Graduation day

Normal mitotic cells duplicate their genome once each cell cycle to ensure its correct transmission. The biochemical process leading to genome duplication has been highly conserved during evolution and can be described by the well-established replicon model (Jacob and Brenner 1963; Newlon and Theis 1993; Gilbert 2004). According to this model, sequence elements (replicators) genetically determine where DNA replication initiates by interacting with trans-acting regulatory factors (initiators, such as the ORC and the MCM complexes) that are cell cycle regulated (Ohta, Tatsumi et al. 2003). Despite the early successes in the identification of microbial eukaryotic origins (Aladjem and Fanning 2004; Gilbert 2004), the available methods for origin identification in mammalian genome and the main results so far generated by their application are sometimes controversial and have actually led to the detailed characterization of just few origins (S. Otha 2003). Only recently the application of genome wide techniques for the hybridization to tiling micro-arrays of neo-synthesized DNA has allowed the identification of a larger number of replication origins, but only in a small portion of the human genome (Lucas, Palakodeti et al. 2007; Cadoret, Meisch et al. 2008). To better understand the regulation of DNA replication, I developed a new method to map replication origins on the basis of their chromatin structure. In this work, I first isolated origin-rich DNA and then it was hybridized onto a chromosome 19 microarray: 217 replication origins were identified, characterized by an open chromatin structure. Among them, I selected 101 replication origins (with a validation rate of 95%), all containing a CCAAT box motif and bound by the NFY transcription factor. 77% of the newly identified origins were located near transcription start sites of expressed genes. I finally used massive sequencing of the same origin-rich DNA and identified 1,846 putative replication origins along the entire human genome. Most of them were located close to gene TSSs, confirming the likelihood of a correlation with gene transcription.

Kinetochores are highly conserved proteinacious structures that assemble on centromeresfollowing DNA replication. Kinetochores attach sister chromatids to the mitotic spindle and orchestrate chromosome segregation. An erroneous kinetochore can lead to aneuploidy, a hallmark of cancer. An amenable organism to study kinetochore assembly and activity is the budding yeast Saccharomyces cerevisiae. Cnn1p, a non-essential kinetochore protein of S. cerevisiae, was recently co-purified with conserved kinetochore subunit Nnf1p. This project aims to genetically, biochemically and functionally analyse Cnn1p to understand its role in chromosome segregation. We determined that Cnn1p localises at low levels only to centromeres in an Ndc10p and Ndc80p dependent fashion. Cnn1p affects chromosome segregation as its overexpression causes chromosome loss. A strong interaction of Cnn1p with members of the Ndc80 complex was shown via yeast two-hybrid analysis and co-purifications, potentially implicating the Ndc80 complex as a direct interaction partner of Cnn1p. The possibility that Cnn1p might be a checkpoint protein was excluded by performing a viability assay in nocodazole. Cnn1p strongly genetically interacts with essential kinetochore proteins Nnf1p, Okp1p, Spc105p and Dam1p. Deleting CNN1 increases mono-orientation of nnf1-17 mutants at semi-permissive temperature, yet after treating cells with and releasing them from nocodazole this is not the case, indicating that Cnn1p may be involved in kinetochore assembly and/or integrity rather than in microtubule capture per se. This function is further supported by the fact that Cnn1p is phosphorylated by Clb5p-Cdk1p in S-phase (Loog & Morgan 2005) when yeast kinetochores assemble. The function of Cnn1p is therefore likely in establishing and/or maintaining kinetochore integrity.

Recent findings suggest that the increased tumor suppression activity of p53 can promote aging (Rodier et al., 2007; Serrano and Blasco, 2007). P53 mutant mice characterized by chronical p53 activation (p44-Tg and p53m/+ model) exhibit increased cancer resistance but a shortened lifespan in association with early aging-associated phenotypes (Maier et al., 2004; Tyner et al., 2002). The lifespan determinant p66Shc was the first mammalian gene whose mutation was demonstrated to extend lifespan by conferring resistance to oxidative stress-induced apoptosis without increase in cancer risk (Migliaccio et al., 1999; Trinei et al., 2002). Ideed, p66Shc is a mitochondrial red-ox enzyme that, in response to a variety of oxidative stresses, generates reactive oxygen species (ROS) as mediators of apoptosis (Giorgio et al., 2005). Indirect evidences suggest that p53 is implicated in the p66Shc apoptotic-signalling pathway and that p66Shc is an oxidative-stress downstream target of activated p53 (Trinei et al., 2002). Since the p53-p66Shc pathway is specifically involved in the propagation of pro-apoptotic oxidative signals (p53-dependent apoptosis in response to gamma-irradiation or adryamicin proceeds normally in p66ShcKO cells), our working hypothesis is that distinct p53-pathways might regulate tumour suppression and aging and p66Shc might be a selective downstream-target of p53 not involved in tumor suppression (Trinei et al., 2002). To further characterize the p53-p66Shc-ROS-signalling pathway we investigated whether the p53-p66Shc-ROS pathway, besides apoptosis, regulates a specific transcriptional program involved in physiological aging. In summary I found that: i) p53 and p66Shc repress a mitotic signature of genes in H₂O₂ treated MEFs as well as physiologically in thymus; ii) the up-regulation of the mitotic signature genes in p66ShcKO thymus correlates with a retarded aging-associated involution and senescence of the organ; iii) the p66Shc deletion in the progeric p44-Tg mouse model rescues the boosted repression of the mitotic signature as well as the accelerated thymus involution; iv) p66Shc might regulate the transcriptional-inhibitory function of p53 on the mitotic signature genes through a ROS-dependent effect on the life span determinant Sirt1.

The TOCA (Transducer of Cdc42 dependent actin assembly) family of F-BAR (Bin, Amphiphysin, Rvs) -containing proteins binds to and remodels lipid bilayers via their conserved F-BAR domains, and regulates actin dynamics via their N-WASP (Neuronal Wiskott–Aldrich syndrome protein) binding SH3 (Src homology 3) domains. Thus, these proteins are predicted to play a pivotal role in coordinating membrane traffic with actin dynamics during cell migration and tissue morphogenesis. By combining genetic analysis in Caenorhabditis elegans with cellular biochemical experiments in mammalian cells, we showed that: i) Loss of CeTOCA proteins reduced the efficiency of Clathrin-mediated endocytosis (CME) in oocytes. Genetic interference with CeTOCAs interacting proteins WSP-1 (WASp homolog) and WVE-1 (WAVE (WASP-family verprolin) homolog), and other components of the WVE-1 complex, produced a similar effect. Oocyte endocytosis defects correlated well with reduced egg production in these mutants. ii) CeTOCA proteins localize to cell-cell junctions and are required for proper embryonic morphogenesis, to position hypodermal cells and to organize junctional actin and the junction-associated protein AJM-1 (Adherens junction marker 1). iii) Double mutant analysis indicated that the toca genes act in the same pathway as the nematode homologue of N-WASP/WASP, wsp-1. Furthermore, mammalian TOCA-1 and C.elegans CeTOCAs physically associated with N-WASP and WSP-1 directly, or WAVE2 indirectly via ABI-1 (Abl-interactor 1). Thus, we propose that TOCA proteins control tissues morphogenesis by coordinating Clathrin-dependent membrane trafficking with WAVE and N-WASP-dependent actin-dynamics.

Giuliani C, Troglio F, Bai Z, Patel FB, Zucconi A, Malabarba MG, Disanza A, Stradal TB, Cassata G, Confalonieri S, Hardin JD, Soto MC, Grant BD, Scita G "Requirements for F-BAR proteins TOCA-1 and TOCA-2 in actin dynamics and membrane trafficking during Caenorhabditis elegans oocyte growth and embryonic epidermal morphogenesis". PLoS Genet. 2009 Oct;5(10):e1000675

In Saccharomyces cerevisiae the conserved and essential Rio1p kinase regulates 20S rRNA processing during ribosome synthesis. However, shutting-off Rio1p expression leads to an accumulation of the Rio1p-depleted population in G1 and in metaphase, suggesting an involvement of Rio1p in mitosis. In Aspergillus nidulans growth of a sudD (rio1) temperature-sensitive mutant was rescued by overexpressing SUDA (SMC3), which encodes cohesin subunit Smc3. These observations and the recent identification of Rio1p, kinetochore proteins, and cohesin components in immunopurifications of kinetochore proteins Mtw1p and Nnf1p (De Wulf lab), further indicated a possible involvement of Rio1p in chromosome segregation by acting at kinetochore or cohesin level. The aim of my project was to analyse the functional involvement of the kinase at this cell-cycle stage. We found that Rio1p localises both to centromeres and cohesin binding sites along the chromosome arms. Epistatic analyses performed both with a strain reduced in Rio1p function and with a strain containing enhanced Rio1p levels, revealed strong, positive interactions between Rio1p and some kinetochore complexes and with most components of the cohesin pathway, except for Esp1p with which it negatively interacted. To functionally dissect the involvement of Rio1p in the regulation of budding yeast mitosis, we constructed a yeast strain in which two TEV protease cleavage sites were introduced in the Rio1p protein sequence. The strain was then transformed with a high-copy plasmid expressing TEV protease from the galactose-inducible GAL1 promoter. Cells in which Rio1p was conditionally cleaved showed a 15 minute delay in G1 and a 30-40 minute delay in metpahase. The later was not the result of spindle assembly checkpoint response. A retarded degradation of Pds1p-3HA and Scc1p-3HA confirmed the metaphase delay and points to a delayed cleavage of cohesin complex. Taken together our genetic and cell cycle analyses propose a hitherto unknown role of Rio1p in regulating mitosis and the correct timing of sister chromatid separation.

The aim of this thesis is to investigate two important networks involved in the cell cycle progression: the spindle assembly checkpoint (SAC), and the regulation of the phosphatase Cdc14. The SAC is active at the metaphase-to-anaphase and its target is APCCdc20, which triggers anaphase onset via Securin and Cyclin B degradation. By inhibiting Cdc20, the SAC delays sister chromatids separation until the very last of them attached to the spindle. First, we focused our attention on the interaction between Cdc20 and Mad2, a key component of the SAC. Studying this interaction in vitro, we found that the basal rate of Mad2 binding to Cdc20 is slow, and Mad2 dimerization accelerates it via a catalytic step. Interestingly, we showed that catalysis does not modify the equilibrium of the Mad2:Cdc20 interaction, and suggest that the release of Mad2 from Cdc20 is an energy-driven process. Second, starting from the in vitro results, we developed a mathematical model describing SAC activation and disengagement. The model reproduced the main systems level proprieties of the SAC: a sharp and irreversible threshold set at one unattached kinetochore, robustness to variations in proteins concentration, and fast dynamics of activation and inactivation. Lastly, we studied the regulation of Cdc14 in budding yeast where the phosphatase is known to have a pivotal role in mitotic exit. During a regular cell cycle, Cdc14 is kept sequestered and inactive in the nucleolus. We clarified that its release in anaphase requires Cdc5 and either one between MEN and CDKs kinase. Once activated, Cdc14 triggers a negative feedback loop, which culminates in Cdc5 degradation by APCCdh1, and thus in Cdc14 re-sequestration in the nucleolus. We showed that, in the presence of high levels of mitotic Cyclins, the negative feedback loop is blocked in an oscillatory regime where we could observe periodic cycles of Cdc14 release and sequestration. Both the SAC and Cdc14 contribute to induce the irreversible switch from a mitotic to a G1 state of the cell cycle via the inactivation of mitotic CDKs. At the metaphase-to-anaphase transition, the SAC assures that the APCCdc20 starts Cyclin B degradation in coordination with sister chromatids separation. Soon after, Cdc14 allows the completion of mitotic Cyclins degradation and the dephosphorylation of CDKs substrates.

Simonetta M*, Manzoni R*, Mosca R, Mapelli M, Massimiliano L, Vink M, Novak B, Musacchio A, Ciliberto A. “The influence of catalysis on Mad2 activation dynamics”. PLoS Biol. 2009 Jan 13;7(1):e10. *equal contribution

Homeodomain proteins constitute a large class of eukaryotic DNA-binding proteins that regulate transcription of a broad range of developmentally important genes. These proteins share a 60 amino acid DNA-binding domain which has been conserved in sequence, structure and mechanism of DNA-binding. While monomeric homeodomain proteins exhibit a limited ability to discriminate between different DNA sequences, their specificity is significantly enhanced through the cooperative binding to DNA with other DNA binding partners. Pbx1 (pre-B-cell leukemia homeobox 1) and Prep1 (Pbx-regulating protein 1), both belonging to the TALE (three amino acids loop extension) family of homeodomain proteins, forms a strong and stable complex with each other, independent of DNA binding. Pbx1 contains a nuclear localisation signal and carries Prep1 into the nucleus. Prep1 and Pbx1 form trimeric complexes with HoxB1 on target enhancers. HoxB1 is known to play an important role in development. The aim of this project was to characterise the structural and functional properties of the homeodomain proteins Prep1 and Pbx1 and their interaction with DNA in order to reach an understanding of the structural basis underlying the functional DNA target specificity. In a broader perspective, the determination of these structures will be valuable in the characterisation and understanding of the functions of this class of transcription factors during processes of development and cancer. I have therefore set up methods and conditions to purify crystallisable Prep1/Pbx1 and Prep1/Pbx1/DNA complexes. I have optimised protein constructs, expression conditions, purification methods and characterized the DNA binding in order to select the optimal DNA oligonucleotides for co-crystallisation and performed crystallisation screenings of a range of DNA oligonucleotides with Prep1and Pbx1.

Villaescusa JC, Buratti C, Penkov D, Mathiasen L, Planagumà J, Ferretti E, Blasi F. "Cytoplasmic Prep1 interacts with 4EHP inhibiting Hoxb4 translation". PLoS One. 2009;4(4):e5213

Inflammation is a response of a tissue to injury, often caused by invading pathogens. Inflammation is considered one of the most important factors contributing to tumorigenesis and tumor progression. Here we have shown that both TLR4 and MyD88 are essential for DMBA-croton oil induced mouse skin tumorigenesis. By contrast, TLR2 and TLR9 do not seem to contribute to skin tumorigenesis. Absence of TLR4 or MyD88 leads to reduced inflammation after croton oil/TPA treatment, suggesting that inflammation plays an important role in TLR4/MyD88 mediated tumor development. Both bone marrow derived and radio-resistant cells are required for carcinogenesis as bone marrow chimeras where TLR4 is missing in either one cell types do not develop carcinomas. Eliminating skin-colonizing bacteria with antiseptics or blocking bacterial lipopolysaccharide (LPS, a ligand for TLR4) in the skin does not reduce inflammation after croton oil/TPA treatment, suggesting that skin-colonizing bacteria are not involved in TLR4-dependent tumorigenesis. In contrast, blocking of high mobility group box-1 protein (HMGB1), one of the endogenous ligands for TLR4, inhibits the recruitment of inflammatory cells and croton oil/TPA-induced inflammation, which shows that TLR4 signaling is triggered by HMGB1 protein released in situ. These results suggest that the initial release of HMGB-1 triggers an inflammatory response that is dependent on TLR4 and leads to tumor development.

Avogadri F, Mittal D, Saccheri F, Sarrafiore M, Ciocca M, Larghi P, Orecchia R, Rescigno M. "Intra-tumoral Salmonella typhimurium induces a systemic anti-tumor immune response that is directed by low-dose radiation to treat distal disease". Eur J Immunol. 2008 Jul;38(7):1937-47.

Acute promyelocytic leukaemia (APL) occurs as a consequence of a chromosomal translocation involving the retinoic acid receptor alpha (RARα) and, in most of the cases, the promyelocytic leukemia protein (PML). Other proteins, however, were found translocated near to the receptor in APL and, strikingly, all the identified RARα partners bear a self-association domain that has been shown to be responsible for altered transcriptional properties of the fusion protein. It is known the dominant negative behaviour of PML-RARα on both PML and RARα coming from the remaining wild type alleles. Indeed, upon the expression of the chimeric oncogene, PML-RARα homo-oligomers are formed, which constitutively repress the transcription from the Retinoic Acid Responsive Elements (RAREs), together with the deregulation of PML protein, due to NBs delocalization. In the present study we analyze more in details the contribution of the PML oligomerization domain (a Coiled-Coil region), and of the recruitment of the Corepressor complexes to leukemogenesis. We used two chimeric proteins containing alternatively the homologous PML Coiled-Coil domain (CC-RAR) or a heterologous oligomerization domain derived from the yeast GCN4 transcription factor (GCN4-RAR). We evaluated their transforming potential in hematopoietic progenitors and in transduction/transplantation experiments in wild type mice. We thus demonstrated that, despite the transforming potential of both the oligomerization constructs, only the presence of the PML Coiled-Coil domain triggers RARα to exert its full leukemogenic potential. Additionally, we found that the PML-RAR/AHT mutant, unable to recruit the N-CoR/HDAC complex, and thus impaired in blocking myeloid differentiation, is otherwise able to trigger an APL-like leukemia in mice. Therefore we proved that other mechanisms of transcriptional repression, beyond HDACs recruitment, are involved in the transcriptional inactivation of RARα target genes. In conclusion, the present structure-function study of the chimeric oncogene PML-RARα has expanded our comprehension of the structural determinants of PML-RARα oncogenic action, and is beginning to unravel alternative views on the mechanism of action of the fusion protein, with respect to the actual knowledge regarding both the PML and RAR moieties.

Villa R, Pasini D, Gutierrez A, Morey L, Occhionorelli M, Viré E, Nomdedeu JF, Jenuwein T, Pelicci PG, Minucci S, Fuks F, Helin K, Di Croce L. "Role of the polycomb repressive complex 2 in acute promyelocytic leukemia". Cancer Cell. 2007 Jun;11(6):513-25.

Eps15 and Eps15R are homologous proteins involved in clathrin and non-clathrin mediated endocytosis of RTKs. I aimed to understand the physiological role of these proteins taking advantage of knockout (KO) mice. Eps15KO mice were viable and had an increased number of red blood cells (RBCs), with decreased size and mean corpuscular hemoglobin. Since Eps15KO RBCs lost transferrin receptor (TfR) protein expression faster than wild type (WT) once, lack of Eps15 might impair TfR endocytosis and/or trafficking. Eps15KO mice showed also an increased number of splenic marginal zone (MZ) B cells. In competitive repopulation of the immune system, Eps15KO cells have an advantage over WT once in producing MZ B cells and thymic T cells, but a disadvantage to form B1 B cells. Since the development of these populations is regulated by Notch signaling, it is possible that in the immune system Eps15 is involved in the Notch pathway. Eps15RKO mice were postnatal lethal and lethality was not explained but, based on the function of Eps15/R orthologue in lower organisms and of their mammalian interactors, a role for Eps15R in synaptic vesicle recycling is compatible with the observed lethality. Combined lost of Eps15/R let to an embryonic lethal phenotype similar to that of Numb, Notch and epsin1-2 KO mice. Since Eps15/R interact with Numb and Epsins and that endocytosis and Notch signaling are linked mechanisms, it is possible that Eps15/R interact with Notch pathway during organogenesis. This project pointed out the possibility that Eps15/R are involved in different physiological processes that require them to act in non-redundant and redundant cellular processes. Further experiments will be necessary to elucidate the exact nature of these cellular processes.

Offenhäuser N, Castelletti D, Mapelli L, Soppo BE, Regondi MC, Rossi P, D'Angelo E, Frassoni C, Amadeo A, Tocchetti A, Pozzi B, Disanza A, Guarnieri D, Betsholtz C, Scita G, Heberlein U, Di Fiore PP. "Increased ethanol resistance and consumption in Eps8 knockout mice correlates with altered actin dynamics". Cell. 2006 Oct 6;127(1):213-26.

Starting from two datasets of cancer related genes, we have analyzed genomic and network properties of ~600 genes mutated in different cancer types. Genes mutated in cancer are prevalently singletons at the genomic level and encode for central hubs of highly interconnected modules within the protein–protein interaction networks (PIN). We have compared the duplicability rate and the network topology of cancer genes with the rest of human genes, showing that genes related to cancer progression have particular genomic and network properties. According to these results, we depict cancer genes as fragile components of the human gene repertoire, sensitive to dosage modification. In contrast to what has been observed in yeast, where all essential genes are prevalently singletons that encode for highly interconnected hubs, in the human PIN the nodes with similar characteristics are rare and probably enriched in candidate cancer genes. We have then extracted from the human protein interaction network genes with similar properties, which may be involved in cancer progression.

Rambaldi D, Ciccarelli FD. "FancyGene: dynamic visualization of gene structures and protein domain architectures on genomic loci". Bioinformatics. 2009 Sep 1;25(17):2281-2.
Rambaldi D, Giorgi FM, Capuani F, Ciliberto A, Ciccarelli FD. "Low duplicability and network fragility of cancer genes". Trends Genet. 2008 Sep;24(9):427-30. Review.
Fumasoni I, Meani N, Rambaldi D, Scafetta G, Alcalay M, Ciccarelli FD. "Family expansion and gene rearrangements contributed to the functional specialization of PRDM genes in vertebrates". BMC Evol Biol. 2007 Oct 4;7:187.

Histone deacetylases (HDACs) are members of an ancient enzyme family, that catalyze the removal of acetyl groups from lysine residues in both histone and non-histone proteins. Among the 18 HDACs encoded in the mammalian genome, many studies suggest that HDAC1, HDAC2 and HDAC3 are critically involved in cell proliferation, but it is still unclear their individual contribution. To address this point, we analyzed the proliferation of HDAC1, HDAC2 and HDAC3 depleted cells. We show here that in the absence of HDAC1, cells can arrest either at the G1 phase or at G2/M transition of the cell cycle, resulting in the loss of mitotic cells, cell growth inhibition and increase in the percentage of apoptotic cells. On the contrary, HDAC2 knockdown showed no effect on cell proliferation, unless we concurrently knocked down HDAC1. Given the defective cell cycle progression of HDAC1 depleted cells, we asked whether and how its cell cycle related function are regulated. Our data revealed that HDAC1 is subjected to a hyperphosphorylation in mitosis. A mitotic hyperphosphorylation of HDAC2 has already been reported and was confirmed also in our analysis. In vitro phosphorylation assays show that HDAC1 and HDAC2 are phosphorylated by Aurora kinases, mitotic-specific serine-threonine kinases that orchestrate proper mitotic progression and are found deregulated in many cancers. Treatment of synchronized cells with different inhibitors of Auroras as well as siRNA experiments confirm that Aurora kinases are responsible for HDAC1 and HDAC2 phosphorylation in vivo, while inhibition of other mitotic kinase cascades, such as the Plk-1 pathway, has no effect. S406 HDAC1 and S407 HDAC2 were identified as phospho-acceptors sites by mass spectrometry analysis and confirmed by site directed mutagenesis in vitro and in vivo. A phosphopeptide-specific antibody recognizing phosphorylated S406 HDAC1 (anti-pS406-HDAC1) revealed a stronger HDAC1 hyperphophorylation in prophase that quickly diminished in prometaphase. In metaphase a crown-like distribution of pS406-HDAC1 was also reported. Our data globally suggest a mechanistic link between HDACs and Aurora kinases in mitosis. Since both HDACs and Aurora kinases are involved in human tumorigenesis and their inhibitors represent promising anticancer drugs, elucidating this relationship could provide new insights for the application of combinatorial approaches in cancer therapy.

1) Zupkovitz G, Grausenburger R, Brunmeir R, Senese S, Tischler J, Jurkin J, Rembold M, Meunier D, Egger G, Lagger S, Chiocca S, Propst F, Weitzer G, Seiser C. "The cyclin dependent kinase inhibitor p21 is a crucial target for Histone deacetylase 1 as regulator of cellular proliferation.". Mol Cell Biol. 2009 Dec 22.
2) Senese S, Zaragoza K, Minardi S, Muradore I, Ronzoni S, Passafaro A, Bernard L, Draetta GF, Alcalay M, Seiser C, Chiocca S. "Role for histone deacetylase 1 in human tumor cell proliferation". Mol Cell Biol. 2007 Jul;27(13):4784-95.

There is a large body of literature that describes the geometry and the physics of filopodia using either stochastic models [1, 2] or partial differential equations and elasticity theory [3]. Comparatively, there is a paucity of models focusing on the regulation of the network of proteins that control the formation of different actin structures [4]. In this thesis, using a combination of in-vivo and in-vitro experiments, and in-silico modeling we characterized a network of interacting molecules that differentially regulate filopodia formation in multiple cell lines. The components of our network are the actin remodeler Eps8, whose capping and bundling activities are a function of its ligands, Abi1 and IRSp53, respectively; VASP and Capping Protein, which exert antagonistic functions in controlling filament elongation. We connected the protein network to a simple system of actin polymerization, and determined most of the parameters governing the interactions of this minimal system to build a mathematical model. This model accurately accounted for all observations, including a seemingly paradoxical result whereby genetic removal of Eps8 reduced filopodia in HeLa, but increased them in hippocampal neurons, and generated quantitative predictions, which we experimentally verified. The model further permitted us to explain how filopodia are generated in different cellular context, depending on the dynamic interaction established by Eps8, IRSp53 and VASP with actin filaments, thus revealing an unexpected plasticity of the signaling network that governs the multifunctional activities of its components in the formation of filopodia.

The control of cell growth, proliferation and differentiation on biomaterials surfaces is of fundamental importance for regenerative medicine, prosthetics, or cell-based assays. The microfabrication of cell-on-chip platforms based on a new family of poly (amidoamine) hydrogels, are promising for in vitro and in vivo applications. Hydrogels present characteristics that mimics biological environments, such as the cross-linked nature of the extracellular matrix, the tissue properties (high water content), and the permeability to oxygen and metabolites. Hydrogels based on poly (amidoamine) results in an optically transparent, biocompatible and fully biodegradable substrates, recommended for body implants that are minimally invasive, and naturally eliminated by human body. In my PhD work I intended to use microfabricated hydrogels for fine-tuning the contact guidance of cells. As microfabrication tools I set up reaction injection moulding for producing features down to 100 µm and developed a novel approach relying on electron beam lithography. This innovating microfabrication consists in the ability of directly writing patterns on already cross-linked hydrogels, with the capability of producing structures at sub-micrometric scale. The exposure to the electron beam produces particular modifications enabling the control of physico-chemical properties of irradiated area. I obtained a selective attachment of proteins as a function of the electron-beam dose; an exclusive adhesion and growth of neural cells on the exposed surfaces; and the control of neurite outgrowth guidance along a microfabricated network. These results offer new perspectives to build physiological microenvironments or cell-on-chip platforms, based on a novel class of microfabricated hydrogels.

A paradigm in transcriptional regulation is that a graded increase in transcription factor (TF) concentration is translated into a digital on/off transcriptional response by cooperative TF binding to adjacent cognate sites in cis-regulatory sequences. This digital type of response underlies the sharp definition of boundaries among body parts during development. Here we show that NF-kB, a key TF controlling the inflammatory and immune transcriptional responses, is instead an analogical transcriptional regulator relying on the usage of homotypic clusters of high-affinity binding sites. We show that a widespread feature of NF-kB target genes is the homotypic clustering of high-affinity binding sites in their promoter-proximal regions. Contrarily to what expected, we observed that gradually increasing concentrations of NF-kB in the cell nucleus are translated into gradually increasing levels of transcriptional activity of NF-kB inflammatory target genes. We provide a thermodynamic interpretation of the experimental observations by combining quantitative measurements and a minimal physical model of an NF-kB dependent promoter. We show that NF-kB does not bind cooperatively to adjacent sites in a cluster; conversely, independent binding of NF-kB to adjacent sites promotes RNA Pol II recruitment in an additive fashion, thus resulting in a graded transcriptional response. These findings reveal a novel paradigm in the usage of clustered TF binding sites, which may be relevant in all biological conditions in which the transcriptional output must be proportionate to the strength of an environmental input.

Licciulli S, Luise C, Zanardi A, Giorgetti L, Viale G, Lanfrancone L, Carbone R, Alcalay M. "Pirin delocalization in melanoma progression identified by high content immuno-detection based approaches". BMC Cell Biol. 2010 Jan 20;11(1):5.
Giorgetti L, Bongiorno G, Podestà A, Berlanda G, Scopelliti PE, Carbone R, Milani P. "Adsorption and stability of streptavidin on cluster-assembled nanostructured TiOx films". Langmuir. 2008 Oct 21;24(20):11637-44.
Giorgetti L, Zanardi A, Venturini S, Carbone R. "ImmunoCell-Array: a novel technology for pathway discovery and cell profiling". Expert Rev Proteomics. 2007 Oct;4(5):609-16.

Kinetochores are multiprotein scaffolds that connect centromeric DNA to spindle microtubules, thus contributing to the mechanism of sister chromatid separation during anaphase. Besides, kinetochores also ensure the fidelity of sister chromatid separation through the spindle assembly checkpoint, a surveillance mechanism that delays anaphase until all sister chromatids are bipolarly attached. The kinetochore subcomplex RZZ, whose expression is limited to metazoans, contains at least three subunits named Rod, Zw10 and Zwilch. It provides a link between kinetochore-microtubule binding and the spindle assembly checkpoint. The RZZ complex is required for the recruitment of the spindle assembly checkpoint complex Mad1-Mad2 and of the dynein- dynactin complex to the kinetochores. The latter interaction is important for the establishment of initial lateral interactions between kinetochores and microtubules and for the stripping of Mad1-Mad2 from kinetochores to silence the mitotic checkpoint. Mechanisms through which the RZZ complex mediates these different functions are still unclear. My PhD project aimed to characterize both the structural and the functional properties of the RZZ complex together with its core kinetochore interactor, Zwint-1. By using recombinant approaches we found that Zwilch is a globular protein formed from two stable domains that strongly interact with each other. Further, we crystallized a Zwilch construct comprising the full protein sequence, and obtained a high-resolution X-ray diffraction dataset. In addition, we identified Zw10 and Zwilch interacting regions of Rod, a necessary step towards reconstituting the RZZ complex in vitro to study its structural organization by X-ray crystallography. By using a proteomic approach, we identified a novel Zw10 interactor named Nag (neuroblastoma amplified protein), whose overall structural organization is similar to Rod’s: in fact, like Rod, Nag is predicted to contain an N-terminal ?-propeller domain followed by ARM repeats. The similarity is particularly high in a region comprising approximately 200 amino acids following the ?-propeller domain. We show that Nag is an interactor of Zw10 during interphase. Nag engages in a complex with Zw10 and Rint-1. We refer to this new complex as the NRZ complex. Our data suggest that the two distinct Zw10-containing complexes are structurally similar.

Civril F, Musacchio A. “Spindly attachments”. Genes Dev. 2008 Sep 1;22(17):2302-7

Neural Cell Adhesion Molecule (NCAM), which is a member of the Immunoglobulin family of Cell Adhesion Molecules (Ig-CAMs), is mainly expressed in the central nervous system, where it mediates cell-cell adhesion and promotes neurite outgrowth. NCAM is also found in non-neural tissues, where its function is not well characterized yet. Our group has previously reported that in pancreatic beta tumor cells NCAM is able to activate FGFR signaling. To investigate the mechanisms underlying NCAM-mediated modulation of FGFR activity, I first asked whether the interaction of NCAM with FGFR has any impact on the cellular response to FGF. My data demonstrate that, in the presence of NCAM, both FGF-induced signaling and proliferation were significantly reduced and that inhibitory effect of NCAM on FGF functions depends on the ability of NCAM to prevent FGF binding to FGFR. Since NCAM per se is also able to activate FGFR, I next exploit the possibility that NCAM is a non-canonical ligand for FGFR. I found that NCAM not only stimulates some of the known FGFR-mediated pathways, but it is also able to trigger FGFR-mediated cellular events that are remarkably distinct from those elicited by FGFs. While FGF promotes cell proliferation via the classical Ras-Erk1/2 cascade, NCAM induces FGFR-dependent Src activation, resulting in cell migration. Moreover, NCAM, unlike FGF, induces the sustained activation of most FGFR effectors. I next asked whether at NCAM and FGF differentially affect the intracellular fate of FGFR. Internalized FGFR is not ubiquitinated and degraded upon prolonged NCAM treatment, but it recycled back to the cell surface in a Rab11-dependent manner. Finally, FGFR-1 recycling is required for NCAM-induced sustained signaling and cell migration. These results show that NCAM acts both as an antagonist of FGF and as a non-conventional FGFR ligand and suggest that NCAM is able to switch FGFR activity from a proliferative to a pro-migratory one. The NCAM/FGFR interplay could be a potential player (and therapeutic target) in those diseases characterised by aberrant FGFR signaling, such as certain tumor types.

Francavilla C, Loeffler S, Piccini D, Kren A, Christofori G, Cavallaro U. “Neural cell adhesion molecule regulates the cellular response to fibroblast growth factor”. J Cell Sci. 2007 Dec 15;120(Pt 24):4388-94

MicroRNAs are a class of evolutionally conserved small non coding RNAs that regulate basic cellular functions including proliferation, differentiation and cell death. MicroRNAs are generated from primary RNA transcripts that undergo a series of catalytic processing events that require the RNAse type III enzymes DROSHA and DICER. Abnormalities in microRNA expression impact on the normal regulation of these processes and thus can influence tumorigenesis. Indeed, a deregulation of different microRNAs has been found in a variety of cancer types but the role of microRNAs in controlling cellular transformation still remains to be elucidated. The transformation process leading to cancer is often preceded by the accumulation of DNA damage. When DNA damage occurs, discontinuities in the DNA double helix are promptly detected triggering a set of events, the DDR, that leads to a transient arrest of cell proliferation and allows DNA damage repair. The inefficiency in the activation of DDR is one of the major causes of genomic instability associated with transformation. Up to now, many different factors have been shown to modulate the DDR but, so far, RNA has never been suggested to play a direct role in this process. I investigated the involvement of RNA molecules in DDR activation. Upon X-ray (IR)exposure proteins involved in repair and checkpoint signal transduction are recruited to DNA lesions. Their local accumulation forms cytologically detectable structures called DDR foci which contains different DDR factors among which the phosphorylated histone variant ?H2AX, 53BP1, MDC1, ATM and other ATM target proteins. I first studied if DDR foci formation involves an RNA component as previously observed for 53BP1. To do so, I treated irradiated human cells with RNAse A and I monitored DDR foci stability by immunofluorescence. I observed that 53BP1, MDC1, ATM and ATM’s targets foci are sensitive to RNAse A treatment. Importantly, ?H2AX foci are not affected suggesting that chromatin structure and nuclear integrity is preserved upon RNAse A treatment. Moreover, in irradiated and RNAse A-treated cells, I was able to restore DDR foci by addition of exogenous RNA. Strikingly, the same complementation experiment performed with RNA extracted from DICER-inactivated cells, which are unable to generate mature microRNA, failed to allow DDR foci formation. Thus, I decided to study DDR foci formation in DICER-inactivated cells post irradiation. I observed that the intensity and the abundance of DDR foci for MDC1, ATM, and ATM’s target proteins, are reduced in DICER-inactivated cells. This was observed both in transformed DICER-hypomorphic cells and in normal human fibroblasts transfected with an siRNA against DICER. I also found that 53BP1 foci formation is delayed after DICER inactivation. Importantly, I observed the same defect in DDR foci formation in cells knocked down for DROSHA, the endonuclease working upstream of DICER in microRNAs processing suggesting that microRNAs are the candidate RNA molecules required for DDR foci formation. Furthermore, I investigated if the defect observed in DDR foci formation was associated with an impaired signal transduction to the downstream checkpoint kinases that do not form foci, such as CHK2. By immunoblot analysis I observed that CHK2 activation was indeed impaired in DICER inactivated cells. Furthermore, I tested if the defect in DDR foci formation and DDR signal transduction observed in DICER-inactivated cells was functionally relevant causing an impaired checkpoint response after irradiation. Depending on the cell cycle phase in which the DNA is damaged, cells stop cell-cycle progression either at the G1/S or at the G2/M transition. I observed that the G1/S transition checkpoint is impaired in DICER-hypomorphic cells and in three different human fibroblasts cell lines transiently knocked down for DICER and DROSHA by siRNA transfection. Moreover, I observed that also the G2/M transition checkpoint is impaired in DICER-inactivated cells by shRNA transfection. Thus, I conclude that DICER and DROSHA, and therefore most likely microRNAs, are strictly required for DDR foci formation, DDR signal transduction and cell cycle arrest upon irradiation in human cells. These results reveal an uncharacterized role of microRNAs in regulating the cellular response to DNA damage.

Francia S, Weiss RS, Hande MP, Freire R, d'Adda di Fagagna F. “Telomere and telomerase modulation by the mammalian Rad9/Rad1/Hus1 DNA-damage-checkpoint complex”. Curr Biol. 2006 Aug 8;16(15):1551-8.

A major goal in genomics is to understand how genes are regulated in different tissues, stages of development and diseases. The key feature distinguishing epigenomic from genomic information is the dynamic pattern of epigenetic features, which impose cell-type specific expression, regulating cell function. Genome accessibility in chromatin is an established epigenetic feature of active regulatory DNA. We combined restriction enzymes and next generation sequencing to isolate accessible DNA elements in primary human CD34+ cells during myeloid differentiation. NA-Seq can be used to study regulatory networks since we demonstrate that nuclease accessible sites (NAS) serve as a predictive marker of in vivo DNA binding by several classes of transcription factors. We identified known and novel cis-regulatory elements, including enhancers, silencers, insulators and promoters. Furthermore, unsupervised clusterization of histone modifications detected in a set of NAS representative of different cis-regulatory elements suggests that acetylated lysine 5 of histone H2A may be considered as a distinctive and predictive marker of cis-regulatory activity at non-promoter DNA elements. Differentiation of hematopoietic cells is accompanied by changes in gene expression and gross nuclear structure, whose topology and inter-relationships are unknown. We found that promoter accessibility, regulatory potential and gene expression pattern were all restricted upon differentiation. NAS were also found at promoter of genes poised for activation in CD34+, and these promoters were invariably associated with di-methylation of lysine 4 (H3K4me2). Finally, we mapped large chromosomal domains showing differential accessibility between progenitors and maturing cells, non-randomly distributed and decreasing in number in CD34- cells. Our results are consistent with a model of a permissive chromatin structure in multipotent cells, becoming progressively and selectively closed during differentiation.

Gargiulo G, Levy S, Bucci G, Romanenghi M, Fornasari L, Beeson KY, Goldberg SM, Cesaroni M, Ballarini M, Santoro F, Bezman N, Frigè G, Gregory PD, Holmes MC, Strausberg RL, Pelicci PG, Urnov FD, Minucci S. “NA-Seq: a discovery tool for the analysis of chromatin structure and dynamics during differentiation”. Developmental cell 2009;16(3):466-81.

Dendritic cells (DCs) isolated from the gut have been shown to display specialized functions, including the ability to promote gut tropism to B and T lymphocytes, to polarize non-inflammatory responses and to drive the differentiation of adaptive Foxp3+ T regulatory (Treg) cells. However, very little is known on what drives the mucosal phenotype of DCs. In our laboratory it was recently described that the cross-talk between human intestinal epithelial cells (IECs) and DCs helps maintaining gut immune homeostasis via the induction of non-inflammatory DCs. In this thesis we present evidence that both in human and mouse the local microenvironment, and in particular intestinal epithelial cells, promote the differentiation of tolerogenic DCs able to drive the development of adaptive Foxp3+ Treg cells. EC-derived TGF-ß and retinoic acid (RA), and in humans also Thymic-stromal lymphopoietin (TSLP), are required for DC conversion. Accordingly, functionally altered human ECs derived from Crohn’s disease patients, have reduced ability to induce tolerogenic DC with Treg polarizing capacity. A subpopulation of gut CD103+ DCs is known to induce the expression of gut homing receptors on T cells and to drive Treg cell development in mice. We extended these studies and found that in both, human and mice, IECs are able to promote CD103 expression on DCs. Importantly, we found that also human primary CD103+ DCs isolated form mesenteric lymph nodes (MLNs) play tolerogenic function by converting naïve T cells into Foxp3+Treg cells. This subset of DCs express CCR7 and seem to represent a LP-derived migratory population. Finally, EC- DC induced Treg cells display gut-homing properties and when adoptively transferred, protect mice from experimental colitis. Thus, we describe an important check of ECs on DC function that is required for tolerance induction. In addition, we identified a population of tolerogenic CD103+ DCs in the human gut that likely differentiate in response to IEC-derived factors and drive Treg development.

Iliev ID, Matteoli G, Rescigno M. “The yin and yang of intestinal epithelial cells in controlling dendritic cell function”. J Exp Med. 2007 Oct 1;204(10):2253-7.

Maintenance of chromosome integrity during DNA replication is essential for preventing genome rearrangements and cancer. When cells experience replication stress, the stability of stalled forks is controlled by the checkpoint response, which in the yeast Saccharomyces cerevisiae requires Rad53 kinase. Hence Rad53 deficient cells are unable to recover from fork stalling induced by the replication inhibitor hydroxyurea (HU). We designed a genome-wide screen aimed to identify pathways contributing to the viability of rad53 deficient cells experiencing replication stress. We generated an arrayed library containing all viable Knock Out (K.O.) strains of S. cerevisiae carrying an inducible dominant negative allele of RAD53 (GALrad53-D339A) and screened for mutants able to grow in the presence of HU under inducing conditions. By this procedure we have identified 18 mutants, out of ~ 4300 represented in the library, whose deletion rescues the checkpoint deficiency under replication stress. These suppressors could be grouped into two classes according to their ability to rescue the kinase deficient allele rad53-K227A. Class I rad53 suppressors allow growth of both rad53–K227A and rad53–D339A overexpressing cells in the presence of HU and belong to pathways regulating CDK activity, G1/S specific transcription regulation and messenger ribonucleoprotein (mRNP) particle biogenesis. Class II rad53 suppressors rescue the HU sensitivity of rad53–D339A overexpressing cells only, most likely by upregulating wild type Rad53 kinase activity. Accordingly, we identified genes belonging to phosphatase complexes known to influence the phosphorylation status of Rad53. Besides the checkpoint phosphatases Pph3/Psy2 and Ptc2, we identified other factors involved in checkpoint inactivation: 1) Nma111, an evolutionary conserved serine-protease involved in apoptotic processes, that might influence DNA damage checkpoint inactivation by degrading factors that sustain checkpoint signaling. 2) Bul2, a specificity factor of the Rsp5 ubiquitin ligase, that is required for efficient checkpoint inactivation likely by promoting inhibitory ubiquitylation of Mec1 kinase.

The aim of the present thesis lies in the study of protein complexes by mass spectrometry. We report the results of a combined approach of cross-linking, mass spectrometry, and bioinformatics relative to the analyses of three human complexes: the Eps8-Actin dimer, the Ndel1 homodimer and the Ndc80 heterotetramer. An important limitation of the cross-linking approach, so far, was the identification of cross-linked peptides from fragmentation spectra. Our novel approach overcomes the data analysis bottleneck of cross-linking and mass spectrometry. We have constructed a purpose-built database to match spectra with cross-linked peptides, defined a score that expresses the quality of our identifications, and estimated false positive rates. Our analysis has shed light on critical structural parameters such as the mechanism of binding of Eps8 to actin, the directionality of the homo-dimeric coiled-coil of Ndel1 and the register of the hetero-dimeric coiled-coils of the Ndc80 complex. In addition, it revealed important structural details regarding the organization of its tetramerization domain. This study nicely complemented structural data obtained by classical techniques such as X-ray crystallography and electron microscopy, and it was instrumental to define the functional role of the three complexes in the cell. Finally we report preliminary results from an investigation of the function of a novel interactor of the spindle assembly checkpoint (SAC) protein Bub3, named Sjogren syndrome scleroderma autoantigen 1 (SSSCA1). We identified SSSCA1 while attempting to elucidate the mechanisms of SAC proteins recruitment to the kinetocore during mitosis. Using mass spectrometry and classical biochemical and cell biology approaches, we started defining the role of SSSCA1 in this process.

Maiolica A, Cittaro D, Borsotti D, Sennels L, Ciferri C, Tarricone C, Musacchio A, Rappsilber J. “Structural analysis of multiprotein complexes by cross-linking, mass spectrometry, and database searching”. Mol Cell Proteomics. 2007 Dec;6(12):2200-11.
Maiolica A, Cittaro D, Borsotti D, Sennels L, Ciferri C, Tarricone C, Musacchio A, Rappsilber J. “Structural analysis of multiprotein complexes by cross-linking, mass spectrometry, and database searching”. Mol Cell Proteomics. 2007 Dec;6(12):2200-11.

Stimulation of cells with epidermal growth factor (EGF) results in the activation of the intrinsic tyrosine kinase of the receptor (EGFR) and the establishment of a complex signalling network leading to a variety of biological outcomes. Eps15 and other adaptor proteins involved in EGFR internalization are monoubiquitinated upon EGF stimulation by a mechanism that involves the E3 ligase Nedd4. I have studied two major questions during this thesis work. First, I was interested in understanding how RTK-originated signals are delivered to the ubiquitination machinery. Although Nedd4 can monoubiquitinate eps15 efficiently, it does not stably associate with it or with the EGFR. To gain insight on this issue, I used an integrated approach based on specific pharmacological inhibitors and molecular genetics tools. In this thesis, I showed that EGF-mediated monoubiquitination of eps15 is dependent on the kinase of the receptor and on the EGFR-mediated activation of Src kinase with the latter being responsible for the tyrosine phosphorylation of Nedd4. Indeed, in cells treated with specific Src inhibitors or stable KD for Src eps15 monoubiquitination does not occur and concomitantly Nedd4 protein is not phosphorylated. This modification appears to modulate the catalytic activity of Nedd4. My second goal was to understand why do Ub receptors tend to be monoubiquitinated rather than polyubiquitinated. The E3 effector machinery, represented by Nedd4, is equally competent to both mono- and polyubiquitinate substrates. Is this a characteristic of the Ub machinery (E3 or E2–E3 complex) or does it depend on intrinsic properties of the substrate? Trying to answer these questions I investigated the molecular details of Nedd4 catalysis. With a series of biochemical experiments I was able to prove that the HECT domain of Nedd4 is able to bind non-covalently to Ub. Results demonstrated that the Ub binding ability resides in the HECT catalytic domain of the protein, more precisely in the N-lobe subdomain, that also contacts the E2 conjugating enzyme. A crystallization approach, still ongoing, was pursued in order to map the interaction surface. Up to know I was able to obtain crystals for the HECT alone. Phasing and refinement on the collected datasets are ongoing. In addition, to identify human E2s that can interact specifically with Nedd4, I used the isolated HECT domain in a yeast two-hybrid screening with a library of E2s. A subset of E2s binds specifically and was characterized further for their ability to work in concert with the E3 Nedd4.

Woelk T, Oldrini B, Maspero E, Confalonieri S, Cavallaro E, Di Fiore PP, Polo S. “Molecular mechanisms of coupled monoubiquitination”. Nat Cell Biol. 2006 Nov;8(11):1246-54

Lymphatic vessels control cell traffic and fluid exchanges in tissues, playing a crucial role in the inflammatory processes accompanying several pathologies and acting as the preferential route of metastatic spread of many solid tumors. As cell-to-cell contacts contribute to regulate fluid fluxes and trafficking of circulating cells, targeting the poorly known lymphatic endothelial junctions could be a useful tool to control inflammation and metastasis. In collaboration with D. Mc Donald’s group at UCSF, we have reported the presence of novel, discontinuous junctions between endothelial cells of initial lymphatics (buttons) differing from continuous junctions in collecting lymphatic and blood vessels (zippers). Both junctions in lymphatics were composed of VE-cadherin and tight junction–associated proteins, including ZO-1 and JAM-A. Our findings suggest that fluids enter through initial lymphatics via openings between buttons, without disrupting junctional integrity. In addition, we applied a subtractive immunization approach to identify novel lymphatic-specific or -enriched junctional molecules, using HUVEC and hDLEC as the tolerogen blood endothelial and the immunogen lymphatic endothelial cell model, respectively. Four out of ten of the isolated monoclonal antibodies recognized CD73, a GPI-anchored ectoenzyme catalyzing the extracellular conversion of 5’-AMP to adenosine, a physiological regulator of endothelial barrier. Opposite to the in vitro data, CD73 was found expressed also in lymphatic and blood vessels of various mouse tissues in vivo. We found that CD73-mediated production of adenosine, besides leading to a decrease in paracellular permeability, results in appearance of actin belt at the cell borders and linearization of the junctional markers VE-cadherin, ?-catenin and ZO-1. Finally, a direct link between CD73 function and activation of the junctional regulator Rap1 was proved for the first time. This link could represent an interesting pathway contributing to the dynamic and fine tuning of the endothelial barrier function.

Cera MR, Fabbri M, Molendini C, Corada M, Orsenigo F, Rehberg M, Reichel CA, Krombach F, Pardi R, Dejana E. “JAM-A promotes neutrophil chemotaxis by controlling integrin internalization and recycling”. J Cell Sci. 2009 Jan 15;122(Pt 2):268-77

For many vertebrate cell types the G0?S-phase progression is the major decision point regulating cell proliferation and execution of cell differentiation programs. For example, the quiescent state is important for maintenance of self-renewal potential in stem cell compartments. The G0?S-phase cell cycle progression can be induced in resting fibroblasts by restoration of mitogens. Several immediate-early serum-responsive genes, such as c-myc, c-fos, or c-jun, encode transcription factors that are believed to drive induction of secondary mitogen-responsive genes. Consistent with this view, Myc function as a sequence-specific transcription factor that has been shown to be essential for this response, but the set of genes directly regulated and required in such conditions is still largely obscure. The purpose of this study is to identify the target genes that are directly regulated by Myc and that are required for the proper execution of the G0?S-phase progression. We took advantage of c-mycflox/flox 3T9 fibroblasts. In these cells, the c-myc open reading frame (ORF) is flanked by loxP sites in order to obtain its deletion using an inducible Cre-ER recombinase activity. Using DNA microarrays, a gene expression profiling of serum response in murine fibroblasts either expressing or not expressing Myc was performed. We found a group of 93 genes induced by serum only in Myc-expressing cells. Most importantly, seventy-eight of these genes were directly bound by Myc at their promoters during mitogenic response, as determined with a quantitative chromatin immunoprecipitation assay. Gene ontology analysis revealed that the group of genes was enriched for genes involved in metabolism of nucleotides and for those coding for ribosomal proteins and regulators of translation. This group of direct Myc-target genes likely constitutes the core transcriptional response built up by Myc during the G0?S-phase progression. Next goal will be to understand if any of those genes is necessary for the cell cycle re-entry governed by Myc in response to mitogens. Accordingly a functional screen, based on a retroviral shRNA expression system, is on the way to knock-down the identified Myc-target genes.

Smith AP, Verrecchia A, Faga G, Doni M, Perna D, Martinato F, Guccione E & Amati B “A positive role for Myc in TGFbeta-induced Snail transcription and epithelial-to-mesenchymal transition”. Oncogene. 2009 Jan 22;28(3):422-30

Transforming growth factor-? (TGF-?) is a multifunctional growth factor that regulates proliferation and migration of many cell types. In endothelial cells (ECs), TGF-? mediates its cellular effects through three serine/threonine kinase receptors: a TGF-? type II receptor (T?RII) and two distinct TGF-? type I receptors (ALK1 and ALK5). The co-receptor endoglin modulates ALK1 and ALK5 signaling. Ligand-dependent assembly of these receptors results in activation of the Smad proteins which regulate target gene transcription. Endothelial adherens junctions (AJs) are junctional protein structures that regulate vascular homeostasis. Vascular endothelial cadherin (VE-cadherin) is the main component of endothelial AJs that mediates cell-to-cell homophilic adhesion and transfers intracellular signals to promote the steady state of the endothelium. However, the molecular mechanisms by which VE-cadherin signals are still poorly understood. The notion that TGF-? receptors knock-out (KO) mice phenotype strikingly resembles that of VE-cadherin KO mice prompted us to investigate a possible crosstalk between TGF-? and AJ pathways in ECs. Here we report that VE-cadherin expression and junctional clustering are required for optimal TGF-? signaling in ECs. TGF-? anti-proliferative and anti-migratory responses are increased in the presence of VE-cadherin. ECs lacking VE-cadherin are less responsive to TGF-?/ALK1- and TGF-?/ALK5-induced Smad phosphorylation (in in vitro and ex-vivo models) and target gene transcription. Proper VE-cadherin clustering at interendothelial junctions is required for efficient TGF-?-dependent Smad phosphorylation. We found that VE-cadherin physically associates with all the components of the TGF-? receptor complex, T?RII, ALK1, ALK5, and endoglin. Clustered VE-cadherin recruits TGF-? receptors at the cell-to-cell contacts and may promote TGF-? signaling by enhancing T?RII/ALK assembly into active receptor complexes. Taken together our data indicate that VE-cadherin is a positive and EC-specific regulator of TGF-? signaling. This suggests that reduction or inactivation of VE-cadherin may contribute to progression of diseases where TGF-? signaling is impaired.

Rudini N, Felici A, Giampietro C, Lampugnani M, Corada M, Swirsding K, Garrè M, Liebner S, Letarte M, ten Dijke P, Dejana E. “VE-cadherin is a critical endothelial regulator of TGF-beta signalling”. EMBO J. 2008 Apr 9;27(7):993-1004.

Homologous Recombination is a fundamental cellular process that is required for the repair of double strand breaks and the gaps arising during DNA replication. Inappropriate commitment of recombination can lead to genome instability and cell lethality. The UvrD DNA helicase family plays a pivotal role in the regulation of the recombination processes. In particular, UvrD helicases can antagonize unscheduled recombination events by disrupting Rad51 nucleofilaments at diverse steps during the process. Srs2 (Suppressor of Rad6 Sensitivity), a member of the UvrD family in S. cerevisiae, is phosphorylated in DNA damaging conditions by the cyclin dependent kinase CDK1 and contains seven putative phosphorylation sites at the C-terminus. CDK1 controls cell cycle progression and is a key regulator of the recombination process, although most of the repair targets are unknown. To address the role of CDK1-dependent phosphorylation of Srs2, we have created two alleles that mimic the unphosphorylated and the phosphorylated forms of Srs2. We found that the phosphorylation of Srs2 is required for Rad51-dependent Double Stran Break repair. Moreover, Srs2 phosphorylation controls the proper recruitment of the protein at DSBs and reduces its affinity for PCNA (Proliferating Cell Nuclear Antigen). Finally, the inability to phosphorylate Srs2 stimulates its SUMOylation, a modification that might be important in regulating protein stability. Srs2 orthologues in higher eukaryotic organisms have not been unequivocally identified although the importance of Srs2 in the regulation of the recombination processes suggests that this activity could be conserved throughout the evolution. We found that the UvrD helicase domain of Srs2 is highly conserved to that of human F-box protein hFBH1. Moreover, hFBH1 inactivation in human cells induces accumulation of spontaneous Rad51nucleofilaments, suggesting that it might play an important role in recombination. Finally,hFBH1 suppresses DNA damage sensitivities of srs2? mutants; this suppression requires afunctional helicase domain but also the F-box domain, that is not conserved in SRS2 and itallows self-regulation of hFBH1 protein level by an SCF-dependent mechanism. Altogether these findings indicate that both Srs2 and hFBH1 UvrD helicases are subjected to post-translational protein modifications that might be fundamental for their recombination roles.

Chiolo I, Saponaro M, Baryshnikova A, Kim JH, Seo YS, Liberi G. “The human F-Box DNA helicase FBH1 faces Saccharomyces cerevisiae Srs2 and postreplication repair pathway roles”. Mol Cell Biol. 2007 Nov;27(21):7439-50.

Acute myeloid leukemia (AML) is sustained by a group of cells, known as leukemic stem cells (LSC) that resemble normal hematopoietic stem cells (HSC) but present alterations in their growth properties and increased cell survival. The persistence of LSC after cytotoxic and antiproliferative therapy is thought to be responsible for disease relapse, which occurs in approximately 70% of AML cases. In an effort to characterize specific features of leukemia initiating cells we decided to directly assess their transcriptional status in two different murine AML models, and compare their global gene expression profile to that of the normal counterparts. Leukemic mice were generated by retroviral transduction of bone marrow cells with PML/RARa or AML1/ETO oncogenes. Different cellular subpopulations were purified using antibodies against specific surface markers (Mac1, c-kit and Sca1), and global gene expression analysis was performed using Affymetrix microarray technology. Our results led to the characterization of gene expression profiles in different cellular subpopulations from normal, preleukemic and leukemic BM and to the identification of putative LSC markers. In particular, we focused our attention on the aberrant expression of the Ncam1 adhesion molecule in LSC derived from the APL model, since its expression was confirmed in 50% of AML patients, regardless of disease subtype. Our studies suggest that Ncam1 is expressed in the LSC compartment, and may therefore represent a relevant target for designing novel therapeutic strategies in AML.

Fumasoni I, Meani N, Rambaldi D, Scafetta G, Alcalay M, Ciccarelli FD. “Family expansion and gene rearrangements contributed to the functional specialization of PRDM genes in vertebrates”. BMC Evol Biol. 2007 Oct 4;7:187

EPN3 belongs to the Epsin family of endocytic proteins. Unlike other Epsin members which are ubiquitously expressed, EPN3 is known to be prevalently expressed under pathological conditions, such as wound healing and ulcerative colitis. Furthermore, EPN3 has also been found to be present in a breast cancer ‘metastasis signature’ predictive of metastatic relapse in lymph node negative breast cancer patients. In this study, I have analyzed the expression of EPN3 in human breast cancers by tissue microarray analysis and have demonstrated that it is indeed overexpressed in approximately 30% of the breast tumors. Although EPN3 expression at the mRNA level does not have prognostic value it significantly correlates with clinical indicators indicative of an aggressive phenotype, such as poorly differentiated and highly proliferating tumors. Furthermore, I have also shown that the EPN3 gene is amplified in 18% of human primary breast tumors and that EPN3 amplification correlates with transcript and protein levels in primary breast tumors and breast cancer cells, respectively. EPN3 is located at 17q21.33 on chromosome 17. The ERBB2 locus, also found on chromosome 17, is commonly amplified in breast cancer. Although EPN3 and ERBB2 amplification significantly correlates, I have also shown that EPN3 amplification occurs independently from ERBB2 amplification. The physiological function of EPN3 is largely unknown. In this study, I have developed appropriate cell models for functional studies of EPN3, the main focus being on uncovering an endocytic function for EPN3. I characterized the localization of EPN3 in specific early endocytic compartments and employed both the epidermal growth factor receptor and transferrin receptor as models to assess the impact of EPN3 on ligand-induced and constitutive internalization, respectively. I found that EPN3 overexpression and/or ablation does not affect the endocytosis/signaling of either receptor. Since Drosophila Epsin participates in the NOTCH pathway, I analyzed the possible involvement of EPN3 in NOTCH signaling by measuring the transcription levels of selected NOTCH targets of the HES and HEY families in breast cancer cell lines. Initial data indicate that EPN3 might affect the mRNA levels of one such target, HEY2, suggesting that EPN3 might be involved in the regulation of NOTCH signaling in cancer cells. In conclusion, this project has demonstrated overexpression of EPN3 in breast cancer and has begun to characterize an endocytic role for EPN3. The project has also unveiled previously unknown links between this endocytic protein and the regulation of NOTCH, a protein that may act as an oncogene in breast cancer.

Vecchi M, Nuciforo P, Romagnoli S, Confalonieri S, Pellegrini C, Serio G, Quarto M, Capra M, Roviaro GC, Contessini Avesani E, Corsi C, Coggi G, Di Fiore PP, Bosari S. “Gene expression analysis of early and advanced gastric cancers”. Oncogene. 2007 Jun 21;26(29):4284-94

Mad2 is a key component of the spindle assembly checkpoint, a safety device ensuring faithful sister chromatid separation in mitosis. The target of Mad2 is Cdc20, an activator of the anaphase promoting complex/cyclosome (APC/C). Mad2 binding to Cdc20 is a complex reaction that entails the conformational conversion of Mad2 from an open (OMad2) to a closed (C-Mad2) conformer. Previously, it has been hypothesized that the conversion of O-Mad2 is accelerated by its conformational dimerization with C-Mad2. This hypothesis, known as the Mad2-template hypothesis, is based on the unproven assumption that the natural conversion of O-Mad2 required to bind Cdc20 is slow. Here, we provide evidence for this fundamental assumption and demonstrate that conformational dimerization of Mad2 accelerates the rate of Mad2 binding to Cdc20. On the basis of our measurements, we developed a set of rate equations that deliver excellent predictions of experimental binding curves under a variety of different conditions. Our results strongly suggest that the interaction of Mad2 with Cdc20 is rate limiting for activation of the spindle checkpoint. Conformational dimerization of Mad2 is essential to accelerate Cdc20 binding, but it does not modify the equilibrium of the Mad2:Cdc20 interaction, i.e. it is purely catalytic. These results surpass previously formulated objections to the Mad2-template model and predict that the release of Mad2 from Cdc20 is an energy-driven process.

Simonetta M, Manzoni R, Mosca R, Mapelli M, Massimiliano L, Vink M, Novak B, Musacchio A, Ciliberto A. “The influence of catalysis on mad2 activation dynamics”. PLoS Biol. 2009 Jan 13;7(1):e10.

Tumor suppressor genes are frequently affected by somatic alterations in cancer, and the impairment of their normal function provides a strong contribution to tumorigenesis. Short-hairpin (sh) RNA libraries are powerful genetic tools to uncover important new players in human cancer. To identify potential novel tumor suppressor genes acting in the p53 and pRB pathway we performed an shRNA screen using two different cell-based models in which only a single additional genetic event disrupting either the p53 or the pRB pathway is required to obtain in vitro transformation. I report here on the identification of the Frizzled-ligand Norrin (Norrie disease protein), identified by this approach, as a candidate tumor suppressor. Norrin, is a secreted protein, which it was originally described to be mutated in a X-linked disease characterized by incomplete retinal vascularization leading to blindness, and by frequent hearing loss and mental disorders (Norrie disease). It was only recently discovered that Norrin, although structurally unrelated to WNT ligands, binds the FZD4 receptor with high affinity, and activates the canonical WNT/?-catenin pathway leading to ?-catenin stabilization. Unexpectedly inhibition of Norrin expression promotes anchorage-independent growth, substituting either for p53 or pRB inactivation during cellular transformation. These evidences, indicate Norrin as a novel putative tumor suppressor. Accordingly the aim of my project was to better understand the involvement of Norrin in tumor suppression and to reveal through which mechanism it acts. Importantly Norrin depletion confers a strong growth advantage to cells and causes a proteasome-dependent reduction of p53 protein levels. Conversely, recombinant human Norrin increases p53 levels in a ?-catenin dependent fashion. These data show that Norrin and WNT signaling are fundamental regulator of p53 stability. My data suggest that Norrin affect also the pRB pathway. Norrin knock down (k.d.) leads to upregulation of some E2F target genes. My preliminary data suggest that increased pRB phosphorylation could be involved. H-RASV12 overexpression leads to accumulation of hypophosphorylated form of pRB and induction of SAHF. Significantly, I observed that upon Norrin depletion, in RASV12 overexpressing cells, pRB is still highly phosphorylated. Conversely, rhNorrin administration leads to downregulation of several E2F target genes, to the induction of p16 and SAHFs formation, which could be due to decrease in pRB phosphorylation. Interestingly these data indicate that Norrin could modulate pRB activity. Furthermore my findings suggest that RAS pathway may activate pRB through Norrin. Notably it was not previously shown that Norrin and WNT/?-catenin signaling could regulate the pRB/E2F pathway. Importantly, I discovered that Norrin expression is stimulated by oncogenic H-RAS and B-RAF, mostly through the MEK-ERK pathway. In addition I observed Norrin upregulation during cellular senescence triggered by different stimuli (oncogenes, DNA damage, replicative exhaustion). Notably Norrin depletion allow bypass of oncogene induced senescence (OIS) and senescence induction by irradiation. I show that Norrin is activated upon oncogenic stress and during cellular senescence, similar to other tumor suppressor genes, and that loss of Norrin allows escape from senescence, which is commonly depicted as a barrier for tumor progression. All together these data suggest that Norrin is part of an early fail-safe mechanism to suppress transformation, and that mutation or down regulation of Norrin could contribute to tumor progression. Coherently, we found that Norrin expression is significantly decreased in human melanoma, breast, prostate, ovarian cancer and during chemically induced skin carcinogenesis in mice. I present here data showing that the candidate tumor suppressor Norrin, is involved in regulation of the p53 and pRB pathway. Furthermore Norrin is induced by oncogenic RAS/RAF signaling and during senescence while is downregulated in several tumor types. All together, these data indicate Norrin as a component of a previously unknown tumor suppressive circuit. Indeed these findings support the existence of a novel autocrine/paracrine feedback loop that constrains tumorigenesis, in which, strikingly, the crosstalk between the RAS and ?–catenin pathways plays an unanticipated role. Interestingly, since Norrin is an extracellular ligand, after its further validation as a tumor suppressor, a novel therapeutic approach could involve the use of Norrin or the design of Norrin agonists.

Intercellular junctions mediate adhesion and communication between adjoining cells. Although formed by different molecules, tight junctions (TJs) and adherens junctions (AJs) are thought to be functionally and structurally connected. AJs are known to organize earlier than TJs during cell-cell contact establishment and intracellular TJ molecules such as zona occludens (ZO)-1 localize at AJs at early stages of contact formation and redistribute at TJs only after junction stabilization. Despite this indirect evidence linking these distinct junctional complexes, the signalling pathways behind AJ/TJ interaction are unknown. We describe in our work a novel and cell-specific mechanism of crosstalk between these two types of structures. We showed, in several in vitro models of endothelial cells (ECs) and in mouse allantoises, that the expression and clustering of VE-cadherin at AJs is needed for the upregulation of the TJ adhesive protein claudin-5. This mechanism may have a relevant impact on endothelial physiological functions, as claudin-5 has been shown to regulate endothelial permeability. Consistently, we observed that the increased permeability of VE-cadherin null cells in vitro is mainly due to their impaired claudin-5 expression. Claudin-5 upregulation needs the release of the inhibitory activity of the transcription factor FoxO1 which, by binding claudin-5 promoter region in two distinct sites, inhibits gene expression. Using several approaches aimed at increasing FoxO1 transcriptional activity in ECs we extensively proved that this is a transcriptional repressor of claudin-5 gene expression. VE-cadherin clustering at AJs activates the phosphatidylinositol-3-OH kinase [PI(3)K]/Akt pathway leading to FoxO1 phosphorylation. This impairs FoxO1 DNA binding and induces the nuclear exclusion of the factor inhibiting its transcriptional activity and allowing claudin-5 upregulation. Furthermore, we defined the role of ?-catenin in the regulation of claudin-5 gene expression. We observed that clustered VE-cadherin is able to limit ?-catenin nuclear localization, and consequently decrease Tcf/?-catenin transcriptional activity, by binding this molecule and sequestering it at the plasma membrane. Since increased nuclear ?-catenin proved able to cause claudin-5 downregulation, we hypothesized that ?-catenin could act in concert with FoxO1. We observed that in VE-cadherin null cells, ?-catenin is able to translocate to the nucleus and form a complex with Tcf-4 and FoxO1, cooperating in claudin-5 repression. Consistently, the FoxO1/?-catenin/Tcf-4 multiprotein complex is much more abundant in the nuclei of VE-cadherin null cells than in those of VE-cadherin positive, and is able to bind claudin-5 promoter only in the lack of VE-cadherin clustering. Furthermore, the inhibition of ?-catenin/DNA binding strongly weakens FoxO1/DNA interaction, suggesting that ?-catenin could act by stabilizing FoxO1 interaction with the claudin-5 promoter. In summary, VE-cadherin acts by inducing FoxO1 phosphorylation through Akt activation and by limiting ?-catenin nuclear translocation, preventing in this way the binding of the FoxO1/?-catenin/Tcf-4 complex to the claudin-5 promoter, and allowing the expression of the TJ molecule. These results offer a molecular basis for the link between AJs and TJs and explain why VE-cadherin inhibition may cause a dramatic increase in endothelial permeability.

Taddei A, Giampietro C, Conti A, Orsenigo F, Breviario F, Pirazzoli V, Potente M, Daly C, Dimmeler S, Dejana E. “Endothelial adherens junctions control tight junctions by VE-cadherin-mediated upregulation of claudin-5”. Nat Cell Biol. 2008 Aug;10(8):923-34.

All organisms respond promptly to nuclear DNA breaks by launching a set of coordinated actions collectively known as the DNA-damage response (DDR). This response integrates cell-cycle checkpoints and DNA repair pathways to fix DNA breaks. The aim of my PhD work was to elucidate DDR events at the level of a whole organism: the nematode Caenorhabditis elegans (C. elegans). The minuscule organismal structure of C. elegans contains proliferating, terminally differentiated and stem cells, thus providing a wide and easily accessible pool of distinct cell types in different cellular contexts for investigation of DDR. It has been previously shown that worm germline cells undergo two spatially separated responses to DNA damage: a cell-cycle arrest of mitotic cells and apoptosis of meiotic cells [1]. The matter of DDR in somatic cells, however, has so far remained unaddressed. In my thesis I show that, following exposure to ionizing irradiation, DNA breaks are generated uniformly in all worm cells but strikingly only germline cells mount a DDR, as detected by the activation of phosphorylation events initiated by upstream elements of the DDR machinery. I sought to understand the molecular mechanisms underlining checkpoint non responsiveness of somatic cells and I discovered that they do not express checkpoint genes in the first place. Surprisingly, and contrary to mammalian systems, the key elements of the DNA checkpoint machinery are transcriptionally silenced in worm somatic cells. I further demonstrate that even if somatic cells do not activate DDR in the presence of DNA breaks, they do retain the ability to repair them. These observations suggest the novel concept of uncoupling of checkpoint and DNA repair pathways in somatic cells in organisms that are devoid of somatic stem cell pools. In addition to DDR analysis in the worm somatic cells, I investigated the mechanisms of checkpoint activation in the germline. I discovered that a differential regulation exists in the activation of the upstream DDR factor ATM-1, in the two distinct germline compartments: the mitotic and the meiotic one. Cells in the mitotic compartment show reduced ATM-1 kinase activation in the presence of DNA breaks while meiotic cells show stronger activity. I provide evidence that this differential activity is regulated by the GLP-1/Notch signaling pathway that maintains stemness in the mitotic compartment. These results were further strengthen by the observations I made with cultured human cells that show downregulation of DDR in the presence of activated Notch. Therefore Notch controls DDR by mechanisms conserved in different systems like worms and human cells. In addition, I have shown that ATM-1 activation coincides with CEP-1 expression in the pachytenee region where cells undergo apoptosis upon DNA damage. I would like to propose that ATM-1 activity is required for apoptotic events. Conversely, I infer that the lack of ATM-1 activity in the mitotic compartment prevents these cells from undergoing CEP-1 mediated apoptosis. Finally, a part of results in my thesis is focused on the endogenous events leading to ATM-1 activation that are related to the process of double strand break generation and homologous recombination during meiosis.

• Current students
• Graduation day