The Metabolic Link between E. coli and C. hutchinsonii
When C. hutchinsonii is grown on pure crystalline cellulose, soluble sugars (including glucose, cellobiose, cellotriose, and cellotetraose) accumulate extracellularly. It is hypothesised that C. hutchinsonii is not able to efficiently use all of the soluble sugars it generates, such as xylose (Xie et al. 2007) (see the C. hutchinsonii page for more information).

E. Coli is capable of using some of these sugars, such as glucose and xylose (Desai and Rao 2010), for its own growth. Thus, there is a hypothesized metabolic link between C. hutchinsonii and E. Coli. We engineered E. Coli to produce fatty acid ethyl esters from these sugars alone, enabling direct conversion of lignocellulose to biodiesel by the C. hutchinsonii and modified E. Coli co-culture.

Co-culture Instability
Zuroff et al. (2013) predicted that “synthetic syntrophic interactions may be unstable” since at least one organism does not necessarily rely on the other for survival. Without an additional level of control the community may breakdown. Obligate mutualisms such as those described by You et al. 2004 and Shou et al. 2007 may be a more stable approach for consortia-mediated lignocellulosic ethanol production” (Zuroff 2013).

Our naive co-culture experiment showed that the C. hutchinsonii and E. Coli co-culture is neither stable nor efficient without the addition of a population control system - C. hutchinsonii vastly dominates the community, not allowing E. Coli enough access to sugars. Our task was then to design a population control circuit to address this problem.
Naive Co-culture Experiment
We needed to examine how E. Coli and C. hutchinsonii grow together naturally so as to know in what way we should engineer them to improve the efficiency and stability of the co-culture. The hypothesis was that due to C. hutchinsonii's slower growth rate, E. Coli would grow too fast, use up all the glucose and die, possibly also critically decreasing the C. hutchinsonii population size in the process.
Methods for Naive Co-culture
The preliminary co-culture experiment served to determine how unmodified C. hutchinsonii and E. Coli grow in mono-cultures versus co-cultures over time. Five cultures with co-culture media, filter paper as a carbon source, and variations of E. Coli C. hutchinsonii populations were prepared, with replicates of each, as shown in Table 1. The conditioned media was co-culture media that contained filter paper and C. hutchinsonii for three days. The cultures were incubated in a 30ºC incubator shaking at 250 rpm. Samples were taken before incubation, every 3 hours for the first 12 hours and once every 24 hours on days 4-8. The cells and supernatant from each sample were isolated via the co-culture protocol. We used the flow cytometer to measure the size of cell populations from the glycerol stocks of the cells, as detailed in the dry notebook.

Table 1
Contamination control in unconditioned media Only E. coli in unconditioned media Only C. hutchinsonii in unconditioned media E. coli and C. hutchinsonii in conditioned media E. coli and C. hutchinsonii in unconditioned media
Media 20ml unconditioned co-culture media and filter paper 20ml unconditioned co-culture media and filter paper 20ml unconditioned co-culture media and filter paper 8ml conditioned co-culture media, 12ml unconditioned co-culture media, and filter paper 20ml unconditioned co-culture media and filter paper
C. hutchinsonii None None
E. coli None None
In the future, we would further tests to measure robustness which include defining parameters for stability in different bioreactor conditions, such as varying temperatures and pH. In addition, we would build a modular circuit design allows for interchangeability between species of bacteria with different growth conditions, inputs, or outputs.
The data from the preliminary co-culture showed the sugar levels (from the conditioned media) initially sharply dropping and the E. Coli population increasing and overtaking the C. hutchinsonii population. This is expected, because E. Coli naturally divides more quickly than C. hutchinsonii. However, the E. Coli population levels out after approximately 10 hours due to sugar depletion, meanwhile the C. hutchinsonii population begins to increase around this time as it starts breaking down the filter paper and consuming the sugars it releases through this process. C. hutchinsonii soons overtakes the E. Coli and the carbohydrate level increases again as the filter paper gets degraded into polysaccharides and simple sugars. The E. Coli population then increases more slowly, as it is able to feed off the simple sugars that escape the C. hutchinsonii, but it cannot consume the polysaccharides.
These results show that the C. hutchinsonii is very efficient at consuming the sugars it generates before it can diffuse into the media, leaving little left for the E. Coli. In order to achieve more efficient bioprocessing, we’d have to introduce a circuit that would limit the growth of C. hutchinsonii according to the amount of E. Coli.
Desai and Rao, Appl Environ Microbiol., 76(5):1524-32 (2010)

Shou et al., Proc Natl Acad Sci USA, 104: 1877–1882 (2007)

Xie et al., J Exp Bot., 58(11):2799-810 (2007)

You et al., Nature, 428(6985):868-71 (2004)

Zuroff et al., Biotechnology for Biofuels, 6:59 (2013)