This web page was produced as an assignment for an undergraduate course at Davidson College.

Design and Synthesis of a Minimal Bacterial Genome

test cycle

Figure 1. Process used to design, build, and test minimal genomes. Figure reproduced from Hutchinson et. al., 2016.

Project Summary and Discovery Science:

    Researchers used a cyclic design process to synthesize a minimized version of the Mycoplasma mycoides genome. Beginning with a previously-designed synthetic M. mycoides genome, referred to as syn1.0, they used transposon data to classify genes; genes that could be interrupted by a transposon without lethality were nonessential, while those that caused lethality when interrupted were essential. Synthesis of a hypothetical minimal genome of only essential genes failed to produce a viable genome because it did not retain genes that were unnecessary for life, but necessary for robust growth. A refined transposon method identified these "quasi-essential" genes.

    The genome was designed in eight segments, each of which could be tested independently within a seven-eighths syn1.0 genome. When quasi-essential genes were retained, a combination of redesigned segments and syn1.0 segments produced a viable genome, then refined into syn2.0. Finally, syn2.0 was optimized to produce syn3.0, a 531 kilobase construct containing 473 genes. Of those 473 genes, researchers could not determine the biological function of 149, implying yet-unknown functions that are essential to life.

    This paper demonstrates discovery science, because the researchers were not testing a falsifiable hypothesis. They created a novel iterative design-build-test cycle, and used this cycle to progressively minimize a bacterial genome.

Genomic Technology:

    In order to classify genes as essential, quasi-essential, or nonessential, researchers used Tn5 transposon mutagenesis analysis. Co-transformation of the syn1.0 genome and the Tn5 transposon produced thousands of colonies, each with a unique transposon insertion, making up the P0 population. Insertion sites in these cells were profiled using inverse PCR and DNA sequencing. These cells were then serially passaged for 40 generations to select for cells with faster growth rates, forming the P4 population. Transposon insertion sites were profiled in both populations.

    Genes with little to no transposon insertion in either the P0 or P4 populations were classified as essential, because any interruptions in these regions were strongly selected against. Genes with extensive insertion in both populations were nonessential, because the cells tolerated disruption of these genes. Genes with insertions in the P0 generation but not the P4 generation were classified as quasi-essential, or unnecessary for life but necessary for robust growth. Interruptions were tolerated, but not when robust growth had been selected for.

figure 5
Figure 1. Comparison of syn3.0 proteins with homologs. BLASTP was used to compare proteins encoded by the syn3.0 genome with homologs
in 14 other species. The x-axis categories correspond to ability to determine gene/protein function. Equivalog indicates complete certainty;
 confidence progressively diminishes from left to right. White space indicate no identifiable homolog. Figure reproduced from Hutchinson et. al., 2016.

    Researchers also used BLASTP alignment to analyze the similarity of syn3.0 genes to wildtype genomes of a variety of species. The visual alignment in Fig. 1 (Hutchinson et. al., 2016) shows that equivalogs of many genes retained in the minimal genome are widely represented across domains of life.

Take Home Message:

    Using transposon mutation data, researchers were able to build syn3.0, a functional synthetic genome smaller than any that has been discovered in nature. They retained many genes that they classified as unnecessary to life, but minimizing the genome further would have repercussion in growth rate. Of the 473 genes retained in the final construct, almost a third have unknown biological functions, suggesting further research is necessary to know more about these genes that are essential to life. Additionally, the cyclic design-build-test model and technology to synthesize a complete, error-free artificial genome could be used in projects beyond genome minimization.


    Overall, I found the project fascinating, specifically because of their simple approach to a complex system. I find it surprising they ever achieved a functioning genome, given the straightforward set of guidelines for deletions. My impression was that genetic code, specifically non-coding regions, is not well understood and thereby very sensitive; one wrong base in a promoter, terminator, or other regulator sequence could render the entire system non-functional. However, this paper implies that we understand most of the organizational principles of genetic code, and can delete entire non-coding segments without doing any damage.

    At the same time, it is surprising that we still cannot identify the functions of almost a third of the essential genes. This lack of information begs the question of what these genes are involved in. Do these proteins perform yet-unexplored roles in processes like transcription or translation, or are there still fundamental biological processes, essential to life, that have not been discovered?

    It makes sense that further minimization has growth-rate consequences, because it explains a natural-selection-enforced minimum on genome size. However, it does not explain why the M. mycoides genome retains non-essential genes. It seems intuitive that eliminating extraneous genes would benefit an organism by minimizing the energy required to replicate its genome. However, there must not be a very strong selection pressure, or non-essential genes would not exist in wildtype organisms. I would be interested to see an artificially-applied selection pressure or directed evolution to approach genome minimization.


1. Gibson, Daniel G., John I. Glass, Carole Lartigue, Vladimir N. Noskov, Ray-Yuan Chuang, Mikkel A. Algire, Gwynedd A. Benders, et al. “Creation of a Bacterial Cell Controlled         by a Chemically Synthesized Genome.” Science 329, no. 5987 (July 2, 2010): 52–56.
2. Hutchison, Clyde A., Ray-Yuan Chuang, Vladimir N. Noskov, Nacyra Assad-Garcia, Thomas J. Deerinck, Mark H. Ellisman, John Gill, et al. “Design and Synthesis of a Minimal Bacterial Genome.” Science 351, no. 6280 (March 25, 2016): aad6253.

Genomics Page
Biology Home Page
Hartlee's Home Page

Email Questions or Comments:

© Copyright 2017 Department of Biology, Davidson College, Davidson, NC 28035