Diagnostic methods for determining the health of germplasm of clonal crops
Contributors to this section: CIP, Lima, Peru (Carols Chuquillanqui, Segundo Fuentes, Ivan Manrique, Giovanna Muller, Willmer Pérez, Reinhard Simon, David Tay); CIP, Nairobi, Kenya (Ian Barker); FERA, UK (Derek Tomlinson, Julian Smith, David Galsworthy, James Woodhall).
Germplasm of clonal crops is typically exchanged through the transfer of in vitro material. A number of viral, viroid, phytoplasma and bacterial pathogens may be present in such material, in some instances as symptomless, latent infections and these can be transmitted through exchange of in vitro material. The possible movement of such pathogens around the world may have implications for plant quarantine regulations and ultimately consequences for food security, livelihoods and economies.
This document outlines some of the diagnostic methods which can be used to identify pathogens in germplasm material, and highlights some examples. A range of diagnostic methods is typically available for any given pathogen, therefore guidelines on identifying the most suitable method with consideration of infrastructure and human resource capacity are outlined. Critical here is a clear appreciation on the question to be answered about the level of confidence in the identification required, the timeframe within which you are operating and the number of samples that are requiring the test (diagnosis). The choice of method is often a trade-off of between different factors effecting cost, specificity and numbers of samples.
A key to any detection system is measured specificity, sensitivity and proficiency in user competence. Whilst specificity and sensitivity are widely referred to in scientific papers, proficiency in use is often overlooked. Approaches to measuring proficiency are available and may take the form of Ring Testing and Proficiency Testing, which whilst similar, aim to do slightly different things. Ring testing is more aimed at validation of a method amongst laboratories to see if comparable results are achieved across labs and is often time limited; whereas Proficiency Testing aims to show that comparable standards of detection are achieved across laboratories by using test-standards of known status and concentration. With Proficiency Testing the user is free to use any method of analysis; it is the outcome that is important. Proficiency Testing is often part of a laboratories continuous monitoring as is important for designation of laboratory competence in support of an ISO standard (competent laboratory status). More information on Proficiency Testing can be obtained from the PhytoPas (http://www.fapas.com/phytopas.cfm).
Commonly used methods for the diagnosis of plant pathogens
Conventional bioassay testing
These are typically more time and labour consuming compared to the laboratory assays described below. Nevertheless, they can still be very important in the diagnosis, for example, host indexing of viruses and viroids are still often required to verify the result of a laboratory test. Bio-assays are usually carried out in insect-proof compartmented glasshouses or in tropical climates, screen houses are required for raising and holding the plants. Possibly for year round work, growth cabinets, able to provide controlled temperature, light and day length are required. Selection of test plants for routine diagnostic use is dependent upon the type of plants tested. The continued maintenance of reference material is a consideration.
Metabolite analysis (mainly bacterial identification)
The majority of bacterial pathogens can be isolated and grown on media and then subjected to analysis of metabolite composition and properties that provide taxonomic information. Traditionally these methods have involved a sequence of tests, mainly substrate utilization tests, which followed a dichotomous key (See Bacterial ID Key) and an identification would take weeks rather than days to achieve. The methods also required a high level of technical competence for reproducible data to be generated. Alternative formats for substrate utilization assessments have subsequently been developed such as the Biolog system (http://www.biolog.com) that allows for quicker and more reliable identification, often within 48hours of having obtained a pure culture.
Another method for identification of bacteria is the Midi-system (http://www.midi-inc.com) that relies on the identification and quantification of cell wall fatty acids. As with Biolog, an identification can be achieved within 48hrs of obtaining a pure culture. Both Biolog and the Midi-system are commercial products and come with extensive libraries on bacterial species. The level of taxonomic identification achieved by these systems is similar. As a general rule both Biolog and the Midi system allow for reliable species level identification for plant bacterial pathogens. More confident and lower taxonomic (i.e. pathovar) identification can be achieved in some species through the addition of samples to the libraries of these systems
A number of plant pathogens can be detected using serological or immunological methods. This technology is based on the specific reaction between an antibody and its corresponding antigen. It is particularly important for the diagnosis and identification of plant viruses; the molecular simplicity of viruses allows a highly specific response. However, such methods have also been developed for bacteria and more complex organisms such as fungi. Serological methods are based on the unique properties of mammalian and avian immune systems. When foreign material (termed antigen) such as a micro-organism, protein or complex carbohydrate is injected into an animal, the animal immune system responds by making antibodies in its blood serum. These antibodies bind specifically with the antigen that triggered their production.
Antibodies can be polyclonal or monoclonal. Polyclonal antibodies are so called because they consist of many antibodies each with a different specificity, which bind to several different epitopes (binding sites) of the antigen. Monoclonal antibodies contain multiple identical copies of just one antibody binding to one specific epitope. Because of these properties a generalized view on mono and polyclonal antibodies is that monoclonal antibodies provide greater specificity than polyclonal antibodies. The merits of this difference then depend on the intended purpose and target pest and if high specificity is needed or otherwise. Immunoassays can visualize, the antibody-antigen binding, either directly or indirectly. Detecting the quantity of antibody or antigen can be achieved by a variety of methods. One of the most common is to label either the antigen or antibody. The label may consist of an enzyme, colloidal gold (lateral flow assays), radioisotopes, magnetic labels or fluorescence. Other techniques include agglutination, nephelometry, turbidimetry and Western blot.
In laboratory based immunoassays, most of these formats have been superceded by ELISA (enzyme-linked immunosorbent assay). In ELISA, an unknown amount of antigen is fixed to a surface, such as a 96-well plate. This is followed by a blocking step to prevent non-specific antibody binding. A specific antibody is then washed over the surface in order to bind to the antigen. The antibody is linked to an enzyme, and in the final step a substance is added that the enzyme can convert to some detectable signal. Such as in fluorescence ELISA, when the sample is exposed to light of an appropriate wavelength, any antigen/antibody complexes will fluoresce. The amount of fluorescence recorded can be used to determine the amount of antigen present in a sample. ELISA remains one of the most widely used methods for the routine detection of plant viruses although nucleic acid amplification technologies are typically becoming more routine. ELISA are highly sensitive assays that are easily replicated, they can quantify levels of the pathogen and can be automated. ELISA is a robust test, it can be performed in almost any laboratory and requires minimum training.
Although it has many advantages, the drawbacks of ELISA include that the test may only be strain specific and ELISA kits have not been developed for many targets. The production of antibodies can also be resource intensive for example, with the use of live animals, unlike PCR techniques where a wealth of sequence data is available and assays can be designed using a personal computer with internet access. ELISA is also less sensitive than PCR.
Nucleic acid based detection methods
The science about nucleic acid based detection of plant pests has progressed rapidly in the past 20-30 years and many different methods populate the literature. However, in very simplistic terms the vast majority of these methods rely on knowledge of the nucleic acid sequence of an organism and inference of taxonomic specificity (e.g. family, genus, species, strain). The choice of method is strongly influenced by the outcome required; mainly the level of specificity required and the volume of material for testing.
The nucleic acid sequence of any organism comprises regions of varying levels of evolutionary conservation: high, moderate and low sequence conservation. Typically genes that code for essential processes are highly conserved (such as rRNA genes), whereas less critical genes may exhibit moderate conservation and non-coding regions low conservation. These differences in sequence have provided supporting evidence for the current taxonomies of organisms, or otherwise led to taxonomic revisions. These same structures also provide the knowledge base for designing short nucleic acid sequences that are taxon specific (genus, species etc) and can be used either as a probe in hybridization or as a primer in PCR based methods as outlined below. With the revolution in whole genome sequencing now in-train, the level of sequence data available in GenBank is ever increasing and allowing highly intelligent diagnostics to be developed.
The below identifies some of the main and latest nucleic acid based methods as applied to pest diagnostics:
Hybridisation based detection
Based on the principle of bringing nucleic acid sequences of the target (unknown) and test (known) organism together to allow annealing (hybridization) of homologous sequences if present and the production of a detection signal. Generally requires either the target or test nucleic acid to be anchored to a physical matrix, such as nitrocellulose membrane. Detection signal can be radiation or, more recently, a dye-based chemistry. Recent advances have seen major changes in the nature of the physical matrix, and the miniaturization of the detection unit. The microarray exemplifies the current design of hybridization platforms where many 10- 1000s of hybridization units can be contained on the ‘footprint’ of a eppendorf. In these microarray systems the sequence of each unit can be designed and thus microarrays can be tailored to detection systems as required. An example of this type of technology for plant viruses is the Bio-chip project (http://www.bio-chip.co.uk) and on a commercial basis with ClonDiag (http://www.clondiag.com). In these examples, the microarrays are looking for identifications from a single sample.
In many cases the question to be answered requires looking at many samples for a specific pest of quarantine concern. In these circumstances a microarray is not suitable. A more appropriate method for high throughput analysis of many samples is Nucleic Acid Spot Hybridisation (NASH). With this method many samples (often a stem imprint) are spotted onto a matrix (nitrocellulose membrane) and then a test probe specific to the pest of concern is applied under conditioned suitable for hybridization and development of a signal. The method and application is highly analogous to ELISA and micro-plates. Experience with NASH is required in interpretation the signal and the difference between a very weak positive and a mark due to discolouration caused from the sample.
Conventional and Real-Time Polymerase Chain Reaction (PCR) based detection methods
In the context of plant pathology, the application of conventional and real-time PCR to pest diagnosis and characterization has been substantial.
The polymerase chain reaction (PCR) pertains to the directed ‘multiplication’ of template nucleic acid sequence by thermostable enzymes such as Taq DNA polymerase under cycles of heating and cooling that drives DNA denaturation and annealing, respectively. The directed nature of PCR is determined mainly by the nucleic acid sequence of short oligonucleotides, called primers, which work in pairs and ‘prime’ the PCR reaction, and the annealing temperature. The primer sequence dictates the region of the template DNA to be amplified. Where sequence information is known, primers can be designed to amplify target regions, or otherwise they can be designed to amplify randomly. PCR of RNA targets, such as viruses, requires a reverse transcription step to generate cDNA from the RNA template prior to DNA amplification. An overview of PCR is provided here.
The key outcome of PCR is the production of many copies of DNA of identical sequence to that of the template. Depending on the design of the primers, either a single or many amplification products of varying size can result from a single PCR. The bulked DNA can then be the subject of further processing (restriction analysis, sequencing, labeling) or visualized. Traditionally, PCR products have been visualized on an agarose gel. However, an evolution of conventional PCR is real-time PCR that allows real-time product analysis through the monitoring of a chemical reaction that parallels amplification. In real-time PCR, a probe labeled with a reporter dye, locates between the primers and when amplification occurs the dye is activated resulting in reporter signal.
In most examples a PCR based diagnostic is based on the production of a single amplification product, seen on an agarose gel or by a fluorescence reading and recorded as a positive or negative outcome. In addition, more considered PCR diagnostics will include an internal control to check that the reaction has preceded as expected. With conventional PCR the internal control has to yield a fragment of a different size to the pest fragment, and with real-time PCR make use of a different dye chemistry. Some PCR methods also include quantification of the pathogen based on the intensity of the single by comparison to the internal standard.
Other PCR based diagnostics methods have made use of a fingerprint of amplification fragments that is characterized and known to be specific to an organism. A further and major application of PCR and diagnostics is to sequence the amplified fragment and then compare, blast, this sequence against known sequences on GenBank or an in-house DNA sequence library.
Many studies have looked to compare PCR and ELISA based methods. The experience of most laboratories has been that real-time PCR is an order of magnitude more sensitive than conventional PCR and two orders of magnitude more sensitive that ELISA. However, PCR and real-time PCR are both requiring of specialist equipment and consumables; and whilst these costs are coming down they exceed ELISA and can be a constraint in terms of initial capital outlay and running costs.
Loop-mediated isothermal amplification (LAMP)
As mentioned above a disadvantage of PCR is the cost of the specialized equipment required to perform accurately controlled thermal cycling, and in the case of real-time PCR, concurrent monitoring of fluorescence. Loop-mediated isothermal amplification (LAMP) is a method for the detection of specific nucleic acid sequences and has the potential to overcome many of the limitations of PCR-based methods. The ability of LAMP to amplify a target nucleic acid sequence under isothermal conditions removes the need for thermal cycling equipment, allowing testing to be carried out with minimal equipment (a water bath or heated block). Furthermore, simplified methods for the detection of amplification products facilitate the use of LAMP-based methods in the field or in less well-resourced settings.
LAMP is an amplification method which uses 2 pairs of primers (internal and external primers) and a DNA polymerase with strand displacing activity to produce amplification products containing loop regions to which further primers can bind, allowing amplification to continue without thermal cycling. Amplification is accelerated by the use of an additional set of primers (loop primers) that bind to those loops which are of the incorrect orientation for the internal primers to bind. A high level of specificity results from the requirement for primers to bind to up to 8 regions of the target sequence. The use of LAMP has previously been described for the detection of a range of plant pathogens.
LAMP reactions generate a large amount of amplification product that can be detected by conventional agarose gel electrophoresis, by the use of spectrophotometric equipment to measure turbidity, in real-time using intercalating fluorescent dyes, or by visual inspection of turbidity or colour changes. While detection methods based on visual inspection have the advantage of requiring no equipment, assessment of colour or turbidity with the unaided eye is potentially subjective. Equipment-free methods for unambiguous detection of LAMP products would increase the feasibility of using LAMP for detection of phytopathogens outside the laboratory. One such method is the use of lateral flow devices (LFDs) for the detection of labels incorporated into the amplification products.
Like PCR, LAMP can be used to detect RNA targets by incorporating a reverse transcription step to generate cDNA from the RNA template prior to amplification; reverse transcription and LAMP can be carried out in one tube, at a single temperature. The significant advantages of LAMP are therefore: (i) the ability to perform amplification reactions under isothermal conditions, obviating the need for thermal cycling equipment; (ii) the high specificity inherent in a mechanism that requires the recognition of 6 regions (or 8 regions if loop primers are used) of the target sequence for amplification to progress; and (iii) a high efficiency of amplification which generates a very large amount of product in less than 1 hour, allowing the use of novel detection strategies.
Pyrosequencing and next generation sequencing
The above PCR approaches have a reliance on sequence knowledge, designing appropriate primers and a basic appreciation of what the causal organism is. A limitation of the reliance on sequence knowledge is when trying to identify a new pest, or to verify that material is clean of pests. In these situations it would be required to apply one or more primer sets of broad taxonomic specificity in the expectation that one of these would detect a pest if present. However, with this approach a negative result does not remove all uncertainty as an unknown pest may be genetically very distinct and the target sites of the primers may lack sufficient homology for PCR to initiate. A new approach to diagnostics that has particular merit in identifying unknown pests and verifying material is without infection is with next generation sequencing. With these methods universal and randomly targeting primers are used to massively amplify all nucleic acid content within a sample, with the short amplification products then sequenced. Using sophisticated informatic software, these sequences are then analyzed, joined-up to give sequence lengths of potential taxonomic value and the identity of sequence tested for by blast searching. This technology is highly specialized and would be limited to a few laboratories and/or justified only in exceptional cases.
Selection of an appropriate diagnostic method
Once a pathogen is required to be tested, consideration should be given to choosing an appropriate diagnostic method. However, assessing the resources the laboratory presently has, or could have, for testing should also be given. Although whether possessing specific items of equipment e.g. a real-time PCR machine, may be one of the initial considerations, additional though should be given other wider resource related factors, including availability of staff and relative competencies of staff, laboratory space and continued maintenance of capital equipment. Then specific requirements of the test should be considered such as sensitivity, specific and robustness or the assay relative to the data out put of the assay (sequence data, viability data, and qualitative/quantitative results). Finally practical concerns such as time taken to obtain a result and cost of individual test should also be considered. An overview of the relative individual attributes of each type of test is given in Table 1. With these factors in mind, literature searching can then begin on the search for the range of diagnostic assays available. A useful start is an internet search, and then a bibliographic database search (Table 2), usually with the type of method required and the target pathogen.
Click here for the flowchart for identifying and implementing a new diagnostic assay.
References and further reading
Albrechtsen SE. 2006. Testing Methods for Seed-Transmitted Viruses: Principles and Protocols. CABI Publishing, UK. 268 pp.
Boonham N, Tomlinson J, Mumford R. 2007. Microarrays for rapid identification of plant viruses. Annual review of Plant Pathology, 45:307-28.
Boonham N, Glover R, Tomlinson J, Mumford R. 2008. Exploiting generic platform technology for the detection and identification of plant pathogens. European Journal of Plant Pathology, 121:355-363.
Monger W, Alicai T, Ndunguru J, Kinyua Z, Potts M, Reeder R, Miano D, Adams I, Boonham N, Glover R, Smith J. 2010. The complete genome sequence of the Tanzanian strain of Cassava brown streak virus and comparison with the Ugandan strain sequence. Archives of Virology, 155:429-433.
Smith J, Waage J, Woodhall J, Bishop S, Spence N. 2008. The challenge of providing plant pest diagnostic services for Africa. European Journal of Plant Pathology, 121:365-375.
Tomlinson J, Boonham N. 2008. Potential of LAMP for detection of plant pathogens. CAB Reviews: Perspectives in Agriculture, Veterinary Science, Nutrition and Natural Resources 3: 066.