This book is a product of the international VIDE (Virus Identification Data Exchange) project (Boswell, Dallwitz, Gibbs and Watson, 1986; Boswell and Gibbs, 1986; Watson, Dallwitz, Gibbs and Pankhurst, in press). This project is collecting diagnostic information on all plant viruses using the DELTA DEscription Language for TAxonomy) database system (Dallwitz, 1980). The advantages of the DELTA system are, in brief, that the system is specifically designed to handle all forms of taxonomic information in a useful, `user-friendly' and flexible way. Data stored by this system are readily transferred to taxonomic programmes (e.g. key-generating or interactive identification programmes) and to automatic typesetting or microfiche-producing facilities.
Information on over 500 characters is currently sought for each virus (Appendix 5.1) and stored in the VIDE database; over two- thirds of these characters record the susceptibility, or otherwise, of a range of commonly used test plant species. The remainder represent the great range of other characters used for virus identification. The way in which these are determined has been described in many books and reviews (Bos, 1983; Gibbs and Harrison, 1976; Maramorosch and Koprowski, 1967 et seq; Noordam, 1973). In many laboratories, workers rely on simple tests such as the resistance of the infectivity of sap from infected plants to ageing, heat, etc., and on the symptoms produced in indicator plants. By contrast those in well equipped laboratories rely on electron microscopy and serological tests. The database includes host range information even though many virologists probably agree with Hamilton et al. (1981) when they stated that host range "is generally not a very precise or reliable guide to virus identification, and exhaustive host range tests are usually no longer appropriate." In practice, host range studies are still an important component of plant virus diagnosis in many laboratories, and, with standardization, they could become more useful.
The VIDE database includes host range information of five sorts:-
These species (Appendix 5.1) comprise all test species recorded in the CMI/AAB `Descriptions of Plant Viruses' No 1-309, together with some commonly available plants which add taxonomic balance to the sample of the Angiosperm species. Comparative information based on this sample is of value as it gives some indication of the host family preferences of each virus, and hence may be used to predict alternate natural hosts or species that might be used as diagnostic or assay hosts.
The list of test plants recorded in the CMI/AAB Descriptions was sent to several VIDE collaborators, who indicated for each species whether they used it in host range tests. Twenty eight of them, (from seven countries) replied, and the species printed in boldface in the Appendix 5.1 are those used occasionally, at least, by more than a quarter of those virologists. It would be valuable if these species were always tested for susceptibility when characterizing a novel virus.
At present most host range information is very incomplete, however some comparative data are available for related viruses, and for viruses from the same or related plant species. Though limited, these data are particularly useful when handled by an interactive identification programme, such as INTKEY. Recently Johnstone et al. (1984) have shown that comparative host range studies done in different countries, and involving shared seed lots of a large number of plant species, may be used to characterize and compare luteoviruses of legumes. They showed that the relationships between these viruses computed from simple `susceptibility/resistance' data correlated well with relationships indicated by serological tests of their particles.
Thus it is likely that, if more virologists used a standard range of species such as that in Appendix 5.1, the value of host range tests would be greatly increased.
To identify an unknown virus, one determines its characters and then attempts to match these with information about previously described viruses.
The identification process may follow one or other of two strategies, the `pragmatic' and the `logical', although in practice a mixture of the two is usually adopted. These strategies, and the relationships of some of the sources of information, is shown as a flow diagram in Figure 1.1. The central whorl of this diagram lists (from the top and in an anticlockwise direction) some of the techniques used, in approximate order of increasing specificity.
The `pragmatic' approach to identification depends on the fact that each virus has a definite host-range that may be confined to one or a small number of plant families, and that plants are usually easier to identify than viruses.
First the natural host or hosts of the virus are identified, next a check of the records is made to establish which viruses have previously been isolated from those species (or their close relatives), and then their properties compared with those of the unknown virus. Suspects can then be specifically checked using serological or nucleic acid hybridization tests. Thus, in Figure 1.1, this identification process `enters' from the top of the diagram, but then moves directly to the area of specific comparitive tests to the right of the diagram.
To aid this process, Chapter 2 of this book lists all the known natural hosts of plant viruses in Australia; "natural hosts" merely means all known hosts, excluding those infected experimentally. Section 2.1 list the botanical names of the natural hosts in alphabetical order, and gives the common name and the plant family to which they belong, and the virus or viruses that have been isolated from them in Australia. Section 2.2 lists the same information, but with the plant species grouped into their families. Thus using these two lists one can establish which viruses have been isolated from a particular species or from its close relatives. Section 3.1 then gives, in alphabetical order, full descriptions of individual viruses against which the properties of the unknown can be checked, and Section 3.2 gives additional, specifically Australian, information about each virus, including references to all known Australian work.
Figure 1.1 Flow Diagram of Strategies to Identify Virus Relationships and Specificities.
The `logical', or taxonomic, approach, to identification, first assesses to which group a virus belongs by determining some of its `group-specific' characters, such as the shape and size of its particles. Then, as above, more specific tests, including host range tests, are used to check whether the unknown is an already described member of the likely group.
Thus, in Figure 1.1, this strategy involves moving around the main identification whorl in an anti-clockwise manner, and requires some understanding of virus classification, which will be briefly outlined below. Chapter 4 is designed to aid this process, it gives information about each group and lists all known members of that group.
No keys are included in this book as an interactive identification programme, INTKEY, together with the current version of the VIDE database, will be available soon for use in micro or personal computers.
There is no universally applicable set of characters for defining all virus species. In practice, a virus species is a collection of virus isolates whose known properties are so similar that there is no reason to distinguish between them and give them separate names (Gibbs and Harrison, 1976). Harrison et al. (1971) found that many of the best studied viruses fell into 16 groups, each of which was best defined using a set of characters, because, for most, no single character uniquely defined the group. This principle has since been used to define additional groups of plant viruses, some monotypic, so that there are now 25 well defined groups (Matthews, 1982), and at least 10-15 others in the process of being defined.
Thus virus groups are collections of virus species that share most of their properties, and have probably evolved from a common ancestor. The most useful level to define such groups is one that permits usefully accurate predictions of the properties of partially described members of the group from those of the well-known members of the group (Gibbs and Harrison, 1976). Thus, for example, the bromovirus, cucumovirus and ilarvirus groups are useful to define as separate groups because members of each of these groups share most of the characters that seem to be biochemically and ecologically important. The same viruses have also been grouped as members of a larger group, the `tricornaviruses' (van Vloten-Doting et al. 1981), and sequencing studies have confirmed that some, at least, of their genes probably have a common progenitor, however different members of this `supergroup' have particles of different shapes, and have quite different vectors and ecological life-cycles, so placing the three groups in this larger group has little practical value.
Viruses in each of the accepted groups usually have:
Thus the taxonomically logical way to identify an unknown virus is to determine some of the properties listed in a, b and c above, before distinguishing between different members of the group, by determining `species'-specific characters, such as those listed in d, e and f, and finally confirming the diagnosis using very specific tests such as serological tests on viral proteins or nucleic acid hybridization tests.
Often the quickest way to obtain `group-specific' information is to examine sap from the infected plant in an electron microscope. This should give information on the shape, size and concentration of virus particles in the plant and, together with information on, for example, the type of symptoms and inclusions it causes, its mode of spread, or the sedimentation coefficients of its particles, should unequivocally indicate to which group it belongs. Researchers without access to an electron microscope and a high- speed centrifuge will instead have to rely on the results of a greater number of less discriminating tests, such as assessments of the stability of infectivity of sap, and the symptoms shown by a range of experimentally infected plants.
In practice, of course, several tests will be done concurrently, and eventually, if these indicate the possible identity or relationships of the unknown virus, this will be confirmed by serological or nucleic acid hybridization tests.
If the unknown virus seems to be unlike any other previously described, it is important to collect and store information on as many of its characters as possible to help in future comparitive studies. The bare minimum of data required is the shape, size and infectivity of its particles, for characteristic particles identify it as a virus and not some other kind of pathogen such as a bacterium, mycoplasma or viroid.
No data collection system is perfect, and that used to compile the VIDE database is no exception. Hence the data in this book has various limitations, some of which are known to us and are worth outlining briefly. For example, although the very best available group of experts collated the information, the quality, accuracy and `age' of the data available to them varied, as probably did their interpretation of it. Also recording the data in the database may have led to inaccuracies, despite various iterative stages of checking.
It is very important to note information in the Australian data that enables one to assess the certainty with which each virus has been identified, and its presence in Australia established. Some viruses have been fully characterized, but others have only been identified by the symptoms they cause in a natural host, and by examining the sap of that plant for virus-like particles; for many of these records no confirmatory tests were done. Other records are merely of quarantine `intercepts' when the infected plants were promptly destroyed, and so it is unlikely that such viruses are still present in Australia. Thus it might have been more accurate for the title of this book to have been "Plant Viruses Allegedly found in Australia on at Least One Occasion". Such a title might have allayed justifiable fears of international quarantine repercussions resulting from the publication of a list of pathogens of this sort.
The geographical distribution data are mostly very sketchy except where the hosts are contract-grown crops, that are frequently inspected, such as sugarcane. The population density of Australian plant virologists is small and, at present, all, except one, work in the south-east quarter of the continent between Adelaide and Brisbane!
Most of the viruses recorded in this book are worldwide viruses of crop species. With one or two notable exceptions, such as lettuce necrotic yellows and barley yellow dwarf viruses, little work has been done to identify the important reservoir hosts of crop viruses, or to study viruses of native plants.
Finally, it should be noted that Australian information about viroids has been included in Section 3.2, but there is no general information of each viroid in Section 3.1.
|
|
|
|