I) Realtime imaging of DNA damage induced protein redistribution
Understanding the genome surveillance remains incomplete without extending the ‘traditional’ testtube approaches by a direct analysis of the key underlying molecular processes in their natural environment, the nucleus of a living cell. Only an intact cell provides truly physiological conditions for the balanced enzymatic reactions, and allows to address key questions such as: What is the order of protein assembly on damaged DNA and/or the key effector structures; what are the temporal differences in recognition of various types of DNA lesions, and which of the signaling components serve as molecular ‘messengers’ to support intracellular communication between the focal DNA lesions and diverse nuclear subcompartments.
To address these questions, we have constructed an integrated imaging workstation allowing microlaserassisted generation of DNA damage in spatiallyrestricted nuclear compartments coupled with an immediate image acquisition. This technique enables interactive in vivo analysis of the most proximal stages of the genotoxic stress response in which some of the key cell fate decisions are being taken. We are using this technology to investigate recognition, signaling and repair of the most deleterious type of DNA lesions, double strand DNA breaks (DSBs), generated by a UVA microlaser illumination of DNA presensitized by halogenated thymidine analogs). The realtime experiments are based on genetic tagging of the key genome surveillance factors with the Green Fluorescent Protein (GFP) or its spectral variants, and on generating cell lines expressing low, physiological levels of such cytologically tractable proteins. An important aspect of our livecell imaging is processing and further analysis of the quantitative fluorescence data by mathematical modelling.
Specific projects in this area:
Assessment of temporal assembly of the DSB sensors and early transducers
Our previous work identified two distinct modes of protein recruitment to the sites of DNA damage. While one class of proteins (represented by the Nbs1 component of the MRN nuclease complex) becomes concentrated at and around DSBs through a dynamic exchange between the damaged sites and the closely surrounding nucleoplasm, the other class (represented by the Chk2 kinase) becomes only very transiently associated with DSB from where (after phosphorylation by ATM and other autophosphorylaion events) it rapidly distributes throughout the nucleus.
We are exploiting these findings and our ambition is to assign a defined spatio-temporal pattern to all currently known major DSB transducers and candidate sensors. We expect that new spatiotemporal modes will emerge during these studies and that these will be instrumental to predict whether a given protein functions largely in local DNA transactions (such as sensing and repairing the primary lesion), in processes ongoing within the DSB-flanking chromosomal microenvironment (such as chromatin remodelling, replication fork protection), or in disseminating the initial signal to the effectors outside the primary sites of DNA lesions (thereby coordinating processes such as regu-lation of replication origins, gene expression, chromatin compaction).
Given the vast number of genes involved in various aspects of genome integrity maintenance in yeast, and the overall high degree of phyllogenetic conservation of key biological mechanisms, it is obvious that there are many more human genes affecting the response to genotoxic insults yet to be discovered. In collaboration with EMBL, Heidelberg, we are performing a largescale survey based on the realtime imaging and aiming at identification of new human proteins involved cellular re-sponse to DNA damage.
Export of DNA damage-induced checkpoint signaling beyond the nucleus.
This line of research has been inspired by recent reports of Chk2 involvement in coupling DNA damage with centrosome integrity in lower eukaryotes (Drosophila melanogaster), and by our own discovery of a specific fraction of human Chk1 kinase specifically residing at interphase but not mitotic centrosomes. Together, these findings provide a precedent that the function of the key DNA damageactivated kinases is not restricted to the cell nucleus. To further explore these findings, we are trying to determine how are centrosomeassociated Chk1 (and perhaps also Chk2) regulated by various forms of genotoxic stress, and assess how these parameters impact on the timing of centrosome separation, mitotic entry, chromosomal ploidy, and genetic stability.
II) Post-translational modifications in response to genotoxic stress
The key prerequisite for dealing with the DNA lesions promptly is to spread the damage signal and coordinate the responses of DNA repair with cell cycle checkpoints and cell death pathways. These rapid cascades of events appear to be driven by diverse posttranslational modifications, most notably phosphorylations. Despite an overwhelming evidence for the central role ATM/ATR and Chk1/Chk2 kinases in DNA damage signaling, and the growing number of their substrates along diverse effector pathways, the exact role(s) of these phosphorylations and the way they change the properties of the downstream effectors is not well understood.
We are using isotopelabelling of cultured cells, phosphopeptide mapping, phosphoaminoacid analysis, and mass spectrometry (in collaboration with Dr. Julio Celis department) to uncover new functional means of regulation and coordination within the genome surveillance network of DNA repair, cell cycle checkpoints, or cell death, and to pinpoint the role of protein kinases and phosphatases in these processes. Moreover, collaborative links with other Scandinavian labs (e.g. Lars Rönnstrand, Lund; and Karl Henrik Heldin, Uppsala) allow us to use complementary methods such as Edman cycle sequencing of radio-labelled phosphopeptides.
Specific projects in this area:
Mapping and functional analysis of DSB-induced posttranslational modifications of checkpoint mediators
The importance of checkpoint mediators (Mdc1, 53BP1, BRCA1, TopBP1, claspin) has been underscored by the growing evidence that these large proteins function as ‘molecular matchmakers’ to assemble diverse enzymatic activities at and around the sites of DNA lesions. It appears that the majority of such interactions are phosphorylation dependent, and ATM/ATR-controlled. Indeed, most checkpoint mediators typically contain domains (such as the FHA or BRCT domains) recognizing phosphates on other proteins, and there is evidence that they all are phosphoproteins themselves (although only a few phosphorylation sites have been mapped).
We are systematically mapping the in vivo phosphoacceptor sites of the main checkpoint mediators. In parallel, these biochemical assays are complemented by realtime imaging of how checkpoint mediators (including their phospho-deficient versions) interact with the DBS-flanking chromatin regions.
III) Coordination of the DSB checkpoint signaling and repair pathways.
Like the DNA damage sensors, some components of the DSB repair machinery (such as the Rad52 epistasis group of proteins) undergo a rapid assembly followed by a dynamic exchange at the sites of DNA lesions. However, how the recruitment and assembly of the repair ‘factories’ communicate with, and are influenced by, the checkpoint signaling remains largely unknown.
Our studies in this area include detailed spatiotemporal analysis of how signaling and repair components interact with ionizing radiationinduced foci (IRIF) and/or lasergenerated DSB tracks. We are also interested in how do these parameters differ in distinct cell cycle stages and how do they correlate with the prevalent mode of DNA repair (non-homologous end joining in G1 versus homologous recombination in G2). We are also investigating how genetic ablation of the key components of the signaling module alters the kinetics of intra-nuclear redistribution of the repair module, and vice versa. To achieve these goals, we employ both cells derived from patients with various genome instability syndromes and the siRNA technology.
IV) Chromatin as an emerging target for the genotoxic stress response
Recent evidence indicates that modulation of chromatin topology and histone phosphorylation, acetylation, and methylation are critical for detection, signaling and repair of DNA lesions. We have contributed to this concept by discovering a new DSB-activated pathway initiated by ATM, propagated by Chk1, and culminating at phosphorylation and inhibition of the S-phase kinase Tlk1. As the only known substrate for Tlk1 is the histone chaperone Asf1, these findings indicate that the ATM/Chk1-controlled signaling is involved in important ‘epigenetic’ processes such as the timing and velocity of histone deposition to the nascent DNA strands.
In addition, in collaboration with Genevieve Almouzni group (Institute Curie) we have recently uncovered a novel role of human Asf1 histone chaperone in regulating the flux of S-phase histones during the replicational stress. Also here we intensively explore these findings further by trying to identify the complex molecular assembly that works with Asf1 to maintain a critical pool of histones, to support resumption of DNA replication during the recovery from the replicational stress.
Additional interest in chromatin biology stems from our recent discovery that the Mdc1 checkpoint mediator induces and/or stabilizes the higher order structure of chromatin in the vicinity of the primary DNA lesions. This appears to be essential to unmask certain histone modifications and thereby allow recruitment of other key proteins (such as 53BP1) involved in the DSB-induced genome surveillance program. We are currently investigating how Mdc1 and 53BP1 regulate local changes in chromatin architecture around the DNA breaks and how this contributes to maintain chromosomal stability in cells exposed to genotoxic stress.
V) Analysis of DNA damage in human tissues at different stages of tumour development and at various ages
One of the burning gaps in our research on DNA damage responses and its potential applications is the lack of suitable ‘markers’ sufficiently simple and robust to be used for analysis of human specimens. We have pioneered some approaches along this line (by introducing functional analysis of activatory phosphorylation of the Chk2 kinase directly on archival human biopsies, using meticulously characterized phosphospecific antibodies). Most recently, our analyses of human tumours derived from different tissues and from distinct stages of tumour development revealed that activated DNA damage response could be detected already in the early precursor lesions. These findings have important conceptual ramifications in identifying the DNA damage response machinery as a potent anti-cancer barrier activated early during the multistep tumorigenesis.
We are continuously optimizing immunohistochemical protocols for reliable detection of the key components of the genome surveillance network and we use these tools to analyze large panels of diverse human tumours for the integrity and/or aberrations of the DNA damage pathways. We also use these assays to analyze series of skin biopsies from donors of different ages, to search for signs of spontaneous DNA damage with age, to support or question the attractive hypothesis that ageing may reflect, at least in part, accumulation of DNA damage.
VI) Modulating the outcome of DNA damage responses
Experiments in this area are based on manipulating the different regulators and/or effectors in an attempt to direct the cellular response towards a desired outcome. We compare normal and transformed human cells, and apply small molecule inhibitors (for DNA damage kinases such as ATM and Chk1) before, or at various intervals after irradiation. The goal of these experiments is to establish whether such time and target-selective modulation of the DNA damage response network could help design more rational schedules for optimized radiation effects by selectively promoting survival and transient cell cycle arrest with repair (protect) versus cell death. These results could help design rational strategies to protect normal cells and selectively eliminate (sensitize) genetically unstable and/or transformed cells.
In addition, in a set of rather longterm projects, we use the functional assays described in the previous sections to assess how the severity of DNA damage affects cell cycle arrest, repair, survival versus cell death, and senescence as alternative outcomes of the global checkpoint response.
VII) Cell cycle as the determinant of the DNA damage response
One longterm interest in our laboratories involves elucidating the basic mechanisms of cell cycle progression in mammalian cells, especially how the cyclindependent kinases (CDKs) and phoshatases regulate the major cell cycle transitions (initiation of DNA replication, entry into and exit from mitosis, cytokinesis). In this regard, we have recently expanded our scope by a systematic effort to elucidate whether and how CDKs modulate the outcome of DNA damage response and/or replicational stress on various levels. To enter this area, we have been currently constructing cellular sys-tems where we can conditionally silence (either alone or in combination) the key mammalian CDKs and/or their regulatory subunits (cyclins) by RNA interference. We will use these tools to screen which of the DNA damage sensors, transducers and effectors are physiological CDK targets, and to elucidate how these phosphorylations impinge of the effectiveness of the genome surveillance program.