A survey was conducted of the processing and storage technologies for sago starch in selected areas of the Sepik and Western Provinces. The surveys were conducted by the commissioned organization (James Cook University) and the University of Papua New Guinea (collaborating institution). These data showed that a mixture of old and emerging practices were being used as far as storage was concerned. Processing methods were largely untouched by modern technology. The processing survey results showed weaknesses in a number of areas of human activity where microbial contamination could enter the food chain.
Samples of sago were collected from villages surveyed and the microbial contamination was assessed in terms of well-known food poisoning organisms, the presence of haemolytic microbes and the presence of common mycotoxins. Five well-known food poisoning bacteria were identified, but, except for one organism, were generally present in low numbers. Various haemolytic bacteria, yeasts and filamentous fungi have been isolated. Some of these were represented in high numbers. Some have been identified and further work is planned on the metabolic products they elaborate, as some species isolated are potentially associated with human illness. Sago samples analysed to date for the presence of mycotoxins commonly associated with human illness have given negative results. This does not mean that mycotoxins are not present, but just that those tested for are not present.
The ability of food-borne bacteria of public health significance to survive in sago has been tested under laboratory conditions. This involved the seeding of both fresh and older sago starch samples with known populations of selected bacteria, and following their history. Surprisingly, all the pathogens selected showed very low survival rates over the time frames studied. This is presumably due to the strong fermentation conditions that develop in bundles of sago with the vigorous growth of lactobacilli and associated decline in pH. These studies are continuing so as to gain a more realistic idea of the major microbial interactions that occur in sago bundles.
The survey of production and storage practices, when combined with the results of the subsequent scientific information, has led to the identification of some weak areas where microbial contaminants enter the food chain. These preliminary observations will form the basis of a more structured study at selected villages to refine the scientists' concept about the feasibility of making a difference to the microbiological flora of sago at the time of eating by manipulating the critical control points associated with the microbial hazards. They believe that some diminution of microbial hazards is possible through alterations to harvesting practices.

No proposed revisions to the objectives have been formulated.

Sociological surveys of sago producing areas were continued to assess the impact of collection and storage techniques on the fermentation events occurring in the sago starch. A composite flow chart of sago production has been created. Analysis of the flow chart, in conjunction with the sociological data, has enabled the identification of several critical control points at which microbial contamination can be minimized. These have been established at trunk storage, pith removal, starch extraction and storage and food preparation. All these points represent category 2 controls, which mean that food borne illnesses can be minimized but not eliminated by attending to various guidelines. Not all activity areas represent critical control points for all village situations. The greatest single factor contributing to sago starch contamination during harvesting and extraction was the quality of the wash water used. Suggested improvements in the location of some the extraction areas and restricting access of animals to the sites should lead to a decrease in certain types of contamination. In the absence of clean water, perhaps the most useful suggestion is to store of sago under fermenting conditions for several days to one month. This is perhaps a practical recommendation that can be made to minimize some populations of food borne bacteria of public health significance. These preliminary observations are being supplemented by laboratory experiments where the status of specific bacteria are being followed in time course studies.

Identification of fungal, yeast and bacterial isolates were completed on samples taken from the Sepik and Western Provinces. Survey data were particularly focused on the identification of fungal and bacterial isolates capable of haemolytic activity on the one hand and food poisoning bacteria on the other. All potentially significant bacteria and filamentous fungi were identified and enumerated in the samples. A number of strongly haemolytic organisms were isolated and they were present in considerable numbers in some samples. Correlations were sought between their occurrence and a number of physical and chemical parameters. Fungal population levels, particularly as measured by ergosterol content, increased with storage time. Population levels shown by this group of microbes was more pronounced when sago was stored in leaves and woven baskets than in earthen pots or plastic containers. Classical bacteria of public health significance were also identified. The behaviour of Escherichia coli and Staphylococcus aureus were of particular interest. The former showed the greatest survival ability of any organism investigated. Survival curves for a number of these bacteria were established in both freshly harvested and older sago. Numbers of most organisms were reduced to low levels generally by 10 days, except for E. coli. Strain characteristics of this organism are yet to be investigated.

Sago samples were analysed for well known mycotoxins capable of causing human disease. These samples were of varying ages and organoleptic states. Several samples tested positive for the presence of zearalenone from the Western Province and one sample from the Sepik Province tested positive for cyclopiazonic acid. However, considering the number of samples analysed, contamination of sago by the common mycotoxins was considered of little overall practical significance. The causal agent of sago haemolytic disease is probably a microbe producing a novel haemolysin. Thus, unusual mycotoxins and haemolysins are being sought actively among the bacteria and fungi isolated and some promising leads are being obtained. Aspects of the ecology of these organisms will be investigated to complement the information gathered already on the behaviour of well known food borne bacteria.

Conditions favouring fermentation are associated generally with the storage of sago, for the moist sago is commonly stored in bulk in containers (leaves, plastic or clay pots) thereby ensuring anaerobic conditions. The identification of a number of different bacterial activity groups is being actively pursued in order to give a clearer idea of the microbes essential to rapid fermentation. The main microbial metabolites are in the process of being identified. The groups of significance, so far identified, go beyond the lactobacilli. The fermentation process appears to function to encourage a decline in certain food borne bacteria of public health significance. We are examining the dynamics of this relationship and attempting to nominate optimum storage periods for the decline of the most common unwanted bacteria. We are also looking at the possibility that emerging bacteria of public health significance may also be present, as early indicators are pointing in this direction.

Sociological surveys of sago producing areas were continued to assess the impact of collection and storage techniques on the fermentation events occurring in the sago starch. A composite flow chart of sago production has been created. Analysis of the flow chart, in conjunction with the sociological data, has enabled the identification of several critical control points at which microbial contamination can be minimized. These have been established at trunk storage, pith removal, starch extraction and storage and food preparation. All these points represent 'category 2 controls', 'meaning that food borne illnesses can be minimized but not eliminated by attending to various guidelines. Not all activity areas represent critical control points for all village situations. The greatest single factor contributing to sago starch contamination during harvesting and extraction was the quality of the wash water used. Suggested improvements in the location of some the extraction areas and restricting access of animals to the sites should lead to a decrease in certain types of contamination. In the absence of clean water, perhaps the most useful suggestion is to store of sago under vigorous fermenting conditions for several days to one week. This is a practical recommendation that can minimize some hazardous populations of food borne bacteria. These preliminary observations follow from laboratory experiments which show that specific food-poisoning bacteria decline rapidly in numbers in the first week of storage. The superiority of fermentation in sago-leaf bound bundles and in clay pots is emphasized over other methods, such as in open baskets.

Identifications of fungal, yeast and bacterial isolates were completed on samples from the Western Province and the identification of isolates from food associated with a fatal case of sago haemolytic disease was commenced. We focused on the identification of fungal and bacterial isolates capable of haemolytic activity on the one hand and food poisoning bacteria on the other. Significant bacteria and filamentous fungi were identified and enumerated. A number of strongly haemolytic organisms were isolated and they were present in considerable numbers in some samples. Fungal population levels, particularly as measured by ergosterol content, increased with storage time. Population levels shown by this group of microbes was more pronounced when sago was stored in leaves and woven baskets. Classical bacteria of public health significance were also identified. Escherichia coli showed the greatest survival ability of any organism investigated. Survival curves for a number of bacteria were established in both freshly harvested and older sago. Numbers of most organisms were reduced to low levels by 10 days, except for E. coli. Strain characteristics of this organism are yet to be investigated to determine the incidence of enterohaemorrhagic isolates.

Sago samples were analysed for well known mycotoxins capable of causing human disease. These samples were of varying ages and organoleptic states. Cyclopiazonic acid was detected in two sago samples and ochratoxin was produced after prolonged storage of one sample in the refrigerator. However, considering the number of samples analysed (58), contamination of sago by the common mycotoxins (aflatoxins, zeralenone) was considered of little overall practical significance. The causal agent of sago haemolytic disease is probably a microbe producing a novel haemolysin. One promising haemolysin has strong lipid affinity and has the ability to lyse human red blood cells rapidly. Some of the isolates of Penicillium and Aspergillus showing strong haemolytic activity were also capable of citrinin production. The possible role of this toxin in human illness or sago haemolytic disease is unknown. It is possible that it may potentiate the action of other toxins.

Conditions favouring fermentation are associated generally with the storage of sago, for the moist sago is commonly stored in bulk in containers (leaves, plastic or clay pots) thereby ensuring anaerobic conditions. The identification of a number of different bacterial activity groups is being actively pursued in order to give a clearer idea of the microbes essential to rapid fermentation. The main microbial metabolites are in the process of being identified. The groups of significance, so far identified, go beyond the lactobacilli. The fermentation process appears to function to encourage a decline in certain food borne bacteria of public health significance. We are examining the dynamics of this relationship and attempting to nominate optimum storage periods for the decline of the most common unwanted bacteria. We are also looking at the possibility that emerging bacteria of public health significance may also be present, such as Enterobacter sakazakii and Citrobacter freundii. The latter has been confirmed as present in some samples.

To publicise project findings, a Symposium on Microbial contaminants associated with sago processing and storage in Papua New Guinea and their association with life threatening Sago Haemolytic Disease will be held in Port Moresby during September 2005 and Government officials and medical practitioners will be invited. Project findings will also be presented at the PNG Medical Conference during September 2005.

Work to determine the significance and characterisation of haemolytic compounds derived from fungi isolated from sago starch has commenced. To date, the techniques to fractionate crude haemolytic extracts from Penicillin steckii have resulted in isolated bands of haemolytic activity which are awaiting further characterisation with High Performance Liquid Chromatography. Techniques based on these methods have been prepared from other fungi with haemolytic properties. Techniques to determine the activity of these haemolytic fractions on red cell membranes have been optimised and await to be utilised with purified compounds. A Hazard Analysis and Critical Control Point (HACCP) study which guides the safe preparation and storage of sago starch in PNG has been completed and initial plans for the publication of these points for village-based use is underway.

Objective 3. Assessment of mycotoxin contamination in sago samples and the potential of thee mycotoxins to cause haemolytic disorders.

Activity 1. Activity directed fractionation
Development of purification and separation method by C18 SPE cartridges using 2 solvent systems one of Acetonitrile and the other with Methanol based on crude P. steckii hexane fraction. These separations provided a series of partially purified fractions which have been assayed for hemolytic activity. Hemolytic activity has been shown to be confined to 2 suspected bands on TLC plate. Further purification of these bands by HPLC has commenced.

Activity 2. Toxin characterization
At this stage little spectral characterization of the hemolytic compound(s) have been undertaken, but will eventuate as HPLC fractions are available.

Activity 3. Crude extraction
Crude extraction of hemolytic compound using P.steckii culture to develop an appropriate extraction and purification protocol for application to other fungi in sago. A working protocol using the non-polar fraction unlike polar fraction for other mycotoxins showing hemolytic activity has been adapted.

Activity 4 Activity directed fractionation
Initial screening of crude extracts for bands and hemolytic activity only. Investigations on hemolytic compounds to be pursued when suitable separation and detection method on HPLC developed.

Activity 5. Toxin Characterisation
To be undertaken when activity 4 is complete.

Activity 6. Red Cell Membrane Ghost SDS PAGE electrophoresis
The technique of red cell ghost preparation and protein separation with SDS PAGE has been optimised. The assay system is ready to test haemolytic fraction for specific red cell membrane activity when fractions become available (activity 2 and 4).

Activity 7 - 10.
No progress, work planned outside this reporting period.

Activity 11. Transfer of technical skills to partner organisation
Mr Aisak Pue (University of Technology, Lae) commenced an AusAID funded PhD program based at University of Queensland and DPI&F in March 2006. Aisak has received technical skills training in laboratory and bench work research. This involved extraction, purification and included blood hemolytic activity assaying skills in PC 2 lab. Learning of instrumentation on HPLC and associated issues of efficient operation, time management and laboratory troubleshooting.

Activity 12. Transfer of appropriate technology to partner organisation
Aisak Pue has undertaken a series of courses involving learning of new methodology and research skills to be transferred in form of knowledge and skills transfer to PNG. This included learning literature searching using the University of Queensland subscribed web linkages for internet access of journals and other materials. It also included computer literature writing with Endnote linked referencing features. These skills have been demonstrated in the preparation of a comprehensive Literature Review (first draft submitted in June). Aisak also undertook AQIS training in April for Certificate of Awareness and Certificate of Approved Premises. These skills begin the technology transfer resulting from this project. Other skills / technology transfer will result from activity outside this reporting period.

Objective 4: Enhancement of HACCP training for strategic Unitech staff
Activity 13. HACCP training
Dr Betty Amoa successfully completed the Food Safety Management Systems Auditor/Lead Auditor Course (NCSI Training and Development) on 2 June 2006. Mr Pue will complete a similar program in 2007.

Activity 14.
No progress, work planned outside this reporting period

Objective 5: Recommend practical measures for reducing contaminants risk for consumers

Activity 15. Develop and distribute a poster describing HACCP in relation to sago starch processing and storage in such a way that individuals within village communities can understand safe sago starch processing and storage

HACCP training and analysis has been completed. Dr Betty Amoa is engaging the services of the PNG Dept of Health to formulate appropriate language that can be used for poster development for village distribution.

Activity 16 - 20
No progress, work planned outside this reporting period