A series of novel fluorescent BODIPY-anionic boron cluster conjugates bearing [BH] (5, 6), [3,3'-Co(1,2-CBH)] (7, 8), and [3,3'-Fe(1,2-CBH)] (9) anions have been readily synthesized from meso-(4-hydroxyphenyl)-4,4-difluoro-4-bora-3 a,4 a-diaza- s-indacene (BODIPY 4), and their structure and photoluminescence properties have been assessed. Linking anionic boron clusters to the BODIPY (4) does not alter significantly the luminescent properties of the final fluorophores, showing all of them similar emission fluorescent quantum yields (3-6%). Moreover, the cytotoxicity and cellular uptake of compounds 5-9 have been analyzed in vitro at different concentrations of B (5, 50, and 100 μg B/mL) using HeLa cells. At the lowest concentration, none of the compounds shows cytotoxicity and they are successfully internalized by the cells, especially compounds 7 and 8, which exhibit a strong cytoplasmic stain indicating an excellent internalization efficiency. To the best of our knowledge, these are the first BODIPY-anionic boron cluster conjugates developed as fluorescent dyes aiming at prospective biomedical applications. Furthermore, the cellular permeability of the starting BODIPY (4) was improved after the functionalization with boron clusters. The exceptional cellular uptake and intracellular boron release, together with the fluorescent and biocompatibility properties, make compounds 7 and 8 good candidates for in vitro cell tracking.

N-methyl D-aspartate receptor (NMDAR) encephalitis is an immune-mediated disorder characterized by a complex neuropsychiatric syndrome together with a reduction of NMDAR. Although in most patients the life-threatening symptoms of the acute stage resolve with immunotherapy, memory and executive functions remain altered for several months or years. A mechanistic explanation for these long-lasting cognitive effects is still lacking and previous animal models have not explored this effect. Here, we combined repeat calcium imaging of the same population of hundreds of hippocampal CA1 neurons for three months along with two behavioral tasks to assess retrograde and anterograde memory loss using a reported mouse model of cerebroventricular transfer of patients' CSF antibodies. We measured how memory-related neuronal activity is affected by the presence of NMDAR antibodies during the induction of the model and its long-lasting recovery. In addition, we developed a computational model that provides a mechanistic explanation for the long-term antibody-mediated impairment of memory. The findings show that the presence of antibodies leads to an increase of CA1 neuronal firing rate, resulting in a reduction of the amount of information encoded by these cells. Furthermore, the antibodies cause a degradation of the hippocampal neuronal response stability over time, providing a neural correlate of memory dysfunction. All these neuronal alterations span the 3 months of recordings, and in some cases beyond the last recording point. The computational model shows that a reduction of NMDAR is sufficient to cause the changes observed in neuronal activity, including the different involvement of excitatory and inhibitory inputs to CA1 neurons. Altogether, we show that the antibody-mediated reduction of NMDAR leads to long-term changes in hippocampal neuronal activity which extend far beyond the antibody clearance, providing a mechanism that can account for the cognitive deficits observed in the protracted recovery of patients with anti-NMDAR encephalitis.