Research

At each cell division, chromosomes should be duplicated and also maintain memory of the specific transcription programs that have been previously established in the embryo. Our objective is to understand how DNA replication can be integrated with transcriptional controls during development. We characterize DNA replication origins and analyze how they are set according to development and differentiation. We also characterize the DNA replication initiation complexes and analyze how epigenetic mechanisms control the organization of chromatin domains for replication.

We have shown that the localization of DNA replication origins is developmentally regulated in Xenopus. DNA replication origins are located with no specificity during early development, when genes are not yet expressed in the embryo (Hyrien and Méchali, 1993). During this phase, asymmetric AT rich sequences are favorized as initiation sites, although they do not increase the overall efficiency of DNA synthesis (Stanojcik et al, 2008).

However, when genes become competent for transcription, DNA replication origins are set at specific sites in the corresponding domains (Hyrien et al, 1995 ; Girard-Reydet et al, 2004). This correlation between the onset of transcription programs and a specific localization of DNA replication origins has been experimentally reproduced using plasmid DNA templates injected into Xenopus eggs and embryos (Danis et al, 2004). These DNA molecules replicate in a regulated manner, but with no site specificity, as for the endogenous DNA. However, if such DNA molecules are organized into mini chromosomes competent for transcription (Prioleau et al, 1994), a specific DNA replication origin is now found, close to the promoter region (Danis et al, 2004). 

This coupling between DNA replication and gene expression was also observed during the differentiation of pluripotent teratocarcinoma cells into neural cells. P19 mouse cells can be induced to differentiate by retinoic acid and the HoxB domain is specifically expressed during differentiation. We observed multiple DNA replication origins along the 120 Kbp HoxB domain when cells are in their pluripotent stage. Expression of the HoxB domain after retinoic acid induction requires the S phase (Fisher and Méchali, 2003), suggesting that the replication fork could license the domain for transcription. Induction of the HoxB domain results in the restriction of DNA replication origins in the domain (Gregoire et al, 2006), confirming our data in Xenopus.

If differentiation processes are associated with restriction in the positioning of DNA replication origins, dedifferentiation processes should have the reverse effect. This is what we found in experiments which mimic nuclear transfer experiments. Sperm nuclei introduced into Xenopus egg extracts replicate very efficiently whereas differentiated cell nuclei replicate poorly. However, if differentiated cell nuclei are first incubated in mitotic egg extracts, then induced in S phase, they replicate as efficiently as sperm nuclei (Lemaitre et al, 2005). Mitotic conditioning of differentiated cell nuclei allows the reprogramming of the inter origin spacing from 120 Kbp to 20-25 Kbp, allowing nuclei to be adapted for the rapid cell divisions occurring during the early development. This replicon remodelling involves a dramatic change in the chromosome structure and is DNA topoisomerase II dependant. It also involves a higher recruitment of ORC to DNA replication origins.

Single-molecule analysis of the inter-origin spacing by molecular combing of DNA. The DNA molecule is in red and DNA replication origins are in green. Replication origins are spaced at large intervals if differenciated nuclei are incubated in an S phase extract (top), but are reprogrammed at close intervals by incubation in a mitotic egg extract (bottom).

DNA Replication initiation complexes

A second objective of our laboratory is to characterize the replication initiation complex and understand how DNA replication origins are regulated by DNA replication origin licensing factors. During recent years, we have characterized three new factors involved in DNA replication : Cdt1, MCM8, and MCM9. 

Cdt1 is recruited to origins in an ORC dependent manner and allows the recruitment of the MCM2-7 helicase (Maiorano et al, 2000). Once the helicase is assembled at origins, Cdt1 is displaced and not required anymore for further stages of DNA synthesis (Maiorano et al, 2004). We also found that Cdt1 and geminin form a complex acting as an ON/OFF switch at replication origins (Lutzmann et al, 2006). This complex rapidly binds to chromatin during the pre RC assembly and activates DNA replication origins. Further geminin is then recruited at the origin, changing the Cdt1-geminin stoichiometry and turning the activity of this complex off. This negative regulation of Cdt1 by geminin is essential to activate DNA replication origins only once in each cell cycle. If this stoichiometry is altered, for example by increasing the level of Cdt1, re-replication occurs (Maiorano et al, 2005). 

We have also characterized two new MCM family members, involved in DNA replication. The first, MCM8, is acting at the replication fork (Maiorano et al, 2006). MCM8 is an ATP-dependent DNA helicase which binds to chromatin after the preRC has been assembled, and facilitates processive DNA synthesis during DNA chain elongation. MCM8 is found co-localized with RPA at replication foci. The second is MCM9 and is involved in the formation of the pre RC (Lutzmann et al, 2008). MCM9 is required for the loading of the MCM2-7 helicase at DNA replication origins. It forms a stable complex with Cdt1 and prevents an excess of geminin to be bound onto chromatin during the licensing reaction. Thus, Cdt1, together with its two opposing regulating factors MCM9 and geminin, forms a major platform regulating the pre-replication complex (preRC). 

The dissociation of replication complexes has also been analyzed at the G2-M transition. We found that RPA is hypophosphorylated during the whole S phase and is hyperphosphorylated on its 34 kDa subunit only at mitosis (Francon et al, 2004). Hypophosphorylation of RPA34 correlates with the disassembly of RPA from chromatin. Interestingly, the dissociation of RPA foci is critically required for proper chromosome assembly and segregation at mitosis (Cuvier et al, 2006). The ORC complex, in addition to its known role in the assembly of the replication initiation complex, is also necessary to recruit cdc2-kinase which phosphorylates RPA34 and allows the disassembly of replication foci before mitosis can occur (Cuvier et al, 2006). We also found that topoisomerase II couples termination of DNA replication with the clearing of the replication complexes at the end of S phase (Cuvier et al, 2008).

We are currently characterizing in more details and at a large scale DNA replication origins. We also wish to understand the epigenetic mechanisms leading to the formation of DNA replication initiation complexes in-vivo both in undifferentiated and differentiated cells. At the same time, we are more deeply characterizing proteins involved in the pre-RC using both in vitro systems derived form Xenopus eggs and cell cultures. We welcome enthusiastic post doctoral scientists who wish to share our interests in these areas.