Cancer Diagnosis Program | Biorepositories and Biospecimen Research Branch

Search BRD Papers

Pre-profiling factors influencing serum microRNA levels.

Author(s): MacLellan SA, MacAulay C, Lam S, Garnis C

Publication: BMC Clin Pathol, 2014, Vol. 14, Page 27

PubMed ID: 25093010 PubMed Review Paper? No

Purpose of Paper

Conclusion of Paper

Studies

1. Study Purpose

   This study investigated potential effects of technical variability, the degree of hemolysis, intra-individual variability, and donor fasting and smoking on miRNA levels in serum from healthy individuals. To investigate potential effects of the degree of hemolysis, blood was collected from ten healthy individuals in two serum separator vacutainers tubes (SST). One tube was immediately passed through a 20 gauge needle several times while the other was left undisturbed until serum separation. To investigate the effects of donor fasting, SST blood was obtained from seven individuals one hour after eating a fatty meal and again 3 weeks later after an overnight fast.  To investigate the effects of smoking, blood was collected into SST tubes from ten current smokers and 10 healthy donors matched for age and gender that never smoked. To investigate the impact of intra-individual variability in miRNA levels, blood was collected into SST twice from 12 healthy volunteers at an interval ranging from 2 months 14 days to 17 months 3 days. All specimens were allowed to clot for 30 minutes before separation of serum by centrifugation at 1500 g for 15 min. Serum was immediately aliquoted and frozen at -80°C within 2 h of venipuncture. RNA was isolated from serum using the miRNeasy Mini Kit and in some cases the spike-in control cel-miR-39-3p was added during lysis.  A total of 742 miRNA were quantified by real-time PCR using miRNome real-time PCR panels. To assess the technical variability, two aliquots of serum from one “never smoker” were processed separately. To identify potential stable endogenous miRNA, the miRNA profile of serum from 26 lung adenocarcinoma patients, 50 oral carcinoma in situ (CIS)/oral squamous cell carcinoma (OSCC) patients, 8 healthy non-smokers, 3 healthy smokers, and the specimens included in the other studies in this paper were compared. Hemoglobin levels (g/L) in serum were calculated spectrophotometrically at an absorbance of 415 (A415), 380 (A380) and 450 nm (A450) using the following equation: (167:2 * A415− 83:6 * A380− 83:6* A450)/1000. Serum triglyceride levels were quantified using Triglyceride Assay Kit.

   Summary of Findings:

   A total of 157 miRNA were detected in both serum aliquots from the same healthy non-smoking individual. Of these, 48 miRNAs had ≥2-fold difference in levels between technical replicates, 14 miRNAs had ≥3-fold difference in levels between technical replicates and 5 miRNAs had ≥ 4-fold differences between replicates. Importantly, all of the miRNAs with a difference ≥2-fold had CT values >28. Using this information, the authors set a threshold of ≥3-fold difference between the samples compared. Hemoglobin concentrations were <0.1 g/L in all non-hemolyzed serum and >0.1 g/L in all hemolyzed serum specimens.  A total of 109 miRNA were detected in all hemolyzed and non-hemolyzed specimens, and 181 miRNAs were only detected in hemolyzed specimens and 116 miRNAs were only detected in non-hemolyzed specimens. Of the 109 miRNAs detected in both specimen types, only two miRNAs (miR-122-5p and miR-1266-5p) displayed a <2-fold difference between hemolyzed and non-hemolyzed specimens. Importantly, there was no difference in levels of the spike-in control (cel-miR-39-3p) between hemolyzed and non-hemolyzed specimens, indicating that differences with hemolysis were not attributable to differences in extraction efficiency. Although 231 miRNAs were quantified at levels that were ≥3-fold higher in at least one hemolyzed specimen compared to the matched non-hemolyzed specimen, only 162 were identified to be higher in 5 or more hemolyzed specimens compared to the control specimen. Further, only four miRNAs were observed at ≥ 3-fold higher levels in all of the hemolyzed sera compared to the matched non-hemolyzed sera. Hemoglobin concentrations were significantly correlated with serum levels of 177 miRNAs (P<0.05, all).  A total of 46 miRNA were detected in all 154 serum samples, of which miR-139-5p and miR-99a showed the lowest standard deviation and were consequently used as normalizers for the study of the effects of fasting and smoking. The number of miRNAs detected in serum of fasting patients was non-significantly higher than in the same patients after consuming a high fat meal, but no individual miRNA displayed a ≥3-fold difference in more than half of the cases. Although none of the miRNAs identified had levels that differed significantly between smokers and “never smokers” using the Mann-Whitney U test, normalized levels of miR-128-3p were slightly higher in smoker than "never smokers"(uncorrected P=0.001).  miRNA levels were strongly to very strongly correlated between serum specimens collected from the same individual approximately 2-17 months apart (R²=087-0.99). A total of nine miRNAs were found approximately 50% of the time in only one of the case-matched specimens and of these four were known to be affected by hemolysis.

   Biospecimens

   • Bodily Fluid - Serum
   • Bodily Fluid - Blood

   Preservative Types

   • Frozen

   Diagnoses:

   • Neoplastic - Carcinoma
   • Normal

   Platform:

   Analyte	Technology Platform
   Protein	Spectrophotometry
   RNA	Real-time qRT-PCR

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Preaquisition	Diagnosis/ patient condition	Fasting 1 h after consuming a high fat meal
   Biospecimen Aliquots and Components	Hemolysis	Fine needle aspiration-induced Not induced
   Biospecimen Acquisition	Time of biospecimen collection	Serum collected at two timepoints with an interval of from 2 months 14 days to 17 months 3 days 1 h after consuming a high fat meal 3 weeks later after overnight fast
   Preaquisition	Other drugs	Smoker Never smoker

   View more