedings of the Eleventh Conference of the International Organization of Citrus Virologists, 1989. Riverside, CA, USA:IOCV, University of California, 386-90.

García ML, Arrese EL, Grau O, Sarachu AN, 1991a. Citrus psorosis disease agent behaves as a two-component ssRNA virus. In: Timmer LW, Garnsey SM, Navarro L, eds. Proceedings of the Eleventh Conference of the International Organization of Citrus Virologists, 1989. Riverside, CA, USA: IOCV, University of California, 337-44.

García ML, Grau O, Sarachu AN, 1991b. Citrus psorosis is probably caused by a bipartite ssRNA virus. Research in Virology 142, 303-11.

García ML, Dal Bo E, Grau O, Milne RG, 1994. The closely related citrus ringspot and citrus psorosis viruses have particles of novel filamentous morphology. Journal of General Virology 75, 3585-90.

Garnsey SM, Timmer LW, 1980. Mechanical transmissibility of citrus ringspot virus isolates from Florida, Texas, and California. In: Calavan EC, Garnsey SM, Timmer LW, eds. Proceedings of the Eighth Conference of the International Organization of Citrus Virologists, 1980. Riverside, CA, USA: IOCV, University of California, 174-79.

Navas -Castillo J, Moreno P, 1993. Biological diversity of citrus ringspot isolates in Spain. Plant Pathology 42, 347-57.

Navas -Castillo J, Moreno P, 1995. Filamentous flexuous particles and serologically related proteins of variable size associated with citrus psorosis and ringspot diseases. European Journal of Plant Pathology 101, 343-48.

Navas -Castillo J, Moreno P, Cambra M, Derrick KS, 1993. Partial purification of a virus associated with a Spanish isolate of citrus ringspot. Plant Pathology 42, 339-46.

Roistacher CN, 1993. Psorosis - a review. In: Moreno P, da Graça JV, Timmer LW, eds. Proceedings of the Twelfth Conference of the International Organization of Citrus Virologists, 1992.Riverside, CA, USA: IOCV, University of California, 139-54.

Sambrook J, Fritsch EJ, Maniatis T, 1989. Molecular Cloning. A Laboratory Manual. 2nd edition, Cold Spring Harbor Laboratory Press.