Applied 3/ftcrobiology Biotechnology
Appl Microbiol Biotechnol (1990) 32:572-576
© Springer-Verlag 1990
Short contribution
Highly sensitive and fast detection of Phoma tracheiphila by polymerase chain reaction Franco Rollo, Roberto Salvi, and Pietro Torchia Dipartimento di Biologia Cellulare, via F. Camerini 2, 1-62032 Camerino, Italy Received 27 April 1989/Accepted 18 August 1989
Summary. A new method for the diagnosis of the plant pathogenic fungus Phoma tracheiphila has been developed. The method takes advantage of the enzymatic amplification of a specific 102 bplong target sequence of fungal DNA by the polymerase chain reaction (PCR) using Thermus aquaticus DNA polymerase. The amplified DNA was characterized by agarose-gel electrophoresis, molecular hybridization using a synthetic oligonucleotide probe and direct sequencing. The application of the new method makes possible fast and direct detection of the pathogen in lignified plant tissues, a goal not previously achieved when a cloned probe and a dot-blot test were employed. In addition the PCR test can be used to advantage as a particularly simple and fast way of typing fungal isolates. This is achieved by submitting to DNA amplification crude homogenates of fungal mycelium and analysing the amplified DNA on an agarose mini-gel.
ing vein chlorosis, wilt, shedding of leaves and ultimately die-back of twigs and branches (Nachmias et al. 1979). In recent years we developed a detection assay for this fungus. Our test was based on the use of a selected tract of fungal DNA as a probe (plasmid pPhoB25) and a dot-blot hybridization protocol (Rollo et al. 1987). The dot-blot test assay using plasmid pPhoB25 proved its usefulness in a variety of experimental situations. In particular the test was employed to formulate a precocious diagnosis of the disease in symptomless plants. However, in the last few years the technology of molecular probes has made dramatic progress thanks to the discovery of the polymerase chain reaction, PCR, (Saiki et al. 1988). Because of their superior sensitivity, detection tests based on the in vitro amplification of DNA by PCR are becoming the method of choice for diagnosis of the most elusive infectious agents (Loche and Mach 1988). In the present paper the construction, evaluation and application of a PCR test for P. traeheiphila are described.
Introduction
Detection tests based on the use of specific and sensitive nucleic acid probes are becoming increasingly important as fundamental tools in a number of applied researches aimed at tracking pathogenic microorganisms on a variety of substrates. The so-called "mal secco" is a serious wilt disease which hits lemon orchards in the Mediterranean countries and is caused by the fungus Phoma tracheiphila (Punithalingham and Holliday 1973). After penetration into the canopy through natural wounds and leaf scars the vegetative hyphae of the fungus reach the vascular tissue causOffprint requests to: F. Rollo
Materials and methods Fungal strains, growth conditions and DNA isolation were as previously reported (Rollo et al. 1987).
DNA sequence analysis. Cloned DNA was sequenced as follows: plasmid pPhoB25 was digested with the restriction enzyme EcoRI to excise the fungal insert. The insert was purified on low-gelling agarose, digested to completion with Sau3AI and the fragments shot-gun recloned into BamHI-digested M13mp8. Single-stranded DNA was purified from Escherichia coli HB101 cultures inoculated with recombinant (white) plaques and submitted to DNA sequence analysis using the dideoxynucleotide chain termination method (Sanger et al. 1977). For direct sequencing amplified DNA was fractionated on 2.5% low-gelling agarose. The gel was excised in correspond-
573 B
A
&
C
&
A: 5'GATCCGTACGCCTTGGGGAC3' B: 5'GATCCGAGCGGGACGAGCAG3' C: 5' TCTCTTGCTAAGCCACGCAA3'
Fig. 1. The nucleotide sequence of the 102 bp D N A target derived from the fungal insert of plasmid pPhoB25. The binding sites for the oligonucleotides A, B and C are evidenced ence with the 102 bp band, melted at 65 ° C and 2 Ixl of the melted agarose submitted to 30 amplification cycles using the normal PCR protocol except that one of the two primers used for the initial amplification was absent (unbalanced priming method; Gyllensten and Erlich 1988). This procedure generates a single-stranded D N A preparation suitable for sequencing. Single-stranded D N A was annealed at 50 ° C for 2 h with the amplification primer not used in the previous step, then sequenced as indicated above. Dideoxynucleotide sequencing was carried on using an MI3 sequencing kit (Amersham International, Amersham, Bucks, England). Electrophoresis in 6% polyacrylamide gels and autoradiography were according to standard techniques.
DNA amplification. The reaction mixture for D N A amplification contained 16.6 m M (NH4)2SO4, 67 m M TRIS-HC1, pH 8.0, at 25°C, 6.7 m M MgC12, 1 0 m M /%Mercaptoethanol, 200 ktM each dATP, dCTP, dGTP, dTTP, 170 Ixg/ml bovine serum albumin, 0.3 Ixg each amplification primer, variable amounts of P. tracheiphila D N A as described in the text, and 2.5 units Thermus aquaticus D N A polymerase (Biolabs, New England Biolabs, Beverly, Ma, USA) per 50 ~tl reaction. The amplification was performed in 500-~tl eppendorf-type vials using an automatic D N A amplifier designed and built in our laboratory (Rollo et al. 1988). The amplification cycles were set at 42°C for 6 min (oligonucleotide annealing + elongation) and 90°C for 1.4 min ( D N A denaturation). According to the different experiments the samples were submitted to 30 or 40 amplification cycles. Gel electrophoresis. Following amplification a portion (usually one-fifth or one-tenth) of the reaction mixture was mixed with a loading buffer containing glycerol and bromophenol blue and fractionated on a 2.5% agarose mini-gel in TRIS-borate buffer containing 0.5 txg/ml ethidium bromide. After the run the gel was destained for 30 min .with distilled water, then photographed under UV light and blotted onto a nylon membrane. Oligonucleotide synthesis and labellin#. Oligonucleotide primers of known sequence were synthesized using an Applied Biosystems (Applied Biosystems, Foster City, Calif) 380 A D N A synthesizer. After removal of the protecting groups the oligonucleotides were made radioactive by T4 polynucleotide kinase and 7/[32p]ATP (5' end labelling). Routinely about 50 pmol oligonucleotides were labelled using 100 ktCi 7/[32P]ATP. Southern blots. The nylon membranes were pre-hybridized for 4 h at 45°C in 6 x standard saline-citrate solution (SSC), 5 x Denhardt's 0.5% sodium dodecyl sulphate (SDS) and 1 m g / m l herring sperm DNA. The hybridization was carried out for 2 h
at 45 ° C in a fresh medium of the same composition containing 50 pmol radioactive oligonucleotide. The membranes were washed .twice in 3 x SSC at room temperature and once in 3 x SSC and 0.5% SDS at 45°C then dried and autoradiographed at - 7 5 ° C using Amersham's/3Max film.
Results and discussion
In a previous paper we described the molecular cloning of P. tracheiphila D N A and the selection of a specific clone (plasmid pPhoB25) suitable for use as a probe for diagnosis of the pathogen. In fact we showed that plasmid pPhoB25 D N A produces strong hybridization signals with P. tracheiphila DNA. On the other hand the same probe does not hybridize appreciably with the D N A isolated from lemon plants nor with the D N A of several fungi known to be parasites or epiphytes of the lemon (Rollo et al. 1987). For this reason we decided to use the fungal insert carried by plasmid pPhoB25 as a starting-point to derive a suitable target for a PCR. To this purpose the fungal insert was excised from the plasmid, digested with the restriction enzyme Sau3AI and shot-gun recloned into BamHI-digested M13mp8. Several M13 sub-clones were submitted to sequence analysis using the dideoxynucleotide chain termination method (Sanger et al. 1977). Eventually a 102 bp sub-clone was selected for use as a PCR target. Its nucleotide sequence is shown in Fig. 1. On the basis of this sequence we synthesized three oligonucleotides (20-mers) code-named respectively A, B and C. The function of oligonucleotides A and B is to act as triggers for the PCR; oligonucleotide C is designed to recognize the amplification product by molecular hybridization. To test the PCR system, serial dilutions of total cellular P. tracheiphila D N A were submitted to 30 amplification cycles under the conditions described above. The results are shown in Fig. 2A. It can be observed that all the tested samples yielded a prominent 102 bp amplification band and fainter low molecular weight bands -- evi-
574
Fig. 2 A, B. Southern blot analysis of the amplification products (one-fifth of the total reaction volume) generated by 1 ng (lane 2), 100 pg (lane 3), 10 pg (lane 4), 1 pg (lane 5), 0.1 pg (lane 6) and 0.01 pg (lane 7) Phoma tracheiphila total cellular DNA. A molecular standard (lane 1) was obtained by adding 1 ~tl (Bethesda Research Laboratories) 123 bp ladder with one-tenth (5 lxl) of the amplification product of the 1 ng fnngal DNA. In the present and in the following pictures the arrowhead indicates the 102 bp amplification band. A Photograph of, the ethidium-bromide-stained gel under UV light. B Radioautogram of the nylon filter hybridized with 32p. labelled oligonucleotide C
dence that the amplification reaction is target-specific. It is also noteworthy that the 102 bp band can be produced by as little as 0.01 pg total fungal DNA. This result is considerably better than the sensitivity obtained using the previously developed cloned probe (20 pg). The specificity of the amplification reaction was further confirmed by blotting the gel onto a nylon membrane and analysing the amplification product using the third oligonucleotide (C). An autoradiograph of the Southern blot shows that the oligonucleotide probe hybridizes to the 102 bp band but not to the low molecular weight bands (Fig. 2B). As a final refinement the 102 bp amplification band was excised from the gel, re-amplified using the unbalanced primers protocol (Gyllensten and Erlich 1988) and submitted to direct nucleotide sequencing. For unclear reasons and despite all our efforts, our direct sequencing gels always showed a certain amount of background. Despite this fact we could verify that the amplified DNA closely fitted the target sequence (Fig. 3). As reported above, in a previous work we employed a cloned probe (pPhoB25) and a dot-blot protocol to detect P. tracheiphila in plant tissues (Rollo et al. 1987). The dot-blot test was evaluated on different substrates and found to perform well when applied to relatively soft plant tissues such as those excised from in vitro grown lemon seedlings or from leaves. However the test performed poorly on lignified branch samples. This was appearently due to the difficulty of extracting from this material the amount of DNA needed by this type of test. This problem was circumvented by preincubating the branch samples for 3 days on solidified Sabouraud medium in order to favour growth of the parasite and, consequently, the
Fig. 3. Sequence analysis of the 102 bp band produced by the amplification of total cellular P. tracheiphila DNA. Compare this sequence with the portion of the target sequence (Fig. 1) defined by the two arrowheads. The letter x indicates an unclear site
575
Fig. 4 A, B. Southern hybridization analysis of the amplification products of the DNAs directly isolated from lignified branches of P. tracheiphila-infected (lanes 2~7) lemon plants. Molecular standard (lane 1) as in Fig. 2. A Photograph of the ethidium-bromide-stained gel under UV light. B Radioautogram of the nylon filter hybridized with 32p-labelled oligonucleotide C
amount of DNA extractable. Such a modified protocol achieved reasonably good results. However it required considerable time, about 1 week, due to a number of steps such as sample preincubation, DNA extraction, membrane hybridization and film exposure. For these reasons we thought that the PCR system could be used to advantage to overcome the drawbacks posed by the relatively low sensitivity of the dot-blot test. Lignified branches excised from P. tracheiphi/a-infected or healthy lemon plants Were split in two halves lengthwise and the exposed surfaces scraped with a razor blade. The small splinters of wood obtained in this way were homogenized by mortar and pestle in the presence of a phenol
mixture and submitted to total nucleic acid extraction. Ethanol-precipitated fractions were resuspended in 20 lxl sterile distilled water and 1 p~l aliquots were submitted to DNA amplification (40 cycles) as reported in the Materials and methods. The products of the amplification reaction were analysed by gel electrophoresis and Southern blot. The results are shown in Fig. 4. Specific amplification bands (102 b) were produced by some of the branches excised from the infected plants (Fig. 4A). On the other hand no specific amplification signal was produced by the healthy plants. The Southern blot analysis (Fig. 4B) confirmed the specificity of the 102 bp bands produced by the infected wood samples. During a survey of a lemon orchard severely stricken by the "mal secco" it is not infrequent to spot with the naked eye patches of fungal mycelium growing on trunks, branches or leaves of lemon plants affected by the disease. Small samples of mycelium can then be easily collected. They can be used to confirm the exact nature of the disease and to set up in vitro collections of the pathogen. We wondered whether our PCR test could be adapted to obtain a very simple and fast way of typing the mycelial isolates. To this end very small ( < 1 mg) samples of P. tracheiphila mycelium were put in a 0.5-ml eppendorf-type vial, 5 lxl sterile water was added and the suspension was smashed with a micro-pestle obtained from a pasteur pipette. The vial was put into boiling water for 3 min and then submitted to 1 min centrifugation in a bench centrifuge. Aliquots of the supernatant were submitted to 40 cycles of DNA amplification according to the standard protocol. The amplification products were analysed by agarose gel electrophoresis. All the tested strains of P.
Fig. 5. Amplification of cell sap (2 ~tl) obtained from different field isolates of P. tracheiphila (lane 2-17) or from a P. linoham (lane 18) isolate. The molecular standard (lane 1) was obtained by adding 1 lxl of the BRL 1 kb ladder with 5 ktl of the amplification product of 1 ng purified P. tracheiphila DNA
576
tracheiphila yielded sharp amplification bands -fairly comparable with those produced by purified DNA preparations. On the other hand, no specific amplification band was produced by a control fungus (P. lingham). As the preparation of the fungal supernatant requires a few minutes work only, dozens of samples can be easily and rapidly identified using this procedure (Fig. 5). All of our experiments always evidenced a strict correlation between the presence of the 102 bp band in the amplified DNA and presence of the pathogen in the samples. For this reason we have good grounds for believing that the more sophisticated, costly and time-consuming final checking of the amplified DNA by Southern blot and direct sequencing can be omitted in routine analysis. This type of analysis can be carried out reliably merely by checking the PCR products for the presence of the 102 bp band using a mini-gel system. In conclusion the PCR test displays considerably higher sensitivity than the dot-blot test. It additionally enables the operator to save considerable time (about 5 days) in the analysis of lignified specimens. Finally, thanks to the possibility of utilizing total cell sap for the amplification reaction, testing of field or laboratory mycelial isolates has now become a particularly straightforward operation. Acknowledgements. The authors wish to thank Isabella Di Silvestro, A. I. D. Research Center, Catania, Italy for providing
fungal and plant samples. This research was supported by the Italian Ministr3~ of Agriculture and Forestry, national project "Tecnologie avanzate applicate alle piante'" research project no. 14: "Basi biochimiche e moleeolari della interazione ospite-parassita e metodi di selezione per resistenza".
References Gyllensten UB, Erlich HA (1988) Generation of singlestranded DNA by the polymerase chain reaction and its application to direct sequencing of the HLA-DQA locus. Proc Natl Acad Sci USA 85:7652-7656 Loche M, Math B (1988) Identification of HIV-infected seronegatival individuals by a direct diagnostic test based on hybridisation to amplified viral DNA. Lancet 11:418-421 Nachmias A, Barash I, Buchner V, Solel Z, Strobel GA (1979) A phytotoxic glycopeptide from lemon leaves infected with Phoma tracheiphita. Physiol Plant Pathol 14:135-140 Punithatingham E, Hotliday P (1973) Deuterophoma tracheiphila. CMI descriptions of pathogenic fungi and bacteria no. 399. Commonwealth Agriculture Bureau, 2. Rollo F, Amici A, Foresi F, Di Silvestro I (1987) Construction and characterization of a cloned probe for the detection of Phoma tracheiphila in plant tissues. Appl Microbiol Biotechnol 26:352-357 Rollo F, Amici A, Salvi R (1988) A simple and low cost DNA amplifier. Nucleic Acids Res 16/7:3105-3106 Saiki RK, Gelfand DH, Stoffel S, Scharf SJ, Higuchi IL Horn T, Mullis KB, Erlich HA (1988) Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239:487-49t Sanger F, Nicklen S, Couison AR (1977) DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74:5463-5467
Appl Microbiol Biotechnol (1990) 32:572-576
© Springer-Verlag 1990
Short contribution
Highly sensitive and fast detection of Phoma tracheiphila by polymerase chain reaction Franco Rollo, Roberto Salvi, and Pietro Torchia Dipartimento di Biologia Cellulare, via F. Camerini 2, 1-62032 Camerino, Italy Received 27 April 1989/Accepted 18 August 1989
Summary. A new method for the diagnosis of the plant pathogenic fungus Phoma tracheiphila has been developed. The method takes advantage of the enzymatic amplification of a specific 102 bplong target sequence of fungal DNA by the polymerase chain reaction (PCR) using Thermus aquaticus DNA polymerase. The amplified DNA was characterized by agarose-gel electrophoresis, molecular hybridization using a synthetic oligonucleotide probe and direct sequencing. The application of the new method makes possible fast and direct detection of the pathogen in lignified plant tissues, a goal not previously achieved when a cloned probe and a dot-blot test were employed. In addition the PCR test can be used to advantage as a particularly simple and fast way of typing fungal isolates. This is achieved by submitting to DNA amplification crude homogenates of fungal mycelium and analysing the amplified DNA on an agarose mini-gel.
ing vein chlorosis, wilt, shedding of leaves and ultimately die-back of twigs and branches (Nachmias et al. 1979). In recent years we developed a detection assay for this fungus. Our test was based on the use of a selected tract of fungal DNA as a probe (plasmid pPhoB25) and a dot-blot hybridization protocol (Rollo et al. 1987). The dot-blot test assay using plasmid pPhoB25 proved its usefulness in a variety of experimental situations. In particular the test was employed to formulate a precocious diagnosis of the disease in symptomless plants. However, in the last few years the technology of molecular probes has made dramatic progress thanks to the discovery of the polymerase chain reaction, PCR, (Saiki et al. 1988). Because of their superior sensitivity, detection tests based on the in vitro amplification of DNA by PCR are becoming the method of choice for diagnosis of the most elusive infectious agents (Loche and Mach 1988). In the present paper the construction, evaluation and application of a PCR test for P. traeheiphila are described.
Introduction
Detection tests based on the use of specific and sensitive nucleic acid probes are becoming increasingly important as fundamental tools in a number of applied researches aimed at tracking pathogenic microorganisms on a variety of substrates. The so-called "mal secco" is a serious wilt disease which hits lemon orchards in the Mediterranean countries and is caused by the fungus Phoma tracheiphila (Punithalingham and Holliday 1973). After penetration into the canopy through natural wounds and leaf scars the vegetative hyphae of the fungus reach the vascular tissue causOffprint requests to: F. Rollo
Materials and methods Fungal strains, growth conditions and DNA isolation were as previously reported (Rollo et al. 1987).
DNA sequence analysis. Cloned DNA was sequenced as follows: plasmid pPhoB25 was digested with the restriction enzyme EcoRI to excise the fungal insert. The insert was purified on low-gelling agarose, digested to completion with Sau3AI and the fragments shot-gun recloned into BamHI-digested M13mp8. Single-stranded DNA was purified from Escherichia coli HB101 cultures inoculated with recombinant (white) plaques and submitted to DNA sequence analysis using the dideoxynucleotide chain termination method (Sanger et al. 1977). For direct sequencing amplified DNA was fractionated on 2.5% low-gelling agarose. The gel was excised in correspond-
573 B
A
&
C
&
A: 5'GATCCGTACGCCTTGGGGAC3' B: 5'GATCCGAGCGGGACGAGCAG3' C: 5' TCTCTTGCTAAGCCACGCAA3'
Fig. 1. The nucleotide sequence of the 102 bp D N A target derived from the fungal insert of plasmid pPhoB25. The binding sites for the oligonucleotides A, B and C are evidenced ence with the 102 bp band, melted at 65 ° C and 2 Ixl of the melted agarose submitted to 30 amplification cycles using the normal PCR protocol except that one of the two primers used for the initial amplification was absent (unbalanced priming method; Gyllensten and Erlich 1988). This procedure generates a single-stranded D N A preparation suitable for sequencing. Single-stranded D N A was annealed at 50 ° C for 2 h with the amplification primer not used in the previous step, then sequenced as indicated above. Dideoxynucleotide sequencing was carried on using an MI3 sequencing kit (Amersham International, Amersham, Bucks, England). Electrophoresis in 6% polyacrylamide gels and autoradiography were according to standard techniques.
DNA amplification. The reaction mixture for D N A amplification contained 16.6 m M (NH4)2SO4, 67 m M TRIS-HC1, pH 8.0, at 25°C, 6.7 m M MgC12, 1 0 m M /%Mercaptoethanol, 200 ktM each dATP, dCTP, dGTP, dTTP, 170 Ixg/ml bovine serum albumin, 0.3 Ixg each amplification primer, variable amounts of P. tracheiphila D N A as described in the text, and 2.5 units Thermus aquaticus D N A polymerase (Biolabs, New England Biolabs, Beverly, Ma, USA) per 50 ~tl reaction. The amplification was performed in 500-~tl eppendorf-type vials using an automatic D N A amplifier designed and built in our laboratory (Rollo et al. 1988). The amplification cycles were set at 42°C for 6 min (oligonucleotide annealing + elongation) and 90°C for 1.4 min ( D N A denaturation). According to the different experiments the samples were submitted to 30 or 40 amplification cycles. Gel electrophoresis. Following amplification a portion (usually one-fifth or one-tenth) of the reaction mixture was mixed with a loading buffer containing glycerol and bromophenol blue and fractionated on a 2.5% agarose mini-gel in TRIS-borate buffer containing 0.5 txg/ml ethidium bromide. After the run the gel was destained for 30 min .with distilled water, then photographed under UV light and blotted onto a nylon membrane. Oligonucleotide synthesis and labellin#. Oligonucleotide primers of known sequence were synthesized using an Applied Biosystems (Applied Biosystems, Foster City, Calif) 380 A D N A synthesizer. After removal of the protecting groups the oligonucleotides were made radioactive by T4 polynucleotide kinase and 7/[32p]ATP (5' end labelling). Routinely about 50 pmol oligonucleotides were labelled using 100 ktCi 7/[32P]ATP. Southern blots. The nylon membranes were pre-hybridized for 4 h at 45°C in 6 x standard saline-citrate solution (SSC), 5 x Denhardt's 0.5% sodium dodecyl sulphate (SDS) and 1 m g / m l herring sperm DNA. The hybridization was carried out for 2 h
at 45 ° C in a fresh medium of the same composition containing 50 pmol radioactive oligonucleotide. The membranes were washed .twice in 3 x SSC at room temperature and once in 3 x SSC and 0.5% SDS at 45°C then dried and autoradiographed at - 7 5 ° C using Amersham's/3Max film.
Results and discussion
In a previous paper we described the molecular cloning of P. tracheiphila D N A and the selection of a specific clone (plasmid pPhoB25) suitable for use as a probe for diagnosis of the pathogen. In fact we showed that plasmid pPhoB25 D N A produces strong hybridization signals with P. tracheiphila DNA. On the other hand the same probe does not hybridize appreciably with the D N A isolated from lemon plants nor with the D N A of several fungi known to be parasites or epiphytes of the lemon (Rollo et al. 1987). For this reason we decided to use the fungal insert carried by plasmid pPhoB25 as a starting-point to derive a suitable target for a PCR. To this purpose the fungal insert was excised from the plasmid, digested with the restriction enzyme Sau3AI and shot-gun recloned into BamHI-digested M13mp8. Several M13 sub-clones were submitted to sequence analysis using the dideoxynucleotide chain termination method (Sanger et al. 1977). Eventually a 102 bp sub-clone was selected for use as a PCR target. Its nucleotide sequence is shown in Fig. 1. On the basis of this sequence we synthesized three oligonucleotides (20-mers) code-named respectively A, B and C. The function of oligonucleotides A and B is to act as triggers for the PCR; oligonucleotide C is designed to recognize the amplification product by molecular hybridization. To test the PCR system, serial dilutions of total cellular P. tracheiphila D N A were submitted to 30 amplification cycles under the conditions described above. The results are shown in Fig. 2A. It can be observed that all the tested samples yielded a prominent 102 bp amplification band and fainter low molecular weight bands -- evi-
574
Fig. 2 A, B. Southern blot analysis of the amplification products (one-fifth of the total reaction volume) generated by 1 ng (lane 2), 100 pg (lane 3), 10 pg (lane 4), 1 pg (lane 5), 0.1 pg (lane 6) and 0.01 pg (lane 7) Phoma tracheiphila total cellular DNA. A molecular standard (lane 1) was obtained by adding 1 ~tl (Bethesda Research Laboratories) 123 bp ladder with one-tenth (5 lxl) of the amplification product of the 1 ng fnngal DNA. In the present and in the following pictures the arrowhead indicates the 102 bp amplification band. A Photograph of, the ethidium-bromide-stained gel under UV light. B Radioautogram of the nylon filter hybridized with 32p. labelled oligonucleotide C
dence that the amplification reaction is target-specific. It is also noteworthy that the 102 bp band can be produced by as little as 0.01 pg total fungal DNA. This result is considerably better than the sensitivity obtained using the previously developed cloned probe (20 pg). The specificity of the amplification reaction was further confirmed by blotting the gel onto a nylon membrane and analysing the amplification product using the third oligonucleotide (C). An autoradiograph of the Southern blot shows that the oligonucleotide probe hybridizes to the 102 bp band but not to the low molecular weight bands (Fig. 2B). As a final refinement the 102 bp amplification band was excised from the gel, re-amplified using the unbalanced primers protocol (Gyllensten and Erlich 1988) and submitted to direct nucleotide sequencing. For unclear reasons and despite all our efforts, our direct sequencing gels always showed a certain amount of background. Despite this fact we could verify that the amplified DNA closely fitted the target sequence (Fig. 3). As reported above, in a previous work we employed a cloned probe (pPhoB25) and a dot-blot protocol to detect P. tracheiphila in plant tissues (Rollo et al. 1987). The dot-blot test was evaluated on different substrates and found to perform well when applied to relatively soft plant tissues such as those excised from in vitro grown lemon seedlings or from leaves. However the test performed poorly on lignified branch samples. This was appearently due to the difficulty of extracting from this material the amount of DNA needed by this type of test. This problem was circumvented by preincubating the branch samples for 3 days on solidified Sabouraud medium in order to favour growth of the parasite and, consequently, the
Fig. 3. Sequence analysis of the 102 bp band produced by the amplification of total cellular P. tracheiphila DNA. Compare this sequence with the portion of the target sequence (Fig. 1) defined by the two arrowheads. The letter x indicates an unclear site
575
Fig. 4 A, B. Southern hybridization analysis of the amplification products of the DNAs directly isolated from lignified branches of P. tracheiphila-infected (lanes 2~7) lemon plants. Molecular standard (lane 1) as in Fig. 2. A Photograph of the ethidium-bromide-stained gel under UV light. B Radioautogram of the nylon filter hybridized with 32p-labelled oligonucleotide C
amount of DNA extractable. Such a modified protocol achieved reasonably good results. However it required considerable time, about 1 week, due to a number of steps such as sample preincubation, DNA extraction, membrane hybridization and film exposure. For these reasons we thought that the PCR system could be used to advantage to overcome the drawbacks posed by the relatively low sensitivity of the dot-blot test. Lignified branches excised from P. tracheiphi/a-infected or healthy lemon plants Were split in two halves lengthwise and the exposed surfaces scraped with a razor blade. The small splinters of wood obtained in this way were homogenized by mortar and pestle in the presence of a phenol
mixture and submitted to total nucleic acid extraction. Ethanol-precipitated fractions were resuspended in 20 lxl sterile distilled water and 1 p~l aliquots were submitted to DNA amplification (40 cycles) as reported in the Materials and methods. The products of the amplification reaction were analysed by gel electrophoresis and Southern blot. The results are shown in Fig. 4. Specific amplification bands (102 b) were produced by some of the branches excised from the infected plants (Fig. 4A). On the other hand no specific amplification signal was produced by the healthy plants. The Southern blot analysis (Fig. 4B) confirmed the specificity of the 102 bp bands produced by the infected wood samples. During a survey of a lemon orchard severely stricken by the "mal secco" it is not infrequent to spot with the naked eye patches of fungal mycelium growing on trunks, branches or leaves of lemon plants affected by the disease. Small samples of mycelium can then be easily collected. They can be used to confirm the exact nature of the disease and to set up in vitro collections of the pathogen. We wondered whether our PCR test could be adapted to obtain a very simple and fast way of typing the mycelial isolates. To this end very small ( < 1 mg) samples of P. tracheiphila mycelium were put in a 0.5-ml eppendorf-type vial, 5 lxl sterile water was added and the suspension was smashed with a micro-pestle obtained from a pasteur pipette. The vial was put into boiling water for 3 min and then submitted to 1 min centrifugation in a bench centrifuge. Aliquots of the supernatant were submitted to 40 cycles of DNA amplification according to the standard protocol. The amplification products were analysed by agarose gel electrophoresis. All the tested strains of P.
Fig. 5. Amplification of cell sap (2 ~tl) obtained from different field isolates of P. tracheiphila (lane 2-17) or from a P. linoham (lane 18) isolate. The molecular standard (lane 1) was obtained by adding 1 lxl of the BRL 1 kb ladder with 5 ktl of the amplification product of 1 ng purified P. tracheiphila DNA
576
tracheiphila yielded sharp amplification bands -fairly comparable with those produced by purified DNA preparations. On the other hand, no specific amplification band was produced by a control fungus (P. lingham). As the preparation of the fungal supernatant requires a few minutes work only, dozens of samples can be easily and rapidly identified using this procedure (Fig. 5). All of our experiments always evidenced a strict correlation between the presence of the 102 bp band in the amplified DNA and presence of the pathogen in the samples. For this reason we have good grounds for believing that the more sophisticated, costly and time-consuming final checking of the amplified DNA by Southern blot and direct sequencing can be omitted in routine analysis. This type of analysis can be carried out reliably merely by checking the PCR products for the presence of the 102 bp band using a mini-gel system. In conclusion the PCR test displays considerably higher sensitivity than the dot-blot test. It additionally enables the operator to save considerable time (about 5 days) in the analysis of lignified specimens. Finally, thanks to the possibility of utilizing total cell sap for the amplification reaction, testing of field or laboratory mycelial isolates has now become a particularly straightforward operation. Acknowledgements. The authors wish to thank Isabella Di Silvestro, A. I. D. Research Center, Catania, Italy for providing
fungal and plant samples. This research was supported by the Italian Ministr3~ of Agriculture and Forestry, national project "Tecnologie avanzate applicate alle piante'" research project no. 14: "Basi biochimiche e moleeolari della interazione ospite-parassita e metodi di selezione per resistenza".
References Gyllensten UB, Erlich HA (1988) Generation of singlestranded DNA by the polymerase chain reaction and its application to direct sequencing of the HLA-DQA locus. Proc Natl Acad Sci USA 85:7652-7656 Loche M, Math B (1988) Identification of HIV-infected seronegatival individuals by a direct diagnostic test based on hybridisation to amplified viral DNA. Lancet 11:418-421 Nachmias A, Barash I, Buchner V, Solel Z, Strobel GA (1979) A phytotoxic glycopeptide from lemon leaves infected with Phoma tracheiphita. Physiol Plant Pathol 14:135-140 Punithatingham E, Hotliday P (1973) Deuterophoma tracheiphila. CMI descriptions of pathogenic fungi and bacteria no. 399. Commonwealth Agriculture Bureau, 2. Rollo F, Amici A, Foresi F, Di Silvestro I (1987) Construction and characterization of a cloned probe for the detection of Phoma tracheiphila in plant tissues. Appl Microbiol Biotechnol 26:352-357 Rollo F, Amici A, Salvi R (1988) A simple and low cost DNA amplifier. Nucleic Acids Res 16/7:3105-3106 Saiki RK, Gelfand DH, Stoffel S, Scharf SJ, Higuchi IL Horn T, Mullis KB, Erlich HA (1988) Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239:487-49t Sanger F, Nicklen S, Couison AR (1977) DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74:5463-5467
