Portal de Periódicos da CAPES

Methodological considerations for circulating long noncoding RNA quantification

2022; Elsevier BV; Volume: 28; Issue: 8 Linguagem: Inglês

10.1016/j.molmed.2022.05.011

1471-499X

David de Gonzalo‐Calvo, Miron Sopić, Yvan Devaux,

Circular RNAs in diseases

In the past decade, significant resources have been invested in long noncoding RNA (lncRNA) research. Despite the knowledge available, we are far from incorporation of lncRNA into clinical practice. Here, we emphasize the technical challenges in the field, hoping to provoke a response leading to new consensus and guidelines. In the past decade, significant resources have been invested in long noncoding RNA (lncRNA) research. Despite the knowledge available, we are far from incorporation of lncRNA into clinical practice. Here, we emphasize the technical challenges in the field, hoping to provoke a response leading to new consensus and guidelines. Although a potential clinical application in medical decision-making has been suggested for circulating cell-free lncRNAs [1.de Gonzalo-Calvo D. et al.Circulating long noncoding RNAs in personalized medicine: Response to pioglitazone therapy in type 2 diabetes.J. Am. Coll. Cardiol. 2016; 68: 2914-2916Crossref PubMed Scopus (16) Google Scholar], the majority of the novel findings have failed to be validated across different investigations. This is in part due to technical challenges and lack of standardized methods. Currently, the available literature does not provide recommendations for preanalytical and analytical variables to be considered during circulating lncRNA quantification. Moreover, highly heterogeneous strategies are often incorrectly used to compensate the methodological challenges. Here, we emphasize problems that could compromise the reproducibility of circulating lncRNA quantification, in an effort to encourage the field to define consensus and guidelines on these issues. The use of high-throughput technologies (i.e., RNA sequencing or microarrays) has enabled the discovery of a large number of disease-specific transcriptomic targets. However, because of the high cost of these methodologies, the new markers are usually validated using RT-qPCR on a larger number of samples. RT-qPCR is the gold standard technique to quantify circulating lncRNAs, yet not much consistency across different protocols can be found regarding the choice of the endogenous/exogenous control(s). Unfortunately, we are not anywhere near global consensus on that matter. Synthetic RNA spike-in templates (synthetic RNAs that lack sequence similarities to known transcripts) may be good exogenous controls to monitor the extraction, reverse transcription, and qPCR efficiency. However, exogenous controls cannot account for variations introduced before RNA isolation. An additional problem lies in the fact that, in most cases, synthetic Caenorhabditis elegans miR-39-3p (cel-miR-39-3p) is used as the spike-in control. Considering that this spike-in control is a miRNA, the differences associated with RNA isolation and miRNA–lncRNA quantification should be taken into account. Different spike-in mixtures for RNA quantification are commercially available, but the number of studies implementing these controls in their protocols is limited. When it comes to the use of endogenous controls, there is a lack of consensus on appropriate stable RNAs (those transcripts with acceptable expression levels that show low variation among experimental conditions) to normalize cell-free lncRNA levels [2.Vanhaverbeke M. et al.Peripheral blood RNA biomarkers for cardiovascular disease from bench to bedside: a position paper from the EU-CardioRNA COST Action CA17129.Cardiovasc. Res. 2021; (Published online October 14, 2021)https://doi.org/10.1093/cvr/cvab327Crossref PubMed Google Scholar]. The use of rRNAs or mRNAs as endogenous controls should be avoided because these transcripts may appear in the circulation due to cell lysis, have different biochemical characteristics from lncRNAs, and display a high expression variability. To our knowledge, one of the most suitable approaches may be to perform preliminary studies using multigene datasets obtained with RNA sequencing, microarray or RT-qPCR, and algorithms such as geNorm or NormFinder. Although this approach is recommended for every new set of experiments, it is time-consuming, is demanding in terms of input biological sample requirement, and can be costly. Endogenous controls with universally stable expression levels in a disease-specific manner remain to be determined. Blood samples are generally obtained following standardized protocols, after 12 h of fasting early in the morning in order to avoid the influence of postprandial state (lipemia) and circadian variation. However, this procedure for sample collection cannot be strictly followed in the case of sudden patient admission. Currently, there are no systematic studies that have assessed the effects of lipemia and circadian variation of circulating lncRNAs. On the one hand, considering the lipidic nature of extracellular vesicles (that are carriers of lncRNAs in the blood), it can be expected that high levels of lipoproteins in a postprandial state may affect their isolation from blood samples [3.Onódi Z. et al.Isolation of high-purity extracellular vesicles by the combination of iodixanol density gradient ultracentrifugation and bind-elute chromatography from blood plasma.Front. Physiol. 2018; 9: 1479Crossref PubMed Scopus (96) Google Scholar]. On the other hand, it is not known if lipemia per se, and to what extent, could influence the isolation of lncRNAs. In addition, some circulating RNAs and extracellular vesicles are implicated in regulatory networks of circadian rhythm, but again, to what extent this could lead to variation in levels of specific lncRNAs remains to be elucidated [4.Tao S.C. Guo S.C. Extracellular vesicles: potential participants in circadian rhythm synchronization.Int. J. Biol. Sci. 2018; 14: 1610-1620Crossref PubMed Scopus (21) Google Scholar]. Whether to use serum or plasma as a biological specimen constitutes a key point in the analysis of circulating lncRNA levels in blood. Decade-long experience in the quantification of circulating miRNAs showed that serum and plasma miRNA profiles can be quite different because of the possible release of platelet-enriched miRNAs during the coagulation process [5.Wang K. et al.Comparing the microRNA spectrum between serum and plasma.PLoS One. 2012; 7e41561Google Scholar]. In the field of lncRNAs, a large-scale deep sequencing analysis of human platelets identified more than 6000 lncRNAs [6.Sun Y. et al.Large-scale profiling of lncRNAs in human non-nucleated cells: Implications in cell function and disease.iScience. 2018; (Published online December 4, 2018)https://doi.org/10.2139/ssrn.3295649Google Scholar]. Consequently, similar to miRNAs, one should pay attention to platelets as a possible source of preanalytical variation when quantifying circulating lncRNAs. Hemolysis due to red blood cell rupture during blood collection and processing is also a potential confounding preanalytical factor that may alter the pool of circulating lncRNAs [7.Iempridee T. et al.Identification of reference genes for circulating long noncoding RNA analysis in serum of cervical cancer patients.FEBS Open Bio. 2018; 8: 1844-1854Crossref PubMed Scopus (15) Google Scholar]. Importantly, lncRNAs are sensitive to degradation by exonucleases (RNases), and it is critical, in order to limit this degradation and ensure good-quality and reproducible results, to process and freeze serum/plasma samples within 2 h after blood draw, avoid freeze–thaw cycles, and use an RNase-free working environment. As far as RNA extraction is concerned, several well-validated approaches are commercially available, based on either column extraction or capture beads. Both approaches can be semiautomated and work fine if handled in an RNase-free environment. Another critical step in the quantification of lncRNAs is the design of PCR primers. lncRNAs are often present in multiple isoforms that exhibit different or even opposite functions [8.Cho H. et al.Splice variants of lncRNA RNA ANRIL exert opposing effects on endothelial cell activities associated with coronary artery disease.RNA Biol. 2020; 17: 1391-1401Crossref PubMed Scopus (10) Google Scholar]. Therefore, the choice to design primers targeting one, multiple, or all isoforms can drastically affect results and their ability to be reproduced in different laboratories [9.Fang J. et al.Multiple non-coding ANRIL transcripts are associated with risk of coronary artery disease: a promising circulating biomarker.J. Cardiovasc. Transl. Res. 2021; 14: 229-237Crossref PubMed Scopus (5) Google Scholar]. Since lncRNAs are often found within protein-coding genes, on a sense or antisense strand, primers need to be carefully designed to specifically target the noncoding sequence and not the coding mRNA sequence derived from the same gene. Because many lncRNAs are not abundant in the circulation, the addition of a preamplification step before quantification may be crucial for lncRNA analysis. Preamplification is one of the points that raises a lot of debate and controversy in lncRNA quantification, with divergent results available in the state-of-the-art probably due to technical biases introduced in the preamplification-quantification workflow. In the study by Schlosser et al. [10.Schlosser K. et al.Assessment of circulating LncRNAs under physiologic and pathologic conditions in humans reveals potential limitations as biomarkers.Sci. Rep. 2016; 6: 36596Crossref PubMed Scopus (48) Google Scholar], with the exception of LIPCAR, the majority of lncRNAs tested in plasma were undetectable or sporadically detectable despite preamplification. Iempridee et al. [7.Iempridee T. et al.Identification of reference genes for circulating long noncoding RNA analysis in serum of cervical cancer patients.FEBS Open Bio. 2018; 8: 1844-1854Crossref PubMed Scopus (15) Google Scholar] demonstrated that target-specific preamplification causes a decrease of approximately five PCR cycles and proposed that this approach could be used to overcome the problem of low abundance. In the study by Jin et al. [11.Jin C. et al.Long non-coding RNA HULC as a novel serum biomarker for diagnosis and prognosis prediction of gastric cancer.Oncotarget. 2016; 7: 51763-51772Crossref PubMed Scopus (102) Google Scholar], preamplification also improved the raw Cq values of HULC lncRNA, but the relative expression levels were not significantly different. In addition to the preamplification protocols, other methodologies could be used for quantification of low-abundance circulating lncRNAs. Digital droplet PCR (ddPCR) is a very sensitive technique that enables precise absolute quantification of low-abundance RNA and DNA targets [12.Munjas J. et al.Placenta-specific plasma miR518b is a potential biomarker for preeclampsia.Clin. Biochem. 2020; 85: 57Crossref PubMed Scopus (1) Google Scholar]. So far, there are not much data on the use of ddPCR in lncRNA quantification. Considering that ddPCR is a much more sensitive tool that also enables absolute quantification of targets, and that absolute quantification is much more suitable for clinical applications, it seems that this technology could facilitate the bench-to-bedside translation of lncRNA biomarkers. In summary, critical factors, from blood collection and processing to quantification, should be taken into account to fully address the potential clinical application of circulating lncRNAs (Figure 1). In this scenario, there is an urgent demand for best practices guidelines and standard operating procedures in order to overcome current weaknesses in terms of reproducibility. Much remains to be done, such as the definition of quality controls and calibrators, which will allow advancing the technology readiness level of lncRNAs as a novel class of biomarkers for clinical use. These collaborative approaches should include not only laboratory specialists but also basic researchers and industry partners, thus creating a multidisciplinary environment that will catalyze the translation of circulating lncRNAs from bench to bedside. This article is based upon work from EU-CardioRNA COST Action CA17129 (www.cardiodrna.eu) supported by COST (European Cooperation in Science and Technology). D.d.G-C. has received financial support from Instituto de Salud Carlos III (Miguel Servet 2020: CP20/00041), cofunded by the European Social Fund (ESF)/'Investing in your future'. This work is supported by Instituto de Salud Carlos III (PI20/00577), cofunded by European Regional Development Fund (ERDF)/'A way to make Europe'. CIBERES is an initiative of the Instituto de Salud Carlos III. Y.D. is funded by the EU Horizon 2020 project COVIRNA (Grant Agreement 101016072), the National Research Fund (grants C14/BM/8225223, C17/BM/11613033, and COVID-19/2020-1/14719577/miRCOVID), the Ministry of Higher Education and Research, and the Heart Foundation-Daniel Wagner of Luxembourg. M.S. is funded by the Ministry of Education, Science and Technological Development, Republic of Serbia, through a grant agreement with University of Belgrade-Faculty of Pharmacy No. 451-03-68/2022-14/200161. Y.D. holds patents related to diagnostic and therapeutic applications of RNAs. The other authors have no interests to declare.