Publikationen

The sustainable removal of iron ions (Fe2+/3+) and sulfate (SO42-) from contaminated water remains a major challenge, particularly in mining-affected regions such as Lusatia (Germany), where both contaminants frequently co-occur at elevated levels. Here, a dual-modification strategy based on chitosan (Cs)-modified yeast cells (Cs@YC) and chitin-glucan complexes (CGC), is presented to enhance biosorption performance. Dried Cs@YC showed improved adsorption capacities compared to unmodified YC, while the combined Cs-functionalized CGC (Cs@CGC) exhibited the highest uptake capacities, reaching 2.09 mmol g-1 for Fe2+/3+ and 0.79 mmol g-1 for SO42- at pH 6. Microscopic and spectroscopic analyses (SEM-EDX, TEM, AFM, XPS) showed that Cs surface functionalization combined with alkaline-induced CGC restructuring increases the density and accessibility of amino and hydroxyl groups and promotes controlled iron (oxyhydr)oxide nucleation and crystallization. Equilibrium modelling using Langmuir, Freundlich, and Sips isotherms demonstrated that Cs@YC provides high-affinity metal-binding sites with positively charged surface domains that facilitate electrostatic association and adsorption of SO42-. To the best of our knowledge, this is the first study to demonstrate the synergistic interaction between Cs modification and CGC-based structural tuning in yeast-derived biosorbents for dual Fe2+/3+/SO42- remediation. The results demonstrate that Cs@YC offers promising adsorption performance under controlled laboratory conditions, warranting further evaluation in real mining-affected waters containing competing ions and organic matter.

We introduce spatially informed variational autoencoders (SI-VAE) as self-supervised deep-learning models that use stochastic point processes to predict spatial organization patterns from images. Existing approaches to learning visual representations based on variational autoencoders (VAE) struggle to capture spatial correlations between objects or events, focusing instead on pixel intensities. We address this limitation by incorporating a point-process likelihood, derived from the Papangelou conditional intensity, as a self-supervision target. This results in a hybrid model that learns statistically interpretable representations of spatial localization patterns and enables zero-shot conditional simulation directly from images. Experiments with synthetic images show that SI-VAE improve the classification accuracy of attractive, repulsive, and uncorrelated point patterns from 48% (VAE) to over 80% in the worst case and 90% in the best case, while generalizing to unseen data. We apply SI-VAE to a real-world microscopy data set, demonstrating its use for studying the spatial organization of proteins in human cells and for using the representations in downstream statistical analysis.

Tumor development and progression involve biophysical changes across spatial scales, from the subcellular to the multicellular tissue scale. While cells are known to dynamically regulate their volumes and mechanics in dependence of cell state and function, it is unclear how these properties are controlled in dense multicellular environments like developing tumors. Here, we quantified cell and nuclear volumes of cancer cells forming multicellular spheroids within mechanically tunable biohybrid polymer hydrogels. We quantitatively showed that formation of multicellular structures is associated with marked reductions of cellular and nuclear volumes, cell cycle delays as well as cell mechanical alterations, and that these changes are coupled. Single-to-multicellular transitions led to up to 60% decreases in median nuclear volumes, which was not explained by growth-induced compressive stress. Instead, nuclear volume reductions in emerging clusters arose from cell cycle adaptations, with accumulation of smaller G1-phase cells-reversed by CDK1 inhibition. Additional nuclear downsizing in forming clusters was associated with cell mass density and stiffness increases and reverted upon cell release. Conversely, multicellular-to-single cell transitions during invasion were accompanied by nuclear volume expansion and cell softening. Together, these findings reveal dynamic regulation of cellular and nuclear volumes, mechanics, and cell cycle progression in response to multicellular state.

Lipids and proteins compartmentalize biological membranes into nanoscale domains, which are crucial for signalling, intracellular trafficking and many other cellular processes. Studying nanodomain function requires the ability to measure protein and lipid localization at the nanoscale. Current methods for visualizing lipid localization do not meet this requirement. Here we introduce a correlative light and electron microscopy workflow to image lipids (Lipid-CLEM), combining near-native lipid probes and on-section labelling by click chemistry. This approach enables the quantification of relative lipid densities in membrane nanodomains. We find differential partitioning of sphingomyelin into intraluminal vesicles, recycling tubules and the boundary membrane of the early endosome, representing a degree of nanoscale organization previously observed only for proteins. We anticipate that our Lipid-CLEM workflow will greatly facilitate the mechanistic analysis of lipid functions in cell biology, allowing for the simultaneous investigation of proteins and lipids during membrane nanodomain assembly and function.

Microcephaly with or without chorioretinopathy, lymphedema, or intellectual disabilities (MCLID) is a rare disease caused by mutations in the mitotic motor KIF11. However, the specific neuronal functions of KIF11, its mechanisms of microtubule (MT) regulation, and the impact of MCLID mutations on KIF11 function remain underexplored. Here, using live-imaging, we find that KIF11 depletion in postmitotic neurons increases minus-end-out MT dynamics in both axons and dendrites. Introducing MCLID-associated KIF11 mutations, KIF11Y82F and KIF11ΔCterm, significantly reduces MT dynamics, impairs dendritic arborization, and decreases mEPSC frequency. Biochemical analyses reveal that the KIF11ΔCterm mutant disrupts tetramer formation and MT crosslinking, while the KIF11Y82F mutant reduces MT sliding velocity and ATP affinity. Temporal inhibition of KIF11 using a photo-controllable KIF11 increases MT dynamics and dendritic growth. Together, these data reveal that KIF11 is a MT dynamics rheostat and regulator of dendritic arborization in mature neurons, providing essential insights into the molecular mechanisms driving MCLID.

Freshwater ecosystems are among the most endangered habitats on Earth, with approximately one-fourth of aquatic species at risk of extinction. Effective conservation efforts require comprehensive monitoring and accurate species identification, including often overlooked groups. Planarian flatworms are one such group that, although commonly present in freshwater ecosystems worldwide, remains understudied even in species-rich areas, e.g. Croatia. As a result, the true extent of planarian diversity often remains underappreciated.

Comparing mathematical models offers a means to evaluate competing scientific theories. However, exact methods of model calibration are not applicable to many probabilistic models which simulate high-dimensional spatio-temporal data. Approximate Bayesian Computation is a widely-used method for parameter inference and model selection in such scenarios, and it may be combined with Topological Data Analysis to study models which simulate data with fine spatial structure.

Microtubules are stiff cytoskeletal polymers whose ability to rapidly switch between growth and disassembly relies on a metastable lattice. This metastability is also reflected in their sensitivity to environmental conditions and in intrinsic lattice dynamics, where spontaneous tubulin loss is balanced by tubulin incorporation from solution-a process that also enables microtubules to self-repair when damaged. Whether such intrinsic self-repair is sufficient to preserve microtubule integrity during dynamic molecular motor-induced buckling, which frequently occurs in cells, remains unclear. Here, we show that kinesin-driven microtubule buckling in vitro induces severe lattice damage, leading to extensive tubulin incorporation. In many cases, however, the damage exceeds the microtubules' capacity for self-repair, resulting in breakage. In contrast, microtubules survive continuous buckling substantially longer in the presence of intracellular factors. Our results identify the limits of intrinsic microtubule self-repair and demonstrate that additional cellular mechanisms are essential to maintain microtubule integrity under sustained mechanical load.

Meritxell Huch spoke with Cell Reports about her scientific journey, mentorship philosophy, and lab culture as well as her research on human liver regeneration. She discussed refined liver organoid models capturing cholangiocyte heterogeneity and cellular plasticity and shared perspectives on challenges in organoid research and modern science.

Lipid vesicles are important as minimal model systems for cellular compartmentalization. They drive major advances in deciphering biological mechanisms by molecular reconstitution; provide rational solutions for primitive compartmentalization in origin-of-life studies; form the basis of synthetic cells and drug delivery vehicles. The emulsion method is a well-established route for producing bilayer lipid vesicles. However, the application of this method in microfluidics requires complex and specialized machinery. The bulk method suffers from the need to physically manipulate the vesicles through oil layers for characterization that can damage the vesicles. Given this, we present a facile and robust method for on-chip production and manipulation of lipid vesicles by the emulsion method. We prepared a simple device that allows preparation, imaging, and collection of activated lipid vesicles. This technique combines minimal processing steps with maximum flexibility in lipid vesicle production and manipulation with direct imaging, thus fast-tracking production lines across disciplines.