First bacterial genome created entirely with a computer April 03, 2019, 12:22

All the genome se­quences of or­gan­isms known through­out the world are stored in a data­base be­long­ing to the Na­tional Cen­ter for Biotech­nol­ogy In­for­ma­tion in the United States. As of to­day, the data­base has an ad­di­tional en­try: Caulobac­ter ethen­sis-2.0. It is the world’s first fully com­puter-gen­er­ated genome of a liv­ing or­gan­ism, de­vel­oped by sci­en­tists at ETH Zurich. How­ever, it must be em­pha­sised that al­though the genome for C. ethen­sis-2.0 was phys­i­cally pro­duced in the form of a very large DNA mol­e­cule, a cor­re­spond­ing or­gan­ism does not yet ex­ist.

C. ethen­sis-2.0 is based on the genome of a well-stud­ied and harm­less fresh­wa­ter bac­terium, Caulobac­ter cres­cen­tus, which is a nat­u­rally oc­cur­ring bac­terium found in spring wa­ter, rivers and lakes around the globe. It does not cause any dis­eases. C. cres­cen­tus is also a model or­gan­ism com­monly used in re­search lab­o­ra­to­ries to study the life of bac­te­ria. The genome of this bac­terium con­tains 4,000 genes. Sci­en­tists pre­vi­ously demon­strated that only about 680 of these genes are cru­cial to the sur­vival of the species in the lab. Bac­te­ria with this min­i­mal genome are vi­able un­der lab­o­ra­tory con­di­tions.

Ra­tio­nal­is­ing the pro­duc­tion process

Beat Chris­ten, Pro­fes­sor of Ex­per­i­men­tal Sys­tems Bi­ol­ogy at ETH Zurich, and his brother, Matthias Chris­ten, a chemist at ETH Zurich, took the min­i­mal genome of C. cres­cen­tus as a start­ing point. They set out to chem­i­cally syn­the­sise this genome from scratch, as a con­tin­u­ous ring-shaped chro­mo­some. Such a task was pre­vi­ously seen as a true tour de force: The chem­i­cally syn­the­sised bac­te­r­ial genome pre­sented eleven years ago by the Amer­i­can ge­net­ics pi­o­neer Craig Ven­ter was the re­sult of ten years of work by 20 sci­en­tists, ac­cord­ing to me­dia re­ports. The cost of the pro­ject is said to have to­talled 40 mil­lion dol­lars.

While Ven­ter’s team made an ex­act copy of a nat­ural genome, the re­searchers at ETH Zurich rad­i­cally al­tered their genome us­ing a com­puter al­go­rithm. Their mo­ti­va­tion was twofold: one, to make it much eas­ier to pro­duce genomes, and two, to ad­dress fun­da­men­tal ques­tions of bi­ol­ogy.

To cre­ate a DNA mol­e­cule as large as a bac­te­r­ial genome, sci­en­tists must pro­ceed step by step. In the case of the Caulobac­ter genome, the sci­en­tists at ETH Zurich syn­the­sised 236 genome seg­ments, which they sub­se­quently pieced to­gether. “The syn­the­sis of these seg­ments is not al­ways easy,” ex­plains Matthias Chris­ten. “DNA mol­e­cules not only pos­sess the abil­ity to stick to other DNA mol­e­cules, but de­pend­ing on the se­quence, they can also twist them­selves into loops and knots, which can ham­per the pro­duc­tion process or ren­der man­u­fac­tur­ing im­pos­si­ble,” ex­plains Matthias Chris­ten.

Sim­pli­fied DNA se­quences

To syn­the­sise the genome seg­ments in the sim­plest pos­si­ble way, and then piece to­gether all seg­ments in the most stream­lined man­ner, the sci­en­tists rad­i­cally sim­pli­fied the genome se­quence with­out mod­i­fy­ing the ac­tual ge­netic in­for­ma­tion (at the pro­tein level). There is am­ple lat­i­tude for the sim­pli­fi­ca­tion of genomes, be­cause bi­ol­ogy has built-in re­dun­dan­cies for stor­ing ge­netic in­for­ma­tion. For ex­am­ple, for many pro­tein com­po­nents (amino acids), there are two, four or even more pos­si­bil­i­ties to write their in­for­ma­tion into DNA.

The al­go­rithm de­vel­oped by the sci­en­tists at ETH Zurich makes op­ti­mal use of this re­dun­dancy of the ge­netic code. Us­ing this al­go­rithm, the re­searchers com­puted the ideal DNA se­quence for the syn­the­sis and con­struc­tion of the genome, which they ul­ti­mately utilised for their work.

As a re­sult, the sci­en­tists seeded many small mod­i­fi­ca­tions into the min­i­mal genome, which in their en­tirety are, how­ever, im­pres­sive: more than a sixth of all of the 800,000 DNA let­ters in the ar­ti­fi­cial genome were re­placed, com­pared to the “nat­ural” min­i­mal genome. “Through our al­go­rithm, we have com­pletely rewrit­ten our genome into a new se­quence of DNA let­ters that no longer re­sem­bles the orig­i­nal se­quence. How­ever, the bi­o­log­i­cal func­tion at the pro­tein level re­mains the same,” says Beat Chris­ten.

Lit­mus test for ge­net­ics

The rewrit­ten genome is also in­ter­est­ing from a bi­o­log­i­cal per­spec­tive. “Our method is a lit­mus test to see whether we bi­ol­o­gists have cor­rectly un­der­stood ge­net­ics, and it al­lows us to high­light pos­si­ble gaps in our knowl­edge,” ex­plains Beat Chris­ten. Nat­u­rally, the rewrit­ten genome can con­tain only in­for­ma­tion that the re­searchers have ac­tu­ally un­der­stood. Pos­si­ble “hid­den” ad­di­tional in­for­ma­tion that is lo­cated in the DNA se­quence, and has not yet been un­der­stood by sci­en­tists, would have been lost in the process of cre­at­ing the new code.

For re­search pur­poses, the sci­en­tists pro­duced strains of bac­te­ria that con­tained both the nat­u­rally oc­cur­ring Caulobac­ter genome and also seg­ments of the new ar­ti­fi­cial genome. By turn­ing off cer­tain nat­ural genes in these bac­te­ria, the re­searchers were able to test the func­tions of the ar­ti­fi­cial genes. They tested each one of the ar­ti­fi­cial genes in a mul­ti­step process.

In these ex­per­i­ments, the re­searchers found out that only about 580 of the 680 ar­ti­fi­cial genes were func­tional. “With the knowl­edge we have gained, it will, how­ever, be pos­si­ble for us to im­prove our al­go­rithm and de­velop a fully func­tional genome ver­sion 3.0,” says Beat Chris­ten.

Enor­mous po­ten­tial for biotech­nol­ogy

“Even though the cur­rent ver­sion of the genome is not yet per­fect, our work nev­er­the­less shows that bi­o­log­i­cal sys­tems are con­structed in such a sim­ple man­ner that in the fu­ture, we will be able to work out the de­sign spec­i­fi­ca­tions on the com­puter ac­cord­ing to our goals, and then build them,” says Matthias Chris­ten. And this can be ac­com­plished in a com­par­a­tively straight­for­ward way, as Beat Chris­ten em­pha­sises: “What took ten years with Craig Ven­ter’s ap­proach, our small group achieved with our new tech­nol­ogy within the time frame of one year with man­u­fac­tur­ing costs of 120,000 Swiss francs.”

“We be­lieve that it will also soon be pos­si­ble to pro­duce func­tional bac­te­r­ial cells with such a genome,” says Beat Chris­ten. Such a de­vel­op­ment would hold great po­ten­tial. Among the pos­si­ble fu­ture ap­pli­ca­tions are syn­thetic mi­croor­gan­isms that could be utilised in biotech­nol­ogy for the pro­duc­tion of com­plex phar­ma­ceu­ti­cally ac­tive mol­e­cules or vi­t­a­mins, for ex­am­ple. The tech­nol­ogy can be em­ployed uni­ver­sally for all mi­croor­gan­isms, not just Caulobac­ter. An­other pos­si­bil­ity would be the pro­duc­tion of DNA vac­cines.

“As promis­ing as the re­search re­sults and pos­si­ble ap­pli­ca­tions may be, they de­mand a pro­found dis­cus­sion in so­ci­ety about the pur­poses for which this tech­nol­ogy can be used and, at the same time, about how abuses can be pre­vented,” says Beat Chris­ten. It is still not clear when the first bac­terium with an ar­ti­fi­cial genome will be pro­duced – but it is now clear that it can and will be de­vel­oped. “We must use the time we have for in­ten­sive dis­cus­sions among sci­en­tists, and also in so­ci­ety as a whole. We stand ready to con­tribute to that dis­cus­sion, with all of the know-how we pos­sess.”