The regulated turnover of synaptic vesicle (SV) proteins is thought to involve the ubiquitin-dependent tagging and degradation through endo-lysosomal and autophagy pathways. Yet, it remains unclear which of these pathways are used, when they become activated, and whether SVs are cleared en masse together with SV proteins or whether both are degraded selectively. Equally puzzling is how quickly these systems can be activated and whether they function in real-time to support synaptic health. To address these questions, we have developed an imaging-based system that simultaneously tags presynaptic proteins while monitoring autophagy. Moreover, by tagging SV proteins with a light-activated ROS generator, Supernova, it was possible to temporally control the damage to specific SV proteins and assess their consequence to autophagy-mediated clearance mechanisms and synaptic function. Our results show that, in mouse hippocampal neurons of either sex, presynaptic autophagy can be induced in as little as 5-10 min and eliminates primarily the damaged protein rather than the SV en masse. Importantly, we also find that autophagy is essential for synaptic function, as light-activated damage to, for example, Synaptophysin only compromises synaptic function when autophagy is simultaneously blocked. These data support the concept that presynaptic boutons have a robust highly regulated clearance system to maintain not only synapse integrity, but also synaptic function.

Anti–N‐methyl‐D‐aspartate receptor (NMDAR) encephalitis is the most common autoimmune encephalitis related to autoantibody‐mediated synaptic dysfunction. Cerebrospinal fluid–derived human monoclonal NR1 autoantibodies showed low numbers of somatic hypermutations or were unmutated. These unexpected germline‐configured antibodies showed weaker binding to the NMDAR than matured antibodies from the same patient. In primary hippocampal neurons, germline NR1 autoantibodies strongly and specifically reduced total and synaptic NMDAR currents in a dose‐ and time‐dependent manner. The findings suggest that functional NMDAR antibodies are part of the human naïve B cell repertoire. Given their effects on synaptic function, they might contribute to a broad spectrum of neuropsychiatric symptoms. Ann Neurol 2019;85:771–776

We thank Thorsten Trimbuch and the Viral Core Facility of the Charité -Universitätsmedizin Berlin for cloning and production of viral constructs; Marisa Brockmann for electrophysiological training; Aleksandra Ichkova for valuable comments on the manuscript; Anny Kretschmer, Christine Bruns, Bettina Brokowski and Katja Poetschke for technical assistance; and the Advanced Medical Bioimaging Core Facility (AMBIO) of the Charité -Universitätsmedizin for support in acquisition of the imaging data.

Autoantibodies against central nervous system proteins are increasingly being recognized in association with neurological disorders. Although a growing number of neural autoantibodies have been identified, a causal link between specific autoantibodies and disease symptoms remains unclear, as most studies utilize patient derived cerebrospinal fluid (CSF) containing mixtures of autoantibodies. This raises questions concerning mechanism of action and which autoantibodies truly contribute to disease progression. To address this issue, monoclonal autoantibodies were isolated from a young girl with a range of neurological symptoms, some of which reacted with specific GABA A receptor subunits, α1-and α1γ2-subunits, which in this study, we have characterized in detail using a combination of cellular imaging and electrophysiological techniques. These studies in neurons from wild-type mice (C57BL/6J) (RRID:IMSR_JAX:000664) of mixed-sex revealed that the α1 and α1ɣ2 subunit-specific antibodies have differential effects on the GABA A receptor. Namely, the α1-antibody was found to directly affect GABA A receptor function on a short time scale that diminished GABA currents, leading to increased network excitability. On longer time scales those antibodies also triggered a redistribution of the GABA A receptor away from synapses. In contrast, the α1γ2-antibody had no direct effect on GABA A receptor function and could possibly mediate its effect through other actors of the immune system. Taken together these data highlight the complexity underlying autoimmune disorders and show that antibodies can exert their effect through many mechanisms within the same disease.

Autoantibodies against the NR1 subunit of NMDA receptors (NMDARs) have been shown to promote crosslinking and internalization of bound receptors in NMDAR encephalitis (NMDARE). This internalization-mediated loss of NMDARs is thought to be the major mechanism leading to pathogenic outcomes in patients.However, the role of bound autoantibody in engaging the resident immune cells, microglia, remains poorly understood. Here, using a patient-derived monoclonal NR1 autoantibody (hNR1-mAb) and a co-culture system of microglia and neurons, we could show that hNR1-mAb bound to hippocampal neurons led to microgliamediated removal of hNR1-mAb bound NMDARs. These complexes were found to accumulate inside endo-lysosomal compartments of microglia. Utilizing another patient isolated monoclonal autoantibody, against the α1-subunit of GABA A receptors (α1-GABA A -mAb), such removal of receptors was found to be specific to the antibody-bound receptor targets. Interestingly, along with receptor removal, we also observed a reduction in synapse number, more specifically in the numbers of post-synaptic proteins like PSD95 and Homer 1, when microglia were present in the culture. Importantly, mutations in the Fc region of hNR1-mAb,