Characterizing AIP Variants in Pituitary Adenomas

The detailed experimental procedures and statistical analyses are provided in Supplemental Methods. In brief, in AIP half-life experiments, HEK293 and EBV-immortalized cells (derived from a control subject [EBV-LC-AIP_WT] and a patient with a heterozygous AIPmut [EBV-LC-AIP_p.R304*]) were treated with cycloheximide (CHX, 100 μg/mL) and total protein was extracted at various time points for Western blot (WB). RNA stability was analyzed using RT-qPCR. Fifteen AIP variant plasmids (obtained by site-directed mutagenesis, Tables 1 and 2) were transfected into HEK293 cells for half-life experiments (CHX, 20 μg/mL, Table 2), in the presence or absence of the proteasome inhibitor MG-132 (20 μM). Genetic and clinical data were collected from 100 pituitary adenoma patients [60 from our cohort (4 (link)) and 40 cases reported in the literature] carrying the missense AIP variants as well as the nonsense mutation we have studied here. AIP interacting partners were identified in a pull-down assay from rat GH-secreting GH3 cells using glutathione-S-transferase-fused WT and mutant AIP synthetic proteins followed by quantitative mass spectrometry. Coimmunoprecipitation and siRNA knockdown (KD) of FBXO3 were used to study the role of FBXO3 in AIP degradation. Half-life experiments were analyzed using a one-phase decay equation, and the degradation speed (K) was compared between each mutant protein and the WT protein using the extra sum-of-squares F test. Correlation between half-life and clinical features was analyzed using data derived from studies listed in Table 3 using the Spearman R test. For other analyses, Kruskal-Wallis, one-way ANOVA, and linear regression tests were used as appropriate, and significance was taken as P < .05.