Team:Groningen/in development



We proved the principle of our Food Warden system by developing a construct that enables Bacillus subtilis to form a pigment as a reaction to rotten meat. We designed different constructs which can improve the function of the Food Warden by tuning the pigment production or turning the Food Warden into a multi-color system.

Tuning system of Pigment Production

The core concept behind the Food Warden is that it should pave the way to a more comprehensive, scientifically informed prediction of food edibility that goes beyond conventional best-before dates. The Food Warden as it is now is only a proof of principle. The goal is to produce a system that is truly more accurate and reliable than the best-before date. The tuning of this device requires a comprehensive study on the relationship between volatile concentration, degree of spoilage health risk and pigment production:

  1. Volatile concentration: Building upon our gas chromatography approach in order to quantitatively assess the volatile production of spoiling meat.
  2. Degree of spoilage health risk: The undefined nature of current assessments of spoiling degrees in food (see Stop the food waste initiative) make this a difficult step. A time resolved total microbial count analysis could be done to assess edibility in terms of this standard for specific types of meat.
  3. Pigment production: The control of pigment production dynamics will depend on the outcome of the previous two aspects of the tuning procedure. Once the relationship between volatile composition/concentration and health risk is elucidated to some degree, the pigment production can be tuned to fit this parameter.

The pigment production can be tuned to a desired speed and sensitivity with different regulating promoters, different rbs and with a positive feedback system to increase the pigment production. One example of the positive feedback system that can be applied to increase pigment production under the regulation of the rotten meat promoter is shown below:

Hover your mouse over the image to see a bigger version!

When the rotten meat promoter is activated, the pigment and inducer will be produced. The positive feedback loop is designed in a way that the pigment and the inducer will be in the loop, increasing the production rate of the pigment. This system is meant to increase the production speed of the pigment. One of the possible set of inducible promoter-inducer is pRE promoter (BBa_K116603) with CII (BBa_K116602).

Multi-colored Pigment System

The pigment system consists of two signals, which are quite simply 'on' or 'off'. There are some disadvantages to this system in terms of user-friendliness that need to be addressed. The Food Warden can only do its job if it can grow properly upon breaking of the inner compartment of the sticker. It is plausible that manufacturing errors during the production of an eventual Food Warden product could lead to issues with the germination of the spores, resulting in a sticker that does not do its job. Eventualities include:

  1. Incorrect medium: In the event that the medium supplied in the sticker is not of the correct composition, the Food Warden will not grow, and therefore will not be able to identify spoiling meat.
  2. Absence of viable spores: A defect sticker could be accidentally produced that either lacks spores capable of germination or lacks spores entirely.
  3. Contaminant organism: It is conceivable that the spores could be outcompeted in a medium that is contaminated with another organism. If the user is not aware of these problems he or she may assume that the lack of pigment production simply means the meat in question is not spoiling yet, whereas it may already be spoiling and indeed will without any warning.

Although these eventualities are a production process concern, it is our job to produce a system that minimizes the risk of these problems affecting the user. Therefore, we would have liked to build in a positive control that allows the user to confirm that the Food Warden in their sticker germinates and grows. The positive control considered was the inclusion of a constitutively produced second pigment. This pigment serves as a signal to the user that the Food Warden is functional upon germination. This positive control pigment should of course not interfere with the visual detection of the warning pigment. To ensure this, the following options were considered:

  1. Choosing the two pigment colors such that the warning pigment color is highly dominant over the control pigment color.
  2. Placing the pigment under the control of a weak constitutive promoter, producing the control pigment at minimum levels required for user detection, therefore more easily being overpowered by the warning pigment.
  3. Placing the pigment under the control of a constitutive promoter, identified in our microarray analysis, that downregulates the gene it controls under spoiling meat conditions, allowing the warning pigment to more easily overpower the control pigment (go to the Sensor page for more information on the downregulated genes and operons).
  4. Including an operator for the transcription of the control pigment that allows repression by a repressor protein. This repressor protein would be under the control of the same promoter responsible for warning pigment transcription. The construct can be put as following:

    Hover your mouse over the image to see a bigger version!

  5. Placing a control pigment that reacts to a specific compound to change its color. The control pigment is under the regulation of a constitutive promoter while the rotten meat promoter regulates the compound needed to change the control pigment's color. The construct can be put as following:

    Hover your mouse over the image to see a bigger version!

    When the rotten meat promoter is activated, the compound needed to change the control pigment's color and the repressor will be produced. The control pigment production will be stopped and the existing control pigment will start to react to the compound, changing its color and becoming the warning pigment. This can be achieved using existing biobrick: BBa_K274100 (lycopene, red pigment) and adding the compound (CrtY) encoded by BBa_K118008 to make the red color from lycopene into yellow color (beta-carotene). Another construct using positive feedback loop for the warning-pigment-compound without stopping the control pigment production is as following:

    Hover your mouse over the image to see a bigger version!

    In this system, the control pigment will still be produced even when the rotten meat promoter activates. The production speed of the second pigment compound which reacts to the control pigment to create new color is enhanced by the positive feedback loop, so the warning pigment color can be immediately produced.

This idea basically mimics the traffic light function: different colors production for every state of the meat. When the meat is still fresh, a specific pigment will be produced. When the meat starts to rot, another pigment will be produced overriding the previous pigment.

Update! (26th October 2012)

After the European regional jamboree, we succeeded to make a new construct: AmilGFP under regulation of PwapA (rotten meat downregulated promoter). The construct is a pilot construct for the following diagram: pigment 1 is regulated by the rotten meat down-regulated promoter (wapA) and pigment 2 is regulated by the up-regulated promoter (sboA).

Hover your mouse over the image to see a bigger version!

When meat is still fresh, the PwapA will regulate a production of pigment 1 while pigment 2 (that is regulated by PsboA) is absent. When the meat is rotten, PwapA will be down-regulated, thus decreasing the production of pigment 1 and PsboA will be upregulated, producing pigment 2. For example: If pigment 1 is a yellow pigment (amilGFP) and pigment 2 is a blue pigment (amilCP), a strong yellow color will be produced when the meat is still fresh and when the meat is rotten, more blue pigment will be produced. So when the meat is rotten, green color pigment is obtained.

Towards a psychrotrophic chassis

A different bacterial chassis is needed for further application of our volatile detection system. This is due to the inability of Bacillus subtilis to grow in a low temperature environment. The main criteria our chassis must:

- Grow in a low temperature environment (psychrotrophic bacteria).
- Form endospores, essential for effective storage and our 'sticker' activation mechanism.
- Exhibit non-pathogenic activity.

Literature study suggests the use of psychrotrophic Gram-positive bacteria from the Bacillus family as the new bacterial chassis for the volatile detection system, especially due to their ability to form spores. Although new strains of psychrotrophic Bacillus, such as the soil dwelling Bacillus hunanensis, have been discovered recently, their possible pathogenic characteristics are not yet described in literature (Chen et al. 2011). Some B. cereus strains are known to be able to grow in <7°C (Dufrenne et al. 1995), while some strains (e.g. B. cereus strain Toyoi) have been widely used as probiotic (Taras et al. 2005) & (Duc et al. 2004). However, no strain of B. cereus was found to fit all the chassis criteria, as the psychrotrophic strains tend to be pathogenic.

On the other hand, a study (Shehata, Collins 1971) reported a B. subtilis strain, RH22, exhibited a cold resistance phenotype when they were grown in 5°C. To further investigate and hopefully find our own psychrotrophic Bacillus subzerus strain (a ** joke), we can combine the sequencing protocols as described in this study, with, in the case of Bacillus cereus relatives, the use of 16S rRNA sequence analysis for detailed identification of microorganisms (Bavykin et al. 2006).

There have been studies where modulation of bacterial temperature tolerance has been achieved by introducing chaperonin genes from arctic bacterium Oleispira antarctica into Escherichia coli(Ferrer et al. 2003). Unfortunately, no published results were found that applied analogous strategies to Bacillus strains.


  1. Bavykin, S.G., Lysov, Y.P., Zakhariev, V., Kelly, J.J., Jackman, J., Stahl, D.A. & Cherni, A. 2006, "Use of 16S rRNA, 23S rRNA, and gyrB gene sequence analysis to determine phylogenetic relationships of Bacillus cereus group microorganisms (vol 42, pg 3711, 2004)", Journal of clinical microbiology, vol. 44, no. 7, pp. 2676-2676.
  2. Chen, Y., Hao, D., Chen, Q., Zhang, Y., Liu, J., He, J., Tang, S. & Li, W. 2011, "Bacillus hunanensis sp nov., a slightly halophilic bacterium isolated from non-saline forest soil", Antonie Van Leeuwenhoek International Journal of General and Molecular Microbiology, vol. 99, no. 3, pp. 481-488.
  3. Duc, L., Hong, H., Barbosa, T., Henriques, A. & Cutting, S. 2004, "Characterization of Bacillus probiotics available for human use", Applied and Environmental Microbiology, vol. 70, no. 4, pp. 2161-2171.
  4. Dufrenne, J., Bijwaard, M., Tegiffel, M., Beumer, R. & Notermans, S. 1995, "Characteristics of some Psychrotrophic Bacillus-Cereus Isolates", International journal of food microbiology, vol. 27, no. 2-3, pp. 175-183.
  5. Ferrer, M., Chernikova, T., Yakimov, M., Golyshin, P. & Timmis, K. 2003, "Chaperonins govern growth of Escherichia coli at low temperatures", Nature biotechnology, vol. 21, no. 11, pp. 1266-1267.
  6. Shehata, T. & Collins, E. 1971, "Isolation and Identification of Psychrophilic Species of Bacillus from Milk", Applied Microbiology, vol. 21, no. 3, pp. 466-&.
  7. Taras, D., Vahjen, W., Macha, M. & Simon, O. 2005, "Response of performance characteristics and fecal consistency to long-lasting dietary supplementation with the probiotic strain Bacillus cereus var. toyoi to sows and piglets", Archives of Animal Nutrition, vol. 59, no. 6, pp. 405-417.