Exploring brain insulin resistance in adults with bipolar depression using extracellular vesicles of neuronal origin

Accumulating evidence suggests that disrupted insulin signaling is involved in bipolar disorder (BD) pathogenesis. Herein, we aimed to directly explore the potential role of neuronal insulin signaling using an innovative technique based on biomarkers derived from plasma extracellular vesicles enriched for neuronal origin (NEVs). We leveraged plasma samples from a randomized, double-blind, placebo-controlled, 12-week clinical trial evaluating infliximab as a treatment of bipolar depression. We isolated NEVs using immunoprecipitation against neuronal marker L1CAM from samples collected at baseline and weeks 2, 6 and 12 (endpoint) and measured NEV biomarkers using immunoassays. We assessed neuronal insulin signaling at its first node (IRS-1) and along the canonical (Akt, GSK-3β, p70S6K) and alternative (ERK1/2, JNK and p38-MAPK) pathways. A subset of participants (n = 27) also underwent whole-brain magnetic resonance imaging (MRI) at baseline and endpoint. Pre-treatment, NEV biomarkers of insulin signaling were independently associated with cognitive function and MRI measures (i.e. hippocampal and ventromedial prefrontal cortex [vmPFC] volumes). In fact, the association between IRS-1 phosphorylation at serine site 312 (pS312-IRS-1), an indicator of insulin resistance, and cognitive dysfunction was mediated by vmPFC volume. In the longitudinal analysis, patients treated with infliximab, a tumor necrosis factor-alpha antagonist with known insulin sensitizing properties, compared to those treated with placebo, had augmented phosphorylation of proteins from the alternative pathway. Infliximab responders had significant increases in phosphorylated JNK levels, relative to infliximab non-responders and placebo responders. In addition, treatment with infliximab resulted in increase in MRI measures of brain volume; treatment-related changes in the dorsolateral prefrontal cortex volume were mediated by changes in biomarkers from the insulin alternative pathway. In conclusion, our findings support the idea that brain insulin signaling is a target for further mechanistic and therapeutic investigations.