A Holin One

by Merry Youle

When time's up, an in­fect­ing phage ly­ses its host cell, thus re­leas­ing the as­sem­bled viri­ons. (See our pre­vi­ous post for more about ly­sis.) This process has been in­ten­sively stud­ied in the col­iphage λ. Here the phage-en­coded holin pro­teins are said to "per­me­abi­lize" the E. coli mem­brane, thus al­low­ing the en­dolysins ac­cu­mu­lated in the cy­to­plasm to pass through and at­tack the cell wall. "Per­me­abi­lize" has such a po­lite sound, sug­ges­tive of an or­derly mod­i­fi­ca­tion of the mem­brane to al­low the folded en­dolysins to qui­etly slip through. But what sort of chan­nels or pores do the holins ac­tu­ally make?

Cry­o­elec­tron to­mog­ra­phy of a λ holin (S105) le­sion. Seg­men­ta­tion of the en­ve­lope den­si­ties shows the outer mem­brane (blue) and in­ner mem­brane (or­ange). Source

First of all, the holes have to be large — large enough to pass the folded 18 kDa en­dolysin pro­tein. That means they have to be way larger than the largest chan­nels known for bac­te­r­ial cell mem­branes, which are the 5 to 9 nm di­am­e­ter chan­nels formed by the twin-argi­nine translo­ca­tion path­way (Tat). To gauge the size of the holin holes, re­searchers fused a se­ries of pro­tein do­mains of vary­ing sizes to the C‑terminus of the en­dolysin and tested the abil­ity of the chimerae to exit through the cell mem­brane. A fu­sion pro­tein al­most 8 times the size of the en­dolysin passed right through and so did tetramers of that pro­tein. The holins aren't very ex­act­ing gate-keep­ers, ei­ther. When the holes made by one phage were tested us­ing en­dolysins with dif­fer­ent folded struc­tures from dif­fer­ent phages that in­fect un­re­lated hosts, the en­dolysins ex­ited as usual.

To­gether with other lines of ev­i­dence, these find­ings sug­gest that the holins don't form a struc­tured chan­nel, but in­stead gen­er­ate a rad­i­cal dis­rup­tion of the mem­brane. Suc­cess in vi­su­al­iz­ing the holes came re­cently when re­searchers com­bined cryo-EM with some other tricks. Cryo-EM re­quires rather thin ob­jects, so they grew their E. coli strains on suc­ci­nate as the car­bon source, thus net­ting an 80% re­duc­tion in cell vol­ume rel­a­tive to that in LB. They en­gi­neered E. coli strains that could be in­duced to ex­press the holin ei­ther with or with­out the en­dolysin. (These strains lacked the phage pro­teins needed to dis­rupt the outer mem­brane, so that mem­brane re­mained in­tact.)

Cryo-EM of λ holin (S105) le­sions in two E. coli cells ex­press­ing holin but not en­dolysin. The white solid lines in­di­cate the lo­ca­tion and ex­tent of the le­sions. Bar = 500 nm. IM = in­ner mem­brane. OM = outer mem­brane. Source

Cells ex­press­ing both holin and en­dolysin lysed about 50 min­utes af­ter in­duc­tion. Those with­out the en­dolysin re­mained in­tact but ceased growth at that same time, sug­gest­ing that the holins had done their work and made holes in the mem­brane. Sam­ples col­lected at 60 min­utes were plunge-frozen into liq­uid ethane and im­aged un­der liq­uid ni­tro­gen tem­per­a­tures with­out any fur­ther ma­nip­u­la­tion. A close look at cells ex­press­ing holin but no en­dolysin re­vealed gaps in the in­ner mem­brane, an av­er­age of two per cell, at ap­par­ently ran­dom lo­ca­tions. These gaps are huge! Their av­er­age size was ~340 nm, the largest was 1.2 µm. This is more than 10-fold larger that the largest mem­brane le­sions known — the ~30 nm pores formed in cho­les­terol-con­tain­ing mem­branes of an­i­mals by cy­tolysins.

Much mys­tery re­mains. As the holins ac­cu­mu­late in the mem­brane, they have no ap­par­ent ef­fect on the host un­til the in­stant when they "trig­ger" hole for­ma­tion. Just how they "trig­ger" or form the hole is not known. How­ever they do it, this is a skill shared by an ex­ceed­ingly di­verse ar­ray of holins be­long­ing to more than 50 un­re­lated pro­tein fam­i­lies. There is as yet no di­rect ev­i­dence that the holins them­selves form the holes, but more ap­peal­ing sce­nar­ios are hard to come by. It is known that about a thou­sand λ holin mol­e­cules are suf­fi­cient to rup­ture the E. coli mem­brane. Based on their size and pre­dicted con­for­ma­tion, a thou­sand of them sit­ting side-by-side could make the two holes these re­searchers ob­served per cell.

So far, the au­thors' at­tempts to lo­cal­ize flu­o­res­cently-tagged holins in the le­sions have been blocked by tech­ni­cal prob­lems. This and other strate­gies are on their agenda. In their words: It is hoped that these ap­proaches will pro­vide fur­ther in­sight into the mys­te­ri­ous mech­a­nism by which the holin man­ages the tem­po­rally sched­uled tran­si­tion be­tween the per­fectly main­tained mem­brane in­tegrity of the pre­hole state and the mas­sive mem­brane holes de­scribed here.

We hope so, too.

Ref­er­ence

Dewey JS, Savva CG, White RL, Vitha S, Holzen­burg A, Young R. (2010). Mi­cron-scale holes ter­mi­nate the phage in­fec­tion cy­cle. Proc. Natl. Acad. Sci. USA, 107 (5), 2219−2223 PMID 20080651