Journal of Virological Merhods, 30 (1990) 161-172

Elsevier 161

VIRMET 01076

Dot hybridization detection of plum pox virus using “P-labeled RNA probes

representing non-structural viral protein genes

T. Wetzel, G. Tavert, P.Y. Teycheney, M. Ravelonandro, T. Candresse and J. Dunez

Station de Pathologie Vkgitale, Institut National de la Recherche Agronomique. Centre de Bordeaux, Villenave d’Omon, France

(Accepted 12 July 1990)

Summary

A cDNA library covering the complete genome of plum pox virus strain D (PPV D) has been obtained, and an endonuclease restriction map derived from it. This map was superposed on the PPV genomic organisation map, established for a non- aphid transmissible strain of PPV (Maiss et al., 1989). This allowed us to select seven probes, corresponding to different regions on the PPV genome. These probes were tested in a dot-blot hybridization assay for the detection of PPV. Probes of various lengths (0.25 to 1.5 kb) were tested and those measuring at least 0.8 kb (4 of the 7 probes selected) proved to be the most sensitive. The detection limit was of about 5 pg of purified virus per assay. Probes representing non-structural viral protein genes were equally sensitive in detecting both serotypes D and M of PPV. The previously described probe pBPPV1 (Varveri et al., 1988), covering the coat protein gene of strain D, was less sensitive, when compared to the above probes, in detecting heterologous strains of PPV. The polyvalence of probes transcribed from non-structural viral protein genes was confirmed by screening isolates of PPV, collected in infected orchards in several Mediterranean countries.

Immunoenzymatic assay; Molecular hybridization

Station de Pathologic VkgCtale, Institut National de la Recherche Agronomique, Cenue de Bordeaux, B.P. 81 33883, Villenave d’Omon, Cedex France.

0168-8510190/%03.50@ 1990 Elsevier Science Publishers B.V. (Biomedical Division)

162

Introduction

Two different serotypes of plum pox virus (PPV), a member of the potyvirus group, have been described (Kerlan and Dunez, 1979). PPV is the causal agent of the devastating Sharka disease of stone fruit trees (Dunez and Sutic, 1988). This disease is currently controlled in France by sanitary selection measures based on ELISA. In spite of a detection level of about 1 ng o~pu~fied virus per assay, the very low titer and the uneven distribution of the virus in the infected trees make it necessary to develop more sensitive detection techniques. Varveri et al. (1988) have developed a molecular hybridization assay for the detection of PPV, using 32P-labeled in vitro transcribed RNA probes, which proved to be more sensitive than ELISA in a large scale field indexing assay with infected apricot samples (Varveri et al., 1988). The probe used for this study (pBPPV1) was derived from the 3’ region of the viral RNA and covered about 65% of the coat protein gene. The detection limit of the assay was 4 pg of purified virus (strain D, homologous strain), but only 20 pg of purified virus for strain M, belonging to the second serotype (Kerlan and Dunez, 1979). We have tested probes covering other regions of the genome in an attempt to find a polyvalent detection test, able to detect with the maximal sensitivity the different strains of PPV. Complementary DNA clones representing non-structural viral protein genes were selected from a cDNA genomic library. These clones were used to synthesize RNA probes for PPV detection. Their performances were tested with purified virus, infected material from the glasshouse and field samples in routine indexing conditions. Probes representing non-s~ctural protein genes proved to be polyvalent, detecting both serotypes D and M with the same sensitivity (5 pg of purified virus per assay).

Materials aud Methods

PPV strains and isolates

Probes were tested against PPV D, an apricot isolate from southeastern France, and also against PPV M, a peach isolate from northern Greece, belonging to the second viral serotype (Kerlan and Dunez, 1979). These virus strains were propagated in Pisum sa&iva~ “Express Genereux”. PPV D and PPV M virus purifications were done as described previously (Varveri et al., 1987). The probes were tested against different PPV isolates, collected from infected peach and apricot trees in France, Spain, Greece, Cyprus and Turkey, and propagated in a glasshouse in peach seedlings GF305.

Synthesis and c&zing of PPV-D COLA

PPV D RNA purification was done as described previously (Varveri et al., 1987). cDNA synthesis and cloning were done according to Gubler and Hoffman

163

(1983), as described (~ave~~~~o et al., 1988). The PPV D geuomic map was established by restriction enzyme digestions and Southern blat analysis using nick-translated inserts (Maniatis et al., 1982). Subcloning was done in plasmid Bluescribe (Stratagene) and Escherichiu coli JM83 cells (Hanaban, 1983).

Probe synthesis and analysis

Probes were synthesized as described previously (Varveri et al., 1988), using the Riboprobe Gemini System @omega), with phage T3 or T7 RNA polymerase, dependiug on the insert o~en~ti~n in the plasmid. The radioactive precursor used was [a- 32P]CTP (800 Cilmmol, Amersham). The probe pBPPV1 (Varveri et al., 1988) was used as a reference.

An aliquot of the in vitro transcribed RNA probes (105 cpm) was heated 10 ruin at ‘70°C in 3 (N morpholino)propanesulfonic acid (MOPS) ethylenediaminete- traacetic acid (EDTA) buffer containing 50% formamide and 6% formaldehyde, and electrophoresed in a 1.5% denaturing agarose horizontal gel (Miller, 1987). The dried gel was exposed to X-ray films (Kodak X AR) at room temperature for 1 h.

Hybridization conditions

All p~hyb~di~tion, hyb~~~tion and washing steps were done as described (Melton et al., 1982; Varveri et al., 1988), except that probes were at 2,106 cpm/ml. Dry processed membranes were exposed to X-ray films at -8OOC for 65 h using intensifying screens. The autoradiograms were scam& using a CAMAGFLC scanner. The values obtained by pea integration were further corrected by deducing the background (determined by scanning buffer control spots) and the results expressed in arbitrary OD units.

Field &de&g trials

Samples were collected in a small (190 trees), isolated, naturally infected orchard in southern France. Three leaves per twig consti~ted one sample. Samples taken from s~ptorn~~~~g areas of the tree were noted symptom positive, regardless of the actual presence of symptoms. Iu experiment I, one hundred and four twigs were assayed on a tree with generalized symptoms. In experiment 2, five trees, in the same naturally infected orchard, bearing for the first time very localized symptoms, were tested. Thirty twigs per tree were assayed, for a total of 150 samples.

Sample prepa~at~~ for dot-blotting

Purified virus in 50 mM t&odium citrate buffer, pH 8.3, was diluted in 12xSSC 6% fo~aldehyde (SSC=150 mM NaCl, 15 mM t&sodium citrate, pH 7). One gram of peach or apricot leaves was ground in 4 ml of 50 mM

trisodium citrate, pH 8.3, 20 mM sodium diethyldithioc~bamate (DIECA), 2% polyvinylpyrrolidone PM 25000 (PVP K25), and diluted in 12xSSC 6% formaldehyde. Five S-fold dilutions were performed on the peach samples (di- luted 1 to 1 (v/v) in 12xSSC 6% f~~aldehyde). One lo-fold dilution was per- formed on each of the apricot samples. From these preparations, 10 ~1 aliquots were spotted on a 20 x SSC saturated nitrocellulose membrane. Membranes were then air dried and baked in vacua at 80°C for 2 h.

E~~yrne-kicked i~~~~sorbe~t assay

The Sanofi commercial PPV detection kit (Sanofi Sante Animale, 37 av. Georges V, 75008 Paris, France) was used. The apricot extracts were diluted lo-fold in 50 mM citrate buffer, pH 8.3, and 250 ~1 was applied. The threshold of detection was established as described previously (Varveri et al., 1987).

Results

Isolation and characterisation of cDNA clones

Recombinant plasmids containing inserts larger than 1.5 kb were selected for the construction of an endonuclease restriction map. Cloned regions identified cover about 9.8 kb, corresponding to 99.6% of the PFV D genume (Fig. IA). The restriction endonucleases BamWI, SstI, SalI, SphI and X/z& revealed single sites in the cloned DNA. Multiple restriction sites were found with EC&I, EcDRV, and Hind111 (Fig. 1A). The comparison of the PPV D restriction map with that of PPV NAT (Maiss et al., 1989), revealed a number of common restriction sites. The differences between the maps of the two strains co~e~nded to three ad~tional restriction sites in the PPV NAT genome (Fig. 1A). The PPV D restriction map was superimposed on the putative PPV genomic organisation map (Fig. 1C) established for PPV NAT (Maiss et al., 1989). To obtain probes corresponding to the various parts of the genome, cDNA fragments ranging from 0.25 to 1.5 kb (Fig. 13) were subcloned in the plasmid Bluescribe. Clones P&25, PO.55 and Pl. 1 corresponded to regions within the putative RNA polymerase coding sequences, Pro.6 to sequences encompassing the NIa protease~A polymerase junction, CI1.2 to a region within the coding region for the cylindrical inclusion protein, HCl , 1 to a region within the coding region for the helper component protein, and S’T1.5 to the 5’ terminal region of the’PPV genome, which covered the coding regions for the 34K protein and the 5’ end of the helper component protein. The G+C content of the probes ranged from approximately 41.5% (HCl. 1) to 45% (CIf .2).

RNA probe synthesis

RNA transcripts complements to the viral RNA were synthetised from each of the sub&ones. The in vitro transcription reaction yielded transcripts of the

165

(A)

PPU-D

(B1 pBPPu1 y

Po.25~ i i

PO.55 ; ; , I.

Pl.l- , I

Pr46;

IC1.2 - I I

m:l.l- I i I ; I

Fig. 1. Probes used for PPV detection. A, The restriction sites were those of the enzymes BarnHI (B), EcoRI (E), EcoRV (Ev), HindIII (H), KpnI (K), WI (S), SphI (Sp), SsrI (Ss), XhoI (X). The boxed sites represent additional sites present in PPV NAT. B, Localisation on the PPV genome of the different RNA probes used for PPV detection. C, Putative PPV genomic organ&ion map, established for PPV

NAT (Maiss et al., 1989).

expected size, as demonstrated by their migration during denaturing agarose gel electrophoresis (Fig. 2). Occasionally, a few smaller products which probably arose from incomplete .polymerizations were detected. The specific activity of the probes was 2.5x 108 cpm/pg.

Comparative sensitivity for detection of purified PPV by dot-blot hybridization using RNA probes representing structural and non-structural viral protein genes

RNA probes representing non-structural protein genes were compared to the previously described probe pBPPV1 (Varveri et al., 1988) for their sensitivity of detecting both serotypes PPV D and PPV M in a dot-blot hybridization assay. At the same time, it was bf interest to determine the influence of probe’ length on the sensitivity of the assay. Ideutical membranes, with purified virus d Wh serotypes D and M, were prepared and hybridized with each of the dif%t%nt p&es. The detection limit of the probe pBPPV1 was 5 pg of PPV-D per assay, and 24 pg of PPV-M per assay (Fig. 3A), as described by Varveri et al. (1988). Increased sensitivities were observed with probes of increasing length: the detection limit was 120 pg of purified virus per assay with probe PO.25 (0.25 kb), 24 pg with probe PO.55 (0.55 kb), and 5 pg with probe Pl. 1 (1.1 kb) (Fig. 3B-D). Similar results were obtained with the other probes tested: Pro.6 (0.6 kb).detected 24 pg of purified virus per assay (Fig. 3E), whereas the probes CD.2 (1.2 kb), HCl. 1 (1.1 kb) and 5’T1.5 (1.5 kb) detected 5 pg of purified virus per assay

166

A 6 C D

Fig. 2. Analysis by denaturing agarose gel electrophoresis of the products of in vitro transcription reactions used as probes in this study. A: Pr 0.6; B: CI 1.2; C: HC 1.1; D: 1 kb ladder.

M PPV

D

M

D

M

D

M

D

k5000 3000 600 120 24 5’pg 15000 3000 600 120 24 5pg

M A E

D

c PA

D D

w

Fig. 3. Detection of purified PPV-D and PPV-M virus by hybridization using RNA probes corresponding to structural and non-structural protein genes. A, probe pBPPV1; B, probe P 0.25; C, probe P 0.55; D,

probe P 1.1; E, probe Pr 0.6; F, probe CI 1.2; G, probe HC 1.1; H, probe 5’T 1.5.

(Fig. 3F-H). Furthermore, in contrast to probe pBPPV1, probes representing non-structural viral protein genes detected both PPV D and PPV M with the same sensitivity (Fig. 3B-H), proving a higher polyvalence for detection of these

167

TABLE 1 Densitometric analysis of the autoradiograms pmsented in Fig. 3, expressed in arbitrary 0-D. units

pBPPV1 PO.25 PO.55 P1.l PdL6 C11.2 HC1.l 5T1.5

15000 Pg D 420 235 450 480 440 450 440 450 M 380 210 450 480 430 440 450 450

3000 Pg D 240 115 M 185 105

600 Pg D 125 M 85

120 Pg D 43 4 M 8 3

24 pg D 12 - M 4 -

5 Pg 2 -

210 240

105 11.5

23 25

5 5

290 300 305 300

180 105 210 190 195 190 100 215 210 185

80 27 85 80 75 65 25 85 80 75

9 5 11 8 9 8 4 11 10 10

3 - 2 3 - 2

different serotypes of the virus. A densitometric analysis of the autoradiograms (Table l), which qu~ti~ted the hyb~~tion results, confirmed the detection limits obtained with the different probes.

Polyvalence of RNA probes representing non-structural viral protein genes

Probes corresponding to non-structural protein genes proved to be capable of detecting both purified serotypes with the same level of sensitivity. Such probes had a higher sensitivity than probe pBPPV1 for the detection of heterologous strains of PPV in infected’ plants. Different isolates of PPV, collected in infected orchards in several Mediterranean countries, were tested in a dot-blot hybridiza- tion assay. Identical membranes were prepared and hybridized with each probe (except PO.25 and PO.55). Because the samples were infected plant extracts, the signals given by the different isolates were correlated with virus con~n~tion in the infected material. Comparisons should be made only between membranes and not within them, each hybridized with the different probes. Most of the plant extracts diluted 1:625 appeared to give a weak signal with probe C11.2 (Fig. 4), whereas the detection limit of probe pBPPV1 was a l/125 dilution for most of the same samples {not shown). Identical results were obtained with the other non- saute probes tested, which suggested that they all were more sensitive than probe pBPPV1. Furthermore, non-specific reactions with healthy plant extracts were not observed with any of the probes.

Large scale application of molecular hybridization using RNA probes represen- ting eon-struc~al viral protein genes

To determine the potential of RNA probes representing non-structural viral protein genes for routine PPV detection, we made comparative field trials, by indexing apricot samples collected in a naturally infected orchard in southern

168

1

5

25

125

625

3125

ABCDEFGHS -Lc

Fig. 4. Detection of PPV isolates using the RNA probe CI 1.2. The isolates came from A: Cyprus; B: Turkey; C: Cyprus; D: France; E,F,G: Spain; H: Greece; S: healthy peach.

France. Since PPV D had been isolated near the orchard chosen, the isolate of PPV present in infected trees is probably, if not the same, closely related to PPV D. Levels of detection of PPV in molecular hybridization experiments can be expected to be almost the same among the different RNA probes. The grinding buffer used allowed the same extracts to be tested by molecular hybridization and ELISA.

Two experiments were done, the first on a tree with generalized symptoms, the second on trees bearing very localized symptoms which had appeared within the last 12 months. The results of these experiments are given in Table 2. A good correlation was observed between the presence of symptoms (as defined in Ma-

TABLE 2 Comparison of the detection of PPV in apricot samples by enzyme-linked immunosorbent assay (ELISA) and molecular hybridization (MH) using the RNA probes pBPPV1 (A) and CIl.2 (B) in two field indexing trials

Symptoms Experi- Nb ELISA positives ELISA negatives mema samp1es MH+ MH- MH+ MH-

A B A B A B A B

+ 1 42 26 26 0 0 11 12 5 4 2 19 18 : 0 0 1 1 0 0

Subtotal 61 44 0 5

(%) (72) (72) YO) (0) 1F9.5) Xl) (8.5) ;6)

- 1 62 2 2 0 0 4 5 56 55 2 131 2 2 1 1 2 126 126

Subtotal 193 4 4 :

(So) (2) (2) lO.5) fO.5) (3) T3.5) ‘& lFi3.5)

Total 254 48

(%) (18.5) FF8.5) tO.5) tO.5) 18 20 187 185

(7) (7.5) (73.5) (72.5)

a Experiments and definition of symptom positive (+) and symptom negative (-) as in Materials and Methods under Field indexing trials.

169

terials and Methods) and the detection of the virus by ELISA (61.5% detection in symptom positive samples in experiment 1, and 94.5%. in experiment 2). By comparison, molecular hybridization (MH+) detected the virus in 88% (experi- ment 1) and 100% (experiment 2) of the same samples with the probe pBPPV1, and in 90% (experiment 1) and 100% (experiment 2) with the probe C11.2. When considering the ELISA positive samples, an almost perfect correlation between molecular hybridization and ELISA was observed (1 MH negative out of 49 ELISA positive). With the symptom ‘negative (as described above) samples, a low detection level was obtained. When considering the ELISA negative sam- ples, molecular hybridization detected the virus in 19.5% (probe pBPPV1) and 22% (probe C11.2) of these samples (experiment 1). These values were lower than those obtained previously in a similar experiment (Varveri et al., 1988). Exceptionally high temperatures in the field in May 1989 when the experiments were done (measured at 31-35OC) may have decreased virus titers in infected plants. Alternatively, the poor detection could be explained by absence or pres- ence at levels below detection limits of PPV in the major parts of trees bearing for the first year very localized symptoms (experiment 2). Except for probe Pr0.6, the same results were obtained with the other probes tested.

An autoradiograph of hybridization with some field samples, using the probe C11.2, is given in Fig. 5. No non-specific reaction with healthy plant material was observed, even after a 10 days exposure of the membranes. However, this longer exposure allowed the detection of the virus in four additional samples with the probe pBPPV1, and five with the probe C11.2 (not shown).

HHHH Fig. 5. Dot-blot hybridization using the RNA probe CI 1.2 to detect PPV in infected apricot samples.

The samples with an arrow are negative in the ELISA test. H: healthy apricots.

170

Discussion

RNA probes (pBPPV1) proved superior for detection of PPV D (Varveri et al., 1988), as compared to DNA probes, or to ELISA. The relative loss of sensitivity observed with probe pBPPV1 when trying to detect heterologous strains such as PPV M, prompted us to select other probes with a higher polyvalence but a similar sensitivity.

Seven probes, corresponding to non-structural viral protein genes of the PPV genome were selected and tested for their polyvalence and their sensitivity. A minimal probe length of approximately 0.8 kb was required to obtain the maximal sensitivity. The probes longer than 0.8 kb proved to be as sensitive as the reference probe pBPPV1 (0.8 kb), detecting 5 pg of purified virus per assay. Furthermore, the loss of sensitivity in detecting the heterologous strain PPV M is abolished when using any of these probes, proving their high polyvalence. The alignment of the nucleotide sequences of PPV D (Teycheney et al., 1989) with those of two other strains of PPV, PPV NAT (Maiss et al., 1989) and PPV Rankovik (Lain et al, 1988), revealed 97% and 97.5% homology, respectively, in the capsid protein gene, and 98% and 99.2% in the putative RNA polymerase gene. However, we have no sequence information for PPV M, nor have PPV NAT and PPV Rankovik been serologically classified. Therefore, it is not possible to determine the homology level required to obtain an identical detection with different serotypes of the virus.

In view of the fact that probes corresponding to non-structural protein genes show a higher polyvalence, it was of interest to determine if they were able to cross hybridize with the RNA of other potyviruses. The alignement of the nu- cleotide sequences of PPV D with those of tobacco etch potyvirus (Allison et al., 1985), and tobacco vein mottling potyvirus (Domier et al., 1986), revealed 61.4% and 62% homology, respectively, in the putative polymerase gene, the most con- served gene among these three viruses. Preliminary results, using the probe P1.l, showed a very low cross-hybridization with a strain of zucchini yellow mosaic potyvirus under reduced stringency hybridization conditions (Wetzel, manuscript in preparation).

In routine testing of apricot samples, probes representing non-structural vi- ral protein genes proved to be as sensitive as the previously described probe pBPPV1. However, whereas probe pBPPV1 could only detect fivefold higher amounts of the heterologous serotype per assay, the probes we have described were able to detect the same amounts per assay of purified PPV D and PPV M. Furthermore, they were able to recognize in tissue extracts, with the maximum sensitivity, isolates from different Mediterranean countries representing a wide variability of PPV isolates. Collection of samples from trees and detection of very low titers of virus remain a problem in the diagnosis of PPV infections, how- ever the improvement in detection of PPV obtained with the polyvalent probes reported here is a step forwards to the development of a better diagnosis system.

171

References

Allison, R.F., Sorenson, J.C., Kelly, M.E., Armstrong, F.B. and Dougherty, W.G. (1985) Sequence determination of the capsid protein gene and flanking regions of tobacco etch virus: evidence for synthesis and processing of a polyprotein in potyvirus genome expression. Proc. Natl. Acad. Sci. USA 82, 3969-3972.

Domier, L.L., Franklin, K.M., Shahabuddin, M., Hellmann, G.M., Overmeyer, J.H., Hirematb, S.T., Siaw, M.F.E., Lomonossoff, G.P., Shaw, J.G. and Rhoads, R.E. (1986) The nucleotide sequence of tobacco vein mottling virus RNA. Nucleic Acids Res. 14, 5417- 5430.

Dunez, J. and Sutic, D. (1988) Plum pox virus. In: Smith, Dunez, Lelliot, Philips and Archer (Eds), European Handbook of Plant Diseases. Blackwell Scientific Publications, Oxford, pp. 44-46.

Gubler, U. and Hoffman, B.J. (1983) A simple and very efficient method for generating cDNA libraries. Gene 25, 263- 269.

Hanahan, D. (1983) Studies on transformation of Escherichia coli with plasmids. J. Mol. Biol. 166, 557-580.

Kerlan, C. and Dunez, J. (1979) Differentiation biologique et serologique de souches du virus de la Sharka. Ann. Phytopathol. 11, 241-250.

Lain, S., Riechmann, J.L., Mendez, E. and Garcia, A.G. (1988) Nucleotide sequence of the 3’ terminal region of plum pox potyvirus RNA. Virus Res. 10, 325-342.

Maiss, E., Timpe, II., Brisske, A., Jelkmann, W., Casper, R., Himmler, G., Mattanovich, D. and Katinger, H.W.D. (1989) The complete nucleotide sequence of plum pox virus RNA. J. Gen. Viil. 70, 513-524.

Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual. Cold Spring Ha&or Laboratory, Cold Spring Harbor, NY.

Melton, D.A., Krieg, P.A., Rebagliati, M.R., Maniatis, T., Zinn, K. and Grenn, M.R. (1984) Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmid containing a bacteriophage SP6 promoter. Nucleic Acids Res. 12, 7035-7056.

Miller, K. (1987) Gel electrophoresis of RNA. Focus 9, 3. Ravelonandro, M., Varveri, C., Delbos, R. and Dunez, J. (1988) Nucleotide sequence of the capsid

protein gene of plum pox potyvirus. J. Gen. Viil. 69, 1509-1516. Teycheney, P.Y., Tavert, G., Delbos, R.P., Ravelonandro, M. and Dunez, J. (1989) The complete

nucleotide sequence of Plum Pox virus RNA (strain D). Nucleic Acids Res. 17, 10115-10116. Varveri, C., Candresse, T., Cugusi, M., Ravelonandro, M. and Dunez, J. (1988) Use of 32P labeled

transcribed RNA probe for dot hybridization detection of plum pox virus. Phytopathology 78, 1280-1283.

Varveri, C., Ravelonandro, M. and Dunez, J. (1987) Construction and use of a cloned cDNA probe for the detection of plum pox virus in plants. Phytopathology 77, 1221-1224.