Fig 1.
Summary of experimental design for Figs 2, 3 and 4.
(A) Left ventricular biopsies were taken from three pigs undergoing cardio-pulmonary bypass (CPB). 7 biopsies were taken per pig. RNA was isolated from the biopsies using four commercially available kits; RNAqueous micro, mirVana, miRCURY tissue, miRNeasy micro. Three kits (RNAqueous micro, mirVana, miRNeasy micro) were tested with and without an additional proteinase K digestion step, giving 7 protocols in total. The miRCURY tissue protocol includes a proteinase K digestion, therefore this kit was not tested without this. RNA samples were characterised by spectrophotometry and capillary electrophoresis. (B) Blood samples were taken from three pigs undergoing CPB and processed to plasma. RNA was isolated from the plasma using four commercially-available kits; miRNeasy serum/plasma, mirVana, miRCURY biofluids, Norgen Biotek plasma/serum. Each kit was tested with and without glycogen as a co-precipitant, giving 8 protocols in total. RNA samples were evaluated by RT-qPCR for the exogenous spike-in cel-miR-39-3p, and endogenous miRNAs miR-16-5p, miR-21-5p, miR-92a-3p.
Fig 2.
Spectrophotometry analysis of pig LV biopsy RNA.
RNA extracted according to each manufacturer’s protocol, with the addition of a proteinase K digestion in half the samples. n = 3 biopsies/experimental condition, mean±SEM. N.B. Exiqon miRCURY tissue protocol requires proteinase K, therefore no–proteinase K condition was done with this kit. (A) RNA yield, assessed as ng RNA recovered per mg input tissue, was not significantly different between the protocols (Kruskal-Wallis test with Dunn correction), and additional proteinase K digestion in the RNAqueous micro, mirVana and miRNeasy micro protocols did not significantly improve RNA yield (Kruskal-Wallis test with Dunn correction). (B) 260/280 and 260/230 ratios provide an assessment of the purity of the RNA. Ratio <2.0 indicates possible contamination with protein, phenol or salts. All protocols produced an acceptable 260/280 ratio. RNA extracted using the mirVana protocol had low 260/230 ratios, indicating probable phenol or guanidine carryover. Numbers in brackets indicate the number of samples in each group.
Fig 3.
Capillary electrophoresis analysis of pig LV biopsy RNA.
(A) Bioanalyzer nano chip traces of all tested samples. Inspection of electropherograms provides additional information to assess RNA quality and protocol performance. (B) Representative traces from each protocol with respective RIN values. Numbers 1–4 correspond to numbered boxes in A. Note that trace 3 has a high RIN despite a large smear in the gel. (C) Summary of RIN values of all the samples tested (mean±SEM). The RNAqueous protocol had undetectable RINs. The mirVana and miRCURY tissue protocols gave consistently high RIN values, while despite high average RIN, the samples from the miRNeasy micro protocol are more variable. Numbers in brackets indicate the number of samples in each group. (D) Bioanalyzer small RNA chip traces for each kit. Although the RIN of samples marked * were similar (8.6–8.7), the spectrum of small RNAs recovered is highly heterogeneous.
Fig 4.
Evaluating plasma RNA yield using RT-qPCR for miRNAs.
As biofluid RNA is below the detection limit of standard quantification tools, all RT-qPCRs were performed using 1.5 μL RNA extracted from 200 μL plasma. Lines indicate paired samples i.e. aliquots of plasma from the same blood draw. (A) RT-qPCR for the spike-in cel-miR-39-3p using the four kits with and without glycogen. (B-D) RT-qPCR for three endogenous miRNAs isolated using the four kits with and without glycogen. (E) The miRNeasy serum/plasma protocol recommends elution in 14 μL, compared with 100 μL for the mirVana protocol. To control for effects of this, RNA extractions were performed with both protocols, eluting in 30 μL or 50 μL. RT-qPCR for cel-miR-39-3p, hsa-miR-16-5p and hsa-miR-21-5p were performed to evaluate the different elution volumes.
Fig 5.
Decision tree for optimising appropriate protocols for RNA extraction from challenging samples.