Skip to main content

Preanalytical considerations in quantifying circulating miRNAs that predict end-stage kidney disease in diabetes.

Satake E, Krolewski B, Kobayashi H, Md Dom ZI, Ricca J, Wilson JM, Hoon DS, Duffin KL, Pezzolesi MG, Krolewski AS

Publication: JCI Insight (2024) • Vol 9PMID: 38912578View on PubMed

Purpose

This paper compared the microRNA (miRNA, miR) profile obtained by next-generation sequencing (NGS) of whole plasma using the HTG EdgeSeq platform with the profile obtained by sequencing extracted RNA using the HTG EdgeSeq, SeqMatic, or LC Sciences platforms or by real-time RT-PCR. Additionally, the authors investigated if plasma miRNA concentrations are correlated (i) with plasma storage duration, (ii) between specimens freeze-thawed once versus three times, (iii) between collection timepoints in patients who did or did not progress to end-stage kidney disease (ESKD), and (iv) with patient age, body mass index (BMI), or Hemoglobin A1c (HbA1c) levels.

Conclusion

Using the HTG EdgeSeq platform, 2002 of 2083 miRNAs were detected directly in whole plasma, but only 930 miRNAs were detected after RNA extraction using the same HTG EdgeSeq platform, and 262 and 439 miRNAs were detected when RNA was sequenced with the SeqMatic and LC Sciences platforms, respectively. Importantly, detected miRNAs were found at both higher and lower levels in extracted RNA compared to direct detection in whole plasma. 1442 miRNAs found in whole plasma were included in the real-time RT-PCR assay, of which 379 were detected in extracted RNA, but once again expression levels differed between the sample types/platforms.  While all eight high-risk ESKD miRNAs were detected in extracted RNA using the HTG EdgeSeq platform, levels were significantly lower and not correlated with those in plasma, and the majority of the eight miRNAs were not detectable in extracted RNA using the other sequencing platforms. All 9 protective ESKD miRNAs were detected in extracted RNA using the HTG EdgeSeq platform, 7 were detected using either the LC Sciences RNA Seq platform or real-time RT-PCR, and 6 were detected using the SeqMatic RNA Seq assay. Importantly, levels of some miRNAs were correlated between plasma and extracted RNA, but the authors state the strength of the correlations was highly variable. 
While significant correlations were observed between levels of some miRNAs and the duration of frozen storage, the majority of miRNAs were not significantly correlated with plasma storage duration, and levels of miRNAs were strongly correlated between specimens that were freeze-thaw cycled once and three times (r=0.882). Importantly, none of the 17 ESKD-associated miRNAs were significantly affected by storage duration or freeze-thaw cycling. miRNA levels did not differ significantly between the baseline and follow-up visits in plasma from the 52 patients with type I diabetes who did not progress to ESKD, but significant increases in five of the eight ESKD-associated miRNAs and decreases in four of the nine ESKD protective miRNAs were observed among the 44 patients who did progress to ESKD; no other miRNAs were significantly increased.  The concentration of the 2002 detected miRNAs or the 17 ESKD-associated miRNAs in whole plasma were not associated with patient age, BMI, or HbA1c nor did they differ between specimens from men and women. 

Studies (2)

Purpose

This study compared the miRNA profile obtained by NGS of whole plasma using the HTG EdgeSeq platform with the profile obtained by sequencing extracted RNA using the HTG EdgeSeq, SeqMatic, or LC Sciences platforms or by real-time RT-PCR. This study used plasma and previously extracted RNA collected from eight patients (7 men and 1 woman) with type I diabetes and extracted RNA from plasma pools that included specimens from 40 type 1 diabetes patients.  RNA extraction was performed using the miRNeasy Serum/Plasma Kit. Levels of 2083 miRNAs were quantified in matched plasma and extracted RNA specimens from eight patients using the HTG EdgeSeq miRNA sequence platform. miRNAs were also quantified by sequencing the extracted RNA from those same eight patients by SeqMatics using SeqMatic’s TailorMix miRNA Sample Preparation Kit V2 Kit and a HiSeq 2000 DNA sequencer and by LC Sciences using the TruSeq Small RNA Sample Prep Kit and a HiSeq 2500 sequencer.  Levels of 1805 miRNAs were quantified using real-time RT-PCR and the miRNome miScript miRNA PCR Array and the RNA extracted from the plasma pools. 

Summary of Findings

Using the HTG EdgeSeq platform, 2002 of 2083 miRNAs were detected directly in whole plasma; however, only 46% (930) of those miRNAs were detected when isolated RNA was sequenced using the HTG EdgeSeq platform, with 53 miRNAs (5.7%) expressed at lower levels and 46 (4.9%) expressed at higher levels in extracted RNA compared to levels in whole plasma. Of the 2002 miRNAs detected by HTG EdgeSeq in whole plasma, only 13% (262) and 22% (439) were detected in extracted RNA that was sequenced using the SeqMatic and LC Sciences RNA-Seq platforms, respectively, with detected miRNAs found at both higher and lower levels in isolated RNA compared to whole plasma. A total of 1442 miRNAs found in whole plasma were included in the real-time RT-PCR assay, of which 379 (27%) were detected in RNA extracted from plasma, but once again, expression levels differed between the sample types/platforms.  All eight high-risk ESKD miRNAs were detected in extracted RNA using the HTG EdgeSeq platform, but levels were significantly lower and not correlated with those in plasma. Only two of the eight high-risk ESKD miRNAs were detected in extracted RNA using either the LC Sciences RNA Seq or the real-time RT-PCR assay, and only one of the eight was detected in extracted RNA using the SeqMatic RNA Seq assay. All nine protective ESKD miRNAs were detected in extracted RNA by HTG EdgeSeq, seven were detected using either the LC Sciences RNA Seq or the real-time RT-PCR, and six were detected using the SeqMatic RNA Seq assay. Importantly, levels of some miRNAs were correlated between plasma and extracted RNA, but the authors state the strength of the correlations was highly variable.  

  • Biospecimens

    FluidPlasma
  • Preservation Types

    Frozen
  • Diagnoses

    Diabetes Type 1
  • Platforms

    RNANext generation sequencing
    RNAReal-time qRT-PCR
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Analyte Extraction and PurificationAnalyte isolation method
    No RNA extraction performed
    RNA extraction performed
    Next generation sequencing SpecificTechnology platform
    HTG EdgeSeq platform
    LC Sciences RNA Seq
    Real-time RT-PCR
    SeqMatic RNA seq