浅井 直也

Abstract Objective Loss‐of‐function mutations in the GIRDIN/CCDC88A gene cause developmental epileptic encephalopathy (DEE) in humans. However, its pathogenesis is largely unknown. Global knockout mice of the corresponding orthologous gene (gKOs) have a preweaning lethal phenotype with growth failure, preventing longitudinal analysis. We aimed to overcome this lethality and elucidate DEE pathogenesis. Methods We developed a novel lifelong feeding regimen (NLFR), which consists of providing mash food from postnatal day 14 (P14) until weaning (P28), followed by agar‐bound food exclusively after weaning. Videography, electroencephalography (EEG), and histological analyses were performed. Conditional Girdin/Ccdc88a knockout mice (cKOs) of variable lineages (Nestin, Emx1, or Nkx2‐1) were generated to identify the region responsible for epilepsy. Results Under the NLFR, gKOs survived beyond 1 year and displayed fully penetrant, robust epileptic phenotypes, including early‐onset (P22.3 in average) generalized tonic–clonic seizures (GTCSs) (averaging eight per day), which were completely synchronized with fast rhythms on EEG, frequent interictal electroencephalographic spikes (averaging 430 per hour), and progressive deformation of visceral organs. In addition, gKOs had absence seizures, which were not always time‐locked to frequent spike waves on EEG. The frequent GTCSs and interictal spikes in gKOs were suppressed by known antiepileptic drugs. Histologically, bilateral hippocampi in gKOs exhibited congenital cornu‐ammonis splitting, granule cell dispersion, and astrogliosis. Furthermore, analysis of conditional knockouts using multiple Cre‐deleters identified a defect in the delivery of interneuron precursors from the medial ganglionic eminence into the hippocampal primordium during embryogenesis as a major cause of epileptogenesis. Significance These findings give rise to a new approach of lifelong caregiving to overcome the problem of preweaning lethality in animal models. We propose a useful model for studying DEE with hippocampal sclerosis and interneuronopathy. gKOs with NLFR combine the contradictory properties of robust epileptic phenotypes and long‐term survivability, which can be used to investigate spontaneous epileptic wave propagation and therapeutic intervention in hippocampal sclerosis.

Adipocyte-cancer cell interactions promote tumor development and progression. Previously, we identified adipsin (CFD) and its downstream effector, hepatocyte growth factor (HGF), as adipokines that enhance adipocyte-breast cancer stem cell interactions. Here, we show that adipsin-dependent adipocyte maturation and the subsequent upregulation of HGF promote tumor invasion in breast cancers. Mature adipocytes, but not their precursors, significantly induced breast tumor cell migration and invasion in an adipsin expression-dependent manner. Promoters of tumor invasion, galectin 7 and matrix metalloproteinases, were significantly upregulated in cancer cells cocultured with mature adipocytes; meanwhile, their expression levels in cancer cells cocultured with adipocytes were reduced by adipsin knockout (Cfd KO) or a competitive inhibitor of CFD. Tumor growth and distant metastasis of mammary cancer cells were significantly suppressed when syngeneic mammary cancer cells were transplanted into Cfd KO mice. Histological analyses revealed reductions in capsular formation and tumor invasion at the cancer-adipocyte interface in the mammary tumors formed in Cfd KO mice. These findings indicate that adipsin-dependent adipocyte maturation may play an important role in adipocyte-cancer cell interaction and breast cancer progression.

Pancreatic stellate cells (PSCs) are stromal cells in the pancreas that play an important role in pancreatic pathology. In chronic pancreatitis (CP) and pancreatic ductal adenocarcinoma (PDAC), PSCs are known to get activated to form myofibroblasts or cancer-associated fibroblasts (CAFs) that promote stromal fibroinflammatory reactions. However, previous studies on PSCs were mainly based on the findings obtained using ex vivo expanded PSCs, with few studies that addressed the significance of in situ tissue-resident PSCs using animal models. Their contributions to fibrotic reactions in CP and PDAC are also lesser-known. These limitations in our understanding of PSC biology have been attributed to the lack of specific molecular markers of PSCs. Herein, we established Meflin (Islr), a glycosylphosphatidylinositol-anchored membrane protein, as a PSC-specific marker in both mouse and human by using human pancreatic tissue samples and Meflin reporter mice. Meflin-positive (Meflin+ ) cells contain lipid droplets and express the conventional PSC marker Desmin in normal mouse pancreas, with some cells also positive for Gli1, the marker of pancreatic tissue-resident fibroblasts. Three-dimensional analysis of the cleared pancreas of Meflin reporter mice showed that Meflin+ PSCs have long and thin cytoplasmic protrusions, and are localised on the abluminal side of vessels in the normal pancreas. Lineage tracing experiments revealed that Meflin+ PSCs constitute one of the origins of fibroblasts and CAFs in CP and PDAC, respectively. In these diseases, Meflin+ PSC-derived fibroblasts showed a distinctive morphology and distribution from Meflin+ PSCs in the normal pancreas. Furthermore, we showed that the genetic depletion of Meflin+ PSCs accelerated fibrosis and attenuated epithelial regeneration and stromal R-spondin 3 expression, thereby implying that Meflin+ PSCs and their lineage cells may support tissue recovery and Wnt/R-spondin signalling after pancreatic injury and PDAC development. Together, these data indicate that Meflin may be a marker specific to tissue-resident PSCs and useful for studying their biology in both health and disease. © 2023 The Pathological Society of Great Britain and Ireland.