And The Win­ner Is…

by Elio

The trans­posases! The con­test, in case you won­der, was for the most abun­dant set of genes in the known uni­verse, or at least in the ge­nomic data banks avail­able on Earth. Aziz, Bre­it­bart, and Ed­wards ex­am­ined some 10 mil­lion genes from 2137 se­quenced genomes (47 ar­chaeal, 725 bac­te­r­ial, 29 eu­kary­otic and 1336 vi­ral) plus 187 metagenomes, in search of the gene set that is both most abun­dant and most ubiq­ui­tous. (Ubiq­ui­tous refers to genes that carry es­sen­tial func­tions and are thus needed by every genome. A gene that is highly abun­dant in some ecosys­tems but not ubiq­ui­tous is likely to play an im­por­tant habi­tat-spe­cific role.) The win­ners, by far, were the genes for trans­posases. One could squawk that the genomes se­lected for se­quenc­ing to date re­flect a bias to­wards "germs" and other mi­crobes of hu­man in­ter­est, and that so far the source en­vi­ron­ments for metagenomes rep­re­sent only a small sam­ple of what's out there. But given the off-scale pre­pon­der­ance of the trans­posase genes, there is lit­tle to make one ques­tion their dom­i­nance.

Abun­dance of dif­fer­ent func­tional roles in 2137 genomes plot­ted against the ubiq­uity of these func­tional roles (de­fined as the num­ber of genomes in which the func­tional role is rep­re­sented at least once). Sources: here and here

Not every­body seems to have trans­posases; only 1/3 of the viruses in this study have them. But those or­gan­isms that do tend to pos­sess mul­ti­ple copies — close to 40 on av­er­age. If you take a set of 2000 ran­domly sam­pled genes (the num­ber found in a typ­i­cal bac­te­r­ial genome), 22 will en­code trans­posases. They are also more com­mon in free-liv­ing bac­te­ria than in ob­lig­ate pathogens or en­dosym­bionts. The cham­pion trans­posase-car­rier? Pos­si­bly the ma­rine cyanobac­terium Cro­cosphaera wat­sonii, which has more than 400 copies.

So, why trans­posases? The au­thors point to their gen­eral use­ful­ness to both viruses and cel­lu­lar or­gan­isms in pro­mot­ing the spread of mo­bile el­e­ments, thus lead­ing to di­ver­si­fi­ca­tion. "Self­ish genes" in­deed! As the au­thors write, By se­cur­ing their own repli­ca­tion and dis­sem­i­na­tion, trans­posases guar­an­tee to thrive so long as nu­cleic acid-based life forms ex­ist. This rather im­plies that they might imag­ine non-nu­cleic acid based life forms ex­ist some­where. Who says that bioin­for­mati­cists are just com­pu­ta­tional nerds?
 

Ref­er­ence

Aziz RK, Bre­it­bart M, Ed­wards RA. (2010). Trans­posases are the most abun­dant, most ubiq­ui­tous genes in na­ture. Nu­cleic Acids Res., 38((13):4207−4217. PMID 20215432