An Icon­o­clas­tic En­dosym­biont

by Merry

The en­dosym­bionts of in­sects pro­vide an un­end­ing pa­rade of nov­elty. They seem to knock on our door, one af­ter the other, each pur­vey­ing some new twist of mi­crobial evo­lu­tion. They break the rules, re­model the ge­netic code, and blur the line be­tween or­gan­ism and or­ganelle. Here is one more that seems to par­tic­u­larly de­light in de­fy­ing con­ven­tion.

The ci­cada Dicero­procta semi­cincta; (male). Size: 3/4 inch. Source

The av­er­age GC con­tent of bac­te­r­ial genomes varies wi­dely, from 17% to 75%. More­over, the vari­a­tion is not ran­dom, i.e., the per­cent GC cor­re­lates roughly with ge­nome size. The larger the genome, the higher the GC con­tent. This shows up most dra­mat­i­cally at the low end of the scale, among the in­sect en­dosym­bionts with greatly re­duced genomes. They ex­hibit the low­est GC per­cent­age. One can't help but won­der why. The argu­ment put for­ward to ex­plain this has gone like so. All DNA is sub­ject to con­tin­ual al­ter­ation. Two of the com­mon chem­i­cal changes are the spon­ta­neous deamina­tion of cy­to­sine and the ox­i­da­tion of guano­sine by re­active oxy­gen species. Both of these re­ac­tions, if not re­paired, change a GC to an AT pair. En­dosym­bionts with re­duced genomes have fewer DNA re­pair genes. Ergo, over time their DNA would ac­cu­mu­late un­re­paired GC to AT shifts, thus low­er­ing the av­er­age GC% of their genomes.

This made a tidy story.

When Car­sonella rud­dii came along, it broke the old records. Weigh­ing in with a genome of only 160 kbp and but 182 pro­tein-cod­ing genes, it hardly seemed to have enough genes to sus­tain its own life. Its 16.5% GC con­tent was also the low­est yet. Thus it still fit the pat­tern (see fig­ure be­low). Now here comes an­other in­sect en­dosym­biont, this one from the Ari­zona ci­cada Dicero­procta se­micincta and given the name Can­di­da­tus Hodgkinia ci­cadi­cola. Its genome is even smaller, a mere 143,795 bps. Its GC con­tent? 58.4%. That puts Hodgkinia way out in left field on the graph.

Re­la­tion­ship be­tween ge­no­me size and GC con­tent for sequ­enced Bac­te­r­ial and Ar­chaeal genomes ≤10 Mb. Red cir­cles = ob­lig­ate in­sect en­do­symbionts; dark blue = ob­lig­ate α‑Pro­teobacteria en­dosym­bionts; pur­ple = red + blue = Hodgkinia; light blue = other α‑Proteobacteria; gray = other Bac­te­ria and Ar­chaea. Source

As if that weren't enough, Hodgkinia has an­other odd­ity. It uses a mod­i­fied form of the "uni­ver­sal" ge­netic code, a ver­sion that has turned up be­fore in some mi­to­chon­dr­ial lin­eages and My­coplasma, all of which ex­hibit genome re­duc­tion and low GC con­tent too. Here, UGA no longer func­tions as a stop codon but in­stead codes for trypto­phan. This had been thought to be a con­se­quence of the drop in GC% in these small genomes. When the G in a UGA codon is changed to an A, pro­tein trans­la­tion isn't af­fected be­cause the new codon, UAA, is also a stop co­don. In the course of time, more UGAs would change to UAAs un­til even­tu­ally, it was imag­ined, no UGA codons would re­main. Then UGA was free to later on be dele­gated a new func­tion, i.e., en­cod­ing tryp­to­phan (other­wise en­coded by UGG).

In­stead, Hodgkinia tells us that it is genome re­duc­tion, not low GC%, that is dri­ving this shift in its ge­netic code. One of the genes lost by Hodgkinia en­codes the trans­la­tional Re­lease Fac­tor TF2 that rec­og­nizes UGA as a stop codon and ter­mi­nates the pro­tein at that point. Hodgkinia gets by per­fectly fine with­out it be­cause it still en­codes Re­lease Fac­tor TF1, the one that rec­og­nizes the other two stop codons. My­coplas­mas that use the same mod­i­fied ge­netic code also lack TF2.

So, now back to the orig­i­nal ques­tion: why the re­duced GC% for all en­dosym­bionts so far ex­cept Hodgkinia? The an­swer seems not to be sim­ply the loss of DNA re­pair en­zymes. Other ev­i­dence has sug­gested that GC% is some­how re­lated to vari­a­tions in a par­tic­u­lar sub­unit of DNA polyme­rase. Since Hodgkinia has but two genes en­cod­ing DNA poly­merase sub­units, it might serve as a sim­pli­fied model sys­tem for in­ves­ti­gat­ing this. The re­searchers will be ques­tion­ing Hodgkinia on this mat­ter. Likely more pa­pers will be forth­com­ing.

Ref­er­ence

Mc­Cutcheon JP, Mc­Don­ald BR, & Moran NA (2009). Ori­gin of an al­ter­na­tive ge­netic code in the ex­tremely small and GC-rich genome of a bac­te­r­ial sym­biont. PLoS ge­net­ics, 5 (7). PMID 19609354