Col­ibactin (cont.)

Note­wor­thy

In do­ing our read­ing for last week's post on "the two faces of col­ibactin," we dis­cov­ered many fas­ci­nat­ing ar­ti­cles in­ves­ti­gat­ing this geno­toxic mol­e­cule. One pa­per, how­ever, struck us so much that we felt it de­served its own "Note­wor­thy" post. Here it is. 

We'll start by say­ing why we liked it so much. Most of the col­ibactin-re­lated re­search seems to come to the sub­ject al­ready with col­ibactin in mind. In con­trast, in this pa­per, the work ran into col­ibactin as the re­sult of an un­bi­ased ge­netic screen. Hav­ing a soft spot for ge­net­ics, our lik­ing it so much should come as no sur­prise.

The ti­tle of the pa­per by Kait­lyn Bar­rack and col­leagues (se­nior au­thor George O'Toole, a long-time friend of STC) nicely sums up the key find­ings: "A geno­toxin as­so­ci­ated with col­orec­tal can­cer linked to gut dys­bio­sis in chil­dren with cys­tic fi­bro­sis." 

Briefly, this is how they got to this con­clu­sion. Want­ing to bet­ter un­der­stand the gut dys­bio­sis ob­served in cys­tic fi­bro­sis (CF) pa­tients – char­ac­ter­ized by an in­crease in pro­teobac­te­ria dri­ven largely by E. coli and con­comi­tant de­creases in anaer­obes such as Bac­teroides spp. – the au­thors set up an in vitro sys­tem to mimic the key vari­ables that make the CF gut dif­fer­ent. Us­ing this sys­tem, they fo­cused on the in­ter­ac­tions be­tween strains of E. coli and Bac­teroides vul­ga­tus in co-cul­ture. The strik­ing find­ing was that, when they added glyc­erol – a sur­ro­gate for in­creased fat – to the growth medium, B. vul­ga­tus was wiped out from co-cul­tures. To find out what was be­hind this phe­nom­e­non, they screened a li­brary of trans­po­son-gen­er­ated mu­tants of E. coli for loss of the abil­ity to elim­i­nate B. vul­ga­tus when in co-cul­ture. Strik­ingly, mu­ta­tions lead­ing to this phe­no­type mapped in the col­ibactin biosyn­thetic gene clus­ter! 

Trans­po­son mu­ta­ge­n­e­sis iden­ti­fied col­ibactin as nec­es­sary in B. vul­ga­tus vi­a­bil­ity re­duc­tion by E. coli in the pres­ence of glyc­erol. (A) Schematic of the pks/clb ge­netic is­land, span­ning 54 kb and con­tain­ing genes clbA–clbS. Black lol­lipops de­note trans­po­son in­ser­tions in the 5′–3′ di­rec­tion, in­verted white and black check­ered lol­lipops de­note trans­po­son in­ser­tions in the 3′–5′ di­rec­tion. There were 17 in­de­pen­dent an­ti­sense hits to clbK, as rep­re­sented by the high­lighted "17." Two in­de­pen­dent can­di­dates have trans­po­son in­ser­tions de­tected in both clbJ and clbK, de­noted with the bracket. (B and C) B. vul­ga­tus clin­i­cal iso­lates (n = 5) grown in mono­cul­ture (first bar, pur­ple) and co-cul­ture with E. coli 139H WT (sec­ond bar, navy) and rep­re­sen­ta­tive E. coli clb::TnM mu­tants for 72 h in (B) MiPro pH 7 and © MiPro pH 7 + 1% glyc­erol. Sta­tis­ti­cal analy­sis per­formed us­ing one-way ANOVA with Dunnett's mul­ti­ple com­par­isons be­tween each co-cul­ture con­di­tion with the mono­cul­ture con­trol (*, P < 0.05; **, P < 0.01; ***, P < 0.005; ****, P < 0.001). Er­ror bars rep­re­sent mean +/− SEM. Source

The pa­per, of course, con­tains many more re­sults pro­vid­ing fur­ther ev­i­dence that col­ibactin me­di­ates the ob­served mi­cro­bial com­pe­ti­tion. And a lot more on the co-cul­ture ex­per­i­ments them­selves. Do read it in its en­tirety. We are con­fi­dent that, like us, you'll be thrilled by "the awe­some power of ge­net­ics."