Urinary mRNA for the Diagnosis of Renal Allograft Rejection: The Issue of Normalization

Urinary messenger RNA (mRNA) quantification is a promising method for noninvasive diagnosis of renal allograft rejection (AR), but the quantification of mRNAs in urine remains challenging due to degradation. RNA normalization may be warranted to overcome these issues, but the strategies of gene normalization have been poorly evaluated. Herein, we address this issue in a case-control study of 108 urine samples collected at time of allograft biopsy in kidney recipients with (n = 52) or without (n = 56) AR by comparing the diagnostic value of IP-10 and CD3ε mRNAs-two biomarkers of AR-after normalization by the total amount of RNA, normalization by one of the three widely used reference RNAs-18S, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and Hypoxanthine-guanine phosphoribosyltransferase (HPRT)-or normalization using uroplakin 1A (UPK) mRNA as a possible urine-specific reference mRNA. Our results show that normalization based on the total quantity of RNA is not substantially improved by additional normalization and may even be worsened with some classical reference genes that are overexpressed during rejection. However, considering that normalization by a reference gene is necessary to ensure polymerase chain reaction (PCR) quality and reproducibility and to suppress the effect of RNA degradation, we suggest that GAPDH and UPK1A are preferable to 18S or HPRT RNA.