MicroRNAs are small noncoding molecules that regulate gene expression following transcription and are involved in the control of various cellular mechanisms, including tumorigenesis and the development of various types of cancers.1,2 A wide variety of miRNAs have been found in body fluids, including blood, CSF, saliva, amniotic fluid, breast milk, and urine,3–7 and differential expression of specific miRNAs, compared with expression in healthy individuals or control patients with other diseases, has been reported for patients with various cancers.5,8,9 Therefore, circulating miRNAs have been proposed as potential biomarkers for early detection of cancer as well as for prognostication and disease monitoring. Most miRNAs in blood are bound to plasma proteins or contained within EVs,10,11 which are small membrane vesicles that are secreted by various types of cells and that transport their contents to local or distant recipient cells.11,12 It has been suggested that EVs contain miRNA profiles that may be reflective of their cell of origin.13 Additionally, the slower degradation of miRNAs within EVs, compared with degradation of miRNAs circulating in plasma, suggests that assessment of EVs for miRNAs may be advantageous clinically.14,15 To the authors’ knowledge, however, direct comparisons of miRNAs detected in plasma versus EVs in dogs have not been reported.
Glioma is one of the most common primary CNS tumors in dogs, representing 20% to 40% of primary intracranial tumors in dogs,16,17 and certain brachycephalic breeds, including Boxers, Bulldogs, and Boston Terriers, have been reported to have an increased risk of developing gliomas.17,18 In veterinary medicine, the diagnosis of intracranial tumors has improved as a result of the greater availability of advanced imaging techniques such as MRI and CT.19 Still, methods for differentiating various types of intracranial tumors prior to surgery would be helpful when advising owners.
Altered amounts of certain miRNAs have been identified in human glioma tumor tissue and cell lines20–23 and have been associated with prognostic factors such as overall survival rate and progressionfree survival time.20–22 In addition, certain characteristic miRNAs, including miR-15b, miR-21, miR-155, and miR-342-3p, have been identified in human patients with glioma and proposed as noninvasive biomarkers.22,24–26 Circulating miRNAs have been described as possible biomarkers for various neoplastic conditions and myxomatous mitral valve disease in dogs.27–29 However, the authors are not aware of any data on circulating miRNAs in dogs with glioma.
The purpose of the study reported here was to measure expression of miRNAs in plasma and in EVs derived from plasma of dogs with gliomas and dogs with various brain diseases other than glioma. We hypothesized that miRNA profiles for dogs with glioma would differ from the profiles for dogs with other brain diseases and that miRNA profiles in EVs would be more useful than plasma miRNA profiles for differentiating glioma from other brain diseases.
Materials and Methods
Plasma samples were obtained from a convenience sample of dogs with brain disease that were examined at the Gifu University Veterinary Teaching Hospital between April 2016 and March 2018. Owners of all dogs included in the study provided informed consent. For all dogs, the diagnosis of brain disease was made on the basis of history, clinical signs, and results of MRI and CSF analysis. For dogs for which biopsy or necropsy samples were not available, a presumptive diagnosis was made on the basis of signalment, diagnostic imaging findings, results of CSF analysis, and response to treatment. For dogs for which brain tissue was available, a histologic diagnosis was made on the basis of the 2007 World Health Organization classification scheme.
Sample collection and processing
For dogs included in the study, a 2-mL blood sample was collected from a jugular or cephalic vein with a 23-gauge needle into an evacuated tube containing EDTA.a Samples without hemolysis were used for routine examination, and the remainder of each sample was centrifuged within 3 hours at 800 × g for 20 minutes. Plasma was harvested and stored at −80°C until analyzed.
Extracellular vesicles were isolated from plasma samples by means of ultracentrifugation, as described previously.30 Briefly, 0.3 mL of plasma was mixed %%%with 1.2 mL of PBS solution. The mixture was placed in a centrifuge with a swinging bucket rotorb and ultracentrifuged at 100,000 × g at 4°C for 1 hour. The resulting pellet was resuspended in 0.3 mL of nuclease-free water.
Total RNA extraction and quantification of miRNA expression
The RNA was extracted from each plasma (0.3 mL) and plasma EV (0.3 mL) sample with a commercially available kitc used in accordance with the manufacture's instruction. Five microliters of total RNA was reverse transcribed with a commercially available kitd in accordance with the manufacturer's instructions. With the use of custom-designed primers,e expression of candidate reference miRNAs (let-7a,f miR-16,g miR-26a,h and miR-103i) and candidate target miRNAs (miR-15b,j miR-21,k miR-155,l and miR-342-3pm) was assessed in duplicate by means of real-time qRT PCR assays. The miRNAs let-7a, miR-16, miR-26a, and miR-103 were selected because they have been commonly used as endogenous control genes for circulating EVs in humans.31–33
Thermal cyclern conditions consisted of 1 cycle at 95°C for 10 minutes, followed by 60 cycles at 95°C for 5 seconds and 60°C for 30 seconds. Negative controls (ie, without template) were included in several runs to confirm that specificity was high. For each reaction, the Ct value was calculated as described previously30 with a threshold set manually at 10 fluorescent units, which fell within the exponential phase of the reaction for all assays. For all samples, a target miRNA was regarded as undetectable if the Ct was > 40. For each miRNA, all samples were examined in the same run. Relative expression of each target miRNA was calculated with the 2−ΔΔCt method. Expression of a target miRNA was calculated as the fold difference in expression relative to the mean expression of the reference gene.
Expression stability of candidate reference genes was calculated with 3 widely used software programs (geNorm,o NormFinder,p and BestKeeperq), as described.30 In summary, the geNorm software calculated the pairwise variation in expression of candidate reference genes as the SD of the logarithmically transformed expression (Ct) ratios. The candidate gene stability measure, M, was then calculated as the mean of these pairwise variations, with a value < 1.5 interpreted as indicating a stably expressed gene. The NormFinder software calculated stability by means of the combined estimates of the intra- and intersample expression of each candidate gene.34 The BestKeeper software generated an index based on calculation of the geometric mean of all the candidate gene Ct values and determined the stability of those genes on the basis of repeated pairwise correlation analysis.35 After statistics for individual candidate genes were computed, expression of the candidate reference genes was determined by inspecting the calculated variations in SD.
For each software program, reference genes were ordered from most stable (ie, with the lowest variation) to least stable (ie, with the highest variation).35 Then, for each gene, the geometric mean of the rank obtained with each of the 3 programs was calculated, and genes were reranked on the basis of those geometric mean values. The gene with the lowest geometric mean value across all samples was considered the most stable reference gene.33
Statistical analysis
Relative expressions of the target miRNAs were imported into a statistical software programr for analysis. Bivariate analyses were conducted with Mann-Whitney tests. Values of P < 0.05 were considered significant.
Results
Plasma samples were obtained from 11 dogs with a histologic or presumptive diagnosis of glioma and 19 dogs with various other brain disorders (idiopathic epilepsy, n = 5; meningoencephalitis of unknown origin, 4; meningioma, 4; cerebral infarction, 4; choroid plexus papilloma, 1; and cerebral hemorrhage, 1). Median age of the dogs with a glioma was 9.7 years (range, 6 to 11 years), and median body weight was 10.4 kg (range, 5.6 to 16.4 kg). There were 5 females (3 neutered and 2 sexually intact) and 6 males (3 neutered and 3 sexually intact). Eight of the dogs were French Bulldogs, and 3 were Boston Terriers. The diagnosis was confirmed by means of histologic examination in 8 of the dogs; 4 of these dogs had an astrocytoma, 3 had an oligodendroglioma, and 1 had a glioblastoma.
For the 19 dogs with brain disorders other than glioma, median age was 8.5 years (range, 10 months to 14 years) and median body weight was 5.0 kg (range, 2.2 to 27.8 kg). There were 9 females (5 neutered and 4 sexually intact) and 10 males (6 neutered and 4 sexually intact). There were 5 Toy Poodles, 3 Chihuahuas, 3 Miniature Dachshunds, 2 Pomeranians, and 1 each of the following breeds: Labrador Retriever, Border Collie, Papillon, Borzoi, Welsh Terrier, and French Bulldog. In 6 of these dogs, the diagnosis was made on the basis of histologic examination of surgical biopsy specimens; 3 of these dogs had a meningioma, 1 had meningoencephalitis, 1 had a choroid plexus papilloma, and 1 had cerebral hemorrhage.
The Ct values for the 4 candidate reference genes (let-7a, miR-16, miR-26a, and miR-103) in plasma and in EVs derived from plasma were tabulated (Supplementary Tables S1 and S2, available at avmajournals.avma.org/doi/suppl/10.2460/ajvr.81.4.355). For plasma samples, the geNorm programo identified let-7a and miR-103 as the most stable, the NormFinder programp identified miR-16 as the most stable, and the BestKeeper programq identified miR-16 and miR-26a as the most stable (Table 1). For samples of EVs, the geNorm programo identified let-7a and miR-26a as the most stable, the NormFinder programp identified let-7a as the most stable, and the BestKeeperq program identified miR-16 as the most stable. When geometric mean ranks were calculated, the candidate reference genes with the lowest rank (ie, the most stably expressed) were miR-16 for plasma samples and let-7a for samples of EVs, and these 2 genes were used as the reference genes for determining expression of the candidate target genes in plasma and EVs, respectively.
Expression stability (M) and stability rank calculated with 3 software programs for 4 candidate reference genes in plasma and in EVs derived from plasma for 11 dogs with glioma and l9 control dogs with other brain diseases.
| geNorm | NormFinder | BestKeeper | ||||||
|---|---|---|---|---|---|---|---|---|
| miRNA | M | Rank | M | Rank | M | Rank | Mean* | Overall rank |
| Plasma | ||||||||
| let-7a | 0.059 | 1 | 0.758 | 3 | 2.020 | 3 | 2.08 | 3 |
| miR-16 | 0.078 | 4 | 0.618 | 1 | 1.910 | 1 | 1.59 | 1 |
| miR-26a | 0.066 | 3 | 0.719 | 2 | 1.910 | 1 | 1.82 | 2 |
| miR-103 | 0.059 | 1 | l.838 | 4 | 2.220 | 4 | 2.52 | 4 |
| EVs | ||||||||
| let-7a | 0.031 | 1 | 0.359 | 1 | 1.440 | 3 | 1.44 | 1 |
| miR-16 | 0.057 | 4 | 0.731 | 4 | 1.220 | 1 | 2.52 | 4 |
| miR-26a | 0.031 | 1 | 0.674 | 3 | 1.680 | 4 | 2.29 | 2 |
| miR-103 | 0.047 | 3 | 0.378 | 2 | 1.240 | 2 | 2.29 | 2 |
Geometric mean value of ranks assigned by the 3 software programs.
For the candidate target genes, median relative expression of miR-15b was significantly higher in both plasma (P = 0.010) and EVs (P = 0.002; Figure 1) from dogs with glioma than in samples from dogs with other brain diseases. Median relative expression of miR-342-3p was also significantly (P = 0.031) higher in EVs from dogs with glioma than in EVs from dogs with other brain diseases. Expression of the other candidate target genes did not differ significantly between groups.
Box-and-whisker plots of relative expression (REL) of 4 miRNAs (miR-15b, miR-21, miR-155, and mi-R-342-3p) in plasma and in EVs derived from plasma for 11 dogs with glioma and 19 control dogs with other brain diseases. Expression of the miRNAs was calculated relative to expression of miR-16 for plasma and relative to expression of let-7a for EVs. For each plot, the box represents the interquartile (25th to 75th percentile) range, the horizontal line within each box represents the median, and the whiskers represent the largest and smallest observed RELs that fell within the 75th and 25th percentiles plus and minus, respectively, 1.5 times the interquartile range; dots represent values for individual dogs. *†Values differed significantly (*P < 0.05; †P < 0.01) between groups.
Citation: American Journal of Veterinary Research 81, 4; 10.2460/ajvr.81.4.355
Discussion
In the present study, relative expression of 4 target miRNAs that have been reported to be useful for identifying glioma in human patients22,24–26 was assessed, and expression of miR-15b in both plasma and EVs derived from plasma was significantly higher in dogs with glioma than in dogs with other brain diseases. This finding was similar to findings for human patients with glioma25 and was consistent with the observation that miR-15b has a regulatory role in cellular proliferation, invasion, apoptosis, and angiogenesis in human and rat glioma cells.36,37 In addition, miR-15b expression is reportedly significantly upregulated in human glioma tissue, compared with expression in normal brain tissue.38
We also found in the present study that expression of miR-342-3p in EVs derived from plasma was significantly higher in dogs with glioma than in dogs with other brain diseases. This was contradictory to a previous finding that human glioma cells inhibit expression of miR-342-3p,23 with expression of miR-342-3p significantly downregulated in human glioma tissue, compared with expression in normal brain tissue.23 Further studies are needed to evaluate expression of miR-15b and miR-342-3p in canine glioma tissues.
Expression of the other target miRNAs (miR-21 and miR-155) evaluated in the present study did not differ between dogs with glioma and dogs with other brain diseases. In contrast, expression of these miRNAs has been shown to be altered in human patients with glioma or in human glioma tissue,23–26 and these miRNAs have been suggested to have potential value as circulating biomarkers of disease.21,22 The lack of difference in expression of these 2 miRNAs %%%between groups in the present study may have been a result of species differences in miRNA expression39 or a consequence of limitations of the present study, including the heterogenous tumor types and the limited overall sample size.
When analyzing expression of target miRNAs, assessing the stability of reference genes is highly recommended because variations in expression stability have been documented for different sample populations. For example, miR-16 has been shown to be stably expressed in EVs derived from serum samples obtained from breast cancer patients, but it was found to have unstable expression in EVs derived from serum samples obtained from human patients with hepatitis B or hepatocellular carcinoma.31,40 We previously compared expression of 4 reference candidate miRNAs in EVs derived from canine plasma and found that miR-103 and let-7a were stable reference genes,30 and a previous study41 that analyzed the stability of miRNAs in plasma from dogs with malignant endothelial proliferative diseases found that miR-16 was more stably expressed than RNU6B, RNU19, RNU48, and miR-1228. These findings were similar to findings for the present study. However, the fact that we identified different reference genes for plasma versus EVs emphasized the need for individual optimization of sample sets.
The use of 2 or more reference genes has been recommended for accurate and robust normalization of miRNA expression data, especially when measuring subtle differences in expression.42 Expression of miR-16 was downregulated in serum from human patients with esophageal squamous cell carcinoma, gastric cardia adenocarcinoma, or gastric non-cardia adenocarcinoma, compared with expression in control patients without cancer, but was upregulated in serum from patients with pancreatic ductal adenocarcinoma.43 Hemolysis can alter measured expression of various miRNAs, including miR-16, miR-92a, miR-451, and miR-486-5p, in plasma,44 and the ratio of miR-451a to miR-23a-3p expression could be useful for detecting low levels of hemolysis.45 Additional research is needed to investigate the usefulness of other reference genes.
There have been several comparisons of miRNA expression in plasma and EVs from healthy humans and humans with various diseases. Some studies46,47 used the same reference gene to calculate relative expression of miRNAs in plasma and EVs. In light of this, an interesting finding in the present study was the differing stabilities of candidate reference genes in plasma and EVs, even though these samples were from the same dogs. Other authors48,49 have also suggested that miRNA profiles in plasma might differ from those in EVs.
Limitations of the present study included the small sample size, lack of a histologic diagnosis for some dogs, use of only 4 candidate target miRNAs, and lack of information about miRNA expression in glioma and control tissues. The method used to isolate EVs differed from that used in previous studies,47,50 and it is known that miRNA expression in EVs may be influenced by the method of isolation.30,51 For our qRT PCR assays, we used 60 amplification cycles; however, 40 cycles are recommended to prevent nonspecific amplification. Direct comparison of our results with results of previous studies may therefore be limited; however, conclusions based on comparisons between study groups should be valid. Additional studies are needed to further compare miRNA profiles in plasma and EVs to determine the optimal sample source.
Finally, our data suggested that miR-15b and miR-342-3p have potential as noninvasive biomarkers for differentiating glioma from other intracranial diseases in dogs. However, more extensive analysis of expression in specific glioma subtypes and grades, compared with expression in more defined control populations, will be necessary to assess their potential as clinically relevant diagnostic biomarkers for glioma in dogs.
Acknowledgments
Funded in part by a Grant-in-Aid for Young Scientists (grant No. 17K15378) from the Japan Society for the Promotion of Sciences and a research grant from Gifu University.
The authors declare that there were no conflicts of interest.
The authors thank Yukina Kuwahara for technical assistance.
ABBREVIATIONS
| Ct | Threshold cycle |
| EV | Extracellular vesicle |
| miRNA | MicroRNA |
| qRT | Quantitative reverse transcription |
Footnotes
Venoject II, VP-NA050K, Terumo Corp, Tokyo, Japan.
S52ST rotor, Hitachi Inc, Tokyo, Japan.
NucleoSpin miRNA plasma kit, Machery-Nagel GmbH, Düren, Germany.
TaqMan miRNA reverse transcription kit, ThermoFisher Scientific, Waltham, Mass.
TaqMan small RNA assays, ThermoFisher Scientific, Waltham, Mass.
cfa-let-7a, assay 000377, ThermoFisher Scientific, Waltham, Mass.
cfa-miR-16, assay 000391, ThermoFisher Scientific, Waltham, Mass.
cfa-miR-26a, assay 000405, ThermoFisher Scientific, Waltham, Mass.
cfa-miR-103, assay 000439, ThermoFisher Scientific, Waltham, Mass.
cfa-miR-15b, assay 000390, ThermoFisher Scientific, Waltham, Mass.
cfa-miR-21, assay 000397, ThermoFisher Scientific, Waltham, Mass.
cfa-miR-155, assay 002623, ThermoFisher Scientific, Waltham, Mass.
cfa-miR-342-3p, assay 002260, ThermoFisher Scientific, Waltham, Mass.
PCR thermal cycler Dice TP800, Takara Bio Inc, Shiga, Japan.
geNorm, version 3.5, Center for Medical Genetics, Ghent, Belgium. Available at: genorm.cmgg.be. Accessed Apr 2, 2018.
NormFinder Excel add-in, version 0.953, Aarhus University Hospital, Aarhus, Denmark. Available at: moma.dk/normfinder-software. Accessed Jun 22, 2018.
BestKeeper, version 1. Available at: www.gene-quantification.de/bestkeeper.html. Accessed Apr 2, 2018.
JMP, version 13.2, SAS Institute Inc, Cary, NC.
References
1. Bartel DP. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 2004;116:281–297.
2. He L, Hannon GJ. MicroRNAs: small RNAs with a big role in gene regulation. Nat Rev Genet 2004;5:522–531.
3. Hata T, Murakami K, Nakatani H, et al. Isolation of bovine milk-derived microvesicles carrying mRNAs and microRNAs. Biochem Biophys Res Commun 2010;396:528–533.
4. Keller S, Ridinger J, Rupp AK, et al. Body fluid derived exosomes as a novel template for clinical diagnostics. J Transl Med 2011;9:86.
5. Mitchell PS, Parkin RK, Kroh EM, et al. Circulating micro RNAs as stable blood-based markers for cancer detection. Proc Natl Acad Sci U S A 2008;105:10513–10518.
6. Baraniskin A, Kuhnhenn J, Schlegel U, et al. Identification of microRNAs in the cerebrospinal fluid as biomarker for the diagnosis of glioma. Neuro Oncol 2012;14:29–33.
7. Rao P, Benito E, Fischer A. MicroRNAs as biomarkers for CNS disease. Front Mol Neurosci 2013;6:39.
8. Lawrie CH, Gal S, Dunlop HM, et al. Detection of elevated levels of tumour-associated microRNAs in serum of patients with diffuse large B-cell lymphoma. Br J Haematol 2008;141:672–675.
9. Chen X, Ba Y, Ma L, et al. Characterization of microRNAs in serum: a novel class of biomarkers for diagnosis of cancer and other diseases. Cell Res 2008;18:997–1006.
10. Arroyo JD, Chevillet JR, Kroh EM, et al. Argonaute2 complexes carry a population of circulating microRNAs independent of vesicles in human plasma. Proc Natl Acad Sci U S A 2011;108:5003–5008.
11. Valadi H, Ekström K, Bossios A, et al. Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells. Nat Cell Biol 2007;9:654–659.
12. Lotvall J, Valadi H. Cell to cell signalling via exosomes through esRNA. Cell Adh Migr 2007;1:156–158.
13. Mathivanan S, Simpson RJ. ExoCarta: a compendium of exosomal proteins and RNA. Proteomics 2009;9:4997–5000.
14. Endzelnš E, Melne V, Kalnica Z, et al. Diagnostic, prognostic and predictive value of cell-free miRNAs in prostate cancer: a systematic review. Mol Cancer 2016;15:41.
15. Valentino A, Reclusa P, Sirera R, et al. Exosomal microRNAs in liquid biopsies: future biomarkers for prostate cancer. Clin Transl Oncol 2017;19:651–657.
16. Schwartz M, Lamb CR, Brodbelt DC, et al. Canine intracranial neoplasia: clinical risk factors for development of epileptic seizures. J Small Anim Pract 2011;52:632–637.
17. Song RB, Vite CH, Bradley CW, et al. Postmortem evaluation of 435 cases of intracranial neoplasia in dogs and relationship of neoplasm with breed, age, and body weight. J Vet Intern Med 2013;27:1143–1152.
18. Truvé K, Dickinson P, Xiong A, et al. Utilizing the dog genome in the search for novel candidate genes involved in glioma development—genome wide association mapping followed by targeted massive parallel sequencing identifies a strongly associated locus. PLoS Genet 2016;12:e1006000.
19. LeCouteur RA, Withrow SJ. Tumours of the nervous system. In: Withrow SJ, Vail DM, eds. Withrow and MacEwen's small animal clinical oncology. 4th ed. Philadelphia: Saunders Elsevier, 2007;659-685.11
20. Sun G, Yan S, Shi L, et al. Decreased expression of miR-15b in human gliomas is associated with poor prognosis. Cancer Biother Radiopharm 2015;30:169–173.
21. Hermansen SK, Dahlrot RH, Nielsen BS, et al. MiR-21 expression in the tumor cell compartment holds unfavorable prognostic value in gliomas. J Neurooncol 2013;111:71–81.
22. Sun J, Shi H, Lai N, et al. Overexpression of microRNA-155 predicts poor prognosis in glioma patients. Med Oncol 2014;31:911.
23. Zhang W, Bi Y, Li J, et al. Long noncoding RNA FTX is upregulated in gliomas and promotes proliferation and invasion of glioma cells by negatively regulating miR-342-3p. Lab Invest 2017;97:447–457.
24. Wang Q, Li P, Li A, et al. Plasma specific miRNAs as predictive biomarkers for diagnosis and prognosis of glioma. J Exp Clin Cancer Res 2012;31:97.
25. D'Urso PI, D'Urso OF, Gianfreda CD, et al. miR-15b and miR-21 as circulating biomarkers for diagnosis of glioma. Curr Genomics 2015;16:304–311.
26. Roth P, Wischhusen J, Happold C, et al. A specific miRNA signature in the peripheral blood of glioblastoma patients. J Neurochem 2011;118:449–457.
27. Fujiwara-Igarashi A, Igarashi H, Mizutani N, et al. Expression profile of circulating serum microRNAs in dogs with lymphoma. Vet J 2015;205:317–321.
28. Heishima K, Ichikawa Y, Yoshida K, et al. Circulating microRNA-214 and −126 as potential biomarkers for canine neoplastic disease. Sci Rep 2017;7:2301.
29. Yang VK, Loughran KA, Meola DM, et al. Circulating exosome microRNA associated with heart failure secondary to myxomatous mitral valve disease in a naturally occurring canine model. J Extracell Vesicles 2017;6:1350088.
30. Narita M, Nishida H, Asahina R, et al. Identification of reference genes for microRNAs of extracellular vesicles isolated from plasma of healthy dogs by ultracentrifugation, precipitation, and affinity chromatography methods. Am J Vet Res 2019;80:449–454.
31. Li Y, Zhang L, Liu F, et al. Identification of endogenous controls for analyzing serum exosomal miRNA in patients with hepatitis B or hepatocellular carcinoma. Dis Markers 2015;2015:893594.
32. Gotanda K, Hirota T, Saito J, et al. Circulating intestine-derived exosomal miR-328 in plasma, a possible biomarker for estimating BCRP function in the human intestines. Sci Rep 2016;6:32299.
33. Li Y, Xiang GM, Liu LL, et al. Assessment of endogenous reference gene suitability for serum exosomal microRNA expression analysis in liver carcinoma resection studies. Mol Med Rep 2015;12:4683–4691.
34. Latham GJ. Normalization of micro RNA quantitative RT-PCR data in reduced scale experiment designs. Methods Mol Biol 2010;667:19–31.
35. Pfaffl MW, Tichopad A, Prgomet C, et al. Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper—Excel-based tool using pairwise correlations. Biotechnol Lett 2004;26:509–515.
36. Sun G, Shi L, Yan S, et al. MiR-15b targets cyclin D1 to regulate proliferation and apoptosis in glioma cells. Biomed Res Int 2014;2014:687826.
37. Zheng X, Chopp M, Lu Y, et al. MiR-15b and miR-152 reduce glioma cell invasion and angiogenesis via NRP-2 and MMP-3. Cancer Lett 2013;329:146–154.
38. Guan Y, Mizoguchi M, Yoshimoto K, et al. MiRNA-196 is upregulated in glioblastoma but not in anaplastic astrocytoma and has prognostic significance. Clin Cancer Res 2010;16:4289–4297.
39. Wagner S, Willenbrock S, Nolte I, et al. Comparison of noncoding RNAs in human and canine cancer. Front Genet 2013;4:46.
40. Eichelser C, Stückrath I, Müller V, et al. Increased serum levels of circulating exosomal microRNA-373 in receptor-negative breast cancer patients. Oncotarget 2014;5:9650–9663.
41. Heishima K, Mori T, Ichikawa Y, et al. MicroRNA-214 and microRNA-126 are potential biomarkers for malignant endothelial proliferative diseases. Int J Mol Sci 2015;16:25377–25391.
42. Hellemans J, Mortier G, De Paepe A, et al. qBase relative quantification framework and software for management and automated analysis of real-time quantitative PCR data. Genome Biol 2007;8:R19.
43. Huang Z, Chen W, Du Y, et al. Serum miR-16 as a potential biomarker for human cancer diagnosis: results from a large-scale population. J Cancer Res Clin Oncol 2019;145:787–796.
44. Pritchard CC, Kroh E, Wood B, et al. Blood cell origin of circulating microRNAs: a cautionary note for cancer biomarker studies. Cancer Prev Res (Phila) 2012;5:492–497.
45. Shah JS, Soon PS, Marsh DJ. Comparison of methodologies to detect low levels of hemolysis in serum for accurate assessment of serum microRNAs. PLoS One 2016;11:e0153200.
46. Tian F, Shen Y, Chen Z, et al. No significant difference between plasma miRNAs and plasma-derived exosomal miRNAs from healthy people. Biomed Res Int 2017;2017:1304816.
47. Uratani R, Toiyama Y, Kitajima T, et al. Diagnostic potential of cell-free and exosomal microRNAs in the identification of patients with high-risk colorectal adenomas. PLoS One 2016;11:e0160722.
48. Cheng L, Sharples RA, Scicluna BJ, et al. Exosomes provide a protective and enriched source of miRNA for biomarker profiling compared to intracellular and cell-free blood. J Extracell Vesicles 2014;3:10.3402/jev.v3.23743.
49. Zhao K, Liang G, Sun X, et al. Comparative miRNAome analysis revealed different miRNA expression profiles in bovine sera and exosomes. BMC Genomics 2016;17:630.
50. Endzelinš E, Berger A, Melne V, et al. Detection of circulating miRNAs: comparative analysis of extracellular vesicle-incorporated miRNAs and cell-free miRNAs in whole plasma of prostate cancer patients. BMC Cancer 2017;17:730.
51. Ding M, Wang C, Lu X, et al. Comparison of commercial exosome isolation kits for circulating exosomal microRNA profiling. Anal Bioanal Chem 2018;410:3805–3814.
