🧬 Aging Wiki

Mitochondrial oxidative stress caused by Sod2 deficiency promotes cellular senescence and aging phenotypes in the skin

8 min read

Mitochondrial oxidative stress caused by Sod2 deficiency promotes cellular senescence and aging phenotypes in the skin

Velarde MC, Flynn JM, Day NU, Melov S, Campisi J — Aging (Albany NY) 2012;4(1):3–12

TL;DR

Mitochondrial superoxide dismutase (SOD2 / MnSOD) deficiency in mouse skin recapitulates key hallmarks of aged skin — p16^INK4a accumulation, nuclear DNA damage, impaired mitochondrial complex II activity, epidermal thinning, and increased terminal differentiation — without triggering apoptosis. Natural aging in C57BL/6J mice shows a parallel decline in mitochondrial complex II activity. In culture, mitochondrial ETC complex I inhibition (rotenone) induces senescence in human keratinocytes at lower doses than in dermal fibroblasts, placing the epidermis rather than the dermis as the primary site of mtROS-induced senescent-cell accumulation. Together the data argue that mitochondrial oxidative stress is a driver — not merely a correlate — of cellular senescence and skin aging phenotypes in vivo. Wrinkled appearance is not described in this paper and should not be attributed to it.

Background

cellular-senescence can be triggered by DNA damage, telomere shortening, oncogene activation, or oxidative stress. The mitochondrial-dysfunction hallmark posits that accumulation of mitochondrial reactive oxygen species (mtROS) drives progressive molecular damage during aging. However, direct in-vivo causal evidence linking mtROS to senescent-cell accumulation in a tissue context was limited at the time of publication.

sod2 (MnSOD) is the primary scavenger of superoxide generated at the inner mitochondrial membrane by complexes I and III. Homozygous deletion is lethal within days of birth 1, requiring neonatal rescue or tissue-specific inactivation for adult studies. This paper used a partial-rescue model (EUK-189, a synthetic SOD/catalase mimetic) to extend survival of Sod2-/- animals into early adulthood, enabling skin phenotyping.

The skin is a useful tissue for senescence biology: it is large, accessible, stratified (epidermis vs. dermis), and displays measurable aging phenotypes (thinning, altered differentiation) amenable to histological quantification. Predominant senescent cell types in this study are epidermal, not dermal fibroblasts — a point with implications for targeting strategies in skin aging.

Methods summary

In-vivo: Sod2-/- mouse model

  • Strain: CD1 background; homozygous Sod2 null (Sod2-/-)
  • Rescue: EUK-189 (synthetic SOD/catalase mimetic) at 1 mg/kg/day, administered from postnatal day 3 as maintenance throughout the experiment
  • Harvest age: 17–20 postnatal days
  • Groups for histology (SA-βgal, H&E, IHC): n=8 WT, n=9 KO (CD1)
  • Groups for SOD2-confirmation Western (Fig 2A): n=8 WT, n=9 KO (CD1)
  • Groups for senescence Westerns (p16^INK4a, ÎłH2AX, PARP/CASP3): n=6 WT, n=6 KO (CD1)
  • Clonogenicity assay: n=3 WT, n=3 KO
  • Endpoints: SA-β-gal activity (stratum corneum); p16^INK4a and p21^CIP1 protein (Western blot); ÎłH2AX foci (nuclear DNA damage); mitochondrial complex II (succinate dehydrogenase) activity; Ki-67 proliferation index; keratin 10 (terminal differentiation marker); epidermal thickness; stratum corneum thickness

In-vivo: natural aging cohort (C57BL/6J)

  • Strain: C57BL/6J; three age groups: n=6 at 4 months, n=10 at 8 months, n=13 at 24 months
  • Endpoints: SA-β-gal; p16^INK4a; mitochondrial complex II activity
  • These n’s must NOT be conflated with the Sod2-/- experiment groups above

In-vitro: rotenone-induced senescence

  • Cell lines: AG21837 (human neonatal keratinocytes); HCA2 (human foreskin fibroblasts)
  • Perturbation: rotenone (complex I inhibitor), 0–200 nM dose range; proliferation assessed at 4 days; SA-βgal assessed at 9 days; key comparison at 100 nM
  • Endpoint: SA-β-gal+ fraction (quadruplicate measurements); population doublings over 4 days
  • Statistical test: Student’s t test; all measurements in quadruplicates

In-vivo findings: Sod2-/- skin phenotype

Senescence markers elevated in Sod2-/- epidermis

Sod2-/- mouse skin showed two-fold higher p16^INK4a protein relative to WT littermates — this is the explicit text quote from the paper (“p16INK4a protein levels were two-fold higher in Sod2-/- relative to WT skin”) — Western blot, n=6/group 2. SA-β-gal activity was detectable in the stratum corneum of all 9 Sod2-/- mice examined (9/9 penetrance), whereas only minimal activity was observed in 1 of 8 WT mice (stratum corneum specificity confirmed; p<0.05 by Fisher Exact test). The senescent cells were localized to the epidermis, not the dermis — consistent with the greater rotenone sensitivity of keratinocytes demonstrated in vitro (see below).

p21^CIP1 protein was also elevated in Sod2-/- skin, though the paper’s primary CDK-inhibitor marker for the senescence phenotype is p16

Nuclear DNA damage without apoptosis

γH2AX foci (double-strand break marker) were increased in Sod2-/- epidermis, consistent with mtROS-driven nuclear DNA damage as an upstream trigger of the p16^INK4a induction. Crucially, the increase in p16^INK4a was not accompanied by elevated apoptosis — the cells are arrested/senescent rather than eliminated, directly paralleling the accumulation model of senescent cells in aged tissue 2.

Mitochondrial complex II activity impaired

Complex II (succinate dehydrogenase) activity was reduced in Sod2-/- skin relative to WT — establishing that mtROS impairs the ETC in this model, not merely correlates with it 2.

Epidermal structural changes

Histologically, Sod2-/- skin showed (n=8 WT / n=9 KO):

  • Reduced epidermal cell number and decreased epidermal thickness
  • Increased stratum corneum thickness (hyperkeratosis pattern)
  • Reduced Ki-67+ proliferating cells (consistent with senescence-mediated proliferative arrest)
  • Elevated keratin 10 (marker of terminal differentiation), suggesting premature commitment to differentiation at the expense of basal layer renewal

No “wrinkled appearance” is described in this paper. gap/unsourced if claimed elsewhere without independent citation.

Natural aging cohort findings (C57BL/6J)

In naturally aged C57BL/6J mice, both p16^INK4a accumulation and SA-β-gal positivity increased progressively with age (4 → 8 → 24 months; n=6, 10, 13 respectively) 2. In parallel, mitochondrial complex II activity declined with age in the same cohort, mirroring the complex-II impairment observed in Sod2-/- skin and implicating the same enzymatic node across models. This convergence between the genetic-deficiency model and natural aging is a key internal consistency argument of the paper.

DimensionStatus
Pathway conserved in humans?yes (SOD2, complex II, p16^INK4a all conserved)
Phenotype conserved in humans?partial — epidermal thinning and p16+ accumulation occur in aged human skin; direct mtROS-causal evidence lacking in vivo
Replicated in humans?no gap/needs-human-replication

In-vitro findings: keratinocyte vs. fibroblast rotenone sensitivity

Human neonatal keratinocytes (AG21837) treated with 100 nM rotenone for 9 days showed approximately ~80% SA-β-gal+ cells, versus approximately ~55% SA-β-gal+ for human dermal fibroblasts (HCA2) at the same dose — both significantly elevated over vehicle controls (Supplementary Fig 1B; quadruplicate measurements; p<0.05 by Student’s t test) 2. Both cell types showed significant reductions in population doubling at ≥20 nM rotenone over 4 days (Supp Fig 1A). This differential SA-βgal induction places keratinocytes as the more vulnerable cell type to complex-I-driven mtROS, consistent with the predominantly epidermal localization of senescent cells in the Sod2-/- in-vivo model.

Mechanistic model proposed

The paper’s proposed causal chain:

  1. Elevated superoxide at the inner mitochondrial membrane (whether from Sod2 loss or ETC dysfunction in natural aging)
  2. Oxidative damage to nuclear DNA → γH2AX foci
  3. p16^INK4a/CDK4/6-Rb axis activation → stable cell-cycle arrest
  4. Senescent cells accumulate in epidermis → reduced proliferative renewal → epidermal thinning + increased terminal differentiation

This connects the mitochondrial-dysfunction hallmark directly to the cellular-senescence hallmark via the dna-damage-response pathway, with the p16^INK4a → Rb arm as the effector. See also p53-pathway and p21 for the parallel p53/p21 arm of senescence enforcement.

The sasp is not analyzed in depth in this paper; the focus is on cell-intrinsic senescence markers.

Limitations

  1. Rescue model caveats. EUK-189 is not physiological; it may partially rescue non-SOD2-related phenotypes or incompletely rescue others, potentially confounding the Sod2-/- phenotype interpretation.
  2. CD1 background for KO, C57BL/6J for aging cohort. The two in-vivo arms use different genetic backgrounds. Effect sizes and senescence dynamics may differ by strain.
  3. No in-vivo human data. All causal evidence for mtROS → senescence is in mice or cell culture. Direct measurement of mtROS-induced p16+ accumulation in human dermis/epidermis in vivo is absent. gap/needs-human-replication
  4. SA-β-gal in stratum corneum. The stratum corneum is a terminally differentiated, anucleate layer; SA-β-gal positivity in this compartment requires careful interpretation relative to the granular/spinous layers where nucleated senescent cells would reside.
  5. Rotenone in-vitro vs. physiological mtROS. Rotenone inhibits complex I at pharmacological concentrations; it is unclear whether the dose used (100 nM) approximates the superoxide flux generated by Sod2 deficiency or natural aging in the same cells.
  6. n-sizes. Histology at n=8–9 and Western at n=6 per group are standard for this era and organism model; statistical power to detect modest effect sizes is limited.
  7. Skin aging scope. Skin aging involves multiple processes beyond mtROS and senescence (UV damage, collagen cross-linking, ECM remodeling). The paper explicitly addresses only the mtROS–senescence axis.

Significance

This paper was among the first to demonstrate in-vivo that mtROS drives cellular senescence accumulation in a specific tissue (skin epidermis), using a genetically tractable model with internal consistency to the natural-aging trajectory. It directly links two hallmarks — mitochondrial-dysfunction and cellular-senescence — with mechanistic evidence rather than correlation, and positions the epidermal keratinocyte as the primary cellular substrate for this process in skin.

The 276 citations (OpenAlex) and 100th citation-percentile (FWCI 12.8) reflect substantial downstream influence in the skin-aging, senescence, and mitochondrial-ROS literature.

For downstream signaling from senescent epidermal cells, see sasp and altered-intercellular-communication. For the broader mtROS-aging hypothesis context, see mitochondrial-dysfunction and free-radical-theory-of-aging.

Footnotes

Footnotes

  1. doi:10.1038/ng1295-376 · Li Y, Huang TT, Carlson EJ, Melov S, Ursell PC, Olson JL, Noble LJ, Yoshimura MP, Berger C, Chan PH, Wallace DC, Epstein CJ · Nat Genet 1995;11:376–381 · dilated cardiomyopathy and neonatal lethality in mutant mice lacking manganese superoxide dismutase; establishes necessity of SOD2 for early survival · in-vivo · model: Sod2-/- (strain not specified in Velarde 2012 citation) ↩

  2. velarde-2012-mitochondria-skin-senescence · Sod2-/- histology/SOD2-Western n=8 WT/9 KO; senescence Westerns (p16/γH2AX/PARP/CASP3) n=6 WT/6 KO; clonogenicity n=3/3; aging cohort n=6/10/13 at 4/8/24mo; in-vitro keratinocyte/fibroblast quadruplicates · in-vivo + in-vitro · model: CD1 Sod2-/- mouse (17–20 days postnatal) + C57BL/6J aging cohort + human AG21837 keratinocytes + HCA2 fibroblasts ↩ ↩2 ↩3 ↩4 ↩5

Graph View

Backlinks