Kato
Pollet
ab,
Nathalie
Garnier
ab,
Sabine
Szunerits
b,
Annemieke
Madder
c,
Didier
Hober
a and
Ilka
Engelmann†
*a
aUniv. Lille, CHU Lille, Laboratoire de Virologie ULR3610, 59000 Lille, France. E-mail: ilka.engelmann@chu-montpellier.fr
bUniv. Lille, CNRS, Centrale Lille, Univ. Polytechnique Hauts-de-France, UMR 8520 – IEMN, 59000 Lille, France
cOrganic and Biomimetic Chemistry Research Group, Department of Organic and Macromolecular Chemistry, Ghent University, Krijgslaan 281-S4, 9000 Gent, Belgium
First published on 26th October 2022
MicroRNAs (miRNAs), small non-coding RNAs that regulate gene expression, have received increasing attention as potential biomarkers of different diseases, including viral infections. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and the disease it is causing, coronavirus disease (COVID-19), has affected health, society and life worldwide since its pandemic spread. Differential expression of miRNAs in COVID-19 patients compared to healthy controls and also between different severity grades of COVID-19 has been described in several recent studies. In this review, we discuss in detail studies that investigated miRNA expression in body fluids of COVID-19 patients. Several studies found a different miRNA expression profile in COVID-19 patients compared to controls but also in different severity grades of the disease. We compared the main findings of the studies in order to identify miRNAs that have been identified as differentially expressed by more than one study and could serve as diagnostic or prognostic biomarkers of COVID-19. Finally, we highlight the challenges and perspectives associated to the use of miRNAs as biomarkers of COVID-19.
SARS-CoV-2 causes COVID-19, a disease presenting symptoms including fever or chills, cough, shortness of breath, fatigue, nausea, muscle ache, headache, nausea or vomiting, diarrhoea, sore throat, congestion or running nose, loss of smell (anosmia) and loss of taste (ageusia). Severity ranges from asymptomatic to critical illness with multi-organ failure, respiratory failure or shock and possible fatal outcome.4
Severe COVID-19 is associated with severe pulmonary disease and can present acute respiratory distress syndromes (ARDS), cardiac impairment, liver dysfunction, acute kidney injury and coagulopathy.5 A hyperinflammatory state termed cytokine storm contributes further to severe COVID-19.6,7 An imbalanced host immune response characterized by reduced type I interferon production and high expression of other cytokines has been suggested to drive severe COVID-19.8,9 TNF-α, IL-6, IL-8 and IL-10 are significantly elevated in patients with severe COVID-19.6,7,9
miRNAs are small, noncoding RNAs that regulate gene expression. They bind to sequences that are often localized in the 3'untranslated region (3'UTR) of a messenger RNA (mRNA) and inhibit translation or induce degradation of this mRNA.10–12 Biogenesis of miRNAs includes transcription by RNA polymerase II to pri-miRNAs that are further processed by Drosha giving rise to pre-miRNAs. Export to the cytoplasm is followed by Dicer processing, which generates miRNA duplexes that are loaded into an Argonaute protein to form a miRNA induced silencing complex (miRISC).12,13 miRNAs are involved in virtually all physiological and pathological processes, including viral infections and the antiviral immune response.13 As discussed in some recent work, viral infections change the expression profile of cellular miRNAs (Fig. 1A).14–16 We recently reviewed data showing that acute and persistent enterovirus infection induced distinct miRNA expression changes in infected cells.15,17 Furthermore, different enterovirus types induced different miRNA expression changes and these were dependent on the cell type and the time point after infection.15 Concerning respiratory virus infections, Martinez-Espinoza and colleagues showed that infections with influenza A virus, human respiratory syncytial virus (HRSV), human metapneumovirus (HMPV), adenovirus and SARS-CoV-2 influenced the miRNA expression profile of different cell types in vitro but also in the blood of patients.16 Wu and colleagues reviewed data of HRSV and HMPV infections, demonstrating that these viruses induced distinct miRNA expression changes in different cell types in vitro and in clinical samples.14 These findings all point to miRNAs being important biomarkers for viral infections and being involved in the pathophysiology of viral infections.
In the recent literature, studies reporting miRNA expression in COVID-19 patients have been published (Tables 1 and 2). Some of these aimed to identify miRNAs as biomarkers, whereas others included miRNA expression analysis as part of Omics approaches. Several studies also investigated a small number of preselected miRNA candidates. The objective of this review was to summarize data on the expression of miRNAs in body fluids of COVID-19 patients. These miRNAs have the potential to represent candidate biomarkers. We systematically searched for studies that describe human miRNA expression in body fluids of COVID-19 patients and reviewed studies that specifically investigated miRNAs as biomarkers for the diagnosis of SARS-CoV-2 infection and those that described miRNAs associated with different clinical courses of COVID-19.
Studies with miRNA screening | ||||||||
---|---|---|---|---|---|---|---|---|
Main findings | Patient numbers | Specimen type | Pre-analytical procedures | Extraction methods | miRNA quantification methods | Normalization methods | Statistical methods | Ref. |
hsa-miR-17-5p, hsa-miR-142-5p down-regulated hsa-miR-15a-5p, hsa-miR-19a-3p, hsa-miR-19b-3p, hsa-miR-23a-3p, hsa-miR-92a-3p and hsa-miR-320a up-regulated in SARS-CoV-2-infected patients | 33 COVID-19 patients, 10 healthy controls | Plasma | Centrifugation within 2 hours | miRNeasy serum/plasma isolation kit (Qiagen) | miRCURY LNA miRNome qPCR panels/miRCURY LNA miRNA PCR assay; validation by RT-qPCR | miR-502-5p | Unpaired student's t-test/ROC; not mentioned whether correction for multiple testing was performed | 40 |
20 miRNAs up-regulated in COVID-19: hsa-miR-103a-3p, hsa-let-7f-5p, hsa-miR-423-5p, hsa-miR-320a-3p, hsa-miR-92a-3p, hsa-let-7a-5p, hsa-miR-148a-3p, hsa-miR-142-3p, hsa-miR-30a-5p, hsa-miR-320c, hsa-miR-320b, hsa-let-7e-5p, hsa-miR-197-3p, hsa-miR-576-5p, hsa-miR-1290, hsa-miR-195-5p, hsa-miR-483-5p, hsa-miR-193a-5p, hsa-miR-6721-5p, hsa-miR-206, hsa-miR-27a-5p, hsa-miR-2116-3p, hsa-miR-4742-3p, hsa-miR-3125, hsa-miR-31-5p; 30 miRNAs down-regulated in COVID-19: hsa-miR-1275, hsa-miR-3617-5p, hsa-miR-500b-3p, hsa-miR-3684, hsa-miR-627-5p, hsa-miR-651-5p, hsa-miR-18a-3p, hsa-miR-3115, hsa-miR-589-3p, hsa-miR-664b-3p, hsa-miR-548k, hsa-miR-769-3p, hsa-miR-1226-3p, hsa-miR-873-5p, hsa-miR-5189-3p, hsa-miR-3198, hsa-miR-4772-3p, hsa-miR-6772-3p, hsa-miR-145-3p, hsa-miR-3913-5p, hsa-miR-6503-3p, hsa-miR-210-3p, hsa-miR-766-3p, hsa-miR-3065-3p, hsa-miR-551b-3p, hsa-miR-28-5p, hsa-miR-491-5p, hsa-let-7i-3p, hsa-miR-4662a-5p, hsa-miR-150-5p; differentially expressed in COVID-19 patients with and without oxygen requirement: hsa-let-7e-5p, hsa-miR-651-5p, hsa-miR-766-3p, hsa-miR-4433b-5p | 10 COVID-19 patients, 10 healthy controls | Plasma | Lysis with Qiazol | miRNeasy Micro Kit (Qiagen) + glycogen | Next-generation sequencing: QIAseq miRNA Library Kit and QIAseq miRNA NGS 48 Index IL (Qiagen) | DESeq2 package in R (not detailed) | DESeq2 package in R (not detailed), FDR; machine learning analysis using the scikit-learn module in python; ROC/AUC | 41 |
hsa-miR-16-2-3p, hsa-miR-6501-5p, hsa-miR-618 upregulated in COVID-19; hsa-miR-183-5p, hsa-miR-627-5p, hsa-miR-144-3p downregulated in COVID-19 | 10 COVID-19 patients, 4 healthy controls | Whole blood | PAXgene blood RNA tubes (BD) | PAXgene Blood miRNA Kit (Qiagen) | High-throughput sequencing: model PE75 using Illumina Hiseq X Ten | Reads per million | Package DESeq2 of R software, not mentioned whether FDR was used | 42, 43 |
hsa-miR-19b-1-3p, hsa-miR-96, hsa-miR-19b-2-5p, hsa-miR-451a, hsa-miR-451b, hsa-miR-194-1-5p, hsa-miR-144, hsa-miR-486-2-3p, hsa-miR-15a, hsa-miR-29c-3p downregulated in COVID-19; hsa-miR-3609, hsa-miR-1244-1, hsa-miR-663a, hsa-miR-3916, hsa-miR-3687-2, hsa-miR-7846-3p, hsa-miR-5047, hsa-miR-3184-5p, hsa-miR-1248, hsa-miR-6891-3p upregulated in COVID-19 | 5 COVID-19 patients, 3 healthy controls | Plasma | Centrifugation | TRIzol Reagent | PALM-seq protocol; BGISEQ-500 sequencer; single-end 100 base reads | Transcripts per kilobase of exon model per million mapped reads | R package (DESeq2), p-value adjusted < 0.1 (no details on p-value adjustment given) | 44 |
hsa-miR-150-5p, hsa-miR-375, hsa-miR-122-5p, hsa-miR-494-3p down-regulated in COVID-19 patients; hsa-miR-3197, hsa-miR-4690-5p, hsa-miR-1915-3p, hsa-miR-3652 upregulated in COVID-19 patients | 12 COVID-19 patients with moderate–severe disease, 8 healthy controls | Plasma | Lysis buffer, proteinase-K with constant shaking at 50 °C for 3 h | Automated extraction-free chemistry of HTG EdgeSeq | NGS (HTG EdgeSeq miRNA Whole Transcriptome Assay, HTG Molecular Diagnostics, TruSeq Small RNA Prep kit, Illumina), validation by RT-qPCR | Not indicated | Student's t-test; (no details on p-value adjustment given) | 45 |
Downregulated in COVID-19: hsa-let-7i, hsa-let-7g, hsa-let-7f2, hsa-let-7f1, hsa-let-7d, hsa-miR-103a1, hsa-miR-126, hsa-miR-139, hsa-miR-15b, hsa-miR-16-1, hsa-miR-16-2, hsa-miR-181b1, hsa-miR-191, hsa-miR-21, hsa-miR-221, hsa-miR-222, hsa-miR-224, hsa-miR-23a, hsa-miR-23b, hsa-miR-25, hsa-miR-26a1, hsa-miR-30c1, hsa-miR-30c2, hsa-miR-30d, hsa-miR-320a, hsa-miR-92a1, hsa-miR-92a2, hsa-miR-93, hsa-miR-98, hsa-miR-151a, hsa-miR-340, hsa-miR-423, hsa-miR-425, hsa-miR-451a, hsa-miR-409, hsa-miR-454, hsa-miR-374b, hsa-miR-1244-1, hsa-miR-3609, hsa-let7a2, hsa-let7a1; upregulated in COVID-19: hsa-miR-663b, hsa-miR-3687-2, hsa-miR-3648-2 | 19 mild and 18 severe COVID-19 patients, 8 controls | Plasma | 1600g for 10 min at 4 °C and 16000g for 10 min at 4 °C | HiPure Liquid RNA Mini kit (Magen) | RNA library constructed using the PALM-seq protocol, DNBSEQ platform | Transcripts per kilobase of exon model per million mapped reads | DESeq, no correction for multiple testing mentioned | 46 |
Both NGS and RT-qPCR: upregulated in COVID-19 pneumonia compared to controls: hsa-miR-193a-5p, hsa-miR-197-3p; upregulated in COVID-19 ARDS compared to COVID-19 pneumonia: hsa-miR-206 | 18 healthy controls, 15 patients with COVID-19 pneumonia, 15 patients with COVID-19 ARDS | Extracellular vesicles isolated from serum | Centrifuged at 3400 × g for 10 min at room temperature | Extracellular vesicles isolation kit (miRCURY Exosome Isolation Kit, Qiagen) followed by RNA extraction | NEBNext Multiplex Small RNA Library Prep Set for Illumina (New England Biolabs); miRCURY LNA RT kit, miRCURY LNA SYBR Green PCR kit, LNA miRNA PCR Assays (Qiagen) | RNA seq: DESeq2; RT-qPCR: geometric mean of reference miRNAs (hsa-miR-30d-5p, hsa-miR-30e-5p, hsa-let-7i-5p, hsa-miR-148b-3p, hsa-miR-146b-5p, hsa-miR-425-5p, hsa-miR-24-3p, hsa-miR-125a-5p) | DESeq2, Benjamini–Hochberg FDR correction | 47 |
hsa-miR-146a-5p, hsa-miR-21-5p, hsa-miR-142-3p downregulated in COVID-19 compared to healthy controls; hsa-miR-3605-3p upregulated in COVID-19 compared to healthy controls; hsa-miR-15b-5p, hsa-miR-486-3p, hsa-miR-486-5p upregulated only in patients with severe COVID-19; hsa-miR-181a-2-3p, hsa-miR-31-5p, hsa-miR-99a-5p downregulated only in severe cases | 6 severe COVID-19 patients, 6 moderate COVID-19 patients, 4 healthy controls | Whole blood | Erythrocyte removal: erythrocyte lysis buffer | NucleoSpin miRNA kit (Macherey-Nagel) | NEBNext Multiplex Small RNA library Prep set for Illumina (NEB) | Number of mapped reads per kilobase per million reads | One-way ANOVA/Tukey's test (no details on p-value adjustment given) | 54 |
Down in severe COVID-19: hsa-miR-335-5p, hsa-miR-24-3p; down in COVID-19: hsa-miR-181a-5p | 6 moderate, 4 severe COVID-19 patients, 4 controls | Whole blood | Red blood cell lysis | TRIzol (Life Technologies) | NEXTflex Small RNA-Seq Kit v3 (Bio Scientific Corporation) | miRDeep2 reads per million mapped reads | Limma package in R, FDR | 56 |
Downregulation of hsa-miR-320a, hsa-miR-320b and hsa-miR-320c, hsa-miR-4747-3p, hsa-miR-4429, hsa-miR-6729-3p and hsa-miR-1908-5p in patients with severe respiratory failure; upregulation of hsa-miR-374a-3p, hsa-miR-15a-3p, hsa-miR-3688-5p, hsa-miR-4721 in COVID-19 patients with severe respiratory failure | 11 COVID-19 patients with moderate respiratory failure, 10 COVID-19 patients with severe respiratory failure, 8 healthy controls | Whole blood | PAXgene Blood RNA System for miRNA | PAXgene Blood miRNA Kit (Qiagen) | QIAseq miRNA Library Kit (Qiagen)/next-generation sequencing: MiSeq Reagent kit v3, the PhiX Sequencing Control v3 and the MiSeq Desktop Sequencer | Not indicated | One-way ANOVA withBonferroni corrected post-hoc analysis | 58 |
Discovery: downregulated in non survivors: hsa-miR-8061, hsa-miR-181c-3p, hsa-miR-410-3p, hsa-miR-101-5p, hsa-miR-339-3p, hsa-miR-28-5p, hsa-miR-17-3p; upregulated in non survivors: hsa-miR-1285-5p, hsa-miR-221-3p, hsa-miR-203a-3p, hsa-miR-100-5p; validation: upregulated in non-survivors: hsa-miR-320b, hsa-miR-483-5p | Discovery: 12 COVID-19 patients (6 survivors, 6 non survivors); validation: 116 COVID-19 patients (75 survivors, 41 non survivors) | Not clear whether plasma or serum | Not detailed | Not detailed | Discovery: TruSeq Small RNA Library Prep Kit v2 (Illumina); validation: TaqMan miRNA Reverse Transcription Kit (Life Technologies), miRNA assays (Applied Biosystems) | Discovery: DESeq2; validation: cel-miR-39-3p, hsa-miR-92a and hsa-miR-103 | Discovery: t-test with Benjamini–Hochberg correction; validation: unpaired Student t-test or Mann–Whitney U test, Kaplan–Meier curves, log-rank test for equality of survivor functions, Cox proportional hazards analysis | 59 |
hsa-miR-155 and hsa-miR-130a higher in mild cases compared to severe/critical disease and healthy controls | 16 severe/critical and 30 mild/moderate recovered COVID-19 patients, 24 healthy controls | Whole blood | — | TRIzol Reagent | Not clearly indicated for miRNAs (RT2 Profiler PCR array (Qiagen)?) | Not indicated | One-way ANOVA/two-tailed t-test | 60 |
hsa-miR-27a-3p, hsa-miR-27b-3p, hsa-miR-148a-3p, hsa-miR-199a-5p, hsa-miR-491-5p upregulated in ICU patients; hsa-miR-16-5p, hsa-miR-92a-3p, hsa-miR-150-5p, hsa-miR-451a; hsa-miR-486-5p downregulated in ICU patients; hsa-miR-16-5p, hsa-miR-92a-3p, hsa-miR-98-5p, hsa-miR-132-3p, hsa-miR-192-5p, hsa-miR-323a-3p downregulated in patients who did not survive the ICU stay | 43 COVID-19 patients (hospitalized bot not ICU), 36 COVID-19 patients (ICU) | Plasma | Centrifugation | miRNeasy serum/plasma Advanced Kit (Qiagen) | RT-qPCR: miRCURY LNA Universal RT microRNA PCR system (Qiagen) for 41 selected miRNAs | cel-miR-39-3p | Limma, ROC | 61 |
hsa-miR-4454, hsa-miR-7975 upregulated in severe COVID-19; hsa-miR-451a, hsa-miR-323a-3p, hsa-miR-188-5p, hsa-miR-432-5p, hsa-miR-433-3p, hsa-miR-371a-5p downregulated in severe COVID-19 | Screening: 20 mild, 21 moderate, 17 severe COVID-19 patients | Plasma | Centrifugation | Plasma/Serum RNA Purification Mini kit (Norgen) | Nanostring human v3 miRNA for 827 human miRNAs (Nanostring) | osa-miR-414 | Mann–Whitney U test, FDR correction using the Benjamini–Hochberg method | 62 |
hsa-miR-1246, hsa-miR-4532, hsa-miR-145-5p, hsa-miR-3651 upregulated in severe versus asymptomatic COVID-19; hsa-miR-3180-3p, hsa-let-7i-5p downregulated in severe versus asymptomatic COVID-19; hsa-miR-3609, hsa-miR-199a-5p, hsa-miR-139-5p, hsa-miR-145-5p, hsa-miR-3651, hsa-miR-1273h-3p upregulated in severe versus mild COVID-19; hsa-miR-4632-5p, hsa-miR-6861-5p, hsa-miR-6802-5p, hsa-miR-5196-5p, hsa-miR-92b-5p, hsa-miR-6805-5p, hsa-miR-98-5p, hsa-miR-3185, hsa-miR-572, hsa-miR-371b-5p, hsa-miR-3180, hsa-miR-8073, hsa-miR-4750-5p, hsa-miR-6075, hsa-let-7i-5p, hsa-miR-1231, hsa-miR-885-3p downregulated in severe versus mild COVID-19 | 9 severe COVID-19 patients, 10 mild COVID-19 patients, 10 asymptomatic COVID-19 patients | Blood | PAXgene Blood RNA Tubes (PreAnalytiX) | Blood miRNA Kit (Qiagen) | Affymetrix GeneChip miRNA 4.0 array using FlashTag Biotin RNA Labeling Kit (Genisphere) | Multiarray average (RMA) method | Limma, FDR | 63 |
Upregulated in severe COVID-19: hsa-miR-4516, hsa-miR-362-5p, hsa-miR-548k, hsa-miR-320a-3p, hsa-miR-320b, hsa-miR-320c, hsa-miR-320d, hsa-miR-185-5p, hsa-miR-629-5p, hsa-miR-1180-3p, hsa-miR-502-3p; downregulated in severe COVID-19: hsa-miR-454-3p, hsa-miR-625-3p, hsa-miR-30b-5p, hsa-miR-192-5p, hsa-miR-451a, hsa-miR-197-3p, hsa-miR-29b-3p, hsa-miR-126-3p, hsa-miR-146b-5p, hsa-miR-30c-5p, hsa-miR-144-5p, hsa-miR-29a-3p, hsa-miR-363-3p, hsa-miR-99a-5p, hsa-miR-342-3p, hsa-miR-193b-3p, hsa-miR-190a-5p, hsa-miR-365b-3p, hsa-miR-122b-5p, hsa-miR-122-3p | 3 mild COVID-19 patients, 5 severe COVID-19 patients, 2 negative controls | Plasma | Centrifugation 15 min at 2000g, heat inactivation | Internal purification protocol of GenXPro GmbH, based on silica columns | TrueQuant small RNA kit (GenXPro GmbH), Illumina sequencing | Not indicated | DESeq, no correction for multiple testing mentioned | 64 |
Large number of differentially expressed miRNAs between the different groups; hsa-miR-22-3p and hsa-miR-3180-3p associated with lower probability of survival at 90 days and hsa-let-7f-1-3p, hsa-let-7g-5p, hsa-miR-1255a, hsa-miR-140-3p, hsa-miR-20a-5p, hsa-miR-363-5p, hsa-miR-4510, hsa-miR-6130 associated with better prognosis | 32 severe, 52 moderate, 12 asymptomatic/mild COVID-19 patients, 13 controls | Plasma | Not indicated | miRNeasy Serum Plasma Advanced kit (Qiagen) | NEBNext Multiplex Small RNA Library Prep for Illumina (New England Biolabs), Illumina sequencing | DeSeq2 | Generalized linear model with negative binomial distribution, FDR correction using the Benjamini–Hochberg method | 65 |
Up in COVID-19: hsa-miR-6780b-3p, hsa-miR-6883-3p, hsa-miR-4769-5p, hsa-miR-6873-3p, hsa-miR-320b, hsa-miR-7111-3p, hsa-miR-4755-3p, hsa-miR-320c, hsa-miR-6511a-3p, hsa-miR-320d, hsa-miR-5187-3p, hsa-miR-4508, hsa-miR-4659a-5p; down in COVID-19: hsa-miR-4433b-5p, hsa-miR-16-2-3p, hsa-miR126-3p, hsa-miR-150-5p, hsa-miR-224-5p | 4 mild COVID-19 patients, 4 severe/critical COVID-19, 4 controls | Plasma | Centrifugation at 2500 rpm at 4 °C for 10 min | miRNeasy Serum/Plasma Kit (Qiagen) | QIAseq miRNA Library Kit (Qiagen), Illumina sequencing | DESeq2 | No correction for multiple testing mentioned | 66 |
Down in severe versus mild COVID-19: hsa-miR-451a, hsa-miR-101-3p, hsa-miR-185-5p, hsa-miR-30d-5p, hsa-miR-25-3p, hsa-miR-342-3p, hsa-miR-30e-5p, hsa-miR-150-5p, hsa-miR-15b-5p, hsa-miR-29c-3p, hsa-miR-10b-5p, hsa-miR-16-2-3p, hsa-miR-186-5p, hsa-miR-16-5p, hsa-miR-425-5p, hsa-miR-187-3p, hsa-miR-125a-5p, hsa-miR-106b-3p, hsa-miR-22-3p, hsa-miR-144-5p, hsa-miR-151a-3p, hsa-miR-30a-5p, hsa-miR-92a-3p, hsa-miR-15a-5p, hsa-miR-195-5p, hsa-miR-4508, hsa-miR-194-5p, hsa-miR-140-3p, hsa-miR-142-5p, hsa-miR-99a-5p, hsa-miR-363-3p, hsa-let-7g-5p, hsa-miR-20a-5p, hsa-miR-144-3p, hsa-miR-10a-5p, hsa-miR-378a-3p, hsa-let-7f-5p, hsa-miR-660-5p, hsa-miR-3135b, hsa-miR-96-5p, hsa-miR-125b-5p, hsa-let-7i-5p | 4 mild COVID-19 patients, 4 severe/critical COVID-19 | Plasma | Centrifugation at 2500 rpm at 4 °C for 10 min | miRNeasy Serum/Plasma Kit (Qiagen) | QIAseq miRNA Library Kit (Qiagen), Illumina sequencing | DESeq2 | No correction for multiple testing mentioned | 66 |
Differentially expressed according to severity of COVID-19: hsa-miR-133a-3p, hsa-miR-122-5p, hsa-miR-126-3p, hsa-miR-21-5p, hsa-miR-197-3p, hsa-miR-320a-3p, hsa-miR-223-3p, hsa-miR-210-3p, hsa-miR-192-5p | Screening: 18 COVID-19 patients with mild to moderate disease, 18 with severe disease, 11 controls; validation: 6 mild, 39 moderate, 16 severe COVID-19 patients | Citrate/CTAD plasma | Centrifugation at 2000g for 15 min, 12000g for 5 min | RNAseq: Maxwell RSC miRNA Tissue kit (Promega), RT-qPCR: miRNeasy Mini kit (Qiagen) | RNA-seq: RealSeq-Biofluids Plasma/Serum miRNA Library kit for Illumina sequencing (RealSeq Biosciences); RT-qPCR: miRCURY LNA RT kit (Exiqon), miRCURY LNA miRNA PCR Assays (Qiagen) | RNA-seq: not detailed; RT-qPCR: cel-miR-39-3p (Qiagen) | RNA-seq: the quasi-likelihood negative binomial generalized log-linear model functions of edge R v3.2828, DESeq2, Benjamini–Hochberg FDR correction; RT-qPCR: Student's t-tests/Mann–Whitney U tests, using ANOVA tests for continuous variables, Benjamini–Hochberg FDR correction | 67 |
Discrimination between mild and severe COVID-19, discovery: hsa-miR-122-5p, hsa-let-7c-5p, hsa-miR-21-5p, hsa-miR-140-3p | 31 mild COVID-19 patients (9 who progressed to severe/critical disease) | Serum | Not detailed | miRNeasy Mini Kit (Qiagen) | QIAseq miRNA Library Kit (Qiagen) | Not indicated | One-way ANOVA, Fisher linear discriminant analysis, ROC analysis, Kaplan–Meier curves | 68 |
hsa-miR-146a-3p, hsa-miR-126-3p, hsa-miR-151-3p, hsa-miR-126-5p downregulated in severe COVID-19; hsa-miR-15a, hsa-miR-424, hsa-miR-627-5p, hsa-miR-145, hsa-miR-205-5p, hsa-miR-200c upregulated in severe COVID-19 | 13 mildly ill, 17 severely ill patients | Small extracellular vesicles isolated from serum | Specific isolation procedure “EV-CATCHER” | miRNeasy Serum/Plasma kit (Qiagen) | Small-RNA sequencing (laboratory developed protocol) | Read counts were normalized to total counts | DESeq2 and edgeR; one-way analysis of variance (ANOVA) with Bonferroni post hoc test | 74 |
hsa-miR-550-5p, hsa-miR-629-3p upregulated in severe COVID-19 | 13 mildly ill, 17 severely ill patients | Serum | Centrifugation | miRNeasy Serum/Plasma kit (Qiagen) | Small-RNA sequencing (laboratory developed protocol) | Read counts were normalized to total counts | DESeq2 and edgeR; one-way analysis of variance (ANOVA) with Bonferroni post hoc test | 74 |
hsa-let-7a-5p, hsa-let-7d-5p, hsa-let-7f-5p, hsa-miR-98-5p, hsa-miR-340-5p, hsa-miR-378a-3p predictive of clinical outcome | 50 mild COVID-19, 16 severe COVID-19 patients, 17 healthy controls | Whole blood | None | miRNeasy Serum/Plasma Advanced Kit (Qiagen) | Small RNA sequencing (NEXTflex Small RNA-seq Kit v3, PerkinElmer) | Not indicated | Student's t-test, no correction for multiple testing mentioned | 76 |
Studies that measured selected miRNAs | ||||||||
---|---|---|---|---|---|---|---|---|
Main findings | Patient numbers | Specimen type | Pre-analytical procedures | Extraction methods | miRNA quantification methods | Normalization methods | Statistical methods | Ref. |
ICU: intensive care unit. FDR: false discovery rate. ROC: receiver operating characteristics. AUC: area under the curve. | ||||||||
Higher expression of hsa-miR-29a-3p, hsa-miR-146a-3p, hsa-miR-155-5p, hsa-let-7b-3p in COVID-19 patients | 18 COVID-19 patients, 15 healthy controls | PBMC | Ficoll Hypaque density gradient centrifugation | miRNeasy Mini Kit (Qiagen) | RT-qPCR: miScript® II RT Kit (Qiagen), miScript SYBR Green PCR Kit (Qiagen) | SNORD47 RNA | t-Test, Mann–Whitney U test, ROC, FDR (Benjamini and Hochberg procedure) | 48 |
hsa-miR-21-5p, hsa-miR-155-5p, hsa-miR-499-5p upregulated in COVID-19 patients compared to healthy controls and influenza-ARDS patients. hsa-miR-126-3p downregulated in COVID-19 patients compared to healthy controls and influenza-ARDS patients. hsa-miR-208a-3p upregulated in both severe COVID-19 and influenza-ARDS groups compared to healthy controls | Discovery cohort: mechanically-ventilated COVID-19 (n = 18) and healthy controls (n = 15); validation cohort: mechanically-ventilated COVID-19 patients (n = 20), influenza-ARDS patients (n = 13), healthy controls (n = 32) | Serum | — | miRNeasy Serum/Plasma Advanced Kit (Qiagen) | TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems), RT-qPCR with specific TaqMan miR assays (Applied Biosystems) | cel-miR-39-3p | Mann–Whitney U test, Dunn's multiple comparisons test, ROC | 49 |
hsa-miR-10b-5p downregulated in COVID-19 patients | 33 patients with COVID-19, 29 controls | Plasma | Centrifugation 2500 rpm for 10 min | miRNeasy Serum/Plasma Kit (Qiagen) | miScript II RT Kit, miScript SYBR Green PCRKit (Qiagen) | U6 snRNA | Mann–Whitney U-test | 50 |
hsa-miR-29b-3p and hsa-miR-1246 levels higher in COVID-19 patients; hsa-miR-186-5p and hsa-miR-15a-5p lower in COVID-19 patients | 29 COVID-19 patients; 29 healthy controls | Plasma | Centrifugation at 3000g for 10 min; plasmas were inactivated in a water bath at 56 °C for 30 min | RNA extraction and purification kit (Machery-Nagel) | miDETECTA Track miRNA qRT-PCR Starter Kit and primers (RiboBio) | cel-miR-39 | Mann–Whitney U test | 51 |
hsa-miR-2392 higher in COVID-19 (serum and urine); hsa-miR-1-3p and hsa-miR-155-5p lower in the serum of COVID-19 patients | Serum: 20 COVID-19, 10 non COVID-19; urine: 25 COVID-19, 21 non COVID-19 patients | Serum, urine | Centrifugation | QIAGEN miRNeasy Serum/Plasma kit, Norgen urine microRNA Purification Kit | QIAGEN miRCURY LNA RT Kit, droplet digital PCR | Not detailed | Student's t-test (unadjusted), ROC | 52 |
Downregulation of miR-200c-3p and miR-421-5p in COVID-19 | 30 COVID-19 patients, 18 controls | Blood (not detailed) | Not indicated | Total RNA isolation kit (Yekta Tajhiz Azama, Tehran, Iran) | Laboratory developed RT-qPCRs; SYBR Green Master Mix (Ampliqon) | U6 | Student t-test | 53 |
Elevated in COVID-19: hsa-miR-146a-5p, hsa-miR-122-5p | 14 COVID-19 patients, 10 controls | Serum | Not indicated | Total RNA Purification Kit (Norgen) | TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems), Taqman assays (Applied Biosystems) | RNU6B | Mann–Whitney test | 55 |
hsa-miR-155-5p upregulated in COVID-19 versus controls, severe versus moderate COVID-19 and non-survivors versus survivors | 98 moderate and 52 severe COVID-19 patients, 50 controls | Plasma | Centrifugation 10000 rpm for 10 min | miRNeasy Mini Kit (Qiagen) | MiScript II reverse transcription Kit, miScript primer assay for miR-155, MiScript SYBR Green PCR Kit (Qiagen) | RNU6-2 | Post hoc test (LSD) for normal distributed variables, ROC | 69 |
Up in COVID-19: hsa-miR-146b; down in COVID-19: hsa-miR-155 | 22 mild, 15 severe COVID-19 patients, 15 controls | Plasma | Centrifugation at 4 °C for 10 min at 1200g, and for 10 min at 10000g | TRI-Reagent-LS (Sigma) | TaqMan MicroRNA-RT-Kit and Taqman assays (Thermo Fisher Scientific) | cel-miR39 | Univariate and multivariate logistic regression analysis, ROC | 70 |
Down in patients who died: hsa-miR-155 | 6 COVID-19 patients who died and 31 who survived | Plasma | Centrifugation at 4 °C for 10 min at 1200g, and for 10 min at 10000g | TRI-Reagent-LS (Sigma) | TaqMan MicroRNA-RT-Kit and Taqman assays (Thermo Fisher Scientific) | cel-miR39 | Univariate and multivariate logistic regression analysis, ROC | 70 |
hsa-miR-31-3p, hsa-miR-29a-3p, and hsa-miR-126-3p decreased in more severe disease grades; hsa-miR-17-3p increased in more severe disease grades | Grade 1 (n = 21), grade 2 (n = 20), grade 3 (n = 20), grade 4 (n = 21), grade 5 (n = 21) COVID-19 patients, 20 healthy controls | Serum | Not indicated | mirPremier microRNA isolation kit (Sigma-Aldrich) | Mir-X miRNA First-Strand Synthesis kit (Takara Bio Inc.), Mir-X miRNA qPCR SYBR (Invitrogen) | RNU 48 | One-way ANOVA | 71 |
hsa-mir-21, hsa-mir-124, hsa-mir-146a downregulated, hsa-mir-326, hsa-mir-155, hsa-mir-27b upregulated in COVID-19 patients with increase of disease grade | Grade 1 (n = 21), grade 2 (n = 20), grade 3 (n = 20), grade 4 (n = 21), grade 5 (n = 21) COVID-19 patients, 20 healthy controls | Serum | Centrifugation at 3000 rpm for 10 min | mirPremier microRNA isolation kit (Sigma-Aldrich) | Mir-X miRNA First-Strand Synthesis kit (Takara Bio Inc.), Mir-X miRNA qPCR SYBR (Invitrogen) | RNU 48 | One-way ANOVA | 72 |
hsa-miR-4257 downregulated in COVID-19 versus control and in severe versus mild COVID-19 | 59 mild COVID-19, 41 severe COVID-19, 100 healthy controls | Serum | Centrifugation at 4000 rpm for 20 min | miRNeasy extraction kit (Qiagen) | TaqMan MicroRNA Reverse Transcription Kit (Thermo Fisher Scientific), miR-4257 TaqMan probe, universal TaqMan master mix | U6 snRNA | One-way ANOVAs, Mann–Whitney test, ROC, multivariable logistic regression analysis | 73 |
Main findings | Patient numbers | Specimen type | Pre-analytical procedures | Extraction methods | miRNA quantification methods | Normalization methods | Statistical methods | Ref. |
---|---|---|---|---|---|---|---|---|
Associated with severe COVID-19: hsa-miR-125a-5p, hsa-miR-1290, hsa-miR-15b-5p, hsa-miR-491-5p, hsa-miR-532-3p, hsa-miR-200b-3p, hsa-miR-629-5p, hsa-miR-103a-2-5p, hsa-miR-340-5p, hsa-miR-455-5p | 21 non severe COVID-19 patients, 20 severe COVID-19 patients, 20 controls | Nasopharyngeal swabs | None | MagMAX mirVana Total RNA Isolation Kit (Thermo Fisher Scientific) | TaqMan Advanced miRNA cDNA Synthesis Kit (Thermo Fisher Scientific), TaqMan Fast Advanced Master Mix, TaqMan™ Advanced miRNA Human A and B Cards (Thermo Fisher Scientific) | Global mean normalization | Mann–Whitney U test, sparse partial least squares-discriminant analysis, ROC, FDR (Benjamini and Hochberg procedure) | 39 |
hsa-miR-200c-3p overexpressed in severe COVID-19 | 39 patients without COVID-19, 37 symptomatic not hospitalized COVID-19 patients, 21 hospitalized COVID-19 patients, 14 patients with severe COVID-19 | Saliva/sublingual smear (severe group) | Collecting tube with saline solution for sublingual smears | EasyExtract DNA–RNA Kit (Interprise) | TaqMan miRNA reverse transcription kit (Applied Biosystems), hsa-miR-200c-3p (478351_mir) qPCR with HOT FIRE Pol Probe Universal qPCR Mix (Solis BioDyne) | miR-191 | Students' t-test/Mann–Whitney test, logistic regression | 77 |
Deregulation of five miRNA ratios: hsa-miR-122-5p/hsa-miR-199a-5p, hsa-miR-125a-5p/hsa-miR-133a-3p, hsa-miR-155-5p/hsa-miR-486-5p, hsa-miR-214-3p/hsa-miR-222-3p, hsa-miR-221-3p/hsa-miR-27a-3p between COVID-19 and non-COVID-19 ICU patients | 18 COVID-19 ICU patients; 14 non-COVID-19 ICU patients | Bronchial aspirates | None | miRNeasy Mini Kit (Qiagen) | miRCURY LNA RT Kit (Qiagen); miRCURY LNA SYBR Green PCR Kit (Qiagen) in 384-well miRCURY LNA miRNA Custom PCR Panels (Qiagen) | Ratios | Linear models for arrays, not mentioned whether FDR was used | 78 |
Deregulation of five miRNA ratios: hsa-miR-1-3p/hsa-miR-124-3p, hsa-miR-125b-5p/hsa-miR-34a-5p, hsa-miR-126-3p/hsa-miR-16-5p, hsa-miR-199a-5p/hsa-miR-9-5p, hsa-miR-221-3p/hsa-miR-491-5p between COVID-19 ICU survivors and nonsurvivors | 39 COVID-19 ICU survivors and 18 nonsurvivors | Bronchial aspirates | None | miRNeasy Mini Kit (Qiagen) | miRCURY LNA RT Kit (Qiagen); miRCURY LNA SYBR Green PCR Kit (Qiagen) in 384-well miRCURY LNA miRNA Custom PCR Panels (Qiagen) | Ratios | Linear models for arrays, not mentioned whether FDR was used | 78 |
Upregulated in COVID-19: hsa-miR-142-3p, hsa-miR-93-5p, hsa-miR-486-5p, hsa-miR-451a, hsa-miR-19a-3p; downregulated in COVID-19: hsa-miR-3065-3p, hsa-miR-3065-5p, hsa-miR-628-3p | 12 COVID-19 positive, 8 uninfected patients | Anterior nares swabs | None | miRNeasy micro kit (Qiagen) | QIAseq miRNA Library Kit with QIAseq miRNA NGS 48 Index IL (Qiagen), sequencing on the NovaSeq 6000 (Illumina) | Not detailed | DESeq2; adjusted FDR of a p-value of < 0.05, machine learning | 79 |
Upregulated in COVID-19: hsa-miR-100, hsa-miR-34b-5p, hsa-miR-200a, hsa-miR-34c-5p, hsa-miR-342-5p, hsa-let-7i, hsa-miR-29a | 10 SARS-CoV-2-positive and 10 SARS-CoV-2-negative patients | Nasopharyngeal swabs | None | In-house method using Sera-Mag beads (GE Healthcare) | Small RNA library construction: NEBNext® Small RNA Library Prep Set for Illumina® (NEB), purification: TailorCut Gel Extraction Tool Set (SeqMatic); MiniSeq High Output Reagent Kit, 75 cycles (Illumina); validation by RT-qPCR (TaqMan MicroRNA Assays, Applied Biosystems) | Normalized to 1 million reads (RPM); hsa-mir-148a miRNA for RT-qPCR | Principal component analysis, Kruskal–Wallis test, not mentioned whether FDR was used | 80 |
Up-regulated in COVID-19: hsa-miR-4443, hsa-miR-12116, hsa-miR-765, hsa-miR-1224-3p, hsa-miR-6880-3p, hsa-miR-6886-3p, hsa-miR-6758-5p, hsa-miR-4716-5p, hsa-miR-1281, hsa-miR-6741-3p, hsa-miR-769-5p, hsa-miR-877-3p, hsa-miR-1469, hsa-miR-10401-5p, hsa-miR-7111-3p, hsa-miR-4646-3p, hsa-miR-204-3p, hsa-miR-7107-5p, hsa-miR-6510-5p, hsa-miR-6823-3p, hsa-miR-3196, hsa-miR-665, hsa-miR-7847-3p, hsa-miR-1268b, hsa-miR-139-3p; downregulated: hsa-miR-34b-3p | 4 SARS-CoV-2-positive and 4 SARS-CoV-2-negative patients | Nasopharyngeal swabs | None | mirVana PARIS kit (Invitrogen) | Small RNA libraries were prepared using NEB Next Multiplex Small RNA Library Prep Set (Illumina) | DEseq2 median of ratios method | Unpaired two-tailed Mann–Whitney U test, not mentioned whether FDR was used | 81 |
Sequencing was used in several studies to profile miRNA expression. Farr et al. reported 50 differentially expressed miRNAs between COVID-19 patients and healthy controls. They proposed three miRNAs (hsa-miR-423-5p, hsa-miR-23a-3p, hsa-miR-195-5p) that classified early SARS-CoV-2 infection accurately with an AUC = 1.0.41 A limitation of the study is the small patient number, especially for the comparison of COVID-19 with and without oxygen that was only based on 3 and 4 patients, respectively (Table 1). An additional comparison between COVID-19 patients and healthy controls was published by Li et al. in two articles.42,43 As far as miRNA expression is concerned, both articles seem to show the same data. The most differentially expressed miRNAs were hsa-miR-16-2-3p, hsa-miR-6501-5p, hsa-miR-618, hsa-miR-183-5p, hsa-miR-627-5p and hsa-miR-144-3p. However, fold-changes were moderate ranging from 1.3 to 2.3.42 This study is further limited by the small patient number and lack of description of whether a correction for multiple testing was used (Table 1).
Yang et al.44 found differentially expressed miRNAs in COVID-19 patients compared to healthy controls: hsa-miR-19b-1-3p, hsa-miR-96, hsa-miR-19b-2-5p, hsa-miR-451a, hsa-miR-451b, hsa-miR-194-1-5p, hsa-miR-144, hsa-miR-486-2-3p, hsa-miR-15a, hsa-miR-29c-3p were downregulated in COVID-19 whereas hsa-miR-3609, hsa-miR-1244-1, hsa-miR-663a, hsa-miR-3916, hsa-miR-3687-2, hsa-miR-7846-3p, hsa-miR-5047, hsa-miR-3184-5p, hsa-miR-1248, hsa-miR-6891-3p were upregulated in COVID-19 patients. The small patient number remains a limitation of this study (Table 1).
Akula et al. found 8 differentially expressed miRNAs in plasma of COVID-19 patients compared to healthy controls by using sequencing and validation by RT-qPCR (Table 1).45 Interestingly, hsa-miR-150-5p was subsequently shown to inhibit SARS-CoV-2 replication. Therefore, the observed downregulation of hsa-miR-150-5p in COVID-19 patients may promote SARS-CoV-2 replication.45
A study with relatively large patient numbers identified several differentially expressed miRNAs between COVID-19 patients and controls (Table 1). Of note, the vast majority was downregulated in COVID-19: hsa-let-7i, hsa-let-7g, hsa-let-7f-2, hsa-let-7f-1, hsa-let-7d, hsa-miR-103a1, hsa-miR-126, hsa-miR-139, hsa-miR-15b, hsa-miR-16-1, hsa-miR-16-2, hsa-miR-181b1, hsa-miR-191, hsa-miR-21, hsa-miR-221, hsa-miR-222, hsa-miR-224, hsa-miR-23a, hsa-miR-23b, hsa-miR-25, hsa-miR-26a1, hsa-miR-30c1, hsa-miR-30c2, hsa-miR-30d, hsa-miR-320a, hsa-miR-92a1, hsa-miR-92a2, hsa-miR-93, hsa-miR-98, hsa-miR-151a, hsa-miR-340, hsa-miR-423, hsa-miR-425, hsa-miR-451a, hsa-miR-409, hsa-miR-454, hsa-miR-374b, hsa-miR-1244-1, hsa-miR-3609, hsa-let-7a2, hsa-let-7a1. Only three miRNAs were upregulated: hsa-miR-663b, hsa-miR-3687-2 and hsa-miR-3648-2.46
Meidert et al. used NGS of serum-derived extracellular vesicles and identified several miRNAs dysregulated in patients with COVID-19 pneumonia compared to healthy controls. When RT-qPCR was used to validate the findings, hsa-miR-193a-5p and hsa-miR-197-3p were upregulated in COVID-19 pneumonia patients (Table 1).47
Four studies investigated the expression of selected miRNAs by RT-qPCR: Donyavi et al. found that four specific miRNAs, hsa-miR-29a-3p, hsa-miR-146-3p, hsa-miR-155-5p and hsa-let-7b-3p were upregulated in PBMCs of COVID-19 patients compared to healthy controls.48 ROC curve analysis revealed that hsa-miR-29a-3p, hsa-miR-146a-3p and hsa-miR-155-5p allowed discrimination between controls and COVID-19 patients. On the other hand, to distinguish post-acute from acute phase COVID-19, hsa-miR-29a-3p and hsa-miR-146-3p may be used.48
Garg et al. included influenza infected patients in addition to COVID-19 and healthy controls. They found hsa-miR-21-5p, hsa-miR-155-5p, hsa-miR-499-5p upregulated in COVID-19 patients compared to healthy controls and influenza-ARDS patients. Hsa-miR-208a-3p was upregulated in both severe COVID-19 and influenza-ARDS groups compared to healthy controls. This shows that some miRNAs may be commonly dysregulated by different viral infections whereas others are dysregulated specifically.49
In a large study, hsa-miR-10b-5p was found downregulated in the plasma of COVID-19 patients.50 Wu et al. analyzed expression of 12 IFN signaling-related miRNAs in plasma of 29 COVID-19 patients and 29 controls and found higher levels of hsa-miR-29b-3p and hsa-miR-1246 levels in COVID-19 patients and lower hsa-miR-186-5p and hsa-miR-15a-5p lower in COVID-19 patients.51 One should note that plasma samples were heated in this study. These two studies report differential expression but did not perform ROC analysis to evaluate diagnostic performance of these miRNAs as biomarkers.
McDonald et al. measured hsa-miR-2392 expression by droplet digital PCR and found it upregulated in COVID-19 patients, in both serum and urine, and hsa-miR-2392 expression was higher in serum of COVID-19 patients necessitating ICU care.52 Hsa-miR-1-3p and hsa-miR-155-5p were downregulated in the serum of COVID-19 patients.52
Abdolahi and colleagues used laboratory-developed RT-qPCR assays to quantify miRNA expression and found a downregulation of miR-200c-3p and miR-421-5p in COVID-19 patients.53
Using sequencing, Tang et al. found that hsa-miR-146a-5p, hsa-miR-21-5p and hsa-miR-142-3p were downregulated whereas hsa-miR-3605-3p was upregulated in COVID-19 compared to healthy controls in whole blood after red blood cell lysis.54 In contrast, hsa-miR-146a-5p was found upregulated in the serum of COVID-19 patients in another study.55 Some miRNAs were upregulated only in patients with severe COVID-19, including hsa-miR-15b-5p, hsa-miR-486-3p and hsa-miR-486-5p, and some miRNAs were downregulated only in severe cases, including hsa-miR-181a-2-3p, hsa-miR-31-5p, and hsa-miR-99a-5p.54 Another study limited by the small patient numbers and performing red blood cell lysis and sequencing, found hsa-miR-335-5p and hsa-miR-24-3p downregulated in severe COVID-19 (ref. 56) (Table 1). This leads to release of red blood cell miRNAs into the blood57 and the authors thus measured a combination of miRNAs in blood cells, red blood cells and extracellular miRNAs.
Two studies used NGS and validation by RT-qPCR to identify miRNAs differentially expressed in different severity grades of COVID-19: a striking downregulation of the miR-320 family members including hsa-miR-320a, hsa-miR-320b and hsa-miR-320c was found in whole blood of patients with severe respiratory failure. Other strongly downregulated miRNAs were hsa-miR-4747-3p, hsa-miR-4429, hsa-miR-6729-3p and hsa-miR-1908-5p, with hsa-miR-374a-3p, hsa-miR-15a-3p, hsa-miR-3688-5p, hsa-miR-4721 being strongly upregulated in COVID-19 patients with severe respiratory failure.58 The second study found that high serum levels of hsa-miR-320b and hsa-miR-483-5p were associated with a higher risk of mortality.59 The discordance of the results concerning hsa-miR-320b may be due to the use of whole blood versus serum for miRNA expression measurements. This shows that the choice of the specimen type is very important and that expression of miRNAs differs between different blood fractions.
More so, Li et al. made a comparison between severe/critical and mild/moderate COVID-19 patients and investigated COVID-19 patients after recovery. This resulted in the observation that hsa-miR-155 and hsa-miR-130a were remarkably higher in the mild cases compared to severe/critical disease and healthy controls.60 This study lacks on the other hand a detailed description of methods used for quantification and normalization of miRNA expression.
A study performed on 79 COVID-19 patients found that five miRNAs were up- and five miRNAs were downregulated in ICU COVID-19 patients compared to non ICU patients by using RT-qPCR61 (Table 1). However, the fold changes were modest. Hsa-miR-148a-3p, hsa-miR-486-5p and hsa-miR-451a were associated with ICU stay in multivariable analysis. Hsa-miR-16-5p, hsa-miR-92a-3p, hsa-miR-98-5p, hsa-miR-132-3p, hsa-miR-192-5p, hsa-miR-323a-3p were downregulated in patients who did not survive the ICU stay. The expression of two miRNAs, hsa-miR-192-5p and hsa-miR-323a-3p, predicted mortality during the ICU stay with an AUC of 0.80.61
Wilson et al. analyzed miRNA expression in plasma of 58 patients with mild, moderate or severe COVID-19 with Nanostring technology. A dysregulation of several miRNAs in severe COVID-19 was found, of which most were downregulated in severe COVID-19 (Table 1).62
Parray et al. compared differential expression of miRNAs between three different COVID-19 severity groups. Multiple miRNAs were found upregulated and downregulated in severe versus asymptomatic COVID-19 as well as between severe and mild COVID-19 (Table 1).63
Grehl and colleagues found 11 upregulated and 20 downregulated miRNAs in severe COVID-19 cases by next generation sequencing (Table 1).64 This study was performed on a small number of patients and no correction for multiple testing was detailed in the article.
Another study compared the expression of miRNAs in severe, moderate, mild/asymptomatic COVID-19 patients and controls. They performed several comparisons between the different groups and found a large number of differentially expressed miRNAs. They further built a score based on the expression of ten miRNAs (hsa-miR-22-3p, hsa-miR-3180-3p, hsa-let-7f-1-3p, hsa-let-7g-5p, hsa-miR-1255a, hsa-miR-140-3p, hsa-miR-20a-5p, hsa-miR-363-5p, hsa-miR-4510 and hsa-miR-6130) that allowed to classify patients into low or high risk for mortality.65
de Nicoletti et al. found several differentially expressed miRNAs between COVID-19 patients and uninfected controls as well as depending on COVID-19 severity (Table 1).66 Only four patients per group were analyzed and it was not mentioned whether correction for multiple testing was applied.
A thoroughly conducted study that included a screening and a validation cohort found several differentially expressed miRNAs in COVID-19 patients with different disease severities67 (Table 1). After adjustment for age and sex, hsa-miR-122-5p, and hsa-miR-133a-3p were the only miRNAs besides platelet- and endothelium-derived hsa-miR-126-3p that showed a significantly different expression in mild, moderate and severe COVID-19.67 RT-qPCR measurements of hsa-miR-122-5p and hsa-miR-133a-3p classified patients into severe (n = 16) and non-severe (n = 45) COVID-19 with AUCs of 0.75 and 0.79, respectively.67 Interestingly, hsa-miR-122-5p was also found in extracellular vesicles of COVID-19 patients and allowed to differentiate mild COVID-19 patients from those who progressed to severe or critical disease with an AUC = 0.81.68
Two studies specifically investigated the expression of hsa-miR-155. One found that hsa-miR-155 was overexpressed in COVID-19 versus controls, severe versus moderate COVID-19 and non-survivors versus survivors. AUC to distinguish COVID-19 patients from controls was 0.986 whereas for the distinction of severe and moderate patients AUC = 0.75.69 The other study found hsa-miR-155 underexpressed in patients with COVID-19 compared to controls and underexpressed in patients who died compared to survivors with an AUC = 0.83 to differentiate survivors and non survivors.70 These studies seem to have observed an opposite regulation of the same miRNA.
Keikha and coworkers published two articles on the expression of selected miRNAs in COVID-19. The first article reports that hsa-miR-31-3p, hsa-miR-29a-3p and hsa-miR-126-3p were decreased in more severe COVID-19 disease grades, whereas hsa-miR-17-3p was increased in more severe disease grades.71 Expression in healthy controls is not shown in the article. The second article, probably on the same cohort, showed that hsa-miR-21, hsa-miR-124, hsa-miR-146a were downregulated, whereas hsa-miR-326, hsa-miR-155, hsa-miR-27b were upregulated in COVID-19 patients with increased of disease severity.72
Hsa-miR-4257 was downregulated in COVID-19 patients versus controls and in severe versus mild COVID-19 as reported by Agwa and colleagues.73 ROC analysis revealed that hsa-miR-4257 expression could be used as biomarker to distinguish COVID-19 patients from controls with an AUC of 0.911.73
Differential expression of two miRNAs was found by using whole serum of patients with severe and non-severe COVID-19, whereas ten distinct differentially expressed miRNAs were identified when using extracellular vesicles isolated from sera (Table 1).74 This shows that the choice of the specimen type is very important and that expression of miRNAs differs between different blood fractions. Direct comparison of miRNA expression in different specimen types is therefore not recommended.75
By using RNA-seq, another study found 20 differentially expressed miRNAs in serum-derived extracellular vesicles from patients with COVID-19 pneumonia compared to patients with COVID-19 ARDS. However, only one of these (hsa-miR-206) showed the same differential expression when RT-qPCR was used (Table 1).47
In a relatively large study, six miRNAs, namely hsa-let-7a-5p, hsa-let-7d-5p, hsa-let-7f-5p, hsa-miR-98-5p, hsa-miR-340-5p, hsa-miR-378a-3p were found predictive of clinical outcome of COVID-19 among other biomarkers.76
One study measured expression of hsa-miR-200c-3p in saliva of symptomatic non-hospitalized, hospitalized COVID-19 patients and patients without COVID-19 as well as in sublingual smears of ICU COVID-19 patients (severe group).77 A higher expression of hsa-miR-200c-3p was found in the severe group (Table 2). However, given the difference in specimen types used in the different groups, this difference may be due to the different specimen type rather than the COVID-19 severity.77 Of note, hsa-miR-200c-3p was found downregulated in the blood of COVID-19 patients in an independent study.53
Molinero et al. studied miRNA expression in bronchial aspirate specimens of COVID-19 and non COVID-19 ICU patients. Rather than comparing the expression of single miRNAs, they compared expression ratios and found several differences of the expression of certain miRNA ratios when COVID-19 and non-COVID-19 as well as ICU survivors and nonsurvivors were compared (Table 2). They constructed a model for ICU mortality prediction and found that the best combination of miRNA ratios (hsa-miR-125b-5p/hsa-miR-34a-5p, hsa-miR-199a-5p/hsa-miR-9-5p, and hsa-miR-221-3p/hsa-miR-491-5p) had an AUC = 0.85 for predicting ICU mortality. The hsa-miR-199a-5p/hsa-miR-9-5p ratio showed an AUC = 0.80 for predicting ICU mortality.78
Farr et al. investigated miRNA expression in anterior nares swabs and found six differentially expressed miRNAs between COVID-19 and uninfected patients (Table 2). Supervised machine learning identified a three-miRNA signature consisting of hsa-miR-30c-2-3p, hsa-miR-628-3p and hsa-miR-93-5p independently classified COVID-19 cases with 100% accuracy.79
Eichmeier and colleagues identified 7 miRNAs in nasopharyngeal swabs by sequencing that were differentially expressed in COVID-19 positive and negative patients (Table 2). However, RT-qPCR comparison of the miRNA expression yielded different results.80
Wu et al. studied miRNA expression in nasopharyngeal swabs specimens of 4 COVID-19 patients and 4 controls and identified several differentially expressed miRNAs, where most of them were upregulated in COVID-19 compared to controls (Table 2).81
We recently investigated miRNA expression in COVID-19 patients and found ten miRNAs associated with severe COVID-19 (Table 2). Interestingly, expression of most of these miRNAs was lower in severe than in non-severe COVID-19 patients. ROC analysis revealed that three of these miRNAs are promising candidate biomarkers for severe COVID-19, namely hsa-miR-125a-5p, hsa-miR-491-5p and hsa-miR-200b-3p. These miRNAs discriminated severe from non-severe cases with AUCs ranging from 0.76 to 0.79.39
The reasons for the small overlap of miRNAs identified as differentially expressed in the different studies are multiple. First, the choice of the patient population differs. Next, the definition of mild versus severe COVID-19 also is not uniform in the different studies. Another main factor is certainly the choice of the initial specimen: the use of whole blood, plasma, serum or extracellular vesicles. miRNAs found associated with severe COVID-19 were distinct when using serum or extracellular vesicles from serum.74 Pre-analytical aspects and techniques used for miRNA extraction and expression profiling were also different (Tables 1 and 2). Normalization strategies and statistical methods were also diverse. Many studies did not report correction for multiple analyses and most did not validate their results in a validation cohort or by using a different technique. All these factors influence on the results.
Whereas the published studies reported relative quantification of miRNAs, the question of quantification for routine diagnostic use arises. Potentially, absolute quantification of miRNAs would be more useful in the routine diagnostic setting with cut-offs of miRNA expression levels associated with diagnostic or prognostic performance of the given miRNA to be defined.
The next steps may be to investigate influence of physiological factors on the expression of candidate biomarker miRNAs in order to determine whether age or sex influence their expression. The impact of infections with other viruses on their expression level also needs to be investigated in order to ascertain that the observed changes are specific to SARS-CoV-2 infection.
While a vast amount of data on miRNA expression in body fluids of COVID-19 patients is available, before the application of miRNAs as diagnostic or prognostic biomarkers, a number of technical and analytical issues have to be resolved first. Further studies with larger patient cohorts allowing for multivariable analysis taking into account confounding factors are necessary.
Footnote |
† Present address: Pathogenesis and Control of Chronic and Emerging Infections, INSERM, EFS, Univ Antilles, Univ Montpellier, Laboratoire de Virologie, CHU Montpellier, France. |
This journal is © The Royal Society of Chemistry 2023 |