Cited by 968MicroRNA in disease regulationGene expression and cancer classificationPlant Molecular Biology Research

The DOI matches its Crossref record; 29 of the 29 references it cites were found in a registry or an index; 14 of the 34 sources its claims cite were checked for support.

Distribution of miRNA expression across human tissues

Open access

Abstract

We present a human miRNA tissue atlas by determining the abundance of 1997 miRNAs in 61 tissue biopsies of different organs from two individuals collected post-mortem. One thousand three hundred sixty-four miRNAs were discovered in at least one tissue, 143 were present in each tissue. To define the distribution of miRNAs, we utilized a tissue specificity index (TSI). The majority of miRNAs (82.9%) fell in a middle TSI range i.e. were neither specific for single tissues (TSI > 0.85) nor housekeeping miRNAs (TSI < 0.5). Nonetheless, we observed many different miRNAs and miRNA families that were predominantly expressed in certain tissues. Clustering of miRNA abundances revealed that tissues like several areas of the brain clustered together. Considering -3p and -5p mature forms we observed miR-150 with different tissue specificity. Analysis of additional lung and prostate biopsies indicated that inter-organism variability was significantly lower than inter-organ variability. Tissue-specific differences between the miRNA patterns appeared not to be significantly altered by storage as shown for heart and lung tissue. MiRNAs TSI values of human tissues were significantly (P = 10(-8)) correlated with those of rats; miRNAs that were highly abundant in certain human tissues were likewise abundant in according rat tissues. We implemented a web-based repository enabling scientists to access and browse the data (https://ccb-web.cs.uni-saarland.de/tissueatlas).

Overview

AI snapshot

Written from the paper’s full text.

TL;DR

This paper presents a human miRNA tissue atlas by profiling 1,997 mature miRNAs across 61 post-mortem tissue biopsies from two male individuals, with 1,364 miRNAs detected in at least one tissue and 143 detected in all tissues. It reports tissue-specific miRNA patterns, a web repository, and conserved tissue specificity between human and rat.

Key findings

  • Of 1,997 mature miRNAs, 633 (31.7%) were not detected in any tested tissue; 1,364 were detected in at least one tissue, and 143 (10.5%) were found in all tissues.
  • Most detected miRNAs had intermediate tissue specificity: 82.9% fell in a middle TSI range, neither tissue-specific (TSI > 0.85) nor housekeeping (TSI < 0.5); in quantile-normalized data, 83.7% of miRNAs in the first body and 88.8% in the second body had intermediate TSI values.
  • Tissue-specific examples included miR-1–3p as the overall most tissue-specific miRNA, high in muscle and myocardium; miR-122–5p in liver, miR-7–5p in pituitary gland, miR-205–5p in skin, and miR-507/miR-514a-3p/miR-509–5p in testis.
  • Human and rat tissue specificity was significantly correlated (TSI correlation 0.362, P = 9 × 10−8; permutation average correlation 0), and inter-individual variability was significantly lower than inter-organ variability in lung and prostate analyses (P < 10−16).

Methods

  • Profiled 61 tissue biopsies from two male cadavers (65-year-old with multiple myeloma; 59-year-old with no signs of cancer) using Agilent SurePrint 8 × 60K Human V19 miRNA microarrays, covering 1,997 human miRNAs present in both miRBase V19 and V21.
  • Pre-analytics included 44 tissue miRNomes for reproducibility and storage/degradation analyses, 16 additional normal lung biopsies from different individuals, and GEO datasets from normal gastric and prostate tissues.
  • Calculated a tissue specificity index (TSI) for each miRNA, performed hierarchical clustering, miRNA family and co-expression analyses, and compared human TSI values with rat data from GEO series GSE52754; analyses used raw, quantile-normalized, and variance-stabilized normalized data.

Limitations

  • The paper states that autopsy samples from only two male individuals provide only a snapshot of full miRNA expression variability, and that age, sex, and disease status can influence miRNA patterns.
  • The authors state they cannot exclude that some tissue-specific miRNAs might be affected by sample degradation, although storage between 1 and 14 days at 4°C had limited overall impact on tissue miRNA patterns.
  • The paper notes that microarray analysis cannot detect presently unknown miRNAs or miRNA isoforms, and that normalization choice can substantially influence results; these are stated limitations of the platform and analysis approach.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1093/nar/gkw116

Cites

29 of 29 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1. Hierarchical clustering of the 44 samples included in the stability and reproducibility study. Quantile normalized and log2 transformed expression intensity values were used for clustering. The intensity values and distribution are presented in the upper left corner. In the present heat map, heart and lung tissues cluster together on the right-hand side. Technical replicates (marked by ‘TR’ in the labels below the heat map) of other organs cluster together in each case in the left-hand side. For VSN-normalized data the same representation is provided in Supplementary Figure S1.
Figure 1. Hierarchical clustering of the 44 samples included in the stability and reproducibility study. Quantile normalized and log2 transformed expression intensity values were used for clustering. The intensity values and distribution are presented in the upper left corner. In the present heat map, heart and lung tissues cluster together on the right-hand side. Technical replicates (marked by ‘TR’ in the labels below the heat map) of other organs cluster together in each case in the left-hand side. For VSN-normalized data the same representation is provided in Supplementary Figure S1.CC BY-NC · PubMed Central
Figure 2. Histogram plot for the frequency of TSI of miRNAs in different tissues. Panel A represents TSI of the first, panel B of the second body. The vertical dotted lines correspond to the threshold originally proposed for defining housekeeping and specifically expressed miRNAs of <0.15 and >0.85. The same representation for VSN-normalized data is presented in Supplementary Figure S2.
Figure 2. Histogram plot for the frequency of TSI of miRNAs in different tissues. Panel A represents TSI of the first, panel B of the second body. The vertical dotted lines correspond to the threshold originally proposed for defining housekeeping and specifically expressed miRNAs of <0.15 and >0.85. The same representation for VSN-normalized data is presented in Supplementary Figure S2.CC BY-NC · PubMed Central
Figure 3. Bar plots for all 61 samples for miR-1–3p, the miRNA with highest overall TSI in the first and second body. The vertical dashed line separates the first from the second body. TSI values for both bodies are highlighted in the figure. The miRNA is high expressed in muscle and myocardium. Raw-, quantile- and VSN-normalized expression intensities for this miRNA match well across all different tissues.
Figure 3. Bar plots for all 61 samples for miR-1–3p, the miRNA with highest overall TSI in the first and second body. The vertical dashed line separates the first from the second body. TSI values for both bodies are highlighted in the figure. The miRNA is high expressed in muscle and myocardium. Raw-, quantile- and VSN-normalized expression intensities for this miRNA match well across all different tissues.CC BY-NC · PubMed Central
Figure 4. Hierarchical clustering of all tissues in both bodies. Log2 transformed quantile normalized intensity values were used for clustering. The intensity value distribution is shown in the upper right corner of the figures. Panel A shows significantly expressed miRNAs, while panel B focuses on the 100 miRNAs with overall highest data variance. The respective representation for VSN-normalized data is presented in Supplementary Figure S3.
Figure 4. Hierarchical clustering of all tissues in both bodies. Log2 transformed quantile normalized intensity values were used for clustering. The intensity value distribution is shown in the upper right corner of the figures. Panel A shows significantly expressed miRNAs, while panel B focuses on the 100 miRNAs with overall highest data variance. The respective representation for VSN-normalized data is presented in Supplementary Figure S3.CC BY-NC · PubMed Central
Figure 5. Heat map for the 25 miRNAs that have TSI values of >0.85 in both bodies. Log2 transformed expression intensities of quantile normalized expression values are presented. To facilitate the interpretation of specific miRNAs in organs or organ groups low expressed miRNAs were greyed out (see also color distribution scheme in the upper left corner). The analysis highlights tissue-specific miRNAs that are exemplarily presented on the right-hand side of the plot, such as hsa-miR-1–3p that has already been described in Figure 3 as most specific miRNA overall.
Figure 5. Heat map for the 25 miRNAs that have TSI values of >0.85 in both bodies. Log2 transformed expression intensities of quantile normalized expression values are presented. To facilitate the interpretation of specific miRNAs in organs or organ groups low expressed miRNAs were greyed out (see also color distribution scheme in the upper left corner). The analysis highlights tissue-specific miRNAs that are exemplarily presented on the right-hand side of the plot, such as hsa-miR-1–3p that has already been described in Figure 3 as most specific miRNA overall.CC BY-NC · PubMed Central
Figure 6. A: Average and standard deviation of TSI value in different miRNA families. For each miRNA family with at least five members the mean and standard deviation of all family members TSI is presented as bar plot. Families are sorted with decreasing average tissue specificity from left to right. Highest tissue specificity was observed for the miR-378 family, predominantly being specific for myocardium and muscle. The number of mature family members is shown above the columns with balloons, representing the family size. B: Co-expression network of miRNAs. Each miRNA pair connected by an edge has co-expression across all samples with Spearman correlation coefficient above 0.95.
Figure 6. A: Average and standard deviation of TSI value in different miRNA families. For each miRNA family with at least five members the mean and standard deviation of all family members TSI is presented as bar plot. Families are sorted with decreasing average tissue specificity from left to right. Highest tissue specificity was observed for the miR-378 family, predominantly being specific for myocardium and muscle. The number of mature family members is shown above the columns with balloons, representing the family size. B: Co-expression network of miRNAs. Each miRNA pair connected by an edge has co-expression across all samples with Spearman correlation coefficient above 0.95.CC BY-NC · PubMed Central
Figure 7. Conservation of tissue-specific expression of miRNAs in human and rat. Matching miRNAs (100% matching of mature miRNA sequence) from organ expression in rats and humans were calculated. For each miRNA in rats and humans the TSI was calculated and highly specific miRNAs were clustered. Since overall expression in humans and rats varied, the maximal intensity of each miRNA in the two organs was set to 100% and all other miRNAs were linearly scaled. All miRNAs with below 10% expression of maximal intensity are shown in grey to facilitate data interpretation (see also colour gradient presented in the upper right corner). On the right-hand side the human/rat miRNA identifiers are shown, below the heat map the matched tissues are presented (HU for human; RA for rat). For rat tissues the average intensity of replicated measurements is presented.
Figure 7. Conservation of tissue-specific expression of miRNAs in human and rat. Matching miRNAs (100% matching of mature miRNA sequence) from organ expression in rats and humans were calculated. For each miRNA in rats and humans the TSI was calculated and highly specific miRNAs were clustered. Since overall expression in humans and rats varied, the maximal intensity of each miRNA in the two organs was set to 100% and all other miRNAs were linearly scaled. All miRNAs with below 10% expression of maximal intensity are shown in grey to facilitate data interpretation (see also colour gradient presented in the upper right corner). On the right-hand side the human/rat miRNA identifiers are shown, below the heat map the matched tissues are presented (HU for human; RA for rat). For rat tissues the average intensity of replicated measurements is presented.CC BY-NC · PubMed Central

Verification

Claims the paper makes

83 Own result · 15 No full text of the sources · 7 Supported · 6 Own method · 6 Own interpretation · 5 Partly supported · 1 Not checked yet · 1 No basis given · accuracy not yet measured

Show all 73 claims

  • Probably not a substantial claim. 19
  • It may state two separate claims as one. 35
  • The quote only partly states this claim. 4
51 of 51
  • The paper relies on this without support.

    The two post-mortem male individuals are representative of the general human population for tissue miRNA distribution.

    “Tissues analysed in this study originated from two male bodies.”
    “we measured all miRNA profiles from only two different individuals to minimize inter-individual variability.”
  • The paper relies on this without support.

    The first individual's multiple myeloma did not substantially alter the miRNA profiles of his non-tumor tissues.

    “The first body was from a 65-year-old male patient, who suffered from multiple myeloma, a cancer that forms in a type of white blood cells (plasma cells).”
  • The paper relies on this without support.

    RNA integrity values between 1.8 and 2.7 in autopsy samples are sufficient for reliable miRNA profiling.

    “As expected for autopsy samples, the RNA integrity values (RIN) ranged between 1.8 and 2.7.”
  • The paper relies on this without support.

    The majority of studies investigate differential expression in only one organ/tissue.

    “Although there is ample evidence for differential miRNA expression in different human tissues, the majority of studies investigate differential expression in only one organ/tissue.”
  • The paper relies on this without support.

    Results of studies are not easily comparable due to different identification methods and normalization strategies, limiting their value for comparison of miRNA expression in different tissues.

    “Due to the different identification methods and normalization strategies, the results of these studies are not easily comparable limiting their value for comparison of miRNA expression in different tissues.”
  • The paper relies on this without support.

    Biopsies will be mostly taken from patients with affected organs.

    “biopsies will be mostly taken from patients with affected organs”
  • The paper relies on this without support.

    High inter-individual differences can mask tissue-specific differences of miRNA abundances.

    “high inter-individual differences can mask tissue-specific differences of miRNA abundances”
  • The paper relies on this without support.

    A bias is likely introduced by multiple centres involved in tissue collections.

    “a bias is likely introduced by multiple centres that are involved in tissue collections”
  • The paper relies on this without support.

    Samples of vital organs (e.g., thalamus, spinal cord, cerebellum) are not available from biopsies.

    “samples of vital organs, e.g. thalamus, spinal cord or cerebellum, are not available”
  • The paper relies on this without support.

    RNA degradation in autopsy samples due to storage duration, advanced age, or disease status of body donors.

    “with the disadvantage of RNA degradation in the samples due to the storage duration of the body and the advanced age or the disease status of the body donors”
  • The paper relies on this without support.

    Many normalization techniques can have a substantial influence on the results.

    “Many techniques that are frequently applied such as variance stabilizing normalization or quantile normalization can have a substantial influence on the results.”
11 of 11

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

4 of 4

Citations

Cites

Show all 29 references

Cited by

Sorted by citations

Items 801–900 of 968

Metadata

Authors
Ludwig, N., Leidinger, P., Becker, K., Backes, C., Fehlmann, T., Pallasch, C., Rheinheimer, S., Meder, B., Stähler, C., Meese, E., & Keller, A.
Affiliations
  • Saarland University (DE)
  • Cologne Excellence Cluster on Cellular Stress Responses in Aging Associated Diseases (DE)
  • University Hospital Heidelberg (DE)
  • Siemens Healthineers (Germany) (DE)
Corresponding author
Keller, Andreas
Venue
Nucleic Acids Research
Published
2016-02-25
Volume
44
Issue
8
Pages
3865-3877
ISSN
  • 0305-1048 (ISSN-L)
  • 1362-4962
OpenAlex
W2288482887
Cited by
968
References
29
Citations by year
2026: 45 · 2025: 73 · 2024: 93 · 2023: 97 · 2022: 104 · 2021: 138 · 2020: 141 · 2019: 109 · 2018: 88 · 2017: 61 · 2016: 17
FWCI
32.73
Type
journal article
Language
en
Open access
Locations
  • Nucleic Acids Research (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • Kölner Universitäts PublikationsServer (Universität zu Köln) (publishedVersion)
  • HighWire Press Open Archive (submittedVersion)
Funders
Universität des Saarlandes
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Cancer Research › MicroRNA in disease regulation
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene expression and cancer classification
  • Life Sciences › Agricultural and Biological Sciences › Plant Science › Plant Molecular Biology Research
Keywords
microRNA expression, human tissue atlas, miRNA abundance, miR-150, brain, lung, prostate, heart, rat comparison, postmortem tissue, web-based repository
MeSH terms
Adult, Aged, Animals, Humans, Male, Organ Specificity, Tissue Distribution, Reproducibility of Results, Databases, Nucleic Acid, MicroRNAs, Rats

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-11-06, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-06 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review