The bacterial nucleoid: nature, dynamics and sister segregation

Abstract

Recent studies reveal that the bacterial nucleoid has a defined, self-adherent shape and an underlying longitudinal organization and comprises a viscoelastic matrix. Within this shape, mobility is enhanced by ATP-dependent processes and individual loci can undergo ballistic off-equilibrium movements. In Escherichia coli, two global dynamic nucleoid behaviors emerge pointing to nucleoid-wide accumulation and relief of internal stress. Sister segregation begins with local splitting of individual loci, which is delayed at origin, terminus and specialized interstitial snap regions. Globally, as studied in several systems, segregation is a multi-step process in which internal nucleoid state plays critical roles that involve both compaction and expansion. The origin and terminus regions undergo specialized programs partially driven by complex ATP burning mechanisms such as a ParAB Brownian ratchet and a septum-associated FtsK motor. These recent findings reveal strong, direct parallels among events in different systems and between bacterial nucleoids and mammalian chromosomes with respect to physical properties, internal organization and dynamic behaviors.

Figure 1

Bacterial nucleoids and their organization. (A, B) 3D images of E.coli G1 nucleoids illuminated with Hup1-mCherry (A; [3]) or GFP-FIS (B; [2]) illustrating curved ellipsoid shape with variable handedness. (C, D) Elongating E.coli nucleoids exhibit thin fingers and progressively protruding ellipsoid ends, implying a strong tendency for chromosomal material to coalesce into defined shapes; imaged as in (A) [3]. Blue arrows in (C) indicate periodic lengthening reflecting the effects of longitudinal density waves in a period of chromosome elongation. (E, F) Complex replicating E.coli shapes; note well-separated bundles (arrow in (F)), imaged as in (A); from [3]. (G) C. crescentus G1 nucleoid structure defined by chromosome capture analysis: dual longitudinal bundles corresponding to left and right repilchores with variable relationships giving curved ellipsoid shape with variable handedness as in (A. parS marks the origin of replication. From [4]. (H, I) Proposed structure of C. crescentus longitudinal bundles: radial array of plectoneme loops giving a bottle brush pattern. (H) Each plectoneme is composed of two DNA duplexes running in opposite directions with compaction as indicated in inset. (I) Plectonemes are packed into fibers that are separated by plectoneme-free regions. (details in Supplemental Figure S12 of ref. [5]). (J) Images of B.subtilis nucleoids labeled by incorporation of fluorescent nucleotides [6]. Green ball marks Spo0J-GFP that binds to origin(s) of replication.

Figure 2

Proposed analogy between post-initiation transitions in C. crescentus and E.coli. Top: parS (origin) dynamics in Caulobacter [48]. Bottom: T1 and T2 transitions of E.coli defined by origin and nucleoid dynamics [3,26,28]. A first transition separates the two origins at a significant distance. A second transition separates the two origins to distant positions. In both cases, the latter transition sets up the ori - ter - ori disposition that permits integration of terminus events with cell division.

Figure 3

Two global nucleoid dispositions. (A) Nucleoids can exist in two basic types of configurations. In the fold-back (Type I) configuration, origin and terminus regions are at the ends of the nucleoid and left and right replichores (arms) are parallel. In the linear (Type II) configuration, the nucleoid comprises a single linear object with left and right replichores flanking the origin and terminus material stretched from one end to the other. (B) The first description of Type I and Type II configurations and the transition from Type I to Type II preceding replication initiation in E.coli by Niki and colleagues [8]. This transition was proposed to occur by rotation. The Type II configuration was proposed to be a ring, rather than the linear form subsequently described [30,55, 56]. (C) Type I and Type II configuration patterns for three types of bacteria. C. crescentus remains in Type I throughout (discussion in [53]). B. subtilis cells are multi-nucleate and oscillate between Type I and Type II patterns [54]. For E.coli growing with a linear cell cycle (no overlapping rounds of DNA replication) a synthetic proposal is presented to explain patterns observed in two slightly different growth cycles: Condition 1 (from [3,8,26,28]) and Condition 2 (from [30,55, 56]) (see also Figure 4). The two cycles differ with respect to the phasing between the cell division and chromosome replication/segregation cycles. Type I and Type II configurations alternate analogously in both cases; the only difference is whether the transition from Type I to Type II occurs after or before division (Conditions 1 and 2, respectively). Note that in all cases shown, the primary outcome of the replication/segregation program is a Type I configuration (in bold).

Figure 4

Proposed synthetic view of E.coli nucleoid disposition dynamics in two different conditions with differing relationships between the cell division and chromosome replication/segregation cycles. (A) Summary of proposal. Images for Condition 1 are from live cell imaging of strains fluorescently-labeled at the origin (green) and terminus (red) regions with timing defined by cell length [56]. Data for Condition 2 are from synchronous cell populations examined for origin and terminus disposition by FISH and for nucleoid status by DAPI staining [28]. Individual dots are positions of the indicated loci in independent analyzed cells from each time point. These and other studies agree that a Type II configuration exists at the onset of replication (text). (B, C) For Condition 1, a mirror-symmetric Type I configuration is documented for the indicated intermediate stage by analysis of oriC and two right replichore loci, which occur in inverted orientation at this stage (Panel B; Panel A “evolving Type I”). This configuration is inferred to persist until the end of the cell division cycle for three reasons. First, origin and terminus positions are stable from the T2 transition onward (Panel A). Second, disposition of the origin near one end of the nucleoid in about-to-divide cells corresponds to a Type I transition, not a Type II transition where the origin would be located in the middle of the nucleoid (Panel C). Third, the Type I configuration by analysis of interstitial loci in another study of cells with the same Condition [8]. These patterns imply that, in Condition 1, a transition from Type I to Type II occurs immediately following cell division. In Condition 2, cells exhibit the same program of origin and terminus dynamics observed for Condition 1 except that, at the end of the replication/segregation cycle, the nucleoids switch from Type I to Type II prior to cell division. Analysis of interstitial loci [29,30] documents the Type II configuration and but may be less sensitive for detecting the Type I configuration which, in this Condition, represents a smaller fraction of cells. Thus, the only difference between the two programs would be the timing of cell division relative to the time of the switch from Type I to Type II (see also Figure 3 and text).