Inaccurate primer design and guideline non-compliance limit translational validity of long non-coding RNA research

Abstract

Long non-coding RNAs (lncRNAs) are at the research spotlight as inflammation regulators. We systematically searched the asthma animal model literature and evaluated the accuracy of the reported real-time quantitative polymerase chain reactions (RT-qPCRs) on lncRNAs and guideline compliance, selecting 32 among 1204 articles retrieved from PubMed, EMBASE, Scopus and Web of Science (highly biased according to SYRCLE). Primer specificity was analysed using Primer-BLAST, UCSC in-silico PCR (GRCm38 and GRCm39), and the Ensembl BLAST/BLAT tools (GRCr8). We found 47 lncRNAs, five mentioned in at least 2 of the 12 articles (Tug1, Pvt1, Bcyrn1, Gas5 and H19). Less than 35% of the MIQE criteria were met by 10/12 articles, whereas at least 50% of both the ARRIVE and GCCP 2.0 criteria were met simultaneously by 7/12 articles. Five did not present primer sequences, and only 3/8 presented primers with specific alignment (to H19 and Gas5), with the highest MIQE compliance (circa 40%). All other primers were predicted to anneal elsewhere in the genome, albeit with mismatches. Based on the critical flaws found in primer design and low to moderate guideline adherence (MIQE, ARRIVE and GCCP 2.0), we strongly suggest checking for primer specificity and general caution in appreciating published results on lncRNA levels in animal models, as well as reinforcement of methodological rigour to assure faster translation of lncRNA findings into clinical practice.