The switch from proteasome to immunoproteasome is increased in circulating cells of patients with fast progressive immunoglobulin A nephropathy and associated with defective CD46 expression.

The proteasome to immunoproteasome (iPS) switch consists of ß1, ß2 and ß5 subunit replacement by low molecular weight protein 2 (LMP2), LMP7 and multicatalytic endopeptidase-like complex-1 (MECL1) subunits, resulting in a more efficient peptide preparation for major histocompatibility complex 1 (MHC-I) presentation. It is activated by toll-like receptor (TLR) agonists and interferons and may also be influenced by genetic variation. In a previous study we found an iPS upregulation in peripheral cells of patients with immunoglobulin A nephropathy (IgAN). We aimed to investigate in 157 IgAN patients enrolled through the multinational Validation Study of the Oxford Classification of IgAN (VALIGA) study the relationships between iPS switch and estimated glomerular filtration rate (eGFR) modifications from renal biopsy to sampling. Patients had a previous long follow-up (6.4 years in median) that allowed an accurate calculation of their slope of renal function decline. We also evaluated the effects of the PSMB8/PSMB9 locus (rs9357155) associated with IgAN in genome-wide association studies and the expression of messenger RNAs (mRNAs) encoding for TLRs and CD46, a C3 convertase inhibitor, acting also on T-regulatory cell promotion, found to have reduced expression in progressive IgAN. We detected an upregulation of LMP7/ß5 and LMP2/ß1 switches. We observed no genetic effect of rs9357155. TLR4 and TLR2 mRNAs were found to be significantly associated with iPS switches, particularly TLR4 and LMP7/ß5 (P < 0.0001). The LMP7/ß5 switch was significantly associated with the rate of eGFR loss (P = 0.026), but not with eGFR at biopsy. Fast progressors (defined as the loss of eGFR >75th centile, i.e. -1.91 mL/min/1.73 m2/year) were characterized by significantly elevated LMP7/ß5 mRNA (P = 0.04) and low CD46 mRNA expression (P < 0.01). A multivariate logistic regression model, categorizing patients by different levels of kidney disease progression, showed a high prediction value for the combination of high LMP7/ß5 and low CD46 expression.