Replication initiation in vertebrates (Marie-Noëlle Prioleau)

In eukaryotes, the spatiotemporal program controls the positioning and the activation time of the origins of replication during the S phase of the cell cycle. We are. More recently, we are also trying to understand. Indeed, it is now clearly established that conflicts between these two machineries can be a source of genomic instability.

In order to get a general view of human genome replication, we are developing high-throughput analyses to map the origins of replication as well as their activation time (Figure 1).

Figure 1. Representative features of early and late replication domains

UCSC genome browser visualization of a 3Mb genomic region of the avian chromosome 1. Data were obtained in the DT40 avian cell line. Tracks of nascent strands (NS) enrichments in the four S-phase fractions, from early to late, are shown separately (S1 to S4). NS-enriched and depleted regions for each fraction are shown in red and blue, respectively. This data provides information on the temporal program of DNA replication. Single reads from nascent RNA-seq data and aligned SNS (spatial program of DNA replication) are reported and between tracks of annotated genes (RefSeq genes) and CpG Islands. The bottom track shows GC content (GC percent).

Collaboration with statisticians and bioinformaticians has allowed us to link these maps to genomic data on chromatin structure and gene expression. We also use an avian cell model (the DT40 cell line) which has the unique property of performing homologous recombination very efficiently. This powerful genetic model allows us to test very efficiently different hypotheses extracted from genomic analyses. We were able to show that G-quadruplexes play a key role in the definition of replication origins. We have just identified a complex motif that is found in half of the strong origins in humans, mice and chickens. Future projects will focus on understanding how this motif ensures the efficient recruitment of the replication machinery.

In a second line of research, we are analyzing how the origin signal carried by this complex motif interacts with the transcription process that takes place on the same substrate, DNA. These analyses will allow us to test a well-supported hypothesis that transcription is able to shift the starting points of replication and thus generally promote replication initiation in intergenic regions. Common Fragile Sites (CFSs) are recurrent sites of chromosomal rearrangements in cancers and some neurological diseases. They are found within large (> 300 kb) genes transcribed and replicated at the end of S phase. One hypothesis regarding their formation is that the lack of replication initiation events along these genes results in incomplete replication of these regions before mitosis. Transcription would remove the pre-replication complexes located within the bodies of these genes, thus inducing a depletion of replication origins. We seek to test the ability of RNA polymerase II to displace pre-replication complexes by inserting a strong or inducible promoter upstream of a highly efficient minimal model origin lacking transcriptional activity. This project will explore an important hypothesis on the formation of CFS. In the longer term, we will test the hypothesis that recurrent break sites observed during the proliferation of neuronal precursors have the characteristic properties of CFS. For this purpose, we will use human brain organoids as a model system.

The understanding of the duplication mode of eukaryotic genomes is essential. Indeed, replication not only ensures the maintenance of genome integrity, but also coordinates the establishment of expression programs during development.

Replication initiation in bacteria (Jean-Luc Ferat)

In bacteria, replication is initiated at a single origin by the initiator protein, DnaA. The ordered binding of DnaA at the origin leads to the local unwinding of DNA and the direct recruitment of replicative helicases on single-strand DNA. This step is highly regulated and assisted by a co-helicase, which synchronizes the loading of the replicative helicases, ensuring Bidirectional Replication Initiation (BRI). We are interested in understanding 1) how does the co-helicase DciA load and activate synchronously the replicative helicases during replication initiation and 2) how does a replication initiated in absence of DciA result in a pathological degradation of the whole chromosome. We developed a large spectrum of experimental approaches (cytological, genetic, genomic, bioinformatic) to address these two questions in the model organisms Vibrio cholerae and Escherichia coli.

In bacteria, replication is initiated at a single origin by the initiator protein, DnaA. The ordered binding of DnaA at the origin creates a local unwinding of the double helix that enables the recruitment of the replicative helicase, DnaB. Replisomes are then assembled on this nucleoprotein platform to trigger Bidirectional Replication Initiation. Contrary to eukaryotes, bacteria load directly their helicase on single strand DNA during replication initiation. This process is highly regulated and assisted by co-helicases in the three domains of life. Bacterial co-helicases are diverse. The first family that was studied, DnaC/I (DnaC in Escherichia coli and DnaI in Bacillus subtilis), is present in only a few orders of bacteria and results from the domestication of a phage gene (Brézellec et al., 2016). The main family of co-helicase is DciA. It was discovered only a few years ago although it is present in the vas majority of bacterial orders (Figure 1). We established that DciA, is the ancestral bacterial co-helicase (Brézellec et al., 2016, 2017).

Fig. 1 Distribution of domesticated dnaC and dciA in the bacterial domain. Distribution of the bacterial phyla investigated on a 16S rRNA-based circular phylogenetic tree (Methods). When the presence of dnaC/I or dciA within a phylum (bold) was not homogeneous, classes (underlined) or orders (normal) are specified. Numbers boxed refer to the number of species’ genomes (and not strains) containing domesticated dnaC/I (fuchsia), dciA (turquoise) or neither (black). fuchsia / turquoise circles: orders in which dciA and dnaC/I genes were identified. The order Thiotrichales is identified with a green/black circle because most species lack dciA and dnaC/I. ① Gammaproteobacteria, ② Firmicutes, ③ Aquificae, ④ Chloroflexi and ⑤ Tenericutes.

We conducted a biochemical investigation of DciA together with its cognate replicative helicase to compare its activity to that of DnaC of E. coli. This study revealed surprisingly that contrary to DnaC/I the loading of the replicative helicase is not dependent on DciA. In other words, the replicative helicase of dciA-sepcies can self-load on DNA. At most, DciA improves the loading efficiency (Figure 2) (Marsin et al., 2021).