To the editor — Aan die redakteur Procedures for the diagnosis of virus particles by electron microscopy

With reference to a recent research note on viral pathogens in piglet faeces samples and a similar report in children, we would like to address the methods used by the authors in these surveys. Electron microscopic viral diagnosis using the negative staining technique is well established. While not as sensitive as some other methods, negatively stained specimens examined for the presence of virus particles remains a rapid alternative to conventional isolation and serology methods in virus diagnosis. However, the success of this diagnostic technique depends on several factors, for example: (a) the removal, as soon as possible after collection of a sample, of unwanted debris and bacterial components that may obscure virus particles; (b) the achievement of a homogeneously distributed monolayer of particles on the substrate of the grids; (c) adequate contrast of the specimen; (d) the use of a suitable magnification to scan samples (it is preferable to use a magnification of, for example, ×25 000 rather than a higher magnification of ×40 000 to ×45 000, which is often used). Although a high magnification is ideal for obtaining suitable micrographs, it is of doubtful value when scanning grids for the presence of virus particles. A simple calculation shows that doubling any magnification will increase the area to be scanned 4-fold, thereby reducing the likelihood of detecting any virus particles 4-fold; (e) experience in identifying and recognising virus-like particles and an ability to distinguish them from structures resembling viruses and (f) achieving a threshold concentration of virus particles by e.g. the use of appropriate centrifugation procedures. In our experience, very few samples can be succesfully examined for the presence of virus particles without centrifugation. Moreover, the application of this technique in a systematic fashion provides a consistent and controlled method of sample preparation irrespective of the personnel involved. When doing a survey over a period of time, predetermined centrifugation times and speed (especially g-values) together with a constant dilution factor provide a reliable method for evaluating and comparing results. The value of using a suitable centrifugation technique is reflected in the results obtained in our laboratory during the past 6.5 years for the diagnosis of parvovirus (20–25 nm) in dogs. Initially, in the first 6 months, a success rate of about 50 % was achieved for samples using only low-speed centrifugation. In order to eliminate possible false negative results, the sample preparation method was modified to include 3 centrifugation steps. Samples were initially centrifuged at 3000 g for 15 min, 7000 g for 30 min and 27 500 g for 1 h. This method resulted in an increase in the success rate to about 70 %. In order to improve contrast quality and minimise unwanted debris and a background film of fine proteinaceous material, further changes were made to the technique. This involved reducing the centrifugation speed and number of centrifugations. To remove bacteria and debris the initial centrifugation was maintained at 3000 g for 30 min. This was followed by high-speed centrifugation (20 000 g for 1 h) to produce a virus-rich fraction suitable for electron microscopy. Using higher g-values to concentrate the virus particles is unnecessary and results in producing protein-like particles that resemble small virus particles, as well as depositing a fine proteinaceous background film that obscures particle detail on the grids. These refinements increased the positive identification of parvovirus to above 90 %, and the technique has been successfully employed to identify a range of virus particles in avian, bovine and equine samples as well. A microlitre head was recently installed on our bench centrifuge, enabling us to decrease the centrifugation time for the 2 steps to 15 min and 30 min respectively, for about the same g-values. The sensitivity of the technique was increased 10-fold as well, as we now need only a 1.5 ml aliquot instead of 15 ml and a ×5 dilution instead of ×20 as used previously. In addition, the sensitivity of the technique can be further increased by any laboratory in possession of an airfuge ultracentrifuge. The disadvantage of not using a centrifugation step is reflected clearly, in our opinion, in the low percentage of positive results reported in some recent publications. An example is the 2.4 % incidence of small-sized virus particles and the 39.8 % incidence of rotavirus reported in piglets suffering from diarrhoea. Apparently these authors also experienced some difficulty in distinguishing entero(28–30 nm), astro(28– 30 nm) and calicivirus (32–38 nm) -like particles from each other. This seems to apply to the recent survey in children as well, in which the authors make unlikely claims about the presence of Breda-like virus particles. This may reflect a problem in recognising or identifying fringed membranous particles (considered to originate from fragmentation of epithelial cell linings) from true virus particles. Our centrifugation procedure provides sufficient contrast and concentration of material to allow us to make, with confidence, positive identifications and distinctions of the abovementioned small viruses, including parvovirus. The diagnoses were based on the appearance and size (irrespective of overlapping size ranges) of the particles, whether or not they occurred as a single or simultaneous infection, provided each virus concentration was above a threshold value necessary for identification. The high specif ici ty and speed of electron microscopy in obtaining a positive virus diagnosis illustrate the value of this technique if the correct sample preparation methods are used. Taking into account the present situation where such electron microscopic diagnoses are no longer made free of charge but involve payment by the patient or customer, the need for a reliable procedure and diagnosis is essential. We recommend that serious thought be given to alternative negative staining preparation methods for virus identification (whether for individual samples or surveys) rather than using a simple staining method that is prone to suffer from false negative results from the outset. Such situations can easily be avoided by following guidelines similar to those suggested here.


Procedures for the diagnosis of virus particles by electron microscopy
With reference to a recent research note on viral pathogens in piglet faeces samples 4 and a similar report in children 8 , we would like to address the methods used by the authors in these surveys.
Electron microscopic viral diagnosis using the negative staining technique is well established 3 .While not as sensitive as some other methods, negatively stained specimens examined for the presence of virus particles remains a rapid alternative to conventional isolation and serology methods in virus diagnosis 2,6 .However, the success of this diagnostic technique depends on several factors, for example: (a) the removal, as soon as possible after collection of a sample, of unwanted debris and bacterial components that may obscure virus particles; (b) the achievement of a homogeneously distributed monolayer of particles on the substrate of the grids; (c) adequate contrast of the specimen; (d) the use of a suitable magnification to scan samples (it is preferable to use a magnification of, for example, ×25 000 rather than a higher magnification of ×40 000 to ×45 000, which is often used).Although a high magnification is ideal for obtaining suitable micrographs, it is of doubtful value when scanning grids for the presence of virus particles.A simple calculation shows that doubling any magnification will increase the area to be scanned 4-fold, thereby reducing the likelihood of detecting any virus particles 4-fold; (e) experience in identifying and recognising virus-like particles and an ability to distinguish them from structures resembling viruses and (f) achieving a threshold concentration of virus particles by e.g. the use of appropriate centrifugation procedures.
In our experience, very few samples can be succesfully examined for the presence of virus particles without centrifugation.Moreover, the application of this technique in a systematic fashion provides a consistent and controlled method of sample preparation irrespective of the personnel involved.When doing a survey over a period of time, predetermined centrifugation times and speed (especially g-values) together with a constant dilution factor provide a reliable method for evaluating and comparing results.
The value of using a suitable centrifugation technique is reflected in the results obtained in our laboratory during the past 6.5 years for the diagnosis of parvovirus (20-25 nm) in dogs.Initially, in the first 6 months, a success rate of about 50 % was achieved for samples using only low-speed centrifugation.In order to eliminate possible false negative results, the sample preparation method was modified to include 3 centrifugation steps 1 .Samples were initially centrifuged at 3000 g for 15 min, 7000 g for 30 min and 27 500 g for 1 h.This method resulted in an increase in the success rate to about 70 %.In order to improve contrast quality and minimise unwanted debris and a background film of fine proteinaceous material, further changes were made to the technique.This involved reducing the centrifugation speed and number of centrifugations.To remove bacteria and debris the initial centrifugation was maintained at 3000 g for 30 min.This was followed by high-speed centrifugation (20 000 g for 1 h) to produce a virus-rich fraction suitable for electron microscopy.Using higher g-values to concentrate the virus particles is unnecessary and results in producing protein-like particles that resemble small virus particles, as well as depositing a fine proteinaceous background film that obscures particle detail on the grids.These refinements increased the positive identification of parvovirus to above 90 %, and the technique has been successfully employed to identify a range of virus particles in avian, bovine and equine samples as well.
A microlitre head was recently installed on our bench centrifuge, enabling us to decrease the centrifugation time for the 2 steps to 15 min and 30 min respectively, for about the same g-values.The sensitivity of the technique was increased 10-fold as well, as we now need only a 1.5 ml aliquot instead of 15 ml and a ×5 dilution instead of ×20 as used previously.In addition, the sensitivity of the technique can be further increased by any laboratory in possession of an airfuge ultracentrifuge 3 .
The disadvantage of not using a centrifugation step is reflected clearly, in our opinion, in the low percentage of positive results reported in some recent publications.An example is the 2.4 % incidence of small-sized virus particles and the 39.8 % incidence of rotavirus reported in piglets suffering from diarrhoea 4 .Apparently these authors 4 also experienced some difficulty in distinguishing entero-(28-30 nm), astro-(28-30 nm) and calicivirus (32-38 nm) -like particles from each other.This seems to apply to the recent survey in children as well 8 , in which the authors make unlikely claims about the presence of Breda-like virus particles.This may reflect a problem in recognising or identifying fringed membranous particles (considered to originate from fragmentation of epithelial cell linings 5,7 ) from true virus particles.Our centrifugation procedure provides sufficient contrast and concentration of material to allow us to make, with confidence, positive identifications and distinctions of the abovementioned small viruses, including parvovirus.The diagnoses were based on the appearance and size (irrespective of overlapping size ranges) of the particles, whether or not they occurred as a single or simultaneous infection, provided each virus concentration was above a threshold value necessary for identification.The high specificity and speed of electron microscopy in obtaining a positive virus diagnosis illustrate the value of this technique if the correct sample preparation methods are used.Taking into account the present situation where such electron microscopic diagnoses are no longer made free of charge but involve payment by the patient or customer, the need for a reliable procedure and diagnosis is essential.
We recommend that serious thought be given to alternative negative staining preparation methods for virus identification (whether for individual samples or surveys) rather than using a simple staining method that is prone to suffer from false negative results from the outset.Such situations can easily be avoided by following guidelines similar to those suggested here.Running a successful veterinary clinic usually implies rendering a sterling service to the community.Such service is usually the result of effective teamwork involving everybody working at the practice.Regardless of the number of people rendering the service, effective communication between them plays a pivotal role in rendering an excellent service.While it is not difficult for the owner or owners of a practice to explain the overall goal of the practice to staff members, it is much more taxing to explain every little detail that goes into consistently providing an excellent service.A clinic procedures manual could fulfil this role and provide all employees with guidelines on all aspects of their daily chores.Dr Fling's booklet provides guidelines for the compilation of a procedures manual, which, as he correctly states, is a living document that requires input from all employees, and should be revised, updated and changed as needed.
The following major points are covered in separate sections of the manual: clinic policies for all employees; clinic operational procedures; job descriptions; client relations and telephone protocol.Starting with a simple mission statement of the practice, the different sections then proceed to elaborate on guidelines regarded as attributes that, when demonstrated by an employee, should result in a superior performance.The different sections are also supplemented with sample pro forma forms that detail various aspects of running a private practice such as staff disciplinary action; hospitalisation and surgery authorisation; permission for euthanasia; daily treatment sheet; veterinary assistant guide to cleaning kennel cages; separating a pet from the client; client takehome literature; telephone standards and first-aid procedures.
Dr Fling sees a procedures handbook as the Bible of a practice which should constantly be studied and consulted by all employees.He clearly states that there is absolutely no right or wrong approach in the preparation of such a handbook but that it is merely a process to get the information floating around in one's head or scribbled on scraps of paper all over the practice, collated into a single document.
This ring-bound soft cover manual is strongly recommended as essential reading for all practice managers as well as for staff employed in veterinary practices.It is bound to trigger a process that may improve the profits of the practice.