Theodor Kocher Institute (TKI)

Research Groups

Research Overview Britta Engelhardt

The central nervous system (CNS) has a unique relationship with the immune system known as CNS immune privilege. We propose that CNS immune privilege is established by the endothelial blood-brain barrier (BBB), the epithelial blood-cerebrospinal fluid barrier (BCSFB), the blood-arachnoid barrier (BAB) and the glia limitans which divide the CNS into compartments that differ with respect to their accessibility to soluble and cellular components of the immune system. CNS brain barriers anatomy thus resembles the architecture of a medieval castle surrounded by two walls bordering a castle moat. Immune surveillance is accommodated in the two-walled rampart at the outer perimeter of the CNS parenchyma and allows for mounting immune responses without directly disturbing communication of neurons in the CNS parenchyma. Disruption or impaired function of the brain barriers will thus translate into clinical disease.

Our research is devoted to understanding the function of the different brain barriers in regulating CNS immune surveillance and how their impaired function contributes to neuroinflammatory diseases such as Multiple Sclerosis (MS) or Alzheimer’s disease (AD). Our laboratory combines expertise in vascular biology, neuroimmunology and live cell imaging and has developed sophisticated in vitro and in vivo approaches to study immune cell interactions with the brain barriers in health and neuroinflammation.

Research Achievements
Ongoing projects

Research Achievements

Multistep T cell extravasation across the BBB

Making use of primary mouse brain microvascular endothelial cells (pMBMECs) we have shown in collaboration with Ruth Lyck that CD4T cell arrest under physiological flow is mediated by endothelial VCAM-1 and ICAM-1, while endothelial ICAM-1 and ICAM-2 are essential for subsequent T-cell polarization and extended crawling against the direction of flow to sites permissive for diapedesis (Abadier et al., EJI, 2015). We identified BBB tricellular junctions as novel sites for T-cell diapedesis across the BBB (Casto Dias et al., J Cell Sci, 2021). In further studies we showed that high levels of endothelial ICAM-1 combined with de novo expression of the atypical chemokine receptor 1 (ACKR1) shuttling chemokines from the abluminal to the luminal side of the inflamed BBB, and/or impaired BBB junctional integrity lead to increased transcellular T-cell diapedesis across the inflamed BBB (Minten et al., Brain 2014; Wimmer et al., Front Immunol, 2019; Marchetti et al., EJI, 2021). Our data thus underscore that BBB junctional integrity is regulated independently of the cellular pathway of T-cell diapedesis across the BBB. This is supported by our observation that inducible endothelial expression of the tight junction sealing claudin-1 reduces BBB leakiness and EAE severity while leaving inflammatory cell recruitment into the CNS undisturbed (Pfeiffer et al., Acta Neuropathol, 2011).

Molecular composition of BBB tight junctions

Analyzing the molecular makeup of BBB tight junctions we clarified that claudin-3 is not expressed in brain endothelial cells (Castro Dias et al., Sci Reports, 2019), and that claudin-12 is not required for BBB integrity (Castro Dias et al., FBCNS, 2019).

The choroid plexus as immune cell entry site into the CNS

Since our original discovery that during experimental autoimmune encephalomyelitis (EAE) the adhesion molecules ICAM-1 and VCAM-1 are upregulated at the BCSFB of the choroid plexus in parallel to their upregulation at the BBB (Steffen et al., Am J Pathol, 1996), we have speculated on a function of the choroid plexus in CNS immunity. In collaboration with Federica Sallusto (Bellinzona/Zürich) we demonstrated that Th17 cells may enter the CNS in a CCR6/CCL20 dependent manner via the ChP to initiate EAE (Reboldi et al., Nat Immunol, 2009). We have succeeded to establish a valid in vitro model of the mouse BCSFB allowing for in depth modeling of the immune function of the BCSFB (Lazarevic et al., FBCNS, 2016). In addition, by using human in vitro BBB and BCSFB models we have shown that human T-cell subsets differ in their ability to cross the BBB versus the BCSFB in vitro (Nishihara et al, FBCNS, 2020).

Drug delivery across the BBB

Making use of nanoparticle chemistry in collaboration with Harm-Anton Klok (EPFL, Lausanne) we identified T cells as novel cellular targets to be explored for cell mediated drug delivery into the CNS (Ayer et al., Bioconjug Chem, 2021; Thomsen et al., Biomacromolecules, 2021; Ayer et al., Adv Healthc Mater, 2021)

Human models of the BCSFB and the BBB

We have established human models for the BCSFB and stem-cell derived in vitro models of the human BBB allowing to study the migration of human immune cells across these barriers (Nishihara et al., FBCNS, 2020). In collaboration with James Mc Grath (Rochester University, NY, USA) we have developed a breakthrough microfluidic device (mSIM-CVB) to investigate the migration of rare patient derived immune cell subsets across the BBB under physiological flow by live cell imaging (Mossu et al, JCBFM, 2019). In collaboration with Renaud Du Pasquier (Lausanne, CH) and Eric Shusta (Madison-Wisconsin, USA) we developed the extended endothelial-cell culture method (EECM) by differentiating human induced pluripotent stem cells (hiPSC) to brain microvascular endothelial (BMEC)-like cells. These BMEC-like cells display an inherent endothelial adhesion molecule phenotype making them the first uniquely suitable model to study human immune cell interactions with the BBB (Nishihara et al., FASEB J, 2020; Nishihara et al., STAR Prot, 2021). Establishing EECM-BMEC-like cells from hiPSC derived from MS patients and healthy controls (HC), we found that MS-derived BMEC-like cells display impaired barrier properties and an inflammatory phenotype with increased immune cell interaction compared to their HC counterparts. Thus hiPSC-derived EECM-BMEC-like cells are suitable to explore the molecular underpinnings of BBB dysfunction in neurological disorders and are thus eligible to study drug delivery across the BBB in neurological diseases.

Ongoing projects

Automated analysis of T cell interaction with the BBB under flow in vitro

With this project we aim to replace the manual cell-by-cell analysis of T cell behavior during their multi-step extravasation across the BBB under physiological flow by an automated and unbiased approach. We have designed an image processing pipeline for T-cell migration analysis in phase contrast microscopic images acquired in such assays. Cell segmentation has been based on a custom designed 2D+T Deep Convolutional Network. To account for the T cell migration under physiological flow a specialized cell tracking algorithm has been employed. This has set the stage for automated analysis of the migration patterns of individual T cells and will thus eventually allow to replace manual analysis of these assays.

Imaging the brain barriers in vivo

Two-photon intravital microscopy (2P-IVM) has advanced our understanding of CNS immune surveillance, but a limitation of current IVM technologies is that they lack simultaneous visualization of the brain barriers and immune cells. We have therefore developed novel fluorescent reporter mice allowing for simultaneous imaging of the arachnoid and pia mater as well as the glia limitans and the immune cells in vivo. Making use of cervical spinal cord and cranial window models and these fluorescent reporters, we currently investigate in collaboration with Benoit Zuber (Institute for Anatomy, University of Bern) and Christer Betsholtz (Uppsala, Sweden) the role of these different barriers in restricting T-cell migration within the spinal cord and brain during health and EAE by 2P-IVM.

Imaging cerebrospinal fluid drainage in vivo

To investigate the afferent arm of CNS immunity we make use of vascular reporter mice allowing to visualize lymphatic and blood vessels. To this end we have established a deep tissue imaging technique employing Gradient Refractive Index (GRIN) lenses allowing to directly visualize CSF drainage from the brain across the cribriform plate to the nasal mucosa.

Cellular and molecular mechanisms regulating CD8 T cell trafficking to the CNS under neuroinflammation.

CD8 T cells are the mayor contributors of the adaptive cellular immune response. When CD8 T cells recognize self-antigens, they differentiate into autoreactive effector CD8 T cells contributing to the initiation, progression, and regulation of autoimmune and inflammatory disorders. In particular, CD8 T cells have been associated to a number of neuroinflammatory and autoimmune disorders affecting the CNS including MS, cerebral malaria, virus-inflammatory brain diseases, and AD. However, the mechanisms regulating CNS entry of CD8 T cells remain to be fully understood. In this project we study the role of BBB endothelial adhesion molecules and MHC-class I mediated antigen presentation on CD8 T cell entry into the CNS. To this end we make use of cervical spinal cord window, cranial window or skull thinning preparations in brain barrier fluorescent reporter mice that allow us to visualize and study CD8 T cell migration across the CNS barriers in vivo by 2P-IVM.

Role of the choroid plexus in immune cell entry into the CNS during EAE

Following up on the role of the chemokine receptor CCR6 and its ligand CCL20 for Th17 cell entry into the CNS, we have investigated the migration of Th17 cell across in vitro models of the BCSFB and have found that CCR6 is not required for Th17 cell migration across the BCSFB. Using fluorescent brain barrier reporter mice developed in our laboratory and light sheet fluorescence microscopy we are currently investigating the anatomical routes of Th17 cell entry into the brain during health and EAE.

Discovering the molecular underpinnings of BBB dysfunction in MS

Here we make use of our recently developed hiPSC-derived in vitro model of the BBB from HC and MS patients, where we differentiate hiPSC into endothelial precursor cells and by the extended endothelial cell culture method (EECM) into brain microvascular like endothelial cells (EECM-BMEC-like cells). In collaboration with Renaud Du Pasquier (Lausanne, CH) and Eric Shusta (Madison, Wisconsin, USA) we have shown that EECM-BMEC-like cells from MS patients show impaired barrier properties and an inflammatory phenotype when compared to their HC counterparts (Nishihara et al., Brain, in press). BBB breakdown is an early hallmark of MS pathogenesis. The mechanisms leading to BBB breakdown are incompletely understood and are generally thought to be a consequence of the autoimmune neuroinflammatory process in MS. In this project we challenge this view and propose that intrinsic alterations in BBB endothelial cells manifested at the genetic or epigenetic, transcriptional and ultimately phenotypic level cause or contribute to altered BBB function, which in combination with additional risk factors are crucial for the development of clinical MS. To this end we have performed transcriptional profiling of MS- and HC- EECM-BMEC like cells and found numerous differentially regulated genes in MS-EECM-BMEC-like cells. We presently analyze the role of differentially regulated signaling pathways as well as individual molecules in contributing to BBB impairment in MS. Identification of the molecular underpinnings of BBB dysfunction in MS will set the stage for the development of an entirely novel generation of therapeutic approaches aiming for BBB stabilization.

Modeling the human BBB in neuroinflammation

BBB characteristics in brain microvascular endothelial cells rely on the continuous crosstalk with pericytes and astrocytes. Alterations in BBB properties lead to impaired barrier characteristics and increased immune cell infiltration into the CNS. The study of the human BBB is limited by the lack of in vitro tools that accurately model all functional aspects of it. In collaboration with James McGrath (University of Rochester, USA) we have proposed to address this need by building an in vitro human BBB on a μSiM (microphysiological system enabled by a silicon membrane) platform. Our “human BBB-on-a-chip” model features ultrathin, transparent, and permeable silicone membranes that mimic the interface between blood and brain. The membranes harbor ~1 billion 60 nm pores, which allow unhindered solute exchange between ‘blood’ and ‘brain’ compartments. The luminal side of the membrane is lined with human induced pluripotent stem cell (hiPSC)-derived BMECs, which mimic the “blood compartment”. The "brain compartment" on the abluminal side of the membrane constitutes of hiPSC-derived pericytes and astrocytes. The BMECs, pericytes and astrocytes -all differentiated from the same hiPSC source- will be incorporated into the chip to build a fully isogenic human neurovascular unit (hNVU). After completing the establishment of the μSiM-hNVU, we will include physiological flow to investigate immune cell trafficking in healthy control-derived and MS patient-derived BBB models.

Studying the effect of MS therapies on T cell interaction with the BBB

We investigate the effect of vitamin D and natalizumab, an a4-integrin blocking antibody, on the cellular and molecular mechanisms involved in the multi-step T-cell extravasation across the BBB. To this end, we employ different in vitro models of the human BBB such as for example CD34⁺ cord blood stem cell derived brain like endothelial cells (in collaboration with Fabien Gosselet, Lens, France) and hiPSC- derived brain microvascular endothelial cells (in collaboration with Renaud Du Pasquier, Lausanne, CH). These in vitro models permit us to study how vitamin D or natalizumab modulate human CD4 and CD8 T cell interaction with the BBB under both static and physiological flow conditions and provide further insights on the molecular mechanisms of action of these therapeutical strategies in MS.

Unravelling consequences of SARS-CoV-2 mediated inflammatory immune responses at the human brain barriers

Neurological symptoms in SARS-CoV-2-infected patients suggest an involvement of the CNS in COVID-19 pathology. Using our human in vitro brain barrier models, we are investigating whether iPSC derived- BMEC-like cells and pericytes forming the BBB or choroid plexus epithelial cells constituting the BCSFB are infected by SARS-CoV-2 and whether infection has an impact on their barrier properties.

Role of the angiogenic factor angiopoietin-2 (Ang-2) and the junctional adhesion molecule A (JAM-A) in monocyte-derived macrophage recruitment into the healthy and inflamed CNS

Several studies have shown a detrimental role of CNS infiltrating monocyte-derived macrophages aggravating disease progression and outcome in experimental autoimmune encephalomyelitis (EAE). We here combine the expertise about dynamics and function of CNS-invading macrophages from the Locatelli lab with cutting-edge techniques including live-cell imaging approaches of an in vitro BBB model together with the use of complex transgenic mouse models from the Engelhardt lab. Our aim is to unravel the mechanisms involved in the migration of monocyte-derived macrophages across the BBB into the CNS by focusing on two endothelial proteins: The angiogenic factor angiopoietin-2 (Ang-2), critical for blood vessel remodeling during pathology. Elevated levels of Ang-2 promote pathological angiogenesis as well as inflammation in various disease settings including ischemia, sepsis and different types of cancer partly by recruiting myeloid cells into peripheral vascular beds. Consequently, we ask whether Ang-2 could play a role in myeloid cell migration into the immune-privileged CNS as well. Using a mouse model with endothelial cell – specific overexpression of Ang-2 crossed with a reporter knock-in mouse line allowing to distinguish CCR2⁺ infiltrating from CX3CR1⁺ CNS-resident macrophages, we study the impact of Ang-2 overexpression on different myeloid subsets during steady state and during EAE. In addition we focus on the junctional adhesion molecule A (JAM-A) and its role in mediating the migration of innate immune cells across the BBB during EAE. We found that JAM-A deficient mice develop ameliorated EAE when compared to wild-type littermates. Lack of JAM-A does not affect CNS infiltration of T cells. Therefore we speculate that JAM-A plays a role in myeloid cell recruitment into the CNS during EAE. To this end we have crossed the above mentioned myeloid cell reporter mouse line into the JAM-A deficient line and currently study how lack of JAM-A affects the composition of myeloid cells in the CNS during EAE