Together, 3BTMs provide a reproducible, scalable and physiologically relevant platform to study human brain tissue, mature microglia and neurodegenerative disease mechanisms, and to test therapeutic interventions in a fully human system. 9/11

As a translational proof of principle, we induced dense-core plaque-like pathology using synthetic Aβ42 and treated with Aducanumab, which engaged Aβ, reduced plaque load over time, and increased microglial Aβ phagocytosis. Multi-omics showed reversal of disease-associated glial signatures. 8/11

Single-cell profiling of AD 3BTMs showed a major shift in microglial states: two genotype-specific clusters emerged, upregulating AD risk genes (TREM2, CD14, SORL1) and patient-derived AD microglia signatures. This confirms disease-relevant neuroinflammatory activation. 7/11

Using #CRISPR knock-in of synergistic APP mutations (NLGF), 3BTMs showed strongly increased Aβ levels, progressive Aβ deposition, and insoluble Aβ42. They also developed increased phospho-tau (AT270, PHF-1), especially when microglia and astrocytes were present. 6/11

#scRNA-seq with @dzne.science revealed that microglia in 3BTMs transition from a proliferative/immature state at 1 month to a resting, immune-sensing state at 3 months. They increasingly resemble adult human microglia and show higher sensome and homeostatic marker expression across lines. 5/11

#Proteomics showed that 3BTMs progressively mature, moving closer to juvenile human brain than to organoids. Synapse proteins, astrocyte maturation markers and brain ECM components (hyaluronan, CSPGs) increased over time. This maturation was highly reproducible across iPSC lines. 4/11

We generated #3BTMs by combining iPSC-derived neurons, astrocytes and microglia. 3BTMs were stable for months and maintained the expected cell ratios. Microglia survived >6 months, adopted ramified morphology, expressed homeostatic markers, and dynamically responded to injury. 3/11

#Microglia are central to brain homeostasis and #neuroinflammation, but mature human microglia are still difficult to model in vitro. This is especially limiting for studies of #Alzheimer’s disease and other #brain disorders. 2/11

🧵New paper from the lab @isd-research.bsky.social @lmu-klinikum.bsky.social @lmu.de published in @natneuro.nature.com: We developed a highly reproducible 3D human brain tissue model (3BTM) that enables studying microglial phenotypes in health and disease! #OA PDF here: rdcu.be/fxLou Thread👇: 1/11

Can this model support translational applications?💊 A human Tau #PET tracer specifically bound to mutant 4R neurons. And we showed proof-of-concept that a drug, the aggregation inhibitor Anle138b, modulated our Tauopathy phenotypes. A human platform for #DrugDiscovery in #Tauopathies! 9/10

We also found synapse loss and ghost tangle-like structures in affected neurons, as well as nuclear lamina deformations linked to microtubule defects. These hallmarks mirror what is observed in patient brains with #Tauopathy. 8/10

Reproducing key late-stage hallmarks found in patient brain, affected neurons accumulated misfolded #Tau hyperphosphorylated at classical pathological epitopes, and formed tangle-like aggregates filling entire somata. All progressive, all cell-autonomous. 7/10

The 4R PLSF neurons developed seeding-competent, Sarkosyl-insoluble Tau that increased progressively over months. Electron microscopy suggested formation of fibril-like structures, confirmed as Tau-positive by immunogold labeling. 6/10

What would happen if only 4R was present? Exclusive expression of mutant 4R Tau disproportionately amplified pathology compared to neurons expressing 3R and mutant 4R Tau. At later stages, most neurons were affected. Spatial clustering analysis suggests inter-neuronal #TauSpreading in culture. 5/10

We combined 3R/4R #Tau expression with two synergistic Tauopathy mutations, P301L and S320F, that promote both nucleation and elongation of Tau fibrils. Neither 4R expression nor the mutations alone induced pathology. Only the combination triggered robust #Tauopathy. 4/10

One key reason: #iPSC neurons express the fetal 3R Tau isoform, but adult human brains have a 1:1 ratio of 3R and 4R. 4R is believed to be more aggregation-prone and central to disease. We used #CRISPR and PiggyBac to scarlessly induce adult-like #Tau splicing from the endogenous MAPT locus. 3/10

The problem: Despite extensive research, no human neuronal model reproducibly develops full late-stage #Tau pathology including seed formation, hyperphosphorylation and aggregation without transgenic overexpression or exogenous seeds. This limits study of disease mechanisms and #DrugDiscovery. 2/10

🧵New paper from the lab @isd-research.bsky.social @lmu-klinikum.bsky.social @lmu.de in STM @science.org! A human #iPSC-derived #Tauopathy model that endogenously develops late-stage Tau pathology. Free personal PDF: www.science.org/eprint/5H5YD.... Here's what we found👇 1/10

Many thanks @nature.com for selecting our image of a fully #iPSC derived #BBB (#blood-brain-barrier) as 'Image of the Week' in Nature Briefings: mailchi.mp/nature/trans... 😀 Find the image in action in our primary research paper here: rdcu.be/eUIl3

PS: If you interested in this kind of work, we are looking for a highly motivated Postdoc to continue in this area of the lab. See our profile and webpage for more details: isd-research.de/paquetlab

Lastly, treatment with AKB-9778, which boosts Tie2 activity by blocking Tie2-dampening vascular endothelial protein tyrosine phosphatase, restored Tie2 signaling both in mice and human endothelial cells, which rescued nitric oxide production and had positive effects on infarct sizes in mice. 6/7

We also found that FoxF2 in endothelial cells facilitated functional hyperemia and limited infarct size in adult mice, again mediated via Tie2 signaling and underscoring the role of FOKF2 in #Stroke and #Small vessel disease. 5/7

In both EC-specific KO mice and our new fully iPSC-derived BBB model (see here: rdcu.be/eUIl3), loss of FOXF2 caused BBB leakage and attenuated TIE2 signaling, pointing to a central role of this pathway in FOXF2-dependent BBB regulation. 4/7

By performing ChIP-seq and omics in complementary mouse and human iPSC-derived models of FOXF2 deficiency we found that FOXF2 acts as a transcriptional activator of Tie2 and other lineage-specific genes in endothelial cells. 3/7

The #blood-brain-barrier (BBB) is a critical structure regulating the brain, but its function is affected by major brain disorders like #Stroke, #Small vessel disease and #Alzheimer’s. Genetics identified risk loci, such as FOXF2, in patients but role and underlying mechanisms remain unclear. 2/7

And another new paper with co-first and -last author contributions from the lab at @natneuro.nature.com! Another great collab with DichgansLab @isd-research.bsky.social @lmumuenchen.bsky.social: Foxf2 maintains brain endothelial cell function via Tie2 signaling! #OA PDF: rdcu.be/eUImz Thread👇 1/7

Lastly, we tested if we could apply the human #BBB model to test treatments restoring #BBB function. We delivered FoxF2 back into disease vessels by perfusing #LNPs #lipid-nanoparticles, and found that this rescued the defects, indicating suitability of the model for translational applications 9/10

Tight junctions, which restrict transport between ECs at the #BBB were also affected, decreasing the functional barrier - the same again was true in vivo, indicating that FOXF2 is a major regulator of #BBB function in ECs and that its dysfunction in patients may lead to NVD via #BBB dysfunction 8/10

We found a major dysregulation of several central #BBB functions, including upregulation of transport via caveolae, which is usually strictly regulated at the BBB. Intriguingly, comparison with data from FoxF2 KO mice showed a full phenocopy (identical phenotype), validating the human model. 7/10

Using #CRISPR/Cas gene editing, we engineered the cells to contain a loss of the FOXF2 gene, a major risk locus found in patients with neurovascular diseases, including #stroke and Small vessel disease, and we investigated occurrence of disease-relevant phenotypes. 6/10