EXPLANATORY DOCUMENT for the Validation of Detection Methods for Plant Pathogens and Pests

Transcription

1 EXPLANATORY DOCUMENT for the Validation of Detection Methods for Plant Pathogens and Pests Dr.Ir. R.A.A. van der Vlugt, Ing. M. Verbeek & Dr. P.J.M. Bonants Report 135EN

2

3 EXPLANATORY DOCUMENT for the Validation of Detection Methods for Plant Pathogens and Pests Dr. Ir. R.A.A. van der Vlugt, Ing. M. Verbeek & Dr. P.J.M. Bonants Plant Research International B.V., Wageningen May 2007 Report 135EN

4 2007 Wageningen, Plant Research International B.V. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of Plant Research International B.V. Copies of this report can be ordered from the (first) author. The costs are 50 per copy (including handling and administration costs), for which an invoice will be included. This Explanatory Document has been drawn up by Plant Research International (PRI, Wageningen, The Netherlands) for the Dutch Ministry of Agriculture, Nature Management and Food Quality and concerns the Validation of Detection Methods for Plant Pathogens and Pests. The document represents a sector-wide discussion on the problems relating to the validation of detection methods for plant pathogens and pests, including viruses, bacteria, fungi, nematodes, insects and other pests. Plant Research International B.V. Address : Droevendaalsesteeg 1, Wageningen, The Netherlands : P.O. Box 16, 6700 AA Wageningen, The Netherlands Tel. : Fax : info.pri@wur.nl Internet :

5 Table of contents page 1. Introduction 1 2. Validation of detection methods for plant pathogens and pests What is validation? Methodology, method and analysis The validation of analyses Procedure Determining the purpose of the analysis Is the analysis to be validated qualitative or quantitative? Is the analysis to be validated a reference method? Determining the performance characteristics 7 3. Definitions Summary of definitions of performance characteristics Summary of other definitions Validation material Reports Validation report List of recommendations 21 Appendix 1. Details of Validation Schedule for Pepino mosaic virus (PepMV) 23 B1.1 DAS-ELISA PepMV 23 B1.2 RT-PCR PepMV 28 B1.3 DAS-ELISA Protocol for PepMV 32 B1.4 RT-PCR Protocol for PepMV 34 Appendix 2. Validation report 35 B2.1 Validation report for DAS-ELISA to detect Pepino mosaic virus in tomato leaf 35 B2.2 Validation report for RT-PCR to detect Pepino mosaic virus in tomato leaf 41 Appendix 3. Terms and definitions NEN Appendix 4. References and Literature consulted 53 B4.1 References 53 B4.2 Literature consulted 53 B4.3 Bibliography (for further study) 54

6

7 1 1. Introduction During the past 20 to 30 years there have been massive developments in the area of research and development of all manner of fast analysis methods, particularly in the food sector. Analysis methods to detect all kinds of ingredients, but also a wide range of pathogens in many different foods. Such methods can provide a considerable advantage compared to existing methods, because they are more sensitive, faster, cheaper etc. For microbiological analysis, the traditional plating methods have been replaced by ELISA and/or molecular methods such as PCR. However, before methods can be used, they must be subjected to a thorough validation procedure. In the Netherlands, many plant pathogens and plant pests occur in the cultivation of all kinds of crops, and our international trading position means that there is a real risk that new pathogens will be introduced. These diseases and plagues form a serious threat to the cultivation of and trade in plants and plant products, to the natural environment and public green spaces. In order to protect crops (or natural ecosystems) and to safeguard trade, the most harmful diseases and plagues are regulated at an international level (quarantine status). Detection and identification methods play a crucial role in monitoring and controlling these diseases and plagues. In the phytopathology sector in the Netherlands, a number of laboratories, services and institutions are active in the development and implementation of such methods for plant diseases. These concern many different organisms such as viruses, bacteria, fungi, nematodes and insects. Within the Netherlands, millions of samples of this wide range of organisms are analysed each year using various methods. The biology of all these organisms is highly diverse, though, as are the methods used to detect them in a wide-ranging series of matrices. The Plant Protection Service plays a key role in preventing the introduction above all of Q(uarantine)-organisms in the Netherlands, whereas the inspection laboratories play a role in assessing the plant (propagating) material for the presence of a variety of organisms. In the Netherlands, Plant Research International (PRI) (previously the Research Institute for Plant Protection (IPO-DLO)) plays an important role in the development of detection methods. Decades of research experience on a variety of plant pathogens have led to the development of many diagnostic and detection methods. These have already been used some considerable time within the Netherlands and abroad by inspection institutions and businesses. In the past, many methods were developed based on the use of antisera (ELISA, IF), particularly against viruses and bacteria (IF). For a large number of viruses and bacteria, these methods are still the reference framework and, in many countries form the cornerstone of phytosanitary policy. For those plant pathogens for which no reliable antisera are available (as is the case for many important fungi), serological detection techniques are not readily applicable. In the last ten years the molecular methods have entered the world of diagnostics and detection methods have been developed based on DNA/RNA. The developments in this area are progressing fast, with a range of new technologies being anticipated on the market during the coming years. Although molecular technologies for the detection and identification of plant pathogens are already (semi-) routinely used in a number of places (and are even available commercially), it is not clear whether for these tests only a comparison on sensitivity has been performed. Before new (molecular) methods can be implemented in practice, such as by the Plant Protection Service and the practical laboratories, such methods need to be properly validated and compared to existing methods. This is necessary because each method, but also each pathogen, has its own characteristics. Criteria that fulfil an important role here are sensitivity, specificity, robustness, the number of false negatives/false positives, the extraction method used and type of plant material (roots, leaf, wood, flowers, etc.). Experience has also shown that not every new (molecular) analysis can be introduced as standard in every laboratory. An implementation phase as well as a much higher level of quality assurance is required compared to ELISA, for example. All the organisations involved in the development, validation, implementation and use of such detection methods for plant diseases have indicated that a uniform system for validating new detection/diagnostic analyses for plant pathogens is required, partly due to the fact that many plant pathogens have their own characteristics.

8 2 Analysis methods offered for accreditation must be validated. For the structure of the validation method, reference is made to the Dutch standard NEN 7777: Environment - Performance characteristics of measurement methods. This standard is intended for the validation of physical and chemical methods. The validation of analysis methods in other areas of application can often not be carried out in entirely the same way as in chemistry or physics. Indications for validations in certain areas of application can be included in an explanatory document (an example of this is Accreditation Council (RvA)-T02, an explanatory document for microbiology). In order to arrive at a generally acceptable validation system of new analyses for plant pathogens, the project group feels it is necessary to involve end users in practice in the development and implementation of such a system. It is also essential to make use of the experiences that have been acquired outside the area of plant pathogens in the validation of PCR-based analyses. Particularly in human and animal virology, PCR diagnostic techniques as well as ELISA are used on a much larger scale. These experiences have been brought together and, together with the literature study carried out and in consultation with the Plant Protection Service and other end users, these should lead to an initial setup for the system to perform validation experiments regarding the detection of plant pathogens. During the course of the Validation Project ( ) and at the organised workshop, intensive discussions were held with the end users of analyses and a clear consensus for the validation system was created. Ultimately this should lead to a system that is acceptable to all parties, in which new analyses can be incorporated safely and responsibly in the existing assessment and inspection systems and under the current quality assurance systems. Furthermore, it should be noted that plant pathogens have their specific characteristics in relation to ecology and their dispersal in the various substrates (plant, soil, air, water) in which they occur. The time of sampling and the nature of the sample to be analysed may also be highly diverse. For this reason, knowledge concerning the biology and epidemiology of the pathogen in question is extremely important. Control samples and necessary reference material is in many cases not available, or only to a limited extent. All these aspects will be taken into account in the way in which new or existing analysis methods must be validated. This document, drawn up by PRI and with additional comments by the supervising committee, comprising subject experts in plant diseases from all relevant laboratories in the Netherlands, meets the need for a further explanation and reference framework with the validation of analyses that will be used for the detection of plant pathogens.

9 3 2. Validation of detection methods for plant pathogens and pests 2.1 What is validation? Validation means demonstrating that the analysis method described is suitable for the intended application, i.e. it meets the requirements set concerning the application. The relevant performance characteristics must be established, depending on the status and type of method. Validation can be divided into two approaches: Process validation and Method validation a. Process validation A good definition of process validation is given by the Food and Drug Administration: Establishing documented evidence which provides a high degree of assurance that a specific process will consistently produce a product meeting its predetermined specifications and quality attributes. The main aspects of this definition are: documented evidence: recording objective evidence in writing; high degree of assurance: based on statistical methods; specific process: describing processes and subprocesses unambiguously; consistently produce a product: the same product/result time after time, within margins; predetermined specifications: set beforehand, not afterwards. With this approach towards validation, it is assumed that the entire process ultimately produces a product. Each sub-process has its own evaluation: a separate validation report or qualification. b. Method validation Another approach is the method validation, whereby it is not so much the entire process that is examined, but specifically the particular method. The definition of method validation is as follows: Validation involves documenting, through the use of specific laboratory investigations, that the performance characteristics of the method are suitable and reliable for the intended analytical purposes (FDA & US Department of Health and Human Services). In practice, often a combination of method and process validation will be used. After all, while you may very well be able to validate a particular analysis, you must also have your equipment in order, employees must be properly trained, the analysis methods will have to be maintained (re-analysing batches, for example) etc. in short, often there will have to be (some form of) accreditation by an institution. This explanatory document focuses in particular on method validation. Process validation is a condition which an accredited institution must meet and is described in detail in ISO/IEC Methodology, method and analysis This document assumes that the method validation is used on analyses and not on methodologies or methods. Even though the methodology (the principle on which the analysis is based) or the method (procedure) is a general one and well described in the literature, the specific analysis must be validated nevertheless. As an example: an analysis for a specific pathogen is based on the immunology methodology and the DAS-ELISA method. When the performance characteristics of this analysis (a DAS-ELISA with specific antibodies against the pathogens) have not been determined before, the analysis will have to be validated as a new analysis.

10 4 2.3 The validation of analyses The validation of an analysis is defined as demonstrating that the analysis is suitable for the intended application. Analyses must be validated each time it is necessary to verify that the performance characteristics are suitable for the application in relation to a specific analytical problem. This is the case, for example, for any new analysis that has been developed for a particular purpose, but also for existing analyses that are modified or extended, or, according to the quality control, have been changed or are carried out with different equipment. The extend of the re-validation will depend on the changes. A report must be made of each validation study. 2.4 Procedure For validating analyses for plant pathogens, the following validation plan is applicable: 1. Determine the purpose of the analysis (name the measurand, the object to be measured / area of application, the analysis method) 2. State whether the result of a measurement will be interpreted qualitatively or quantitatively 3. Determine whether the analysis is a reference method (i.e. not only the methodology/method, but an analysis applied to the target pathogen) and, if so, whether the performance characteristics of freedom from bias, selectivity, specificity, measurement range and robustness are sufficiently described or generally accepted by the professional group (e.g. washing method for soil samples) 4. Determine on the basis of steps 2 and 3 which performance characteristics must be determined 5. Determine for each performance characteristic the sub-areas of application (e.g. for different matrices) for which a separate validation study is required 6. Find out whether any external requirements have been laid down for certain performance characteristics absolute limit values (fixed requirement) standard values (from an earlier validation study) 7. Find out which samples are needed for the validation study (n 8 and see table 1) 8. Perform the validation study for individual performance characteristics or in a combined test setup (see table 1) 9. Assess the performance observed compared to any external requirements. If there are no external quantitative requirements, assess directly compared to the purpose 10. Present the results in a validation report Figure 1 sets out in brief how the choice is made for determining performance characteristics. The aim is to achieve as clear a flow chart as possible, in which the performance characteristics to be determined can be easily read off after going through the following decision moments. This flow chart (figure 1) explicitly assumes that a clear choice has already been made for the analysis to be used. Adjusting the analysis during the validation process is not recommended. The following paragraphs provide a more detailed explanation of these steps.

11 5 Figure 1. Flow chart of analysis validation.

12 Determining the purpose of the analysis The first step is to determine the purpose of the analysis (scope). Often this depends on the client. An essential requirement is that not too much should be promised and that the analysis must be limited to a specifically described application (definition). A good description needs to be provided of the analysis method to be used for which pathogen, which crop (or part of crop), which matrix, under which circumstances and what the result of the analysis means in the context concerned (determining that pathogen in that matrix with a particular method, for example: the identification of Plum Pox Virus (PPV) in plum sepals, cultivar xxx, using DAS-ELISA (for this, see RvA-T25). In the analyses of plant pathogens, four analysis categories can be identified in practice: 1. As much certainty as possible about presence or absence: Quarantine organisms (Q-organisms) 2. Quality organisms (K-organisms) with a max. % permissible infection (e.g. of a virus in seed potatoes) 3. K-organisms, only determining presence or absence (do I have a problem or not) 4. diagnosis (what is my problem?) A typical feature of the first 3 categories is that you deliberately test for a predetermined pathogen with a test or analysis method suitable for that pathogen. This comes under the term Detection. The fourth category comprises Diagnosis, the attempt to identify the cause of a disease. It makes sense to recognise these analysis categories and take them into account in determining the scope of the analysis. REMARK This document must be regarded as a general blueprint according to which analysis methods can be validated. However, every pathogen/plant combination, because of its specific properties, places special requirements on the analysis and the way in which it must or can be carried out, and possibly too on the way in which certain performance characteristics must be determined. It is not possible to provide a detailed description for each combination in this document Is the analysis to be validated qualitative or quantitative? Often a method (e.g. ELISA) can be used for both a qualitative assessment (pathogen present or absent) and a quantitative assessment (a certain quantity of the pathogen is present). In none of the Dutch-language standards a clear definition is given of qualitative or quantitative, which can lead to confusion. To clarify this matter, we propose defining the terms as follows, regardless of the method used: Qualitative: establishing the presence or absence of a plant pathogen in a sample. Quantitative: establishing a certain quantity of a plant pathogen in a sample Is the analysis to be validated a reference method? Many of the analyses used for plant pathogens can either be found in the literature (scientific or other sources, e.g. EPPO protocols) or are generally accepted in practice (or by standardisation institution = reference lab = Plant Protection Service). Although analyses can be found in scientific journals, often no thorough analyses of performance characteristics are presented. For the performance characteristics selectivity, specificity, measuring and robustness it must be examined whether one or more of these characteristics are determined in the literature (or practice) in the correct way as described in this document. This must also be clearly described for the analysis to be validated (protocol and circumstances when performing the analysis). If this is the case, the performance characteristics do not need to be determined in the validation study. If not, the performance characteristics concerned must be determined.

13 7 REMARK The literature must describe the analysis to be used. Literature on the methodology or method used cannot be regarded as the source for the analysis validation data. REMARK It may also be that an analysis that was already validated is modified on minor points (= in accordance with the reference method). If an analysis is modified it must be revalidated, focusing on those performance characteristics that may reasonably be affected by the change. Therefore describe precisely what the modification involves and what performance characteristics may possibly be affected by this modification. These must then be redetermined. When, for example, another crop is analysed, this means that a number of performance characteristics, particularly selectivity, must be re-examined (different matrix). When the purpose of the analysis (scope) is totally changed, in principle this means a new analysis and should be regarded as a major change Determining the performance characteristics Appendix 1 gives an overview of the Validation Schedule for Pepino mosaic virus and an explanation on determining the performance characteristics for two analyses (DAS-ELISA and RT-PCR) on Pepino mosaic virus in tomato leaf. Sometimes external requirements are imposed on performance characteristics, e.g. because clients set a certain threshold value. If this is the case, the validation report must state that the particular performance characteristic complies with the external requirements. Performance characteristics do not need to be determined in separate studies, but may be determined in a single, combined study. A diagram of an analysis schedule over eight days is given in Table 1 (according to NEN 7777).

14 8 Table 1. Example of schedule for a minimum study to determine jointly the performance characteristics. Laboratory sample Performance Day Day Day Day Day Day Day Day characteristic Laboratory sample I repeatability/ xx x reproducibility Laboratory sample 2 repeatability/ reproducibility x xx Laboratory sample 3 repeatability/ xx x reproducibility Laboratory sample 4 repeatability/ reproducibility x xx Laboratory sample 5 repeatability/ xx x reproducibility Laboratory sample 6 repeatability/ reproducibility x xx Laboratory sample 7 repeatability/ xx x reproducibility Laboratory sample 8 repeatability/ x xx reproducibility Laboratory sample close to detection limit detection limit x x x x x x x x Reference material freedom from bias / x x x x x x x x specificity Standard model deviation x x x x x x x x Sample(s)/ sample(s) with deviating matrix selectivity xx sample(s)/ sample(s) under deviating robustness xx circumstances x : single determination xx : duplo determination in repeatable circumstances See Chapter 3.3 for the choice of the validation material.

15 9 3. Definitions Definitions of performance characteristics play an important role in the process of analysis validation. What do you want to show in what, how and under what conditions? When studying the literature, it becomes apparent that various definitions are used for the same terms. Definitions for the various performance characteristics are not always entirely unambiguous. The committee supervising this project has stated that the NEN7777 standard should be regarded as the guideline. Appendix 3 sets out the definitions as found in this standard. Since NEN7777 focuses on chemical substances and analyses, its definitions do not always comply with analyses from the plant pathogen sector. Some modification to these definitions is needed due to the specific problems of this sector. The definitions in NEN7777 are therefore interpreted by the project group for the plant pathogen sector, taking into account the feasibility and usefulness of determining some performance characteristic (e.g.: can the concentration of a pathogen indeed be determined?). Appendix 3 goes on to give additional terms in respect of detection and diagnosis. These definitions have been drawn up by the Royal Netherlands Society of Plant Pathology (KNPV) and published in the Glossary of Terms: list of crop protection terms. Gewasbescherming year 28, supplement 1, Chapter 3.1 describes the NEN7777 definitions for the relevant performance characteristics, followed by relevant explanations/comments from the field of plant pathogens. This is followed by a new definition for that particular performance characteristic as proposed by the project group. Finally, details are described for each performance characteristic. It is strongly recommended that these new definitions be used in the future. 3.1 Summary of definitions of performance characteristics Freedom from bias, Trueness NEN7777 DEFINITION Ability of a method of measurement to give indications without systematic errors [NPR 2814]. REMARKS The definition of freedom from bias, trueness in a variety of standards assumes an average value of a series of observations. Only NEN7777 avoids using the word average (ability to give indications without systematic errors). With this definition, the question may be asked whether freedom from bias is indeed appropriate for a qualitative test, which does not involve establishing the quantity of a pathogen. The actual question here is whether the analysis does actually do what it says. Is the desired organism actually detected? Have I not missed any isolates? The problem, however, is that many laboratory analyses are used in situations precisely where it is not clear beforehand that the organism is present (e.g. symptomless viruses, cultivars that are difficult to assess) and a comparison with a visual assessment is not possible. Consequently knowledge concerning the pathogen, sampling, matrix effects and the time of the year are extremely important. In order to assess the results of the laboratory analyses for freedom from bias (the number of items in a sample assessed as diseased), a comparison could be made between the analysis to be validated and another analysis based on a different methodology (another analysis, or observations of symptoms, for example). The freedom from bias can then be derived from the percentage of the samples that has the same result in both methods (positive or negative), ideally 100%. Within this framework, we could mention diagnostic sensitivity and specificity (see under specificity), but in that case more than 8 samples are needed to make statistically significant statements. Moreover, these analyses must then

16 10 be made with material for which it has been established that they concern the correct pathogen, such as those that occur in a reference collection. This is not necessary if the second method is the reference method. EDVDP DEFINITION Freedom from bias is the ability of the analysis to do what it says it will do (e.g. detection of organism A in matrix X). PROCEDURE Carry out the analysis (n 8) on the target organism obtained from a high-quality reference collection (if possible), or on a sample for which it has been established (validated and based on another principle) in another test that it contains the target organism. In practice, positive controls are also used that are always included during the performance of the method. See also Tables 1 and 2. Limit of detection NEN7777 DEFINITION Lowest concentration of the component in the laboratory sample, of which the presence still can be established with a certain degree of reliability. REMARK The limit of detection cannot always be established absolutely while detecting plant pathogens. Often the limit of detection is defined in terms of the smallest detectible quantity. For chemical substances (NEN7777) this does not need to be a problem because these are often available in a pure form with a precisely known concentration; however for plant pathogens this is not so clear. Often when establishing the smallest detectible quantity, dilution series are used. However, in that case the start concentration of the substance to be detected must be known. For plant pathogens this is certainly not always possible. Furthermore, there are also organisms that cannot be cultured (obligate pathogens), which are only present in the plant and which cannot be purified. For this reason the exact concentration of plant pathogens can not or hardly ever be established accurately and so estimates have to be used. Even with plant pathogens which can be purified (many bacteria and viruses), the concentration determined is no more than a good estimate (e.g. CFUs or mg/ml). This is because determining the concentration is often based on an indirect measurement. REMARK When detecting cysts in soil samples, you can add 1 cyst to the soil sample and wash it: limit of detection is 1 cyst per xx cc soil. For detecting free-living nematodes, you can remove 1, 2 or 5 of them with a pipette, process and test them in a PCR: detection in that case is 1, 2 or 5 nematodes of the type to be named. For detecting viruses in leaves, you can pick a single leaf from a positive plant, and add it to 3, 9 or more leaves from a healthy plant. If this sap gives a positive reaction in ELISA, the detection of the virus in the matrix concerned is 1 in 4, 10 or more. Many more examples can be thought of. In this way, it may be possible to determine the limit of detection for plant pathogens better. A decision threshold, for example, is that the extinction of a sample must be larger or the same as 0.15 or 0.20 or similar, or the Ct value must be lower than 35 in the case of an RT-PCR. REMARK When establishing the limit of detection, it in fact involves the decision whether the pathogen is present or absent in a particular sample. Is my sample healthy or not? At which point do I decide that my sample is no longer healthy? The matrix material often plays an important role here. Unlike chemical analyses, analyses of plant pathogens often take place in very complex matrices. Background reactions with other components can quickly lead to false-positive reactions. The extent of background reactions forms an important criterion in establishing the limit of detection. You can establish this based on healthy plant material in which you want to show that the plant pathogen is present (the same matrix). With ELISA, for example, the average background of healthy samples can be determined, whereby often a certain safety margin is added in. This together gives a value that is adopted as a decision threshold.

17 11 EDVDP DEFINITION The limit of detection is: The detection limit (if the pathogen is available in a known concentration), or the limit whereby the risk of a false positive is minimal (if the concentration of the pathogen is not known). PROCEDURE Concentration known: Analyse eight dilution series in steps of two of the pathogen in the matrix. Determine the lowest concentration that still gives a signal above the background of the healthy control. Determine the average of these concentrations Determine the standard deviation of these measurements Calculate the limit of detection as the average of the measurement values plus 3 x the standard deviation: AG = x + 3δ Compare the maximum dilution whereby the signal disappears with the dilution used in the analysis method Concentration not known: Analyse eight dilutions of healthy plant material (from the same host plant species that is/are to be analysed), in the standard dilution with the standard analysis method. Determine the average of the healthy host series. Determine the standard deviation of the healthy host series. Set the limit of detection at the average of the measurement values plus 3 x the standard deviation: AG = x + 3δ According to NEN7777, after establishing the limit of detection the presence of the pathogen must be checked by analysing eight times a laboratory sample close to the limit of detection. See also tables 1 and 2. Repeatability NEN7777 DEFINITION Degree of concurrence between the results of successive measurements of the same measurand, which are carried out under the same measurement circumstances [NPR 2814]. REMARK When establishing the repeatability of an analysis, it must be determined to what extent the analysis is sensitive or not to the variations that may occur in the way the analysis is carried out under routine circumstances. In practice this means that it must be checked whether, and to what extent, there is any variation in the results of the analyses carried out within a single laboratory, by the same person with the same equipment. REMARK The repeatability of a PCR is established by analysing 8 sub-samples from a large, homogenous positive laboratory sample (both DNA/RNA extraction and (RT-) PCR). The repeatability of washing soil samples, for example, could be established (is in fact not always possible or sensible) by analysing a sub-sample from a large homogenous sample eight times. EDVDP DEFINITION Degree of concurrence between the results of successive measurements of the same measurand, which have been carried out under the same circumstances by the same person in a single laboratory within a certain period of time. PROCEDURE The repeat analyses (n 8) are carried out within a single laboratory by the same person with the same equipment. Repeatability is determined by examining identical (well-homogenised), practical samples (or 8 sub-samples of a single, well-homogenised practical sample) in a single measurement series. In day-to-day practice, the repeatability

18 12 of analyses is measured continually. Based on differences in observations, the measurement distribution will be determined. See also Tables 1 and 2. Reproducibility NEN7777 DEFINITION Degree of concurrence between the results of a measurement of the same measurand, obtained under varying measurement conditions [NPR 2814] REMARK The reproducibility that is stated in official validation reports is usually determined based on an intralaboratory study that is carried out in accordance with ISO 5725 (this is often not possible for plant pathogens). Another option is to establish the interlaboratory reproducibility based on a number of analyses carried out on naturally infected samples at different laboratories at different times. EDVDP DEFINITION Degree of concurrence between the results of a measurement of the same measurand, obtained under varying, specified conditions, preferably by different persons within a single laboratory. PROCEDURE Within a single laboratory, the reproducibility is determined by carrying out analyses on different days, and if possible also by different persons. Properly homogenised practical samples (within sufficient concentration of the pathogen to be detected) or control samples, to which a known quantity of pathogen has been added, are used for this. See also Tables 1 and 2. Specificity NEN7777 DEFINITION NEN7777 does not have a definition, but contains a remark on it: Specificity is a term that describes the same phenomenon as selectivity but in a different way. REMARK Identification of organisms/pathogens and establishing the relationships between organisms/pathogens is carried out by establishing morphological, biochemical and recently also molecular characteristics, often sequences. Using such characteristics, it can therefore also be established which organism is closely related to the organism to be detected, and for which it can be expected that there is a (theoretical) possibility of a cross-reaction in the analysis. In addition, other pathogens (organisms) not directly taxonomically related will occur in a certain crop (e.g. mixed infections with several viruses). In that case, a possible cross-reaction in the analysis to be validated with these pathogens (organisms) must be ruled out. Also the variation within the species must be carefully examined. Does the analysis show all variants (serotypes) within the species? For PCR, sequence information must therefore also be available for several isolates of the target organism. We refer here to analytical specificity. Reference is often made in diagnostics to diagnostic sensitivity and/or specificity, in which: Diagnostic sensitivity = A/(A+C); in which A = no. of genuine positives and C = no. of false negatives Diagnostic specificity = D/(D+ B); in which B = no. of false positives and D = no. of genuine negatives These are determined based on a gold standard. The gold standard is an analysis (or a combination of analyses) that is or are generally accepted. Criteria need to be established for this. EDVDP DEFINITION Specificity is the ability of a detection method to distinguish the pathogen from other organisms, whether related or not, and the extent to which the analysis can distinguish known or unknown variants of the organism.

19 13 PROCEDURE Carry out the analysis (if possible or applicable) on a number of variants (isolates) of the species, and also on a number of species closely related phylogenetically. Assume existing knowledge and establish which other plant pathogens may occur in the crop to be tested. Test the most frequently occurring organisms. See also Tables 1 and 2. Selectivity NEN7777 DEFINITION Dependency of the result of a measurement on a variable other than the measurand [ISO 6879] REMARK Plant pathogens often occur in several crops and/or cultivars of a crop. In practice, one needs to take account of the fact that these different crops, or even cultivars, can influence the reliability of the result of the analysis. In fact, the aim is to establish the possible influence of the material in which the pathogen has to be detected ( the matrix ) on the reliability of the analysis result. Selectivity should not be confused with specificity. Selectivity concerns only the matrix, whereas specificity concerns the organism. What background is possible? It is important to determine this where a laboratory-specific/new method is involved. EDVDP DEFINITION Selectivity is the ability of a detection method to distinguish the pathogen from other components of the sample (matrix effects). PROCEDURE Determine the healthy value for various cultivars of the crop to be analysed. Moreover, determine the influence of the matrix after adding the positive sample to sap from various cultivars of the crop. See also Table 1 and 2. Measuring range (working area, range) NEN7777 DEFINITION None REMARK This performance characteristic is often determined by the user and client together. The measurement range is defined for the entire method of measurement (all the different steps of the method together). If dilution of the sample is an explicit part of the established method of measurement, this should also be included in quantifying the measurement range. In some analyses the analysis does not work properly below or above a certain concentration of the pathogen. It is therefore necessary to establish within which range (lower and upper limit) the analysis functions. Pure reference material or infected practical material is necessary for this. EDVDP DEFINITION Limits within which the analysis can be applied reliably. PROCEDURE Make dilutions of the target in the matrix and determine within which limits (lower and upper limit) the test is adequate. See also Tables 1 and 2.

20 14 Uncertainty of measurement NEN7777 DEFINITION A parameter related to the result of a measurement, which characterises the distribution of values that can be reasonably assigned to the measurand [NPR 2814] REMARK This is important only for quantitative analyses. With the detection of plant pathogens, it is not possible or useful to determine the uncertainty of measurements of qualitative procedures, because the exact concentration of the plant pathogens can often not or hardly be accurately established (see limit of detection). EDVDP DEFINITION Uncertainty about the result of the measurement. PROCEDURE To determine the uncertainty of measurement of detection methods for plant pathogens, we refer to the design standards NEN7779 (Environment-Uncertainty of measurement results) and NEN8004 (Microbiology-Determination of measurement uncertainty). See also Table 1 and 2. Model error NEN7777 DEFINITION Deviation from the assumed link between measurand and measurement signal REMARK Model error is only relevant in the case of a quantitative analysis method, whereby the concentration of the pathogen under investigation is exactly known. This can be possible, for example, for the number of cysts of nematodes or fungal spores but not for the concentration of a virus. NEN7777: Model error can only be established based on a calibration standard. If this is not on sale (almost always) a standard must be created/determined independently (e.g. add cysts to the soil). EDVDP DEFINITION The deviation from the value found compared to the standard value. PROCEDURE You make the standard yourself using a known quantity/concentration of the pathogen in a certain matrix. You then establish by testing what the deviation and distribution is in the analysis compared to the standard (on 8 different days in accordance with NEN7777). Robustness NEN7777 DEFINITION Degree of insensitivity of the results of a measurement to deviations in procedure, circumstances and nature of materials as these may occur in practice. REMARK Robustness is an important performance characteristic in the plant pathogen sector. Particularly with the introduction of new, large-scale analyses, it must be demonstrated that the analysis is insensitive to variations in procedure, persons, circumstances and the nature of materials. The extent to which unintended variations in circumstances (which could also reasonably occur) can influence the reliability of the analysis is often examined on an experimental basis. The results are useful in determining the required accuracy of stoves, fridges, the composition and ph of buffers, the condition of the samples, the extent to which they can be stored, etc.

21 15 The stability of for instance a virus in a certain extraction buffer can vary widely and can affect the result of the analysis, but in principle has nothing to do with the validation of the analysis. It is important, however, to know the stability of the virus. EDVDP DEFINITION Degree of insensitivity of the result of a measurement to deviations in procedure, circumstances and nature of materials. PROCEDURE Estimate which parameters could reasonably influence the result of the analysis in the laboratory with the equipment present and personnel available. Test with homogenised laboratory samples of naturally infected or artificially inoculated and healthy plant material in which form variations in these parameters affect the result. See also Tables 1 and 2. The analysis is the sample preparation plus the extraction plus the specific analysis (e.g. PCR, ELISA, etc.). Robustness therefore concerns all these steps. 3.2 Summary of other definitions Analysis / Test Conducting a method to detect a certain target (in our case a particular plant pathogen). e.g. DAS-ELISA for PepMV in leaf material. Analysis sample Sample that has been prepared from the laboratory sample in accordance with the measurement procedure. Detection Detection is an activity that focuses on showing the presence (or absence) of a certain pathogen which is known or suspected to occur. This can be both quantitative and qualitative detection. Routine detection has the tendency to miss the unknown. In this respect, monitoring or screening must be viewed as detection because in such a case one also goes in search of particular pathogens in plant and vector populations or specific plants or vectors within the context of ecological or epidemiological studies. REMARK A distinction must be made between the diagnostics or recognition of the pathogen or parasite as the cause of a disease or damage, and the most routine detection of a known pathogen or parasite for quality and health certification. Diagnostics Recognition of diseases and damage based on the history (anamnesis) and of the characteristic symptoms and as much as possible based on the causing agent or cause (aetiological diagnostics, see also aetiology). REMARK Then the organism is often already available in pure form. Is the organism that has been isolated from an infected plant really the quarantine organism or not? In principle the postulates of Koch must be proven. In crop protection, it is desirable to reserve the terms diagnosis and diagnostics as much as possible for the recognition of diseases and damage. For the distinction between diagnostics and detection, see the explanation.

22 16 Own method (NEN7777) A method that has been developed entirely by the laboratory itself. Aetiology Knowledge of the causes of a disease and of its nature and relationships with the host. REMARK Interpretations of the term vary from merely identifying the cause of a disease to studying the entire progression of events that lead to a disease. Identification Determining the identity of the organism under study. REMARK Then the organism is often already available in pure form. Is the organism that has been isolated from an infected plant really the quarantine organism or not? In principle the postulates of Koch must be proven. Laboratory sample Quantity of material destined for examination, in the form and condition in which it is delivered to the laboratory. Method A method describes a procedure/working method to achieve a certain objective. A method is a way of doing something. It comes from the Greek word methodos, met' hodos. e.g. DAS-ELISA or PCR. Methodology Methodology means a general principle of methods. e.g. immunological or molecular methods. Monitoring Identifying on a large scale the presence of a number of organisms in place and time. See also Detection, Screening. REMARK Monitoring is often carried out to determine the presence of a particular organism in a particular crop (or group of crops) in a particular country (pest status). Reference method (RvA T01) This is taken to mean a method that: has been established nationally or internationally by a standardisation institution, the competent authority or equivalent, or, is generally accepted as a method within a sector, or has been published in national and international scientific literature. See also in this context T01 of the Accreditation Council, particularly concerning in accordance with reference method, equivalent to reference method and own method.

23 17 Screening Detecting particular organisms in a large number of samples. See also Detection, Monitoring. Standard method (NEN7777) A method accepted and validated by the user group. 3.3 Validation material In order to validate a detection method for a plant pathogen, the availability of the target organism is extremely important. For this reason, collections have been assembled in the world at various places to store organisms and groups of organisms (reference material). However, this is no guarantee that the organism required is available and that its identity is correct. Validation is therefore not always possible. For analysing plant pathogens in plant material, sufficient material is not always available or not at the right time of the year. The distribution of the pathogen in the material to be analysed must also be included in the nature and extent of the sample. A good evaluation of this material demands expert knowledge. Which objects to be measured are required for the validation study? a. The size of the laboratory sample is sufficient to obtain at least 2 analysis samples. b. The size of the laboratory sample only permits a single analysis. It is recommended taking mixed samples because of the non-uniform distribution of the pathogen in plant material, and to take sufficient material (if available) for several analyses. The mixed sample must however be properly homogenous so that a sub-sample can be taken from it for further analysis. For example, to homogenise infected leaf material it will be mixed and ground in liquid nitrogen or in a buffer (from a kit, for example). A sub-sample of 200 mg will be used for RNA extraction (RT-PCR) or for the ELISA. Table 6 (NEN7777) gives additional indications with the choice of material for validation. For plant pathogens, this concerns: the organism in a pure form, plant material spiked with the organism or naturally or artificially infected plant material The table below shows the order of validation material.

24 18 Table 2. Indications with the choice of validation material. Performance characteristics Freedom from bias, trueness Limit of detection Reproducibility Repeatability Specificity (analytical) Selectivity Robustness Measuring range Uncertainty of measurement Model error Indications Use in order of availability a. representative reference material from national and international collections b. naturally infected or inoculated plant material, for which it has been established using a different test that the plant pathogen is present a. healthy homogenised laboratory samples from practice b. dilution series of homogenised laboratory samples with known concentration (if limit of detection = detection limit) a. homogenised laboratory samples (positive and negative) from practice a. several isolates of the target species b. species related to pathogen obtained from reference and other collections and c. other organisms that can occur in the crop to be analysed originating from reference and other collections a. homogenised laboratory samples of naturally infected or inoculated and healthy plant material of various cultivars of the crop to be analysed a. dilution series of reference samples (make dilutions in matrix of crop to be analysed) b. dilution series of homogenised laboratory samples of naturally or artificially infected plant material; make dilutions in matrix of crop to be analysed) a. homogenised laboratory samples (positive and negative) from practice a. standard with known concentration of the pathogen in the matrix

25 19 4. Reports 4.1 Validation report NEN 7777 describes clearly how to write a validation report. As a minimum the following must be included in the validation report: A) the validated measurement procedure or an unambiguous reference to it; B) a description of: a. the purpose of the measurement (fitness for purpose) b. the measurand c. the object to be measured d. the area of application (matrix/measurement range) e. the result of a measurement, in so far as not contained in the measurement procedure C) reasons whether or not to include the performance characteristics that are not compulsory, but whose relevance in this standard needs to be considered; D) a description of any division into sub-areas of application, for which there is individual validation, with reasons; E) a description of the external requirements that are applicable, with their status (absolute limit value, estimated limit value, standard value); F) description of the test setup of the validation study, including the study period, and a description of the laboratory samples used; G) the deviation from this standard; H) the original results of the measurement; I) the calculation method followed, if this does not follow unambiguously from this standard; J) the calculated value of the performance characteristics, together with the circumstances for which these are valid, if applicable; K) optional: descriptions of performance characteristics spanning sub-areas of application*); L) the evaluation of performance characteristics against external requirements; if there are no external requirements: your own estimate of the fitness for purpose of the method for the measurement objective; M) a reference to standard NEN7777; N) the name of the researcher responsible and the institution. *) If there are no indications for differences in the value of a performance characteristic between different subareas of application (matrix/measurement range), a pooled value can be calculated and reported. If a functional link between performance characteristic and the value of the measurand can be assumed, this link can be presented for the purpose of estimating the performance characteristics at levels other than the one being studied. Appendix 2 contains two validation reports: DAS-ELISA and RT-PCR for Pepino mosaic virus in tomato leaves.

26 20

27 21 5. List of recommendations The Explanatory Document for the Validation of Detection Methods for Plant Pathogens and Pests has been discussed with the Accreditation Council. The Accreditation Council underlines the importance of such a document and expressly applauds its creation. However, it advises that the document be regularly maintained and updated. We share this recommendation and underline its importance. The Explanatory Document must be distributed among as many interested parties as possible and brought to their attention. This should include the Plant Protection Service, inspection services, institutions, laboratories, breeding companies, seed companies etc. In view of further internationalization, the document should be distributed within a European framework, and possibly even further. An English translation will allow it to play a role within Europe, and will establish the Netherlands as an important player in this field. We therefore strongly recommend that this document be translated. The importance of reference collections should have become clear from this document. We therefore emphasise the need to maintain good reference collections for plant pathogens.

28 22

29 23 Appendix 1. Details of Validation Schedule for Pepino mosaic virus (PepMV) B1.1 DAS-ELISA PepMV 1. Objective The objective is to detect the presence of PepMV with the aid of a double-antibody sandwich (=DAS-ELISA, hereinafter referred to as ELISA) in leaf sap pressed from a tomato leaf. The measurand is the extinction at 405 nm (A 405 ) as measured with a spectrophotometer. 2. DAS-ELISA DAS-ELISA is a known and existing methodology. Implementation in accordance with the standard DAS-ELISA method (see Appendix 1.3) using an antiserum produced against the tomato strain of the virus (Van der Vlugt et al., 2002). R.A.A. van der Vlugt, C. Cuperus, J. Vink, C.C.M.M. Stijger, D.-E. Lesemann, J.Th.J. Verhoeven & J.W. Roenhorst (2002). Identification and characterization of Pepino mosaic virus in tomato. EPPO Bulletin 32: Performance characteristics to be determined The result of the ELISA test will be interpreted qualitatively because it only concerns detecting the presence of the virus in the leaf material, not its exact quantity. Based on the flow chart (Figure 1 from 2.4) the performance characteristics to be determined are: Freedom from bias, Trueness Limit of detection Repeatability Reproducibility For the sake of completeness, it will also be stated how the extra performance characteristics that apply for a new analysis will be determined. Specificity Selectivity Measuring range Robustness The validation of each performance characteristic is rounded off with a conclusion. 4. Details of performance characteristics Freedom from bias, Trueness The definition of freedom from bias is described in 3.1. In practice, this means that one must show that the analysis does demonstrate the presence of Pepino mosaic virus. We note here that this can be done by analysing reference material present in national and foreign reference collections, for example. In these validation tests we use the type material of the tomato strain of PepMV (PD , Van der Vlugt et al., 2002) which is present in the collection of Plant Research International. In addition, a negative

30 24 control must always be included. For this, leaf material must be used that originates from tomato plants that have been grown from seed from healthy tomato plants in virus-free conditions. Procedure Grind PepMV-infected tomato leaf 1: 10 (w/v) with ELISA Sample Buffer (ESB; see Appendix 1.3). Grind healthy tomato leaf 1: 10 (w/v) with ESB as negative control. Make a 2x dilution series (1:2; 1:4; up to 1:128 (v/v)) in ESB from both leaf saps. Coat an ELISA plate with a standard (1:1000 in coating buffer) dilution of PepMV coating antiserum. Analyse all plant sap dilutions on the same ELISA plate, each dilution being applied in duplicate. For control purposes, apply at least 8 samples of the 1:2 dilution of healthy leaf sap on each plate to determine the limit of detection. Perform the ELISA according to the standard protocol using the standard antiserum against the tomato strain of the PepMV (Prime Diagnostics). Determine the limit of detection (being the average of the healthy series plus 3x the standard deviation of this series). Measure the A and 60 minutes after addition of the substrate. Determine the presence of a positive signal in the reference samples and the absence of a positive signal in the healthy control material. Prove, by using another method, that the pathogen is indeed present in the leaf material, e.g. RT-PCR. Limit of detection The definition of limit of detection is described in 3.1. This states that for a standard DAS-ELISA the limit of detection is: the average A 405 of the healthy controls, plus 3x the standard deviation of the A 405 extinction values of the healthy control series. Procedure Grind healthy tomato leaf 1: 10 (w/v) with ELISA Sample Buffer (ESB; see Appendix 1.3) as negative control. Grind PepMV-infected leaf material 1: 10 (w/v) with ESB. Make a dilution series (1:10; 1:20 to 1:1280) in ESB from both leaf saps. Coat an ELISA plate with PepMV coating (Prime Diagnostics) in the standard dilution in coating buffer. Analyse all dilutions on the same ELISA plate whereby each dilution of the healthy leaf sap is applied 8 times and each dilution of the diseased leaf sap is applied twice. Perform the ELISA according to the standard protocol. Measure the A and 30 minutes after addition of the substrate. Determine the limit of detection (being the average of the extinction values (A 405 ) of the healthy series plus 3x the standard deviation of the extinction values (A4 05 ) of that series ( AG = x + 3δ ). Repeatability The definition of repeatability is described in 3.1. In practice, this means that the entire procedure for the DAS-ELISA analysis must be repeated by the same person within a laboratory several times, in order to obtain an understanding of the degree of deviation. The repetition tests must be carried out on different days by the same person with the same equipment. Preferably use infected practical (laboratory) samples for the tests. The NEN 7777 standard stipulates a standard schedule for repeating this experiment (table 18). When determining the repeatability, assume an analysis of 8 samples on 8 different days. Bear in mind that the schedule states that a duplicate-procedure must be carried out on certain days (total 24 repeats). Procedure Grind PepMV-infected tomato leaf 1: 10 (w/v) with ELISA Sample Buffer (ESB; see Appendix 1.3). Grind healthy tomato leaf 1: 10 (w/v) with ESB as negative control. Make 2x dilution series (1:10; 1:20; to 1:1280) in ESB from both leaf saps.

31 25 Coat an ELISA plate with a standard dilution of PepMV coating antiserum (Prime Diagnostics) in coating buffer. Analyse all dilutions on the same ELISA plate whereby each dilution is applied in duplicate. Perform the ELISA according to the standard protocol using the standard antiserum against the tomato strain of PepMV (Prime Diagnostics). Measure the A and 30 minutes after addition of the substrate. Determine the presence of a positive signal in the reference samples and the absence of a positive signal in the healthy control material. Reproducibility The definition of reproducibility is described in 3.1 In practice, this means that the entire procedure for the DAS-ELISA test must be repeated several times, without the results deviating between the repeats. The repeats must be carried out on different days by different people, possibly also with different equipment. Preferably use infected practical laboratory samples and other samples for the tests. Procedure Grind PepMV-infected tomato leaf 1: 10 (w/v) with ELISA Sample Buffer (ESB; see Appendix 1.3). Grind healthy tomato leaf 1: 10 (w/v) with ESB as negative control. Make 2x dilution series (1:2; 1:4; to 1:128 (v/v)) in ESB from both leaf saps. Coat an ELISA plate with a standard dilution of PepMV coating antiserum (Prime Diagnostics) in coating buffer. Analyse all dilutions on the same ELISA plate whereby each dilution is applied in duplicate. Perform the ELISA according to the standard protocol using the standard antiserum against the tomato strain of PepMV (Prime Diagnostics). Measure the A and 60 minutes after addition of the substrate. Determine the presence of a positive signal in the reference samples and the absence of a positive signal in the healthy control material. NEN7777 stipulates a standard schedule for repeating this experiment (Table 18). When determining the reproducibility, assume an analysis of 8 samples on 8 different days. Bear in mind that the schedule states that a duplicate procedure (repeatability) must be carried out on certain days. Specificity The definition of specificity is described in 3.1. Analysing according to specificity means that it must be demonstrated that in the ELISA: positive reactions occur with the various known strains and isolates of PepMV. This may be demonstrated through the analysis of reference material, present in national or international reference collections, if available no cross-reactions occur with related viruses no cross-reactions occur with positive controls of other viruses that may also be present in tomato. This can be done by using certified antiserum, i.e. the antiserum that is used for the ELISA (coating and conjugate) comes with a quality certificate that shows that it demonstrates no cross-reactivity with other tomato viruses. If no certified antiserum is available, any cross-reaction with other tomato viruses must be excluded by carrying out these tests oneself. Positive control material can be used for this that can be obtained from the supplier of the antisera. Viruses on which cross-reactivity must at least be tested are: PVX, PAMV, TMV, ToMV, TSWV, CMV Procedure Grind PepMV-infected leaf material 1: 10 (w/v) with ELISA Sample Buffer (ESB; see Appendix 1.3). Grind healthy tomato leaf 1: 10 (w/v) with ESB as negative control. Grind infected material 1:10 (w/v) of the viruses to be tested in ESB or use positive controls of these viruses (used in this example: PVX, PAMV, TMV, ToMV, TSWV, CMV, TAV and HVX). Make a dilution series (1:10; 1:20, 1:40) in ESB from the leaf saps.

32 26 Coat an ELISA plate with PepMV coating (Prime Diagnostics) in the standard dilution in coating buffer. On the same plate, also coat wells in the standard dilution with antiserum against the viruses to be tested (4 wells for each virus to be tested). Analyse all dilutions on the same ELISA plate, whereby each dilution is applied in duplicate in each column. For control purposes, apply at least 8 samples of the 1:10 dilution of healthy leaf sap to the plate to determine the limit of detection. Perform the ELISA according to the standard protocol. Determine the limit of detection (being the average of the healthy series plus 3x the standard deviation of this series). Measure the A and 30 minutes after addition of the substrate. Determine the absence of a positive signal in the healthy control material and any presence of a positive signal in the samples to be analysed. Selectivity The definition of selectivity is described in 3.1 Plant pathogens often occur in several different crops and/or cultivars of a particular crop. In practice, these different crops or even cultivars may affect the reliability of the result of the analysis. This means that the possible influence of the material in which the pathogen must be detected ( the matrix ) on the reliability of the analysis result should be established. For this particular analysis one must make sure that no false positive reaction with healthy tomato leaf will occur. This can be demonstrated by testing leaf material in the ELISA that originates from plants that have been grown from seed from healthy plants under virus-free conditions. This negative control must be included on each ELISA plate. Procedure Grind healthy tomato leaf 1: 10 (w/v) with ELISA Sample Buffer (ESB; see Appendix 1.3) of at least 6 different cultivars. Add to this healthy sap a quantity of virus or diseased plant sap. Coat an ELISA plate with PepMV coating (Prime Diagnostics) in the standard dilution in coating buffer. Analyse healthy and spiked leaf sap on the same ELISA plate, whereby each sample is applied in duplicate. Perform the ELISA according to the standard protocol. Measure the A and 30 minutes after addition of the substrate. Compare the values obtained from all series. Determine to what extent the changes in cultivar have an effect on the measurement values. Measuring range The definition of measuring range is described in 3.1 It must be demonstrated in practice that extremely low or high concentrations of the pathogen to be analysed do not lead to major deviations in the reliability of the result. It is therefore necessary to establish within what range (lower and upper limit) the analysis functions. To do so, reference material or infected practical material is required. In fact, the measuring range can only be determined reliably if a direct comparison can be made between infected material in practice and a dilution series of purified pathogen. Procedure Make a purified solution of PepMV with a known concentration, a dilution series in ELISA Sample Buffer (ESB; see Appendix 1.3). Grind healthy tomato leaf 1:10 (w/v) with ESB as negative control. Grind PepMV-infected leaf material 1: 10 (w/v) with ESB. Make a dilution series in steps of 2 (v/v) in ESB from both leaf saps. Coat an ELISA plate with PepMV coating (Prime Diagnostics) in the standard dilution in coating buffer.

33 27 Analyse all dilutions of purified virus, healthy and diseased leaf sap on the same ELISA plate whereby each dilution is applied in duplicate. Perform the ELISA according to the standard protocol. Measure the A and 30 minutes after addition of the substrate. Plot the measured dilutions and accompanying A 405 extinction values on a graph and draw a line through the measurement points of the dilution series of the purified virus. Determine with the very low dilutions and the very high dilutions to what extent the measurement values deviate from the reference line for the purified PepMV. Determine from the analyses to establish repeatability to what extent these measurement values for PepMV fall within the linear part of the measuring range curve. Robustness The definition of robustness is described in 3.1. In practice, this means that it must be demonstrated that the reliability of the results of a measurement is not affected by (small) deviations from the standard protocol. In principle, this can be measured by comparing under varying circumstances a dilution series of purified virus on the same ELISA plate each time with a dilution series of one or more practical samples. Procedure Make a purified solution of PepMV with a known concentration, a dilution series in ELISA Sample Buffer (ESB; see Appendix 1.3). Grind healthy tomato leaf 1: 1 (w/v) with ESB as negative control. Grind PepMV-infected leaf material 1: 1 (w/v) with ESB. Make a dilution series in steps of 2 (v/v) in ESB from both leaf saps. Coat an ELISA plate with PepMV coating (Prime Diagnostics) in the standard dilution. Analyse all dilutions of purified virus, healthy and diseased leaf sap on the same ELISA plate, whereby each dilution is applied in duplicate. Perform the ELISA according to the standard protocol. Measure the A and 60 minutes after addition of the substrate. Compare the values obtained from all series under varying conditions. Determine to what extent the changes in conditions and/or procedure have an effect on the measurement values. 5. Reports and explanations As part of the validation report, the tests for establishing the various performance characteristics, as determined based on the above-mentioned validation schedule, must be worked out in detail and explained. In particular, what the result of determining the performance characteristic means for the routine application of the analysis is something that deserves attention. Each confirmation of the performance characteristic will be rounded off with a conclusion. 6. Final conclusion Draw up a final conclusion based on the results of the confirmation of all performance characteristics.

34 28 B1.2 RT-PCR PepMV 1. Objective To detect the presence of PepMV using a Reverse Transcriptase PCR (RT-PCR) in tomato leaf. The measurand is the presence of a PCR product of the correct size as established using agarosis gel electrophoresis followed by DNA staining using ethidium bromide. 2. RT-PCR RT-PCR is a known and existing methodology. It is performed according to the procedure as described in Verhoeven et al, The RT-PCR method for PepMV is therefore regarded as a reference method. As a primer combination in the RT-PCR, though, the following primers are used: 1 st strand cdna primer: 5 -CTGTATTGGGATTTGAGAAGTC-3 (PepMV-15RC) 2 nd strand PCR primer: 5 -GTCCCACATTACACTTCCTTCAG-3 (PepMV-11) Upon amplification this primer combination produces a DNA fragment of 973 bp. J.Th.J. Verhoeven, R.A.A. van der Vlugt & J.W. Roenhorst (2003). High similarity between tomato isolates of Pepino mosaic virus suggests a common origin. European Journal of Phytopathology 109: Performance characteristics to be determined The result of the RT-PCR test will be interpreted qualitatively because it only concerns detecting the presence of the virus in the leaf material, not the exact quantity. The performance characteristics to be determined are: Freedom from bias, Trueness Limit of detection Reproducibility Repeatability For the sake of completeness, it will also be stated how a number of extra performance characteristics that apply for a new analysis will be determined. Selectivity Specificity 4. Details of performance characteristics In this test, only the presence or absence of a PepMV-specific PCR product on gel will be determined as positive or negative. If the result of the test is not clear enough because the PCR product on the gel is not clearly visible, the result will be assessed as dubious. The assessment must always be made by two persons. Freedom from bias, Trueness The definition of freedom from bias is described in 3.1. In practice, this means that you must show that the analysis does detect what you want to detect. The freedom from bias of the PCR method can be checked both theoretically ( in silico ) and in practice, e.g. by using reference material available in national or foreign reference collections, for example. In these validation tests, we use the type material of the tomato strain of PepMV (PD , vd Vlugt et al, 2002) as available in the collection of Plant Research International. In addition, a negative control must be included at all times, for which leaf material must be used that originates from tomato plants that have been grown from seed from healthy tomato plants under virus-free conditions.

35 29 Procedure According to the instructions, a standard RT-PCR for the presence of PepMV was performed on leaf material originating from the reference collection of Plant Research International and infected with the tomato strain of PepMV. This applies to the entire procedure including RNA extraction. The same standard RT-PCR was also carried out on healthy tomato leaf (= healthy control). Proof, by using a different method (e.g. DAS-ELISA) that the pathogen is indeed present in the leaf material. Limit of detection The definition of limit of detection is described in 3.1. Contrary to that stated in relation to DAS-ELISA, it is impossible to determine the average of this PCR on healthy controls. In this case, the limit of detection can also be interpreted as a detection limit or as the analytical sensitivity. For the RT-PCR this can be established by means of a comparison with the limit of detection of a reference test, in this case the DAS-ELISA. This can be done by determining in analogy with the DAS-ELISA test with which dilution of the leaf material the result of the analysis becomes negative. Since the extraction of RNA from the leaf material is a fixed element of the RT-PCR analysis, the entire RT-PCR protocol must be applied to all dilutions from the dilution series and/or positive material in negative matrix. Procedure Analogous to the test for the DAS-ELISA, a dilution series will be made assuming RNA purified from undiluted leaf sap. The dilution series is analysed using the standard RT-PCR method for the presence of a PCR product of the correct size. The highest dilution of the RNA that can still be called positive in the RT-PCR is determined. The highest dilutions that are still positive in the RT-PCR analysis and in the DAS-ELISA. From this test a comparison of values is obtained between the RT-PCR analysis and a reference method (in this case DAS-ELISA). Repeatability / Reproducibility The definition of repeatability/reproducibility is described in 3.1. In practice, this means that the entire procedure for the RT-PCR analysis must be repeated several times. The repeats must be performed on different days by different people, possibly with different equipment if necessary. For the tests, preferably use infected field samples. NEN7777 stipulates a standard schedule for performing the test (Table 18). When determining repeatability, assume an analysis of 8 samples on 8 different days. Bear in mind that the schedule states that a duplicate procedure must be carried out on particular days. Procedure Perform a standard RT-PCR test for the presence of PepMV according to the instructions on 8 sub-samples on 8 different days, from a large field sample infected with PepMV. This applies for the entire procedure, including RNA extraction. Assume a large mixed sample that is divided into eight sub-samples. The tests must be performed by different people at different laboratories, using different equipment each time. At the same time, a healthy control sample originating from a healthy tomato plant must be included in the analysis with each infected sub-sample. Specificity The definition of specificity is described in 3.1. In practice, specificity is closely connected to (but not the same as) selectivity; the ability of a detection method to distinguish the pathogen from other components of the sample (matrix effects).

36 30 The specificity must be investigated in two ways. Firstly, the theoretical possibility that the primers to be used will give a false positive reaction with non-viral RNA. This can be done in silico and gives an indication of the crossreactions to be expected. The second way is a practical one, also verifying the absence of a false-positive signal with related viruses and nonrelated viruses from tomato. Concerning specificity, however, it is also important to check to what extent known isolates and strains of the pathogen to be analysed react in the analysis. Procedure In silico Compare the primers to be used with a BLAST search on the EMBL-EBI server ( for possible homologies with related or unrelated sequences, and establish the number of matches and mismatches with other related sequences present in the NCBI database. Practical test A. Reaction with isolates and strains of PepMV Find out whether positive reactions do indeed occur with the various known strains and isolates of PepMV. This can be done by analysing reference material present in national and foreign reference collections, for example, when available. B. Reaction with other tomato viruses Use reference material from the virus collection of Plant Research International of the following viruses that may actually occur in tomato but are not the same as PepMV: Potato virus X (PVX), Tomato mosaic virus (ToMV), Tobacco mosaic virus (TMV), Potato aucuba mosaic virus (PAMV), Cucumber mosaic virus (CMV), Tomato aspermy virus (TAV) and Tomato spotted wilt virus (TSWV). Extract total RNA from this reference material and perform a standard RT-PCR on it using the PepMV-specific primers according to Appendix 4.2. (NEN7777 shows only one test in table 18, no repeats) Check using agarosis gel electrophoresis and the right marker for the presence of the correct size of the PCR product. Selectivity The definition of selectivity is described in 3.1. Plant pathogens often occur in several crops and/or cultivars of a crop. In practice, these different crops, or even cultivars, can influence the reliability of the result of the analysis. In fact, it concerns here establishing the possible influence of the material where the pathogen has to be shown ( the matrix ) on the reliability of the analysis result. You must therefore establish that no reaction with healthy tomato leaf will occur. This can be demonstrated by testing leaf material in the RT-PCR that originates from plants that have been grown from seed from healthy plants under virus-free conditions. This negative control must be included in each experiment. Procedure Extract total RNA from healthy tomato leaf and perform a standard RT-PCR test on it. (NEN7777 shows only one test in table 18, no repeats). Check using agarosis gel electrophoresis for the absence of a PCR product. Robustness The definition of robustness is described in 3.1. In practice, this means that it must be demonstrated that the reliability of the results of a measurement is not affected by (small) deviations in the standard protocol. In principle, this can be measured by making deliberate changes in the performance of the RT-PCR analysis and examining the extent to which this affects the result. The standard protocol and a positive virus sample must be used as reference each time.

37 31 Examples of deviations from the standard protocol that could be analysed: pipette errors: deliberate deviations in the volumes of RT-mix and RNA are made. batches RT-PCR kit: analyse the effect that different batches of RT-PCR kit can have on the result of the analysis. PCR equipment: examine whether different PCR equipment affects the result of the analysis. Depending on the individual situation, other points of attention may also be analysed. 5. Reports and explanations As part of the validation report, the tests for establishing the various performance features, as determined based on the above-mentioned validation schedule, must be worked out in detail and explained. In particular, what the result of determining the performance characteristic means for the routine application of the analysis is something that deserves attention. Each confirmation of the performance characteristic should be rounded off with a conclusion. 6. Final conclusion Draw up a final conclusion based on the results of the confirmation of all performance characteristics.

38 32 B1.3 DAS-ELISA Protocol for PepMV 1. Materials used in ELISA 1.1 Microtitre plates 1.2 Buffers Coating buffer (ph 9.6) 1.59 g sodium carbonate (Na 2 CO 3 ) 2.93 g sodium bicarbonate (NaHCO 3 ) 0.20 g sodium azide (NaN 3 ) Dissolve in 900 ml H 2 O, adjust ph to 9.6 with HCl and make up to 1 l PBS (ph 7.4) phosphate buffer saline 8.0 g sodium chloride (NaCl) 0.2 g monobasic potassium phosphate (KH 2 PO 4 ) 1.15 g dibasic sodium phosphate (Na 2 HPO 4 ) 0.2 g potassium chloride (KCl) 0.2 g sodium azide (NaN 3 ) Dissolve in 900 ml H 2 O, adjust ph to 7.4 with NaOH or HCl and make up to 1 l PBS-Tween (PBST) PBS ml Tween 20 per litre Sample extraction buffer (ph 7.4) PBST + 2% PVP (Sigma PVP-40 polyvinyl pyrrolidone) Conjugate buffer PBST + 2% PVP + 0.2% egg albumin (Sigma A-5253) Substrate buffer 97 ml diethanolamine 600 ml H 2 O 0.2 g sodium azide (NaN 3 ) Adjust to ph 9.8 with HCl and make up to 1 litre with H 2 O Buffers can be stored at 4-10 degrees Celsius for at least 2 months. Warm to room temperature before use. 1.3 Substrate p-nitrophenyl phosphate disodium (Sigma stock no ; Sigma 104 phosphatase substrate tablets 5 mg/tablet) Make to 0.75 mg/ml in substrate buffer.

39 33 2. Procedure 2.1 Add 200 µl of purified IgG (coating) diluted in coating buffer (most common dilution is 1000x = 1 μl/ml) to each well of a microtitre plate. 2.2 Incubate at 37 degrees Celsius for 2-4 h or overnight at 4 degrees C. 2.3 Wash plate with PBS-Tween using wash bottle, soak for a few minutes and repeat washing two times. Blot plates by tapping upside down on tissue paper. 2.4 Add 200 μl aliquots of the test sample (extracted in sample extraction buffer) to duplicate wells. 2.5 Incubate overnight at 4 degrees C. 2.6 Wash three times as in step Add 200 µl anti-virus conjugate diluted 1:1000 in conjugate buffer to each well. 2.8 Incubate at 37 degrees for 4 hours. 2.9 Wash three times as in step Add 200 µl aliquots of freshly prepared substrate (7.5 mg p-nitrophenyl phosphate [Sigma ] dissolved in 10 ml of substrate buffer) to each well. Incubate at room temperature for min, or as long as necessary to obtain clear reactions Assess results by: a) Visual observation or b) Spectrophotometric measurement of absorbance at 405 nm in ELISA plate reader. 3. General remarks Use row 1 as negative control (negative sample or no sample added). Preferably always include a positive control (in duplicate) on each plate Reference Clark, M.F. and A.N. Adams Characteristics of the microplate method of enzyme-linked immunosorbent assay for the detection of plant viruses. Journal of General Virology 34:

40 34 B1.4 RT-PCR Protocol for PepMV Extract total RNA from mg of tomato leaf using a Qiagen Plant RNEasy kit or other equivalent commercial RNA extraction kit, according to the manufacturer s protocol. Preferably use fresh leaf material to isolate the RNA. If this is not available, material can also be used that has been stored at -20 C. Take µl total RNA and use this in a one-tube RT-PCR reaction (Access RT-PCR System, Promega or an equivalent system from another manufacturer) according to the relevant protocol. First strand cdna primer: PepMV-15RC: Second strand PCR primer: PepMV-11: 5 -CTGTATTGGGATTTGAGAAGTC-3 5 -GTCCCACATTACACTTCCTTCAG-3 This primer combination, after amplification, produces a DNA fragment of 973 bp. PCR programme: min 48 C 2. 2 min 94 C sec 94 C 4. 1 min 54 C 5. 1 min 30 sec 72 C 6. Go to step 3, 40 x min 72 C 8. overnight 10 C Check the presence of the specific PCR product using agarosis gel electrophoresis. Use a standard marker to check the expected size of the PCR product. Stain the DNA using ethidium bromide.

41 35 Appendix 2. Validation report B2.1 Validation report for DAS-ELISA to detect Pepino mosaic virus in tomato leaf Date of report: 6 October 2006 Carried out in period: September October 2006 Carried out by: M. de Weerdt, M. Verbeek, R. van der Vlugt Introduction Pepino mosaic virus (PepMV) is a viral disease that is mostly found in tomato plants. The virus has a regulated status on tomato seed. According to EU Directive 2004/200/EC, commercial consignments of seed must be demonstrably free of the virus. The virus can be detected by means of a classic DAS-ELISA or RT-PCR. DAS-ELISA is a serological method used to detect the presence of the coat protein of the virus. 1. Objective: Demonstrate the presence of PepMV using a double-antibody sandwich (=DAS-ELISA, hereinafter referred to as ELISA) in leaf sap pressed from a tomato leaf. The measurand is extinction at 405 nm (A 405 ) as measured on a spectrophotometer. 2 DAS-ELISA is a known and existing methodology. It is carried out in accordance with the standard DAS-ELISA protocol (see Appendix 1) using an antiserum produced against the tomato strain of the virus (Van der Vlugt et al, 2002). R.A.A. van der Vlugt, C. Cuperus, J. Vink, C.C.M.M. Stijger, D.-E. Lesemannn, J.Th.J. Verhoeven & J.W. Roenhorst (2002). Identification and characterization of Pepino mosaic virus in tomato. EPPO Bulletin 32: Establishing the performance characteristics The result of the ELISA will be interpreted qualitatively because it only concerns the detection of the presence of the virus in the leaf material, not the exact quantity. The ELISA is published but not validated, and so must be validated as a new method. Using the flow diagram from the explanatory document, the performance characteristics to be determined are: Freedom from bias (trueness) Limit of detection Repeatability Reproducibility Specificity Selectivity Measuring range Robustness

42 36 Procedure, results and conclusions Freedom from bias, Trueness The definition of freedom from bias is described in 3.1 The procedure for freedom from bias is described in Appendix 1.1 Result Healthy material (tomato, variety Moneymaker) that is used as a negative control gave an average A 405 of (after 30 min substrate incubation) with a 1:10 dilution, and with a standard deviation of The limit of detection was calculated from this (av. healthy value + 3 x standard deviation) = (3 x 0.002) = The average value of diseased material (tomato variety Moneymaker, infected with the tomato strain of PepMV) of the 1:10 dilution was With this dilution, the diseased:healthy ratio is 3.15:0.053 = With the highest dilution (1:1280) of the infected material, this ratio is still In addition, a second method was used to detect PepMV in the tomato material. The analysis using RT-PCR is described further on in the validation report for RT-PCR, which shows that the analysis used does indeed detect PepMV. Conclusion There is a clear signal difference between diseased and healthy material. This analysis method detects the presence of PepMV in infected tomato material correctly. Limit of detection The definition of limit of detection is described in 3.1. The procedure for limit of detection is described in Appendix 1.1. Result The average A 405 after 30 minutes of substrate incubation of the 8 negative controls (healthy tomato leaf 1:10) is The standard deviation is The limit of detection can be calculated from this: (3*0.002) = Figure 2. Limit of detection of ELISA for PepMV.

43 37 Conclusion The limit of detection established for the ELISA for PepMV is clearly much lower than the ELISA values found for the PepMV-infected tomato leaf. Repeatability/Reproducibility The definition of repeatability/reproducibility is described in 3.1. The procedure for repeatability/reproducibility is described in Appendix 1.1. Since the analyses in this validation were carried out at a research laboratory, whereby insufficient variation could be achieved as regards equipment and manpower, only repeatability has been determined in this example report. Result The results of the 24 repeats are shown in Table 3. The average healthy value was (SD 0.023) and the average diseased value was (SD 0.645). Table 3. Average healthy value and average diseased value of the 1:10 dilution as A 405 after 30 min. substrate incubation. Repeat Day Lab Sample Av. healthy value (1:10) Av. diseased value (1:10) Conclusion The repeatability of the ELISA on PepMV is 100%, i.e. all diseased samples produce a positive result in the ELISA.

44 38 Specificity The definition of specificity is described in 3.1. The procedure for specificity is described in Appendix 1.1. Result The analysis shows no (cross-) reactions with the tested viruses (Figure 3). In the positive PAMV control, the virus could no longer be detected. Figure 3. The specificity of the ELISA for PepMV. Conclusion The analysis method is highly specific for PepMV, i.e. no cross-reactions were observed with the other viruses included in this test. These are potexviruses related to PepMV (HVX, PVX) and viruses that may occur in tomato (PVX, TMV, ToMV, TSWV, CMV, TAV). (Remark: this test could be further extended with several viruses that may occur in tomato and various isolates of PepMV). Selectivity The definition of selectivity is described in 3.1. The procedure for selectivity is described in Appendix 1.1. Result None of the 6 cultivars tested (see Figure 4) caused false-positive signals or a marked reduction of the positive signal with the samples that were spiked with purified virus (0.3 ng per well).

45 39 Figure 4. Selectivity of the ELISA for PepMV. Conclusion No matrix effects were observed in six different tomato cultivars on the results of measurements. Measuring range The definition of measuring range is described in 3.1. The procedure for measuring range is described in Appendix 1.1. Result First and foremost, an ELISA was carried out in accordance with the above-mentioned protocol with a dilution series of purified virus. This was done to determine to what extent high concentrations affect the ELISA values and which dilution series had to be used in the comparison with the plant material. High virus concentrations lead to maximum measurement values. The A 405 values measured only fall when the concentration of purified virus is less than approximately 3 ng per well. The ELISA in which the purified virus is added to healthy leaf sap shows that even the highest dilution of the infected tomato leaf still gives a measurement value that is comparable with 0.75 ng of purified virus per well. Figure 5. Results of measurements in ELISA of a dilution series of purified PepMV.

46 40 Figure 6. Measuring range of the ELISA for PepMV. Conclusion The measuring range of the ELISA for PepMV is very large. There is no clear upper limit, and no signal reduction is measured with high concentrations of virus. The lower limit is lower than 24 pg/well. Robustness The definition of robustness is described in 3.1. The procedure for robustness is described in Appendix 1.1. Since most factors that can be attributed to robustness are determined by local conditions at the laboratory that ultimately carries out the analysis and has to be determined anew for each lab, it has been decided here not to carry out any analyses in relation to robustness. Final conclusion The DAS-ELISA analysis for detecting Pepino mosaic virus in tomato leaf is suitable for the intended purpose.

47 41 B2.2 Validation report for RT-PCR to detect Pepino mosaic virus in tomato leaf Date of report: November 2006 Carried out in period: September October 2006 Carried out by: M. de Weerdt, R. van der Vlugt Pepino mosaic virus (PepMV) is a viral disease that is found mostly in tomato plants. The virus has a regulated status on tomato seed. According to EU Directive 2004/200/EC, commercial consignments of seed must be demonstrably free of the virus. The virus can be detected by means of a classic DAS-ELISA or RT-PCR. DAS-ELISA is a serological method used to detect the presence of the coat protein of the virus. Introduction Pepino mosaic virus (PepMV) is a viral disease that is found mostly in tomato plants. It is not an official quarantine organism, but is on the EPPO Alert List. According to EU Directive 2004/200/EC, commercial consignments of seed must be demonstrably free of the virus. The virus can be detected by means of a classic DAS-ELISA or a RT-PCR. DAS-ELISA is a serological method used to detect the presence of the coat protein of the virus. RT-PCR (reverse transcriptase polymerase chain reaction) is a detection method used to establish the presence of the viral RNA, the genetic material of the virus. The RT-PCR test for detecting Pepino mosaic virus (PepMV) in tomato leaf is carried out according to the procedure as described in Verhoeven et al, 2003.( J.Th.J. Verhoeven, R.A.A. van der Vlugt & J.W. Roenhorst (2003). High similarity between tomato isolates of Pepino mosaic virus suggests a common origin. European Journal of Phytopathology 109: ) As a primer combination in the RT-PCR, however, the following primers are used: 1 st strand cdna primer: 5 -CTGTATTGGGATTTGAGAAGTC-3 (PepMV-15RC) 2 nd strand PCR primer: 5 -GTCCCACATTACACTTCCTTCAG-3 (PepMV-11) This primer combination produces, after amplification, a DNA fragment of 973 bp. The procedure involves the isolation of total RNA from leaf samples, performing RT-PCR, gel electrophoresis to make the PCT product visible and the assessment of the test results. In this test, only the presence or absence of a PepMV-specific PCR product on gel is determined as positive or negative respectively. If the result of the test is not clear enough because the PCR product on the gel is not visible enough, the result will be assessed as dubious. The assessment must always be carried out by two persons. Determining the performance characteristics The result of the PCR test will be interpreted qualitatively because it only concerns the detection of the presence of the virus in the leaf material, not the exact quantity. Using the flow chart from the explanatory document, the performance characteristics to be determined are: Freedom from bias, Trueness Limit of detection Repeatability Reproducibility Selectivity Specificity Measuring range Robustness The last four performance characteristics are included as they are not properly described in the publication.

48 42 Procedure, results and conclusions Freedom from bias, Trueness The definition of freedom from bias is described in 3.1. The procedure for freedom from bias is described in Appendix 1.2. Tomato leaf material, infected with the tomato strain of PepMV, originating from the virus collection of Plant Research International was tested for the presence of the virus using an RT-PCR method published earlier. Result The RT-PCR test on the reference isolate of PepMV was clearly positive with a PCR product of the expected size, whereas the negative control showed no PCR fragment. The presence of PepMV in the reference isolate could also be detected with ELISA. Conclusion The RT-PCR method used, including the primer set, is correct because it does indeed show the presence of PepMV in tomato leaf. Limit of detection The definition of limit of detection is described in 3.1. The procedure for limit of detection is described in Appendix 1.2. Result Figure 7 shows the results of determining the limit of detection of the analysis. To the left is the marker as control on the fragment size of the PCR products. Then from left to right the RT-PCR products obtained from a dilution series (undiluted, 2x, 4x, 8, 16x, 32x, 64x, 128x) of RNA purified from an infected plant. A RNA dilution of 64x still shows a detectable product in the RT-PCR analysis. M Figure 7. Agarose gel of RT-PCR test to determine the limit of detection. M = 100 bp marker. From left to right the RT-PCR products obtained from a dilution series (undiluted, 2x, 4x, 8, 16x, 32x, 64x, 128x) of RNA purified from an infected plant. A comparison with the values obtained earlier from ELISA for determining the limit of detection is shown in the following table.

49 43 Table 4. Dilution RNA ELISA Undiluted x x x x x x x ± Conclusion A dilution of 64x of the RNA purified from a PepMV-infected tomato leaf, still shows a positive signal in a RT-PCR analysis for PepMV. The limit of detection for the RT-PCR analysis is at a leaf sap dilution of 64x. REMARK With subsequent analyses it is recommended not to carry out the RT-PCR on dilutions of the purified RNA, but first to purify RNA from dilutions of the infected leaf sap and then perform a RT-PCR on these RNA extracts. Repeatability/Reproducibility The definition of repeatability/reproducibility is described in 3.1. The procedure for repeatability/reproducibility is described in Appendix 1.2. Since the analyses in this validation were carried out at a research laboratory, whereby insufficient variation could be achieved as regards equipment and manpower, only repeatability has been developed further in this example report. Result Figure 8 shows the results of the test for repeatability of the analysis. On the left is the marker as control of the fragment size of the PCR products. Then from left to right the negative control next to the practical sample for each day. M PC Figure 8. Agarose gel of RT-PCR test for repeatability. M = 1 Kb marker. From left to right, alternating results of the analysis on 8 different days of healthy and PepMV-infected tomato leaf. To the far right the positive control (PC) on the action of the RT-PCR kit.

50 44 Conclusion The repeatability of the analysis is 100%. For each day on which the analysis is carried out the result is identical, i.e. PepMV could be detected. Specificity The definition of specificity is described in 3.1. The procedure for specificity is described in Appendix 1.2. Result In silico In the BLAST searches, only homologies (100%) with PepMV isolates from the EMBL database were found. Homologies were not found in any other sequences in the database. Practical test The RT-PCR analysis did not give a PCR product with any virus, except with PepMV as positive control (see Figure 9). M PC Figure 9. Agarose gel of RT-PCR test for specificity. M= 100 bp marker, 1 = TAV, 2= CMV, 3= PVX, 4= ToMV, 5= TMV, 6= TSWV, 7= PepMV, PC= positive control on action of RT-PCR kit. Conclusion The RT-PCR analysis is specific for PepMV. Selectivity The definition of selectivity is described in 3.1. Determining selectivity for the analysis is described in Appendix 1.2 and consists of determining the possibility of false-positive background reactions when carrying out the analysis on various tomato cultivars.

51 45 Result Total RNA was isolated from the following healthy tomato cultivars and tested for possible false-positive signals; cultivars Extase, Roma vf, San Marzano, Fabiola, VW700, VW63 and E114. As a control on the correct action of the RT-PCR, RNA was also isolated from a tomato leaf (PC) infected with PepMV, control RNA from the RT-PCR kit and water (MQ) included in the analysis (Figure 10). Conclusion The RT-PCR analysis is selective for PepMV. M PC C MQ Figure 10. Agarose gel of the RT-PCR analysis to determine selectivity. M = 100 bp marker. From left to right, the results of the RT-PCR analysis carried out on the tomato cultivars 1 =Extase, 2=Roma vf, 3=San Marzano, 4=Fabiola, 5=VW700, 6=VW63 and 7=E114. PC = tomato plant infected with PepMV; C = control RNA from RT-PCR kit; MQ = water control. Robustness The definition of robustness is described in 3.1. The procedure for robustness is described in Appendix 1.2. Since most factors that can be attributed to robustness are determined by local conditions at the laboratory that ultimately carries out the analysis and has to be determined once again for each lab, it has been decided here not to carry out any analyses in relation to robustness. Final conclusion The RT-PCR analysis for detecting Pepino mosaic virus in tomato leaf is suitable for the intended purpose.