CoChat
Sign Up
Cited by 126Muscle Physiology and DisordersGenetics, Aging, and Longevity in Model OrganismsNutrition and Health in Aging

The DOI matches its Crossref record; 106 of the 107 references it cites were found in a registry or an index.

Skeletal muscle transcriptome in healthy aging

Open access
Tumasian, Robert A., Harish, Abhinav, Kundu, Gautam, Yang, Jen-Hao, Ubaida-Mohien, Ceereena, Gonzalez-Freire, Marta, Kaileh, Mary, Zukley, Linda M., Chia, Chee W., Lyashkov, Alexey, Wood, William H., Piao, Yulan, Coletta, Christopher, Ding, Jun, Gorospe, Myriam, Sen, Ranjan, De, Supriyo, Ferrucci, Luigi·2021·Nature Communications
DOI 10.1038/s41467-021-22168-2PMC PMC8016876
Open this paper in CoChat — free accountOpen PDF

Abstract

Age-associated changes in gene expression in skeletal muscle of healthy individuals reflect accumulation of damage and compensatory adaptations to preserve tissue integrity. To characterize these changes, RNA was extracted and sequenced from muscle biopsies collected from 53 healthy individuals (22-83 years old) of the GESTALT study of the National Institute on Aging-NIH. Expression levels of 57,205 protein-coding and non-coding RNAs were studied as a function of aging by linear and negative binomial regression models. From both models, 1134 RNAs changed significantly with age. The most differentially abundant mRNAs encoded proteins implicated in several age-related processes, including cellular senescence, insulin signaling, and myogenesis. Specific mRNA isoforms that changed significantly with age in skeletal muscle were enriched for proteins involved in oxidative phosphorylation and adipogenesis. Our study establishes a detailed framework of the global transcriptome and mRNA isoforms that govern muscle damage and homeostasis with age.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41467-021-22168-2

Cites

106 of 106 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Methodology overview, model overlap, and heat map of significant changes in RNA (ENSG) levels. a Flowchart of sample preparation (orange) and analytical methods (green to gray) used in this study. b Volcano plots capture all significant (p < 0.01 from two-sided Wald test, unadjusted) RNAs identified by our models: linear model including all RNAs, standard negative binomial model including only RNAs with all non-zero CPB values, and zero-adjusted negative binomial model including only RNAs with at least one value of zero CPB. Red and blue points indicate significant RNAs with positive and negative β-values for age, respectively. Black points denote RNAs with nonsignificant β-values for age. c Venn diagram displaying the overlap between all RNAs significantly associated with aging identified by the linear (blue) and negative binomial (yellow) models (overlap shown in brown). Black denotes total significant RNAs, whereas red and blue values represent significant RNAs with positive and negative β-values for age, respectively. d Heat map representing the expression levels of the 506 significant RNAs shared between the linear and negative binomial models with age. Row-wise z-scores were calculated and color-coded to display changes in RNA expression with increasing age (22–83 years old). Light yellow and dark blue bars indicate RNAs with very low and very high expression levels, respectively, at a particular age. In addition, a correlation-based distance method was used to cluster the RNAs obtained from the linear model with positive and negative β-values for age. RNAs with positive β-values are shown in the top of the heat map, containing RNAs with low to high expression (yellow to blue rows) with increasing age. Similarly, RNAs with negative β-values are shown on the bottom of the heat map, containing RNAs with high to low expression (blue to yellow rows) with increasing age.
Fig. 1 Methodology overview, model overlap, and heat map of significant changes in RNA (ENSG) levels. a Flowchart of sample preparation (orange) and analytical methods (green to gray) used in this study. b Volcano plots capture all significant (p < 0.01 from two-sided Wald test, unadjusted) RNAs identified by our models: linear model including all RNAs, standard negative binomial model including only RNAs with all non-zero CPB values, and zero-adjusted negative binomial model including only RNAs with at least one value of zero CPB. Red and blue points indicate significant RNAs with positive and negative β-values for age, respectively. Black points denote RNAs with nonsignificant β-values for age. c Venn diagram displaying the overlap between all RNAs significantly associated with aging identified by the linear (blue) and negative binomial (yellow) models (overlap shown in brown). Black denotes total significant RNAs, whereas red and blue values represent significant RNAs with positive and negative β-values for age, respectively. d Heat map representing the expression levels of the 506 significant RNAs shared between the linear and negative binomial models with age. Row-wise z-scores were calculated and color-coded to display changes in RNA expression with increasing age (22–83 years old). Light yellow and dark blue bars indicate RNAs with very low and very high expression levels, respectively, at a particular age. In addition, a correlation-based distance method was used to cluster the RNAs obtained from the linear model with positive and negative β-values for age. RNAs with positive β-values are shown in the top of the heat map, containing RNAs with low to high expression (yellow to blue rows) with increasing age. Similarly, RNAs with negative β-values are shown on the bottom of the heat map, containing RNAs with high to low expression (blue to yellow rows) with increasing age.CC BY · PubMed Central
Fig. 2 Top 20 significant (p < 0.01) RNAs (ENSGs) identified by linear regression with positive β-values for age. a Linear model p-values (two-sided Wald test, unadjusted) and β-values, and RNA characteristics obtained from literature review, are shown. RNAs that failed to have at least ten samples with CPM values above −3 were removed. b Linear plots representing log2(CPM) values with age for the top ten RNAs ranked by the lowest p-value (two-sided Wald test, unadjusted)36,37,84–96.
Fig. 2 Top 20 significant (p < 0.01) RNAs (ENSGs) identified by linear regression with positive β-values for age. a Linear model p-values (two-sided Wald test, unadjusted) and β-values, and RNA characteristics obtained from literature review, are shown. RNAs that failed to have at least ten samples with CPM values above −3 were removed. b Linear plots representing log2(CPM) values with age for the top ten RNAs ranked by the lowest p-value (two-sided Wald test, unadjusted)36,37,84–96.CC BY · PubMed Central
Fig. 3 Top 20 significant (p < 0.01) RNAs (ENSGs) identified by linear regression with negative β-values for age. a Linear model p-values (two-sided Wald test, unadjusted) and β-values, and RNA characteristics obtained from literature review, are shown. RNAs that failed to have at least ten samples with CPM values above −3 were removed. b Linear plots representing log2(CPM) values with age for the top ten RNAs ranked by the lowest p-value (two-sided Wald test, unadjusted)97–108.
Fig. 3 Top 20 significant (p < 0.01) RNAs (ENSGs) identified by linear regression with negative β-values for age. a Linear model p-values (two-sided Wald test, unadjusted) and β-values, and RNA characteristics obtained from literature review, are shown. RNAs that failed to have at least ten samples with CPM values above −3 were removed. b Linear plots representing log2(CPM) values with age for the top ten RNAs ranked by the lowest p-value (two-sided Wald test, unadjusted)97–108.CC BY · PubMed Central
Fig. 4 Validation of RNA (ENSG) expression patterns with age using RT-qPCR analysis. Relative expression levels of five young (20–34 years) and five old (65–79 years) participants were calculated relative to GAPDH mRNA levels, and were compared using a two-sided Kruskal–Wallis test. Of the eight RNAs selected for analysis, statistically significant changes in expression with age were identified for FAM83B (p = 0.0472), LGI1 (p = 0.0283), NR2F2 (p = 0.0278), and C12orf75 (p = 0.0163) mRNAs, denoted by a red asterisk. All p-values are unadjusted. Error bars represent mean ± SEM.
Fig. 4 Validation of RNA (ENSG) expression patterns with age using RT-qPCR analysis. Relative expression levels of five young (20–34 years) and five old (65–79 years) participants were calculated relative to GAPDH mRNA levels, and were compared using a two-sided Kruskal–Wallis test. Of the eight RNAs selected for analysis, statistically significant changes in expression with age were identified for FAM83B (p = 0.0472), LGI1 (p = 0.0283), NR2F2 (p = 0.0278), and C12orf75 (p = 0.0163) mRNAs, denoted by a red asterisk. All p-values are unadjusted. Error bars represent mean ± SEM.CC BY · PubMed Central
Fig. 5 Significant (p < 0.05) downstream protein abundance and log2(CPM) correlations. a Distribution of 4281 protein–RNA correlation coefficients. b Scatterplot of the age β-coefficients for the protein–RNA matches limited to the 122 mRNAs significantly associated with age (p-value from two-sided Wald test, unadjusted). c Heat map of 122 mRNAs significantly associated with age and their correlations with proteins. The mRNA with the strongest negative correlation (H2AZ) is shown in the inset (p-value from two-sided Wald test, unadjusted). d Among all the RNAs in the four top 20 lists, ten RNAs were identified to have statistically significant correlations between relative protein abundance and age (red and blue points indicate mRNAs with positive and negative correlations, respectively). Of these ten RNAs, nine had statistically significant correlations between relative protein [log2(Protein Abundance)] and RNA abundance [log2(CPM)] (p-values from two-sided Wald test, unadjusted).
Fig. 5 Significant (p < 0.05) downstream protein abundance and log2(CPM) correlations. a Distribution of 4281 protein–RNA correlation coefficients. b Scatterplot of the age β-coefficients for the protein–RNA matches limited to the 122 mRNAs significantly associated with age (p-value from two-sided Wald test, unadjusted). c Heat map of 122 mRNAs significantly associated with age and their correlations with proteins. The mRNA with the strongest negative correlation (H2AZ) is shown in the inset (p-value from two-sided Wald test, unadjusted). d Among all the RNAs in the four top 20 lists, ten RNAs were identified to have statistically significant correlations between relative protein abundance and age (red and blue points indicate mRNAs with positive and negative correlations, respectively). Of these ten RNAs, nine had statistically significant correlations between relative protein [log2(Protein Abundance)] and RNA abundance [log2(CPM)] (p-values from two-sided Wald test, unadjusted).CC BY · PubMed Central
Fig. 6 Differential transcript usage analysis for top RNAs (ENSGs). Among the RNAs in the four top 20 lists, CFAP61, ESRRG, TET2, and PLAG1 mRNAs were identified to have splice variants with statistically significant changes in differential transcript usage with age. a Change in differential transcript usage for all significant variants of the four mRNAs (all p < 0.05 from two-sided, unadjusted Wald tests). b Change in differential transcript usage for all nonsignificant variants of the four RNAs, provided for comparison (all p > 0.05 from two-sided, unadjusted Wald tests).
Fig. 6 Differential transcript usage analysis for top RNAs (ENSGs). Among the RNAs in the four top 20 lists, CFAP61, ESRRG, TET2, and PLAG1 mRNAs were identified to have splice variants with statistically significant changes in differential transcript usage with age. a Change in differential transcript usage for all significant variants of the four mRNAs (all p < 0.05 from two-sided, unadjusted Wald tests). b Change in differential transcript usage for all nonsignificant variants of the four RNAs, provided for comparison (all p > 0.05 from two-sided, unadjusted Wald tests).CC BY · PubMed Central
Fig. 7 Changes in the differential exonic usage of SMIM11A mRNA in young (20–34 years, n = 12) compared to old (80+ years, n = 5) age groups. a Red and blue bars represent the average exon usage for all 17 exons of SMIM11A mRNA in the young and old groups, respectively. DEXSeq analysis identified one exonic region (E011) that significantly changed in abundance between the age groups (pink in the gene model). The average exonic usage in this region is higher in the young age group and lower in the old age group. Of the seven variants of SMIM11A mRNA, two have statistically significant changes in expression continuously with age according to our linear model. Linear model p-values (two-sided Wald tests, unadjusted) are displayed for these variants. The horizontal axis below the variant models denotes the positions in the genome. b Ages of participants in the young (red) and old (blue) age groups are shown on the vertical axis. UCSC browser tracks indicate that the exons displaying differential expression (location on each transcript identified by red bar) identified in our DEXSeq analysis are more expressed in the young age group than in the old age group, indicated by the presence of larger blue peaks in younger participants. UCSC browser tracks support our DEXSeq findings.
Fig. 7 Changes in the differential exonic usage of SMIM11A mRNA in young (20–34 years, n = 12) compared to old (80+ years, n = 5) age groups. a Red and blue bars represent the average exon usage for all 17 exons of SMIM11A mRNA in the young and old groups, respectively. DEXSeq analysis identified one exonic region (E011) that significantly changed in abundance between the age groups (pink in the gene model). The average exonic usage in this region is higher in the young age group and lower in the old age group. Of the seven variants of SMIM11A mRNA, two have statistically significant changes in expression continuously with age according to our linear model. Linear model p-values (two-sided Wald tests, unadjusted) are displayed for these variants. The horizontal axis below the variant models denotes the positions in the genome. b Ages of participants in the young (red) and old (blue) age groups are shown on the vertical axis. UCSC browser tracks indicate that the exons displaying differential expression (location on each transcript identified by red bar) identified in our DEXSeq analysis are more expressed in the young age group than in the old age group, indicated by the presence of larger blue peaks in younger participants. UCSC browser tracks support our DEXSeq findings.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

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.

The self-check needs the paper’s full text.

Citations

Cites

Show all 107 references

Cited by

Sorted by citations

Items 101–125 of 125

Related

See all

Metadata

Authors
Tumasian, R. A., Harish, A., Kundu, G., Yang, J.-H., Ubaida-Mohien, C., Gonzalez-Freire, M., Kaileh, M., Zukley, L. M., Chia, C. W., Lyashkov, A., Wood, W. H., Piao, Y., Coletta, C., Ding, J., Gorospe, M., Sen, R., De, S., & Ferrucci, L.
Affiliations
National Institutes of Health (US)
Corresponding author
Ferrucci, Luigi
ORCID
Venue
Nature Communications
Publisher
Springer Science and Business Media LLC
Published
2021-04-01
Volume
12
Issue
1
Pages
2014-2014
ISSN
2041-1723 (ISSN-L)
DOI
10.1038/s41467-021-22168-2
PMC
PMC8016876
OpenAlex
W3142606313
Cited by
126
References
107
Citations by year
2026: 24 · 2025: 26 · 2024: 26 · 2023: 22 · 2022: 23 · 2021: 5
FWCI
5.96
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
U.S. Department of Health & Human Services | National Institutes of Health (Z01-AG000996)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Muscle Physiology and Disorders
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Aging › Genetics, Aging, and Longevity in Model Organisms
  • Health Sciences › Medicine › Physiology (medical) › Nutrition and Health in Aging
Keywords
skeletal muscle, healthy aging, transcriptome, gene expression, age-related changes, cellular senescence, insulin signaling, myogenesis, mRNA isoforms, oxidative phosphorylation, adipogenesis, muscle damage, muscle homeostasis, non-coding RNA
MeSH terms
Healthy Aging, Adult, Aged, Aged, 80 and over, Female, Homeostasis, Humans, Male, Middle Aged, Muscular Diseases, RNA, Messenger, Muscle, Skeletal, Gene Expression Profiling, RNA, Untranslated, Young Adult, RNA Isoforms, Transcriptome

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.

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