A Whiff of Repli­ca­tion – Atrib­ac­ter lam­i­na­tus

by Christoph

When mi­cro­bi­ol­o­gists de­scribe a pre­vi­ously un­known bac­terium, they are mainly in­ter­ested in its lifestyle, me­tab­o­lism, mor­phol­ogy and cell cy­cle (see part 1). For me, "get­ting to know" a new­bie al­ways in­volves – let's call it an ad­dic­tion – read­ing its genome se­quence and find­ing out how it ini­ti­ates the repli­ca­tion of its chromosome(s) (see here ️‍in STC for the Lil­liputians – for­merly CPR bac­te­ria, now Patescibac­te­ria), here for Plancto­mycetota, here for Bor­re­lia). This is par­tic­u­larly in­trigu­ing, of course, if it is the first char­ac­ter­ized bac­terium of an en­tire phy­lum, the Atrib­ac­te­ria (now: Atrib­ac­terota).

Fig­ure 1. Gene or­der in the oriC re­gion of Atrib­ac­ter lam­i­na­tus RT761 and Atrib­ac­ter sp. iso­late AMR_MDS _4709 (phy­lum Atrib­ac­terota), Sul­fu­ri­monas denitrifi­cans DSM 1251, and Bacil­lus sub­tilis 168. Gene sizes are to scale, and the di­rec­tion of their tran­scrip­tion is indi­cated by "<" for counter-clock­wise (ccw), or ">" for clock­wise (cw). Red boxes mark in­ter­genic re­gions con­tain­ing DnaA bind­ing sites (DnaA boxes). Or­ange boxes in­di­cate the DUE (DNA un­wind­ing el­e­ment, ref.) in oriC. Both At­ri­bacter oriCs are pre­dic­tions by the au­thor; the S. deni­trificans oriC was con­firmed ex­per­i­men­tally (ref.) as was B. sub­tilis oriC (ref.). Fig­ure by the au­thor. Fron­tispiece: Con­fo­cal-laser mi­croscopy show­ing the lo­cal­iza­tion of DNA and RNA within the in­tra­cy­to­plas­mic mem­brane struc­ture. DNA, RNA, and mem­brane lipids were stain­ed by Hoechst (blue), SYTO RNAs­e­lect (green) and FM4-64 (red), re­spec­tively. Out­lines of the cell from a are in­cluded in all pan­els. a Phase con­trast im­age. b–d Con­fo­cal-laser im­ages. Scale bars, 1 μm. Source

Could I pos­si­bly "read" the repli­ca­tion ori­gin, oriC, and the ini­tia­tor pro­tein, DnaA, from the genome se­quence of Atrib­ac­ter lam­i­na­tus RT761 and of Atrib­ac­ter sp. iso­late AMR MDS 4709 for a com­par­i­son? The dnaA genes are an­no­tated in both genome se­quences. The pro­tein se­quences share 95% iden­tity, point­ing to a close re­la­tion­ship on the (tax­o­nomic) fam­ily level, sim­i­lar to the 96% iden­tity of E. coli K‑12 MG1655 and Sal­mo­nella Ty­phimurium LT2 DnaA. An aside: the DnaA pro­teins of A. lam­i­na­tus RT761 and E. coli K‑12 MG1655 share only 42% iden­tity, which is low but in the typ­i­cal range for DnaA pro­teins of species belong­ing to dif­fer­ent phyla.

Ori­gin se­quences can­not sim­ply be "read" from genome se­quences be­cause there is no "oriC code" as for pro­tein-cod­ing se­quences and they are there­fore, with few ex­cep­tions, not in­cluded in the an­no­ta­tions of se­quenced genomes. But oriCs are of­ten found in the dnaA·dnaN in­ter­genic re­gion of bac­te­r­ial genomes and can be de­tected by con­served struc­tural el­e­ments: counter-clock­wise tran­scrip­tion of the left-flank­ing gene, a DNA-un­wind­ing el­e­ment (DUE), DnaA-trio mo­tifs, and mul­ti­ple DnaA bind­ing sites. *)

Both Atrib­ac­ters have an oriC struc­ture lo­cated in the dnaA·dnaN in­ter­genic re­gion and, in ad­di­tion, a clus­ter of DnaA-bind­ing sites lo­cated in the dnaA up­stream re­gion (Fig­ure 1). The over­all arrange­ment is iden­ti­cal to that of the oriC re­gions of bac­te­ria from dif­fer­ent phyla, Sul­fu­ri­monas den­i­tri­f­i­cans (Campy­lobac­terota, for­merly Ep­silon­pro­teobac­te­ria) and B. sub­tilis (Bacil­lota, for­merly Fir­mi­cutes). For B. sub­tilis, it is known that for full ori­gin func­tion­ing in vivo both the incA re­gion and oriC are nec­es­sary. Pre­sum­ably, DnaA monomers bound to each of the two re­gions in­ter­act and lead to DNA loop­ing, al­though the func­tion of this loop­ing is not un­der­stood. Whether also the Atrib­ac­ters have such a bi‑partite repli­ca­tion ori­gin or whether the DnaA-bind­ing sites up­stream of the dnaA gene con­tribute to the re­gulation of its ex­pres­sion would have to be stud­ied ex­per­i­men­tally. The way I have got­ten to know the Small Things, I would not be sur­prised if they use both op­tions, si­mul­ta­ne­ously or al­ter­nately.

Fig­ure 2. Par­tial DNA se­quence and struc­tural el­e­ments in the oriC re­gion of Atri­bacter lam­i­na­tus RT761, At­ribacter sp. iso­late AMR_MDS_4709, E. coli K‑12 MG16­55, and B. sub­tilis 168. Or­ange box: DUE (DNA un­wind­ing el­e­ment, ref.) pre­dicted us­ing the SIST soft­ware. Dots above (up­per strand) or be­low (lower strand) the se­quence mark the bases where Krause & Messer (1999) found un­wind­ing ex­per­i­men­tally (red dots: with SSB; pale red: with­out SSB). The se­quences were alig­ned on the DnaA box R1 (TGTGNAWAA, rev) in E. coli oriC (red box). DnaA boxes with a close match to the con­sen­sus se­quence TGTGNAWAA are shown in red, while those de­vi­at­ing from the con­sen­sus (<4 mismat­ches) are shown in pale red. DnaA boxes (TTWTNCACA, fwd) are shown in turquoise, re­laxed boxes in light tur­quoise. Dark blue: IHF-bind­ing site in E. coli oriC (ref.). Vi­o­let boxes and bold vi­o­let A residues: (NAN)x DnaA-trio mo­tifs to which DnaA binds to ss­DNA dur­ing un­winding ac­cord­ing to Richard­son et al. (2016). Fig­ure by the au­thor

A closer look at the Atrib­ac­ter oriC se­quences shows that key struc­tural el­e­ments are present and, de­spite non-sig­nif­i­cant ho­mol­ogy at the nu­cleotide level, are arranged sim­i­larly to E. coli oriC and B. sub­tilis oriC (Fig­ure 2). The DNA-un­wind­ing el­e­ment (DUE) is a ~30-bp short and mostly AT‑rich se­quence that re­acts to in­creas­ing neg­a­tive su­per­he­lic­ity by lo­cal­ized unwind­ing, that is, as­sum­ing a sin­gle-stranded con­for­ma­tion. This con­for­ma­tional switch has been shown ex­per­i­men­tally for E. coli and B. sub­tilis upon DnaA bind­ing to oriC, but can also be sim­u­lated with ap­pro­pri­ate soft­ware.

The ini­tia­tor pro­tein DnaA binds via its C‑terminal do­main 4 to dou­ble-stranded DNA at se­quences re­sem­bling the non‑palindromic "DnaA box" TTWTNCACA (re­verse com­ple­ment: TGTGNAWAA), and the DnaA box marked "R1" in Fig­ure 2 is lo­cated at a dis­tance of 1–2 he­lical turns from the DUE in all known oriCs. We had found long ago that in E. coli tin­ker­ing with DnaA box R1 in­evitably leads to in­ac­ti­va­tion of oriC. Lo­cated be­tween the DUE and DnaA box R1 is the DnaA-trio mo­tif (NAN)x to which oligomer­ized DnaA binds via its cen­tral do­main 3 when it is sin­gle-stranded dur­ing un­wind­ing as was re­cently shown for B. sub­tilis oriC by Heath Murray's lab at New­cas­tle Uni­ver­sity, UK. Both Atrib­ac­ter oriCs fit ex­tremely well into this scheme and an ex­per­i­men­tal con­fir­ma­tion of their ori­gin func­tion looks straight­for­ward (but needs to be done).

The ini­ti­a­tion phase of chro­mo­some repli­ca­tion in E. coli cul­mi­nates with the load­ing of the replica­tive he­li­case, DnaB, onto the sin­gle-stranded re­gion in the DUE in oriC, by phys­i­cal in­ter­ac­tion of DnaA with DnaB. Briefly, DnaB then un­winds DNA and re­cruits the pri­mase, DnaG, which in turn trig­gers the for­ma­tion of a com­plete repli­some. As far as ini­ti­a­tion is con­cerned, A. lam­i­na­tus as the first char­ac­ter­ized mem­ber of the phy­lum Atrib­ac­terota does not ap­pear to de­vi­ate from paths known for other bac­te­ria. This does not rule out smaller and larger sur­prises in the fur­ther phases of chro­mo­some repli­ca­tion. Just a small one here: the genome of A. lam­i­na­tus en­codes ho­mologs of DnaB and DnaG, but not of the E. coli he­li­case loader, DnaC. In­stead of DnaC, A. lam­i­na­tus may use as he­li­case loader DciA (DUF721) as most bac­te­ria do that do not have a dnaC gene. And most bac­te­ria do not have a dnaC gene, only En­ter­obac­ter­aceae (see here in STC).

 

Do you want to com­ment on this post? We would be happy about it! Please send us an email, or com­ment on Mastodon, or on Bluesky.

*) If one can­not de­tect an oriC struc­ture in the dnaA·dnaN in­ter­genic re­gion of a com­plete genome, one can make use of the GC-skew for its de­tec­tion. OriC is com­monly lo­cal­ized at a dis­tance of ~10 kb from the min­i­mum in­flec­tion point of the GC-skew in a ~100–500 bp-long in­ter­genic re­gion. For the com­plete genome of Atrib­ac­ter lam­i­na­tus RT761 [NZ_CP065383.1], I ob­tained a sec­ondary GC-skew min­i­mum at pos. 852932 us­ing the GenSkew web­server, which ac­counts for a dis­tance of 13.8 kb from the pre­dicted po­si­tion of oriC (pos. 839078..839268).