Livrables
We have extensively studied Hirschsprung disease (HSCR) as a model complex genetic disorder. In particular,
(1) genetic analysis at the RET locus has been carried out to identify variants conferring a modified risk of HSCR development.
We have confirmed the view that RET variants defining common haplotypes act as low penetrant alleles in the development of the disease, even in the absence of clear causative mutations, by SNPs genotyping, haplotype reconstruction and search for RET functional/hypomorphic variants. A high LD in the region, defining two distinct blocks, and a clear association of the 5 region of the gene with HSCR disease have been reported. In addition, our two groups participated to an International HSCR Consortium aiming to genotype a number of SNPs across the RET locus and using the transmission disequilibrium test (TDT) to determine transmission from heterozygous parents to affected offspring. In this study the relative contributions of CDS mutations and of an enhancer mutation at risk of HSCR already described have been examined with respect to various HSCR-associated features in a total of 1,000 patients (as trios) providing a consistent explanation of the genetic features of a classical multifactorial disease. In addition, we have obtained results suggesting that the HSCR predisposing haplotype is associated with low RET gene expression, a finding in agreement with the overrepresentation and non random transmission of this haplotype to HSCR patients.
We have also investigated a HSCR protective RET haplotype finding association with two nucleotide substitutions located nearby the third and fourth polyadenylation sites. Data obtained in vitro, using proper constructs containing the 3 terminal sequences of the RET gene, clearly demonstrate that one of these variant alleles interferes with normal RET mRNA half-life thus allowing to conclude that a more stable RET mRNA can lead to an increase of endogenous transcript, and probably of the amount of total protein produced, and providing a consistent explanation of the protective effect previously described.
(2) molecular characterization of the genetic defect underlying a HSCR associated phenotype: the Congenital Central Hypoventilation Syndrome (CCHS)
Between our two groups, we have collected so far a total of more than 180 CCHS patients, 28% of whom showing association with HSCR. Mutation screening of the PHOX2B gene has allowed to detect heterozygous frameshift mutations and in-frame changes leading to 5 to 13 polyalanine expansions of the PHOX2B gene, vastly occurred de novo, in over 85 % CCHS patients studied, thus concluding that PHOX2B is the major gene in CCHS pathogenesis, with autosomal dominant inheritance and reduced penetrance. In addition, we have shown i) a particular susceptibility to neuroblastoma in CCHS patients with truncating mutation of the PHOX2B gene as opposed to those with polyalanine expansion mutations, ii) the possibility of germline PHOX2B gene mutation predisposing to neuroblastoma independent of the CCHS phenotype and that iii) the two sets of PHOX2B mutations found in CCHS patients play different roles in the transcriptional regulation of different target genes. In particular, we have shown a correlation between the length of polyalanine expansions and the severity of reduced transcriptional activity caused by retention of the mutated protein in the cytoplasm or in the nuclear aggregates. On the other hand, frameshift mutations did not impair the PHOX2B nuclear income, suggesting a different mechanism through which they would exert the observed effects on target promoters
(3) functional reconstruction of the gene network underlying correct development of Autonomic Nervous System as an approach to identify novel candidates for ENS dysfuntions and/or HSCR modifier genes
We aimed to isolate factors regulating expression and targets of two homeobox genes, Tlx2 and Tlx3, involved in the development of branches of autonomic nervous system, and to identify genes cooperating in the RET transcriptional regulation. In particular, i) transient transfections, electrophoretic mobility shift and chromatin immunoprecipitation assays have all indicated that PHOX2B is able to bind sequences of a cell-specific element in the 5-flanking region of the human TLX2 gene, determining its transactivation in neuroblastoma cells, and confirming a physiological role of the PHOX2B-TLX2 promoter interaction in the transcription factors cascade underlying the differentiation of autonomic neuronal lineages during human embryogenesis, ii) after identifying the transcription initiation points, we have demonstrated that the 5-untranslated region of the TLX3 gene is necessary for the basal promoter activity in cell lines from different origin. Moreover, two tandem CCAAT boxes have been localized as critical elements, also showing their interaction with the nuclear factor Y (NFY).
This part of the project concerned the study of mouse intestinal aganglionosis by evaluating the roles of RET and interacting molecules in the enteric nervous system development mainly through the analysis of their expression and proliferative/differentiative activities in neural crest derivatives obtained from wild type and RET C620R mutated mice.
Study of retC620R/C620R mice:
In order to address whether the intestinal aganglionosis in retC620R/C620R pups is a consequence of a growth defect of the Ret-expressing cells, neural crest cells (NCCs) destined to colonize the bowel were isolated from 8.5-9.0 d.p.c. embryos, cultured in vitro, and labeled with an anti-Ret antibody. Fluorescent immunohistochemistry and ultrastructural analysis by electronic microscopy demonstrated a significantly lower percentage of Ret-positive cells in ret C620R/C620R primary cultures compared to controls. Furthermore, in ret+/+ and ret+/C620R cells, the Ret-specific signals are often found in vesicles, near the cell membrane or in the cytoplasm. In the retC620R/C620R cells only a faint labeling was observable, restricted to the cytoplasm near neurotubules and endoplasmic reticulum. Notably, while the Ret-specific signals were observed in clusters on the cell membrane of control cells, the retC620R/C620R cells were consistently devoid of membrane labeling.
This result supports the in vitro biochemistry findings that ret C620R impairs maturation of the Ret protein and prevents it from being transported to the cell membrane, where its signal transduction function is effected (Yin et al. under revision by the Intl. J. Cancer).
Possible RET pathway interacting molecules:
During mice breeding RET C620R mouse carriers demonstrated very low fertility. Since RET has been shown to be expressed in testis, the low fertility of mutated mice could reveal a feature specific of the RET C620R mutation and linked to RET function during spermatogenesis.
Cytoskeletal components have been involved in differentiation/migration of NCC derivatives upon RET activation. TSGA10 is a new gene with putative functions in cell division during spermatogenesis and neurogenesis which has been recently proposed to have a role in neural crest development based on embryonic expression data. We demonstrated the presence of TSGA10 in mouse primary neural crest cultures at both RNA, by RT-PCR, and at the protein level, by immunocytochemistry (Behnam, Modarressi, Conti, Taylor, Puliti, Wolfe. Biochem Biophys Res Commun. 2006 344(4):1102-10).
Recent genetic and molecular data suggest the interaction between RET and EDNRB pathways and indicate that both are essential for the proliferation/differentiation of enteric NCCs. Although it is now clear that RET and EDNRB pathways control the enteric innervation development, the molecular mechanisms that mediate such control remain to be fully understood. We used a neuroblastoma cell line as model of NCCs in order to clarify the role of GDNF and ET3 on the fate of RET positive cells and on possible effects of RET and EDNRB pathways interaction on cell proliferation and differentiation. We verified expression of Ret, Gdnfra, and Ednrb in the neuroblastoma cells by RT-PCR analysis, and confirmed the presence of the RET protein by immunocytochemistry.
We used single or double treatments with GDNF and/or ET3 in synthetic medium in triplicated experiments. Cell growth was evaluated by counting cells obtained at the end of a 2/3 days treatment. Cell differentiation was evaluated by a detailed analysis of cell morphology by phase-contrast microscopy. A statistically significant increase of proliferation was observed in cell cultures after 2 days treatments with both GDNF and ET3. Cell morphology analysis indicated an increase of undifferentiated cells and of cells with only one neurite-like process (monopolar). Three days treatments of cells showed a statistically significant increase of differentiated cells after GDNF/ET3 double treatment.
Overall our data indicate a synergic effect of GDNF and ET3 on undifferentiated and monopolar cells proliferation as well as on the branching processes cells differentiation. These results are in agreement with the hypothesis of an interaction between the GDNF/RET and ET3/EDNRB pathways and confirm the use of the cell line as useful tool to further investigate on the complex relationships between RET and EDNRB molecular pathways.
Germline RET mutations are responsible for different inherited disorders: Hirschsprung disease (congenital aganglionic megacolon), caused by loss of function mutations, familial medullary thyroid carcinoma (FMTC) and multiple endocrine neoplasia type 2 (MEN2A), caused by gain of function mutations. Intriguingly, some RET mutations, including C620R, are associated with both types of diseases. In order to investigate the dual role of such RET mutations, a mouse model with a targeted mutation retC620R was generated. retC620R/C620R offspring die during the first postnatal day, and show kidney agenesis and intestinal aganglionosis.
Decreased outgrowth of the Ret-positive cells was observed in retC620R/C620R neuronal cell cultures, which is suggestive of an impaired migration, proliferation, or survival of the Ret-expressing cells. Electronmicroscopy revealed the absence of membrane-bound Ret in retC620R/C620R cells as compared to ret+/+ and ret+/C620R cells. On the other hand, aged ret+/C620R mice develop pre-cancerous lesions in the adrenal gland or in the thyroid. In humans, RET+/C620R is associated with FMTC and MEN 2A. In mice, however, the same genotype does not produce the above phenotypes.
It is worth noting that mice carrying the heterozygous MEN2B-specific mutation M918T (ret+/M918T) obtained by replacing the endogenous wild-type copy of the gene by ES mediated homologous recombination, display merely C-cell hyperplasia (CCH; 55%) and nodular chromaffin cell hyperplasia (16-17%). In addition, when the M918T mutation was targeted to be expressed under the calcitonin gene promoter specifically in thyroid C-cells, in transgenic mice, nodular CCH was observed only in some mice from 8 months on and MTC was detected in 13% of mice from the age of 11 months on. However, no thyroid dysplasia was observed in transgenic mice carrying the wild-type ret allele or non-transgenic controls of same age. MEN 2B manifests a more complex phenotype than MEN 2A and RETM918T has an in vitro tumorigenic capacity much higher than RETC620R. It is therefore not surprising that the C620R mutation in heterozygous mice is not fully capable of transforming the neural crest derivatives.
An increase in C-cell hyperplasia, was observed in 48.4% of ret+/C620R mice of 129/Sv background versus the 22.8% of ret+/+ mice. Although the differences in frequency of C-cell hyperplasia between the two groups do not reach statistical significance (P= 0.077), probably because of the relatively small number of observations allowed by the material available, they suggest that the C620R mutation is predisposing to tumor formation only at late stages in mice. In a recent work by Carniti and coworkers no C-cell hyperplasia or tumor formation was detected in heterozygous mice carrying the same C620R mutation, which were analysed at earlier stages, whereas results similar to ours were reported for the aganglionosis found in the homozygous mice.
It is quite interesting that ret+/C620R mice, in both the129/Sv and 129/Sv X C57BL/6 backgrounds, develop thyroid dysplasia and adenoma, and adrenal gland cortical nodular regeneration or hyperplasia. The above result indicates that the retC620R mutation also causes susceptibility to some cancers of epithelial origin, which are not associated with the human equivalent genotype. How the retC620R transforms the mouse epithelium, which is normally devoid of Ret expression, remains to be elucidated.
In conclusion, we have observed that (1) the retC620R mutation in homozygotes displays a loss of function effect, manifested by the lack of mature Ret protein on the cell membrane, a growth defect of the retC620R/C620R neuronal primary cultures, and consequent total intestinal aganglionosis and kidney agenesis; (2) the retC620R mutation in heterozygotes exerts a possible dominant gain of function effect with low penetrance, which might lead to pre-cancerous lesions in the thyroid and/or adrenal gland.
The low penetrance of these pre-cancerous lesions in ret+/C620R mice suggests that there is a requirement for additional factors, genetic as well as environmental, in order to develop the complete tumour phenotype. Therefore, further studies of ret+/C620R mice will hopefully provide clues for the identification of these additional components.
We investigated whether RET-induced cell death was dependent on one of the two classical pathways for apoptosis induction (i.e. the death receptor pathway involving caspase-8 activation or the mitochondria-dependent pathway requiring cytochrome c release and caspase-9 activation) and demonstrated that RET induces cell death through an independent pathway requiring caspase-9 probably by forming a caspase-activating complex. In meantime, we are now interested in the importance of caspase-2 in this caspase activating complex.
The biggest part of the work performed during project period was dedicated to the relevance in vivo of the pro-apoptotic function of RET. As stated in the project, P. Mehlen in collaboration with G. Romeo was initially supposed to generate knock-in mice in which RET would have been mutated in its caspase site, a mutation that blocks RET-induced pro-apoptotic activity. The construct for homologous recombination was prepared using the initial construct used by G. Romeo to generate its C620R knock-in mice and further sequenced for verification.
However, while performing the homologous recombination in ES cells, this part of the project was stopped because P. Mehlen was contacted by Dr. Costantini (Columbia University, USA) that has already generated the mice and observed a very interesting phenotype in homozygous D707N/D707N. Indeed, as expected the number of enteric neurons is strongly increased in the small intestine while the distal colon shows aganglionosis. To further demonstrated that apoptosis is required for the appropriate development of the enteric nervous system, P. Mehlen decided to go for an alternative model in chicken. Indeed, P. Mehlen, in collaboration with N. Le Douarin (Collège de France, France), have electroporated neural crest cells from quail embryos with a dominant negative mutant of RET, -i.e. the intracellular cellular domain of RET mutated in D707N behaves like an inhibitor of RET-induced cell death (see the recent paper from P. Mehlen, Thibert et al., 2003, Science, 301, 84-846). The electroporated quail neural crest cells were then introduced in chicken embryos and the migration of quail cells were then followed in the chicken embryos. P. Mehlen has then observed that a delay in neural crest cells migration occurs in electroporated embryos. This delay in a large number of embryos is associated with a dramatic phenotype the in enteric nervous with either ectopic plexi or a complex absence of neuronal differentiation. P. Mehlen is now performing a few sets of accessory experiments i.e., trying to show that the dominant negative electroporated neural crest cells are more resistant to apoptosis that the mock electroporated when cultured ex-vivo- before trying to publish this approach in back-to-back manuscripts together with the mice approach of Dr. Costantini.
P. Mehlen also initiated an alternative approach to show the in vivo importance of RET-induced cell death. In collaboration with Dr. Arumae in Helsinki, P. Mehlen has taken advantage of the powerful system of primary sympathetic neurons that can be committed to apoptosis via withdrawal of GDNF. Up to now, the overall idea says that this death commitment is related to a loss of survival signals (PI3K, MAPK) normally provided by the canonical GDNF/RET pathways resulting in a by default cell death. However, P. Mehlen took advantage of the dominant negative mutant of RET (see above). Microinjection of the dominant negative mutant was performed in GDNF-withdrawn neurons and cell death was analyzed.
While the injection of this dominant negative mutant has no effect on cell death observed after NGF withdrawal, this mutant strongly inhibited the death of sympathetic neurons withdrawn in GDNF, thus demonstrating that the so-called death by default observed after GDNF loss is related to the pro-apoptotic effect of RET. This is an important observation in regards to the classic cancer therapy strategy that deals with inhibiting kinase activity to induce cancer cell death i.e. see Glivec with c-Kit-. This result suggests that in the case of RET, at least, inhibiting kinase activity will probably not be sufficient to push RET-mutated tumor cell to undergo apoptosis. Thus inhibiting kinase activity should cytostatic but not cytotoxic. If so therapeutic strategy should focus on molecules affecting both the kinase activity but also the pro-apoptotic face of tyrosine kinase receptors.
Mice of both inbred strains CBA/J and C57/Bl, were analysed, with the COMET-FISH technique specifically designed to detect presence and amount of genomic fragmentation of RET, c-Abl, and p53 genes, following 0, 1, 2 or 4 Gy of X rays exposure. Analyses were performed on PBL withdrawn 30 , 24 hours, and 30 days after exposure, and compared to data coming from COMET determination on the same samples.
Thyroid tissues and whole body examinations of both strand of mice (eight mice for each) were taken from long term 3 Gy irradiated animals.
COMET values were determined by the dedicated freeware software CASP 2.22 (Comet Assay Software Project, University of Wroclaw, Poland). Data have been scored as the mean +/-95% confidential intervals (95% CI), and the 90 percentile algorithm has been applied to the tail moment (TM), percentage of DNA content on tail (%DNA), and tail length (TL). In this study, we stated that only difference between untreated vs. irradiated PBL were discriminated by mean of COMET assay, and no statistical dose response effects can be further recognizable. Moreover, analyses delivered on single mice confirmed both an inter-individual variability and a fluctuation of experimental procedures that hamper the use of COMET for in vivo experiments.
Data coming from experimental points taken from animals sacrificed 24 hours and 30 days after exposure were undistinguishable from controls, thus confirming prompt DNA repair and clone selection of viable cells. On the contrary, COMET-FISH technique specifically directed to detect presence and amount of genomic fragmentation of RET, c-Abl, and p53 genes was able to discriminated in a statistically significant way all doses delivered, when fragmentations were analyzed contextually for RET, cAbl, and TP53 genes. These data were confirmed by Laser confocal microscopy determination. Data coming from experimental points taken at the later points of 24 hours and 30 days after exposure were undistinguishable from controls, probably for viable clone selection, and a prompt DNA repair.
One year following 3 Gy exposure, no evidence of RET cAbl and p53 genes positive fragmentation resulted from thyroid and bone marrow tissues by COMET-FISH. Accordingly, COMET assay alone was absolutely negative to detect any difference between controls and long term irradiated mice. Histochemistry carried out on thyroid samples gave no evidence of tumor progression, both in C57Bl/6 and CBA/J strain of mice, thus hamper further investigation, at least on the basis of this experimental protocol.
Mechanisms of stress-induced activation of Rai. 1.1 We have analyzed the effects of H2O2 (an inducer of oxidative stress) and CoCl2 (an inducer of cellular hypoxia) on Rai protein expression using primary neuronal mouse cortical cultures. Notably, H2O2 and CoCl2 treatments consistently induced an acceleration of the mobility (shift-down) of endogenous Rai proteins in polyacrilamide gels. The shift down of Rai polypeptides appeared at late time points after treatment how 1 hr and 2 hr for H2O2 and CoCl2 respectively and returned to the basal levels within 24 hours. The same modification was also detected in a transformed cell line (PFSK1) and in vivo in different areas of the brain mouse following 15 minutes of ischemia and 24 hours of reperfusion. In vitro 32P labeling experiments and usage of dual specificity or tyrosine-specific phosphatases indicated that the stress-induced gel-shift of Rai proteins is due to dephosphorylation of serine-threonine residues. Rai contains 40 conserved serine and threonine residues in the CH1 region. Using site-directed mutagenesis we have restricted the putative phosphorylation-sites to two serine residues. These two serines are predicted to be phosphorylation sites for the CaMKII (Ca2+/calmodulin-dependent protein kinase).
CaMKII is enriched in neuronal tissue (up to 2% of total protein) and highly concentrated in the post-synaptic densities and mediates a variety of different cellular response to Ca2+ influx. It is involved in synaptic plasticity and has a role in learning and memory. So far we have demonstrated that the kinase is able to phosphorylate in vitro Rai proteins. 2.2 Identification of Rai interactors. To characterize the Rai-signalling potential (and upstream effectors), we attempted purification of Rai-complexes. To this end we expressed N-terminally or C-terminally Flag-HA-epitope-tagged Rai at stable and low levels in the neuroectoderma cell line PFSK-1 endogenously expressing Rai.
We isolated protein complexes by affinity purification and analyzed them by mass spectrometry. The experiments have been performed in growing cells and after stress stimuli (H2O2 treatment and hypoxia). From the mass-spectrometry analysis we identified the protein PP2C. PP2C is a Mg2+ dependent serine/threonine phosphatase present in the brain. It is activated by cellular stress and is known to dephosphorylate CaMKII. By in vitro coimmunoprecipitation studies using specific antibodies we validated the Rai-PP2C association. Rai is constitutively bound to PP2C and the binding increases after 5 minutes of H2O2 treatment. We are investigating the protein domain of interaction between the two proteins. To demonstrate the role of Rai dephosphorylation on activation of its biological activity we are performing PP2C RNA interference in Rai expressing cells and we are evaluating the cellular resistance to stress stimuli.
The study of the molecular mechanisms involved in Rai-dependent induction of anti-apoptotic activity, the identification of the main phosphorylation sites and of the kinase/ phosphatase involved can identify Rai as a new potential diagnostic/prognostic marker and, eventually, a target of therapeutic intervention for neurodegenerative diseases. Moreover, the identification of other proteins that interact with Rai will improve our knowledge on which pathways are activated in different stress conditions, such as neuronal injury or neurodegenerative diseases.
We identified C. elegans homologues of three genes related to the RET signalling pathway: ECE1 homologue NEP1, SGK1 (serum- and glucocorticoid-responsive kinase) and T16A9.4, another NEP gene similar to ECE1. Knockout of nep-1 leads to reduced locomotion, when compared to wild type worms, suggesting a neuronal dysfunction of the ECE1 mutant (Spanier et al., manuscript is in preparation). SGK1 acts downstream of PDK-1.
Loss of SGK-1 in C.elegans results in extended generation time, increased life span, and, in combination with mutations in the Akt/PKB homologs AKT-1 and AKT-2, causes constitutive developmental arrest. The SGK/AKT/PKB complex characterized is a downstream effector of RET/Rai/GDNF signalling. We generated a model that can be used to study the downstream effectors of RET/RAI signaling, and the genes regulated inappropriately in mutations of RET/RAI.
Furthermore, interaction studies have been conducted to identify RET interacting partners using a split-ubiquitin yeast interaction screening system and a human interactor libary.
A semi-automated screening platform was established to screen for candidate genes involved in the RET pathway in high-throughput. This platform uses an automatic worm-sorting device to distribute, analyze and redistribute C. elegans test animals in a microtiter format. Several phenotypic parameters were established that allow the quantitative screening of mutational consequences in factors of the RET/RAI pathway.
Recherche de données OpenAIRE...
Une erreur s’est produite lors de la recherche de données OpenAIRE
Aucun résultat disponible