What's Old is New

Genome Wide Ma­nip­u­la­tion of the Bac­te­r­ial Chro­mo­some in Vivo

by Michael G. Schmidt

Every now and then you come across an ar­ti­cle that makes you sit back and hear your­self say Wow! This is what hap­pened to me re­cently while prepar­ing to dis­cuss a pa­per on This Week in Mi­cro­bi­ol­ogy (TWiM), a bi-weekly pod­cast pro­duced by Vin­cent Racaniello and ASM, or as I think of it, a jour­nal club open to any who are cu­ri­ous about all as­pects of mi­cro­bi­ol­ogy. The pa­per by Church and col­leagues, pub­lished in Sci­ence, de­scribes the pre­cise ma­nip­u­la­tion of bac­te­r­ial chro­mo­so­mal se­quences on a genome-wide scale. Ti­tles mat­ter, and this one, Pre­cise Ma­nip­u­la­tion of Chro­mo­somes in Vivo En­ables Genome-Wide Codon Re­place­ment, spoke vol­umes. Be­fore even read­ing the ab­stract, it caused me to ask, Have these au­thors found the Holy Grail that syn­thetic bi­ol­o­gists have been search­ing for?I'll let you be the judge as I take you through the world of Mul­ti­plex Au­to­mated Genome En­gi­neer­ing (MAGE) and Con­juga­tive As­sem­bly Genome En­gi­neer­ing (CAGE).

Mul­ti­plex Au­to­mated Genome En­gi­neer­ing (MAGE) is the ap­pli­ca­tion of in vivo oligonu­cleotide mu­ta­ge­n­e­sis on a grand scale. In this ex­am­ple, a col­lec­tion of oli­gos con­tain­ing ten pre­cisely lo­cated TAG→TAA changes is added to a pop­u­la­tion of cells. The cells are al­lowed to repli­cate in the pres­ence of the oli­gos un­til the ma­jor­ity of the pop­u­la­tion con­tains all ten sub­sti­tu­tions. This process was re­peated for each of the 32 chro­mo­so­mal re­gions, and the 32 re­sul­tant strains were then con­sol­i­dated us­ing CAGE. Source

Here's what they did: they re­placed all of the 314 TAG am­ber stop codons, scat­tered across the 4.6 megabase genome of Es­cherichia coli with the syn­ony­mous TAA ochre codon. This means that they would then have strains with a free codon that can be re­pro­grammed to do other things. You might be ask­ing, What things? This re­ally is a sub­ject for an­other STC piece on cel­lu­lar en­gi­neer­ing writ­ten from the per­spec­tive of a syn­thetic bi­ol­o­gist. But, in keep­ing with the theme of What's Old Is New, one could en­vi­sion the al­ter­ation of a tRNA gene, sneak­ing it into the chro­mo­some via a lim­ited ver­sion of MAGE, such that its prod­uct will now rec­og­nize the TAG codon and sub­sti­tute a novel or unique amino acid into a pro­tein re­sult­ing in the desired/engineered phe­no­type. Mother na­ture has al­ready pro­vided the proof of con­cept: ex­tra­genic sup­pres­sion of ex­ist­ing am­ber mu­ta­tions.

This was done in vivo, not via the in vitro ap­proach used to en­gi­neer a bac­te­r­ial cell con­tain­ing a chem­i­cally syn­the­sized genome. The pro­ce­dures to re­place the nat­u­rally oc­cur­ring TAG codons were re­mark­ably well-planned in that the re­searchers elicited the help of Mother Na­ture. In E. coli, ter­mi­na­tion of pro­tein syn­the­sis is at one of the three stop codons, TAG, TAA, and TGA, and is me­di­ated by re­lease fac­tors 1 and 2 (RF1 and RF2). The TAG codon is rec­og­nized as a stop codon only by RF1, while the TAA stop codon is rec­og­nized by both RF1 and RF2. Con­se­quently, they could for­mu­late a straight­for­ward and eas­ily testable hy­poth­e­sis: Would re­plac­ing all of the TAG codons with TAA in one cell abol­ish the need for RF1, a testable phe­no­type? The an­swer has to wait un­til they fin­ish con­struct­ing such a strain (al­though they are close).

The el­e­gance of their ap­proach re­sides in their com­bin­ing two processes: MAGE, the use of mul­ti­ple syn­thetic oligonu­cleotides to mod­ify a num­ber of tar­gets in the genome in vivo (in this case spec­i­fied TAG codons within a sec­tor of the chro­mo­some), and CAGE, which re­lies on tricks for bac­te­r­ial con­ju­ga­tion (as I'll ex­plain be­low).

A Hitch­hik­ers Guide to Genome Wide En­gi­neer­ing: The Util­ity of MAGE & CAGE.in vivo oligonu­cleotide mu­ta­ge­n­e­sis, al­beit on a grand scale, us­ing the spe­cial hy­per-mu­ta­ble strain. Here MAGE is car­ried out on 32 sep­a­rate pop­u­la­tions of cells, one for each of the 32 chro­mo­so­mal re­gions de­fined. For each pop­u­la­tion, a mix­ture of oli­gos are added, DNA repli­ca­tion takes place, the sub­sti­tu­tions are in­cor­po­rated, and the changes are ver­ti­cally in­her­ited. This yields 32 re­coded strains, each car­ry­ing one chro­mo­so­mal re­gion in which all the TAG codons have been re­placed by TAA. A se­ries of con­ju­ga­tions be­tween pairs of re­coded strains (CAGE) con­sol­i­dates the TAG→TAA codons into fewer and fewer strains, ul­ti­mately down to just one strain. The strain is made de­fec­tive in Re­lease Fac­tor 1 (ΔprfA), thus can grow if the TAG codons have been sub­sti­tuted for TAA, whereas wild type strains can­not. Source

The ef­fi­ciency of MAGE is achieved through the suc­ces­sive in vivo in­tro­duc­tion of mul­ti­ple oligonu­cleotides with the de­sired mu­ta­tions. To do this, they used a strain of E. coli that was amenable to such ma­nip­u­la­tion. For ge­net­ics afi­ciona­dos, here are the de­tails. This strain car­ries a lambda (λ) prophage, termed Red, that has three mu­tant pro­teins: the lambda ex­onu­cle­ase, which pro­ces­sively di­gests the 5'-ends of ds­DNA; the beta pro­tein, which binds to sin­gle stranded DNA (think oli­gos) and pro­motes strand an­neal­ing; and the gamma pro­tein, which binds to the bac­te­r­ial RecBCD en­zyme and in­hibits its ac­tiv­i­ties (pre­vents the cell from cor­rectin­gits repli­ca­tion mis­takes). The strain also is de­fi­cient in mis­match re­pair (mutS). If you skipped these de­tails, just be cog­nizant that these mu­ta­tions re­sult in a hy­per-re­com­bi­na­tion state for E. coli, thus al­low­ing a quick way to achieve high fre­quency al­lelic re­place­ment (TAG→TAA).

As an aside, this be­ing a hy­per-mu­ta­ble strain, un­wanted sec­ondary mu­ta­tions may oc­cur at high fre­quency. To min­i­mize this like­li­hood, they used a di­vide-and-con­quer strat­egy to re­duce the num­ber of genome repli­ca­tions re­quired to reach their goal. Since the 314 codons are dis­trib­uted through­out the chro­mo­some, they di­vided the genome into 32 re­gions: 31 of them con­tain­ing ten TAG codons each and one con­tain­ing the re­main­ing four. The in­tended sub­sti­tu­tions within each re­gion were made in sep­a­rate strains in one MAGE cy­cle each. Thus, 'genome-wide' en­gi­neer­ing could be done on a small scale—and quickly. This bril­liant strat­egy re­duced the num­ber of cell di­vi­sions re­quired to typ­i­cally 6 to 8 per MAGE cy­cle. In this man­ner, they gen­er­ated 32 strains that col­lec­tively con­tained all 314 de­sired al­lelic sub­sti­tu­tions.

Hi­er­ar­chi­cal CAGE method­ol­ogy for con­trolled genome trans­fer. Two pairs of strains il­lus­trate the de­sign and method­ol­ogy of CAGE, with re­coded ge­nomic re­gions in red. Par­tially re­coded strains are split into con­ju­ga­tion pairs. Af­ter con­juga- tion, a spe­cific set of three si­mul­ta­ne­ous se­lec- tions is ap­plied to yield a re­com­bi­nant strain that con­tains the re­coded ge­nomic frag­ment from the donor strain while re­tain­ing the other re­coded re­gions in the re­cip­i­ent genome. All con­ju­ga­tion fac­tors are main­tained epi­so­ma­lly on F', so only a ~2 kb oriT se­quence must be in­serted onto the genome to gen­er­ate a highly con­trol­lable Hfr donor strain. This con­ju­ga­tion as­sem­bly-se­lec­tion strat­egy is im­ple­mented in stages to merge the genomes of 32 re­coded strains into a sin­gle strain. Source

The next chal­lenge was to com­bine those 32 strains into one that car­ries all these sub­sti­tu­tions. They al­most ac­com­plished this through a five-stage hi­er­ar­chi­cal as­sem­bly process, CAGE, which re­lies on con­ju­ga­tion by HFR strains to ef­fi­ciently trans­fer large seg­ments of the chro­mo­some from one bac­terium to an­other. The whole process is nicely sum­ma­rized in their fig­ure shown be­low.

More ge­netic de­tails. In HFR con­ju­ga­tion, the cis-act­ing ori­gin of trans­fer (oriT) within the chro­mo­some re­lies on trans-act­ing fac­tors pro­duced by the F fac­tor epi­some, also in­te­grated in the chro­mo­some, to move large amounts of ge­netic con­tent from the donor to the re­cip­i­ent. Use of coun­ter­s­e­lec­table mark­ers, one from the donor and one from the re­cip­i­ent, fa­cil­i­tates the easy iso­la­tion of the re­com­bi­nant strain. CAGE builds on the fun­da­men­tal prin­ci­ples re­quired for HFR con­ju­ga­tion but adds a few twists. CAGE donor strains were con­structed by first mov­ing the genes re­quired for the trans­fer fac­tors to a plas­mid lack­ing an ori­gin of trans­fer. The cis act­ing ge­netic el­e­ment oriT was fused to a se­lec­table marker, kanamycin (oriT-kan). This cas­sette was placed im­me­di­ately up­stream of the re-coded sec­tor of the chro­mo­some within each of the 16 donor strains at an ap­pro­pri­ate 'safe in­ser­tion re­gion' (in­ter­genic re­gions that have no cod­ing or reg­u­la­tory func­tion). Us­ing clever ge­netic tricks, they ob­tained donor strains that en­abled them to "start" and se­lect for con­ju­ga­tion from a de­sired lo­ca­tion.

Us­ing this strat­egy, they cre­ated 16 re­cip­i­ent strains with com­pat­i­ble se­lec­table mark­ers near the de­sired re­gion of the chro­mo­some. They car­ried out mat­ing, fol­lowed by pos­i­tive se­lec­tion and neg­a­tive se­lec­tion by plat­ing onto agar plates with a com­bi­na­tion of drugs and sug­ars to in­sure get­ting the right re­com­bi­nant. Through the sub­se­quent it­er­a­tive processes of mat­ing with ap­pro­pri­ate strains, they should ul­ti­mately end up with a sin­gle strain that will have each TAG codon re­placed by a TAA se­quence. This is nicely il­lus­trated in Fig­ure 4A from their pa­per, re­pro­duced here. (In ad­di­tion, pan­els C and D in the orig­i­nal fig­ure ex­plain the need for pos­i­tive and pos­i­tive-neg­a­tive se­lec­tions.)

Us­ing CAGE, they are al­most there. They have gone from 32 strains to 4, each of which con­tain 80 of the de­sired 314 TAG→TAA sub­sti­tu­tions. Not all that easy, be­cause they en­coun­tered is­sues with some of the strains and the se­lec­tion pro­ce­dures. Nev­er­the­less, com­bin­ing MAGE with CAGE has shown that, at least in E. coli, the bac­te­r­ial chro­mo­some is both ed­itable and evolv­able in a man­ner that will al­low many what if ex­per­i­ments to be done.

Will these clever tech­niques al­low us to take big leaps in the large-scale as­sem­bly of mod­i­fied genomes? Time and tal­ent will tell, but one thing is cer­tain: what's old will be new again. Who knew that bac­te­r­ial con­ju­ga­tion would help in­tro­duce and fa­cil­i­tate the era of syn­thetic bi­ol­ogy!

 

Ref­er­ence

Isaacs FJ, Carr PA, Wang HH, La­joie MJ, Ster­ling B, Kraal L, Tolo­nen AC, Gi­anoulis TA, Good­man DB, Rep­pas NB, Emig CJ, Bang D, Hwang SJ, Jew­ett MC, Ja­cob­son JM, Church GM. (2011). Pre­cise ma­nip­u­la­tion of chro­mo­somes in vivo en­ables genome-wide codon re­place­ment. Sci­ence, 333(6040), 348−353 PMID. 21764749

 

Michael G. Schmidt is Pro­fes­sor and Vice Chair­man of Mi­cro­bi­ol­ogy and Im­munol­ogy at the Med­ical Uni­ver­sity of South Car­olina.