Schedule Slot

Alzheimer’s Disease (AD), the 7th leading cause of death worldwide, is a progressive neurodegenerative disorder characterized by amyloid β-peptide (Aβ) plaques and neuronal loss. The limited efficacy of treatments targeting Aβ has shifted the attention toward cerebrovascular dysfunctions. In particular, damage to the blood-brain barrier (BBB) (e.g. increased permeability, cell-cell junction disruption, vascular Aβ deposition) is now recognized as a key player in AD pathophysiology. Traditional human in vitro models involve the use of immortalized cell lines or simplified monolayers of primary cells, hence showing limited ability to recapitulate the complexity of BBB dysfunction in AD. Our goal is to develop a human induced pluripotent stem cell (hiPSC)-derived 3D model of the BBB to investigate AD-related vascular alterations.

To assess the impact of genetic mutations associated with familial AD on BBB function, we generated brain endothelial-like cells, pericytes and astrocytes from two hiPSC lines homozygous and heterozygous for the APP Swedish mutation, along with the isogenic control line. A combination of flow cytometry, immunofluorescence and qPCR was employed to characterize key identity markers for each cell population. Differentiated cells were embedded in a 3D fibrin matrix within a mesoscale platform providing unidirectional hemodynamic stimuli to recreate a self-assembled model of the BBB.

Optimized differentiation protocols supported the successful generation of all three main cell types of the BBB from each hiPSC source. The expression level and proper localization of specific markers (e.g. ZO-1, Claudin-5, VE-Cadherin, vWF for endothelial cells; PDGFR-beta, NG2 for pericytes; GFAP for astrocytes) confirmed the acquisition of the correct phenotype. Moreover, some of these characterizations suggested cell-specific pathological phenotypes associated with AD. We are currently validating how these features impact endothelial barrier integrity and permeability in the 3D model.

Our hiPSC-derived cells express key cell-identity markers and represent a milestone for the biofabrication of a microphysiological model of the BBB. Future steps include the full molecular and functional characterization of the model, the study of the pathological crosstalk occurring in the perivascular brain microenvironment and its validation by comparison with tissue biopsies

Alzheimer’s Disease (AD) is a neurodegenerative disease, and the most common and severe type of dementia worldwide. In AD the pathological phenomena are well described – extracellular accumulation of amyloid-β (Aβ) peptides as amyloid plaques, and aggregation of abnormally-phosphorylated tau protein into intraneuronal neurofibrillary tangles in the cortex. Recent data point to a role of the adaptive immune system in AD pathogenesis, as clonally expanded effector memory CD8 T cells were found in the cerebrospinal fluid (CSF) and blood samples of AD patients. In the ArcAβ mouse model for AD, infiltration of CD3 T cells into the brain parenchyma was shown. To reach the central nervous system (CNS), T cells have to breach the brain barriers, including the blood-brain barrier (BBB), the blood-cerebrospinal fluid barrier (BCSFB), or the arachnoid barrier. Our goal is to investigate T cell entry into the CNS in AD and how different T cell subsets may be involved in AD pathogenesis.

To explore this, young and aged in-house developed brain border and brain barrier reporter mice (Aqp4-mRuby3; Cdh5-GFP and Cldn5-GFP; FoxJ1Cre; Ai14) were crossed with a mouse model for AD (ArcAβ) and ex vivo and in vivo imaging techniques were employed. Decalcified heads from age-matched control and ArcAβ animals were collected and stained for different T cell markers. To quantify the number of T cells from each subset present in both young and aged, healthy and ArcAβ mice, mononuclear cells were counted in tissue sections and assigned to different compartments in the cerebrum, ChP, and meninges. Two-photon intravital microscopy (2P-IVM) will be used to study the trafficking of the different T cell subsets across the respective brain borders and barriers.

Different subsets of T cells, including CD45+CD3+, CD45+CD4+ and CD45+CD8+ T cells were detected in perivascular locations, in the parenchyma, choroid plexus, and meninges, in both healthy and ArcAβ mice.

These findings will set the stage for future studies determining the role of T cells in AD pathogenesis and the development of novel therapeutic approaches targeting the adaptive immune system in AD.

This study aims to (1) uncover activity-dependent shifts in pericyte proteomic signatures and (2) spatially resolve emerging phenotypic heterogeneity.

We will combine sensory manipulation (unilateral whisker plucking or patterned stimulation) with a temporally controlled non-canonical amino acid labeling (mouse lines FUNCAT and PheRS) approach in ATP13a5-tdTomato × FUNCAT or ATP13a5-tdTomato × PheRS mice. Cell-type-specific protein labeling is achieved through tamoxifen administration, while temporal control during postnatal development is provided by administering non-canonical amino acids that label newly synthesized proteins. Labeled proteins will be captured by click chemistry and profiled by mass spectrometry, while high-resolution multiplex imaging will map pericyte marker expression under different sensory conditions.

We validated the dynamic expression of angiotensin-converting enzyme 2 (ACE2) in mouse brain pericytes during development. Reanalysis of the adult mouse brain pericyte transcriptome from Vanlandewijck et al. (2018) identified two pericyte populations defined by differential Ace2 expression, a finding supported by an independent dataset (Yao et al., 2023). ACE2 expression distinguishes two subpopulations—ACE2^high and ACE2^low—with region-specific patterns. ACE2^low pericytes are common in deep brain regions such as the septum, hypothalamus, and amygdala, while nearly all pericytes in motor and somatosensory cortex are ACE2^high. Within the somatosensory cortex, ACE2^low cells localize mostly to lower layers. During postnatal development, ACE2 expression appears gradually between days 6 and 13, starting in the cortical plate and spreading inward. By day 13, most cortical pericytes express ACE2, with deep brain pericytes acquiring expression later. By day 21, ACE2 expression resembles adult patterns. These results demonstrate the developmental and regional specificity of ACE2 expression, reflecting distinct pericyte subpopulations.

Pericytes in the mouse brain exhibit transcriptional and phenotypic heterogeneity that is region-dependent and progressively acquired during postnatal development. While our preliminary data confirm distinct ACE2^high and ACE2^low pericyte subpopulations with specific spatial distributions, the factors influencing this developmental diversification are still unknown.

We aim to elucidate the mechanisms underlying the different distribution pattern and resolution for cranial nerve lesions in these autoimmune neuroinflammatory diseases

We first report two unique clinical cases of inflammatory demyelinating disease, both of whom present with extensive optic neuritis and trigeminal nerve lesions at the root entry zone (REZ) with trigeminal nerve lesions appearing to terminate at the TZ. Consequently, we utilized the MOG35-55 induced active EAE model in CX3CR1GFP/CCR2RFP reporter mice, to characterize the cranial nerve lesions and evaluate if CCR2+ immune cell infiltrates recapitulate the clinical observations in these two patients. Subsequently, by characterization the anatomical arrangements of CNS barriers such as astrocytic foot processes and the arachnoid barrier and lymphatic vessels near the nerves, along with a functional analysis of the distribution of intracerebroventrically infused tracer we investigated the potential role of cerebrospinal fluid (CSF) and lymphatic vessels in the development and resolution of inflammatory lesions at cranial nerves during neuroinflammation.

We focused on the CNS-PNS TZ of the trigeminal and cochlear nerves in a the MOG35-55 EAE model. As shown in the graphical summary, these nerves were found to exhibit unique arrangements of anatomical barrier layers including the arachnoid and glia limitans, which affected CSF tracer distribution as well as CCR2+ immune cell infiltration. Our data demonstrated that CCR2+ immune cells accumulate at the TZ on both CNS and PNS side of the trigeminal nerve and cochlear nerve, which mirror the locations of cranial nerve pathology observed clinically in patients with inflammatory demyelinating disease. On the other hand, the optic and olfactory nerves, which both lack a TZ, did not exhibit restrictions in immune cell localization. Using the Prox1-GFP reporter mice, we did not identify lymphatic vessels around the trigeminal or cochlear nerve. Yet, the infiltrated CCR2+ cells seen at the EAE peak stage were significantly reduced at EAE chronic stage.

Overall, our results reconcile with the hypothesis that the segment of the cranial nerve that is exposed to CSF flow is more susceptible to CCR2+ immune cell infiltration.

Tumor vasculature is structurally and functionally abnormal, impeding oxygen and nutrient delivery and contributing to hypoxia, acidosis, high lactate levels, and elevated interstitial pressure. These conditions create an immunosuppressive microenvironment, support tumor progression, and reduce the efficacy of cancer therapies. Current strategies to normalize tumor vasculature primarily target angiogenic factors and are limited by substantial side effects. This study aims to explore an alternative approach by targeting myeloid cell MCT1, particularly in neutrophils, to improve vascular normalization and enhance chemotherapy efficacy.

We employed genetically engineered mouse models with myeloid/neutrophil-specific MCT1 deletion (MCT1ΔLyz2 and MCT1ΔS100a8/+) in subcutaneous melanoma (B16-F10) and lung carcinoma (LLC) models. Tumor progression was monitored, and subsequent analyses included flow cytometry and histological assessment of tumor sections. Vascular structure, immune cell infiltration, and localization were assessed, along with functional assays to evaluate oxygenation and chemotherapeutic response. Neutrophil depletion was achieved through anti-Ly6G antibody administration to examine the mechanistic contribution of neutrophils.

Deletion of MCT1 in myeloid cells led to increased vessel lumen diameter and enhanced pericyte coverage, contributing to improved tumor oxygenation. Notably, this was associated with increased tumor growth. Tumor immune profiling showed elevated neutrophil infiltration in MCT1-deficient tumors. In wild-type mice, neutrophils were primarily perivascular, whereas MCT1-deficient neutrophils localized within hypoxic tumor regions. Neutrophil depletion reversed the vascular normalization effects observed with MCT1 deletion, underscoring their role in the process. Importantly, MCT1 deletion in neutrophils enhanced the therapeutic efficacy of chemotherapy in primary tumors.

These findings identify neutrophil MCT1 as a critical regulator of tumor angiogenesis. Targeting MCT1 in neutrophils improves tumor vascular architecture and potentiates the response to chemotherapy. This strategy represents a promising avenue for improving current anti-cancer therapies by modulating the tumor microenvironment.

The Central Nervous System (CNS) barriers ensure the brain and spinal cord homeostasis by tightly regulating the exchange of immune cells and mediators between the CNS and the periphery. Among these, the leptomeninges and the glia limitans are important sites where immunosurveillance of the CNS border compartments takes place. We recently published our VE-cadherin-GFP reporter mouse line and described its applicability for imaging in vivo the leptomeninges, which constitute an important landmark for anatomical delimitation of the subarachnoid space (SAS). Here, we introduce our novel aquaporin-4 (AQP4)-mRuby3 knock-in mouse reporter to visualize another crucial CNS barrier: the glia limitans.

Using confocal and two-photon intravital microscopy (2P-IVM) in the brain and spinal cord of this mouse reporter, we demonstrated AQP4-mRuby3 colocalization with astrocytes and parenchymal basement membrane covering the surface of the CNS parenchyma and the perivascular spaces. Moreover, by 2P-IVM of VE-cadherin-GFP x AQP4-mRuby3 CNS border reporter mice we could simultaneously visualize blood vessels, the leptomeninges and the glia limitans along with their resulting compartments.

We confirmed a normal CNS water content in our heterozygous reporter, considering the role of AQP4 as a water channel. Crossing our reporter into the CX3CR1-GFP mouse line, we distinguished GFP+ microglia from border associated macrophages based on their location respective to the glia limitans. In vivo trafficking studies in the CNS border reporter during immunosurveillance and using different neuroinflammation models enabled the allocation of immune cells to the dura mater, SAS, blood vessels, subpial space and parenchyma. Furthermore, imaging the distribution of fluorescent tracers of varying sizes allowed to analyze the barrier properties of the glia limitans.

We here show that the glia limitans forms a barrier for soluble mediators, beads and immune cells. Combining the Aqp4-mRuby3 reporter with additional reporters for vascular, leptomeningeal or myeloid cells ensures precise localization of immune cells to CNS borders versus the CNS parenchyma allowing to assign functional roles in CNS immune surveillance versus neuropathology. Availability of the Aqp4-mRuby3 reporter mouse will further advance our understanding of the active role of the glia limitans in CNS immune privilege.

To investigate the role of Connexin 43 (Cx43)-mediated Gap Junction Intercellular Communication (GJIC) between tumor cells (TCs) and endothelial cells (ECs) during metastasis.

RNA-sequencing was performed on TCs and ECs isolated from lung metastatic foci to identify differentially expressed genes related to cell-cell communication. GJIC between ECs (bEnd3) and TCs (MC-38, LLC1.1) was assessed in vitro using the Calcein AM permeation assay. Cx43 was knocked down in TCs to examine its impact on GJIC with ECs. Cx43-knock-down TCs were tested in vivo, upon intravenous injection.

Cx43 was found to be overexpressed in both ECs and TCs. The knock-down of Cx43 in TCs led to a reduced ability to communicate with ECs in vitro. Mice injected with Cx43-knock-down TCs developed fewer and smaller metastatic lesions.

Cx43 mediates TCs-ECs communication during lung metastasis, promoting metastatic lesion formation. Further studies using Cx43-knock-out TCs and endothelial-specific deletion mouse models will delineate how Cx43 contributes to metastasis.

Afferent lymphatic vessels (LVs) are crucial for transporting leukocytes and antigens to draining lymph nodes (dLNs). Intravital imaging in murine skin has revealed that in the capillaries leukocytes actively migrate on the lymphatic endothelial cells (LECs). Rapid and passive transport to dLNs occurs only in collecting vessels because of the increased flow generated by contractions of surrounding lymphatic muscle cells (LMCs).
Using 3D imaging of whole-mount human skin biopsies, the Halin-lab has found that human skin contains mainly pre-collectors with lymphatic valves while lacking LMC coverage. Considering the expected low fluid flow in pre-collectors and the inefficient migration observed in capillaries, we hypothesize that the valves in pre-collectors might not only serve for fluid transport but also help to direct crawling leukocytes in the downstream migration, aiding the induction of adaptive immunity. Single-cell RNA sequencing of human dermal LECs uncovered a valve cluster enriched with genes connected with cell adhesion and immune regulation. Thus, we hypothesise that highly specific adhesion molecules like CD24, expressed in the valve can support the migration of dendritic cells (DCs) across the valve.

We are investigating whether the pre-collector valves influence the directionality of DC migration using time-lapse imaging in vivo or in ear skin explants using Prox1-mOrange x CD11c-YFP mice with fluorescent LVs and DCs. We are using an established algorithm in the lab to quantify the directionality of DC migration within lymphatic pre-collectors. In addition, we explore CD24 contribution to valve development and DC migration by taking advantage of CD24 knock-out (KO) mice and performing various in vitro and in vivo experiments, including confocal imaging, adoptive transfer and FITC painting.

Our experiments confirmed that CD24 is expressed on murine LN LECs, mouse dermal LECs, bone-marrow derived DCs and endogenous migratory DCs. Our research reveals that CD24 may impact the valve development in organ-specific manner. Ongoing research will reveal if CD24 has a role in regulating DC migration.

Our preliminary project holds the potential to uncover an unexpected new function of lymphatic valves in guiding intra lymphatic leukocyte migration and to identify novel strategies to manipulate this process by targeting valve-specific adhesion molecules.

The blood-brain barrier (BBB) is a protective cellular layer that regulates the passage of substances between the bloodstream and the brain, maintaining central nervous system (CNS) homeostasis. It is formed by brain microvascular endothelial cells (BMECs). Under physiological conditions, the BBB also controls immune cell trafficking into the CNS. Barrier properties are not intrinsic to BMECs but rely on the continuous crosstalk with pericytes and astrocytes which form the neurovascular unit (NVU).
In neurodegenerative and neuroinflammatory diseases such as Alzheimer’s disease (AD) and multiple sclerosis (MS), the integrity of the BBB is compromised.
In the recent years, human induced pluripotent stem cells (hiPSCs) have been increasingly used in disease modeling and in vitro drug discovery.
The main susceptibility gene for AD, apolipoprotein E4 (ApoE4) has recently been shown to play a role in BBB dysfunction. Our group has established differentiation of hiPSC into BMECs, pericytes and astrocytes allowing us to create fully isogenic NVU models. We have started to establish isogenic NVU models from hiPSCs derived from healthy controls (HC) homozygous for ApoE3 and from hiPSCs from AD patients homozygous for ApoE4 to investigate the role of ApoE4 on BBB integrity. This will be achieved by culturing our NVU models microscale devices harboring nanoporous ultrathin silicon nitride membranes (μSiM). The modular μSiM design allows for co-culture of BMECs, pericytes and astrocytes and thus for in vitro modelling of the NVU.

The differentiation of the hiPSC cells into BMEC cells, pericytes and astrocytes.
Characterisation of the differentiated cells via immunofluorescence staining and multi-color flow cytometry.
Investigating barrier properties of the BMEC monolayer by measuring transendothelial electrical resistance (TEER) via continuous impedance spectroscopy and permeability using clearance of low molecular weight tracers.
BMEC cells are co-cultured with isogenic pericytes and astrocytes in a μSiM device.

Optimization of the differentiation protocols for the clones included in the project.
Initial characterisation of the BMEC cells and monolayer differentiated from those clones.

The project is ongoing. Further characterization of BMEC cells and barrier function is needed, including repeating the experiments under co-culture conditions.