Biology subjects

Eichin, F.

Publications and source records attributed to Eichin, F..

7 recordsLinked to original sources

bioRxiv2026

Plasmalogen metabolism shapes germinal center immunity

Plasmalogens are membrane lipids whose roles in immunity are emerging through identification of the PEDS1 gene and advances in lipid analytics. Because PEDS1 catalyses the final step of plasmalogen biosynthesis, Peds1 knockout models plasmalogen loss. Building on reports of a haematopoietic phenotype in Peds1-deficient mice, we performed extended profiling to characterise haematological and immune phenotypes. In line with these previous investigations, mice showed reduced red blood cell counts and haematocrit, with increased mean platelet volume. At steady state, plasmalogen deficiency was associated with reduced germinal centre (GC) B cell frequencies in mucosal lymphoid tissues and lower basal serum IgG1 levels. Following immunisation, GC B cell frequencies and ovalbumin-specific IgG1 were similar across genotypes, yet GC B cells exhibited elevated mitochondrial superoxide and plasma cells displayed increased lipid peroxidation. To test for B cell-intrinsic effects, we used an in vitro GC culture, where Peds1-deficient B cells showed reduced class switching to IgG1, intracellular IgG1 accumulation and diminished IgG1 secretion. Together, these findings support a role for plasmalogens and PEDS1 as a lipid-metabolic modulator of GC dynamics and IgG1 output, consistent with broader ether lipid perturbations. Plasmalogen metabolism may therefore offer a potential entry point to modulate humoral immunity or haematopoiesis.

immunology↗

bioRxiv2026

Ex Vivo Assay for Organ-Specific Cancer Cell Invasion

BackgroundMetastasis is the leading cause of cancer-related mortality, yet experimental models often fail to recapitulate the tissue-specific microenvironments that shape metastatic dissemination. While in vivo systems provide physiological relevance, they are poorly suited for mechanistic studies. Conversely, conventional in vitro assays lack the organ-specific extracellular matrix (ECM) context that critically regulates invasive behavior. There is therefore a need for accessible models that balance biological relevance with experimental feasibility. ResultsIn this study, we developed an ex vivo invasion platform based on mild detergent decellularization of mouse organs followed by vibratome slicing. This approach generates optically transparent lung, liver, and intestine ECM scaffolds that preserve native matrix architecture, mechanical properties, and retain biochemical hallmarks of their tissues of origin. Organ-derived matrices were integrated into standard microfluidic channels and analyzed using conventional fluorescence microscopy to enable quantitative assessment of cancer cell invasion. Benchmarking with breast cancer cell lines of defined invasive capacity, we could demonstrate the robustness and biological relevance of the system. Non-invasive MCF7 cells failed to infiltrate any organ scaffold, whereas highly invasive MDA-MB-231 cells exhibited pronounced organ-specific invasion. These cells preferentially invaded lung and liver ECM while showing minimal invasion of intestinal scaffolds, recapitulating clinically observed metastatic tropism. Quantitative invasion rates closely matched values previously reported in vivo using intravital microscopy. ConclusionsThis ex vivo organ-derived ECM platform provides a scalable, cost-effective, and experimentally accessible system to study ECM-driven determinants of metastatic invasion. By preserving tissue-specific matrix cues while reducing reliance on animal models, this approach provides a powerful tool to interrogate ECM-driven determinants of metastasis and to evaluate potential therapeutic interventions.

cancer biology↗

bioRxiv2025

tRNA thiolation defects disrupt cellular proteostasis and tissue homeostasis in mammals

Sulfur modification of tRNA wobble uridines is an evolutionarily conserved mechanism that ensures efficient protein synthesis. In humans, loss of this anticodon modification due to mutations in CTU2 (cytosolic thiouridylase 2) causes DREAM-PL syndrome, a severe congenital disorder often leading to early postnatal death. However, the mechanisms by which loss of tRNA thiolation drives pathology remain unclear. Here, we show that loss of CTU2 triggers significant cellular proteostasis defects in patient cells and model cell lines. Structural and biochemical analyses reveal that the pathogenic CTU2L63P mutation destabilizes the CTU1/CTU2 complex and abolishes tRNA binding and thiolation. Acute loss of CTU2 caused codon-specific ribosome pausing at A-ending codons decoded by thiolated tRNAs, and decreased ribosome occupancy of A-rich transcripts in a dosage-dependent manner. Codon-biased mRNAs transcribed from genes critical for ciliogenesis are predicted to be most affected, linking their reduced translation to DREAM-PL etiology in humans. Surprisingly, Ctu2L63P mice display severe thiolation defects, but develop normally, are viable and fertile. Our findings highlight the importance of functional tRNA thiolation for organismal health in humans and identify species-specific vulnerabilities during embryonic development in mammals.

cell biology↗

bioRxiv2025

Sequential PIDD1 auto-processing is essential for ploidy control in liver and heart

Polyploidization refers to the balanced increase in gene copy number and is a feature of specialized cells in different mammalian tissues, including the liver and the heart. During organogenesis, hepatocytes and cardiomyocytes undergo scheduled polyploidization events to increase their cellular or nuclear DNA content. This is thought to improve cellular output and enable for rapid genetic adaptation in response to stress. Yet, excessive increases in ploidy can also be disadvantageous and increase the risk of genome instability. Hence, a dedicated machinery, the PIDDosome multi-protein complex, has evolved to prevent exacerbated increases in DNA content. Using targeted mutagenesis in mice, we show that the PIDDosome controls hepatocyte ploidy in a cell-autonomous manner and that sequential and quantitative auto-processing of PIDD1 is key for accurate control of ploidy in postnatal development of liver and heart. Stoichiometric imbalances in bioactive PIDD1-fragments impair p53-dependent and independent cell cycle arrest responses during organogenesis, as well as caspase-2-dependent apoptosis caused by centrosome amplification. Strikingly, targeted mutagenesis of the caspase cleavage motif in the critical E3-ligase controlling p53 protein levels, Mdm2, impairs ploidy control in hepatocytes, but not in cardiomyocytes, indicative of the existence of alternative caspase-2 substrates that help to restrict ploidy in the heart.

cell biology↗

bioRxiv2024

The PIDDosome controls cardiomyocyte polyploidization during postnatal heart development

The adult mammalian heart is characterized by post-mitotic polyploid cardiomyocytes (CMs). Understanding how CMs regulate cell cycle exit and ploidy can help developing new heart regenerative therapies. Here, we uncover that the PIDDosome, a multi-protein complex activating the endopeptidase Caspase-2, helps to implement a CM-specific differentiation program that limits ploidy during postnatal heart development. DNA content analyses show that PIDDosome-loss causes a cell-autonomous increase in nuclear and cellular CM ploidy. Remarkably, increased ploidy does not affect cardiac structure nor function. PIDDosome-imposed ploidy restriction commences at postnatal day 7 (P7), reaching a plateau on P14. PIDDosome activation requires ANKRD26, targeting PIDD1 to mother centrioles. Opposite to prior observations in liver development, the PIDDosome limits CM polyploidization in a p53-independent manner but reliant on p21/Cdkn1a, a notion supported by nuclear RNA sequencing and genetic deletion experiments. Our results provide new insights how proliferation of polyploid CMs is restricted during postnatal heart development.

developmental biology↗

bioRxiv2021

Incorporating structural knowledge into unsupervised deep learning for two-photon imaging data

Live imaging techniques, such as two-photon imaging, promise novel insights into cellular activity patterns at a high spatio-temporal resolution. While current deep learning approaches typically focus on specific supervised tasks in the analysis of such data, we investigate how structural knowledge can be incorporated into an unsupervised generative deep learning model directly at the level of the video frames. We exemplify the proposed approach with two-photon imaging data from hippocampal CA1 neurons in mice, where we account for spatial structure with convolutional neural network components, disentangle the neural activity of interest from the neuropil background signal with separate foreground and background encoders and model gradual temporal changes by imposing smoothness constraints. Taken together, our results illustrate how such architecture choices facilitate a modeling approach that combines the flexibility of deep learning with the benefits of domain knowledge, providing an interpretable, purely image-based model of activity signals from live imaging data. Teaser sentenceUsing a neural network architecture that reflects domain knowledge provides an interpretable model of live cell imaging data.

bioinformatics↗

bioRxiv2019

SAFB2 enables the processing of suboptimal stem-loop structures in clustered primary miRNA transcripts

MicroRNAs (miRNAs) are small noncoding RNAs that post-transcriptionally silence most protein-coding genes in mammals. They are generated from primary transcripts containing single or multiple clustered stem-loop structures that are thought to be recognized and cleaved by the DGCR8/DROSHA Microprocessor complex as independent units. Contrasting this view, we here report an unexpected mode of processing of a bicistronic cluster of the miR-15 family, miR-15a-16-1. We find that the primary miR-15a stem-loop is a poor Microprocessor substrate and is consequently not processed on its own, but that the presence of the neighboring primary miR-16-1 stem-loop on the same transcript can compensate for this deficiency in cis. Using a CRISPR/Cas9 screen, we identify SAFB2 (scaffold attachment factor B2) as an essential co-factor in this miR-16-1-assisted pri-miR-15 cleavage, and describe SAFB2 as a novel accessory protein of DROSHA. Notably, SAFB2-mediated cluster assistance expands to other clustered pri-miRNAs including miR-15b, miR-92a and miR-181b, indicating a general mechanism. Together, our study reveals an unrecognized function of SAFB2 in miRNA processing and suggests a scenario in which SAFB2 enables the binding and processing of suboptimal DGCR8/DROSHA substrates in clustered primary miRNA transcripts. HighlightsO_LIthe primary miR-15a stem-loop structure per se is a poor Microprocessor substrate C_LIO_LIcleavage of pri-miR-15a requires the processing of an additional miRNA stem-loop on the same RNA C_LIO_LIsequential pri-miRNA processing or "cluster assistance" is mediated by SAFB proteins C_LIO_LISAFB2 associates with the Microprocessor C_LI

molecular biology↗