Some Like It Curved

by Elio

Think about it: un­der­stand­ing what goes on in­side any cell is hard enough. Add to that the com­plex­i­ties of be­ing rod-shaped, and the mind starts to bog­gle. Peo­ple have won­dered about this aplenty. For sure, many ex­pe­di­tions  have set out to ex­plore the poles of rod-shaped bacilli — how they are cre­ated, what mol­e­cules pre­fer to re­side at this site and how this in­flu­ences the phys­i­ol­ogy of the cell. In truth, de­fin­i­tive an­swers are gen­er­ally want­ing, but re­cent pa­pers help us un­der­stand some of the com­plex­i­ties of rod-ness. A most read­able re­view of the whole is­sue of po­lar lo­cal­iza­tion is this one by Bardy and Mad­dock.

E. coli Cell En­ve­lope Op­er­at­ing In­struc­tions. (a) Pro­teins dis­trib­uted in the in­ner mem­brane are found in dis­tinct do­mains e. g., po­lar, lat­eral, and sep­tal. Along the cy­linder, some pro­teins are found at spe­cific lo­ca­tions, per­haps by as­so­ci­at­ing with fu­ture cell di­vi­sion sites, or be­cause of the MreB he­li­cal fil­a­ment. The mem­brane com­po­si­tion varies, with CL sig­nif­i­cantly en­riched at the cell poles (yel­low shad­ing). (b) The pep­ti­do­gly­can layer is in­ert at the poles (darker shad­ing). MreB-de­pen­dent and FtsZ-de­pen­dent in­ser­tion of new PG oc­curs along la­teral edge and sep­tum, re­spec­tively (green dia­monds). (c) Po­lar outer mem­brane pro­teins are rela­tively im­mo­bile, as in­di­cated by the darker shad­ing. (d) Com­pos­ite over­lay of im­ages (a–c). Source

A re­spectable num­ber of re­ports de­scribe the se­lec­tive lo­cal­iza­tion of pro­teins and lipids at or near bac­te­r­ial poles and the septa. (See re­cent re­views by Kirk­patrick and Vi­o­l­lier and Shapiro, McAdams and Losick.) Pro­teins lo­cated at the poles in­clude chemo­taxis re­cep­tors (the first to be so lo­cal­ized), pro­teins in­volved in po­lar chro­mo­some at­tach­ment, con­stituents of the cell di­vi­sion ap­pa­ra­tus and their pro­teases, and many more. And, of course, there are such things as po­lar fla­gella, pili, and mechanosen­si­tive chan­nels. Over­pro­duced pro­teins form ag­gre­gates (in­clu­sion bod­ies) that end up at or near the poles, but that's an­other story. Pep­ti­do­gly­can also has a po­lar wrin­kle in that the lay­er­located at the poles dif­fers from the rest in be­ing meta­bol­i­cally in­ert; that is, it does not un­dergo turnover like that in the cylin­dri­cal part of the cell. So, bac­te­ria are not just bags of en­zymes. But how many times have you heard that?

Many ques­tions arise, no­tably what cer­tain pro­teins are do­ing at the poles. Here we will not deal with this ques­tion but will fo­cus on how these pro­teins find their po­lar sites. Bac­te­ria lack the vesi­cle sys­tems used by eu­kary­otic cells for pro­tein traf­fic, so how do they do it? The sim­plest thought is that pro­teins dif­fuse freely, ei­ther through the cy­to­plasm or along the mem­brane, even­tu­ally to be­come cap­tured at magic sites con­tain­ing al­lur­ing pro­teins. This, of course begs the ques­tion: how did those pro­teins get there?

Be­fore go­ing on, a note of cau­tion: When anom­alous pro­teins, such as those with GFP or other fu­sions, are ex­pressed at high lev­els, they tend to ag­gre­gate into in­clu­sion bod­ies that then end up near the poles. (See here.) This does not mean that the nor­mally ex­pressed pro­teins be­long there. Thus, re­sults with such fu­sions should be taken with a grain of salt. For more on flu­o­res­cent tech­nol­ogy, see a re­view by Kent­ner and Sour­jik.

Surely there are di­verse mech­a­nisms for po­lar lo­cal­iza­tions, but one stands out for its ap­par­ent sim­plic­ity. It's the no­tion that phos­pho­lipids in­volved are curved in shape. Such phos­pho­lipids nat­u­rally pre­fer curved sur­faces and there­fore in­ter­ca­late at sites where the mem­brane is  curved. These are, of course, the septa and the poles (keep­ing in mind that septa even­tu­ally be­come poles). When in­ter­ca­lated, these phos­pho­lipids con­tribute to mak­ing even greater cur­va­ture. "Po­lar pro­teins," as the thought goes, then an­chor to such curved sites. The cue to mak­ing poles may there­fore not be chem­i­cal but geo­met­ric! Note that this works both for in­sert­ing ma­te­r­ial at the in­ner con­cave face of the pole or at the con­vex out­side.

Key phys­i­cal char­ac­ter­is­tics for quan­ti­fy­ing cur­va­ture sen­si­tiv­ity of sin­gle car­di­olipin mol­e­cules. A. Cardio­lipin's unique dimeric struc­ture, with two phos­phatidyl head groups and four acyl chains, is in­dica­tive of a con­i­cal shape with a small ra­dius of cur­va­ture R_lipid and hence a pref­er­ence for highly curved geome­tries. B. The en­ergy scale κ is set by the stiff­ness of the mem­brane for bend­ing. C. The en­hanced cur­va­ture of the pole is the dif­fer­ence be­tween the cur­va­ture at the pole and at the cylin­dri­cal mid­cell, ΔC_cell = 1/R. The com­bi­na­tion of κ, C_lipid,ΔC_cell and the sur­face area A_lipid de­ter­mines the ex­tent to which the mol­e­cule will pref­er­en­tially lo­cal­ize at the pole. Source

But here's the ac­tual story. One of the main lipids in the E. coli in­ner mem­brane, car­di­olipin (CL), lo­cal­izes at the poles and septa of the cells. How come? First, a bit about CL. It makes up about 5% of the to­tal phos­pho­lipids and dif­fers from other phos­pho­lipids in be­ing hav­ing a dimeric struc­ture (of phos­phaty­dyl­glyc­erol. See the fig­ure). This un­usual con­fig­u­ra­tion makes CL as­sume the shape of a con­i­cal wedge (with a cur­va­ture value of ~1–5 nm⁻¹), which fa­vors its in­ser­tion into curved por­tions of the mem­brane. Cal­cu­la­tions show that af­fect­ing mem­brane cur­va­ture re­quires the co­op­er­a­tion of sev­eral mol­e­cules of CL to­gether be­cause the bend­ing strength of a sin­gle one is too weak. CL mol­e­cules found at non-pole lo­ca­tions may be kept from mak­ing large ag­gre­gates be­cause — it is thought —  os­motic forces ex­erted on the mem­brane and cell wall keep high-cur­va­ture lipids from ag­gre­gat­ing into do­mains. (See here and here.)

Of course, the pro­teins des­tined for lo­cal­iza­tion must be able to rec­og­nize ei­ther CL mol­e­cules di­rectly or the cur­va­ture they cre­ate. In fact, pro­teins with such in­trin­sic prop­er­ties are known to ex­ist, a fact well es­tab­lished for eu­kary­otes. In Bacil­lus sub­tilis, oligomers of the di­vi­sion pro­tein called Di­vIVA in­sert at re­gions of neg­a­tive mem­brane cur­va­ture, those of the sporu­la­tion fac­tor SpoVM, at pos­i­tive ones. For a dis­cus­sion, see a re­view by Huang and Ra­ma­murthi.

How has it been learned that CL likes be­ing lo­cal­ized at the poles? This has been shown in two ways. First, the car­di­olipin-spe­cific dye 10-N-nonyl acri­dine or­ange (NAO) was used to demon­strate that car­di­olipin is en­riched at the cell poles and near po­ten­tial di­vi­sion sites in E. coli, Pseudomonas putida and B. sub­tilis. Sec­ond, E. coli mini­cells (small round cells formed by di­vi­sion near the poles) are en­riched for CL. Now the clincher: When the po­lar lo­cal­iza­tion of CL was di­min­ished by mu­ta­tion or os­motic ef­fects, the po­lar lo­cal­iza­tion of a pro­line trans­port pro­tein, ProP, and of a mechanosen­si­tive chan­nel pro­tein also de­creased.

CL and MinD colo­cal­iza­tion in E. coli spher­o­plasts con­fined in mi­crocham­bers. (A) In spher­i­cal micro­cham­bers. (B) In rod-shaped mi­cro­­chambers.BF = phase con­trast. CL = CL la­beled with NAO. MIN D = MinD-YFP. Merge = over­lay of the CL and MinD im­ages. DNA = DAPI la­beled DNA. Dashed cir­cles in­di­cate the peri­meter of the mi­crocham­bers. Source

But when you think of it, all these lovely find­ings rep­re­sent cor­re­la­tions. More di­rect ev­i­dence would be wel­come. Such is pro­vided in a re­cent study by Ren­ner and Weibel. They made large spher­o­plasts of E. coli by first in­duc­ing them to grow as fil­a­ments and then treat­ing them with lysozyme. The cells be­came big spheres 3–4 μm in di­am­e­ter. They now placed these in spe­cial mi­crocham­bers of dif­fer­ent di­am­e­ters, some where the spher­o­plasts had to squeeze in and thus be­come elon­gated. (This re­minds me of a song from the '40's Please No Squeeze De Ba­nana.) On squeez­ing in one di­men­sion, CL be­came lo­cal­ized at the re­gions of great­est cur­va­ture, the ar­ti­fi­cial "poles." MinD, a pro­tein in­volved in cell di­vi­sion and known to os­cil­late from pole to pole co-lo­cal­ized with CL. Now, are you con­vinced that CL is in­volved in the lo­cal­iza­tion of pro­teins at the bac­te­r­ial poles? I am (for now).

The in­tri­ca­cies of sub­cel­lu­lar or­ga­ni­za­tion are now emerg­ing fast and fu­ri­ous. Note, for ex­am­ple the re­cent dis­cov­ery by Lopez and Kolter of the bac­te­r­ial equiv­a­lent of the lipid rafts in eu­kary­otic mem­branes. These spa­tially de­lin­eated mem­brane re­gions are surely in­volved in spe­cial func­tions. An ex­am­ple al­ready re­ported is that mu­tants lack­ing some pro­teins needed for lipid raft for­ma­tion are de­fec­tive in a sig­nal trans­duc­tion path­way whose sen­sor ki­nase is nor­mally lo­cated in the rafts. And if that were not enough to point to the in­tri­ca­cies of bac­te­r­ial struc­tures, Sun and col­leagues (re­viewed here) re­cently found that glid­ing motil­ity in Myx­o­coc­cus de­pends on the first known bac­te­r­ial mo­tor able to move in a di­rected man­ner be­tween re­gions of the cell! Bags of en­zymes, bah!

Ref­er­ence

Ren­ner LD, Weibel DB. (2011). Car­di­olipin mi­crodomains lo­cal­ize to neg­a­tively curved re­gions of Es­cherichia coli mem­branes. Proc Natl Aca Sci USA, 108 (15), 6264–6269. PMID 21444798