PRJ-0336: Diagnostic investigation of sporidesmin toxicity: Histological study


  • Sample collection: Organ samples are collected from abattoirs and stored in formalin when possible. Samples are collected in three different numbered bottles. The first bottle contains the left liver lobe (randomly numbered bottle), the second the right liver lobe, and the third contains a liver biopsy punch to sample a piece of lung tissue. These samples are batched and sent through to Onderstepoort Pathology Department.
  • The microclimate sensors worked well initially, but by late February/March, they gave sporadic readings which became unreliable. The service was terminated, and alternative options are sought together with FABI and satellite sensing.
  • The spore counts this year started to reduce towards the end of March, so the service was terminated. The spore count service is due to start again soon and will be done the same way as last year.


PRJ-0362: Exploring the facial eczema problem in dairy cattle in the Eastern Cape of South Africa, with a focus on the fungus Pseudopithomyces chartarum.

Aim 1: To resolve the taxonomy of the genus Pseudopithomyces with focus on Pseudopithomyces chartarum and to determine which species are associated with Sporidesmin Induced Liver Disease.

Progress: An additional 58 Pseudopithomyces strains from the CN culture collection were re-plated and DNA was extracted for all strains. There is now 105 strains from the CN and 48 strains from the CBS culture collections. An additional 250 sequences have been generated for existing and new strains, and an additional 384 PCRs are in the process of being sequenced.

Aim 2: To determine whether Pseudopithomyces chartarum is seed-borne.

Progress: The results from grass samples showed that fungal growth was mainly observed on plates where seeds were plated out with filter paper. Genera identified include Fusarium, Cladosporium and some Alternaria. No Pseudopithomyces was observed. This probably indicates that Pseudopithomyces chartarum is not a seed-borne fungus. However, it can still be an endophytic species, which need to be established.

Aim 3: To determine which Pseudopithomyces species predominate the outbreak areas in the Eastern Cape.

Progress: Grass samples were taken from five farms. In the July trip, 15 samples were collected from all of the sampling sites. The first box of microscope slides from the spore traps (Sarnia) were collected for processing. During the August trip, 13 samples were collected from all the sampling sites. The second box of microscope slides from the spore traps (Sarnia) were collected for processing. The counts from both boxes are available from the official report.

Aim 4: To study the relationship between sporidesmin production and the population diversity of Pseudopithomyces chartarum.

Progress: Two genomes of Pseudopithomyces chartarum were sequenced and compared with the New Zealand genome. This research was presented at the 14th World Mycotoxin Forum in Antwerp, Belgium, 9-11 October. The project leader was invited to publish this genome comparison in the 25th anniversary of ‘’Toxin’’. In addition, she has established a strong national and international team of scientists to cooperate in this context.

PRJ-0360: Use of NIR to detect and quantify mastitic bacteria in cow’s milk.

Background: Recent advances in portable NIR (Near-Infrared Spectroscopy) technology and cloud-based Machine Learning application development have made the NIR technology affordable and accessible. The InnoSpectra range of portable devices, along with the Frequai software will allow dairy farmers, processors and consultants to analyse their products for as many parameters as necessary on-site and if necessary in real time. Whereas the present project is to develop NIR calibration equations for the major mastitis pathogens, the usual quality parameters can also be monitored regularly with the equipment as can be seen below.

Aim 1: Optimizing NIR models of milk quality parameters.

Progress: This was for the development of the software models for the analysis of major milk components, i.e., protein, fat and sugar, on an inexpensive InnoSpectra T-11 device. To that effect in order to produce the necessary range of fat, sugar and protein composition, mixtures of low fat milk, full cream milk and cream were created. The prediction models created showed exceptional accuracy with R2s in excess of 0.9 [Y = predicted value; X = tested value]:

Fat: Y = 0.812X + 1.51; R2 = 0.953

Sugar: Y = 0.923X + 0.370; R2 = 0.980

Protein: Y = 0.749X + 0.725; R2 = 0.931

The models were developed rapidly and easily. Therefore, the performance indicates that the InnoSpectra T-11 is a suitable device, coupled with the Frequai software, for rapid and inexpensive milk analysis.

Aim 2: Optimizing NIR models of Staphylococcus aureus infection of milk.

Progress: The inoculum level for S. aureus was optimized and the coagulation point determined. Inoculated milk is used to create a standard curve of S. aureus as a function of time, over a 24-hour period, sampling every 15 minutes in order to build a model with a much bigger sample size. Each sample is scanned by NIR, and then used in a dilution plate count to get an accurate count of cells per ml in each sample. These results are then used to create the calibration curve of NIR for the pathogen population in milk. Currently, the samples are scanned with the InnoSpectra T-11 device and plate counts are done using a dilution series for each sampling time. A problem encountered is inconsistencies in the plate counts, but should be solved soon. If all data is satisfactory, it will be used to create an NIR calibration model using the Machine Learning routine.

Aim 3: Creating NIR models of infection of milk by Streptococcus spp and E. coli.

This will be done once the S. aureus model is satisfactory.