Three-Cohort Centenarian Atlas (NK rejuvenation)

eBioMedicine · 2025. Raw sequencing is not on this host. What follows is the open paper layer: access route, figure captions with PMC originals, and 1–2 core tables transcribed from the publication.

Controlled access PMID 40945050 PMC12571583 CC BY-NC-ND 4.0

Access truth Lingao scRNA is GSA-Human HRA011570 (controlled). Rugao HRA002867 is also controlled. Japan RIKEN SC2018 and age-control GSE158055 are public reuse, not a local centenarian matrix.

Non-derivative licence. LACA shows captions and links to the publisher/PMC originals. Do not recolour or rehost these UMAPs as LACA figures.

Technology

scRNA-seq + TCR + CyTOF + flow

Material

Peripheral blood immune cells

Local matrix

None. This page is not a donor browser.

Source note

No local single-cell matrix. Use the published-evidence page for figures, sample accounting and paper tables. Exact participant-level age, sex and sample accession are not mapped.

Published sample accounting

Study-reported counts, not deduplicated LACA donors.

Site / layer	CEN / SC	Offspring	Control	Note
Lingao (Hainan)	17	9	14	Discovery scRNA / TCR
Rugao (Jiangsu)	7	6	9	Reuse of HRA002867
Japan (RIKEN)	7	2	3	Public SC2018
Paper headline total	31	17	26	Includes CyTOF / flow subsets; not a single matrix

Paper figure gallery

Captions are curated locally. The images themselves stay on PMC / the publisher. This article is ND-licensed, so LACA does not rehost or recolour the panels.

Fig.1 Global PBMC immune landscape UMAP of 10 major clusters with group proportions and box plots across CEN / offspring / control. Fig.2 CyTOF immune landscape 11 CyTOF clusters, t-SNE and percentage box plots validating the scRNA shifts. Fig.3 CD4+ T phenotypes Nine CD4 subsets, trajectories, TF networks and RUNX3 flow validation. Fig.4 CD8+ T phenotypes Six CD8 subsets, cytotoxic / MHC programmes and delayed naive-to-CTL polarisation. Fig.5 NK and γδ T phenotypes Seven NK / γδ subsets with RUNX3 / KLF4 and flow confirmation. Fig.6 Centenarian NK signatures Cytotoxic genes, granzyme B / perforin after stimulation, and rejuvenated membrane receptors. Fig.7 NK–T cell–cell communication CellChat heatmaps: enhanced MHC-I, CD99 and MIF between NCAM1-low NK and T cells.

Core published tables

Transcribed facts from the open full text. Not a re-analysis.

CEN vs control composition shifts (paper-reported)
Compartment	Direction in CEN vs CT	Reported P
B cells	Decreased	P<0.05
CD4+ T	Decreased	P<0.001
NK cells	Increased	P<0.01
MAIT	Decreased	P<0.01
Proliferative T / NK	Increased	P<0.01
CD4+ CTL	Increased	paper-reported
CD27+ Tcm / Th17 / Tn	Decreased	paper-reported
CD8+ CTL / CMC1+	Increased	paper-reported
RUNX3 in CD56dim NK	Increased	P<0.001
RUNX3 in CD4+ T	Increased	P<0.05

Source: Main text; Kruskal–Wallis + Dunn. Exact percentages live in the figure box plots, not a numeric table.

Where the 31 / 17 / 26 headline counts come from
Site	CEN	Offspring	Control	Note
Lingao	17	9	14	Discovery
Rugao	7	6	9	HRA002867
Japan	7	2	3	RIKEN SC2018
Headline total	31	17	26	Not one open matrix

Source: Methods / results sample accounting in PMC12571583.

Accessions and exits

PMC full text ↗ PubMed ↗ Publisher ↗ HRA011570 · Lingao discovery (controlled) ↗ HRA002867 · Rugao reuse (controlled) ↗ RIKEN SC2018 · Japan cohort ↗ GSE158055 · Age-control reuse ↗