🧬 Aging Wiki

Survey of senescent cell markers with age in human tissues

9 min read

⚠️ Auto-extracted by Claude on 2026-06-25. Full text read from PMC OA version (PMC7093180). Quantitative values sourced from the PMC full text — figures are approximate bar-chart readings from the text descriptions. Verify percentages against Figures 1–5 in the primary PDF before relying on exact values. Wikilinks use existing wiki pages; foundational references (Krishnamurthy 2004, Ressler 2006) are cited by DOI pending their own study pages.

Idda et al. 2020 — Survey of Senescent Cell Markers with Age in Human Tissues

Open-access primary study. doi:10.18632/aging.102903 · PMID 32160592 · PMC7093180. Gorospe laboratory, National Institute on Aging, NIH. Published Aging (Albany NY) 2020; 12(5):4052–4066.

This is one of the most comprehensive cross-sectional immunohistochemical surveys of senescent cell accumulation across human organs, directly confirming the age-dependence of p16INK4a and p21CIP1 in human tissue rather than inferring it from model organisms.

TL;DR

Idda et al. surveyed p16INK4a and p21CIP1 immunostaining in 10 human organs across three age groups (Young 13–35 yr, Middle 40–59 yr, Old >65 yr; n=5 donors per organ per group) using commercial formalin-fixed, paraffin-embedded (FFPE) tissue arrays. They found “unexpectedly distinct patterns” — no single tissue behaved the same way, and the two markers were frequently discordant. The endocrine pancreas showed the highest p16 burden in elderly donors (~35% positive cells). Lung and skeletal/cardiac muscle showed no age-related increase in either marker. The study establishes that senescent cell accumulation in humans is tissue-specific and marker-specific, cautioning against the assumption that p16 and p21 are universally co-expressed surrogates.

Study Design

Tissue arrays

Commercially sourced, custom-designed human normal tissue arrays (FFPE). Tissues obtained from 10 organs, each with 5 independent tissue cores per age group. Organs:

  • Pancreas (exocrine vs. endocrine islets assessed separately)
  • Skin (epidermis vs. dermis assessed separately)
  • Kidney
  • Liver
  • Intestine (colon)
  • Spleen
  • Brain cortex
  • Lung
  • Skeletal muscle
  • Cardiac muscle

Age groups

GroupAge rangen per organ
Young13–35 yr5 independent donors
Middle40–59 yr5 independent donors
Old>65 yr5 independent donors

Markers and quantification

Immunohistochemistry using antibodies against p16INK4a (CDKN2A) and p21CIP1 (CDKN1A). Positive cells were counted within a defined area and divided by total cells in the same area to yield a percentage of positive cells. Statistics: one-way ANOVA with Tukey adjustment for multiple comparisons. Significance thresholds: p<0.05 (*), p<0.01 (**).

Key technical limitation: SA-β-galactosidase (SA-β-gal), the original and widely-used senescence biomarker (Dimri 1995 1), is enzymatically inert in FFPE tissue and could not be used; the study is therefore limited to the two CDK-inhibitor markers and cannot confirm co-localization with the full canonical senescence signature 2.

Per-tissue Quantitative Results

Endocrine pancreas (islets of Langerhans) — highest p16 burden

  • p16: ~1.5% in Young rising to approximately 35% in Old donors — a large magnitude increase; the authors report a 1.96-fold increase in the fraction of p16-positive cells, though absolute % reaching ~35% in the oldest group is the headline finding
  • p21: Minimal increase with age; p21 does not co-accumulate with p16 in islets

This is the tissue with the most striking p16 accumulation. Endocrine pancreatic senescence has been independently linked to age-related beta-cell dysfunction and reduced insulin secretory reserve 3. gap/needs-replication — the 5-donor-per-group design limits confidence in the ~35% estimate.

Skin — divergent compartment-specific patterns

Epidermis

  • p16: Young ~0.2% → Old ~4% (~21-fold increase, p<0.01)
  • p21: Young ~1.25% → Old ~6% (~5-fold increase)

Both markers increase significantly in the epidermis. Epidermal p16 accumulation with age has been independently validated in human skin biopsies by Ressler 2006 using a larger n 4.

Dermis

  • p16: Undetectable in all age groups — no signal
  • p21: Young ~1% → Old ~7.7% (~7.7-fold increase)

The dermis shows a p21-only pattern — p16 is absent but p21 rises substantially. This compartment-level divergence within a single organ illustrates that the two markers are not co-regulated in all cell types.

Kidney

  • p16: Young ~0.03% → Old ~0.2% (~7-fold increase, p<0.05)
  • p21: Young ~0.1% → Old ~1% (~11-fold increase, p<0.05)

Both markers increase, though at low absolute percentages. The proportional increase is large despite the small absolute numbers.

Liver

  • p16: Modest but statistically significant increase with age (~2-fold); absolute levels remain <1% in all groups
  • p21: <1% throughout; minimal change

p16-predominant pattern. The low absolute burden is notable given the liver’s high regenerative capacity.

Intestine (colon)

  • p16: Increased with age (~1.7-fold), primarily in lamina propria cells; low absolute percentages
  • p21: Minimal increase (~1.7-fold), low absolute percentages

Both markers modestly elevated but at very low absolute levels.

Spleen

  • p16: Young <0.2% → Old ~0.8% (~4-fold increase)
  • p21: No positive cells detected in any age group

p16-only pattern. No p21 immunoreactivity observed in splenic tissue.

Brain cortex

  • p16: Glial cells show an increase with age (~3-fold, Young ~0.3% → Old ~1%); neuronal staining not clearly separated
  • p21: No positive cells detected

p16-only pattern in glial cells. Neurons did not show detectable p16 or p21 staining, consistent with their post-mitotic identity (post-mitotic cells do not require CDK-inhibitor-mediated cell-cycle arrest in the conventional sense).

Lung

  • p16: Present across all age groups (~0.5%) but did not increase with age
  • p21: Present but also did not increase with age; some nonspecific background noted

Lung is the sole organ with no age-related change in either marker, despite being a tissue with well-documented age-related functional decline. The authors note this is unexpected. gap/no-mechanism — whether lung senescence operates through markers not surveyed here (e.g., SA-β-gal, γH2AX, lamin B1 loss) is unknown.

Skeletal muscle and cardiac muscle

  • p16: Not detected in skeletal or cardiac muscle in any age group
  • p21: Not detected in skeletal or cardiac muscle in any age group

Both muscle types showed no detectable p16 or p21 immunoreactivity. The authors interpret this to mean either (1) canonical CDK-inhibitor-mediated senescence does not accumulate in muscle with normal aging, or (2) these markers are inappropriate for muscle-resident senescent cells. Notably, satellite cells (the muscle stem cell population implicated in sarcopenia) represent a small fraction of total muscle cells and may be below detection limits at n=5 per group. gap/no-mechanism gap/needs-replication

Summary Table

Organp16 with agep21 with agePattern
Endocrine pancreas (islets)Large increase (~35% in Old)Minimalp16-predominant
Exocrine pancreasModest increase (<0.5% in Old)Modest increase (~2%)Both, low burden
Skin epidermisLarge increase (~4% in Old, ~21-fold)Large increase (~6%, ~5-fold)Both
Skin dermisUndetectableLarge increase (~7.7%, ~8-fold)p21-only
KidneyModest increase (~0.2% in Old, ~7-fold)Modest increase (~1%, ~11-fold)Both, low burden
LiverModest increase (<1%, ~2-fold)Minimalp16-predominant
Intestine (colon)Modest increase (~1.7-fold)Minimalp16-predominant
SpleenModerate increase (~0.8%, ~4-fold)Undetectablep16-only
Brain cortex (glial)Moderate increase (~1%, ~3-fold)Undetectablep16-only
LungNo changeNo changeNeither
Skeletal muscleNot detectedNot detectedNeither
Cardiac muscleNot detectedNot detectedNeither

Key Conclusions

  1. Tissue-specific marker patterns. Senescent cell accumulation is not uniform — p16 and p21 show distinct, organ-dependent trajectories. The assumption that both markers co-accumulate everywhere is incorrect.
  2. p16 is the more broadly informative marker across the tissues surveyed; however, p21 provides complementary signal particularly in the skin dermis where p16 is absent.
  3. Post-mitotic and rapidly self-renewing tissues (muscle, lung) showed the least evidence of canonical CDK-inhibitor-positive senescence, raising the question of whether different markers are needed for different tissue types.
  4. The endocrine pancreas stands out as a tissue with unusually high p16 burden in elderly humans — a finding with potential mechanistic relevance for age-related diabetes risk.

Significance as Human Evidence

This paper serves as direct human-tissue confirmation that senescent cell accumulation increases with age across multiple organs in humans. It complements:

  • Dimri 1995 1 — established SA-β-gal as a senescence marker; first cross-sectional human skin data
  • Ressler 2006 4 — p16INK4a in human skin biopsies across age; larger n validation of skin epidermal p16 accumulation
  • Krishnamurthy 2004 3 — p16INK4a in multiple rodent tissues and first human pancreas data
  • Intervention studies (Hickson 2019 dasatinib+quercetin; Justice 2019) that measure p16/p21 reduction in human adipose and skin — those studies rely on the assumption, now confirmed cross-sectionally here, that these markers are genuinely elevated at baseline in aged humans

Limitations

  • Small n per group (n=5 donors). Underpowered for organ-level subgroup comparisons; the ~35% p16 estimate in aged endocrine pancreas, while striking, carries wide uncertainty at n=5. gap/needs-replication
  • Two markers only. SA-β-gal is not measurable in FFPE, and γH2AX, lamin B1 loss, and SASP cytokine staining are not included. The study cannot confirm that p16+/p21+ cells are bona fide senescent by multi-marker criteria. Current consensus recommends multi-marker assessment 2.
  • Cross-sectional design. Different donors in each age group; inter-individual biological variability may confound the age signal at small n.
  • Donor heterogeneity. Medical history, cause of death, tissue handling, and fixation duration vary across array donors; not all are documented.
  • Binary IHC threshold. Positive/negative calling on IHC may miss cells with low but biologically relevant p16 or p21 expression.
  • Limited organ coverage. 10 organs from a universe of >70 tissue types; thymus, bone marrow, aorta, adipose, and reproductive tissues — all with independent evidence of senescent cell accumulation — are absent.
  • No co-staining. The two markers were assessed sequentially on serial sections or the same section, not simultaneously; co-localization at the single-cell level is inferred, not confirmed.

Cross-references

  • cellular-senescence — primary atomic page; this study provides direct human accumulation evidence cited in the senescence burden section
  • p16-rb-pathway — p16INK4a biology; this study confirms its in-vivo human age-dependent expression across tissues
  • p21 — p21CIP1 biology; this study’s dermis-p21-only finding is a key in-vivo human data point
  • sasp — SASP not directly measured here; the p16/p21+ cells are presumed to secrete SASP but this is not confirmed in this study
  • cellular-senescence — hallmark page; this study is the cross-sectional human-tissue anchor for the hallmark’s evidence base
  • cherqui-2025-senescence-burden-organs — more recent mouse + human PBMC study using multi-marker single-cell quantification; complementary scope

Footnotes

Footnotes

  1. doi:10.1073/pnas.92.20.9363 · Dimri GP et al. · PNAS 1995 · observational · model: human fibroblasts + skin biopsies · original establishment of SA-β-gal as senescence biomarker; includes cross-sectional human skin data ↩ ↩2

  2. idda-2020-senescent-markers-human-tissues · n=5 donors/organ/age-group · observational (cross-sectional IHC) · model: human FFPE tissue arrays; 10 organs; 3 age groups (Young 13–35 / Middle 40–59 / Old >65 yr) · one-way ANOVA with Tukey correction · gold OA, PMC7093180 ↩ ↩2

  3. krishnamurthy-2004-ink4a-arf-aging-biomarker · doi:10.1172/JCI22475 · PMID 15520862 · Krishnamurthy J et al. · J Clin Invest 2004;114(9):1299–1307 · in-vivo (rodent multi-tissue) · model: mouse/rat multi-organ + compiled human literature · Ink4a/Arf locus expression as biomarker of aging across tissues (~9.7-fold geometric-mean p16 rise); caloric restriction attenuates the rise; first systematic rodent age-stratified multi-organ survey ↩ ↩2

  4. ressler-2006-p16-human-skin-biomarker · doi:10.1111/j.1474-9726.2006.00231.x · PMID 16911562 · Ressler S et al. · Aging Cell 2006;5(5):379–389 · observational · model: human skin biopsies stratified by age · p16INK4a as robust in-vivo biomarker of cellular aging in human skin; epidermal + dermal p16 accumulation with age, BMI1 downregulation ↩ ↩2

Graph View

Backlinks