What was tested
Niyaz Al-Sharabi and six colleagues collected conditioned medium from cultured human dental pulp stem cells (DPSC-CM) after 24, 48, and 72 h of serum-free conditioning and asked whether the conditioning duration changes what the medium contains and what it does (preprint, posted 3 September 2026). The medium was profiled by label-free LC-MS/MS proteomics with donor-blocked differential abundance analysis. Producer-cell stress was checked by Annexin V/PI flow cytometry. Two conditioned-medium groups, 48 h and 72 h, were tested for angiogenesis in the ex ovo chick chorioallantoic membrane (CAM) assay. All three groups were applied to stem cells from human exfoliated deciduous teeth (SHED) for metabolic activity and osteogenic readouts (alkaline phosphatase staining and Alizarin Red S mineralization). A single PI3K/Akt western blot was run, and only on the 48 h group.
What they found
Conditioning up to 72 h did not significantly affect DPSC viability or apoptosis. Proteomics identified 1562 proteins, with 914 differentially abundant between the 48 h and 24 h medium, 745 between the 72 h and 24 h medium, and none between the 72 h and 48 h medium: by this measure the proteome reaches a plateau after 48 h. Enriched processes across the changing sets included extracellular matrix organization, cell-substrate adhesion, cytoskeletal organization, protein folding, and angiogenesis-related pathways. In the CAM assay, the 48 h medium increased total vessel area, while the 72 h medium increased total vessel length. The 48 h medium increased SHED metabolic activity at day 3. Alkaline phosphatase staining did not differ among groups; Alizarin Red S showed greater matrix mineralization with the 48 h medium than with the 24 h medium. PI3K inhibition reduced Akt phosphorylation as expected, but the 48 h medium itself did not significantly move the p-Akt/Akt ratio.
What it does not show
There is no tooth repair model anywhere in the design: no dentin-pulp assay, no mammalian in vivo work, only cultured cells and a chick-membrane angiogenesis readout. Two features limit the headline comparison. First, the 24 h group was not run in the CAM assay and was not probed by western blot, so part of the case for 48 h reflects which assays were run on which groups rather than a measured inferiority of 24 h. Second, the CAM endpoints are not comparable across groups: the 48 h group was read on vessel area and the 72 h group on vessel length, so no angiogenesis superiority of one duration over the other can be concluded from the stated data. The PI3K/Akt arm is a single negative readout on one group; it does not establish that the medium acts independently of PI3K/Akt. The abstract reports no donor count, no group sizes, and no statistical detail, so the load-bearing numbers cannot be weighed for precision. The work is a preprint and has not been peer reviewed.
Where it sits in the field
For the pulp and dentin repair program, cell-free preparations (conditioned medium, secretome, and extracellular vesicles from pulp and deciduous-tooth stem cells) are a live alternative to cell transplantation, and production standardization is one of their main translational problems: duration, serum exposure, and donor state all change the product. This preprint contributes a direct, proteome-scale datapoint that most of the compositional change happens between 24 and 48 h of serum-free conditioning, which is a useful manufacturing parameter if it holds up in review. It sits alongside, and does not resolve, the harder question the program page tracks: whether any cell-free pulp preparation produces organized pulp-dentin repair in a mammalian model. The current field assessment stands at /field/.
Where we differ from the coverage
We found no press or popular coverage of this preprint to differ from. Against the abstract’s own conclusion, we decline the shorthand that 48 h is “best.” On the stated data, 48 h is the earliest duration at which the proteome reaches the plateau, and the functional readouts for 48 h and 72 h were measured on different endpoints, so the preprint supports “48 h is enough to reach the mature profile,” not “48 h outperforms 72 h.”
Provenance: grounded in the abstract of record (Crossref, DOI 10.20944/preprints202609.0224.v1, posted 3 September 2026; also indexed by Europe PMC as PPR1311257); the full text is free but blocked automated retrieval, so it was not read, and every number above comes from the abstract. The work is a preprint and has not been peer reviewed. Method and sourcing standard at /method/.