Hey fla­gel­lum, shift into re­verse gear!

by Christoph

In a clas­si­cal man­ual gear­box, a gear­wheel – oc­ca­sion­ally called "idler gear" –  is in­serted be­tween the coun­ter­shaft and the drive shaft. This re­verses the di­rec­tion of ro­ta­tion of the drive shaft. Fla­gella of E. coli and Sal­mo­nella have found a smart so­lu­tion to do with­out an ad­di­tional gear­wheel in their gear­box: they run their en­gine along the out­side of a gear­wheel or, if they re­verse the di­rec­tion of ro­ta­tion, on its in­side.

We of­ten do not con­sider small things like bac­te­r­ial fla­gella with the at­ten­tion they de­serve (there are def­i­nitely too many small things). The last time Mike Der­mont did this was six years ago in his piece Fla­gel­lar Mo­tors: How a Bac­terium Shakes its Groove Thing. When I saw the video shown as Fig­ure 6 here in a be­hind the pa­per ar­ti­cle by Steven John­son and Justin Deme a few days ago, I thought it was time to turn my at­ten­tion to fla­gella again. So, here it goes…  and I thought that this time I would walk you through six il­lus­tra­tions.

Fig­ure 1. See text for de­tails. Source. Fron­tispiece: Scre­en­shot from an­i­ma­tion demon­strat­ing gear­ing in the Sal­monella fla­gel­lum. Source

Fig­ure 1  Who in the cell man­u­ally shifts into re­verse gear? Changes in chemo-ef­fec­tor lev­els in the cell's periplasm are de­tected by trans­mem­brane chemo­receptors (usu­ally arranged as "che­mosensory ar­rays", see here in STC). In the cy­to­plasm, these "sig­nal" by se­quen­tially in­duced con­for­ma­tional changes via CheW to the chemo­taxis his­ti­dine ki­nase CheA. In re­sponse to de­creased at­trac­tant con­cen­tra­tion, the chemore­cep­tors trig­ger CheA auto­phos­phorylation. Phos­phorylated CheA (CheA–P) in turn phos­pho­ry­lates ei­ther of its cog­nate re­sponse regula­tors, CheB and CheY.

CheY‑P is the key mol­e­cule here as it binds to the fla­gel­lar mo­tor and pro­motes the switch in the di­rec­tion of ro­ta­tion from an­ti­clock­wise (ccw) to clock­wise (cw). As with all bi­o­log­i­cal sig­nal trans­mis­sions, there are also built-in safe­guards for sig­nal quench­ing: CheB–P is a me­thylesterase that me­di­ates adap­ta­tion in con­junc­tion with the chemo­taxis methyl­trans­ferase, CheR, for ex­ample re­cep­tor clear­ance. CheZ is a spe­cific phos­phatase that de­phos­phorylates CheY–P, al­low­ing rapid sig­nal ter­mi­na­tion.

Fig­ure 2. See text for de­tails. Mod­i­fied from Source

Fig­ure 2  The coun­ter­clock­wise (ccw) and clock­wise (cw) con­for­ma­tions of the mo­tor ex­ist in equi­librium. In the ab­sence of CheY–P, ccw ro­ta­tion is strongly fa­vored. CheY–P bind­ing to FliM and FliN com­po­nents of the switch com­plex shifts the equi­lib­rium in fa­vor of clock­wise rota­tion. There are as­sumed to be ~34 bind­ing sites for CheY–P, and as more of these sites be­come oc­cupied, the prob­a­bil­ity of cw ro­ta­tion in­creases co­op­er­a­tively (but see be­low).

What is la­beled "switch com­plex" in the fig­ure is now re­ferred to as the C‑ring. In E. coli and Sal­mo­nella, the C‑rings, which lie di­rectly un­der the in­ner mem­brane, are al­most with­out ex­cep­tion col­lars with 34 "chain links" in a cir­cu­lar arrange­ment; the chain links are the FliG, FliM and FliN sub­units stacked on top of and in­side each other. FliF sub­units cou­ple the C‑ring to the over­ly­ing MS ring (la­beled "ro­tor") by in­ter­act­ing with FliG in a sto­i­chiom­e­try that has not yet been con­clu­sively deter­mined. Note that the MotAB com­plexes, of­ten re­ferred to as "sta­tor" in di­a­grams, are con­nected to the (periplas­mic) pep­ti­do­gly­can layer via their MotB sub­units, but have no con­nec­tion to the C‑ring. They have, and this will be­come im­por­tant, cru­cially im­por­tant, fur­ther down...

Fig­ure 3. a Us­ing AlphaFold2 to pre­dict the in­ter­ac­tions be­tween CheY–P (yel­low sur­face, phos­phate shown as space fill­ing atoms) and the C‑ring (sil­ver car­toon) ge­nerates an arrange­ment of the FliMM:FliGM do­mains that matches that seen in the cw state (dark green) but not the ccw state (light green, over­laid based on FliMM to high­light rel­a­tive move­ment of FliGM). The pre­vi­ously de­ter­mined in­ter­ac­tion be­tween the FliM N‑terminus (sil­ver he­lix on CheY–P sur­face) and CheY–P is high­lighted. b Map­ping the lo­ca­tion of CheY–P and the full N‑terminus of FliM from the Al­phaFold model to a set of three neigh­bour­ing cw sub­units re­veals the nestling of CheY–P into the C‑ring at a subunit–subunit in­ter­face and with in­ter­ac­tions spread across two sub­units. Source

Fig­ure 3  To nail down the lo­cal­iza­tion of the bind­ing of CheY–P to the C‑ring, John­son et al. (2024) em­ployed model build­ing with AlphaFold2.0 for var­i­ous com­bi­na­tions of the FliG, FliM and FliN do­mains. All com­bi­na­tions that in­volved the FliMM do­main pro­duced mod­els with a con­sis­tent CheY bind­ing in­terface at the top of FliMM (the in­dices in­di­cate definded do­mains of the pro­teins), with the phos­pho­ry­lated residues (D57) point­ing to­wards the FliGM:FliMM inter­face, com­pat­i­ble with bind­ing the FliM N‑terminus, as seen in crys­tal struc­tures (Fig­ure 3a). In­ter­est­ingly, over­lay of the dif­fer­ent struc­tures via the FliMM do­main demon­strated that the FliGM do­main of the ccw state is shifted ap­prox­i­mately 4 Å com­pared with the cw state. Mod­els of CheY–P bound to FliGM:FliMM pro­duced an in­ter­face con­sis­tent with the cw struc­ture, sug­gesting a mech­a­nism where CheY–P bind­ing sta­bi­lizes the cw state (Fig­ure 3b).

Fig­ure 4. Struc­ture of the ccw ro­tat­ing state (top) and of the cw ro­tat­ing state (bot­tom) of the Sal­mo­nella fla­gel­lar C‑ring. Us­ing a mask around 3 sub­units and re­fin­ing sym­me­try ex­panded par­ti­cles gen­er­ates a higher-res­o­lu­tion vol­ume (ccw: 3.2–4.5 Å across the sub­unit, cw: 3.3–5.2 Å across the sub­unit) for the cen­tral sub­unit, al­low­ing build­ing and re­fine­ment of atomic mod­els. Source

Fig­ure 4  Cry­oET analy­sis of Sal­mo­nella basal bod­ies with in­tact C‑rings iso­lated from strains locked in ei­ther the ccw state or the cw state showed John­son et al. (2024) that the over­all prop­er­ties of the C‑ring struc­ture are con­served be­tween the two ro­ta­tional states, includ­ing the ex­pected C34 sym­me­try (34 "chain links" in the ring), de­spite het­ero­gene­ity in the lo­cal arrange­ment of FliG, FliM, and FliN (com­pare both zooms on FliM in Fig­ure 4).

Fit­ting crys­tal struc­tures into the tomogram‑derived vol­umes shown in Fig­ure 4 re­vealed that the mem­brane prox­i­mal re­gion is formed from FliG pro­tomers, teth­ered via a FliF C‑terminus to the MS‑ring. The FliGM do­main forms the main in­ter­ac­tion with a layer of FliM pro­tomers un­der­neath a FliG "ring", and the most mem­brane dis­tal layer is formed from a con­tin­u­ous "ring" of in­ter­locked FliMC:FliN3 sub­com­plexes. Lat­eral con­tacts in both ccw and cw states are dom­i­nated by do­main swaps at all ma­jor lay­ers, many of which in­volve do­mains of FliG. There is a change in the outer perime­ter wall of FliGM:FliMM:FliMC:FliN3, with the cw con­for­ma­tion (Fig­ure 4, bot­tom) be­ing more up­right than the ccw con­for­ma­tion (Fig­ure 4, bot­tom). The FliGM:FliMM do­mains ro­tate through an an­gle of 22° as a rigid ob­ject (re­flect­ing the bind­ing of CheY?), while the FliMC:FliN3 spi­ral pulls up slightly and hinges in­wards rel­a­tive to FliMM.

Again fit­ting crys­tal struc­tures into the tomogram‑derived vol­umes, John­son et al. (2024) found a sub­stan­tial dif­fer­ence for the FliGM do­main, where the cw-locked state at residues 169–171 leads to a shift of the FliGMC he­lix by al­most a full he­li­cal turn. This shift does change some of the con­tacts at the FliGM:FliMM in­ter­face, but the he­lix re­mains bound to FliMM. The C‑terminal end of the FliGMC he­lix leads into a less well-or­dered linker that con­nects the two do­main-swapped por­tions of the FliGM do­main, and this is a ma­jor hinge point be­tween the two states. This linker, in turn, con­tacts the cor­re­spond­ing re­gion of a neigh­boring FliM copy, thereby high­light­ing the im­por­tance of this re­gion of the struc­ture in ty­ing the ring to­gether and prop­a­gat­ing lat­eral struc­tural changes. Also at the C‑terminus of FliGMC is a ma­jor con­tact to the he­lix that links FliGN and FliGM, termed FliGNM.

The most strik­ing struc­tural change the au­thors ob­served af­fects the po­si­tional change of this FliGNM sub­do­main be­tween the two states. In the ccw form, FliGNM folds back through 180° via a turn in­volv­ing residues 101–106, whereas the cw form is a straight he­lix that runs from 87 to 112. This leads to the en­tire FliGN do­main ro­tat­ing through 180° be­tween the two states. The FliGN do­main main­tains its over­all domain‑swapped con­for­ma­tion, but residues 72–99 make con­tact with N−1 rather than N+1 (N stands for neigh­bor­ing FliG). Known mu­tations that in­duce a strong cw bias are lo­cal­ized to FliGNM and are clus­tered around the hinge point and the FliGMC in­ter­ac­tion point, sup­port­ing the no­tion that the ob­served struc­tural change is one of the key dri­ving forces of switch­ing. You can cer­tainly see this 180° flip very clearly in the com­par­i­son of the ccw and cw con­for­ma­tions in Fig­ure 4, even if you can't see the he­lices in­volved there.

Fig­ure 5. See text for de­tails. Source

Fig­ure 5  It is known for decades that the fla­gel­lar mo­tor cre­ates torque "by in­ter­ac­tions be­tween the ro­tor pro­tein FliG, lo­cated at the in­ter­sec­tion be­tween the MS and C rings, and the sta­tor units at­tached to the cell wall. Each (sta­tor) unit is be­lieved to con­tain two copies of MotB and four copies of MotA in pro­ton-dri­ven mo­tors of E. coli ... and func­tions as an ion chan­nel. Ion flux through these chan­nels pow­ers the mo­tor." (ref.) It was only re­motely in­teresting that over time the "sta­tor" changed its sto­chiom­e­try to MotA5B2, but eye-open­ing that it has to be "con­sid­ered as a mo­tor it­self: it con­verts the elec­tro­chem­i­cal po­ten­tial en­ergy from the ion mo­tive force into me­chan­i­cal torque. Upon re­cruit­ment to the basal body and cell wall bind­ing, the sta­tor units un­dergo a con­for­ma­tional change from an inactive/plugged state into an activated/unplugged state. In the un­plugged state, the flux of ions through the sta­tor unit chan­nel en­er­gizes the ro­ta­tion of the ro­tor." (Hu et al. (2021)). Ion flux from the periplasm into the cy­to­plasm through the (yet to be fully char­ac­ter­ized) ion chan­nel within the MotA5B2 "sta­tor" re­sults in a ro­ta­tion of the MotA sub­units in steps of ~36° per trans­ported pro­ton, which ex­erts torque on the FliG sub­unit in the C‑ring. This mo­tor keeps its name "sta­tor" for his­tor­i­cal rea­sons.

John­son et al. (2024) set out to solve the struc­ture of the MotA in­ter­ac­tion with the FliGC down to the con­tact­ing he­lices, in par­tic­u­lar the FliGC "torque" he­lix. For this, they fused the MotA ho­molog from C. sporo­genes with Sal­mo­nella FliGC sep­a­rated by spac­ers of vary­ing lengths. In these chimeras, FliGC do­mains were only seen to bind be­tween closely spaced MotA pairs, that is, a pin­cer grip of a pair of ad­ja­cent MotA do­mains is ap­par­ently re­quired for the bind­ing of FliGC. This in­ter­ac­tion sand­wiches the FliGC do­main be­tween the MotA cy­to­plas­mic do­mains and presents the "torque" he­lix into a cav­ity. Fur­ther mod­el­ling re­vealed that MotA5B2–FliG on the ccw and cw C‑rings switches the sta­tor com­plex from the out­side to in­side the C‑ring.

In Fig­ure 5ab, the MotA5B2–FliGC com­plex is over­laid on the same FliGC do­main with the ccw (a) and cw (b) C‑ring sub­units. The zoom boxes show that the FliGC do­main is not struc­turally al­tered in the con­text of the C‑ring sub­unit and over­lays well. How­ever, the ro­ta­tion of the FliGC do­main in the con­text of the dif­fer­ent ro­ta­tional states re­sults in the MotA5B2 com­plex be­ing pre­sented on the ex­ter­nal (ccw) or in­ter­nal (cw) face of the C‑ring. In Fig­ure 5c, align­ing the ccw and cw struc­tures based on the in­vari­ant FliF com­po­nent of the ro­tor re­vealed that the Sal­mo­nella C‑ring does not change di­am­e­ter as part of switch­ing (over­laid vol­umes with ccw (light green) and cw (dark green), top) and that, in ad­di­tion, the lower two pan­els show the MotA5B2–FliGC com­plex in the con­text of the ccw and cw C‑rings, shows that the sta­tors move from push­ing the C‑ring on the out­side to the in­side of the C‑ring, thus ex­plain­­ing how a sta­tor that ro­tates in a sin­gle di­rec­tion can power bi-di­rec­tional move­ment of the C‑ring.

An aside thought. Isn't it fun to spec­u­late whether the switch of the C‑ring from the ccw into the cw con­for­ma­tion trig­gered by CheY bind­ing to FliM is con­strained or stim­u­lated by the torque ex­erted on FliGC cap­tured in the MotA tweez­ers?

Note that the cross-sec­tion shown in Fig­ure 5c has only two MotA5B2 sta­tors are in con­tact with the C‑ring via FliGC. It had been found ear­lier that up to at least 11 "sta­tors" can dock onto one C‑ring un­der phys­i­o­log­i­cal con­di­tions. High (me­chan­i­cal) load sta­bi­lizes C‑ring-bound "sta­tor" units in both ccw and cw ro­tat­ing mo­tors, in­creas­ing their num­ber over a timescale of a few min­utes. Con­versely, low torque de­creases the life­time of mo­tor-bound sta­tor units. This process en­ables the "stator"-plus‑C‑ring mo­tor to adapt its torque out­put (via FliF to the MS-ring) to the de­mand placed on it by ex­ter­nal load by mod­i­fy­ing its compo­sition (ref­er­ences in Wad­hwa & Berg (2022)). It's a bit like switch­ing in­di­vid­ual cylin­ders on or off as re­quired in an eight-cylin­der in­ter­nal com­bus­tion en­gine (the tech­ni­cal term is vari­able displace­ment).

Fig­ure 6. An­i­ma­tion demon­strat­ing gear­ing in the Sal­mo­nella fla­gel­lum. A small pen­tameric cog (MotA) ro­tates in re­sponse to the pro­ton-mo­tive force (pmf). The teeth of the small cog drive ro­ta­tion of a large cog formed by the C‑ring of the fla­gel­lar basal body, which in turn dri­ves the ro­ta­tion of the fla­gel­lum. Struc­tural tran­si­tions of the C‑ring in re­sponse to chemo­taxis reg­u­la­tors re­model the teeth of the large cog, re­vers­ing the di­rec­tion of ro­ta­tion. Source

Fig­ure 6  John­son et al. (2024) sum­ma­rize their re­sults in a model for "gear shift­ing" in the fla­gel­lar mo­tor in four steps. In the an­i­mated Fig­ure 6 these four steps can be seen in loop, and you only have to add CheY–P your­self.

1. In the ab­sence of CheY–P, the in­vari­ably clock­wise-ro­tat­ing MotA5B2 mo­tor "sta­tor" in­ter­acts with the ex­ter­nal face of the C‑ring to drive ccw ro­ta­tion and tran­sien­tly ten­sions the in­ter­act­ing sub­unit.

2. CheY–P bind­ing at a tran­siently ten­sioned sub­unit tran­si­tions the FliG do­main, which swaps to the cw state and co­op­er­a­tively dri­ves the spread of a conforma­tional change around the C‑ring.

3. In the pres­ence of CheY–P, the "sta­tor" in­ter­acts with the in­ter­nal face of the C‑ring to drive CW ro­ta­tion and tran­siently ten­sions the in­ter­act­ing sub­unit.

4. CheY–P dis­so­ci­a­tion from a tran­siently ten­sioned sub­unit tran­si­tions the FliG do­main, which swaps to the ccw state and co­op­er­a­tively dri­ves the spread of a con­for­ma­tional change around the C‑ring.

De­spite all the enor­mous progress in cryo-elec­tron to­mog­ra­phy (cry­oET), which now makes it pos­si­ble to track con­for­ma­tional changes down to in­di­vid­ual he­lices and loops of inter­acting pro­teins, there is no tech­nol­ogy to "watch" this in real time. The ac­tual "switch" be­tween two confor­ma­tions, how fla­gella shift gear into re­verse, can only be mod­eled. But still.

I should not fail to men­tion that in the same is­sue of Na­ture Mi­cro­bi­ol­ogy, Singh et al. (2024) pub­lished their study that es­sen­tially came to the same con­clu­sions as the study by John­son et al. (2024) pre­sented here, even if there are dif­fer­ences in de­tails be­tween the two. Both were fa­mil­iar, of course, with the work of Chang et al. (2020), who stud­ied fla­gel­lar move­ment in Bor­re­lia and al­ready pre­sented a model for switch­ing to re­verse gear by po­si­tional swap­ping of the MotA5B2 "sta­tor" on the C‑ring.

If the an­i­ma­tion in Fig­ure 6 is too min­i­mal­is­tic for your taste, I rec­om­mend watch­ing the full-blown 3D an­i­ma­tion C‑ring on the Mo­tor Goes Vroom­VroomVroom by Singh et al. (2024) (but for heaven's sake mute your speaker, the sound of the wooden ratchet is nerve‑wrack­ing). It is not even im­por­tant who pub­lished first. Both stud­ies build on years of work by many other re­search groups in­volved and of course men­tion this. There has al­ways been a stim­u­lat­ing com­bi­na­tion of both col­lab­o­ra­tion and com­pe­ti­tion in the loose "fla­gel­lar com­munity" (not to be con­fused with re­li­gious fla­gel­lants). And when two com­pet­ing teams come to very sim­i­lar con­clu­sions within a year or so, this can only be pos­i­tive for pro­gress in un­der­stand­ing fla­gel­lar move­ment.

Even if we now have a fairly ac­cu­rate idea of how E. coli and Sal­mo­nella "shift into re­verse gear" their fla­gella... other fla­gella-bear­ing bac­te­ria may very well have evolved other, si­mi­lar­ly smart vari­a­tions on this theme. Prob­a­bly none with an ad­di­tional cog­wheel in the gear­box, though.

 

Do you want to com­ment on this post? We would be happy about it! Please send us an email, or com­ment on Mastodon, or on Bluesky.