Non-Pho­to­syn­thetic Bac­te­ria Also Have a Cir­ca­dian Rhythm

by Elio

The very first en­try in this blog, in Decem­ber 2006, was about the amaz­ing fact that some photo­synthetic cyano­bacteria have a cir­ca­dian rhythm. I en­titled it "It Don't Mean a Thing If It Ain't Got That Swing." Now, a group of Ger­man invest­igators re­port this is also true for an ordin­ary non-pho­to­syn­thetic lab rat, Bacil­lus sub­tilis.

I was go­ing to intro­duce the sub­ject anew, when it oc­curred to me that my an­cient text was still quite service­able now, So, by your leave, I will copy parts of it here:

It's not that many years ago that most peo­ple, me in­cluded, thought that bac­te­ria lacked the abil­ity to sense chem­ical gra­di­ents, commun­icate with one an­other, make fancy multi­cellular struc­tures (with a few no­table excep­tions), and do some of the other things attrib­uted to "higher" or­gan­isms. How things have changed! The num­ber of com­plex processes that bac­te­ria can do grows by the month, lead­ing me to won­der at times if eu­karyotes in­vented any­thing!…

To the list of phe­nom­ena that made this jump we can add cir­ca­dian rhythms... It's been known since about 1985 that such rhythms ex­ist in some cyanobac­te­ria. In these or­ganisms, many processes ex­hibit daily fluc­tuations, in­clud­ing global gene ex­pression, oxy­gen up­take, and carbo­hydrate synth­esis. One gratify­ing rea­son for diurnal/nocturnal period­icity is that ni­tro­ge­nase (the en­zyme respon­sible for ni­tro­gen fix­a­tion) is highly sen­si­tive to the oxy­gen gen­er­ated dur­ing photo­synthesis. Un­like or­gan­isms, e.g. An­abaena, where photo­synthesis and ni­tro­gen fix­a­tion are car­ried out in dif­fer­ent cells, uni­cellular cyano­bacteria such as Syne­chococ­cus do both in the same cell. Here, the two processes are not sep­a­rated in space but by time. Dur­ing the day, they photo­synthesize, at night they fix ni­tro­gen.

Sur­prises abound. The genes in­volved in photo­synthesis and ni­tro­gen fix­a­tion are far from be­ing the only ones show­ing cir­ca­dian be­hav­ior. Most of the Syne­chococ­cus' pro­mot­ers show such a rhythm, a global be­hav­ior that may be dic­tated by changes in DNA compac­tion (LINK). Note also that the cell di­vi­sion cy­cle in Syne­chococ­cus varies with light in­ten­sity and can be as short as about 6 hours, so there is noth­ing cir­ca­dian about it (cir­ca­dian, for those that need re­mind­ing, comes from the Latin for circa for about, and dies for day). Period­icity is not de­pen­dent on ex­ter­nal stim­uli such as light or dark pe­ri­ods, but that's im­plied in the de­f­i­n­i­tion of cir­ca­dian rhythms. In the words of Golden et al, "Un­like a metro­nome that merely taps out the beat, the cir­ca­dian clock that times the events in a cell knows the score by heart." So, these or­gan­isms got that swing!

It gets bet­ter. The whole of the cir­ca­dian rhythms has been attrib­uted to a lowly num­ber of pro­teins, four. SasA, a sen­sory ki­nase is the pri­mary trans­ducer, and three oth­ers are cen­tral tim­ing mech­anism. They are called Kai A, B, C ("kai" means cy­cle in Japan­ese). There is, of course more to it and for the whole cell phy­siology to be­have cyclic­ally, many other pro­teins are in­volved as well.

Just as you thought that it couldn't get any more interest­ing, there comes a star­tling discov­ery: the three Kai pro­teins act rhythm­ically in vitro. Mix the three, add ATP, and you see phos­phory­lation of KaiC un­der­go­ing a 21 hour cy­cle at 30° (LINK). What is star­tling is that Kai pro­teins from mu­tants with al­tered period­icity show the same period­icity in vitro as the cells from which they are de­rived. And these are not just flimsy oscill­ations. They are to­tally con­vinc­ing and main­tained for at least four ro­bust cy­cles. This find­ing is in keep­ing with the fact that the pro­teins do not have to be made anew and that the cir­ca­dian rhythm does not de­pend on transcrip­tion and trans­lation.

Now the big ques­tion is, what kind of bio­chemistry op­er­ates over such a long time? The Kai pro­teins in­ter­act with one an­other and one makes hexa­m­ers from monomers that have two po­ten­tial phos­phory­lation sites. This may lead to a com­plex choreo­graphy of auto­phosphory­lation and de­phosphory­lation, which may lead to such un­usual ki­net­ics. I find es­pe­cially puz­zling that the same ki­net­ics is seen in vivo and in vitro, de­spite dif­fer­ences in con­centra­tion of the rel­e­vant pro­teins and in chem­ical and phys­ical mi­lieu. There may be ex­cit­ing ways of explain­ing all this, but for now, to me it seems like a mys­tery.

Fig. 1 En­train­ment by light and a free-run­ning rhythm in B. sub­tilis. Bi­o­lu­mi­nes­cence of P_ytvA::lux (A) and P_kinC::lux (D)un­der 5 days of en­train­ment with cy­cles of dark­ness and blue light (12-hour D/12-hour L) and af­ter re­lease to con­stant dark­ness con­di­tions [DD; (B) and (E) for 5 days. The tem­per­a­ture was kept con­stant at 25.5°C. The de­trended data are pre­sented as means ± SD. The shad­ing in (A) and (D) shows the tim­ing of the LD cy­cle (yel­low, light phase; gray, dark phase) rel­a­tive to the bi­o­lu­mi­nes­cence. The hor­i­zon­tal bar in (B) and (E) (lower left) shows the time win­dow of 48 hours se­lected for the analy­sis of pe­riod length. The cal­cu­lated pe­riod length is plot­ted in © and (F); in­di­vid­ual data points [©, N = 16; (F), N = 7] are shown along with the me­dian and in­terquar­tile range. Source. Fron­tispiece: Cir­ca­dian rhythms. Source

Let me now re­turn to the present. Cyano­bacteria, be­ing photo­synthetic, are well equipped to sense light, but what about non-pho­to­syn­thetic bac­te­ria? A new re­port from the lab of Prof. Martha Mer­row in Mu­nich and oth­ers re­veals that B. sub­tilis can also tell time. They con­structed strains of this or­gan­ism with the gene for the light-emit­ting en­zyme lu­ciferase fused to two genes. One, ytvA, codes for a blue light photo­receptor and the other one for an en­zyme called KinC that is in­volved in biofilm for­ma­tion and sporul­ation. Light emis­sion cor­re­lated with the level of gene ac­tiv­ity. They found that when the or­gan­isms were ex­posed to 12-hour cy­cles of light and dark for sev­eral days, lev­els of the ytvA pro­moter ac­tiv­ity in­creased in the dark and de­creased in the light. When these con­di­tions were in­verted, the pat­tern was re­versed. This is in keep­ing with a cir­ca­dian rhythm be­ing present.

Strains whose rhythms had been syn­chron­ized with the environ­ment were more fit; that is, when grown to­gether un­der the con­di­tions that im­pose pe­ri­od­ic­ity, they out­grew the un­trained ones. This fact may not seem sur­pris­ing, but so far it has only been shown us­ing bac­te­r­ial sys­tems (good for our side!). B. sub­tilis does not have the miracul­ous Kai pro­teins of Syne­chococ­cus, but they have some­thing analog­ous, the so called PAS do­mains, which are struc­tural mo­tifs present in all eu­karyotic cir­ca­dian clocks.

It turns out that estab­lishing that an or­gan­ism has a cir­ca­dian rhythm is not as sim­ple as it may sound. It seems to be general­ly agreed that three cri­te­ria need to be met. First, un­der con­stant con­di­tions, the oscilla­tion should have a pe­riod close to 24 hours (but not ex­actly. Note the "circa" in cir­ca­dian). Sec­ond, the rhythm must be "tem­perature com­pensated," that is, it main­tains its rhythm over a range of temp­eratures. Third, the rhythm is "en­trained" to an appro­priate environ­mental cy­cle. When the tim­ing of light and dark ex­posure is changed, so does the peaks in rhythms of gene ex­pression. The au­thors con­clude that these find­ings con­clusively prove the ex­is­tence of a cir­ca­dian rhythm in this or­gan­ism. They go on to say: "Our work opens the field of cir­ca­dian clocks in the free-liv­ing, non­photo­synthetic pro­karyotes, bring­ing con­sid­er­able po­ten­tial for im­pact upon bio­medicine, ecol­ogy, and in­dus­trial processes." I agree that their work sheds light on novel but ex­cit­ing as­pects of bac­te­r­ial phy­siology.

Af­ter writ­ing this piece, it oc­curred to me to con­sult with a B. sub­tilis ex­pert. None bet­ter than my old col­league and friend, A. L. (Linc) So­nen­shein. This was his re­sponse:

Think­ing about cir­ca­dian rhythms in bac­te­ria, it seems rea­son­able for them to ac­com­mo­date to the vari­a­tion in light and at­mos­pheric tem­per­a­ture dur­ing the day. Since an­i­mals of vir­tu­ally all types must be able to sleep at night and stay awake dur­ing the day, the bac­te­ria that come in con­tact with them could have de­vel­oped sim­i­lar re­sponses. B. sub­tilis does not in­fect hu­mans or an­i­mals but just eat­ing, grow­ing, and sur­viv­ing are its big is­sues. And since plants re­spond to light and tem­per­a­ture, bac­te­ria that in­habit gar­dens and forests pre­sum­ably have to "know" when is the best time to look for food. Was this abil­ity de­vel­oped by bac­te­ria be­fore eu­kary­otes arose? If so, the rea­son lies else­where.