All living organisms are made of cells, within which genetic information is stored on long strands of deoxyribonucleic acid (DNA). The DNA encodes thousands of different genes and provides the blueprint for all of the structures and activities occurring within the cell. The building blocks of DNA are the four deoxyribonucleotides, dATP, dGTP, dTTP, and dCTP, which are collectively referred to as dNTPs. The key enzyme in the production of dNTPs is ribonucleotide reductase (RNR). In the budding yeast Saccharomyces cerevisiae, the concentrations of the individual dNTPs are not equal and it is primarily RNR that maintains this balance. Maintenance of the dNTP pool balance is critical for accurate DNA replication and DNA repair since elevated and/or imbalanced dNTP concentrations increase the mutation rate and can ultimately lead to genomic instability and cancer. In response to DNA damage, the overall dNTP concentration in S. cerevisiae increases. Cell survival rates increase as a result of the elevated concentration of dNTPs, but the cells also suffer from a concomitant increase in mutation rates. When the replication machinery encounters DNA damage that it cannot bypass, the replication fork stalls and recruits specialized translesion synthesis (TLS) polymerases that bypass the damage so that replication can continue. We hypothesized that elevated dNTP levels in response to DNA damage may allow the TLS polymerases to more efficiently bypass DNA damage. To explore this possibility, we deleted all known TLS polymerases in a yeast strain in which we could artificially increase the dNTP concentrations. Surprisingly, even though all TLS polymerases had been deleted, elevated dNTP concentrations led to increased cell survival after DNA damage. These results suggest that replicative DNA polymerases may be involved in the bypass of certain DNA lesions under conditions of elevated dNTPs. We confirmed this hypothesis in vitro by demonstrating that high dNTP concentrations result in an increased efficiency in the bypass of certain DNA lesions by DNA polymerase epsilon, a replicative DNA polymerase not normally associated with TLS activity. We asked ourselves if it would be possible to create yeast strains with imbalanced dNTP concentrations in vivo, and, if so, would these imbalances be recognized by the checkpoint control mechanisms in the cell. To address these questions, we focused on the highly conserved loop2 of the allosteric specificity site of yeast Rnr1p. We introduced several mutations into RNR1-loop2 that resulted in changes in the amino acid sequence of the protein. Each of the rnr1-loop2 mutation strains obtained had different levels of individual dNTPs relative to the others. Interestingly, all of the imbalanced dNTP concentrations led to increased mutation rates, but these mutagenic imbalances did not activate the S-phase checkpoint unless one or several dNTPs were present at concentrations that were too low to sustain DNA replication. We were able to use these mutant yeast strains to successfully correlate amino acid substitutions within loop2 of Rnr1p to specific ratios of dNTP concentrations in the cells. We also demonstrated that specific imbalances between the individual dNTP levels result in unique mutation spectra. These mutation spectra suggest that the mutagenesis that results from imbalanced dNTP pools is due to a decrease in fidelity of the replicative DNA polymerases at specific DNA sequences where they are more likely to make a mistake. The mutant rnr1-loop2 strains that we have created with defined dNTP pool imbalances will be of great value for in vivo studies of polymerase fidelity, translesion synthesis by specialized DNA polymerases, and lesion recognition by the DNA repair machinery.

Le rayonnement ultraviolet (UV) émis par le soleil et qui atteint la peau de chaque individu est composé majoritairement de photons UVA (λ de 315 à 400 nm), le reste (5 à 10 %) étant composé d’UVB les plus longs (λ de 300 à 315 nm), car les radiations de longueur d’onde 300nm, c’est-à-dire les plus toxiques en terme de santé humaine, sont absorbées par la couche d’ozone stratosphérique. Contrairement aux UVB, les radiations UVA sont faiblement absorbées par l’ADN et de fait, génèrent peu de dimères cyclobutaniques de pyrimidines. Néanmoins, un des problèmes majeurs posés par une exposition aux UVA tient à ce que ce rayonnement excite certains composés endogènes photosensibles, inducteurs de la production d’espèces réactives de l’oxygène (ROS) qui peuvent alors endommager les composants cellulaires tels que les lipides,les acides nucléiques et les protéines. De ce fait, si les UVB restent le facteur étiologique majeur contribuant à la cancérogenèse cutanée photoinduite, un rôle des UVA, via la production de ROS, semble également émerger. Des précédents travaux obtenus au laboratoire ont montré que le rayonnement UVA ralentit la réplication de l’ADN, indépendamment de l’activation des points de contrôle du cycle cellulaire. Les auteurs ont émis l’hypothèse que les UVA, via l’oxydation des protéines, pouvaient altérer la machinerie de réplication. Mon travail de thèse a donc consisté à tenter de préciser le mécanisme qui gouverne ce retard de la réplication de l’ADN induit par les UVA dans les cellules de mammifères.Pour étudier au niveau moléculaire les effets des UVA sur la réplication, nous avons tout d’abord mis en place et utilisé au laboratoire la technique du peignage moléculaire (DNA combing) qui permet de mesurer divers paramètres de la réplication. Ainsi, nous montrons que le rayonnement UVA inhibe immédiatement et transitoirement les vitesses de fourches alors que l’inhibition sur l’initiation des origines est plus prolongée. Dans le cadre d’une collaboration, nous montrons également que les radiations UVA induisent une diminution modeste et transitoire du pool de dNTPs intracellulaires. La complémentation en ribonucléosides ne semble pas suffisante pour restaurer une vélocité normale de fourches immédiatement après UVA, ni la réplication dans sa totalité. En parallèle, nous observons l’oxydation réversible de la sous-unité R1 de la ribonucléotide réductase impliquée dans la biosynthèse des dNTPs. Bien que cette oxydation ne puisse expliquer la baisse transitoire du pool de nucléotides après UVA, nous ne pouvons pas exclure que d’autres formes d’oxydation de la RNR puissent affecter son activité.La présence d’azide de sodium (NaN3) au cours de l’irradiation UVA prévient le retard réplicatif, limite l’oxydation de la sous-unité R1 et la diminution du pool de dNTPs, ce qui démontre que ce retard de réplication est totalement dépendant des ROS, principalement de l’oxygène singulet généré pendant l’irradiation.L’ensemble de nos résultats indiquent que les UVA affectent le processus de réplication en modifiant non seulement la vélocité des fourches mais également l’initiation des origines de réplication. Puisqu’une perturbation de la réplication est une cause majeure d’instabilité génétique, il reste à déterminer si, dans nos conditions expérimentales, les radiations UVAfavorisent cette instabilité. Enfin, nous pensons que la ou les cibles des ROS induites par les UVA sont essentiellement cytosoliques et que le mécanisme conduisant à l’inhibition de la réplication n’est pas spécifique de ces ROS mais pourrait s’observer en utilisant d’autres types de stress oxydant. / The solar UV radiation that reaches the earth’s surface is composed of 10 % UVB (280–320 nm) and 90 % UVA (320–400 nm) the main toxic radiations (wavelengths below 300 nm) being blocked by the stratospheric ozone. Unlike UVB, the UVA component of solar radiation is weakly absorbed by DNA. Nevertheless, one of major problems due to UVA exposure is the production of reactive oxygen species (ROS) through the interaction with endogenous and exogenous chromophores. These ROS cause damage to DNA, lipids and proteins. Even if UVB remains the major etiological factor known to be implicated in photoinduced cutaneous carcinogenesis, a novel role for UVA via the production of ROS seems to emerge. In our lab, previous works have provided evidence that exposure of mammalian cells to UVA-induced ROS led to delayed S-phase and reduced DNA synthesis, by a yet unknown process, which does not require a functional DNA damage checkpoint response, despite ATM-, ATR-, p38-dependent pathways activation. The authors proposed that inhibition of DNA replication is due to impaired replication fork progression and/or origins activation, as a consequence of UVA-induced oxidative damage to proteins rather than to DNA. The project for my PhD thesis is to better understand the mechanism underlying this UVA-induced slowdown of DNA replication in human cells.To study at the molecular level the effects of UVA on DNA replication, we used the DNA combing methodology. This technique allows measurement of the fork velocity and of the origins density. We show that UVA-induced ROS inhibit immediately after irradiation, but transiently, the progression of replication forks, while the inhibition on the initiation of originslasts longer. By HPLC-MS, we show that UVA radiation induces a moderate and transient decrease of the level of each intracellular dNTP. The supply of ribonucleosides doesn’t seem to be sufficient to restore neither a normal forks velocity immediately post-UVA nor the overall slowdown of DNA replication. In addition, we observe a reversible oxidation of the subunit R1 of ribonucleotide reductase, an enzyme which is involved in dNTPs biosynthesis. This oxidation cannot explain the transient reduction of dNTPs pool after UVA exposure, but other types of RNR oxidative modification could affect its activity. During UVA irradiation, the presence of the antioxidant sodium azide (NaN3) prevents the delay of DNA replication, limits the oxidation of the subunit R1 and the decrease of dNTPs pool. These results strongly suggest that the slowdown of DNA replication totally depends on ROS, in particular on singlet oxygen production induced by UVA.Altogether, our data indicate that UVA irradiation affects the process of DNA replication by modifying the forks velocity and the activation of origins. As DNA replication impairment is a major cause of genetic instability, it is of importance to determine if UVA irradiation leads to this instability in our experimental conditions. Finally, we suspect that the target of UVAinduced ROS is essentially cytosolic and that the mechanism driving the inhibition of replication is not specific of UVA-induced ROS, but could be also observed with other types of oxidative stress.