Team:ETH Zurich/Project/Motivation

Why in this case simpler is better ...

Building new functions into cells is the goal of synthetic biology. To achieve this, circuits encoding the desired behavior have to be inserted into an existing network of already considerable complexity where many of the interactions are not completely understood. By reducing the complexity of the target organism, these interdependencies can be reduced and more predictable results can be obtained.

One of the main engineering goals in this field is to make the design of new circuits more deterministic, allowing for predictive mathematical models and for simulations of new functions before these are implemented in vivo. Cross-talk between the different pathways in the organism and the additionally implemented circuits can lead to interferences, making the behavior non-deterministic. In such cases "debugging" of the inserted circuit can be extremely hard. A minimal cell would provide a more predictable chassis when used as an engineering fundament.

Minimal genomes

While the human genome contains around 30,000 genes (1), a worm needs 19,099 genes (2) to survive and reproduce, while a fruit fly can live on just 13,600 genes (3). The smallest free-living eukaryote Ostreococcus tauri has a diminutive genome of just 8,166 genes (4), while baker’s yeast has a mere 6,563 genes. Organisms that do not indulge in the luxuries of multicellularity or a nucleus can live a happy life with even less genes. The standard laboratory bacterium Escherichia coli is happy with just 4,467 genes (5, 6) and bacteria that live parasitically inside other cells like Mycoplasma genitalium and Carsonella ruddii have record genomes of just 470 (7) and 182 (8) protein coding genes, respectively.

How many genes does a cell need in order to be able to survive and reproduce? Despite the wealth of genomic information that has been accumulated it is very difficult to answer this question in an absolute manner, since the number of genes required for survival is heavily depending on the conditions of growth. Single gene targeted deletion studies and the comparison between genomes shows that around 200 genes are absolutely essential (9), because they encode proteins that are needed to synthesize and handle DNA (such as DNA topoisomerases, DNA synthase), produce RNA (RNA polymerase) or translate the genetic information into proteins (ribosome, elongation factors, tRNAs). If the cell is provided with all nutrients required to build its DNA and proteins and has a source of energy, those approximately 200 genes are more or less all the cell needs for survival. Indeed, cells that live parasitically within eukaryotic or bacterial cells have some of the smallest genomes reported.

However, as soon as the cell lives in conditions that do not provide it with all the nutrients and energy it needs, it has to have additional genes that encode enzymes for biosynthesis of nutrients and energy conversion in order to survive and reproduce. If different sources of energy are available at different times, it will have to determine concentration of different compounds in the nutrient medium and react to changes in concentration (as E. coli does with the lac operon). If the environment grows more adverse, heat shock proteins, antifreeze proteins and other equipment might be necessary for survival. Therefore, the number of essential genes is directly linked to the growth conditions of the organism.

If a minimal genome for a certain condition is sought, there are in theory two ways to achieve it: bottom-up and top-down. In the bottom-up approach, the absolute minimal set of genes is used to begin with and additional genes are added on to enable the organism to survive under defined environmental conditions. Craig Venter’s recent complete synthesis of the Mycoplasma genitalium genome (10) and experiments towards transformation of whole genomes (11) take this direction. The top-down approach, on the other hand, would start with an established laboratory organism like E. coli and remove gene after gene either until the organism is not viable under the specified culture conditions (if a true minimal genome is sought, the last step before the organism becomes unviable would constitute the minimal genome for the specified culture conditions) or until the organism grows fastest (if elimination of evolutionary baggage and the result of a biotechnologically optimized genome is sought).

E. coli has been living in the intestine of animals for millennia and is perfectly adapted for this environment. Although it also grows extremely fast in shake flasks on standard culture media, the standard laboratory strains of E. coli are not adapted to the artificial environment in the sense that they still contain various biosynthetic pathways that are not needed for growth in standard glucose media and that therefore only add to the costs of protein synthesis without increasing the efficiency of the organism. Extensive research on the experimental evolution of E. coli by Richard E. Lenski have demonstrated that the growth rate of E. coli in shake flasks and glucose medium can already increase dramatically if cells are cultured for just 5,000 generations (13). During this process of adaptation, cells lose the ability to metabolize various energy sources that are not present in the medium and specialize on the sources that are provided. We could, therefore, postulate that it would be possible to create an E. coli with an optimized genome which has lost all biosynthetic pathways except for the ones necessary to metabolize the substances provided in the medium and synthesize the substances that are lacking.

Engineering chasis

Our main motivation is to develop a method to reduce the complexity of the target organism dramatically. This will allow to characterize the biochemical reactions that are necessary to sustain life, eliminating all the "ballast" that may interfere with the inserted functions.

The result of this work work would be a "cell factory" with more predictable and controllable behavior, allowing for a more deterministic approach to metabolic engineering. Many groups have been working on this field [xxx]. Using organisms as varied as obligate parasites like Mycoplasma genitalium (580 kbp), Escherichia coli (4.6 Mbp) to Sacharomices cerevisiae (6Mbp). In all cases, the goal is to obtain a laboratory strain that can be used as a platform in synthetic biology.

While the genomic sequence of E. coli has been determined (5, 6) and single-knockout libraries exist (14), we still are just beginning to understand E. coli as a whole system. Therefore, it is very difficult to decide based on genetic and biochemical information which genes should be deleted. In addition, it can be expected that disrupted biosynthetic pathways reroute and form novel pathways that can not be predicted based on the data available to date.

By reducing the complexity of the organism, we are trying to move from a network with a large number of (unknown) interactions like this:

Fig 1: Metabolic network of e.coli

to well characterized system, in which a complete characterization of possible interactions can be obtained. The result of this effort should provide us with something more in line with this:

Fig 2: Schematic representation of a simpler metabolic network

In this way we are trying to get rid of "annoying" effects like redundancy, unused genes, complex regulation and cross-talk. Our engineering chassis should provide a basis with known functions and a nice platform for the implementation of orthogonal circuits. Recently, a cleaned-up version of E. coli containing a 15% smaller genome in which cryptic sequences, biosynthetic pathways that are not needed under standard culture conditions and other genes had been deleted by targeted large- scale gene disruption has been presented (12). Despite having fewer genes than the standard E. coli, the cells with reduced genome grew slightly denser in certain media, tolerated a greater variety of plasmids and had higher electroporation efficiency. These results can be seen as evidence, that reducing the complexity of the network will provide a neutral environment, where biochemical reactions can take place in a more predictable way.

Goal – self-optimizing E. coli

We postulate that random disruption of genes combined with an evolutionary approach in which growth speed determines fitness and survival would be a viable approach to obtain a cell that is perfectly adapted to laboratory growth conditions and has jettisoned all genes that are not contributing towards growth and reproduction under the specified growth conditions. Furthermore, the cell could be adapted rapidly to different conditions and – with appropriate experimental design - it could be rapidly optimized for recombinant expression of proteins.

In order to prove the viability of the concept, we intend to demonstrate that ..

a) .. excision of genomic sequences by restriction enzymes can be performed within the living cell;

b) .. fragmented genomic DNA can be ligated inside the living cell by simultaneous expression of T4 ligase;

c) .. pulse generators that produce pulses of protein expression of variable length terminated by induction with a second compound can be devised both on the transcriptional as well as on the translational level;

d) .. cells with optimized growth and reproduction can be selected by competition inside chemostat cultures.

References

(1) Lander, E. S. et al. (2001) Initial sequencing and analysis of the human genome. Nature 409, 860-921.

(2) (1998) Genome sequence of the nematode C. elegans: a platform for investigating biology. Science 282, 2012-8.

(3) Adams, M. D. et al. (2000) The genome sequence of Drosophila melanogaster. Science 287, 2185-95. (4) Derelle, E.et al. (2006) Genome analysis of the smallest free-living eukaryote Ostreococcus tauri unveils many unique features. Proc Natl Acad Sci U S A 103, 11647-52.

(5) Blattner, F. R. et al. (1997) The complete genome sequence of Escherichia coli K-12. Science 277, 1453-74.

(6) Riley, M. et al. (2006) Escherichia coli K-12: a cooperatively developed annotation snapshot--2005. Nucleic Acids Res 34, 1-9.

(7) Fraser, C. M. et al. (1995) The minimal gene complement of Mycoplasma genitalium. Science 270, 397-403.

(8) Nakabachi, A. et al. (2006) The 160-kilobase genome of the bacterial endosymbiont Carsonella. Science 314, 267.

(9) Kobayashi, K. et al. (2003) Essential Bacillus subtilis genes. Proc Natl Acad Sci U S A 100, 4678-83.

(10) Gibson, D. G.et al. (2008) Complete chemical synthesis, assembly, and cloning of a Mycoplasma genitalium genome. Science 319, 1215-20.

(11) Lartigue, C.et al. (2007) Genome transplantation in bacteria: changing one species to another. Science 317, 632-8.

(12) Posfai, G.et al. (2006) Emergent properties of reduced-genome Escherichia coli. Science 312, 1044-6.

(13) Cooper, V. S., and Lenski, R. E. (2000) The population genetics of ecological specialization in evolving Escherichia coli populations. Nature 407, 736-9.

(14) Baba, T.et al. (2006) Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2, 2006 0008.