Rolling-Cir­cle Repli­ca­tion

Pic­tures Con­sid­ered #26

Note. Trans­fer­ring our more than 2,000 posts from one plat­form to an­other, Word­Press, is a cum­ber­some task. Thus, many of our posts still con­tain bro­ken links. We will fix this one by one. We ask for your pa­tience.

by Christoph

In the late 1960s, read­ers of struc­tural bi­ol­ogy pa­pers were fre­quently con­fronted with (and con­fused by) elec­tron mi­cro­graphs of heav­ily in­ter­twined DNA mol­e­cules that re­sem­bled a hope­lessly tan­gled ball of wool. They were re­ally hard to "read" even for ex­perts (see Pic­tures Con­sid­ered #6 X for an ex­am­ple), and pic­tures as the one con­sid­ered here were no­table ex­cep­tions (Fig. 1).

Fig­ure 1. Elec­tron mi­cro­graph of rolling-cir­cle rep­lication in­ter­me­di­ate of phage φX174. Con­tour length of the ds­DNA cir­cle (lower left): ~1.65 µm. Credit: David Dressler. Source: Wat­son JD. 1977. Mol­e­c­u­lar Bi­ol­ogy of the Gene. 3rd Ed. W.A. Ben­jamin, Menlo Park, CA

This unas­sum­ing mi­cro­graph of a dou­ble-stranded DNA cir­cle of bac­te­rio­phage ΦΧ174 (length 5386 bp) with its long sin­gle-stranded ex­ten­sion was part of the work that earned the au­thor a doc­tor­ate from Har­vard, a pa­per in Na­ture, and... halted his sci­en­tific ca­reer (tran­siently), as the ex­perts at that time wouldn't be­lieve that this was a true repli­ca­tion in­ter­me­di­ate, stem­ming from what every mol­e­c­u­lar bi­ol­ogy text­book cov­ers with a ded­i­cated chap­ter to­day: rolling-cir­cle repli­ca­tion. Back in the 60s, it was known that the sin­gle-stranded DNA (ss­DNA) in the ΦΧ174 virion is con­ver­ted to dou­ble-stranded DNA (ds­DNA) af­ter en­try into the host cell as the ini­tial step in its repli­ca­tion. It was as­sumed, as was dogma then, that the dou­ble-stranded cir­cle of unit genome length repli­cates in the theta-mode (θ) like bac­te­r­ial chro­mo­somes (see Pic­tures Con­sid­ered #2 and #23).

Fig­ure 2. DNA syn­the­sis ini­ti­ates us­ing the free 3' ‑OH end at the nick as a primer, and a rep­li­ca­tion fork pro­ceeds around the tem­plate. In the process, the newly syn­the­sized strand dis­places the old strand from the tem­plate. In the case of repli­ca­tion of the replica­tive form (RF) of sin­gle-stranded phage genomes and of plas­mids of Gram-pos­i­tive bac­te­ria, the dis­placed old strand is cleaved off af­ter one round of repli­ca­tion and is con­ver­ted into the cir­cu­lar, dou­ble-stranded form. In con­trast, in phage lambda repli­ca­tion, the repli­ca­tion fork pro­ceeds a num­ber of rev­o­lu­tions around the tem­plate with­out cleav­age of the dis­placed strand, and the dis­placed strand be­comes dou­ble-stranded as it is peeled off. Source

The true hall­marks of rolling-cir­cle repli­ca­tion can­not be seen in the elec­tron mi­cro­graph shown in Fig. 1 due to its lim­ited res­o­lu­tion (sic!). For one, ini­ti­a­tion of repli­ca­tion is not brought about by an ini­tia­tor pro­tein that binds to a repli­ca­tion ori­gin and tears the two DNA strands apart to ex­pose sin­gle-stranded DNA as en­try site(s) for he­li­cases, pri­mases, and poly­merases, as in theta-repli­ca­tion. In­stead, the 'ini­tia­tor' in this case has en­donu­cle­ase ac­tiv­ity and makes a "nick", a sin­gle-strand cut, in its dou­ble-stranded bind­ing re­gion that pro­vides the 3'-OH end for repli­ca­tion to start. There­fore, no (host) pri­mase is re­quired here (see Fig. 2). For most ss­DNA phages — their con­ver­sion to ds­DNA is worth a sep­a­rate post, and is there­fore ne­glected here — and small plas­mids that sport rolling-cir­cle repli­ca­tion, the ini­tia­tor pro­tein (REP) re­mains co­va­lently bound (via a ty­ro­sine residue) to the 5'-end of the nicked strand. This nicked strand is suc­ces­sively dis­placed by the (host) repli­some, and since REP re­mains at­tached to it, it will en­counter its orig­i­nal bin­ding site when the repli­some has come full cir­cle. The REP-bound strand is then re­leased as a com­plete, co­­va­­lently-closed sin­gle-stranded mol­e­cule, and REP is free to make the nick again for the next round. Thus, and this is its sec­ond hall­mark, rolling-cir­cle repli­ca­tion is dis­con­ti­nu­ous: the repli­ca­tion of lead­ing and lag­ging strand is un­cou­pled.

Rolling-cir­cle repli­ca­tion — or sigma-repli­ca­tion (σ) — is also an in­te­gral part of bac­te­rio­phage lambda de­ve­lop­ment in the lytic cy­cle: af­ter be­ing in­jected into the host cell, lin­ear dou­ble-stranded (ds) lambda DNA cir­cu­la­rizes via its co­he­sive ends (cos sites). Cir­cu­lar ds­DNA then repli­cates a few times in the theta-mode (θ) be­fore swit­ch­ing to the sigma-mode (σ). An aside: the­ nomen­cla­ture us­ing Greek let­ters sim­ply re­flects the shape of the mol­e­cules in elec­tron mi­cro­graphs. The ss­DNA stretch "peeled off" the cir­cle by dis­place­ment syn­the­sis, as shown schemat­i­cally in Fig. 2, is con­ver­ted to ds­DNA by lag­ging-strand DNA syn­the­sis. It's the lin­ear con­cate­meric (multi-genome) ds­DNA ob­tained by rolling-cir­cle repli­ca­tion that is then processed — by cleav­age at the cos sites — for pack­ag­ing  of lin­ear (sin­gle-genome) ds­DNA into pre­formed lambda phage heads. This said, you might as­sume that lambda repli­ca­tion is fully un­der­stood to­day. Not quite! De­spite decades of work of (to­day few) ded­i­cated lamb­dol­o­gists, the en­donu­cle­ase re­spon­si­ble for mak­ing the es­sen­tial sin­gle-strand cut (Fig. 2, "Nick") at the ori to pro­vide the 3'-OH end for repli­ca­tion to switch from θ‑mode to σ‑mode re­mains elu­sive. A fact Grze­gorz We­grzyn, an ex­pert in this mat­ter, con­firmed to me dur­ing prepa­ra­tion of this post, in May 2015.