Cancer Diagnosis Program | Biorepositories and Biospecimen Research Branch

Search BRD Papers

Normalization strategies differently affect circulating miRNA profile associated with the training status.

Author(s): Faraldi M, Gomarasca M, Sansoni V, Perego S, Banfi G, Lombardi G

Publication: Sci Rep, 2019, Vol. 9, Page 1584

PubMed ID: 30733582 PubMed Review Paper? No

Purpose of Paper

Conclusion of Paper

Studies

1. Study Purpose

   The purpose of this study was to determine the most reliable method to normalize miRNA real-time PCR data. Blood was collected into K₂EDTA tubes from 10 men who exercised at moderate intensity for 150 min a week or were sedentary <30 min, 5 days a week (controls) and from 14 male non-professional athletes who had trained for the mountain ultra-trail race (8-10.5 h per week of running and resistance exercise at 70-90% maximal heart rate) a week after tapering from training. Plasma was obtained by centrifugation at 1300 x g for 10 min at room temperature. Plasma was frozen at -80˚C. miRNA was extracted using the miRCURY RNA Isolation Kit and reverse-transcribed using miRCURY LNA Universal RT. Efficiency was verified by real-time PCR of three synthetic spiked-in miRNAs (cel-39, UniSp2, and UniSp4). Real-time PCR was performed for the 179 miRNAs on the serum/plasma miRCURY LNA miRNA focus panel. Stable endogenous miRNAs and spiked-in controls were identified using the NormFinder algorithm. The relative expression of each miRNA was calculated based on the global mean CT, the mean CT of the most eight most stable miRNAs, the CT of individual endogenous miRNAs, the mean CT of the spike-ins and the CT of individual spiked-in oligonucleotides. Hemolysis was evaluated in specimens with a ΔCT ratio of >7 for miR-23 to miR-451.

   Summary of Findings:

   Of the 179 miRNAs evaluated, 67 were expressed in all specimens and 41 were identified as stable. This list was further narrowed down by only including the most stable miRNA in each cluster, resulting in a final list of 33 miRNAs. The most stable eight miRNAs had an accumulated SD of less than 50% and included miR-21-5p, miR-126-3p, miR-125a-5p, miR-590-5p, miR-26b-5p, miR-320d, miR-23a-3p, and miR-146a-5p and the least stable miRNAs were miR-486-5p, miR-144-3p, and miR-142-3p. The lowest mean CV was obtained when normalized to miR-320d (34.62%). Normalization using the global mean CT and mean endogenous CT resulted in mean CVs of 38.64% and 37.44%, respectively. As expected, normalization to any of the less stable miRNAs resulted in higher CVs (53.89-62.83%). Analysis of the percentage of the frequency distribution of the CV was narrowest when normalization was to the global mean CT or the mean endogenous CT. Further, the median CV value of the frequency distribution was lowest when the mean endogenous CT was used to normalize (median=0.318) followed by using miR-320d (median=0.323) and the global mean CT (median=0.326).

   Normalization of CT based on the mean CT values of cel-39, UniSp2, and UniSp4 was identified as the most suitable exogenous normalization method by Normfinder. The average CV was comparable when data was normalized to the mean CT (47.55%) or to either cel-39 (44.40%), UniSp4 (48.50%), or UniSp6 (48.88%) alone; however, analysis of the percentage of the frequency distribution of CV showed the lowest median value occurred after normalization to the mean rather than the individual spike-ins.

   The number of miRNAs up- and down-regulated in trained subjects versus controls differed among the normalization strategies employed. The expression profile of miRNAs showing a more than 20% difference in trained versus control individuals when normalized to the global mean was very similar regardless of whether normalized to global mean, mean of endogenous miRNAs, or miR-21-5p; but resulted in much larger fold changes in the opposite directions when normalized to the mean spike-in or cel-39 CTs. Two of the evaluated miRNAs showed a significant difference in level between trained and control subjects only when normalized to miR-320d, indicating increased sensitivity for these differences. Further analysis using miR-320d for normalization identified 40 miRNAs to be higher and 17 lower in trained individuals than control. As expected, this included skeletal muscle specific miRNAs and miRNAs known to play a role in myogenic differentiation, atrophy, and osteoblast and osteoclast differentiation.

   Biospecimens

   • Bodily Fluid - Plasma

   Preservative Types

   • Frozen

   Diagnoses:

   • Normal

   Platform:

   Analyte	Technology Platform
   RNA	Real-time qRT-PCR

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Real-time qRT-PCR Specific	Data handling	Normalized to global mean CT Normalized to the mean CT of the most eight most stable miRNAs Normalized to the CT of individual endogenous miRNAs Normalized to the mean CT of the spike-ins Normalized to CT of individual spike in oligonucleotides
   Preaquisition	Diagnosis/ patient condition	Trained for the mountain ultra-trail race Control individual

   View more