What was compared

Amer and colleagues isolated human dental pulp stem cells (hDPSCs) from impacted, caries free third molars, and separately isolated extracellular vesicles (EVs) from two sources: hDPSC conditioned media and human blood plasma, both from healthy, non smoking, non alcohol consuming men aged 20 to 35. A G*Power calculation set a target of 16 donors per group at alpha 0.05 and 80 percent power; the paper does not restate how many donors actually fed the pooled assays reported in the results, which limits how confidently the significance figures below can be read as donor-level, not just well-level, effects.

Both EV types were isolated by differential ultracentrifugation and dosed onto hDPSC cultures at the same protein concentration, 50 micrograms per milliliter. Nanoparticle tracking put the two EV preparations at different particle concentrations at that same protein dose: 8.9 times 10 to the 10th particles per cubic centimeter for hDPSC derived EVs (median diameter 133.2 nm) against 1.8 times 10 to the 11th particles per cubic centimeter for plasma derived EVs (median diameter 117.6 nm). Protein mass, not particle number, was held equal between arms.

Proliferation and mineralization

MTT assay optical density rose across all three groups (control, hDPSC-EV, plasma-EV) from day 1 to day 5. By day 5 the values were 0.442 plus or minus 0.017 (control), 0.791 plus or minus 0.016 (hDPSC-EVs), and 0.546 plus or minus 0.017 (plasma-EVs), with all three pairwise comparisons significant (ANOVA P < 0.001). Alizarin Red staining after 14 days, read as optical density, was 0.088 plus or minus 0.001 in the negative control, 0.211 plus or minus 0.037 in the positive (osteogenic-medium) control, 0.456 plus or minus 0.051 with hDPSC-EVs, and 0.315 plus or minus 0.038 with plasma-EVs (ANOVA P < 0.001, all pairwise Tukey comparisons significant).

The differentiation markers

Dentin sialophosphoprotein (DSPP), measured by quantitative PCR and expressed as fold change against beta-actin, was 1.037 plus or minus 0.318 (negative control), 2.046 plus or minus 0.277 (positive control), 4.220 plus or minus 0.713 (hDPSC-EVs), and 2.648 plus or minus 0.844 (plasma-EVs). The hDPSC-EV group differed significantly from the positive control (p = 0.009) and from the plasma-EV group (p = 0.046); the plasma-EV group did not differ significantly from the positive control (p = 0.617). Osteocalcin, read as percent DAB-positive area by immunocytochemistry, was 0.997 plus or minus 0.097, 4.972 plus or minus 0.853, 22.218 plus or minus 1.614, and 18.198 plus or minus 0.701 for the same four groups in order, with every pairwise comparison significant.

What it does not show

This is entirely an in vitro comparison: isolated cells, isolated vesicles, a 14-day culture window. There is no animal model, no pulp injury, and no dentin bridge measured in a tooth. Donors were restricted to healthy men aged 20 to 35, a narrow slice of the population that eventually needs pulp therapy, and the authors state this limits generalizability. The authors also flag their own central design tradeoff: dosing by protein mass rather than particle count means the plasma-EV arm may have delivered more particles at the same 50 microgram-per-milliliter dose, which the paper’s own numbers support (1.8 times 10 to the 11th versus 8.9 times 10 to the 10th particles per cubic centimeter). A same-lineage advantage measured this way could partly reflect a dosing artifact rather than tissue-specific cargo alone.

Where we differ from the framing

The paper concludes that tissue origin “significantly influences EVs bioactivity” and frames hDPSC-derived vesicles as biologically superior. That is a fair description of what was measured, but the comparison was not particle-matched, so it cannot yet separate a true cargo effect from a dose effect. The paper’s own discussion says as much in its limitations, so this is not a disagreement with the authors, it is a reason to treat “tissue-matched EVs work better” as provisional pending a particle-normalized repeat.

For the pulp and dentin repair program, this adds a cell-free candidate, vesicles rather than transplanted stem cells, to a route whose strongest human evidence remains a 26-patient autologous pulp stem cell graft trial. It does not move that program’s tier: no animal or human pulp-injury data exist yet for either EV type. The current field assessment is at /field/.

Provenance: every number and finding above was checked against the full primary report and its methods, results, and figure captions, per our method at /method/.