Hyperacute Infarct on Intraoperative Diffusion Imaging of Pediatric Brain Tumor Surgery.

BACKGROUND: Brain neoplasms are the second-most prevalent cancer of childhood for which surgical resection remains the main treatment. Intraoperative MRI is a useful tool to optimize brain tumor resection. It is, however, not known whether intraoperative MRI can detect complications such as hyperacute ischemic infarcts. METHODS: A retrospective analysis of pre- and intraoperative MRIs including DWI sequence and correlation with early and 3-month postoperative MRIs was conducted to evaluate the incidence of hyperacute arterial infarct during pediatric brain tumor resection. Patient demographics, pathological type, tumor location, resection type as well as preoperative tumoral vessel encasement, evolution of the area of restricted diffusion were collected and analyzed comparatively between the group with acute infarct and the control group. Extent of the hyperacute infarct was compared to both early postsurgical and 3-month follow-up MRIs. RESULTS: Of the 115 cases, 13 (11%) developed a hyperacute arterial ischemic infarct during brain tumor resection. Tumoral encasement of vessels was more frequent in the infarct group (69%) compared to 25.5% in the control group. Four cases showed additional vessel irregularities on intraoperative MRI. On early follow-up, the infarcted brain area had further progressed in six cases and was stable in seven cases. No further progression was noted after the first week post-surgery. CONCLUSIONS: Hyperacute infarcts are not rare events to complicate pediatric brain tumor resection. Tumoral encasement of the circle of Willis vessels appears to be the main risk factor. Intraoperative DWI underestimates the final extent of infarcted tissue compared to early postsurgical MRI.

Présence d'infarctus ischémiques hyper-aigus révélés par des IRM peropératoires de diffusion dans le cas de chirurgies visant à traiter le cancer du cerveau chez l'enfant.Contexte: Chez l'enfant, les néoplasmes du cerveau viennent au second rang des cancers pour lesquels la résection chirurgicale demeure le principal traitement. L'IRM peropératoire est un outil efficace pour maximiser la résection d'une tumeur cérébrale. On ignore toutefois de quelle façon cet outil peut permettre de détecter des complications telles que les infarctus ischémiques hyper-aigus. Méthodes:Nous avons tout d'abord effectué une analyse rétrospective des IRM préopératoires et peropératoires, ce qui inclut des IRM de diffusion. Nous avons aussi cherché à établir des corrélations avec des IRM post-opératoires, certaines effectuées tout juste après une résection et d'autres après 3 mois, afin d'évaluer l'incidence d'infarctus ischémiques hyper-aigus survenant en cours de résection. Des données portant sur les caractéristiques des patients, sur la pathologie tumorale, sur l'emplacement des tumeurs, sur le type de résection ainsi que sur l'envahissement vasculaire tumoral préopératoire et sur l'évolution de la région de diffusion restreinte ont été collectées. Nous avons ensuite effectué une analyse comparative entre notre groupe de patients et un groupe de témoins. L'ampleur des infarctus ischémiques hyper-aigus a été comparée au moyen d'IRM réalisées tout juste après une résection et d'IRM de suivi après trois mois. Résultats:Sur un total de 115 cas, 13 (11 %) ont donné à voir un infarctus ischémique hyper-aigu au cours d'une résection. L'envahissement vasculaire tumoral était plus fréquent dans le groupe de jeunes patients ayant donné à voir un infarctus (69 %) comparativement à 25,5 % dans le groupe des témoins. Quatre cas ont aussi montré, lors d'IRM peropératoires, des irrégularités additionnelles en ce qui a trait aux vaisseaux sanguins. Lors de suivis effectués rapidement après une résection, les régions du cerveau affectées par un infarctus avaient continué à croître davantage chez 6 cas et étaient demeurées stables chez 7 autres cas. Aucune autre progression n'a été notée à la suite d'une IRM menée une semaine plus tard. Conclusions:Les infarctus ischémiques hyper-aigus ne sont pas des événements inhabituels. Comme on le sait, ils peuvent entraîner des complications au moment d'effectuer la résection d'une tumeur cérébrale chez l'enfant. L'envahissement vasculaire tumoral du polygone de Willis semble être ici le principal facteur de risque. Les IRM de diffusion peropératoires ont tendance à sous-estimer l'étendue finale des tissus affectés par un infarctus ischémique comparativement aux IRM effectuées tout juste après une résection.

A direct interaction between two Restless Legs Syndrome predisposing genes: MEIS1 and SKOR1.

Restless Legs syndrome (RLS) is a common sleep disorder for which the genetic contribution remains poorly explained. In 2007, the first large scale genome wide association study (GWAS) identified three genomic regions associated with RLS. MEIS1, BTBD9 and MAP2K5/SKOR1 are the only known genes located within these loci and their association with RLS was subsequently confirmed in a number of follow up GWAS. Following this finding, our group reported the MEIS1 risk haplotype to be associated with its decreased expression at the mRNA and protein levels. Here we report the effect of the risk variants of the three other genes strongly associated with RLS. While these variants had no effect on the mRNA levels of the genes harboring them, we find that the homeobox transcription factor MEIS1 positively regulates the expression of the transcription co-repressor SKOR1. This regulation appears mediated through the binding of MEIS1 at two specific sites located in the SKOR1 promoter region and is modified by an RLS associated SNP in the promoter region of the gene. Our findings directly link MEIS1 and SKOR1, two significantly associated genes with RLS and also prioritize SKOR1 over MAP2K5 in the RLS associated intergenic region of MAP2K5/SKOR1 found by GWAS.

The planar cell polarity protein Vangl2 is required for retinal axon guidance.

Vangl2 plays a critical role in the establishment of planar cell polarity (PCP). Previously, we detected expression of Vangl2 in the developing retina during late embryogenesis, which led us to investigate the possible role of Vangl2-mediated PCP signaling in eye development. We have generated a Vangl2(BGeo) knock-in mouse allowing us to evaluate Vangl2 mRNA expression during retinal development, and used an isoform-specific antibody to examine Vangl2 protein expression in cryosections. To investigate the role of Vangl2 in retinal development, we examined eyes taken from embryos homozygous for independent alleles of Looptail (Lp, Lp(m1jus) ) mutant mice. We found that Vangl2 mRNA and protein are dynamically expressed in the developing embryonic and postnatal retina, with Vangl2 expression becoming progressively restricted to the ganglion cell layer and optic nerve as the retina matures. The expression pattern of Vangl2 transcript and protein is most prominent in retinal ganglion cells (RGC), and their axons. Additionally, we show that Vangl2 is required for retinal and optic nerve development as Vangl2 (Lp/Lp) mutant embryos display a significantly reduced eye size, marked thickening of the retina, and striking abnormalities in the morphology of the optic nerve (significant hypoplasia, and aberrant exit trajectory). Notably, we identified a salient intraretinal axon guidance defect in Vangl2 (Lp/Lp) mutant embryos through which axon bundles traverse the entire thickness of the retina and become trapped within the subretinal space. Our observations identify a new and essential role for Vangl2-dependent PCP signaling in the intraretinal path-finding of RGC axons.

Transmembrane topology of mammalian planar cell polarity protein Vangl1.

Vangl1 and Vangl2 are membrane proteins that play an important role in neurogenesis, and Vangl1/Vangl2 mutations cause neural tube defects in mice and humans. At the cellular level, Vangl proteins regulate the establishment of planar cell polarity (PCP), a process requiring membrane assembly of asymmetrically distributed multiprotein complexes that transmit polarity information to neighboring cells. The membrane topology of Vangl proteins and the protein segments required for structural and functional aspects of multiprotein membrane PCP complexes is unknown. We have used epitope tagging and immunofluorescence to establish the secondary structure of Vangl proteins, including the number, position, and polarity of transmembrane domains. Antigenic hemagglutinin A (HA) peptides (YYDVPDYS) were inserted in predicted intra- or extracellular loops of Vangl1 at positions 18, 64, 139, 178, 213, and 314, and individual mutant variants were stably expressed at the membrane of MDCK polarized cells. The membrane topology of the exofacial HA tag was determined by a combination of immunofluorescence in intact (extracellular epitopes) and permeabilized (intracellular epitopes) cells and by surface labeling. Results indicate that Vangl proteins have a four-transmembrane domain structure with the N-terminal portion (HA18 and HA64) and the large C-terminal portion (HA314) of the protein being intracellular. Topology mapping and hydropathy profiling show that the loop delineated by TMD1-2 (HA139) and TMD3-4 (HA213) is extracellular while the segment separating predicted TMD2-3 (HA178) is intracellular. This secondary structure reveals a compact membrane-associated portion with very short predicted extra- and intracellular loops, while the protein harbors a large intracellular domain.

Molecular and cellular mechanisms underlying neural tube defects in the loop-tail mutant mouse.

Loop-tail (Lp) mice show a very severe neural tube defect (craniorachischisis) caused by mutations in the Vangl2 gene (D255E, S464N). Mammalian Vangl1 and Vangl2 are membrane proteins that play critical roles in development such as establishment of planar cell polarity (PCP) in epithelial layers and convergent extension movements during neurogenesis and cardiogenesis. Vangl proteins are thought to assemble with other PCP proteins (Dvl, Pk) to form membrane-bound PCP signaling complexes that provide polarity information to the cell. In the present study, we show that Vangl1 is expressed exclusively at the plasma membrane of transfected MDCK cells, where it is targeted to the basolateral membrane. Experiments with an inserted exofacial HA epitope indicate that the segment delimited by the predicted transmembrane domains 1 and 2 is exposed to the extracellular milieu. Comparative studies of the Lp-associated pathogenic mutation D255E indicate that the targeting of the mutant variant at the plasma membrane is greatly reduced; the mutant variant is predominantly retained intracellularly in endoplasmic reticulum (ER) vesicles colocalizing with the ER marker calreticulin. In addition, the D255E variant shows drastically reduced stability with a half-life of approximately 2 h, compared to >9 h for its wild type counterpart and is rapidly degraded in a proteasome-dependent and MG132 sensitive pathway. These findings highlight a critical role for D255 for normal folding and processing of Vangl proteins, with highly conservative substitutions not tolerated at that site. Our study provide an experimental framework for the analysis of human VANGL mutations recently identified in familial and sporadic cases of spina bifida.

Cooperative interactions between the two DNA binding domains of Pax3: helix 2 of the paired domain is in the proximity of the amino terminus of the homeodomain.

Pax3 is a transcription factor that plays an important role during neurogenesis and myogenesis, and Pax3 mutant animals display neural tube defects and lack limb muscles. Pax3 harbors two DNA binding domains, the paired domain (PD) and a paired-type homeodomain (HD). Genetic and biochemical data have (i) identified strong cooperative interactions between the PD and HD domains for DNA binding in the intact Pax3 protein and (ii) suggested an important role for the amino-terminal portions of both domains in such cooperativity. We have studied proximity relationships between the PD and HD of Pax3. For this, we have used a cross-linking strategy with the bifunctional thiol reagent bismaleimidoethane (BMOE) in 21 mutants bearing pairs of cysteine residues (DCM) inserted in strategic locations of a functional Pax3 protein otherwise devoid of endogenous cysteine residues. All 21 DCMs were characterized for protein stability, for DNA binding by the PD and HD, and for the effect of BMOE on protein binding to PD, HD, or PD-HD combined DNA targets. BMOE-induced cross-links in DCMs were detected as slower migrating species on immunoblots. Mutants bearing double cysteine insertions (I59C/S222C, S73C/Q219C, and V78C/K218C) showed the most robust cross-linking upon BMOE exposure. These cross-linking studies suggest that portions of helix 1 (I59), helix 2 (S73), and the loop between helices 2 and 3 (V78) of the PD are in the proximity of the N-terminal segment of the HD (K218, Q219, and S222) in the tertiary structure of Pax3. These results are compatible with a model in which the PD and HD are organized in an everted arrangement, with the N-terminal portion of the PD being in the proximity of the N-terminus of the HD. This arrangement may be important for the noted PD-HD cooperativity in DNA binding.

The paired domain of Pax3 contains a putative homeodomain interaction pocket defined by cysteine scanning mutagenesis.

Pax3 is a transcription factor that plays an important regulatory role during neurogenesis, myogenesis, and formation of neural crest cell derived structures. Pax3 has two DNA binding domains, a paired domain (PD) and paired-type homeodomain (HD) that show complete interdependence for DNA binding, with mutations in one domain impairing DNA binding by the other domain. Cooperative interactions between the PD and HD of Pax3 suggest that the two domains may physically interact for DNA binding. Site-specific modification with thiol reagents in single cysteine Pax3 mutants was used to determine which segment of the PD may interact with the HD. Twenty-four single cysteine mutants were independently introduced in the second alpha-helix (alpha2, positions 59-80) and in the beta-hairpin structure (positions 40-41) at the amino terminal portion of the PD. These mutants were tested for their ability to bind to PD (P6CON, P3OPT) and HD-specific DNA targets (P2), and the effect of treatment with N-ethylmaleimide on these binding properties was established. In the PD, single cysteine mutants CL/Q40C, CL/I59C, CL/V60C, CL/P69C, CL/S70C, CL/I72C, CL/S73C, CL/L76C, CL/V78C, and CL/S79C displayed NEM sensitive DNA binding toward both PD and HD targets. Three PD mutants (CL/L41C, CL/A63C, and CL/H64C) showed unusual behavior, with DNA binding to PD targets being NEM insensitive while DNA binding by the HD was abrogated by NEM treatment. Three-dimensional modeling of the NEM sensitive PD cysteine mutants reveal that they are not randomly distributed, but rather that they cluster in a hydrophobic pocket. We propose that this hydrophobic pocket may serve as a docking site for the HD during DNA binding by the intact protein.

Cross-talk between the paired domain and the homeodomain of Pax3: DNA binding by each domain causes a structural change in the other domain, supporting interdependence for DNA Binding.

The Pax3 protein has two DNA binding domains, a Paired domain (PD) and a paired-type Homeo domain (HD). Although the PD and HD can bind to cognate DNA sequences when expressed individually, genetic and biochemical data indicate that the two domains are functionally interdependent in intact Pax3. The mechanistic basis of this functional interdependence is unknown and was studied by protease sensitivity. Pax3 was modified by the creation of Factor Xa cleavage sites at discrete locations in the PD, the HD, and in the linker segment joining the PD and the HD (Xa172, Xa189, and Xa216) in individual Pax3 mutants. The effect of Factor Xa insertions on protein stability and on DNA binding by the PD and the HD was measured using specific target site sequences. Independent insertions at position 100 in the linker separating the first from the second helix-turn-helix motif of the PD and at position 216 immediately upstream of the HD were found to be readily accessible to Factor Xa cleavage. The effect of DNA binding by the PD or the HD on accessibility of Factor Xa sites inserted in the same or in the other domain was monitored and quantitated for multiple mutants bearing different numbers of Xa sites at each position. In general, DNA binding reduced accessibility of all sites, suggesting a more compact and less solvent-exposed structure of DNA-bound versus DNA-free Pax3. Results of dose response and time course experiments were consistent and showed that DNA binding by the PD not only caused a local structural change in the PD but also caused a conformational change in the HD (P3OPT binding to Xa216 mutants); similarly, DNA binding by the HD also caused a conformational change in the PD (P2 binding to Xa100 mutants). These results provide a structural basis for the functional interdependence of the two DNA binding domains of Pax3.

Site-specific modification of single cysteine Pax3 mutants reveals reciprocal regulation of DNA binding activity of the paired and homeo domain.

The mechanism by which the paired domain (PD) and the homeo domain (HD) act together in the intact Pax3 protein to recognize DNA is unclear and was studied in a Pax3 mutant (Pax3-CL) devoid of cysteines. Pax3-CL binds to PD (P6CON-P3OPT sites) and HD (P2, P1/2 sites) DNA site sequences with near wild-type activity but, contrary to Pax3, in a N-ethyl maleimide (NEM) insensitive fashion. The Pax3-CL backbone was used for cysteine scanning mutagenesis and for site-specific NEM modification. Five single cysteine replacements were independently introduced in the PD, while eight were inserted in the HD. NEM sensitivity of PD and HD DNA binding was investigated in DNA-binding competent mutants. In the PD mutant C82, NEM abrogated DNA binding by the PD but also abolished DNA binding by the Cys-less HD. Likewise, in the HD mutant V263C, NEM modification abrogated DNA binding not only by the HD, but also by the Cys-less PD. The transfer of NEM sensitivity to the PD seen in V263C was specific and not due to simple loss of HD DNA binding since alkylation of adjacent V265C and S268C, although impairing HD DNA binding did not affect PD DNA binding. Thus, the PD and HD do not function as independent DNA binding modules in Pax3 but seem functionally interdependent.(1)