Ki­tasatospora, a bac­terium pro­moted to hon­orary amoeba

by Christoph

A decade ago, Elio wound up a piece on en­do­cy­to­sis by E. coli with the bold re­mark: "If in­deed E. coli is only one pro­tein away from en­do­cy­to­sis, per­haps this work will com­pel us to cross endo­cytosis off the list of physiolo­gical traits that are ex­clu­sive at­trib­utes of the eu­kary­otic cell plan." For sure, ex­press­ing hu­man caveolin‑1 in E. coli to tempt it to endo­cy­tose was an ar­ti­fi­cial sce­na­rio. But al­ready back in 2013, Merry pointed out that while endo­cy­tosis is rare among bac­te­ria, it is not un­heard of, and the en­­do­­cy­­to­­sis-like up­take of Pseudomonas phage φ6 by its host serves as ex­am­ple. Want more ex­am­ples?

En­ter Ki­ta­sa­to­spo­ra vi­ri­di­fa­ciens, a close rel­a­tive of Strep­to­my­ces (see here and here in STC). In a re­cent post, Roberto went into de­tail about the work on Ki­ta­sa­to­spo­ra done in the labs of Den­nis Claessen and Gilles van Wezel at Lei­den Uni­ver­sity, NL. In short, the re­searchers had found that the out­grow­ing mycelial tips read­ily shed their cell wall un­der os­motic stress and con­tinue to live as pro­to­plasts (they call these cells "S‑cells"). Such S‑cells can be se­pa­ra­ted from the mycelium and live on for days, but they pro­lif­er­ate only – and form colonies on solid agar plates – upon restora­tion of their cell wall. Roberto hypo­the­sized that S‑cells could es­cape the im­mune sys­tem of plants and travel through their tis­sue (rather than grow­ing in­side the tis­sue as mycelium). This has not yet been tested but On­ge­nae et al. (2021) found that S‑cells es­cape in­fec­tion by phages! To get rid of their cell wall, at least tem­porar­ily, is cer­tainly a clever strat­egy of bac­te­ria as it is a pre­ferred tar­get of many phages.

Fig­ure 1. DNA up­take by cell wall-de­fi­­cient cells. Myce­lium (n=3), pro­to­plasts (n=5 from two ex­per­i­ments), S‑cells (n=7 from two ex­per­i­ments) and L‑form lines al­pha (n=3 for both 4‑and 6‑­day-old cells), delta (n=2 for both 3‑and 7‑day old cells) and M1 (n=1 for 3‑­day-old cells) were in­cubated with plas­mid DNA (pRed*) for 18–24 h, plated on se­lec­tive me­di­um and in­cubated at 30°C to se­lect for trans­formed cells. A close-up of the colony mor­phol­ogy of the L‑forms is given. Scale bars in­di­cate 1 mm. Note that only L‑ forms show con­sis­tent DNA up­take. Fron­tispiece: Pop­ping en­do­cy­tosed bal­loons (de­tail). By Shrad­dha Shi­tut. Source

Now Renée Kapteijn, Shrad­dha Shi­tut and col­leagues have pub­lished a study show­ing that Ki­ta­sa­to­spo­ra is ca­pa­ble of en­do­cy­to­sis. To get there was a con­vo­luted jour­ney, and I will guide you step­wise along the way. They had no­ticed that rarely and spon­ta­neously some of their care­fully tend­ed S‑cells pro­lif­er­ated with­out hav­ing re­stored the cell wall. They were able to cul­ti­vate some of these weirdos as rea­son­ably sta­ble lin­eages whose sin­gle cells be­have like typi­cal L‑forms, that is, they're de­void of the cell wall (CWD) and re­quire con­ti­nu­ous os­motic sta­bi­liza­tion. They di­vide by bleb­bing, that is, ex­o­cy­to­sis of cy­to­plas­mic con­tent via mem­­brane-bound vesi­cles, rather than by FtsZ‑driven bi­nary fis­sion or or­derly bud­ding as yeasts do. The ploidy of these cells is vari­able, and their chro­mo­somes carry muta­tions, dele­tions, and rear­rangements (the lat­ter was also ob­served for trans­formed pro­to­plasts upon re­ge­ne­ra­tion).

Kapteijn et al. were in­ter­ested to learn if these Ki­ta­sa­to­spo­ra L‑forms were amenable to ge­netic ma­nip­u­la­tions by trans­for­ma­tion. They were in­deed, much like con­ven­tion­ally pre­pared pro­to­plasts or S‑cells, by PEG‑based transforma­tion with plas­mid DNA (no one un­der­stands how ex­actly po­ly­ethyl­ene‑gly­col (PEG) based trans­for­ma­tion works but it usu­ally does work with bac­te­r­ial, fun­gal, and plant pro­to­plasts). Sur­pris­ingly, un­like pro­to­plasts or S‑cells, the L‑form lin­eages could be trans­formed by sim­ply adding "naked" plas­mid DNA and pro­longed in­cu­ba­tion (Fig­ure 1). Could nat­ural com­pe­tence, that is, the abil­ity to take-up ex­tra­cel­lu­lar DNA via type IV pili (T4P) and a ded­i­cated membrane‑bound pro­tein com­plex be at play here?

Fig­ure 2. For­ma­tion of in­ter­nal vesi­cles and up­take of ex­tra­cel­lu­lar ma­te­r­ial in L‑forms. Rep­re­sen­ta­tive fluo­rescence mi­cro­graph of al­pha pI­J82-GFP (cy­to­plas­mic eGFP; green) in­cu­bated with Cy5-la­beled plas­mid DNA (pFL-ssgB; ma­genta) for 3 days (n=3 ob­ser­va­tions from one experi­ment). BF bright­field. Source

The Ki­ta­sa­to­spo­ra genome en­codes bona fide ho­mologs of the B. sub­tilis ComEA (DNA-bind­ing) and ComEC (membra­ne chan­nelling) com­pe­tence pro­teins. When Kapteijn et al. re­placed the neigh­bor­ing comEA and comEC genes simul­taneously with an apramycin re­sis­tance cas­sette in the L‑form lin­eage al­pha, they saw no sig­nif­i­cant ef­fect on the trans­for­ma­tion effi­ciency. This in­di­cated that DNA up­take by L‑forms oc­curs in­de­pen­dent­ly of genes ho­mol­o­gous to the canon­i­cal DNA translo­ca­tion ma­chin­ery. But how?

The re­searchers fol­lowed, us­ing flu­o­res­cence mi­croscopy, the path of ex­ter­nally-added, Cy‑5‑labelled plas­mid DNA into L‑form cells that ex­pressed eGFP in their cy­to­plasm. Fig­ure 2 shows a typ­i­cal re­sult with in­ter­nal­ized plas­mid DNA (ma­genta) con­tained in what re­sembles a vesi­cle in the cy­to­plasm (green). Note that some plas­mid DNA is still out­side the cen­trally lo­cated glob­u­lar cell, and the glob­u­lar in­ter­nal vesi­cle is de­void of eGFP flu­o­res­cence, in­di­cat­ing that it has not merged with the cy­to­plasm. Don't miss the an­i­mated 3D re­con­struc­tion of this mi­croscopy ex­per­i­ment, which is par­tic­u­larly im­pres­sive if you are strug­gling with the 2D im­ages.

Kapteijn et al. rea­soned that such in­ter­nal vesi­cles could have orig­i­nated by an in­vagi­na­tion of the mem­brane, whereby ex­tra­cel­lu­lar ma­te­r­ial be­comes trapped in­side the vesi­cles. To test this, they in­cu­bated eGFP-ex­press­ing L‑forms with the flu­o­res­cent mem­brane dye SynapseRed (equiv­a­lent to FM5‑95). And in­deed, the mem­brane dye stained the cell mem­brane of the L‑forms but also the mem­branes of in­ter­nal vesi­cles af­ter overnight in­cu­ba­tion (see here). Since SynapseRed is mem­brane‑im­per­meable, the con­sis­tent stain­ing of the ve­sicle's mem­brane must have oc­curred while it was still part of the L‑form mem­brane!

Fig­ure 3. For­ma­tion of in­ter­nal vesi­cles and up­take of ex­tra­cel­lu­lar ma­te­r­ial in L‑forms. Rep­re­sen­ta­tive micro­graph of the for­ma­tion of foci and ring-struc­tures of Di­vIVA-eGFP in al­pha pKR2 (green) af­ter overnight in­cu­ba­tion with Dex­tran-Texas Red (D‑TR, ma­genta). More than 50 ob­ser­va­tions from one ex­per­i­ment. Note that L‑forms are able to take up fluo­rescently stained DNA and dex­tran by the for­ma­tion of in­ter­nal vesi­cles. Source

En­do­cy­to­sis re­quires mem­brane re­mod­el­ling, es­pe­cially the for­ma­tion of re­gions of neg­ative mem­brane cur­va­ture (this ap­plies also to bleb­bing). The di­vi­some pro­tein Di­vIVA (pro­nounced "Div‑4‑A") of Fir­mi­cutes and Acti­nobac­te­ria binds pref­er­en­tially to such neg­a­tively curved mem­brane re­gions, even­tu­ally form­ing a neck­lace-type oligomer that en­cir­cles a mem­brane "bottle­neck". Kapteijn et al. found Ki­ta­sa­to­spo­ra Di­vIVA tagged with eGFP as com­plexes close to in­ter­nal­ized ve­sicles and even ob­served said neck­lace struc­tures (Fi­g­ure 3). How­ever, the trans­for­ma­tion ef­fi­ciency of a Δdi­vIVA mu­tant was com­pa­ra­ble to that of the di­vIVA+ par­ent of the L‑form lin­eage al­pha. There­fore, Di­vIVA is not a trig­ger of en­docytosis but is, for fluo­res­cence mi­croscopy, a wel­come cy­to­plas­mic marker for endo­cy­tosis when fused to eGFP.

I men­tion for the sake of com­plete­ness that Kapteijn et al. also demon­strated up­take via endocy­tosis of lipid nanopar­ti­cles and dex­tran in ad­di­tion to DNA up­take. Strik­ingly, they found a signi­ficant re­duc­tion of dex­tran up­take by L‑form cells in the pres­ence of 2.5 mM sodium azide (NaN3), a po­tent in­hibitor of the res­pi­ra­tory chain, and in­cu­ba­tion at 4°C in­hi­bi­ted up­take com­pletely. This sug­gests that the up­take process in L‑forms, the in­ter­na­li­za­tion of ex­ter­nal ma­te­r­ial by mem­brane in­vagi­na­tion, is en­ergy-de­pen­dent like en­do­cy­to­sis in eu­kary­otic cells. Ki­ta­sa­to­spo­ra L‑forms thus dis­pose of a com­plete en­do­cy­to­sis pro­gram that is on par with that of eu­kary­otic amoe­bae.

Fig­ure 4. 3D cryo-flu­o­res­cence and cryo-FIB-SEM of L‑forms re­veal their ultra­structure in high res­o­lu­tion. j–q Analy­sis of the inter­con­nected vesi­cles of the cell in (i) (n=1). j–l Three con­sec­u­tive slices show­ing the in­ter­ac­tion of dif­fer­ent ve­sicles. n–p show higher mag­ni­fi­ca­tion of the re­gions in white boxes in j–l, respec­tively. m,q 3D seg­men­ta­tion of n–p. While some of the vesi­cles are in­tra­cel­lu­lar, oth­ers pro­trude out of the cell. A com­plete con­nected vesi­cle struc­ture is shown in green (m,q) and in­di­cated by white ar­rows (i,j,l,m). The cell in pan­els i–q is around 4 μm in size. Scale bars re­present 500 nm un­less oth­er­wise speci­fied. Source

Fi­nally, the re­searchers wanted to gain a more de­tailed in­sight into ul­tra­struc­ture and com­po­­sition of in­tra­cel­lu­lar vesi­cles in Ki­ta­sa­to­spo­ra L‑forms and ap­plied for this pur­pose ad­vanced elec­tron mi­croscopy, cryo-CLEM and cryo-FIB-SEM. They found that L‑forms mostly con­tain (un­like in the im­ages shown here in Fig­ures 2&3) mul­ti­ple vesi­cles of vary­ing sizes, in some cases form­ing clus­ters that pro­truded out of or in­truded into the cell mem­brane (Fig­ure 4). Fully in­ter­nal vesi­cles, con­spic­u­ous by a de­vi­at­ing elec­tron den­sity with signs of lipid droplet for­ma­tion, may re­lease their con­tents into the cell af­ter vesi­cle degra­da­tion (Fig­ure 4, see here for a par­ic­u­larly clear ex­am­ple). This last point, vesi­cle degra­da­tion in­side L‑form cells, is still opaque. But it clearly hap­pens quite ef­fi­ciently as they could re­pro­ducibly trans­form their L‑form lin­eages via vesi­cles with plas­mids whose genes were ex­pressed af­ter ve­sicle up­take, degra­da­tion, and mix­ing of the con­tents with the cy­to­plasm.

The elec­tron mi­croscopy ex­per­i­ments show that both bleb­bing, that is, ex­pul­sion of cyto­plas­­mic con­tent via mem­bra­ne-bound vesi­cles, and en­do­cy­to­sis are struc­turally re­lated and in­herent prop­er­ties of the mal­leable cell mem­brane. So, "hon­orary amoeba" is all well and good. Still miss­ing are the pseudopo­dia to se­cure Ki­ta­sa­to­spo­ra vi­ri­di­fa­ciens te­nure. As Renée Kap­teijn said on Twit­ter: "It would be very cool if we can make it even more blobby and amoeba-like." So, wait for it...

On a more se­ri­ous note but highly spec­u­la­tive, Kapteijn et al. con­sider it pos­si­ble that the pro­fi­ciency for en­do­cy­to­sis in present-day L‑form bac­te­ria is an evo­lu­tion­ary an­cient – and to­day rarely used – legacy of pri­mor­dial prokary­otic cells that did not yet pos­sess cell walls and re­gularly ex­changed cy­to­plas­mic con­tents among each other. A long, long time be­fore an arch­ae­­on endo­cyto­sed a prop­erly cell-walled pro­teobac­terium en route to be­come a eu­kary­otic amoeba.

 

Other Posts

  • Bdellovibrio's Ap­petite for Metabo­lites

    by Henry N. Williams — Not long ago, Elio said in this blog that pre­da­tion, a ma­jor force in evo­lu­tion, is some­what ne­glected in mi­cro­bi­o­log­i­cal cir­cles. The full im­pli­ca­tions of pre­da­tion are just be­gin­ning to be un­cov­ered as more be­comes known about the ecol­ogy, phys­i­ol­ogy, and ge­nomics of preda­tors against mi­crobes and their in­ter­ac­tions…

  • How to grow shrubs

    by Christoph — Ex­actly eight years ago to­day, we in­tro­duced you to the 'Hedge­hogs In Your Gar­den'. No, we had not turned our backs on the Small Things and be­came gar­den­ers in­stead. Again to­day, it is not about gar­den­ing ac­tiv­i­ties, namely, 'How to grow shrubs'. Let me ex­plain...

  • An Il­lus­trated Tree

    by Roberto — From the fun­da­men­tal con­cepts of bio­lo­gy, the fact that all or­gan­isms now alive share a com­mon an­ces­tor (LUCA) ranks as one of my fa­vo­rites. When I pause to con­si­der that, des­pite four bil­lion years of evo­lu­tion and di­ver­gence, my ri­bo­so­mal RNA shares some stret­ches of se­quence iden­ti­ty with all other au­to­no­mous or­gan­isms, I still get goose bumps...

  • Ar­chaeal Nin­jas

    by Elio — We con­fess to hav­ing missed a re­port of one of the mi­cro­bial world's most un­usual struc­tures. These are pili-like fil­a­ments lo­cated on an ar­chaeon that look for all the world like strands of barbed wire with a three-pronged grap­pling hook at the end. The ar­ti­cle ap­peared in 2005 and has not had any fol­low-up that we could find.

  • No Phos­pho­rus? No Prob­lem!

    (There's More Than One Way to Skin a Phy­to­plank­ton) by Elio — Are you a lipophobe on a low fat diet? Or do you just be­lieve that lipids are bor­ing? Try again. Mo­not­o­nous they are not. There's greater va­ri­ety to be found among the lipids than al­most any other cell con­stituent. It's truly mind-bog­gling. Within the prokary­otes alone, ar­chaeal mem­brane lipids dif­fer from "reg­u­lar" ones in nearly every re­spect...

  • Putting Re­dun­dancy to Work

    by Ka­t­rina Nguyen — Ge­netic re­dun­dancy, when two genes en­code for the same func­tion, is wide­spread among many or­gan­isms. Re­dun­dant genes con­fer an ad­van­tage: if one gene is lost, its part­ner can sub­sti­tute and the phe­no­type of the or­gan­ism will not change. It is un­clear how re­dun­dancy is main­tained dur­ing evo­lu­tion since se­lec­tion against…