<p>Glioblastoma multiforme is a malignant brain tumor characterized by heterogeneity.Interactions between heterogeneous tumor cells are supposed to affect the behavior of awhole tumor cell population. In this thesis an <i>in vitro</i> model system of clonal glioma celllines originating from one glioblastoma tumor was used, and the behavior of cells incocultures was studied and compared the behavior of cells grown separately. The resultsindicate the presence of two types of interactions. In one, paracrine signals acted via extra-cellular media. This was associated with increased growth of the whole co-culture followedby a selective force driving one clone to dominance. In the other type, the cell clones grewside by side without signs of paracrine signalling, in a balance resulting in an increasedterminal cell density. Further investigations focused on mechanisms of interactions in thiscombination.</p><p>Two cell clones were chosen, a GFAP⁺ and a GFAP^-, for further experiments. Withdifferential display PCR it was possible to investigate their specific gene expressionpatterns. Seventeen cDNA fragments were differentially expressed, among them twocorresponded to known transcription factors, ATF3 and prox-1, one to a cytoskeletal protein,α-tropomyosin. The collection also contained eight ESTs (Expressed Sequence Tags) wherethe corresponding genes are unknown at present. Expression of the isolated sequences werealso analyzed in a panel of 12 different glioma cell lines and the results illustrate thecomplexity of gene expression and of tumor heterogeneity. Genes, the expression levels ofwhich were modulated in co-cultures and/or were cell density dependent, were alsoidentified.</p><p>PDGF B is suggested to play a role in sarcomas. The gene codes for an mRNA transcriptwith long UTRs, parts of which are deleted in the homologous oncogene <i>v-sis</i>. The UTRs ofPDGF B mRNAs in human sarcomas were investigated for deletions similar to <i>v-sis</i> thatmight result in increased protein levels. A new transcript variant was identified, lacking a149 base region in the 3'UTR, but its presence was not associated with increased levels ofprotein. Alterations in the 5'UTR were found more likely to be associated with increasedprotein levels.</p>

Glioblastoma multiforme is a malignant brain tumor characterized by heterogeneity.Interactions between heterogeneous tumor cells are supposed to affect the behavior of awhole tumor cell population. In this thesis an in vitro model system of clonal glioma celllines originating from one glioblastoma tumor was used, and the behavior of cells incocultures was studied and compared the behavior of cells grown separately. The resultsindicate the presence of two types of interactions. In one, paracrine signals acted via extra-cellular media. This was associated with increased growth of the whole co-culture followedby a selective force driving one clone to dominance. In the other type, the cell clones grewside by side without signs of paracrine signalling, in a balance resulting in an increasedterminal cell density. Further investigations focused on mechanisms of interactions in thiscombination. Two cell clones were chosen, a GFAP+ and a GFAP-, for further experiments. Withdifferential display PCR it was possible to investigate their specific gene expressionpatterns. Seventeen cDNA fragments were differentially expressed, among them twocorresponded to known transcription factors, ATF3 and prox-1, one to a cytoskeletal protein,α-tropomyosin. The collection also contained eight ESTs (Expressed Sequence Tags) wherethe corresponding genes are unknown at present. Expression of the isolated sequences werealso analyzed in a panel of 12 different glioma cell lines and the results illustrate thecomplexity of gene expression and of tumor heterogeneity. Genes, the expression levels ofwhich were modulated in co-cultures and/or were cell density dependent, were alsoidentified. PDGF B is suggested to play a role in sarcomas. The gene codes for an mRNA transcriptwith long UTRs, parts of which are deleted in the homologous oncogene v-sis. The UTRs ofPDGF B mRNAs in human sarcomas were investigated for deletions similar to v-sis thatmight result in increased protein levels. A new transcript variant was identified, lacking a149 base region in the 3'UTR, but its presence was not associated with increased levels ofprotein. Alterations in the 5'UTR were found more likely to be associated with increasedprotein levels.

Les macrophages (MΦ), des cellules clefs de la réponse immunitaire peuvent répondre à des contaminants environnementaux comme les particules diesel (DEP), des polluants atmosphériques récemment classés cancérigènes pour l'Homme. Les MΦ sont des cellules hétérogènes et plastiques qui s'activent en fonction de leur microenvironnement soit en MΦ M1 (dits classiquement activés ou pro-inflammatoires) sous l'effet de l'INFγ et du LPS soit en MΦ M2 (dits alternativement activés ou réparateurs) sous l'effet de l'IL-4 et/ou de l'IL-13. Les effets des DEP sur la polarisation M1/M2 des MΦ restent peu documentés. Nous avons dans un premier temps caractérisé l'expression des marqueurs des MΦ différenciés in vitro en présence de M-CSF à partir de monocytes humains et polarisés en sous-type M1 ou M2. Nos principaux résultats montrent que les MΦ différenciés au M-CSF considérés comme des MΦ anti-inflammatoires, sont en réalité capables de s'activer vers un phénotype M1 après une stimulation au LPS/IFNγ. De plus, les marqueurs mis en évidence au cours de ce travail ont permis d'évaluer l'impact d'extraits organiques de DEP (DEPe) sur la polarisation des MΦ et plus généralement sur leur physiologie. Les DEPe altèrent l'expression de certains marqueurs M1 et M2 des MΦ, sans toutefois provoquer d'inhibition globale des processus de polarisation M1 et M2 ou de transition d'un phénotype vers un autre. Cette altération du phénotype est associée à une diminution de la réponse inflammatoire LPS-dépendante dans les MΦ M1 et des capacités chimiotactiques des MΦ M2. Les DEPe diminuent la sécrétion de certaines cytokines et chimiokines comme l'IL-6, l'IL-12p40 et le CCL18 via l'activation d'AhR et/ou de Nrf2. Parallèlement, nous montrons que les MΦ M1 et M2 exposés aux DEPe sécrètent le platelet deried growth factor B (PDGF-B), un facteur de croissance profibrosant, via l'activation d'AhR en quantité suffisante pour stimuler la prolifération de fibroblastes pulmonaires. Au total, ces travaux démontrent que les DEP possèdent des propriétés immunotoxiques vis-à-vis de la physiologie des macrophages humains polarisés in vitro. Cette immunotoxicité pourrait participer aux effets délétères de ces contaminants environnementaux urbains sur la santé humaine. / Macrophages (MΦ), well-known to play a key role in immune response, also respond to environmental toxic chemicals such as diesel exhaust particles (DEP), an air pollutant recently classified as carcinogenic to humans. MΦ are heterogeneous and plastic cells which activate according to their microenvironment into either an M1 subtype (so called classically activated or pro-inflammatory) under IFNγ and LPS stimulation or an M2 subtype (so called alternatively activated or anti-inflammatory) under IL-4 and/or IL-13 stimulation. However, potential effects of DEPs on M1/M2 MΦ polarization remain poorly documented. First, we characterized the expression marker of in vitro M-CSF-differentiated MΦ from human monocytes and activated into the M1 or M2 subtypes. Our main results show that M-CSF-generated MΦ considered as anti-inflammatory are actually able to switch to an M1 phenotype after IFNγ/LPS stimulation. Furthermore, the markers identified in this study were used to assess the impact of organic extracts of DEP (DEPe) on MΦ polarization and more generally on their physiology. DEPe alter some M1 and M2 markers expressed by polarized MΦ, without causing the overall inhibition of the M1 and M2 polarization process or the switch to a different phenotype. This phenotype alteration is associated with a decrease in the LPS-dependent inflammatory response in M1 MΦ and the chemotactic capacities in M2 MΦ. DEPe decrease the secretion of some cytokines and chemokines such as IL-6, IL-12p40 and CCL18 via AhR and/or Nrf2 activation. At the same time, we show that M1 and M2 MΦ in response to DEPe are able to secrete a sufficient level of a pro-fibrotic growth factor, the platelet derived growth factor B (PDGF-B) via AhR activation, leading to stimulation of lung fibroblast proliferation. Finally, these works show that DEPe have immunotoxic properties with regards to the physiology of human in vitro polarized MΦ. This immunotoxicity may then contribute to the deleterious effects of these urban environmental contaminants on human health.

Pre-mRNA splicing is an important level in posttranscriptional gene regulation that is essential for accurate protein synthesis and generating protein diversity. The abundance of cryptic splice sites and long intronic DNA sequences makes their splicing a complex one. The identification of correct exons and introns needs additional information in the form of splicing regulatory elements (SREs) along with canonical splice signals. The interplay among these SREs and the trans factors (which bind to SREs) gives the identity to introns and exons which in turn leads to precise pre-mRNA splicing. Previous studies from our laboratory showed, that when expressed in mammalian cells from an expression vector, PDGF-B ORF was re-spliced at 4/5 exon junction with the downstream SV40 splice acceptor site in the vector. However, deletion of the 66-nt PDGF-B 3’ UTR region resulted in about 25% reduction in re-splicing. Sequence analysis of this region revealed presence of binding sites for splicing factors ASF/SF2 and SRp55, and an AU-rich element (ARE), mutation each of which affected re-splicing partially. In mammals, AREs are commonly found in the 3’UTR of mRNAs encoding proteins involved in diverse functions and are involved in selective mRNA degradation. Several ARE binding proteins are crucial for ARE’s function. Since mutation of the single ARE in the 3’UTR region altered the re-splicing efficiency, the role of AU-rich elements and ARE-binding proteins (AU-BPs) in modulation of splicing was investigated using siRNAs against AU-BPs, BRF1, hnRNPD, HuR, GAPDH and TTP. Down regulation of expression of these factors indeed affected the level of re-spliced product. We have studied the interactions between the full-length splicing factors (U1-70K and U2AF35) and the AU-BPs (BRF1, hnRNPD and HuR) as well as among the AU-BPs using three different assay methods: Yeast-two hybrid, co-immunoprecipitation and pull down assays. Our study has revealed that the BRF1 interacts with U1-70K and U2AF35 as well as the other AU-BPs hnRNPD and HuR but with different affinities. We have also analyzed the ability of AU-BPs to interact with SR proteins SRp20 and 9G8. We did find strong interaction of BRF1 with SRp20 and 9G8. Generation of a large number of nested deletion mutants of all the proteins allowed us to identify the interaction regions on the surface of BRF1, U1-70K, hnRNPD, U2AF35 and HuR. The results of Y2H analyses were further confirmed by pull down assay using purified interacting regions. It was found that a single region from aa 181-254 in BRF1 interacts with multiple partners i.e., splicing factors and the AU-BP hnRNPD. However, the RNA-binding zinc-finger domain from residue 120-181 independently interacts with HuR. Further, the multiple protein interacting region (MPIR) (aa 181-254) in BRF1 exhibits different affinities towards its interacting partners with that for U1-70K and hnRNPD being stronger than that for U2AF35 and HuR. This observation suggests that BRF1 activity can be modulated by interaction with different partners at different sites. U1-70K interacted only with BRF1 among the proteins tested in this study and this interaction appears to be RNA independent .This could have implications in splice site selection and RNA stability since BRF1 has been shown to promote RNA degradation. While the Arg/Glu-rich C-terminal region in U1-70K is sufficient for its interaction with BRF1, U2AF35 requires both the zinc-finger 2 and the arg/Gly/Ser-rich C-terminal regions for its association with BRF1. hnRNPD also interacts with multiple partners that include BRF1, HuR and U2AF35 using the N-terminal region that harbors a Ala-rich domain. The interaction of hnRNPD with HuR is RNA dependent while with BRF1 and U2AF35, it is RNA independentt. Further, its interaction with all the partners is equally strong. This suggests that hnRNPD could exert differential influence depending on the context of its interaction and abundance of the interacting partner. HuR, primarily known as an mRNA stabilizing factor, interacts with both BRF1 and hnRNPD with equal affinity involving the hinge region, the interaction with the former being RNA-independent and the later being RNA-dependent. This differential RNA-dependent and independent interactions with the two AU-BPs using a single interacting domain suggests a balancing act of HuR on the activities of BRF1 and hnRNPD. These interactions can further be differentially modulated by posttranslational modifications on one or all of the interacting partners depending on the physiological status of the cell. We have also analyzed the multiple protein complexes formed in absence of cellular RNA. Though we are unable to see direct protein-protein interaction between HuR and U1-70K in Yeast two hybrid analysis, we could detect the presence of U1-70K in HuR immunoprecipitate. It appears that U1-70K associates with HuR via BRF. We also detected the presence of HuR in U1-70K complexes which could be due to its association with BRF1. We are unable to find hnRNPD and U2AF35 in these complexes indicating that they may have been excluded. In anti-U2AF35 immunoprecipitates, we detected the presence of U1-70K as well as hnRNPD but no HuR. This may be due to RNase treatment as hnRNPD and HuR interactions are RNA dependent. Our findings that AU-rich elements in conjunction with AU-BPs function as intronic splicing modulators or enhancers, reveal hitherto unidentified new players in the poorly understood complex mechanisms that mediate alternative splicing. The possibility of dynamic nature of the interactions among splicing factors and AU-BPs mediated by post-translational modifications provide a basis for rapid cellular responses to changing environmental cues through generation of differentially spliced mRNAs and corresponding protein products that differ in their stability and hence their relative abundance. Our results also unfold enormous possibilities for future investigations on interactions among the many splicing factors and AU-BPs, and in understanding these complex interactions in modulation of pre-mRNA splicing, mRNA translation and degradation. The finding of coupling of AU-BPs to splicing machinery could further lead to better understanding of the mechanism of AU-BP-mediated targeting of mRNAs to processing bodies and ultimate degradation of the mRNAs.