Fat­ten­ing Up Mi­cro­bial Ge­o­log­i­cal Bio­mark­ers

by Paula We­lander

First Evolved! Last Ex­tinct! This prokary­otic pride motto was coined by my un­der­grad­u­ate ad­vi­sor (and good friend) Prof. Mark O. Mar­tin. As a mi­cro­bi­ol­o­gist, I love this motto for many rea­sons, but es­pe­cially be­cause it al­ludes to one of the un­der­ly­ing prin­ci­ples of my cur­rent re­search. Mi­crobes were in­deed the first to evolve and the meta­bolic in­ven­tions of an­cient mi­crobes greatly in­fluenced the an­cient Earth's en­vi­ron­ment and the evo­lution of life. The in­ter­ac­tion be­tween the Earth and mi­crobes has been recorded in sed­i­men­tary rocks that are bil­lions of years old.

Fig­ure 1A. The bio­marker prin­ci­ple in­volves the bur­ial and preser­va­tion of bac­te­r­ial lipids such as hopanoids in the sed­i­men­tary rock over bil­lions of years. Dia­genesis de­scribes the chem­i­cal, physi­cal, or bi­o­log­i­cal changes un­der­gone by sed­i­ment af­ter its ini­tial de­position. Bac­teria present in the wa­ter col­umn die and are de­posited and buried in the sed­i­ment where they un­dergo degra­da­tion. Hopan­oid mol­e­cules lose the ma­jor­ity of their func­tional groups but the ba­sic hopane struc­ture is pre­served over bil­lions of years. Geo­chemists are then able to ex­tract and de­tect these lipids in an­cient sed­i­ments. Credit: Paula We­lander

Ac­cess­ing this mi­cro­bial record con­tin­ues to be one of the chal­lenges of ge­omi­cro­bi­ol­ogy. One pow­er­ful stra­tegy for un­der­stand­ing the mi­cro­bial sig­na­tures in the rock record is the use of "mol­e­c­u­lar fos­sils" or biomark­ers, or­ganic com­pounds that are pro­duced by se­lect groups of mi­croor­gan­isms and, amaz­ingly, are pre­served in both mod­ern and an­cient sed­i­ments. Geo­chemists are able to ex­tract these mol­e­cules from very old rocks and, based on their dis­tri­b­u­tion in mod­ern or­gan­isms, link spe­cific groups of bac­te­ria to an­cient en­vi­ron­ments (Fig. 1A+B).

As Tanja Bosak de­scribed in her ear­lier con­tri­bu­tion to this blog, both eu­kary­otic and bac­te­r­ial lipids are pre­served in old rocks, thus both have the po­ten­tial to be "mol­e­c­u­lar fos­sils." One ex­ample of such is the use of 2‑methylhopanoids as bio­mark­ers for cyanobac­te­ria and oxy­genic pho­to­syn­the­sis. Hopanoids are pen­ta­cyclic triter­penoid lipids pro­duced mainly by bac­te­ria and that are struc­turally sim­i­lar to eu­kary­otic sterols (Fig. 2).

Fig­ure 1B. (Leg­end see Fig. 1A)

As the name im­plies, 2‑methylhopanoids are methy­lated at the C‑2 po­si­tion. Al­though seem­ingly a mi­nor modifi­cation, this chem­i­cal al­ter­ation en­dures in an­cient rocks. Stud­ies with mod­ern bac­te­ria showed that cyanobac­te­ria are the pre­dom­i­nant pro­duc­ers of 2‑methylhopanoids. On this ba­sis, 2‑methylhopanoids have been used as bio­mark­ers not only for cyanobac­te­ria but also for oxy­genic pho­to­syn­the­sis . The pres­ence of these mol­e­cules in rocks pro­vided ev­i­dence for the an­tiq­uity of oxy­genic pho­to­syn­the­sis and for the role of cyanobac­te­ria in the an­cient past (click here, here, and here).

The link be­tween cyanobac­te­ria and 2‑methylhopanoids was brought into ques­tion a few years ago when it was dis­cov­ered that the anoxy­genic pho­totroph Rhodopseudomonas palus­tris was ca­pa­ble of pro­duc­ing methy­lated hopanoids un­der anaer­o­bic con­di­tions and in quan­ti­ties sim­i­lar to those made by the cyanobac­te­ria. This find­ing raised doubts about the pre­vi­ous in­ter­pre­ta­tion of the ori­gin of 2‑methylhopanoids in the rock record. It is here where we, as mol­e­c­u­lar microbio­logists and mi­cro­bial phys­i­ol­o­gists, could make a con­tri­bu­tion to the bio­marker field by "fat­ten­ing up" the slim amount of phy­lo­ge­netic and phys­i­o­log­i­cal data avail­able for link­ing hopanoids to spe­cific bac­te­ria and their me­tab­o­lism.

Fig­ure 2. Struc­ture of a 2‑methylhopan­oid. Credit: Paula We­lander

Cur­rently, our un­der­stand­ing of the tax­o­nomic classi­fi­cation of 2‑methylhopanoid pro­duc­ers is based on lipid analy­sis of cul­tured bac­te­r­ial strains. We rea­soned that a more con­struc­tive ap­proach would be to iden­tify a genet­ic fac­tor re­quired for the pro­duc­tion of 2‑methylhopan­oids that could be used to sur­vey the vast amount of avail­able ge­nomic and metage­nomic data. Uti­liz­ing com­parative ge­nomics and ge­netic dele­tion analy­sis, we un­covered the vi­t­a­min B‑12 bind­ing rad­i­cal SAM pro­tein (HpnP) in R. palus­tris re­spon­si­ble for the methy­la­tion at the C‑2 po­si­tion. Only 30 bac­te­r­ial strains (out of ap­prox­i­mately 200 hopanoid pro­duc­ers tested) have a copy of the HpnP methy­lase (Fig. 3). These fall into three phy­lo­ge­netic groups – the cyanobac­te­ria, the Rhi­zo­biales (an or­der of the α‑proteobacteria that con­tains R. palus­tris), and the aci­dobac­te­ria. No­tably, not all cyanobac­te­ria con­tain HpnP. We found no com­mon thread among the bac­te­ria with the methy­lase (e.g., they aren't all pho­totrophs nor are all found in one type of ecosys­tem). Clearly just under­standing the phy­lo­ge­netic dis­tri­b­u­tion of 2‑methylhopanoid pro­duc­ers is not suf­fi­cient to prop­erly "read" these bio­mark­ers in the rock record. Rather, it would be more use­ful to iden­tify the cel­lu­lar func­tion as well as the en­vi­ron­men­tal fac­tors that in­duce the ex­pres­sion of these com­pounds.

Fig­ure 3. Un­rooted tree show­ing the ma­jor clus­ters of HpnP from dif­fer­ent bac­te­r­ial groups. Source

What to do? We turned once again to the anoxy­genic pho­totroph R. palus­tris that pro­duces up to six dif­fer­ent mod­i­fied hopanoids. We started out by char­ac­ter­iz­ing a mu­tant un­able to pro­duce hopanoids methy­lated at the C‑2 po­si­tion, but we found no as­so­ci­ated phe­no­type. It was clear that first we needed to step back and under­stand the func­tion of hopanoids in gen­eral. We con­structed an R. palus­tris mu­tant that no longer pro­duced any hopanoids by delet­ing squa­lene hopene cy­clase, an en­zyme re­quired for the ini­tial cy­cliza­tion of squa­lene to form the ba­sic hopanoid struc­ture. This mu­tant has simi­lar growth char­ac­ter­is­tics to the wild type strain un­der all stan­dard growth con­di­tions in­clud­ing pho­to­syn­thetic growth, un­der­scor­ing that hopanoids do not play a role in pho­to­syn­the­sis – and high­light­ing again that hopanoids are not good bio­mark­ers for any type of pho­to­syn­the­sis! How­ever, the hopanoid dele­tion mu­tant showed a growth de­fect un­der en­vi­ron­men­tal stress con­di­tions such as high pH or high tem­per­a­ture. Its outer mem­brane be­came more per­me­able, sug­gest­ing that en­hanced per­me­abil­ity low­ers re­sis­tance to en­vi­ron­men­tal stress. In­ter­est­ingly, the mu­tant de­fi­cient in only 2‑methylhopanoid syn­the­sis grows nor­mally un­der stress and its outer mem­brane is as im­per­me­able as the wild type. This means that the methy­la­tion at the C‑2 po­si­tion has a unique and dis­tinct pur­pose, sep­a­rate from main­tain­ing outer mem­brane in­tegrity. Seek­ing some clues as to the func­tion of these spe­cific mol­e­cules, we have started work on the biochem­istry of the methy­lase pro­tein and are in­ves­ti­gat­ing the lo­cal­iza­tion of the 2‑methylhopan­oids, their pos­si­ble reg­u­la­tors, and the en­vi­ron­men­tal fac­tors that might stim­u­late their produc­tion.

Al­though we have made some progress in un­der­stand­ing the func­tion of hopanoid mol­e­cules in bac­te­r­ial cells, we feel we have just be­gun to scratch the sur­face. We hope that fur­ther investiga­tions into these po­ten­tially in­ter­est­ing lipids will re­veal novel func­tions that will not only in­form the proper in­ter­pre­ta­tion of hopanoids found in old rocks but will also spark the in­ter­est of microbio­logists and cell bi­ol­o­gists. Hopanoids have been buried for bil­lions of years and it is time for their mys­ter­ies to be un­earthed.

Ref­er­ence

We­lander PV, Cole­man ML, Ses­sions AL, Sum­mons RE, & New­man DK (2010). Iden­ti­fi­ca­tion of a methy­lase re­quired for 2‑methylhopanoid pro­duc­tion and im­pli­ca­tions for the in­ter­pre­ta­tion of sed­i­men­tary hopanes. Proc Natl Acad Sci U.S.A, 107 (19), 8537−42. PMID 20421508

Paula is a post­doc­toral fel­low co-ad­vised by Prof. Roger Sum­mons in the Depart­ment of Earth, At­mos­pheric and Plan­e­tary Sci­ences at MIT, and Prof. Di­anne New­man in the De­part­ment of Bi­ol­ogy at Cal­tech.