Decision letter: Norrin/Frizzled4 signalling in the preneoplastic niche blocks medulloblastoma initiation
Notice bibliographique
Résumé
Full text Figures and data Side by side Abstract Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract The tumor microenvironment is a critical modulator of carcinogenesis; however, in many tumor types, the influence of the stroma during preneoplastic stages is unknown. Here we explored the relationship between pre-tumor cells and their surrounding stroma in malignant progression of the cerebellar tumor medulloblastoma (MB). We show that activation of the vascular regulatory signalling axis mediated by Norrin (an atypical Wnt)/Frizzled4 (Fzd4) inhibits MB initiation in the Ptch+/− mouse model. Loss of Norrin/Fzd4-mediated signalling in endothelial cells, either genetically or by short-term blockade, increases the frequency of pre-tumor lesions and creates a tumor-permissive microenvironment at the earliest, preneoplastic stages of MB. This pro-tumor stroma, characterized by angiogenic remodelling, is associated with an accelerated transition from preneoplasia to malignancy. These data expose a stromal component that regulates the earliest stages of tumorigenesis in the cerebellum, and a novel role for the Norrin/Fzd4 axis as an endogenous anti-tumor signal in the preneoplastic niche. https://doi.org/10.7554/eLife.16764.001 Introduction The tumor microenvironment is comprised of many stromal cell types, including the vasculature, which is well known to promote the growth and propagation of established tumors (Calabrese et al., 2007; Hambardzumyan et al., 2008; Hanahan and Coussens, 2012). It has also become increasingly clear that construction of a tumor-permissive stroma is a dynamic process influencing all stages of malignant progression, through the ongoing extracellular matrix (ECM) remodelling and recruitment or activation of angiogenic vascular cells, fibroblasts and immune cells (Hanahan and Coussens, 2012; Hanahan and Weinberg, 2011). While there is mounting evidence that a permissive niche is required during the earliest steps of tumor initiation, studying this early crosstalk requires adequate multistage models of carcinogenesis (Barcellos-Hoff et al., 2013). Medulloblastoma (MB), the most common malignant brain tumor in children, has become a paradigm for the study of pediatric tumors and primary brain tumors; however little is known about tumor/stromal communication in the early stages of MB. It is challenging to model these early events with human tumor biopsies or MB cell lines, as both represent late stage tumors and the latter frequently fail to maintain their in vivo characteristics (Sasai et al., 2006). An effective surrogate to test these interactions endogenously is the Ptch+/− mouse (Goodrich et al., 1997), a model of the human predisposition to MB (Gorlin syndrome) that, along with a subset of sporadic MB, belong to the Sonic hedgehog (Shh) subgroup (Taylor et al., 2012). Ptch+/− mice progress through well-defined stages of tumorigenesis in the cerebellum, beginning with the ectopic proliferation of granule neuron progenitor cells (GNPs), which form preneoplastic lesions on the cerebellar surface by two weeks of age. While most of these lesions regress, a minority undergo malignant transformation to MB (Kessler et al., 2009; Oliver et al., 2005). The environmental signals that co-operate with Ptch haploinsufficiency to regulate lesion induction and transformation are poorly understood. Here, we modelled pre-tumor/stromal crosstalk during Ptch+/− MB evolution by manipulating the Norrin/Frizzled4 (Fzd4) pathway, an endogenous signalling axis that regulates vascular development in the cerebellum via neural/endothelial interactions (Wang et al., 2012; Xu et al., 2004; Zhou et al., 2014). We demonstrate that the preneoplastic niche is a potent modulator of Ptch+/− tumor initiation. Loss of vascular Norrin/Fzd4 signalling, either genetically or by short-term blockade, creates a tumor-permissive stroma that promotes the formation of preneoplastic lesions and their progression to malignancy. We show that activation of angiogenesis and stromal remodelling are key features of the oncogenic microenvironment that dramatically accelerates tumorigenesis in the Ptch+/ cerebellum. This is the first study to describe a stromal component to the early stages of Shh-driven tumor initiation in the brain. Results Ndp is expressed in GNPs and mouse and human Shh-MB To assess the stromal compartment at early stages of tumorigenesis in the Ptch+/− cerebellum, we sampled entire lesions at postnatal day (P) 14 by Collagen IV+ immunostaining, and observed an invasion of vasculature in a minority (24%; Figure 1A). Furthermore, lesion volume, which is a measure of neoplastic progression in the Ptch+/− model, was statistically larger in vascularized lesions (mean 0.18 mm3) compared to non-vascularized ones (mean 0.029 mm3, Figure 1B). These observations are notable, considering that only a minority of lesions undergo malignant transformation and continue to grow into tumors (Kessler et al., 2009; Oliver et al., 2005). To explore this pre-tumor/blood vessel interaction, we targeted Norrin signalling, a well-characterized regulator of neural-endothelial cell communication in the cerebellum (Wang et al., 2012; Zhou et al., 2014). Norrin (encoded by the X-linked gene Ndp) is a secreted atypical Wnt ligand that signals specifically through the Fzd4 receptor and the Lrp5/6 and Tspan12 co-receptors to activate the canonical Wnt pathway in endothelial cells (Figure 1C), where it is required for blood brain barrier (BBB) integrity in the cerebellum (Junge et al., 2009; Wang et al., 2012; Xu et al., 2004). Using male Ndp-/y mice carrying an Ndp-lacZ knockout (KO) allele (Junge et al., 2009), we examined the cerebellar expression profile of Ndp. Consistent with alkaline phosphatase-based reporter data (Ye et al., 2010a), we detected Ndp expression in the Purkinje cell layer, presumably in Bergmann glia (Figure 1D). We also detected Ndp expression in the cerebellar external granule layer (EGL), where GNPs, the Shh-MB cell of origin, reside throughout postnatal development (Figure 1D). Combined X-gal staining and immunohistochemistry (IHC) during the peak period of GNP proliferation at P7 revealed that Ndp expression is concentrated in the outer region of the Pax6+ EGL, in the proliferative phospho-histone H3 (PH3)+ compartment (Figure 1D). β-gal+ cells also overlapped with myelin basic protein (MBP)+ white matter (Figure 1D) implicating oligodendrocytes as another source of Norrin. To focus our expression analysis on the Ptch+/− tumor-relevant cell type, we examined Ndp expression in GNPs isolated from the cerebellar surface at various stages of tumorigenesis. While Ndp levels in Ptch+/− GNPs from pre-lesion (P7) and early lesion (P14) stages were comparable, Ndp expression exhibited a downward trend in GNPs at a lesion progression stage (P30) that reached significance by the tumor stage (Figure 1E). In human MB, NDP expression is enriched specifically in the Shh subgroup compared to the other three molecularly distinct subgroups: Wnt, driven by aberrant Wnt pathway activation, and Group 3 and 4, which are less clearly defined biologically (Taylor et al., 2012) (Figure 1F). Based on a limited sample size, we observed a trend towards reduced survival in Shh-MB patients with tumors exhibiting the lowest levels of NDP expression compared to those with the highest (Figure 1G). Thus, given its known function in signalling to endothelial cells and its expression pattern in GNPs and MB, the Norrin/Fzd4 axis is well-positioned to mediate neural-endothelial cell crosstalk within the Shh-MB lesion and tumor microenvironment. Figure 1 Download asset Open asset Ndp is expressed in Shh-MB precursors and mouse and human Shh-MB. (A) Representative immunostains for Collagen IV (ColIV) counterstained with hematoxylin, to depict non-vascularized (red outline) and vascularized (green outline) lesions from 37 Ptch+/− mouse cerebellar lesions sampled by serial sections at P14. (B) Boxplot showing statistically larger volume in P14 Ptch+/− lesions assigned as vascularized versus those assigned as non-vascularized, based on immunostaining serial sections of each lesion for ColIV. (C) Schematic illustrating the machinery required for Norrin activation of β-catenin (β-cat)/T-cell factor (TCF)-dependent transcription via the Fzd4 receptor and Lrp5 and Tspan12 co-receptors. (D) X-gal staining (blue) of sagittally sectioned cerebella from male Ndp-/y mice carrying an Ndp-lacZ KO allele, at the ages indicated. Bottom row, combined X-gal staining and immunohistochemistry (IHC, brown) for phospho-histone H3 (PH3), Pax6, and myelin basic protein (MBP). (E) Box plot of qRT-PCR analysis of Ndp expression in mouse Ptch+/− purified GNPs and MB lysate from symptomatic animals ranging in age from 3 to 10 months. (F) Box plot of NDP expression obtained by array profiling of human cerebella and two different cohorts (Boston, left; Toronto, right) of human MB samples, categorized by molecular subgroup. (G) Kaplan-Meier survival curve illustrating overall survival of Shh-MB patients with high versus low NDP expression. EGL, external granule layer; GNP, granule neuron progenitor; PCL, Purkinje cell layer; ML, molecular layer; WM, white matter; IGL, internal granule layer; CB, cerebellum; MB, medulloblastoma. Scale bars, 100 µm. https://doi.org/10.7554/eLife.16764.002 Norrin/Fzd4 loss of function dramatically enhances Ptch+/− MB formation To explore the relevance of the stromal compartment during Shh-MB evolution, we disrupted the Norrin/Fzd4 signalling axis at the level of the ligand or receptor in Ptch+/− mice. Given that MB incidence in Ptch+/− mice is dependent on the genetic background strain (Mille et al., 2014), we performed all long and short-term studies of tumorigenesis by comparing animals within the same breeding cohort. While germ-line deletion of the ligand Ndp (NdpKO) or conditional deletion of Fzd4 in endothelial cells via the Tie2Cre driver (Tie2Cre+;Fzd4fl/fl) alone did not result in tumors, both mutations dramatically accelerated MB formation and significantly reduced survival in Ptch+/− mice (Figure 2A–C), to an extent that has only previously been observed in this model upon DNA damage (Pazzaglia et al., 2006a; 2006b; Wetmore et al., 2001). In both NdpKO;Ptch+/− and Tie2Cre+;Fzd4fl/fl;Ptch+/−compound mutants, MB incidence increased by approximately two-fold and mean latency was reduced more than two-fold compared to Ptch+/− littermates (Figure 2C). In isolated GNPs, we detected the expression of endogenous FZD4 protein (Figure 2D), and transcripts of Fzd4, Lrp5 and Tspan12 (Figure 2E); however, conditional deletion of Fzd4 from GNPs in the Ptch+/− model had no impact on survival and tumorigenesis (Figure 2F). These results reveal a novel tumor inhibitory role for Norrin/Fzd4 signalling in Ptch+/−MB that, strikingly, is mediated by the endothelial cell component of the tumor stroma. Figure 2 Download asset Open asset Norrin/Fzd4 signalling in endothelial cells has a potent tumor inhibitory role in Ptch+/− MB. (A) Kaplan-Meier survival curve to assess the impact of Ndp deletion in Ptch+/− mice. NdpKO animals do not develop tumors but five of 44 animals were euthanized prematurely due to a skin condition. (B) Kaplan-Meier survival curve to assess the impact of endothelial cell-targeted (Tie2Cre-driven) Fzd4 deletion in Ptch+/− mice. Tie2Cre+;Fzd4fl/fl animals do not develop tumors but exhibit reduced survival as reported in Fzd4KO mice, which die with esophageal-related feeding defects and progressive auditory and cerebellar degeneration (Wang et al., 2001). (C) Summary of sample sizes, MB incidence and average latency in all animals from Kaplan-Meier survival curves in A and B. (D) Purified P10 GNPs immunostained for anti-FZD4 and anti-keyhole limpet hemocyanin (KLH) isotype-matched control (green), counterstained with Hoescht (blue). White boxes are magnified at right. (E) Box plots of qRT-PCR analysis of Norrin receptor components in isolated mouse GNPs and Ptch+/− MB. (F) Kaplan-Meier survival curve to assess the impact of GNP-targeted (Atoh1Cre-driven) Fzd4 deletion in Ptch+/− mice. (G) Summary of sample sizes, MB incidence and average latency in all animals from Kaplan-Meier survival curve in F *Died with confirmed MB; †Do not develop MB. WT, wild-type; n.s., not significant. https://doi.org/10.7554/eLife.16764.003 Loss of Ndp alters the stromal gene expression signature in Ptch+/− MB To gain insight into Norrin-mediated effects on Ptch+/− tumorigenesis, we examined established NdpKO;Ptch+/− and Ptch+/− MBs by histology and expression profiling. Both tumor types exhibited classic histology, and similar expression patterns for expected Shh-MB neural markers (Atoh1, TUBB3 and GFAP) and Hh target genes (Gli1, Mycn and Ccnd1) (Figure 3A). Co-immunostaining for the ECM protein laminin and the vascular endothelial cell marker CD31 (PECAM1; platelet/endothelial cell adhesion molecule 1) revealed a stromal component in both tumors (Figure 3B). However, tumor heterogeneity poses challenges to histological-based classification, therefore we turned to whole genome expression profiling, where hierarchical clustering and principal component analysis revealed that NdpKO;Ptch+/− and Ptch+/− tumors have clearly separable gene expression signatures (Figure 3C and Figure 3—figure supplement 1A). Consistent with a stromal requirement for Norrin/Fzd4 signalling, tumors in NdpKO;Ptch+/− mutants were enriched for changes in stromal genes compared to their Ptch+/− counterparts, particularly ECM components (Figure 3D,E). Up-regulated genes in NdpKO;Ptch+/− tumors included components of endothelial cells, including Esm1 (Endothelial cell-specific molecule 1), Plvap (Plasmalemmal vesicle associated protein) and Emcn (Endomucin) (Figure 3—figure supplement 1C,D), which we validated by qRT-PCR (Figure 3F). The up-regulation of Plvap, a marker of fenestrated endothelial cells, was associated with increased vascular permeability in NdpKO;Ptch+/− versus Ptch+/− tumors, demonstrated by enhanced leakage of the serum protein binding dye Evans blue (Figure 3G). Additional genes upregulated in NdpKO;Ptch+/− versus Ptch+/− tumors by qRT-PCR were Pecam1, suggesting increased vascularity in NdpKO;Ptch+/− tumors, and the angiogenic regulator Angpt2 (Angiopoietin-2) (Figure 3F). These results suggest that rather than impacting Shh signalling, loss of Norrin modulates features of the tumor stroma. Figure 3 with 1 supplement see all Download asset Open asset Ptch+/− and NdpKO;Ptch+/− MBs have separable gene expression signatures enriched for stromal gene changes. (A) Sections of Ptch+/− and NdpKO;Ptch+/− established MBs stained by hematoxylin and eosin (H and E), in situ hybridization for Atoh1, Gli1, Mycn and Ccnd1 (purple), or immunostaining for class III β-tubulin (TUBB3) and glial fibrillary acidic protein (GFAP; green) counterstained with Hoescht (blue). n = 3 tumors of each genotype examined. (B) Sections of Ptch+/− and NdpKO;Ptch+/− established MBs co-immunostained for CD31 (green) and pan-laminin (red), counterstained with Hoescht (blue). n = 3 tumors of each genotype examined. (C) Whole genome expression profiles of Ptch+/− and NdpKO;Ptch+/− MBs and P6 WT (wild-type) GNPs were used for principal component analysis performed with the 1500 most variable probes across all samples. (D,E) Cellular component gene ontology (GO) analysis of differentially expressed genes between Ptch+/− and NdpKO;Ptch+/− MBs. (F) qRT-PCR analysis of vascular genes upregulated in NdpKO;Ptch+/− versus Ptch+/− MBs. (G) Wholemount views of Ptch+/− and NdpKO;Ptch+/− MBs in animals injected with Evans Blue dye prior to sacrifice. Scale bars, 100 µm. See also Figure 3—figure supplement 1. https://doi.org/10.7554/eLife.16764.004 Stromal Norrin/Fzd4 signalling regulates lesion induction in the Ptch+/− cerebellum In the Ptch+/− model, MB progression is impacted by both the number of preneoplastic lesions, and the rate at which they transform to established tumors (Corcoran et al., 2008; Pazzaglia et al., 2006a; 2006b; Tanori et al., 2010). To determine the stage of tumorigenesis where stromal Norrin/Fzd4 signalling is critical, we examined lesion formation in cerebella of NdpKO;Ptch+/− and Tie2Cre+;Fzd4fl/fl;Ptch+/− compound mutants. At P14, the earliest stage when lesions are consistently detected in Ptch+/− mice, the number of cerebellar lesions in NdpKO;Ptch+/− and Tie2Cre+;Fzd4fl/fl;Ptch+/− mutants was increased 3.9-fold and 2.4-fold, respectively, compared their Ptch+/− littermates (Figure 4A,C). This effect on lesion number was not associated with changes to the Ptch+/− cerebellum at prior ages, as we failed to detect separable GNP expression profiles or general overgrowth of the EGL in NdpKO;Ptch+/− versus Ptch+/− cerebella at P6 (Figure 4—figure supplement 1). Thus, we have revealed an endogenous signalling axis in the preneoplastic microenvironment that plays a protective role during the early stages of Shh-induced tumorigenesis. Figure 4 with 1 supplement see all Download asset Open asset Loss of Norrin/Fzd4 signalling increases lesion formation in P14 Ptch+/− cerebella. (A) Quantification of lesions from serial sections of cresyl violet stained P14 cerebella from NdpKO, Ptch+/− and NdpKO;Ptch+/− mice. Example lesions are outlined in red, and n indicates the number of mice examined. (B) Quantification of lesion volumes from the lesions in A. (C) Quantification of lesions from serial sections of hematoxylin and eosin (H and E) stained P14 cerebella from Tie2Cre+;Fzd4fl/fl, Ptch+/−, and Tie2Cre+;Fzd4fl/fl;Ptch+/− mice. Example lesions are outline in red, and n indicates number of mice examined. (D) Quantification of lesion volumes from the lesions in C. Means are denoted by black horizontal lines on graphs. Scale bars, 100 µm. See also Figure 4—figure supplement 1. https://doi.org/10.7554/eLife.16764.006 Loss of Norrin/Fzd4 signalling promotes stromal remodelling and angiogenesis in Ptch+/−lesions Together with the altered stromal expression signature in NdpKO;Ptch+/− tumors, our finding of enhanced lesion formation suggests that loss of Norrin/Fzd4 signalling may induce stromal perturbations that impact the earliest stages of MB progression. We therefore compared stromal characteristics in lesions from NdpKO;Ptch+/− and Tie2Cre+;Fzd4fl/fl;Ptch+/− compound mutants and their Ptch+/− littermates. Consistent with our previous observations (Figure 1A), immunostaining for CD31 and laminin showed that the majority of Ptch+/− lesions were poorly vascularized, aside from existing blood vessels associated with the meninges at the cerebellar surface (Figure 5A). In contrast, compound mutant lesions exhibited several hallmarks of angiogenic remodelling, including irregular deposition of ECM (Figure 5A), an increased frequency of mitotic endothelial cells per vessel area (Figure 5B), and an increase in vascular endothelial cell and laminin density (Figure 5C). This vascular remodelling was restricted to lesions, with only rare examples of disrupted morphology associated with the EGL in single Tie2Cre+;Fzd4fl/fl cerebella (Figure 5—figure supplement 1). Interestingly, while the vast majority of compound mutant lesions were vascularized at P14 (78% NdpKO;Ptch+/− and 95% Tie2Cre+;Fzd4fl/fl;Ptch+/− versus 24% Ptch+/−;Figure 5D), compound mutant lesion volume was not increased at this early stage of tumor evolution (Figure 4B,D). Accordingly, we observed vascular remodeling in very small (<0.02 mm3) compound mutant lesions (Figure 5—figure supplement 2). Thus, vascular remodelling induced by loss of Norrin/Fzd4 is independent of the increase in lesion volume. Figure 5 with 2 supplements see all Download asset Open asset Loss of Norrin/Fzd4 signalling in endothelial cells promotes angiogenic remodeling. (A) Co-immunostaining for CD31 and pan-laminin, counterstained with Hoescht (blue), on sections of P14 cerebella from the genotypes indicated. Lesions are outlined in white. Arrow on Ptch+/− lesion denotes meningeal blood vessels. Scale bar, 100 µm. (B) Quantification of mitotic endothelial cells in Ptch+/− and NdpKO;Ptch+/− lesions. Top images show co-immunostaining for CD31 and PH3, counterstained with Hoescht (blue), on vascularized P14 lesion sections. Scale bar, 50 µm. Red squares denote areas shown by confocal scans below, where left image depicts the composite maximum intensity projection and the center and right images show individual z-stack slices. Scale bar, 10 µm. Blue arrows denote a PH3+ cell scored as negative for co-localization, whereas red arrows denote a positive co-localization. Graph on right summarizes quantification of double labelled PH3+CD31+ cells per endothelial area. (C) Quantification of CD31+ vessel density and laminin density in lesions of Ptch+/−, NdpKO;Ptch+/− and Tie2Cre+;Fzd4fl/fl;Ptch+/−. Number of lesions (n) examined is indicated on each graph in B and C, and means are denoted by black horizontal lines. (D) Summary of the proportion of vascularized lesions from each genotype. ****p<0.0001. See also Figure 5—figure supplement 1 and 2. https://doi.org/10.7554/eLife.16764.008 Consistent with the known role for Norrin/Fzd4 in normal development and function et al., 2012) and Figure supplement lesions in compound mutants showed a loss of vascular characterized by induction of (Figure variable loss of the protein (Figure and leakage of Evans blue (Figure In contrast, single Ptch+/− mice exhibited vessels (Figure and Figure supplement which is with the positive of by Hh signalling et al., 2011). Accordingly, the proportion of lesions was significantly increased in compound mutants compared to Ptch+/− (Figure vessels in compound mutants did not show in (Figure supplement however, they did exhibit a high frequency of of compared with Ptch+/− lesions (Figure these results show that loss of Norrin/Fzd4 signalling to endothelial cells accelerates the transition to a tumor-permissive stroma characterized by vascular and angiogenic These changes are of the to angiogenesis that promotes the preneoplastic progression of tumors and 2009), where the of and molecular Angpt2 et al., et al., 2014), with our of increased Angpt2 expression in tumors (Figure 3F). Figure with 2 supplements see all Download asset Open asset Loss of Norrin/Fzd4 signalling in endothelial cells promotes vessel in Ptch+/− lesions. (A) Co-immunostaining for and counterstained with Hoescht (blue), on lesion sections of P14 cerebella from the genotypes indicated. outlined in Top images show to variable in expression (B) Quantification of vessel density in lesions of Ptch+/−, NdpKO;Ptch+/− and Tie2Cre+;Fzd4fl/fl;Ptch+/−. Number of lesions (n) examined is indicated on each and means are denoted by black horizontal lines. (C) P14 Ptch+/−, NdpKO;Ptch+/− and Tie2Cre+;Fzd4fl/fl;Ptch+/− mice injected with Evans Blue dye prior to sacrifice. Sections lesions in show Evans Blue as red by and sections and whole brain images outlined in (D) Summary of the proportion of lesions from each genotype vessels or (E) Co-immunostaining for and Collagen IV (ColIV) on lesion sections of P14 cerebella. Scale bars, 100 µm. See also Figure supplement 1 and 2. suggest that Wnt in the is particularly in the of lesion we of Ptch+/− lesions was associated with altered Wnt signalling by expression of a canonical Wnt target gene and 2004). expression in P14 NdpKO;Ptch+/− lesions was significantly reduced compared to P14 Ptch+/− lesions and EGL (Figure supplement the Ndp of this marker in the While expression in the of non-vascularized Ptch+/− lesions was similar to that in EGL, we observed a of endothelial expression in vascularized Ptch+/− lesions compared to EGL (Figure supplement These data suggest that lesion progression in Ptch+/− mice may associated with a in endogenous canonical Wnt in the which is also with the downward trend in Ndp expression during Ptch+/− lesion progression. The microenvironment has an requirement for Fzd4 deletion of Ndp or Fzd4 cerebellar vasculature throughout development and (Wang et al., 2012; Zhou et al., 2014). Thus, we we the effects of genetic on the development of a tumor-permissive stroma by short-term of with a function Fzd4 which has been previously shown to the and impact vasculature et al., (Figure supplement 1). of at P7 and was to the enhanced tumor initiation of our compound mutant MB incidence from in animals injected with anti-keyhole limpet hemocyanin control to in those injected with and significantly mean tumor latency from to (Figure Furthermore, a single of as late as P7 was to increase number not of lesions at P14 (Figure lesions to a vasculature (Figure and significantly increase the proportion of vascularized lesions compared to (Figure Thus, of Fzd4 promotes the to an pro-tumor stroma within suggesting that of Norrin/Fzd4 signalling creates a microenvironment. Figure with 2 supplements see all Download asset Open asset of Fzd4 promotes the to a pro-tumor stroma (A) Kaplan-Meier survival curve comparing Ptch+/− mice with or control at P7 and mice with confirmed MB. sizes, MB incidence and average latency are indicated at right. (B) Quantification of lesion number and volume from P14 cerebella of Ptch+/− mice at P7 with = mice, lesions or = mice, lesions Means are denoted by black horizontal lines on graphs. (C) images from Ptch+/− mice at P7 with = 5 or = were injected with Evans Blue dye prior to and sampled by and staining by Evans Blue as red on sections. Lesions outlined in (D) images from Ptch+/− mice at P7 with = or = immunostained for and counterstained with hematoxylin, to the of non-vascularized in and vascularized in green) lesions in each (E) Kaplan-Meier survival curve comparing Ptch+/− mice with or control at and The in survival is a result of animals from brain or other animals were euthanized due to or while all other mice in both with confirmed MB. sizes, MB incidence and average latency are indicated at the right. (F) Quantification of lesion number and volume from P14 cerebella of Ptch+/− mice as with = mice, lesions or = mice, lesions (G) images from Ptch+/− mice with = or = as were injected with Evans Blue dye prior to and sampled by immunostaining by Evans Blue as red on sections. Lesions outlined in white. images from Ptch+/− mice with = or = as immunostained for to the of non-vascularized in brown) and vascularized in green) lesions in each Summary of Ptch+/− lesion number and vessel upon with or Scale bars, 100 µm. See also Figure supplement 1 and 2. not Ptch+/− tumorigenesis
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction distillée sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Apprise à partir de 10 348 étiquettes directes de Codex et de 10 348 étiquettes directes de Gemma. Le mode candidate est l'union des têtes enseignantes seuillées; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont ni des étiquettes humaines ni des étiquettes directes de modèles de pointe.
Scores Codex et Gemma par catégorie
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,001 | 0,001 |
| Méta-épidémiologie (sens strict) | 0,001 | 0,000 |
| Méta-épidémiologie (sens large) | 0,001 | 0,000 |
| Bibliométrie | 0,000 | 0,000 |
| Études des sciences et des technologies | 0,000 | 0,000 |
| Communication savante | 0,000 | 0,000 |
| Science ouverte | 0,001 | 0,000 |
| Intégrité de la recherche | 0,001 | 0,000 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,000 | 0,000 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule tête enseignante, pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».