A Vi­ral Pyra­mid Scheme

A mud­pot in the Lower Geyser Basin, Yel­low­stone Na­tional Park, USA. Source

by Merry Youle

In or­der to re­lease their newly as­sem­bled viri­ons, most viruses lyse the cells that have fed and housed them. This ly­sis is not a hap­haz­ard af­fair. Some phages, for ex­am­ple, em­ploy a holin-en­dolysin sys­tem to rup­ture their host's cell mem­brane and di­gest the cell wall at a pre­cisely con­trolled time. (For our ear­lier posts about this, click here and here.) Oth­ers in­stead in­ter­fere with cell wall syn­the­sis — the same strat­egy that we use with our β‑lactam an­tibi­otics. Now yet an­other com­pletely dif­fer­ent scheme has been dis­cov­ered in a virus that in­fects an aci­dophilic hy­per­ther­mophilic Cre­nar­chaeon.

Fig­ure 1. A typ­i­cal S. is­landi­cus cell is an ir­reg­u­larly-shaped coc­cus. Bar = 200 nm. Source

We have come to ex­pect nov­elty in the bizarre zoo of cre­nar­chaeal viruses; wit­ness the Acid­i­anus two-tailed virus that "grows" its two long tails af­ter leav­ing the host cell. Still we weren't pre­pared to find a novel ly­sis strat­egy among them. The ones that in­fect the ex­tremophilic Cre­nar­chaeota weren't even thought to be lytic. In­stead, it ap­peared that they resided long-term within their host in a car­rier state, con­tin­u­ously se­cret­ing viri­ons from the host cell with­out killing it (but also with­out in­te­grat­ing their genome into the host chro­mo­some as do the tem­per­ate phages such as λ). This be­nign strat­egy was, we thought, how the wily virus pro­tected it­self from rapid degra­da­tion by ex­treme en­vi­ron­ments. Not nec­es­sar­ily so.

In 2009, Ar­i­ane Bize, work­ing in David Prangishvili's lab, was in­ves­ti­gat­ing the life cy­cle of viruses in­fect­ing Sul­folobus is­landi­cus, a cre­nar­chaeon iso­lated from vol­canic springs (more ac­cu­rately, sol­fa­taric fields). Cul­tur­able S. is­landi­cus has be­come a model hy­per­ther­mophile, able to grow in a rich medium at 78 °C and pH 3.0, with a gen­er­a­tion time of about 13 h. Fur­ther­more, it is host to viruses from sev­eral di­verse fam­i­lies, in­clud­ing the Rudi­virus SIRV2 (Sul­folobus islandi­cus rod-shaped virus 2). Bize found that SIRV2 de­fied our ex­pec­ta­tions by be­ing a lytic virus. Not only can its free viri­ons sur­vive in such vol­canic en­vi­ron­ments, but they are al­most as sta­ble at 80 °C as are those of mesophilic phages at 37 °C.

Fig­ure 2. A rod-shaped, non-en­veloped Rudi­virus virion. The SIRV2 virion is 23 nm wide and 900 nm long — not count­ing the three 28 nm-long tail fibers at each end. Pack­aged in­side is the 35.5 kb lin­ear ds­DNA genome. Source

Ev­i­dence for ly­sis by SIRV2? If you add SIRV2 viri­ons to grow­ing S. is­landi­cus cells in cul­ture, prog­eny viri­ons be­gin to be re­leased about 8 h later, with peak re­lease fol­low­ing at 13 h. Burst size is ~150. Start­ing al­most im­me­di­ately upon in­fec­tion, the virus de­grades the host chro­mo­some. By 6 h, the num­ber of colony form­ing units (CFU) has dropped 1000-fold, but ac­tual host cell death, as de­tected us­ing the mem­brane po­ten­tial-sen­si­tive dye DiBAC₄ comes later, co­in­cid­ing with vi­ral re­lease.

Fig­ure 3. Elec­tron mi­cro­graphs of SIRV2-in­fected S. is­landicus cells 10 h post in­fec­tion. (B1) SEM. (B2) TEM. Ar­rows in­di­cate in­di­vid­ual pyra­mid-shaped pro­tru­sions (VAPs). Bar = 200 nm. Source

What's novel here is that if you look at a host cell 10 h post-in­fec­tion, you'll see mul­ti­ple pyra­mid-shaped pro­tru­sions on its sur­face (Fig­ure 3). The re­searchers dubbed these struc­tures VAPs (virus-as­so­ci­ated pyra­mids), an in­nocu­ous name that spec­i­fies nei­ther their ori­gin or their func­tion. What might be their role here? Think of each of them as a hol­low 7‑sided cone that pro­trudes from the cell and pen­e­trates both the cell mem­brane and the S‑layer. The in­te­rior of the VAP is con­tin­u­ous with the cy­to­plasm (see B2 in Fig­ure 3). When first formed they are small, each side of the polyg­o­nal base be­ing on the or­der of 15 nm. As the in­fec­tion pro­ceeds, the VAPs grow and this di­men­sion in­creases to as much as 180 nm. At the time of ly­sis, the larger VAPs open up at the tip, each face of the pyra­mid open­ing out­ward. This per­fo­rates the cell en­ve­lope and re­leases the ma­ture viri­ons. What re­mains is an empty ghost of a cell whose cell mem­brane and S‑layer ap­pear in­tact ex­cept for nu­mer­ous dis­crete per­fo­ra­tions where the VAPs had formed.

Fig­ure 4. Neg­a­tive con­trast elec­tron mi­cro­graphs of iso­lated VAPs. In­tact VAPs: (A) top view; (B) side view; © a sin­gle de­tached tri­an­gu­lar face. Open VAP: (D) top view. Bar = 100 nm. Source

This, of course, raises lots of ques­tions. Some an­swers about both the ori­gin and func­tion of these struc­tures can be found in a re­cent pa­per by Tessa Quax, also from the Prangishvili lab. Con­ve­niently, VAPs can be iso­lated for study as dis­crete struc­tures, free of the cell mem­brane or other cel­lu­lar con­stituents. How? Make an­ti­bod­ies against the pro­tein found in the VAPs; fuse the an­ti­bod­ies to mag­netic beads; mix the an­ti­body-coated beads with crude ex­tracts of cells col­lected 10 h post in­fec­tion. The VAPs ob­tained this way are mostly in­tact poly­he­dra com­posed of seven tri­an­gu­lar faces. Seven? Al­though not un­heard of, such sev­en­fold sym­me­try is ex­tremely rare in Na­ture.

Work­ing with these iso­lated VAPs the re­searchers un­cov­ered even more amaz­ing as­pects of the story. Us­ing SDS/PAGE they found that these novel struc­tures are com­posed of only one pro­tein, sub­se­quently iden­ti­fied as the ~10 kDa phage pro­tein, P98. Con­versely, us­ing im­mu­no­elec­tron mi­croscopy they lo­cal­ized the P98 pro­tein specif­i­cally in the VAPs, and found it uni­formly dis­trib­uted over their en­tire struc­ture (Fig­ure 5).

Fig­ure 5. Neg­a­tive con­trast elec­tron mi­cro­graphs and P98 im­muno­la­bel­ing of VAPs. (A and B). Pu­ri­fied VAPs shown from dif­fer­ent an­gles. Bar = 100 nm. (D) Thawed cryosec­tions through SIRV2-in­fected cells. Bar = 200 nm. Source

And, the gene en­cod­ing P98 is not only nec­es­sary to as­sem­ble VAPs, it is also suf­fi­cient! This was demon­strated el­e­gantly first by ex­press­ing the gene in an­other Sul­folobus species, S. aci­do­cal­dar­ius — a species that is re­sis­tant to SIRV2 in­fec­tion and does not carry ho­mologs of any SIRV2 pro­teins. EM ob­ser­va­tions of these P98-ex­press­ing cells showed nor­mal-ap­pear­ing VAPs on the cell mem­branes. Even more sur­pris­ing, E. coli cells ex­press­ing P98 con­struct VAPs on their cell mem­brane, the VAPs pen­e­trat­ing into the periplasm. Al­though these VAPs self-as­sem­bled in vivo in both cases, they were never seen to open and per­fo­rate the cell. Ap­par­ently other fac­tors made dur­ing SIRV2 in­fec­tion are re­quired to trig­ger open­ing of the VAPs, in­clud­ing the factor(s) that de­ter­mine the time of cell ly­sis.

Fig­ure 6. Thin sec­tions through E. coli 1051 ex­press­ing SIRV2-P98. Ar­rows in­di­cate VAPs. Bar = 200 nm. Source

This is as far as the story goes for now. Clearly, there is more to come, as ques­tions abound. How does SIRV2 con­trol when ly­sis will oc­cur? Do all the VAPs open si­mul­ta­ne­ously? How do the VAPs self-as­sem­ble and grow in vivo? Their re­la­tion­ship with the cell mem­brane is es­pe­cially in­trigu­ing given that the same pro­tein is able to as­sem­ble into pyra­mids within both bac­te­r­ial and ar­chaeal mem­branes. And what about the rest of the hy­per­ther­mophilic ar­chaeal viruses? This ly­sis mech­a­nism is not uni­ver­sal among them. Its spotty dis­tri­b­u­tion sug­gests move­ment by HGT, leav­ing us won­der­ing where else it might turn up.

The VAPs carry an­other mes­sage for us. Cur­rently, when one se­quences the genome of an ar­chaeal virus, more than 90% of the pre­dicted genes will be un­knowns, i.e., se­quences with no de­tectable sim­i­lar­ity to any known gene. Even for the bet­ter-stud­ied phage, 80% of the genes are un­knowns. This raises a big ques­tion. Are all these un­knowns truly novel, car­ry­ing out func­tions not ob­served pre­vi­ously? Or are they sim­ply highly di­ver­gent mem­bers of well-stud­ied gene fam­i­lies? Some think that the lat­ter is the case, cit­ing the long evo­lu­tion­ary his­tory of viruses and their higher mu­ta­tion rates. How­ever, the VAPs are proof that at least some novel func­tions are still lurk­ing out there, wait­ing to be dis­cov­ered. And the cre­nar­chaeal viruses are a good place to look. So grab your sam­pling equip­ment and head for your near­est hot spring, fu­ma­role, or mud pot. You will be re­warded by en­coun­ters of a strange kind.

Ref­er­ences

Bize A, Karls­son EA, Eke­fjärd K, Quax TE, Pina M, Pre­vost MC, Forterre P, Tenail­lon O, Bernan­der R, Prangishvili D. (2009). A unique virus re­lease mech­a­nism in the Ar­chaea. Proc Natl Acad Sci USA, 106 (27), 11306−11311. PMID 19549825

Quax TE, Lu­cas S, Reimann J, Pe­hau-Ar­naudet G, Pre­vost MC, Forterre P, Al­bers SV, Prangishvili D. (2011). Sim­ple and el­e­gant de­sign of a virion egress struc­ture in Ar­chaea. Proc Natl Acad Sci USA, 108 (8), 3354−3359. PMID 21282609