Molecular diagnostics of Phytophthora ramorum, causal agent of Sudden Oak Death

Matteo Garbelotto

Introduction and general overview.

Using a molecular diagnostic approach for Phytophthora ramorum can be justified for many reasons. While culturing can be routinely performed with considerable success, several issues related to the traditional isolation-based diagnostics should be mentioned. Isolation success is extremely variable based on plant substrate and time of year. Although isolation techniques are not difficult, they do require special handling of samples (e.g., storage of plates with selective media, freshly obtained samples needed). Also, because new species of Phytophthora are being discovered in exactly the same niches as P. ramorum , molecular confirmation as a back up of morphological identification may be useful anyway. During the cool and rainy winter and spring months, isolation success is high for substrates such as bay laurel leaves, margins of tanoak and coast live oak bole cankers, rhododendron leaves and branches, huckleberry canes, and Pacific madrone leaves and twigs. But as summer progresses, isolation success progressively declines. By the end of summer, for instance, success rate has been shown to have dropped fourfold for one of the easiest available hosts: bay leaves.

The reason for this decline can definitely be attributed to false-negative results, because symptoms are generally still visible on most plant hosts. Besides seasonal false-negatives, several hosts, including bigleaf maple, honeysuckle, and California buckeye, are always extremely hard to isolate from. To date, only one isolation has been made from maple, one from honeysuckle, and none from manzanita, showing the limitations of relying simply on isolation-based approaches.

PCR-based diagnostics generally follow the same trend of success as more traditional isolation-based techniques, but the former provide a more consistent approach because the percentage of false-negatives is much lower. As stated above, for some plant species or for some plant parts for which isolations are still hard to obtain, PCR-based technology provides the only significant approach. Even without considering the difficult hosts, the example of tanoak leaves is most interesting. Tanoak leaves were scored positive for P. ramorum over a year and a half ago, but it took about that length of time to realize that prolonged water soaking was necessary to successfully isolate the pathogen from this substrate. Although baiting can be recommended for soil and water because it takes advantage of the motility of P. ramorum zoospores, we have been able to show that P. ramorum can be diagnosed in these substrates via our PCR-assay.

PCR-based assays are extremely flexible and generally applicable to most substrates. Extracting DNA from plant material in or near the zone of infestation before shipping it to the diagnostic lab may prevent the risks associated with shipment of live material. Another safe way of shipping material is by, rapidly freezing it to -80° C, or freeze-drying it. In that way, it loses its infectiousness, but the DNA of the pathogen will still be viable and detectable through our assay. We have developed approaches that can be executed with minimal PCR technology, currently available in every scientific laboratory. Our traditional PCR approach is subdivided into two components: a) first direct round (moderate sensitivity), and b) second nested round (high sensitivity). We have also modified the basic PCR-based approach to include a quantification component through a SYBR green-mediated Real Time (RT) PCR protocol. Quantification is important for epidemiological and aetiological studies, because it allows us to differentiate between plant material potentially "contaminated' by random deposition of sporangia or other form of inoculum on the plant surface, and plant tissue actually infected by the pathogen. By comparing P. ramorum DNA quantities from samples that were culturable and samples that were unculturable, we found a statistically significant difference. Culturable samples were broadly characterized by larger amounts of DNA of the pathogen. PCR-based quantification thus appears to be extremely more useful than traditional PCR whose output is always of the plus/minus type.

We have also developed a full RT-PCR protocol using a type of assay called Taqman. In Taqman assays a fluorophore reporter and a quencher are attached as modification to a P. ramorum specific DNA probe and will light up only if P. ramorum DNA is present in the sample tested. This approach requires a more sophisticated laboratory setup, but it is much faster and more user friendly. Although extremely sensitive, the detection threshold is lower than that of the traditional PCR approach. To date, our diagnostic assays are based on the ITS sequence of the nuclear ribosomal operon. We are setting up assays based on mitochondrial sequences. Other assays based on mitochondrial sequences, as well as on beta-tubulin nuclear sequences, are available and have been tested experimentally in the laboratory. To date, only our assays have been extensively tested in the field. Other available techniques include traditional PCR primers designed to be specific for P. ramorum , Taqman protocols (RT-PCR), Molecular Beacons (RT-PCR), and a combination of genus specific primers with a single strand conformation polymorphism (SSCP) analysis capable of differentiating P. ramorum DNA from that of all other P. species.

Although we will deal with the drawbacks of taxon-specific primers and relative solutions more extensively below, we summarize here the major issues for all viable molecular detection methods:

• Taxon-specific primers: cross- reactivity if conditions are not stringent or too much DNA templateTaxon-specific primers: cross- reactivity if conditions are not stringent or too much DNA template
• Nested PCR: possibility of contamination when going from first to second round because PCR product of first round can be extremely concentrated and is used as template for second round.
• Taqman and Molecular Beacons: detection threshold in environmental samples is significantly lower than in nested PCR. Sometimes chemistry of reaction is fidgety.
• SSCP: because each Phytopthora species will be identified by a profile of three bands, presence of multiple species in same sample may be confusing. SSCP from environmental samples may also result in spurious bands.

Any assay is only as good as the database that it was created from. All of the listed assays are based on a single genetic locus. One of the intrinsic limitations of these approaches is that they are really detecting an allele rather than an organism. Because different species can exchange genes through introgression and because Phytophthoras are known to do so through interspecific hybridization, it is wise to use a side-by-side approach where morphological traits may indicate genetically different isolates that came out positives using the assay. We are currently trying to develop a multi-locus approach, in which any species will be defined by two or three markers, possibly both nuclear and mitochondrial. Although the major benefit of PCR-based diagnostics is that it can be performed directly on plant material, a culture collection allows for further genetic and manipulation studies.

PCR-based diagnosis is the obvious choice for broad-scale sampling and surveying schemes, because of its sensitivity and reliability, but it need not be the only diagnostic approach. As stated above, culturing has great advantages, and other molecular approaches, such as immunologically based assays, can be extremely useful. ELISA dipsticks, for instance, allow in-field determination of Phytophthora at the genus level. Finally, independently from the PCR assay chosen, the preparation of the sample is critical. Best results are obtained when samples are freeze-dried upon collecting and then pulverized before DNA extractions. Extraction protocols may have to be varied upon substrate. In the case of leaves and twigs, phenol-chloroform extractions followed by chromatography using silica affinity of DNA have produced excellent results. When dealing with wood, protocols for the extraction of DNA from stools appear to be more effective.

Diagnosis of P. ramorum via PCR.

DNA for PCR-based detection can be extracted directly from plant tissue; alternatively, PCR can be used on cultures to confirm their identity. While many different protocols can be adopted for DNA extraction, we describe one that can be applied to all types of plant material, including wood, with a high success rate. Symptomatic leaves, twigs, or bark cankers are first frozen and then lyophilized. As in the case of culturing, plant tissue at the margins of the infection should be selected for the extraction. Approximately 10-100 mg of plant tissue are placed in a 1.5 microfuge tube and ground into a fine powder by placing two stainless steel ball-bearings in the tube and shaking it vigorously in an amalgamator for several seconds. The extraction procedure involves hydrating the plant material in a CTAB extraction buffer, using a phenol-chloroform-isoamyl solution to eliminate the proteins, and binding the DNA to silica beads according to manufacturer's instruction of the Geneclean Turbo Nucleic Acid Purification Kit. DNA extracts are then diluted 1:100-1:200 in 0.1X TE buffer before setting up the PCR reaction. In the case of cultures, it is sufficient to place a minute amount of hyphal tissue in a microfuge tube containing 100-200 uL of .1X TE, vortex it, pellet the cells down by a 2-minute spin in a microfuge, and use the supernatant as template DNA. A single round of 35 cycles is normally sufficient to detect P. ramorum DNA in Rhododendron spp, and bay laurel leaves. Diagnosis from wood and from other hosts requires more sensitive approaches such as increasing the number of cycles to 40 or a second round of nested PCR using the PCR products from the first round as template.

PCR reaction.
For best detection, two rounds of PCR are necessary. The first round is executed using primers Phyto1 (CATGGCGAGCGCTTGA) and Phyto4 (GAAGCCGCCAACACAAG) and reagents described in Table 1.

TABLE 1: Recipe of the cocktail to be used for the PCR-diagnosis of P. ramorum

Reagent	Stock Conc.	Volume (ul)	Final Conc.
PCR Water	N/A	12.25	N/A
PCR buffer (no Mg)	10X	2.50	1X
DNPTs	2mM	2.50	0.2mM
MgCl2*	50mM	0.75	2.0mM
Phyto1	50uM	0.25	0.5uM
Phyto4	50uM	0.25	0.5uM
Taq polymerase	5U/μl	0.25	0.05U/μl
Template DNA		6.25
Total Volume		25

PCR parameters should be as follows:
1 cycle at	94C for 1'25"
34 cycles at	93C for 35"
	62C for 55"
	72C for 50" adding 5' at each cycle
1 cycle at	72C for 10'
Ramp rate:	3.3°C/s heating, 2.0°C/s cooling

The expected PCR product is 687 bp long. The size of the amplicon can be easily verified by agarose gel electrophoresis using the following conditions: 1.5% agarose, 4V/cm, 1 hour run time. For best detection, dilute the PCR product 1:500 and perform a second PCR run using the same parameters described above, but primers Phyto2 (AAAGCCAAGCCCTGCAC) and Phyto3 (GGTGGATGGGGACGTG) instead of Phyto1 and Phyto4. The nested procedure can be four times more sensitive than the single-round approach. Nested PCR products should be 291 bp long. No cross-reactivity with other Phytophthora species has been noted with the exception of P. lateralis and P. cambivora, but only when DNA of these species are present in high concentration. If cross-reactivity with P. lateralis and P. cambivora is an issue, the problem can be easily eliminated by determining concentration of the template DNA and by diluting the sample to concentrations at which no cross-reactivity occurs. This diagnostic protocol, in fact, will only detect P. lateralis and P. cambivora at concentrations of DNA = 96 pg/uL and 56 pg/uL, respectively. These concentrations of pathogen DNA are rather unlikely to occur in plant tissue. P. ramorum DNA, on the other hand, will be detected at concentrations ‹.152 pg/uL, a six-order magnitude difference in detection threshold when compared to the other two species. Template DNA can be quantified before PCR in a variety of ways including spectrophotometry, densitometry, and fluorometry. As a rule of thumb, a 1:1000 dilution of a DNA extraction from plant tissue will ensure pathogen DNA concentrations are below the threshold of 56 pg/uL.

Precautions and controls.
At all stages (material preparation, lyophilization, DNA extraction, first PCR round, nested PCR round), negative control samples should be included. Extreme care should be taken when handling and diluting the first round of PCR because DNA contamination is likely to occur at this stage. If in doubt about the efficiency of either the DNA extraction or of the PCR reaction stages, the following universal primers can be added to the traditional PCR reaction. Control primers will amplify an amplicon approximately 175 bp in size from all plant and fungal material, and will not interfere with P. ramorum primers. Control primers UnivForw (ggaacgtgagctgggtttag) and UnivRev (ttctgacttagaggcgttcag) can be added to reach a final 0.5 uM concentration in the traditional PCR approach. Currently, DNA sequencing is extremely inexpensive and both the 687 bp and the 291 bp amplicons are informative enough to differentiate P. ramorum from all other Phytophthora species. We recommend sequencing confirmation for critical samples. Although the diagnostic procedures here described are highly specific for P. ramorum , unwanted deviations from the described protocols due to varying quality of reagents and DNA extracts, or to pipette and Thermocycler malfunction, may result in non-specific amplification.

Several other PCR-based procedures based on a range of nuclear and mitochondrial sequences are now available, including Real Time PCR , single strand conformation polymorphism (SSCP) detection, as well as other traditional PCR protocols. The procedure described above in detail can be employed by all laboratories, without sophisticated equipment, and is the only technique based on primers fully matching P. ramorum DNA that has been extensively tested with environmental samples.

Shortcomings and pitfalls of PCR-based diagnostics.

Cross-reactivity is a distinct possibility even for the best designed Taxon-specific TS-PCR protocol, especially in the course of molecular ecology studies based on real-world environmental samples. Furthermore, undetected malfunction of the thermalcycler may also result in unwanted cross-reactivity. It is therefore of paramount importance to implement a system of product confirmation, to verify that no cross-reactivity occurs. Several options are available to the researcher. If the target PCR product is unique in size, careful determination of amplicon size through gel electrophoresis may suffice. If amplicon length is not unique, nucleotide sequence needs to be. To ensure the amplicon belongs to the target organism, at least four approaches are currently available. If Real Time PCR technology is available (see below), it may be possible to select an amplicon characterized by a specific melt temperature. If high resolution electrophoresis is available, it may be possible to use single strand conformation polymorphism (SSCP) to ensure product specificity. The amplified fragment may be characterized by the presence of unique endonuclease restriction sites and thus its identity could be verified by RFLP analysis. Finally, sequencing the entire amplicon is now easy, fast , and inexpensive. If cross-reactivity occurred, the amplicon sequence will not match that of the target organism. There are intrinsic limitations to the size of amplicons that can be easily verified with each method. For the melt temperature approach, optimal length is about 200 bp (range 100-300); while for SSCP and sequencing, it is fastest to use amplicons with sizes less than 500 bp. RFLP analysis has no significant size limitations.

Because TS-PCR is designed to specifically amplify target organisms, while basically ignoring all non-target organisms, it can be a somewhat 'reductionist' and potentially misleading tool in the hands of the mycologist. This is particularly true when the TS assay is based on a single locus. There is mounting evidence of horizontal transfer of genes across taxa; therefore, associating the diagnosis of a species to the detection of a single gene may be deceiving, because the gene in question may be found in species different from the target one because of rare introgression events. This becomes a serious issue when diagnosis is done directly on plant tissue, without the ability to support the diagnosis by observation of a live culture. It is now understood that interspecific hybridization is fairly common even amongst the fungi and the oomycetes, and that first generation hybrids often bear phenotypic traits that are very different from those of either parents. Unfortunately, when using a TS assay developed for species A, results will be identical to those obtained when processing individuals belonging to species A, and hybrid AB individuals. Using assays based on multiple and unlinked loci (e.g., a multiplex reaction based on two or more loci that are diagnostic for the target organism) may be helpful, but may ultimately not resolve the problem. A combination of genus-specific (as most hybrids do occur within a genus) and species-specific assays may be required to solve this issue.

To be sure that a negative result is really due to lack of DNA from the target organism, it is necessary to have a positive control primer set in the TS-PCR assay. This primer set should consist of primers that a) will not interfere with the TS-primers, and b) will universally amplify plant hosts and/or the entire group of microorganisms related to the target one. In this case, if no amplification at all will occur, it may be not necessarily due to the lack of target DNA, but simply to a bad DNA extraction.

Finally, although DNA of the target organism is detected in a sample, it is often unclear whether the pathogen may still be viable in the sample. This is an important detail when diagnosis is performed in the context of epidemiological studies or regulatory implementations. Quantification of template target DNA in the sample (see below) may provide a partial solution to this problem. It may be possible to determine, for each pathogen/host combination, thresholds of DNA concentrations that correspond to viable, hence infectious, presence of the microbe.

Conclusions.

TS-PCR is an invaluable tool for the diagnosis of pathogens, fungi in general, and for molecular ecology studies. Because of its recent broad application, TS-PCR still awaits the definition of clear standards and procedural protocols. It is extremely important that researchers exhaust all possible explanations of TS-PCR results, as well as perform all necessary positive and negative controls, before translating such results into diagnostic conclusions. Nevertheless, PCR-assisted diagnostics is the most powerful tool available to us to identify fungi and needs to become the standard of all research or regulatory action in which detection of fungi is involved. The value of diagnosis based on culturing is not in discussion here. Molecular diagnostics are needed especially where culturing is unsatisfactory or impractical. To avoid the risk of creating two systems without cross-referencing, one based on isolations and one on DNA-assays, I suggest a DNA-based assay should be developed and standardized for all organisms of interest. This approach may create a body of compatible and highly comparable diagnostic systems. Other approaches will be needed depending on the type of questions asked, and may be easily superimposed on the molecular approach.