The purpose of this paper was to assess how performing extraction, ligation, and hybridization steps at different sites (using the same protocol) affected miRNA levels of plasma and cystic fluid specimens; potential differences were examined via correlation analysis and hierarchical clustering.
A pilot study to troubleshoot quality control metrics when assessing circulating miRNA expression data reproducibility across study sites.
Permuth JB, Mesa T, Williams SL, Cardentey Y, Zhang D, Pawlak EA, Li J, Cameron ME, Ali KN, Jeong D, Yoder SJ, Chen DT, Trevino JG, Merchant N, Malafa M
Purpose
Conclusion
Levels of housekeeping, spike-in controls and the 55 miRNAs that were expressed in >80% of specimens were strongly to very strongly correlated between replicates when extraction, ligation and/or hybridization were performed at different sites. Between-site correlations for housekeeping genes were higher for plasma than cystic fluid specimens, but correlations for the spike-in controls and the 55 miRNAs expressed in >80% of specimens were higher in cystic fluid specimens than plasma specimens. The 55 miRNAs detected in >80% of specimens, showed similar expression patterns among processing sites and specimens clustered by diagnosis and not site.
Studies (1)
Purpose
The purpose of this study was to assess how performing extraction, ligation, and hybridization steps at different sites (using the same protocol) affected miRNA levels of plasma and cystic fluid specimens; potential differences were examined via correlation analysis and hierarchical clustering. EDTA blood was collected from 3 patients diagnosed with Intraductal Papillary Mucinous Neoplasm (IPMN), 5 patients with pseudocysts occurring in the background of pancreatitis, and 4 patients without pancreatic conditions. Blood tubes were inverted three times and within 2 h of blood collection plasma was separated by centrifugation at 3,600 rpm for 8 minutes. Plasma was aliquoted and stored at -80°C. Cystic fluid was collected from the same patients by endoscopic ultrasound guided fine-needle aspiration (EUS-FNA) or by aspiration during surgical resection and was aliquoted and frozen at -80°C. Plasma and cystic fluid aliquots, as well as high quality commercially obtained RNA, were thawed on wet ice and assessed for hemolysis visually and using a spectrophotometer (absorbance at 414, 541, and 576 nm, with a threshold of >0.2). Specimens were spiked with osa-miR-414, cel-mi-248, and athmiR-159a before RNA isolation using a modification of the Plasma/Serum Circulating and Exosomal RNA Purification Mini Kit at each of three sites. Isolated RNA was concentrated with Amicon Ultra 0.5 mL MWCO 3 kDa centrifugal filters. RNA yield was quantified using a NanoDrop spectrophotometer and integrity was assessed using an Agilent Bioanalyzer RNA Pico Chip. miRNA levels were quantified using the nCounterTM Human v3 miRNA Expression Assay CodeSet, which included 800 human miRNAs, 5 housekeeping RNAs (ACTB, GAPDH, B2M, RPL19 and RPLP0) and the spiked-in RNAs, at each of the three sites.
Summary of Findings
Levels of positive control RNAs were very strongly correlated between replicates when all steps were performed at one site (ρ=0.93-0.97). Between-site correlations for housekeeping genes were strongest among commercial specimens, followed by plasma and cystic fluid specimens; replicate correlations were strong and comparable among the three sites (ρ=0.8-0.95). Quantified levels of spike-in controls were highest in cystic fluid specimens, followed by plasma specimens. Spike-in control miRNA levels were strongly correlated between replicates at each of the three sites (ρ=0.9-1.0), with the strongest correlations observed for cystic fluid specimens. Finally, levels of the 800 miRNAs were strongly correlated among the three sites (median=0.9) with the highest correlation observed for cystic fluid, which the authors attributed to slightly higher yields. The 55 miRNAs detected in >80% of specimens, showed similar expression patterns among the three sites (median correlations= 0.9-1.0) and specimens clustered by diagnosis.
Biospecimens
FluidPlasmaFluidBloodPreservation Types
FrozenDiagnoses
NormalOther diagnosesNeoplasticOtherPlatforms
RNADNA microarrayRNAAutomated electrophoresis/BioanalyzerRNASpectrophotometryPre-analytical Factors
Classification Pre-Analytical Factor Value(s) Preaquisition Diagnosis/ patient condition IPMNNormal pancreasPseudocysts occurring in the background of pancreatitisBiospecimen Acquisition Biospecimen location Pancreatic cystPlasmaAnalyte Extraction and Purification Analyte isolation method At 3 different sites