The research focused on cellular senescence, a state of stable cell cycle arrest triggered by replicative exhaustion, DNA damage, or oxidative stress. Senescent cells are not dead. They remain metabolically active, resist apoptosis, and release a potent cocktail of cytokines, chemokines, growth factors, and matrix-remodeling enzymes known as the senescence-associated secretory phenotype, or SASP. This secretory program is widely believed to drive the chronic, low-grade, sterile inflammation called inflammaging, a hallmark of aging tissues. In the kidney, the cyclin-dependent kinase inhibitor p16Ink4a, encoded by the Cdkn2a gene, serves as one of the most reliable markers of senescence, and its ablation has previously been shown to improve repair after ischemic injury and extend graft survival.

To dissect the relationship between senescent cells and immune cells, the team combined quantitative histology, RNA in situ hybridization for Cdkn2a, NanoString-based transcriptomics of fibrosis and inflammation gene panels, and Visium spatial transcriptomics. They worked with young mice aged three to four months and old mice aged 22 to 24 months on a C57BL/6J background, and their first key observation was one of remarkable heterogeneity. Although Cdkn2a expression was significantly elevated in aged kidneys overall, individual old animals varied enormously. When the researchers stratified old mice into p16-low and p16-high groups, they found that the p16 signal, rather than chronological age, was the strongest predictor of the inflammatory transcriptome, associating with 60 significantly regulated genes involved in cytokine signaling, chemotaxis, and antigen presentation.

Immunostaining confirmed that aging remodels the kidney’s immune landscape. Old kidneys contained significantly more CD45-positive leukocytes, F4/80-positive macrophages, and CD3-positive T cells than young ones, and even rarer populations such as CD138-positive plasma cells and NKp46-positive natural killer cells increased with age. Activated macrophage phenotypes, both pro-inflammatory CD86-positive and pro-fibrotic CD206-positive cells, also rose. But when old kidneys were separated by senescence burden, macrophages stood out as the population most tightly linked to p16 levels, being significantly more abundant in p16-high kidneys, while T cells, plasma cells, and NK cells showed no significant differences between the groups.

The study’s most visually compelling finding came from its spatial analyses. Using senescence-associated beta-galactosidase staining merged with immunofluorescence, and separately using high-resolution RNAscope detection of Cdkn2a transcripts combined with immune markers, the researchers demonstrated that both leukocytes and macrophages were significantly enriched in immediate proximity to senescent tubular structures. The immune cells were not themselves senescent; they clustered around p16-positive tubules like moths around a lamp. Notably, p16 expression was predominantly localized to tubular epithelial cells, with interstitial p16-positive cells being rare and explicitly excluded from the colocalization analysis, addressing the known caveat that macrophages can express p16 independently of senescence.

Supplementary Visium spatial transcriptomics independently reinforced this picture. By integrating the team’s own pilot data with a public mouse kidney dataset, the analysis revealed a cortex-restricted transcriptomic cluster that combined senescence-associated genes such as Cdkn1a, Trp53, Serpine1, Tgfb1, and Ccl2 with immune-associated signatures and a prominent macrophage component. The same cluster also expressed markers of failed-repair proximal tubules, including Havcr1, which encodes kidney injury molecule 1, along with Vcam1 and C3. This convergence of senescence, inflammation, and maladaptive repair signaling within a single spatial neighborhood provides strong evidence that the aging kidney cortex hosts organized inflammatory niches rather than a diffuse, uniform inflammatory haze.

The researchers then asked whether these niches are modifiable. They applied two senolytic strategies: pharmacological clearance using the BCL-2/BCL-xL inhibitor ABT-737 in old mice aged 20 to 22 months, and genetic ablation in middle-aged INK-ATTAC mice, in which administration of AP20187 inducibly kills p16-expressing cells. Both interventions significantly reduced the Cdkn2a signal, and histological analysis showed decreased infiltration of both leukocytes and macrophages in treated kidneys. Critically, colocalization analysis revealed that in vehicle-treated controls, immune cells clustered strongly around senescent cells, whereas after ABT-737 treatment this spatial relationship largely dissolved for leukocytes, which became distributed more evenly throughout the tissue.

One nuance deserves emphasis. Even after senolysis, macrophages remained preferentially enriched near the residual senescent tubules, although their overall numbers fell. The authors propose several explanations: remaining senescent cells may continue to emit chemokine signals, particularly through the Ccl2-Ccr2 axis, which is known to recruit Ccr2-positive myeloid cells during senescence surveillance; macrophages may persist locally as tissue-resident cleaners of damaged cells; or altered immune cell turnover may shape the residual pattern. The p16-high kidneys in the cohort also showed increased Ccr2 expression together with macrophage-associated gene signatures, consistent with this recruitment model. Distinguishing between persistent recruitment, local retention, and altered turnover will require future work.

The findings carry broader implications for how scientists conceptualize biological aging. Because the p16-high and p16-low old mice were genetically identical and chronologically matched, the divergence in inflammatory profiles reflects differences in biological rather than chronological aging. This supports the idea that senescence burden is a more faithful indicator of renal inflammatory remodeling than age itself, echoing previous observations of variable senescence signatures in aging human kidneys. The study also suggests that the inflammatory microenvironments of renal aging are at least partially reversible, and that intervening at middle age, before senescence burden fully accumulates, can already blunt immune cell accumulation, hinting at a preventive window for senolytic therapies.

The authors are candid about limitations. Both senolytic approaches act systemically rather than specifically on renal tubular cells, so the observed reduction in inflammation cannot be attributed exclusively to clearing senescent tubules; ABT-737 also carries known hematologic off-target effects, which is why the genetic models were included as complementary evidence. The short observation period after treatment precludes conclusions about long-term durability, the bulk transcriptomic data cannot assign gene expression to individual cell types, and differences between murine and human immune systems may limit direct translation. Nevertheless, by mapping senescence and immunity with spatial precision, the study provides a framework in which senescent tubular structures serve as nucleation sites for inflammaging in the kidney, and it positions these senescence-associated inflammatory niches as concrete, targetable structures for future therapies against renal aging and chronic kidney disease.

Subject of Research: Spatial organization of senescent cell and immune cell interactions in the aging mouse kidney and their modulation by senolytic therapy

Article Title: Insights Into the Interplay Between the Senescent Cells and Immune Cells

Article References: Jaros, M., Schmidt, M., Neubert, L., Kamp, J.-C., von Vietinghoff, S., Bräsen, J. H., Schmitt, R., & Melk, A. (2026). Insights Into the Interplay Between the Senescent Cells and Immune Cells. Aging Cell, 25(9), Article e70698. https://doi.org/10.1111/acel.70698

Image Credits: AI Generated

DOI: 10.1111/acel.70698

Keywords: cellular senescence, kidney aging, p16Ink4a, macrophages, inflammaging, SASP, senolytics, ABT-737, INK-ATTAC, spatial transcriptomics, chronic kidney disease, renal cortex