Sources to variability in circulating human miRNA signatures.
Author(s): Keller A, Rounge T, Backes C, Ludwig N, Gislefoss R, Leidinger P, Langseth H, Meese E
Publication: RNA Biol, 2017, Vol. 14, Page 1791-1798
PubMed ID: 28820329 PubMed Review Paper? No
Purpose of Paper
This paper investigated the effects of patient age, inter- and intra-patient variability, and serum storage duration on the microRNA (miRNA, miR) profile and investigated if miRNA sequence features were associated with inter- and intra- patient variability.
Conclusion of Paper
The average coefficient of variance attributable to differences in patient age was significantly higher than the CV attributed to differences in serum storage duration but there were some miRNAs which displayed larger differences due to storage duration than patient age. Analysis of miRNA levels from the same individuals over time revealed miRNA that were more or less affected by intra- and inter-individual variability. Sequence analysis revealed that miRNA with higher inter-individual variability had a higher percentage guanine and lower percentage uracil and cytosine than those predominantly affected by intra-individual variability. Further analysis revealed motifs that indicated a miRNA was more likely to be affected by inter- or intra-individual variability. Of the 528 miRNAs identified in serum, only 135 miRNAs were found to not be significantly affected by inter-individual variability, patient age, serum storage duration, or differences attributed to sampling procedure changes over time.
Studies
-
Study Purpose
This study investigated the effects of patient age, inter- and intra-patient variability and serum storage duration on the miRNA profile and investigated if miRNA sequence features were associated with inter- and intra- patient variability. Three serum specimens collected at approximately 5 year intervals from 30 patients without cancer were obtained from the Janus tissue bank. Blood was collected into gel vials, tubes with sodium iodoacetate, or tubes without additive depending on collection year and were allowed to clot for 1 h or 14-28 h depending on specimen. Serum was obtained by unspecified method and was stored frozen at -25˚C for 23.5-40.5 years. As specimens were collected longitudinally, storage duration was modestly, inversely correlated with patient age at sampling (R=-0.54, P=0.00000003). The effect of patient age was studied using 15 specimens with maximal age variation (27.3 years between oldest [57 years] and youngest [30 years]), but storage durations were within 0.9 years (~34 years). Conversely, the effect of storage duration was examined using 15 specimens from patients of similar ages (~39-43 years, actual spread 2.9 years) that had been stored for 23.5-40.5 years. RNA was extracted using miRNeasy and quality was assessed by bioanalyzer. miRNA expression was determined using the SurePrint G3 8 x 60k miRNA microarrays.
Summary of Findings:
The average coefficient of variance in miRNA levels in the 15 specimens with similar storage duration (34 years) but widely different ages (30-57 years) was 0.48. The average CV for miRNA levels for specimens from individuals chosen based on being of similar age (39-43 years) but having been stored for 23.5-40.5 years was 0.38. The CV attributed to differences in patient age was significantly higher than the CV attributed to differences in serum storage duration (P<2.2 x 10-16). However, the CVs attributable to storage duration were higher than those attributed to patient age for some individual miRNAs such as miR-4730 and miR-4306 (1.04 versus 0.7 and 1.12 versus 0.57, respectively). Further analysis of miRNA levels from the same individuals revealed that some miRNAs, such as miR-6740-5p and miR-7847-3p, were not significantly affected by patient but were significantly different between sampling points for the same individual (P=1.2 x 10-7 and P=1.4 x 10-6, respectively). In the same analysis, there were 73 miRNAs, including miR-328-5p and miR144-3p, that varied more between individuals than between specimens from the same individual. Sequence analysis revealed that miRNAs with higher inter-individual variability had a higher percentage guanine than those predominantly affected by intra-individual variability (45% versus 35%, P=7.6 x 10-6) and those with higher inter-individual variability had more cytosine (25% versus 20%, P value not stated) and uracil (20% versus 16%, P value not stated). Further, miRNAs that varied over time rather than by individual were more likely to include CCCA (27% versus 9%), UCCC (26% versus 7%), or CCCC (30% versus 7%) and less likely to include GGGG (20% versus 47%), GGAG (25% versus 48%), or GCGG (9% versus 26%). Of the 528 miRNAs identified in serum, only 135 were found to not be significantly affected by inter-individual variability, patient age, serum storage duration, or differences attributed to sampling procedure changes over time.
Biospecimens
Preservative Types
- Frozen
Diagnoses:
- Not specified
Platform:
Analyte Technology Platform RNA DNA microarray RNA Automated electrophoresis/Bioanalyzer Pre-analytical Factors:
Classification Pre-analytical Factor Value(s) Preaquisition Patient age 30.3-63.8 years
Storage Storage duration 23.5-40.5 years
Biospecimen Acquisition Time of biospecimen collection First collection
5 years after first collection
10 years after first collection