engleski

Determination of the domain structure of DISC1 in mammalian cells identifies a region crucial for its aggregation in schizophrenia

Background To date, the pathophysiology of schizophrenia remains incompletely understood. While traditionally investigated using genetics, disrupted protein homeostasis has been suggested as a complementary approach to determine biological basis for these diseases in at least a subset of patients. Several proteins previously implicated by genetic studies have been shown to form insoluble aggregates in this illness, including Disrupted in Schizophrenia 1 (DISC1). DISC1 is a multi-functional scaffolding protein of significant importance for neurodevelopment and synaptic function. However, delineation of the domain structure of DISC1 has only recently been attempted, using high-throughput analysis in E. coli. This approach identified four distinct stable domains named D, I, S and C. Here, we used this knowledge to confirm and refine the structure of DISC1 in mammalian cells, as well as to identify the region of DISC1 responsible for its aggregation. Methods DISC1 domain borders were refined by combining previously published theoretical data with the recent empirical data. Constructs encoding variants of domains D, I and C were cloned, with various alternative domain boundaries, based on theoretical predictions. These were transfected into HEK293 cells and tested for stability via proteasome inhibition assay. Results were obtained by immunoblotting. In order to investigate which structural region(s) might be responsible for its aggregation, previously cloned fragments of DISC1, their combinations, and interaction partners were expressed in SH-SY5Y neuroblastoma cells. Localization patterns and aggregation were assessed by fluorescent microscopy. Binding to interaction partners was tested using the nanoscale-pulldown assay. Results A proteasome inhibition assay showed that modified versions of D and C domains show improved stability over their empirically-derived counterparts, with evidence that the I region may not represent a stable folded domain by itself. The D, S and C domains were further shown to be functional in isolation, based on interaction with known protein binding partners. Single domains exhibited clear cytoplasmic localization with no aggregate formation. In contrast, the combination of domains D and I showed clear signs of aggregation, with full length DISC1 used as a positive control. Based on this, we hypothesized that the unstructured region between these two domains, D and I, is responsible for DISC1 aggregation propensity, which was verified using further truncation constructs. Discussion Previous theoretical predictions of DISC1 sequence repeats indicate the existence of conserved structural loops. These loops are in overlap with some of the experimentally proposed domains. Here, we have shown that the loops are indeed essential for the domain stability. Moreover, changing the boundaries of some experimental domains to match the end of theoretical structural motifs noticeably increased their expression, suggesting that current DISC1 domain boundaries should be redefined. Additionally, we were able to narrow the aggregating region of DISC1 down to a cluster of amino acids near the center of the protein. Taking into consideration how aggregation affects the brains of patients with neurodegenerative diseases, these findings could present a powerful insight into pathological mechanisms of mental illness, thus enabling future therapeutic development.

Schizophrenia ; Protein aggregation, Domain structure

nije evidentirano