Synchondrosis fusion contributes to the progression of postnatal craniofacial dysmorphology in syndromic craniosynostosis.

Syndromic craniosynostosis (CS) patients exhibit early, bony fusion of calvarial sutures and cranial synchondroses, resulting in craniofacial dysmorphology. In this study, we chronologically evaluated skull morphology change after abnormal fusion of the sutures and synchondroses in mouse models of syndromic CS for further understanding of the disease. We found fusion of the inter-sphenoid synchondrosis (ISS) in Apert syndrome model mice (Fgfr2S252W/+ ) around 3 weeks old as seen in Crouzon syndrome model mice (Fgfr2cC342Y/+ ). We then examined ontogenic trajectories of CS mouse models after 3 weeks of age using geometric morphometrics analyses. Antero-ventral growth of the face was affected in Fgfr2S252W/+ and Fgfr2cC342Y/+ mice, while Saethre-Chotzen syndrome model mice (Twist1+/- ) did not show the ISS fusion and exhibited a similar growth pattern to that of control littermates. Further analysis revealed that the coronal suture synostosis in the CS mouse models induces only the brachycephalic phenotype as a shared morphological feature. Although previous studies suggest that the fusion of the facial sutures during neonatal period is associated with midface hypoplasia, the present study suggests that the progressive postnatal fusion of the cranial synchondrosis also contributes to craniofacial dysmorphology in mouse models of syndromic CS. These morphological trajectories increase our understanding of the progression of syndromic CS skull growth.

Classification of large-scale image database of various skin diseases using deep learning.

PURPOSE: The purpose of this study was to develop a deep learning-based computer-aided diagnosis system for skin disease classification using photographic images of patients. The targets are 59 skin diseases, including localized and diffuse diseases captured by photographic cameras, resulting in highly diverse images in terms of the appearance of the diseases or photographic conditions. METHODS: ResNet-18 is used as a baseline model for classification and is reinforced by metric learning to boost generalization in classification by avoiding the overfitting of the training data and increasing the reliability of CADx for dermatologists. Patient-wise classification is performed by aggregating the inference vectors of all the input patient images. RESULTS: The experiment using 70,196 images of 13,038 patients demonstrated that classification accuracy was significantly improved by both metric learning and aggregation, resulting in patient accuracies of 0.579 for Top-1, 0.793 for Top-3, and 0.863 for Top-5. The McNemar test showed that the improvements achieved by the proposed method were statistically significant. CONCLUSION: This study presents a deep learning-based classification of 59 skin diseases using multiple photographic images of a patient. The experimental results demonstrated that the proposed classification reinforced by metric learning and aggregation of multiple input images was effective in the classification of patients with diverse skin diseases and imaging conditions.

ELAVL2-directed RNA regulatory network drives the formation of quiescent primordial follicles.

Formation of primordial follicles is a fundamental, early process in mammalian oogenesis. However, little is known about the underlying mechanisms. We herein report that the RNA-binding proteins ELAVL2 and DDX6 are indispensable for the formation of quiescent primordial follicles in mouse ovaries. We show that Elavl2 knockout females are infertile due to defective primordial follicle formation. ELAVL2 associates with mRNAs encoding components of P-bodies (cytoplasmic RNP granules involved in the decay and storage of RNA) and directs the assembly of P-body-like granules by promoting the translation of DDX6 in oocytes prior to the formation of primordial follicles. Deletion of Ddx6 disturbs the assembly of P-body-like granules and severely impairs the formation of primordial follicles, indicating the potential importance of P-body-like granules in the formation of primordial follicles. Furthermore, Ddx6-deficient oocytes are abnormally enlarged due to misregulated PI3K-AKT signaling. Our data reveal that an ELAVL2-directed post-transcriptional network is essential for the formation of quiescent primordial follicles.

GBIQ: a non-arbitrary, non-biased method for quantification of fluorescent images.

Non-arbitrary and non-biased quantification of fluorescent images is an essential tool for the data-centric approach to biological systems. Typical application is high-content analysis, where various phenotypic changes in cellular components and/or morphology are measured from fluorescent image data. A standard protocol to detect cellular phenotypes is cell-segmentation, in which boundaries of cellular components, such as cell nucleus and plasma membrane, are first identified to define cell segments, then acquiring various phenotypic data of each segment. To achieve reliable outcome, cell-segmentation requires manual adjustments of many parameters; this requirement could hamper automated image processing in high-throughput workflow, whose quantification must be non-arbitrary and non-biased. As a practical alternative to the segmentation-based method, we developed GBIQ (Grid Based Image Quantification), which allows comparison of cellular information without identification of single cells. GBIQ divides an image with tiles of fixed size grids and records statistics of the grids with their location coordinates, minimizing arbitrary intervenes. GBIQ requires only one parameter (size of grid) to be set; nonetheless it robustly produces results suitable for further statistical evaluation. The simplicity of GBIQ allows it to be readily implemented in an automated high-throughput image analysis workflow.

Segmental border is defined by Ripply2-mediated Tbx6 repression independent of Mesp2.

The precise border of somites formed during mouse somitogenesis is defined by a Tbx6 expression domain, which is established by Mesp2-mediated Tbx6 suppression in the anterior part of the presomitic mesoderm (PSM). Ripply2, a target of Mesp2, is proposed to be involved in this down-regulation because Ripply2 deficiency causes an anterior expansion of the Tbx6 domain, resembling the Mesp2-null phenotype. However, it is unclear whether Ripply2 acts on Tbx6 independently or in association with Mesp2. To address this question, we generated three sets of transgenic mice with the following Ripply2 expression patterns: (1) overexpression in the endogenous expression domain, (2) expression instead of Mesp2 (Ripply2-knockin), and (3) ectopic expression in the entire PSM. We found accelerated Tbx6 degradation in the embryos showing Ripply2 overexpression. In the Ripply2-knockin embryos, the anterior limit of Tbx6 domain was generated by Ripply2 even in the absence of Mesp2. Ectopic Ripply2 expression along the entire PSM suppressed Tbx6 and induced Sox2-positive neural tube formation at the bilateral domain, resembling the Tbx6-null phenotype. This phenotype resulted from Tbx6 protein and not mRNA elimination, suggesting the post-translational down-regulation of Tbx6 by Ripply2. Taken together, our results demonstrate that Ripply2 represses Tbx6 in a Mesp2-independent manner, which contributes to the accurate segmental border formation.

Subcellular distribution of mitochondrial ribosomal RNA in the mouse oocyte and zygote.

Mitochondrial ribosomal RNAs (mtrRNAs) have been reported to translocate extra-mitochondrially and localize to the germ cell determinant of oocytes and zygotes in some metazoa except mammals. To address whether the mtrRNAs also localize in the mammals, expression and distribution of mitochondrion-encoded RNAs in the mouse oocytes and zygotes was examined by whole-mount in situ hybridization (ISH). Both 12S and 16S rRNAs were predominantly distributed in the animal hemisphere of the mature oocyte. This distribution pattern was rearranged toward the second polar body in zygotes after fertilization. The amount of mtrRNAs decreased around first cleavage, remained low during second cleavage and increased after third cleavage. Staining intensity of the 12S rRNA was weaker than that of the 16S rRNA throughout the examined stages. Similar distribution dynamics of the 16S rRNA was observed in strontium-activated haploid parthenotes, suggesting the distribution rearrangement does not require a component from sperm. The distribution of 16S rRNAs did not coincide with that of mitochondrion-specific heat shock protein 70, suggesting that the mtrRNA is translocated from mitochondria. The ISH-scanning electron microscopy confirms the extra-mitochondrial mtrRNA in the mouse oocyte. Chloramphenicol (CP) treatment of late pronuclear stage zygotes perturbed first cleavage as judged by the greater than normal disparity in size of blastomeres of 2-cell conceptuses. Two-third of the CP-treated zygotes arrested at either 2-cell or 3-cell stage even after the CP was washed out. These findings indicate that the extra-mitochondrial mtrRNAs are localized in the mouse oocyte and implicated in correct cytoplasmic segregation into blastomeres through cleavages of the zygote.

Experimental analysis of the transdifferentiation of visceral to parietal endoderm in the mouse.

The visceral endoderm (VE) of isolated extraembryonic regions (ExEmbs) of 7 days postcoitum (dpc) prestreak mouse conceptuses have been shown to convert readily to parietal endoderm (PE). The present study addresses the following three unanswered questions. On what does conversion depend, how rapidly does it occur, and is it an enduring general property of a residual small population of relatively immature cells? In situ hybridization reveals that change in cell state occurs within 2 days of culture. Deprivation of the mesoderm also promotes it in later ExEmbs. Conversely, the conversion to PE in isolated 7 dpc ExEmbs is suppressed by grafting 8 dpc or 9 dpc mesoderm. Hence, the conversion provides an example of transdifferentiation that is promoted by the absence of extraembryonic mesoderm. The presence of mesoderm seems to be necessary to enable the VE to grow rather than convert to PE, as occurs if it retains contact with the extraembryonic ectoderm.

Disappearance of epidermal growth factor receptor is essential in the fusion of the nasal epithelium.

Abstract Epidermal growth factor (EGF) and receptor (-R) signaling pathway is required for epithelial cell growth and differentiation such as the degeneration of the medial edge epithelial cells during the fusion process of secondary palate formation. As epithelial fusion takes place during primary palate formation, we investigated the involvement of the EGF-R in fusion of the medial (MNP) and lateral (LNP) nasal prominences of the mouse embryo was examined. Immunoreactivity of EGF-R was investigated in embryonic day 10 embryos (32-37 somite stages). The EGF-R immunoreactivity was observed in the nasal epithelia of the presumptive fusion area before fusion. It became undetectable just prior to the fusion and faintly reappeared at the time of the fusion. In contrast, the non-fusing epithelial cells of the nasal groove maintained the immunoreactivity throughout these stages. In order to elucidate whether the EGF/EGF-R signaling pathway was involved in nasal epithelial fusion, EGF solution was injected into the exocoelum of explanted mouse embryos, and the embryos were cultured for 18-24 h by whole embryo culture (WEC). This exogenous EGF inhibited fusion of nasal prominences in 66.7-81.5% of the embryos. Treatment with EGF for 4-14 h showed that exogenous EGF disturbed the EGF-R disappearance and normal alteration of epithelial cell morphology in the fusion area. These results suggest that temporal disappearance of the EGF/EGF-R signaling from presumptive fusion of the nasal prominences is required for morphological change of the epithelial cells leading to the fusion of MNP and LNP.

Rhombomere formation and hind-brain crest cell migration from prorhombomeric origins in mouse embryos.

Prior to rhombomere development, structures called prorhombomeres appear in the mammalian hindbrain. This study clarifies the developmental relationship between prorhombomeres and their descendent rhombomeres and hindbrain crest cells in mouse embryos by focal dye injections at various levels of prorhombomere A (proRhA), proRhB, and proRhC, as well as at their boundaries. ProRhA gives rise to two rhombomeres, rhombomeres 1 and 2 (r1 and r2), as well as to crest cells that migrate into the first pharyngeal arch, including the trigeminal ganglion. ProRhB develops into r3 and r4 and produces crest cells populating the second arch and acousticofacial ganglion. The anterior portion of proRhC gives rise to r5 and r6 and to crest cells migrating into the third pharyngeal arch and the IXth ganglion; its posterior portion develops into r7 and releases crest cells into the fourth pharyngeal arch region as well as the Xth ganglion. These results suggest that the boundaries between prorhombomeres serve as lineage restrictions for both hind-brain neuroepithelial cells and for segmental origins of crest cell populations in mouse embryos. The Hox code of the mouse head can be schematized in a much simpler way based on this prorhombomeric organization of the hind-brain, suggesting that prorhombomeres primarily underlie mammalian hind-brain segmentation.