Meta menu:

From here, you can access the Emergencies page, Contact Us page, Accessibility Settings, Language Selection, and Search page.

  • Contact us:

    Campus Charité MitteCharitéplatz 1 (local address: Bonhoefferweg 3)
    10117 Berlin

  • Accessibility:
    Contrast Settings Change contrast
    Font size
    Font size bigger: STRG+ Font size smaller: STRG-

    You can enlarge or reduce the browser window. Please use CTRL and + to zoom in or CTRL and - to zoom out. Press CTRL and 0 to reset your browser window to normal size.

Open Menu

Autoimmunity and Neurodegeneration

We focus on the role of autoantibodies targeting the nervous system in clinical conditions ranging from encephalitis to neurodegeneration.

We are part of the German Center for Neurodegenerative Diseases (DZNE) Berlin.

You are here:


An increasing number of neuropsychiatric diseases are caused by pathogenic anti-neuronal autoantibodies, such as those targeting NMDA receptors, GABA receptors, LGI1, Caspr2, IgLON5, metabotropic glutamate receptors, and entirely novel autoantigens. Associated clinical conditions range from acute encephalitis to slowly progressive dementia. Our research focuses on the detailed analysis of mechanisms of autoantibody-mediated impairment of neuronal function, using recombinant monoclonal autoantibodies derived from patients’ cerebrospinal fluid, but also on the development of novel diagnostics and innovative antibody-selective immunotherapies.


Cloning and recombinant production of patient-derived monoclonal autoantibodies

Autoantibodies targeting a variety of neuronal surface proteins have recently been discovered in a multitude of neurological disorders and continuously define novel autoimmune entities such as NMDA-receptor encephalitis. However, first characterizations of these antibodies were limited to experiments that used bio-samples containing disease-specific but also further antibodies that were present in these patients. Thereby, these studies could not link the observed effects directly to single antibodies. To overcome that limitation and allow the characterization of functional effects for each antibody individually, we have developed a methodology to isolate single (monoclonal) antibodies from patients’ cerebrospinal fluid (Figure). This core technology uses optimized molecular biological techniques together with custom bioinformatic analytics to encode genetic information from large sets of antibody-producing immune cells to then separately reassemble single antibodies in theoretical unlimited amounts. Using this approach, we have successfully isolated disease-defining monoclonal antibodies in multiple forms of autoimmune encephalitis (such as NMDAR, LGI1, Caspr2, IgLON5, mGluRs) and provided proof of their pathogenicity using in vitro and in vivo assays. Besides studying their biological functions, each of these recombinant antibodies can also be used as a tool in scientific applications (e.g., high-resolution microscopy), for optimization of antibody diagnostics or as a starting point for novel antibody-depleting therapies.

In the face of the COVID-19 pandemic, we used the potential of our pipeline to generate SARS-CoV-2 neutralizing monoclonal antibodies to support the global efforts to understand the immune response against the novel coronavirus.

Identifying disease mechanisms of human autoantibodies

Autoantibodies targeting neuronal antigens are one of the diagnostic hallmarks of autoimmune encephalitis. Disease phenotype and severity depend on the underlying antigen. To understand whether such autoantibodies are directly pathogenic and cause disease, our research group uses patient-derived monoclonal antibodies in different in vitro and in vivo models.

Using primary neurons, transfected cells and rodent brain slices, we study the molecular interaction between autoantibodies and their target antigens (Figure A). The specific binding region of the recombinant human autoantibodies can be determined using epitope mapping, where human cell lines are transfected with different target domains (B). Moreover, application of autoantibodies to neuronal cultures provides insight into the effector mechanisms in disease pathogenesis (C).

Autoantibody pathogenicity can be further characterized in vivo, for instance, through intrathecal (cerebroventricular) delivery of antibodies using osmotic pumps (D). Autoantibodies infused into the CNS target their cognate antigen and can reproduce certain clinical features observed in the human disease. Animal MR imaging and behavioral tests allow quantification of the clinical phenotype. Taking brain slices of treated animals, electrophysiology measurements can help to understand how autoantibody-driven alterations of synaptic currents disturb neuronal function (E). Finally, using flow cytometry, western blot and ELISA, analysis of cells and brain proteins provides mechanistic understand of antibody-mediated down-regulation of CNS proteins.

CAAR T-cells (Chimeric autoantibody receptor T cells)

NMDA receptor (NMDAR) encephalitis is the most common autoimmune encephalitis causing psychosis, epileptic seizures and cognitive impairment. The underlying pathogenic autoantibodies lead to internalization of the NMDAR as well as profound synaptic changes. NMDAR encephalitis is currently treated with broad immunosuppression or non-selective antibody removal, resulting in often treatment-limiting side effects or insufficient responses. Thus, there is a strong medical need for the highly selective depletion of monospecific antibody producing cells.

To selectively target NMDAR-specific B cells, we developed NMDAR-specific Chimeric AutoAntibody Receptor (CAAR) T cells. Conventional chimeric antigen receptor (CAR) T cells have recently revolutionized treatment of B cell malignancies (Figure A). A CAR consists of an antibody fragment conferring target recognition fused to intracellular activating and costimulatory domains. The antibody fragment against the tumor antigen (e.g. anti-CD19) mediates the specific antigen recognition by the CAR expressing T cell. In contrast, the target “antigen” for NMDAR-CAAR T-cells is a membrane-bound anti-NMDAR antibody on the surface of disease-related B cells (Figure B). Therefore, CAAR T-cells present a fragment of the NMDAR on their surface and specifically bind and destroy autoantibody-producing B cells, leaving the large number of other (“good”) B cells unaffected (Figure C).

Our proof-of-principle experiments demonstrated that human NMDAR-CAAR T-cells can be selectively activated by autoantibody-producing cells and kill their target cells in vitro and in vivo. Our work currently focuses on the translation of this concept into clinical application, as well as application of the CAAR concept to other neurological autoantibody-mediated diseases.

Identification of novel autoantibody epitopes

The discovery of autoreactive antibodies targeting structures of the central and peripheral nervous system has revolutionized our understanding of pathomechanisms in neuropsychiatric diseases. Despite their potential as disease driving reagents and diagnostic biomarkers, there is a high number of autoantibodies with unknown target epitopes and unclear pathophysiological relevance. In our target identification pipeline, we therefore aim to identify and characterize novel autoantigens. For this, we initially screen patients’ samples for autoantibody binding to different nervous system tissues such as brain and sciatic nerves. In case of characteristic binding patterns (Figure A), we use immunoprecipitation-mass spectrometry (IP-MS) to identify the targets of autoreactive antibodies in an unbiased manner (Figure B-C). Using Crosslinking-IP-MS (CL-IP-MS) we further aim to characterize the exact interfaces between the antibodies and their antigens.

Identified target antigens are then confirmed using overexpression of the respective proteins in HEK cells and antibody binding assays. After confirmation, we investigate the pathophysiological relevance of the autoantibodies as well as their diagnostic potential as possible biomarkers. The findings will not only help to understand the antibody-repertoire in neuropsychiatric diseases, but also pave the way for selective autoantibody-specific treatments in the future.

Maternofetal autoantibody transfer can impair neonatal brain development

During normal pregnancy, antibodies are continuously transferred from the maternal into the fetal circulation. This also includes pathogenic autoantibodies, which can enrich in the fetus and enter the brain, as the blood-brain barrier is not fully developed. Using a murine pregnancy model, we demonstrated that human monoclonal NMDAR autoantibodies can reach the fetal brain, result in reduction of NMDA receptors and long-lasting neuronal dysfunction. The offspring showed behavioral abnormalities and MRI changes throughout life. Pilot data from human mothers suggest that these autoantibodies may be more common in women who have a child with a neurobiological developmental or psychiatric disorder. Ongoing studies using human biosamples and murine models will demonstrate whether maternal autoantibodies are a risk factor for an entire spectrum of neuropsychiatric conditions, including autism, schizophrenia or ADHD. This might also include further autoantibodies targeting neuronal cells. In some mothers, therapeutic removal of autoantibodies may prevent brain diseases in the offspring.

Autoantibody diagnostics in human biosamples (serum and CSF)

Anti-neuronal and anti-glial autoantibodies are increasingly recognized as the autoimmune cause of encephalitis, psychosis and dementia. Some autoantibodies are well-established, their direct pathogenicity has been proven in experimental work, and their routine diagnostics has become clinical standard in neurological operating procedures. The repertoire of diseases-defining autoantibodies is still continuously expanding. To identify yet unknown autoantibodies, indirect immunofluorescence can be applied. For this, unfixed rodent brain sections are incubated with human cerebrospinal fluid or serum. If the samples contain novel autoantibodies targeting neuronal epitopes, the antibodies stick to the section and are visualized with fluorescently labelled secondary antibodies (Figure). In the past, many of these new findings resulted in recognition of an underlying autoimmune etiology, treatment of patients and marked clinical improvement. On a collaborative research basis, this service is available to clinical centers and laboratories (please use the request form for shipment of patient sample).

Selected publications

Noviello C, Kreye J, Teng J, Prüss H*, Hibbs R*. Structural mechanisms of GABAA receptor autoimmune encephalitis. Cell (in press) (2022)doi: 10.1016/j.cell.2022.06.025

Reincke SM, Yuan M, Kornau HC, Corman V, …, Prüss H*, Wilson IA*, Kreye J*. SARS-CoV-2 Beta variant infection elicits potent lineage-specific and cross-reactive antibodies. Science 18;375:782-287 (2022), doi: 10.1126/science.abm5835

Kreye J, Wright SK, van Casteren A, Stöffler L, Machule ML, et al., ...Prüss H. Encephalitis patient-derived monoclonal GABAA receptor antibodies cause epileptic seizures. J Exp Med 218(11):e20210012 (2021)doi: 10.1084/jem.20210012

Prüss H. Autoantibodies in neurological disease. Nat Rev Immunol 1-16, doi: 10.1038/s41577-021-00543-w

Kreye J, Reincke SM, Kornau HC, Sánchez-Sendin E, Corman VM, Liu H, et al., …Prüss H. A Therapeutic Non-self-reactive SARS-CoV-2 Antibody Protects from Lung Pathology in a COVID-19 Hamster Model. Cell 183(4):1058-1069.e19 (2020), doi: 10.1016/j.cell.2020.09.049

Kornau HC, Kreye J, Stumpf A, Fukata Y, Parthier D, Sammons RP, Imbrosci B, Kurpjuweit S, Kowski AB, Fukata M, Prüss H*, Schmitz D*. Human Cerebrospinal Fluid Monoclonal LGI1 Autoantibodies Increase Neuronal Excitability. Ann Neurol 87(3):405-418 (2020)doi: 10.1002/ana.25666

Jurek B, Chayka M, Kreye J, Lang K, Kraus L, Fidzinski P, et al., …Prüss H. Human gestational N-methyl-d-aspartate receptor autoantibodies impair neonatal murine brain function. Ann Neurol 86(5):656-670 (2019)doi: 10.1002/ana.25552

Wenke NK, Kreye J, Andrzejak E, van Casteren A, Leubner J, Murgueitio MS, Reincke SM, Secker C, Schmidl L, Geis C, Ackermann F, Nikolaus M, Garner CC, Wardemann H, Wolber G, Prüss H. N-methyl-D-aspartate receptor dysfunction by unmutated human antibodies against the NR1 subunit. Ann Neurol 85(5):771-776 (2019)doi: 10.1002/ana.25460

Kreye J, Wenke NK, Chayka M, Leubner J, Murugan R, et al., …Prüss H. Human cerebrospinal fluid monoclonal N-methyl-D-aspartate receptor autoantibodies are sufficient for encephalitis pathogenesis. Brain 139:2641-52 (2016)doi: 10.1093/brain/aww208

Group Leader

Prof. Dr. Harald Prüß

Director Department of Experimental Neurology

CCM: Campus Charité Mitte

Team Members

Friederike Antonia Arlt

Clinician Scientist

Sonja Blumenau


Chiara Bode

Personal Assistant to Prof. Prüß

Doreen Brandl


Emilie Buchholz

Medical student

Dr. Maria Buthut

Clinician Scientist

Max-Pelgrom de Haas

Medical student

Dr. Sophie Lan-Linh Duong

Clinician Scientist

Julius Hoffmann

Medical student

Marie-Alice Homeyer

Medical student

Dr. Jakob Kreye

Clinician Scientist

Marie-Luise Machule

Clinician Scientist

Dr. Momsen Reincke

Clinician Scientist

Dr. Rosa Rössling

Clinician Scientist

Finja Staabs

Medical student

Laura Stöffler

Medical student

Elli Trampenau

Medical student

Dr. Niels von Wardenburg

Clinician Scientist