Identification of reference genes for the normalization of retinal mRNA expression by RT-qPCR in oxygen induced retinopathy, anemia, and erythropoietin administration

Abstract

Background Anemia and retinopathy of prematurity (ROP) are common comorbidities experienced by preterm infants, yet the role of anemia on the pathogenesis of ROP remains unclear. Reverse-transcriptase quantitative polymerase chain reaction (RT-qPCR) is a sensitive technique for estimating the gene expression changes at the transcript level but requires identification of stably expressed reference genes for accurate data interpretation. This is particularly important for oxygen induced retinopathy studies given that some commonly used reference genes are sensitive to oxygen. This study aimed to identify stably expressed reference genes among eight commonly used reference genes in the neonatal rat pups’ retina upon exposure to cyclic hyperoxia-hypoxia, anemia, and erythropoietin administration at two age groups (P14.5 and P20) using Bestkeeper, geNorm, and Normfinder, three publicly available, free algorithms, and comparing their results to the in-silico prediction program, RefFinder. Results The most stable reference gene across both developmental stages was Rpp30, as predicted by Genorm, Bestkeeper, and Normfinder. RefFinder predicted Tbp to be the most stable across both developmental stages. At P14.5, stability varied by prediction program; at P20, RPP30 and MAPK1 were the most stable reference genes. Gapdh, 18S, Rplp0, and HPRT were predicted as the least stable reference genes by at least one of the prediction algorithms. Conclusion Expression of Rpp30 is the least affected by experimental conditions of oxygen induced retinopathy, phlebotomy induced anemia and erythropoietin administration at both timepoints of P14.5 and P20.

Commentary: Identification of optimal reference genes for gene expression normalization in human osteosarcoma cell lines under proliferative conditions

“…Hence, three candidate RGs, namely, β-actin, β-2-microglobulin, and GAPDH, displayed satisfactory results (M-values <1.5), except for the 18S rRNA gene, which was considered the least stable (M-value = 2.4) under our test conditions ( Figure 1A ). This finding replicated other studies [( Molomjamts and Ingolfsland, 2023 ; Lewczuk et al, 2023 ; Gubern et al, 2009 ; Julian et al, 2014 ; Zhang et al, 2018 )]. Based on this, we realized that the 18S rRNA gene performed worst during human mRNA analysis in vitro upon drug stimulus due to its unstable expression.…”

“…In summary, it is crucial to carefully evaluate ahead for appropriate RG in expression studies for the accurate detection of meaningful changes in targeted gene transcript levels in human in vitro real-time qRT-PCR studies since normalization with unsuitable housekeeping gene products led to significant misinterpretation of expression profiles and final results ( Köhsler et al, 2020 ). Our data not only support those previous ones ( Molomjamts and Ingolfsland, 2023 ; Lewczuk et al, 2023 ; Gubern et al, 2009 ; Julian et al, 2014 ; Zhang et al, 2018 ) but also add another round of information regarding the selection and validation of ideal RGs for expression analysis. Finally, the ideal RG should have minimal regulation across the disease spectrum and be minimally influenced by patient heterogeneity ( Kumar et al, 2022 ).…”