What the study is
Sebahat Melike Durukan, Mustafa Burak Acar and colleagues, at the Genome and Stem Cell Center (GENKOK) of Erciyes University in Kayseri, with collaborators in Naples and Berlin, report in the Journal of Cellular Physiology (volume 241, issue 4, e70172, April 2026 issue; open access, CC BY; doi:10.1002/jcp.70172) on the biological state of dental pulp stem cells (DPSCs) taken from teeth whose caries has not reached the pulp. The design is a matched comparison. From eight donors aged 18 to 40 they isolated DPSCs from the pulp of clinically healthy semi-impacted third molars, and from eight donors they isolated DPSCs from third molars with deep dentinal caries that showed no clinical or radiographic sign of irreversible pulpitis: no spontaneous pain, no prolonged sensitivity, no swelling, no periapical pathology (ethics approval 2019/573). To get enough material for proteomics, cells from the donors within each group were pooled separately at the end of passage 1. The question is narrow and practical: does a carious lesion that still looks restorable leave the resident stem cell population functionally intact.
The senescence phenotype, measured two ways
Both quantitative senescence assays moved in the same direction. In the C12FDG flow-cytometry assay, senescent cells made up 25.60% ± 1.44 of the caries-group DPSCs versus 9.70% ± 1.66 in the healthy group; in the senescence-associated beta-galactosidase stain, 22.11% ± 2.88 versus 6.78% ± 1.87, both p < 0.0001. The rest of the phenotype is consistent with that reading. Caries-group cells shifted toward G0/G1 (57.73% ± 1.6 versus 54.35% ± 0.88), proliferated more slowly over 72 hours on the MTT assay, and enlarged: nuclear area rose from 102.2 ± 4.69 to 183.4 ± 10.27 square micrometers, and cytoplasmic area from 639.75 ± 41.82 to 1775.3 ± 107.09. The senescence markers p16 and p21 were near background in healthy cells and clearly present in caries-group cells (p16 intensity 28.55 ± 3.65, p21 16.29 ± 6.61). Total apoptotic cells were also higher by Annexin V/7-AAD (7.25% ± 1.63 versus 0.30% ± 0.16), so the carious pulp niche was losing cells to both senescence and apoptosis while still clinically quiet.
The proteome and the secretome
Shotgun LC-MS/MS proteomics separated the groups further. Whole-cell analysis found 612 unique proteins in healthy DPSCs, 386 in caries-group DPSCs, and 1,899 common to both; the secretome showed 58 unique proteins on the healthy side, 90 on the carious side, and 221 shared. Label-free quantification of shared proteins found 8 downregulated and 5 upregulated whole-cell proteins in the caries group (including NQO1 and the stemness-linked IGF2BP3 down, and HIST2H2BE up), and 16 differentially expressed secretome proteins, 12 of them upregulated, dominated by extracellular matrix remodeling factors: TIMP1, MMP2, LOXL1, COL3A1, fibulin-1, PCOLCE, EMILIN1, and MXRA8, plus IGFBP4, a named component of the senescence-associated secretory phenotype (SASP). Pathway analysis (Ingenuity) placed NF-kB, detected only in the caries-group whole-cell proteome, at the center of the IL-1, IL-6, and TNF-alpha acute-phase pathways, alongside a TLR4-containing LPS-stimulated MAPK cascade that the authors read as molecular evidence of past microbial product contact; the retinoblastoma cell-cycle regulator appeared only in healthy cells. Immunofluorescence confirmed the inflammatory signal at the protein level: NF-kB intensity 24.46 ± 5.15 in caries-group cells versus 0.07 ± 0.15 in healthy, and TNF-alpha detectable only in the caries group (14.00 ± 9.37), while IL-1beta and IL-6 showed no significant between-group difference.
The honest boundary
This is an ex vivo association study in 16 pooled donor samples, not a causal demonstration in an animal or a patient. Pooling at passage 1 masks donor-to-donor variation, as the authors themselves flag in their limitations section, and third molars from young adults are one anatomical source, not the general pulp population. The assays show the caries-group cells carry senescence markers and a SASP-like secretome in culture; they do not show that this state blocks tertiary dentin formation or pulp repair in the same teeth, and the suggestion that senescent pulp cells could impose a systemic inflammatory burden is the authors’ extrapolation, not a result. The direction of cause is also underdetermined: the cells may have been driven senescent by the lesion, or a pre-existing subclinical pulp condition may have accompanied the caries. Finally, p values quoted per assay describe the pooled comparison; with eight donors per group merged into one population, donor-level statistics are not available.
Why it matters for pulp regeneration
For the pulp-dentin repair program, the finding is a sourcing-quality caveat rather than a new mechanism. Autologous DPSC-based pulp therapies, and the cell banks that supply allogeneic and off-the-shelf approaches, assume that cells taken from a tooth without pulpitis symptoms are functionally healthy. This record says that assumption fails in at least some caries-adjacent teeth: a clinically restorable lesion can coincide with a quarter of the resident stem cells already senescent and secreting ECM-remodeling, pro-inflammatory factors. It also complements the senescence-mechanism papers the field has produced this year, such as the CPT1A-Parkin mitophagy axis and the SHED senescence screen, by supplying the human isolation-study counterpart: in mice and dishes we know how pulp stem cells are pushed into senescence, and here is evidence that the push has already happened in some human teeth that would pass a routine clinical screen. The field assessment does not change on this record; it adds a donor-screening question to the route’s translation checklist.
Provenance: grounded in the open-access full text of Durukan et al. (2026), Journal of Cellular Physiology 241(4):e70172, doi:10.1002/jcp.70172, retrieved via Europe PMC (PMC13084213) on 2 October 2026. See /method/.