Professor Dr. Son Radu Centre of Excellence for Food Safety Research Universiti Putra Malaysia

1. How safe is our food? Using The Results of a Risk Assessment in Food Safety Risk Management Professor Dr. Son Radu Centre of Excellence for Food Safety Research Universiti Putra Malaysia

2. MRA is a used as a tool at this level Policy Government level Food control system • Determine the policy • Set public health goals • Set standards • Based on risk analysis • Adjust inspection systems to verify FSO/PO are met PH goals Standards FSO, PO Operational level Food safetyassurance HACCP is appliedat this level PC • Design control measures from farm to fork • Responsible for hazard control HACCP GAPs/GHPs/GMPs

3. Public health burden ?? ALOP Food Safety Objectives Safety by design Performance Objectives • GAP • GMP • GHP • HACCP • Code of practicle

4. RISK COMPARISON

5. Growth pattern of a child with frequent episodes of diarrhoea

6. Presentation Topics • Risk assessment of Listeriamonocytogenesin Raw Vegetable • Raw Vegetables Consumption as A Potential Risk Factor for Campylobacteriosis in Malaysia • Biofilm formation and persistence of Salmonella • Malaysia Fishery Products Export to the EU & Issues Encountered – An example of a collaborative work between University and the relevant authority

7. What is the hysteria about Listeria? Listeria is a Gram positive, facultative anaerobic, psychrotrophic, rod shaped bacterium Listeria is a hardy bacterium- grows across a broad pH (4.3-9.8) and temperature (0.5 -45oC) range, and up to 20% salt tolerance Incidence of listeriosis varies between 0.1 to 11.3 per 1,000,000 in different countries Listeriosis has an average case-fatality rate of 20-30% despite adequate antimicrobial treatment Life-threatening illness in three clinical syndromes: maternofetal listeriosis or neonatal listeriosis, blood stream infection, and meningoencephalitis Listeria monocytogenes is widely distributed in soil and water Can enter the VBNC state and pose problem in routine plating detection methods

8. What is the hysteria about Listeria?

9. What is the hysteria about Listeria?

10. Key Questions O% 100% No raw vegetables consumption vs. Raw vegetables consumption • What should be done to intervene? • To what extent does the consumption of raw vegetables contribute to Listeria monocytogenes infections in humans in Malaysia?

11. The different steps of the infectious process at the cellular level: Left panels. Electron micrographs describing the various steps of entry, lysis of the vacuole, intracellular movement, cell-to-cell spread and lysis of the two-membrane vacuole (Cossart and Lecuit, 1998). Right panel. Schematic representation of the infectious process with the proteins involved: internalin, InlB, PI-PLC, ActA, and lecithinase (adapted from (Tilney and Portnoy, 1989)). The infectious process by L. monocytogenesat the cell and tissue levels L. monocytogenesis an invasive bacterium which induces its own entry into cells. Internalization results from the tight apposition of the plasma membrane over the entering bacterium. This process, also called “the zipper mechanism” appears different from the “trigger mechanism” used by bacteria such as Salmonella and Shigelladuring which dramatic membrane ruffles rich in filamentous F-actin engulf the bacterium in a process similar to macropinocytosis. Listeriais then present in vacuole that is lysed in less than thirty minutes. When free in the cytosol, Listeriastarts to replicate while inducing the recruitment and the polymerisation of cellular actin. Actinpolymerisation only occurs at one pole of the bacterial body and allows the bacterium to propel itself inside the cytosol. From time to time, bacteria reach the plasma membrane where they induce the formation of long protrusions containing a bacterium at their tips. These protrusions can invaginate in a neighboring cell and give rise to a two-membrane vacuole that the bacterium lyses to get access to a second infected cell and by doing so disseminate into tissues by a direct cell-to-cell process (Tilney and Portnoy, 1989).

12. The Recipe For Risk • A person must ingest cheese that is contaminated with Listeria monocytogenes in the order of 100-1000 cells • The immunocompromised person that ingests these bacteria must become sick – invade gastro-intestinal epithelium, become bloodborne and associated with monocytes, then subsequently the liver, spleen and lymphatic system and then to the nervous system and placenta barrier

13. FLOWCHART OF SAMPLING, MPN-PLATING AND MPN-PCR FOR LISTERIA MONOCYTOGENES Contaminated sample MPN 9-TUBES DILUTIONS Sample & broth (pre-enrichment) 1:10 RATIO Pre-enrichment 4 hours, 30°C Incubate (30oC, 48 h, aerobic condition) Presumptive Listeria colonies are black centered on PALCAM Agar (selective agar) VBNC MPN-PCR of turbid tubes MPN-PLATING of turbid tubes

14. Detection of Listeria monocytogenes using polymerase chain reaction

15. Consumption data Estimated mean intake (g/day/person) of raw vegetables as ulam in Malaysia according to area, gender and race (Source: MOH, 2008)

16. Prevalence No. of cells Dose of L. monocytogenes in a meal Risk assessment model structure Schematic representation of the model framework for retail-to-table risk assessment of L. monocytogenes in raw vegetables.

17. Exposure Assessment Prevalence • The distribution for the prevalence of L. monocytogenes in Japanese parsley, wild parsley, winged bean and Indian pennywort was estimated from the data collected in this study. The prevalence was described by pert distribution assuming a minimum and maximum prevalence accordingly: Pr = RiskPert( min, max, most likely ) where min and max are minimum and maximum of the prevalence data, respectively.

18. Exposure Assessment Concentration • The distribution of concentration of L. monocytogenes in contaminated raw vegatables was estimated from the data collected in this study and was assumed to follow a lognormal distribution: Cr = RiskPert (min, most likely, max)

19. Exposure Assessment Log reduction of washing practice • Log reduction of washing practice was estimated from the data collected in the kitchen simulation study performed in the laboratory. • The correlation coefficient between initial microbial load on the raw vegetable and log reduction was determined by using StatTools (Palisade Corporation). • The distributions of microbial load on vegetable before washing and log reduction of washing were fitted to the data collected in the study and the correlation between both inputs was defined with the resulted correlation coefficient with @Risk 5.5.

20. Risk Characterization • The output of exposure assessment was combined with the dose response function for hazard characterization to estimate the yearly risk from L. monocytogenes. • The probability of illness per person per year was described by the equation: Pill:year = 1 – ( 1 – Pill )365 where Pill is the probability of illness person per day.

21. Prevalence and MPN count in log10 MPN/g of Listeria monocytogenes in four types of raw vegetables purchased from wet markets and hypermarkets a Numerator: number of positive samples; denominator: total number of samples tested. b NA: Not applicable c ND: Not done

22. (a) (b) (c) (d) Distribution of prevalence of L. monocytogenes in (a) Indian pennywort, pegaga; (b) Japanese parsley, selom; (c) winged bean, kacang botol; and (d) wild parsley, ulam raja. Figure 2: Distribution of prevalence of L. monocytogenes in (a) Indian pennywort, pegaga; (b) Japanese parsley, selom; (c) winged bean, kacang botol; and (d) wild parsley, ulam raja.

23. Distribution of concentration of L. monocytogenes in retail Indian pennywort, Japanese parsley, winged bean and wild parsley.

24. (a) (b) (b) (a) (c) (d) Distribution of microbial load before and after washing of vegetable with tap water. (a) Indian pennywort; (b) Japanese parsley; (c) winged bean; and (d) wild parsley.

25. Raw vegetables (ulam-ulaman) consumption in Malaysia based on estimated mean intake (g/day/person) a a The distribution for the mean intake was described by the following cumulative distribution RiskCummul (1.0,5.5,{1.72,1.74,3.02,3.27,3.62,3.7,3.78,4.41,4.59},{0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9})/ F(x) is the cumulative probability (Vose, 1996), i.e. F(x) = i / (n+1), where i is the rank of the observed data point and n is the total number of data points.

26. The probability of illness per person (high-risk population and low-risk population) per year due to consumption of raw vegetables.

27. Raw Vegetables Consumption as A Potential Risk Factor for Campylobacteriosis in Malaysia

28. FARMS WET MARKET SUPERMARKETS OUR FINDINGS • We found presence of campylobacters especially C. jejuni in farm and retail vegetables. • The prevalence and concentration of C. jejuni in freshly-harvested vegetables are lower compare to retail raw vegetables.

29. Irrigation water Prevalence: non detected Freshly-harvested vegetables Prevalence: 18.8% (3.0-150.0 MPN/g)  Aged manure 3.0% (9.1 MPN/g)  composted manure Soil Prevalence: 30.4% (3.0-9.1 MPN/g)  Aged manure 2.7% (6.1 MPN/g)  composted manure Poultry Manure Prevalence: 57.1% (3.0-9.3 MPN/g)  Aged manure Non detected  composted manure Field study: vegetable farms

30. FARM LEVEL animals? Soil? Irrigation water? Raw Vegetables poultry manure? Environmental? Poultry? Other animals? vehicles? workers? containers? workers? workers? RETAIL LEVEL Cross-contamination from poultry? Meat? Water? Holding time Holding temperature containers? Raw Vegetables KITCHEN Contact with meat? Washing water? Spraying water? Raw Vegetables Consume raw/ salad Cook How to reduce? Washing  rate of reduction Blanching  rate of reduction RISK? NO RISK

31. Simulation of cross-contamination and decontamination of C. jejuni during handling of raw vegetables • The simulation was designed to simulate the real preparation of salad in a household kitchen starting from washing of vegetables in tap water; cutting the vegetable on cutting board; followed by slicing cucumber and blanching (heating in hot water) the vegetables in 85oC water. • Vegetables naturally contaminated with C. jejuni were used throughout the simulation to attain realistic quantitative data. • The mean of the percent transfer rate for C. jejuni from vegetable to wash water was 30.1% - 38.2% ; wash water to cucumber was 26.3% - 47.2%; vegetables to cutting board was 1.6% - 10.3%; and cutting board to cucumber was 22.6% - 73.3%. • The data suggest the wash water and plastic cutting board as potential risk factors in C. jejuni transmission to consumers. • Washing of the vegetables with tap water had a 0.4 log10 reduction of C. jejuni attached to the vegetables (MPN/g); while rapid blanching reduced the number of C. jejuni to an undetectable level.

32. Retail-to-Fork Model: a simple deterministic risk assessment Retail raw vegetables * Consumption * Consumption * Washing * Blanching * Consumption *Washing: wash in a bucket of tap water (30oC) for about 20 min. The log reduction rate is 0.36 log MPN/g. *Blanching: blanching was done by dipping raw vegetables in 85oC hot water for 10 sec. The removal efficiency is 0.95. * Locals consume 3.7g (mean) of raw vegetables per serving (MOH Food Consumption in Malaysia, 2007). * the dose-response model for C. jejuni was adapted from a report by Joint FAO/WHO Expert Consultation on Risk Assessment of Microbiological Hazards in Foods (FAO/WHO 2002).

33. Retails Wet Market Supermarkets WW W WB WW W WB WW W WB WW: Consume without washing W : Consume after washing WB: Consume after washing and blanching

35. The presented risk model is just a preliminary risk assessment aimed to demonstrate the risk of Campylobacteriosis via raw vegetables consumption. • Washing as proposed in this study has a certain efficiency in reducing the risk. However, washing methods with higher reduction efficiency (1 log reduction) are able to reduce the risk by 80%. • Other washing practices like wash under running tap water, wash and soak in salt water, etc. might have higher log reduction rate.

36. Biofilm formation and persistence of Salmonella

37. Salmonella clones can persist in many locations for many years • Persistence not due to resistance against • heat • disinfection • dry conditions

38. BIOFILM I can’t go with the flow anymore I’m thinking of joining a biofilm

39. Bacteria on surfaces Adhesion Microcolonies Structures Matrix

40. Two weeks old biofilm in an industrial condencer Photo from ASM Microbe Library

41. Liquid-air interface (pellicle) Salmonella biofilm Inorganic surfaces

42. Organic surface Close up Under magnifying glass SEM

43. Salmonella biofilm production

44. Biofilm formation at room temperature on different surfaces

45. Correlation between persistence and biofilm forming abilities at room temperature Amount biofilm produced in 48 hours Persistent strains Presumed non- persistent strains

46. Correlation between persistance and pellicle formation at room temperature

47. Long time persistence in biofilm Lg cfu 10 8 6 4 2 4 months dessication and nutrient depletion After 4 months At start

48. Effect of disinfection on salmoella in biofilm

49. Malaysia Fishery Products Export to the EU & Issues Encountered

50. Malaysia Fishery Export • Malaysia fishery products export is valued at RM 2260 million per year • About 50% of Malaysia fishery products are exported to the European Union (EU) and United States of America (USA)