. 2023 Dec;624(7990):164-172.

doi: 10.1038/s41586-023-06802-1. Epub 2023 Dec 6.

Organ aging signatures in the plasma proteome track health and disease

Hamilton Se-Hwee Oh^#^{1

2

3}, Jarod Rutledge^#^{2

3

4}, Daniel Nachun⁵, Róbert Pálovics^{2

3

6}, Olamide Abiose^{3

6}, Patricia Moran-Losada^{2

3

6}, Divya Channappa^{2

3

6}, Deniz Yagmur Urey^{2

7}, Kate Kim^{2

3

6}, Yun Ju Sung^{8

9}, Lihua Wang^{8

9}, Jigyasha Timsina^{8

9}, Dan Western^{8

9

10}, Menghan Liu^{8

9}, Pat Kohlfeld^{8

9}, John Budde^{8

9}, Edward N Wilson^{3

6}, Yann Guen^{6

11}, Taylor M Maurer⁵, Michael Haney^{2

3

6}, Andrew C Yang^{12

13

14}, Zihuai He⁶, Michael D Greicius⁶, Katrin I Andreasson^{3

6

15}, Sanish Sathyan¹⁶, Erica F Weiss¹⁷, Sofiya Milman¹⁶, Nir Barzilai¹⁶, Carlos Cruchaga^{8

9}, Anthony D Wagner^{3

18}, Elizabeth Mormino⁶, Benoit Lehallier⁶, Victor W Henderson^{3

6

19}, Frank M Longo^{3

6}, Stephen B Montgomery^{5

20

21}, Tony Wyss-Coray^{22

23

24}

Affiliations

¹ Graduate Program in Stem Cell and Regenerative Medicine, Stanford University, Stanford, CA, USA.
² The Phil and Penny Knight Initiative for Brain Resilience, Stanford University, Stanford, CA, USA.
³ Wu Tsai Neurosciences Institute, Stanford University, Stanford, CA, USA.
⁴ Graduate Program in Genetics, Stanford University, Stanford, CA, USA.
⁵ Department of Pathology, Stanford University School of Medicine, Stanford, CA, USA.
⁶ Department of Neurology and Neurological Sciences, Stanford University School of Medicine, Stanford, CA, USA.
⁷ Department of Bioengineering, Stanford University School of Engineering, Stanford, CA, USA.
⁸ Department of Psychiatry, Washington University in St Louis, St Louis, MO, USA.
⁹ NeuroGenomics and Informatics Center, Washington University School of Medicine, St. Louis, MO, USA.
¹⁰ Division of Biology and Biomedical Sciences, Washington University School of Medicine, St. Louis, MO, USA.
¹¹ Quantitative Sciences Unit, Department of Medicine, Stanford University School of Medicine, Stanford, CA, USA.
¹² Departments of Neurology and Anatomy, University of California San Francisco, San Francisco, CA, USA.
¹³ Gladstone Institute of Neurological Disease, Gladstone Institutes, San Francisco, CA, USA.
¹⁴ Bakar Aging Research Institute, University of California San Francisco, San Francisco, CA, USA.
¹⁵ Chan Zuckerberg Biohub, San Francisco, CA, USA.
¹⁶ Departments of Medicine and Genetics, Institute for Aging Research, Albert Einstein College of Medicine, New York, NY, USA.
¹⁷ Department of Neurology, Montefiore Medical Center, New York, NY, USA.
¹⁸ Department of Psychology, Stanford University, Stanford, CA, USA.
¹⁹ Department of Epidemiology and Population Health, Stanford University, Stanford, CA, USA.
²⁰ Department of Genetics, Stanford University School of Medicine, Stanford, CA, USA.
²¹ Department of Biomedical Data Science, Stanford University School of Medicine, Stanford, CA, USA.
²² The Phil and Penny Knight Initiative for Brain Resilience, Stanford University, Stanford, CA, USA. twc@stanford.edu.
²³ Wu Tsai Neurosciences Institute, Stanford University, Stanford, CA, USA. twc@stanford.edu.
²⁴ Department of Neurology and Neurological Sciences, Stanford University School of Medicine, Stanford, CA, USA. twc@stanford.edu.

^# Contributed equally.

PMID: 38057571
PMCID: PMC10700136
DOI: 10.1038/s41586-023-06802-1

Organ aging signatures in the plasma proteome track health and disease

Hamilton Se-Hwee Oh et al. Nature. 2023 Dec.

. 2023 Dec;624(7990):164-172.

doi: 10.1038/s41586-023-06802-1. Epub 2023 Dec 6.

Authors

Affiliations

¹ Graduate Program in Stem Cell and Regenerative Medicine, Stanford University, Stanford, CA, USA.
² The Phil and Penny Knight Initiative for Brain Resilience, Stanford University, Stanford, CA, USA.
³ Wu Tsai Neurosciences Institute, Stanford University, Stanford, CA, USA.
⁴ Graduate Program in Genetics, Stanford University, Stanford, CA, USA.
⁵ Department of Pathology, Stanford University School of Medicine, Stanford, CA, USA.
⁶ Department of Neurology and Neurological Sciences, Stanford University School of Medicine, Stanford, CA, USA.
⁷ Department of Bioengineering, Stanford University School of Engineering, Stanford, CA, USA.
⁸ Department of Psychiatry, Washington University in St Louis, St Louis, MO, USA.
⁹ NeuroGenomics and Informatics Center, Washington University School of Medicine, St. Louis, MO, USA.
¹⁰ Division of Biology and Biomedical Sciences, Washington University School of Medicine, St. Louis, MO, USA.
¹¹ Quantitative Sciences Unit, Department of Medicine, Stanford University School of Medicine, Stanford, CA, USA.
¹² Departments of Neurology and Anatomy, University of California San Francisco, San Francisco, CA, USA.
¹³ Gladstone Institute of Neurological Disease, Gladstone Institutes, San Francisco, CA, USA.
¹⁴ Bakar Aging Research Institute, University of California San Francisco, San Francisco, CA, USA.
¹⁵ Chan Zuckerberg Biohub, San Francisco, CA, USA.
¹⁶ Departments of Medicine and Genetics, Institute for Aging Research, Albert Einstein College of Medicine, New York, NY, USA.
¹⁷ Department of Neurology, Montefiore Medical Center, New York, NY, USA.
¹⁸ Department of Psychology, Stanford University, Stanford, CA, USA.
¹⁹ Department of Epidemiology and Population Health, Stanford University, Stanford, CA, USA.
²⁰ Department of Genetics, Stanford University School of Medicine, Stanford, CA, USA.
²¹ Department of Biomedical Data Science, Stanford University School of Medicine, Stanford, CA, USA.
²² The Phil and Penny Knight Initiative for Brain Resilience, Stanford University, Stanford, CA, USA. twc@stanford.edu.
²³ Wu Tsai Neurosciences Institute, Stanford University, Stanford, CA, USA. twc@stanford.edu.
²⁴ Department of Neurology and Neurological Sciences, Stanford University School of Medicine, Stanford, CA, USA. twc@stanford.edu.

^# Contributed equally.

PMID: 38057571
PMCID: PMC10700136
DOI: 10.1038/s41586-023-06802-1

Abstract

Animal studies show aging varies between individuals as well as between organs within an individual^1-4, but whether this is true in humans and its effect on age-related diseases is unknown. We utilized levels of human blood plasma proteins originating from specific organs to measure organ-specific aging differences in living individuals. Using machine learning models, we analysed aging in 11 major organs and estimated organ age reproducibly in five independent cohorts encompassing 5,676 adults across the human lifespan. We discovered nearly 20% of the population show strongly accelerated age in one organ and 1.7% are multi-organ agers. Accelerated organ aging confers 20-50% higher mortality risk, and organ-specific diseases relate to faster aging of those organs. We find individuals with accelerated heart aging have a 250% increased heart failure risk and accelerated brain and vascular aging predict Alzheimer's disease (AD) progression independently from and as strongly as plasma pTau-181 (ref. ⁵), the current best blood-based biomarker for AD. Our models link vascular calcification, extracellular matrix alterations and synaptic protein shedding to early cognitive decline. We introduce a simple and interpretable method to study organ aging using plasma proteomics data, predicting diseases and aging effects.

PubMed Disclaimer

Conflict of interest statement

T.W-C., H.O., J.R., B.L. and Stanford University have filed a patent application related to this work, PCT/US2023/027896. T.W-C., H.O. and J.R. are co-founders and scientific advisors of Teal Omics Inc. and have received equity stakes. T.W.-C. is a co-founder and scientific advisor of Alkahest Inc. and Qinotto Inc. and has received equity stakes in these companies. C.C. has received research support from GSK and EISAI. The funders of the study had no role in the collection, analysis or interpretation of data; in the writing of the report; nor in the decision to submit the paper for publication. C.C. is a member of the advisory board of Vivid Genomics and Circular Genomics and owns stocks in these companies. S.B.M is a consultant for BioMarin, MyOme and Tenaya Therapeutics. All other authors have certified they have no competing interests to declare.

Figures

**Fig. 1. Plasma proteins can model organ aging.**
a, Study design to estimate organ-specific biological age. A gene was called organ-specific if its expression was four-fold higher in one organ compared to any other organ in GTEX bulk organ RNA-seq. This annotation was then mapped to the plasma proteome. Mutually exclusive organ-specific protein sets were used to train bagged LASSO chronological age predictors with data from 1,398 healthy individuals in the Knight-ADRC cohort. An ‘organismal’ model, which used the nonorgan-specific (organ shared) proteins, and a ‘conventional’ model, which used all proteins regardless of specificity, were also trained. Models were tested in four independent cohorts: Covance (n = 1,029), LonGenity (n = 962), SAMS (n = 192) and Stanford-ADRC (n = 420); models were also tested in the AD patients in the Knight-ADRC cohort (n = 1,677). To test the validity of organ aging models, the age gap was associated with multiple measures of health and disease. An example age prediction (predicted versus chronological age) and an example age gap versus phenotype association (age gap versus phenotype, standard boxplot) are shown. b, Individuals (ID) with the same conventional age gap can have different organ age gap profiles. Three example participants are shown. Bar represents mean age gap across n = 13 age gaps. c, Pairwise correlation of organ age gaps from n = 3,774 healthy participants across all cohorts. Distribution of all pairwise correlations is shown in inset histogram, with dotted line median correlation. The control age gap was highly correlated with the organismal age gap (r = 0.98), the sole outlier in the inset distribution plot. d, Identification of extreme agers, defined by a two standard deviation increase or decrease in at least one age gap. A representative kidney ager, heart ager and multi-organ ager are shown. e, All extreme agers were identified (23% of all n = 5,676 individuals) and clustered after setting age gaps below an absolute z-score of 2 to 0. The mean age gaps for all organs in the kidney agers, heart agers and multi-organ agers clusters are shown.

**Fig. 2. Organ age predicts health and disease.**
a, A cross-cohort meta-analysis of the association (linear regression) between the kidney age gap and hypertension (with hypertension n = 1,566, without n = 1,561). False discovery rate (FDR) P value_meta = 4.05 × 10⁻⁴⁰, effect size_meta = 0.486. (Supplementary Table 10). b, As in a, kidney age gap versus diabetes (with diabetes n = 335, without n = 2,839). FDR P value_meta = 1.15 × 10⁻²⁴, effect size_meta = 0.604. c, As in a, heart age gap versus atrial fibrillation or pacemaker (with atrial fibrillation n = 239, without n = 2,936). FDR P value_meta = 5.32 × 10⁻²¹, effect size_meta = 0.657. d, As in a, but for heart age gap versus heart attack (with heart attack history n = 280, without n = 2,904). FDR P value_meta = 1.77 × 10⁻²⁰, effect size_meta = 0.615. e, All kidney aging model coefficients. x axis shows % of model instances in the bagged ensemble that include the protein. Size of bubbles is scaled by the absolute value of the mean model weight across model instances (absolute value of y axis) (Supplementary Table 7). f, Single-cell RNA expression of kidney aging model proteins. Mean normalized expression values shown. g, As in e, but for the heart aging model. h, Human heart single-cell RNA expression of heart. Mean normalized expression values shown. i, Cox proportional hazard regression analysis of the relationship between organ age gap and future congestive heart failure risk over 15 years of follow-up in the LonGenity cohort for those without heart failure history at baseline (n = 26 events in 812 individuals). FDR P value_Heart = 7.07 × 10⁻⁷, hazard ratio_Heart = 2.37. (Supplementary Table 11). j, Cox proportional hazard regression analysis of the relationship between organ age gap and future mortality risk, over 15 years of follow-up in the LonGenity cohort (n = 173 events in 864 individuals). FDR P value_Conventional = 2.27 × 10⁻¹⁰, hazard ratio_Conventional = 1.54. (Supplementary Table 12). All error bars represent 95% confidence intervals.

**Fig. 3. Brain aging in cognitive decline and AD.**
a, FIBA was used to test the contributions of brain aging proteins to associations between brain age gap and global clinical dementia rating (CDRGLOB) (y axis) or chronological age prediction accuracy (x axis). Permutation of some proteins reduced the brain age gap association with CDRGLOB (FIBA+), while permutation of others strengthened it (FIBA−). FIBA+ brain aging proteins were used to train a cognition-optimized brain aging model (CognitionBrain) from cognitively unimpaired individuals in Knight-ADRC. (Supplementary Table 15). FI, feature importance. b, CognitionBrain aging model. Age estimation in all cohorts (ii) and bootstrap aging model coefficients (ii). Size of bubbles is scaled by the absolute value of the mean model weight. (Supplementary Table 15). c, A cross-cohort meta-analysis of the association (linear regression) between the CognitionBrain age gap and AD diagnosis (with AD n = 1,441, without n = 2,052). P value_meta = 9.23 × 10⁻³⁶, effect size_meta = 0.448. (Supplementary Table 15). d, A multivariate cox proportional hazard model of future dementia progression risk over five years in Stanford-ADRC (n = 48 events in 325 individuals). P value_{CognitionBrain} = 8.95 × 10⁻³, hazard ratio_{CongitionBrain} = 1.57. e, Kaplan–Meier curve for the CPH model in f. Risk of dementia progression for different levels of CognitionBrain AgeGap and PlasmaPTau181 while all other covariates are held constant. Displayed hazard ratio is a first-order estimate of the combined hazard ratio. f, Human brain single-cell RNA expression of CognitionBrain aging proteins. Mean normalized expression values shown. Top model proteins and proteins in the GO:CC synapse pathway are highlighted. g, Changes with age and AD of top CognitionBrain proteins across tissues (plasma and brain) and molecular layers (protein, bulk RNAand single-cell RNA). Changes in plasma were assessed using linear models from the Stanford- and Knight- ADRC cohorts (n = 3,226 individuals). Statistics for brain tissue were pulled from refs. ^,. Proteins with significant changes across tissues shown. Asterisks represent FDR-adjusted P value thresholds: *q < 0.05; **q < 0.01; ***q < 0.001. All error bars represent 95% confidence intervals. NS, not significant.

**Fig. 4. Organ aging in cognitive decline and AD.**
a, CDRGLOB FIBA was applied to all organ aging models using the Knight-ADRC (K-ADRC) to understand body-wide contributions to brain aging phenotypes (Supplementary Table 15). b, Associations (linear regression) between AD and the CognitionArtery (P value_meta = 6.02 × 10⁻¹⁶), CognitionBrain (P value_meta = 9.23 × 10⁻³⁶), CognitionOrganismal (P value_meta = 2.03 × 10⁻²⁸) and CognitionPancreas (P value_meta = 1.11 × 10⁻²¹), age gaps replicated in the Stanford-ADRC (S-ADRC) (Supplementary Table 20). c, Associations (linear regression) between organ age gaps and a composite score of overall cognition in the LonGenity cohort (n = 888). P value_{CognitionOrganismal} = 9.58 × 10⁻⁸, P value_{CognitionBrain} = 4.24 × 10⁻⁷, P value_{CognitionArtery} = 2.46 × 10⁻³ and P value_{CognitionPancreas} = 4.8 × 10⁻³ (Supplementary Table 23). d, Cox proportional hazard regression analysis, organ age gap and risk of conversion from cognitively normal to cognitive impairment (CDR-Global 0 → > = 0.5) over 15 years follow-up in the Knight-ADRC (n = 226 events in 940 individuals). P value_{CognitionOrganismal} = 0.02, P value_{CognitionArtery} = 0.04, P value_{CognitionBrain} = 0.14 and P value_{CognitionPancreas} = 0.26 (Supplementary Table 24). e, Aging trajectories of top ten weighted model proteins in healthy individuals (n = 3,774) across the four study cohorts. Top CognitionOrganismal proteins change with age earliest and at the highest rate. f, Changes with age of top cognition-optimized aging model proteins in healthy individuals (n = 3,774) across the four study cohorts. Age effect and negative log₁₀ FDR-corrected P values from a linear model are shown. Size of bubbles is scaled by the absolute value of the average model weight (Supplementary Table 25). g, Left, human brain vasculature single-cell RNA expression of top five CognitionOrganismal aging proteins. Mean normalized expression values and fraction of cells expressing the genes are shown. Right, pericytes, smooth muscle cells (SMC) and fibroblasts are lost in AD. Asterisks represent P value thresholds from a two-tailed t-test: *P < 0.05; **P < 0.01. h, Model of age-related cellular degradation of the human brain vasculature reflected in the plasma proteome. i, StringDB protein–protein interaction network of CognitionArtery and interacting proteins (score ≥ 0.4), and related pathway enrichments (percent overlap between proteins and pathway gene sets). j, Model of age-related vascular calcification and extracellular matrix alterations reflected in the plasma proteome. All error bars represent 95% confidence intervals.

**Extended Data Fig. 1. Identification of organ-enriched plasma proteins.**
a, Plasma proteins for which the gene encoding the protein was expressed at least four-fold higher in one organ compared to any other organ were called “organ-enriched” in line with the definition proposed by the Human Protein Atlas. To calculate organ-level gene expression, the maximum expression of sub-tissues in the Gene Tissue Expression Atlas (GTEx) bulk RNA-seq database was used. An example of this tissue expression aggregation into organ expression CPLX1. (See ST2). b, Organ-wide expression for CPLX1. CPLX1 is expressed over four-fold higher in the brain compared to any other organ and is therefore defined as organ-enriched. c, Organ-level fold-change distribution of SomaScan plasma protein encoding genes. (See ST3). d, Organ-level expression of 843 organ-enriched plasma protein encoding genes. These 843 genes correspond to 893 plasma protein epitopes measured on the SomaScan assay. Some plasma proteins on the assay are quantified multiple times by different aptamers, which target different epitopes of the same protein. e, Top significantly enrichment biological pathways of brain-enriched plasma proteins.

**Extended Data Fig. 2. Aging model training and testing.**
a, A bagged ensemble of least absolute shrinkage and selection operator (LASSO) aging models was trained for each of 11 major organs using the mutually exclusive organ-enriched proteins identified as inputs. An “organismal” aging model using the 3907 organ-nonspecific proteins and a “conventional” aging model using all 4778 QC’ed proteins on the SomaScan assay were also trained. Models were trained from the 1,398 healthy individuals in the Knight-ADRC cohort. To reduce overfitting, the LASSO regularization parameter α was determined with bootstrap resampling by selecting sparser model α that provided 95% of maximum training set performance. An individual’s predicted age was defined as the average predicted age across all bootstrapped models. The entire model training scheme for a single example aging model is shown. b, Models were tested in four independent cohorts (Covance, LonGenity, Stanford-ADRC, SAMS). Age predictions from a single example aging model across test cohorts is shown.

**Extended Data Fig. 3. Aging model prediction and coefficients.**
**a-m**, Aging model age prediction (i), average coefficients across bootstraps (ii) and top 15 coefficients (**iii**) are shown for all aging models in alphabetical order. (See ST7).

**Extended Data Fig. 4. Extreme organ agers are widespread in the population.**
a, Extreme agers were defined as individuals with a 2-standard deviation increase or decrease in at least one age gap. 23% of the individuals (n = 5,676) across the four cohorts were identified as extreme agers. To visualize all extreme agers, age gaps were denoised by setting values below absolute z-score of 2 to zero. Denoised age gaps are shown in the heatmap. b, Extreme ageotypes were defined based on kmeans clustering of individuals based on their denoised age gaps. The mean z-scored age gap per ageotype is shown. c, The percentage of extreme agers is shown across all cohorts. d, A cross-cohort meta-analysis of associations (logistic regression) between extreme ageotypes versus diagnosis of 9 major age-related diseases annotated in at least 2 independent cohorts. Log odds ratios and significance are shown. P-values were Benjamini-Hochberg FDR-corrected. The strongest associations per disease are highlighted with black borders. (See ST9). e, A cross-cohort meta-analysis of associations (linear regression) between organ age gaps versus diagnosis of 9 major age-related diseases annotated in at least 2 independent cohorts. Disease covariate effects and significance are shown. P-values were Benjamini Hochberg FDR-corrected. The strongest associations per disease are highlighted with black borders. (See ST10). Asterisks represent q-value thresholds: *q < 0.05; **q < 0.01; ***q < 0.001.

**Extended Data Fig. 5. Plasma proteomic organ aging models versus established clinical markers of aging, health, and disease.**
a, Phenotypic Age (PhenoAge, Levine et al. 2018) was calculated based on 10 clinical markers in the Covance cohort (n = 1,026). PhenoAge-based age prediction is shown. b, The PhenoAge age gap was calculated and correlated with plasma proteomic organ aging model age gaps. Pairwise correlations are shown. c, Organ age gaps and the PhenoAge age gap were associated with 43 individual clinical markers of health and disease. Phenotype covariate effect sizes and significance based on Benjamini Hochberg FDR corrected p-values for all associations are shown. Asterisks represent q-value thresholds: *q < 0.05; **q < 0.01; ***q < 0.001. (See ST14).

**Extended Data Fig. 6. Feature Importance for Biological Aging (FIBA) to derive a cognition-associated brain aging model.**
a, Schematic for FIBA algorithm, (see methods) an algorithm to assess brain aging model protein contributions to the brain age gap association with cognition and chronological age prediction accuracy. FIBA+ brain aging model proteins were used to train a new cognition-optimized brain aging model (CognitionBrain) from healthy individuals in the Knight-ADRC cohort. b, A cross-cohort meta-analysis of the association (linear regression) between the CognitionBrain, Brain, and Conventional age gaps versus Alzheimer’s disease (with AD n = 1,441, without n = 2,052). CognitionBrain age gap p-value_meta = 9.23 × 10⁻³⁶, effect size_meta = 0.448; Brain age gap p-value_meta = 5.67 × 10⁻¹⁰, effect size_meta = 0.221; Conventional age gap p-value_meta = 1.33 × 10⁻¹³, effect size_meta = 0.270. (See ST10, ST20). c, CognitionBrain age gaps were associated with brain MRI volume in the Stanford-ADRC and SAMS cohorts (n = 469). CognitionBrain associations with individual brain region volumes shown. Bubbles are sized by FDR corrected p-value. (See ST22). d, Pairwise-correlations between the CognitionBrain age gap, plasma pTau-181, and AD polygenic risk score. All error bars represent 95% confidence intervals.

**Extended Data Fig. 7. Feature Importance for Biological Aging (FIBA) plots for all aging models in relation to cognition.**
a, FIBA was applied to all aging models to assess peripheral versus central contributions to brain aging and cognitive decline (CDR-Global dementia rating). For each aging model, proteins were assessed for their contributions to the age gap association with cognition (CDR-Global, y-axis) and chronological age prediction accuracy (x-axis). Proteins for which permutation reduces the age gap association with cognition were termed FIBA+ , while proteins for which permutation strengthens the age gap association with dementia were termed FIBA-. FIBA+ proteins were used to train new cognition-optimized aging models from healthy individuals in the Knight-ADRC cohort. FIBA results for all aging models are shown in alphabetical order. (See ST15).

**Extended Data Fig. 8. Cognition-optimized aging model associations with age and AD.**
a, FIBA+ proteins from each aging model were used to train new cognition-optimized aging models from healthy individuals in the Knight-ADRC cohort. Correlations between predicted vs chronological age in healthy individuals in the training (Knight-ADRC) and test (Covance, LonGenity, Stanford-ADRC, SAMS) cohorts for all aging models are shown. All aging models significantly estimated age across five independent cohorts. Cognition-optimized aging models predicted chronological age slightly worse than their non-optimized counterparts as expected, given the subsetting of proteins. (See ST19). b, Pairwise correlation of all model age gaps in all cohorts. Cognition-optimized aging models predicted similar age gap estimates with their non-optimized models. c, Model age gap associations (linear regression) with Alzheimer’s disease (with AD n = 1,393, control n = 1,680) in the Knight-ADRC cohort. Effect sizes, 95% confidence intervals, and p-values for the Alzheimer’s covariate are shown. Despite decreased associations with chronological age, cognition-optimized models showed substantially stronger associations with Alzheimer’s disease. (See ST20). d, As in c, but in the Stanford-ADRC cohort (with AD n = 48, control n = 372). (See ST20).

**Extended Data Fig. 9. Cognition-optimized aging model associations with cognitive function in non-cognitively impaired individuals.**
a, Associations (linear regression) between organ age gaps and a composite score of overall cognition in the LonGenity cohort (n = 888) shown. p_{CognitionOrganismal} = 9.58 × 10⁻⁸, p_{CognitionBrain} = 4.24 × 10⁻⁷, p_{CognitionArtery} = 2.46 × 10⁻³, p_{CognitionPancreas} = 4.8 × 10⁻³. (See ST23). b, Associations (linear regression) between organ age gaps and a memory test score in the SAMS cohort (n = 160) shown. p_{CognitionOrganismal} = 9.85 × 10⁻³, p_{CognitionBrain} = 2.44 × 10⁻², p_{CognitionArtery} = 0.53, p_{CognitionPancreas} = 0.29. (See ST23).

**Extended Data Fig. 10. Mapping CognitionOrganismal and CognitionArtery proteins to human organs and cell types.**
a, The organ sources of highly weighted CognitionOrganismal proteins were investigated by analyzing their expression levels in the Gene Tissue Expression Atlas (GTEx) bulk RNA-seq database. Organ-level expression of pleiotrophin (PTN), transgelin (TAGLN), WNT1 inducible signalling pathway protein 2 (WISP2), and chordin like 1 (CHRDL1) are shown. Though not organ-specific, these genes were highly expressed in the arteries and brain. b, Single-cell RNA expression (Tabula Sapiens) of highly weighted CognitionOrganismal proteins in human vasculature. Mean normalized expression values and fraction of cells expressing the genes are shown. c, Single-cell RNA expression (Tabula Sapiens) of highly weighted CognitionArtery and StringDB-based “interacting” proteins in human vasculature. Mean normalized expression values and fraction of cells expressing the genes are shown.

See this image and copyright information in PMC

Cited by

Adding intrinsically disordered proteins to biological ageing clocks.
Dormann D, Lemke EA. Dormann D, et al. Nat Cell Biol. 2024 Jun;26(6):851-858. doi: 10.1038/s41556-024-01423-w. Epub 2024 May 23. Nat Cell Biol. 2024. PMID: 38783141 Review.
Brain-Periphery Interactions in Huntington's Disease: Mediators and Lifestyle Interventions.
Burtscher J, Strasser B, Pepe G, Burtscher M, Kopp M, Di Pardo A, Maglione V, Khamoui AV. Burtscher J, et al. Int J Mol Sci. 2024 Apr 25;25(9):4696. doi: 10.3390/ijms25094696. Int J Mol Sci. 2024. PMID: 38731912 Free PMC article. Review.
Not every organ ticks the same.
Talbi K, Melk A. Talbi K, et al. Nat Rev Nephrol. 2024 Jul;20(7):431-432. doi: 10.1038/s41581-024-00842-0. Nat Rev Nephrol. 2024. PMID: 38658670 No abstract available.
Mitochondrial respiration atlas reveals differential changes in mitochondrial function across sex and age.
Sarver DC, Saqib M, Chen F, Wong GW. Sarver DC, et al. bioRxiv [Preprint]. 2024 Mar 29:2024.03.26.586781. doi: 10.1101/2024.03.26.586781. bioRxiv. 2024. PMID: 38586038 Free PMC article. Preprint.
Hungry for biomarkers of aging.
LeBrasseur NK. LeBrasseur NK. Aging Cell. 2024 Apr;23(4):e14158. doi: 10.1111/acel.14158. Epub 2024 Mar 27. Aging Cell. 2024. PMID: 38532727 Free PMC article. No abstract available.

See all "Cited by" articles

References

1. Schaum N, et al. Ageing hallmarks exhibit organ-specific temporal signatures. Nature. 2020;583:596–602. doi: 10.1038/s41586-020-2499-y. - DOI - PMC - PubMed
1. Almanzar N, et al. A single-cell transcriptomic atlas characterizes ageing tissues in the mouse. Nature. 2020;583:590–595. doi: 10.1038/s41586-020-2496-1. - DOI - PMC - PubMed
1. Pálovics R, et al. Molecular hallmarks of heterochronic parabiosis at single-cell resolution. Nature. 2022;603:309–314. doi: 10.1038/s41586-022-04461-2. - DOI - PMC - PubMed
1. Zahn JM, et al. AGEMAP: a gene expression database for aging in Mice. PLoS Genet. 2007;3:e201. doi: 10.1371/journal.pgen.0030201. - DOI - PMC - PubMed
1. Janelidze S, et al. Plasma P-tau181 in Alzheimer’s disease: relationship to other biomarkers, differential diagnosis, neuropathology and longitudinal progression to Alzheimer’s dementia. Nat. Med. 2020;26:379–386. doi: 10.1038/s41591-020-0755-1. - DOI - PubMed

MeSH terms

Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions

Save citation to file

Email citation

Add to Collections

Add to My Bibliography

Your saved search

Create a file for external citation management software

Your RSS Feed

Organ aging signatures in the plasma proteome track health and disease

Affiliations

Organ aging signatures in the plasma proteome track health and disease

Authors

Affiliations

Abstract

Conflict of interest statement

Figures

Similar articles

Cited by

References

MeSH terms

Substances

Grants and funding

LinkOut - more resources

Full Text Sources

Medical

Abstract

Conflict of interest statement

Figures

Similar articles

Cited by

References

MeSH terms

Substances

Related information

Grants and funding

LinkOut - more resources

Full Text Sources

Medical