# Human stem cell–derived neurons and astrocytes to detect novel auto-reactive IgG response in immune-mediated neurological diseases

## TL;DR

This study uses human stem cell-derived neurons and astrocytes to detect new auto-reactive antibodies in patients with immune-related neurological diseases who test negative in standard tests.

## Contribution

The novel contribution is a cell-based assay using hiPSC-derived neurons and astrocytes to identify previously undetected neural antibodies in seronegative patients.

## Key findings

- The assay detected neural-specific antibodies in 19 out of 99 patients, with higher frequency in inflammatory neurological disease cases.
- Antibody reactivity was specific to neural tissue, with no binding to peripheral blood mononuclear cells.
- The assay identified antibodies in seronegative NMOSD and probable autoimmune encephalitis cases.

## Abstract

Up to 46% of patients with presumed autoimmune limbic encephalitis are seronegative for all currently known central nervous system (CNS) antigens. We developed a cell-based assay (CBA) to screen for novel neural antibodies in serum and cerebrospinal fluid (CSF) using neurons and astrocytes derived from human-induced pluripotent stem cells (hiPSCs).

Human iPSC-derived astrocytes or neurons were incubated with serum/CSF from 99 patients [42 with inflammatory neurological diseases (IND) and 57 with non-IND (NIND)]. The IND group included 11 patients with previously established neural antibodies, six with seronegative neuromyelitis optica spectrum disorder (NMOSD), 12 with suspected autoimmune encephalitis/paraneoplastic syndrome (AIE/PNS), and 13 with other IND (OIND). IgG binding to fixed CNS cells was detected using fluorescently-labeled antibodies and analyzed through automated fluorescence measures. IgG neuronal/astrocyte reactivity was further analyzed by flow cytometry. Peripheral blood mononuclear cells (PBMCs) were used as CNS-irrelevant control target cells. Reactivity profile was defined as positive using a Robust regression and Outlier removal test with a false discovery rate at 10% following each individual readout.

Using our CBA, we detected antibodies recognizing hiPSC-derived neural cells in 19/99 subjects. Antibodies bound specifically to astrocytes in nine cases, to neurons in eight cases, and to both cell types in two cases, as confirmed by microscopy single-cell analyses. Highlighting the significance of our comprehensive 96-well CBA assay, neural-specific antibody binding was more frequent in IND (15 of 42) than in NIND patients (4 of 57) (Fisher’s exact test, p = 0.0005). Two of four AQP4+ NMO and four of seven definite AIE/PNS with intracellular-reactive antibodies [1 GFAP astrocytopathy, 2 Hu+, 1 Ri+ AIE/PNS)], as identified in diagnostic laboratories, were also positive with our CBA. Most interestingly, we showed antibody-reactivity in two of six seronegative NMOSD, six of 12 probable AIE/PNS, and one of 13 OIND. Flow cytometry using hiPSC-derived CNS cells or PBMC-detected antibody binding in 13 versus zero patients, respectively, establishing the specificity of the detected antibodies for neural tissue.

Our unique hiPSC-based CBA allows for the testing of novel neuron-/astrocyte-reactive antibodies in patients with suspected immune-mediated neurological syndromes, and negative testing in established routine laboratories, opening new perspectives in establishing a diagnosis of such complex diseases.