(Repetition of  part of Voeding en Gezondheid (Artikelen) Start page section)

Food Microbiology (Revisited)

Preservation, Risk Perception and Risk Taking

To Derrick Kilsby, my mentor

My professional background is food microbiology and preservation. This essay (written in February 2003)  starts from a strong food perspective but will ultimately encompass society. We have enough food for mankind if we know how to preserve whilst guaranteeing safety. A proper balancing of risks will determine whether we shall ultimately survive.

Food should be safe and stable. Consuming food, however, is balanced risk taking. Necessary risk taking: Simply, not to starve, to reproduce and to grow old. Romantic risk taking: Picking raspberries (parasite hazard) in a tick infested forest. Hedonistic risk taking: eating Japanese puffer, - putting one’s life in the hands of an expert cook, - raw oysters, raw milk based cheese or undercooked hamburgers. One may even wish, in future, to create non-addictive thrilling near death sensation, in essence, bungee-jumping from the edge of Martin Cole’s food processing cliff. What about the risk taking by governments? What kind of risks were governments willing to take in the first stage of the BSE crisis in Europe? How many vCJD deaths are now acceptable and how does that compare to Salmonella and Campylobacter?

The dioxin contamination of chicken in Belgium had a health benefit. The accompanying temporary reduction in chicken consumption led to a reduced mortality due to Salmonella and Campylobacter. The media hype due to the lethal outbreak of EHEC in Germany in 2011 will certainly have negative long term health effects because of the steep drop in consumption of fresh fruit and vegetables. How to find the best balance in risk taking? Fish contains healthy omega fatty acids, but also mercury. Fresh produce contains vitamins and antioxidants, but may indeed carry pathogens, parasites. Microorganisms are not only scary but can be modified to deliver vitamins in the GI-tract. How to make objective decisions based to address the world needs?

Lees meer/ READ MORE (Continued)

How can we develop our capability to make objective decisions and gain the trust of the consumer? My training in the logical principles of Biology and Science has helped me to develop Preservation Microbiology (defensive food microbiology). The aim of preservation is to destroy harmful microorganisms or prevent microbial growth, thus assuring the quality and safety of the treated food. I have never lost interest in the other site of the medal. How digestion of microorganisms may contribute to health. How to find and calculate the best balance between quality and safety. This essay is not only aimed at stimulating the interest in the fascinating field of Food Microbiology. It is also a wake-up call, to revisit the assumptions and initiate a debate about the direction. How can we further develop science in an increasingly complex or n-dimensional world in a more vulnerable individualized, ageing or starving world? An n-dimensional approach requires efficient creative scientists with a sense of wonder and objective analytical power. How to find order in complexity, even Chaos, to accept Uncertainty in a quest for solutions?

How to assist in true matters of Life and Death?

Pieter ter Steeg



Food microbiology has been historically empirical and empiric laws (e.g. Low acid canned food regulation requiring an F0 = 3 minutes at 121°C for ambient stable low acid), were set in our minds without connection to the underlying forces/biological laws. Progress can be made in any area if systems have been made transparent and decisions are based on logical principles. Accordingly, the concept of a preservation universe or Preservation Hyperspace was proposed. Microorganisms can be considered as “residing” in an n-dimensional hyperspace.  The number of dimensions is determined by all the factors, which affect the presence, process survival and subsequent spoilage/pathogenicity potential of the microorganisms in foods. Applying this concept and biological principles such as niches, survival of the fittest, and evolution, can create provocative insights and identify research areas and focus university programmes.

Safety or spoilage?

The first question is should one focus on safety or spoilage?

Government and academia tend to focus on safety, because the consumer warrants protection. Also, pathogens receive more, in fact even too much press coverage. But is it also logical, whilst 30-40% of our food is lost due to spoilage. The world would not know starvation, if we addressed spoilage adequately. How many people die from undernourishment or have an impaired quality of life compared to foodborne pathogens? Spoilage microorganisms are more also robust. The preservation hierarchy is:

1) Commercial sterility

2) Microbiological stability (ensured by a combination of preservation dimensions)

3) Safety

If commercial sterility is ensured, or foods are microbiologically stable, pathogenic microorganisms have often implicitly been dealt with in the process and formulation design. For example, conventional industrial sterilization processes should meet a minimal heat process at the coldest point. (an F0 value of 3’ at 121°C) to ensure a 12 log reduction of the target organism for safety: Clostridium botulinum. Conventional industrial retort processes, however, are very often not set for quality or stability. For example, a static retort process can be set at an F0-value of 5 minutes at the coldest spot. Rotary retort is often set at a higher F0-value of 8 minutes.  In rotary retort the coldest spot has a much larger volume containing more spores to be killed. Not surprisingly, these surviving mesophilic thermotolerant spores having nothing to do with safety, but everything with spoilage.

If one is able to put down his safety preconception, questions can pop-up leading to understanding: What is the actual selection pressure in a particular set of dimensions or niche of the hyperspace? Which organism is the fittest and becomes accordingly the process and/or formulation bottle-neck? Which organism should be studied?

A few candidates:

Bacillus sporothermodurans for Ultra-short High Temperature-processes

Thermoanaerobium asaccharolyticum for hot vended drinks (60-70°C)

Alicyclobacillus acidoterrestris in pasteurized acidic fruit juices at elevated temperatures

For any new set of dimensions one cannot guarantee that the rules of the selection game have not changed and other players become more successful and will occupy y the new niche. A historic example has been Deinococcus radiodurans for irradiated foods. What will be the process bottle-neck microorganism for novel technologies like adiabatic sterilization (123°C at 700 MPa) or ultra-short UHT treatments (0.1 second at 160°C). What about emerging pathogens, do they have a chance to occupy a novel niche:

Do emerging pathogens exist?

Emerging pathogens seem to continuously evolve and pop us as “Aliens” in the Hyperspace and cause unsuspected illness or spoilage. As a biologist one can also question, whether emerging pathogens really evolve or is it only about niche amplification (exotic imports, change to large scale manufacture and distribution) or improved methodology. Prions, the causative agent of BSE and the human form of Creutzfeld-Jacob disease have been the most frightening “Aliens” of the last two decades of last millennium and may even start causing problems in the Third World. Prions can be considered one of the most unusual forms of life multiplying via a domino protein conversion mechanism not requiring new protein synthesis and generally “moving” to a new victim via animal feed. One requires extreme heat treatments to eliminate prions in contaminated animal protein (60 minutes at 132°C).

Are Escherichia coli O157 or other enterohemorrhagic non O157 E. coli (EHEC) really emerging pathogens?  Karl Bettelheim (Australia) started detecting these organisms before 1970. He even states that O157 may be just a red herring and may not even be always the cause of disease. Data exist where less pathogenic O157 were co-isolated with non-sorbitol negative more virulent EHEC or viruses. How emerging and new can such a pathogen really be, if the common ancestor of non-pathogenic E. coli and the serotype O157 is dating from 4.5 millions years ago.

Evolution of emerging foodborne pathogens (best examples are multi-drug resistant bacteria) will generally only occur in environments in which there is a continuous selective advantage and re-establishment loops.  Their success depends on hosts. Outside the host they will be out-competed. E.g. the non-pathogenic Listeria innocua does not have to wear the energy burden to maintain pathogenicity islands and will outgrow its virulent sister Listeria monocytogenes in the factory environment or enrichment culture. It also strongly suggests that there should be also true niches for organisms like L. monocytogenes and enterohemorrhagic E. coli in nature.  Eradication strategy can only provide temporary solutions because it is a biological law, if there is a niche out there, it will be filled. Vaccination of flocks against one Salmonella serotype will favour domination by a different serotype.

Review of research priorities

Food microbiology has been a largely empirical field. Observations have often been restricted to the dimensions of local galaxies (product groups, research networks) in the hyperspace without the need to communicate in unifying forces like in physics. Researchers at academia often adopted approaches of mainstream microbiology without realizing the unique opportunities in preservation microbiology. Just consider the flow from farm to fork and apply the hyperspace concept, consider the n-dimensions, which affect the process survival and subsequent outgrowth potential and wonder about what we need to know. A grouping in preservation dimensions is presented below.  The microorganism is regarded as a preservation target with a physiological history which may influence its process susceptibility. The microorganism (vital, injured ord ead) leaving the process has to face the endproduct conditions:

Grouping of preservation dimensions

1)      The input :

  1. The microbial load and diversity in raw materials and ingredients.
  2. The organism’s state (e.g., process susceptibility, injury, acquired stress response, etc.).

2)      The process:

  1. The heating intensity and duration (pasteurization, sterilization, Radiofrequency Heating, Microwave)
  2. Alternative, non or reduced thermal preservation technologies and their parameters (e.g.: High Pressure Processing (HPP), electron beam (EB), filtration)

3)       The product (environment):

  1. Storage temperature, pH, water activity, organic acids, salt, chemical and biological preservatives (sorbic acid, nisin) and “cryptobials” (hidden antimicrobial factors).
  2. Microstructure, e.g., emulsion type, solid matrix’s strength.

4)      Economic and social factors such as:

  1. Consumer use, risk groups
  2. Logistic considerations, mass production, etc.

What kind of knowledge is required to build understanding to determine the success rate of existing and to design new milder preservation systems. One has to start to deviate from traditional approaches.

Ad 1) The input dimensions:

Knowledge or methodology is required to assess the microbial load and process susceptibility distribution. The focus of the methodology should be on the process limiting organisms. Pathogenic organisms that will anyhow die in a controlled process are irrelevant. The physiological prehistory of microorganisms can affect their susceptibility in mild processes. Many preservation processes are targeted against permeability barriers like the membranes or spore cortex. Pre-culture conditions (or growth phase (exponential or stationary)) will directly affect the membrane composition or spore cortex and process susceptibility. Any unconscious choice in studies for data acquisition of cells pre-cultured under optimal conditions like pH 7, 37°C instead of ambient or chill, without an acquired stress response can lead to knowledge of limited use. One has to address the right target species but also their physiological make-up and response in experimental and safe product and process design.

For fresh produce one may still wish to balance such an “ingredient based risk assessment and selection” versus end-product labeling to protect susceptible groups.

Ad 2) The process dimensions:


Start with considering a real process and become aware of the limitations of laboratory studies. Kilsby (2000) did already address methodology in “Food Microbiology: the challenges for the future”. He warned against self-delusion. Laboratory strains have lowered their wild-type constitutive stress-response because of a historic selection pressure of sub-cultivation under optimal conditions. These organisms (e.g. Bacillus subtilis 168) have a reduced genetic variability (loss of plasmids, insertion elements) and often only one prehistory, both implicit ways to control experimental variability and move away from reality: Vegetative cells are generally killed in seconds in conventional pasteurisation processes. How important can the relatively slow process of  stress response acquisition be in survival? It is but it does not warrant the current research efforts. Accept that organisms have their acquired resistance switched on. Anyhow, they will not get the time to rethink and reorganise their survival strategy in a real life process. It is more important to take the prehistory of our targets into account whilst designing mild preservation systems. Inactivation targets for vegetative cells (pasteurisation) are poorly defined (words from Robert Buchanan,  Food & Drug Administration). Prehistory can easily cause a 4 Decimal reductions difference in inactivation. What to do, if the law requires a 5D reduction whilst one is comparing a conventional heat versus pressure pasteurisation process and subcultivation temperature can cause an inverse process susceptibility difference of 4 logs.

The concept of traditional process measures like the Decimal reduction (D) value appears insufficient as inactivation generally follows non log-linear inactivation. There is a clear trend towards empirical distributional models. The challenge, however, is to increase also our understanding by establishing a link with the underlying biological distribution or physiology. Developments in single cell techniques like flow cytometry will increasingly allow the assessment of underlying mechanisms and resistance distributions in bacterial populations.

Ad 3) The product dimensions

Pathogens or surrogate microorganisms?

The probability of outgrowth and growth rate is a function of product (pH, water activity, …) and storage (T,  ..) preservation dimensions. Should we only model outgrowth of specific pathogenic species with a reduced genotype, whilst the species or serotype concept may have had its longest time (Kilsby, 2000), or functional ecophysiology? DNA can be exchanged across the species boundary. The definition of species is not clear. Adopting a strict taxonomic view of a species (e.g.> 70% DNA homology) will not help to select the proper targets for modelling studies but a functional ecophysiologal view might.  Ample evidence exists that the eco-physiological response of different strains of E. coli is not confined to a serotype. A generic model for the growth of a non-pathogenic strain such as M23 from Tom McMeekin’s group is sufficient to describe the growth of many Shiga toxin producing E. coli (STEC).  One can speculate whether a model based on such a single but ecophysiologically representative strain like M23 may last longer than predictive models of specific groups of so called emerging organisms. Many studies have used cocktails of E. coli O157: H7 whilst ignoring the 200 different STEC or very similar organisms like Shigella sonnei. These studies have not taken into consideration relatively simple exchange mechanisms like bacteriophage mediated virulence genes transfer? One may also question how long a model for a pathogen will be valid.  The microorganism might acquire a more efficient way to reproduce its virulence genes or to generate energy via DNA exchange or recombination.

Sublethal injury

Sublethal injury is often mixed up with acquired stress-response. Injured cells are often more susceptible to a subsequent stress.  Sub-lethal genuine injury is in my opinion not a threat, but an enormous research opportunity, as those cells are more susceptible to subsequent stresses/processes. How could a sub-lethally injured cell be more virulent than a healthy cell?

“Structuromics or Peter Setlows “Permeability barriers” (“Permeabolomics”). The first lines of defence for any microorganism is its barriers (cell wall, cytoplasmic membrane + proteins for vegetative cells, spore coat cortex and inner membrane of spores). As such Microbiology should combine their forces with Biophysics and Material Sciences. Peter Setlow made one very significant statement: Biophysicist seemed not to be willing to work longer than two weeks on bacterial spores. Technology has advanced and even for biophysicists microbiology and the intrinsic variability of living and genetically plastic cells should not be only frustrating.

Management tools ( 12 D concept for proteolytic bots and 6D concept for non-proteolytic bots , pasteurisation, 5D reduction of vegetative pathogens in processed juice).

Do we still need such performance criteria, often management tools, or is it possible to replace it by “Q.I.P.” (Quantitative Integrated Preservation), linking microbial susceptibility distributions in ingredients throughout the process to outgrowth probabilities under the final product conditions. How do we establish a 5D or 4D reduction requirement for vegetative cells while it is fairly easy to manipulate the outcome of experiments? Regulations like the Low Acid Canned Foods still require inactivation whilst growth inhibition will do, e.g. a combination of pH and heat and absence of oxygen.

Need to make preservation transparent

A major step forward would be to develop and communicate in a cross-product group scientific unifying language of preservation factors not hiding the active forces in preservation systems. E.g. undissociated acid instead of weight % acid, salts on water instead of % salt and moisture, solids, etc. The same holds for preservation processes. Thermal inactivation is still the preferred method. How can one compare the efficacy of alternative preservation technologies like manothermosonication or pulsed magnetic fields to conventional heat whilst ignoring local heat effects. Hypotheses to start with and proper measurement techniques bring understanding and identify the ways to move forward.

Quality Control versus Quality Assurance

Quality control is very useful when uncontrolled  ingredients like fresh produce enter the food processing chain. Quality Assurance, however, aims at finding the best balance in addressing safety issues, whilst taking into account the preservation and Quality requirements. This requires an effective integrated process and product design, which cannot be effectively controlled by a sampling strategy. Advocates of (on-line) Quality-Control run into the Dead or Alive Question and the uncertainty principle in Quantum Microbiology. Bridson and Gould  (2000) have already described a similarity between physics and microbiology. In both sciences there is an apparent dichotomy between the certainty and stability of the macro-subject and the uncertainty/complexity of the individual atom/cell. Classical physics is to quantum mechanics as classical microbiology is to quantal microbiology. A probability of a 99% reduction or 0.01 of survival will on single cell level be reflected by a quantum outcome of 0 (Dead) or 1 alive. There is not such a thing as 0.01 Alive. How can Quality Control cope with tolerable failure rates of 1 out of 100,000 containers, whilst the proof of the pudding will destroy the pudding. In essence, how to cope with Schrodinger's cat in a box in Quantum Physics.

Chemical or natural antimicrobials, Organic produce. More or less hygiene?

There is a general trend in society towards “organic” risk taking.  Is the consumer sufficiently informed to make a balanced decision in an ageing and more immunocompromised society? What is against a conventional chemical antimicrobial, if there are no negative side-effects on health. What is against eating unpreserved food with an increased probability of carcinogenic mycotoxins or foodborne pathogens? Are people aware that cows in a meadow have a higher probability of carrying EHEC than cows inside because of the diet that creates a niche for EHEC.? What is against hygiene (hand washing with soap) in day-care centres and primary schools, if it prevents viral and bacterial transmission? Who is in favour of wasting 20,000,000,000 euro on unnecessary BSE-measures not saving one life whilst 5,000,000,000 is sufficient to stop malaria globally saving millions of lifes. Besides the unnecessary destruction of 16 million tons of animal protein requires 23 million tons of soy and 64 -112 million tons of grain sufficient to feed > 500 million people,

The goal is objective balanced risk taking in order to be able to survive and not to die or continue the destruction of the world. We need decision making based on objective science accepted by consumers and governments.


Bridson E.Y., Gould GW. (2000) Quantal Microbiology. L. Appl.  Microbiol. 30 (2): 95-98.

Kilsby, D.C. 1999. Food Microbiology: the challenges for the future. Int. J. Food Microbiol. 50, 59-63.

Ter Steeg P.F. ter, Ueckert J.E. (2002) Debating the biological reality of modeling

preservation. Int. J. Food Microbiol. 73, 409–414.