The neurotrophic effects of different human dental mesenchymal stem cells

Abstract

The current gold standard treatment for peripheral nerve injury is nerve grafting but this has disadvantages such as donor site morbidity. New techniques focus on replacing these grafts with nerve conduits enhanced with growth factors and/or various cell types such as mesenchymal stem cells (MSCs). Dental-MSCs (D-MSCs) including stem cells obtained from apical papilla (SCAP), dental pulp stem cells (DPSC), and periodontal ligament stem cells (PDLSC) are potential sources of MSCs for nerve repair. Here we present the characterization of various D-MSCs from the same human donors for peripheral nerve regeneration. SCAP, DPSC and PDLSC expressed BDNF, GDNF, NGF, NTF3, ANGPT1 and VEGFA growth factor transcripts. Conditioned media from D-MSCs enhanced neurite outgrowth in an in vitro assay. Application of neutralizing antibodies showed that brain derived neurotrophic factor plays an important mechanistic role by which the D-MSCs stimulate neurite outgrowth. SCAP, DPSC and PDLSC were used to treat a 10 mm nerve gap defect in a rat sciatic nerve injury model. All the stem cell types significantly enhanced axon regeneration after two weeks and showed neuroprotective effects on the dorsal root ganglia neurons. Overall the results suggested SCAP to be the optimal dental stem cell type for peripheral nerve repair.

Figure 1

Human D-MSCs characterization, osteogenic, and adipogenic differentiation. Flow cytometry analysis of human SCAP, DPSC and PDLSC from two donors (marked as I and II) at passage 2, were positive for expression of CD73, CD90 and CD105 and lacked the expression of the negative markers (HLA-DR, CD45, CD34, CD19 and CD11b) (a). The expression profiles of the MSCs markers were inter- and intra-individually similar among the various D-MSCs (a). Differentiation along osteogenic lineage (Alizarin red) and adipogenic lineage (Oil Red O) is shown in (b) and (c), respectively. Scale bar: 100 μm.

Figure 2

Neurotrophic and angiogenic factor gene expression of D-MSCs. RT-PCR analysis of various unstimulated D-MSCs at passage 2, showed a variability of gene expression in unstimulated D-MSCs. GAPDH is used as a house-keeping gene. For original and uncropped blots, see Supplementary Figure 1. Quantitative-PCR analyses for NANOG (b), BDNF (c), GDNF (d), ANGPT1 (e) and VEGFA (f) with GAPDH, ACTB and RPL13A used for normalisation and relative expressions set to value 1 for the SCAP group. P < 0.01 is indicated by ** and P < 0.001 is indicated by *** (n = 2 patients and each condition was performed with 4 technical replicates).

Figure 3

Quantification of secreted protein in supernatants from unstimulated D-MSCs (cultured in regular MSCs medium) and stimulated D-MSCs (cultured for 14 days in MSCs-medium supplemented with neuregulin1-β1, basic fibroblast growth factor, platelet derived growth factor and forskolin). ELISA analysis of stimulated D-MSCs showed significantly elevated protein levels of BDNF (a) and VEGF-A (e) in SCAP and DPSC groups compared with corresponding unstimulated controls. In contrast, the stimulation protocol did not induce PDLSC to secrete significantly higher levels of BDNF (a). In addition the stimulation of D-MSCs did not lead to elevated secretion of NGF (b), NT3 (c), GDNF (d), or angiopoietin-1 (f). All levels were normalized to 10⁶ cells per group (pg/ml = pg/10⁶ cells). Mean values +/− SD are indicated, n = 6 independent experiments from 2 different donors for each D-MSCs group. P < 0.05 is indicated by *, P < 0.01 is indicated by ** and P < 0.001 is indicated by ***.

Figure 4

Neurite outgrowth of SHSY-5Y neuronal cells exposed to various conditioned media from D-MSCs. SHSY-5Y cells were retinoic acid-differentiated for 48 h (a–c) or 72 h (d–g) prior to exposure to conditioned media. Thereafter, retinoic acid-differentiated SHSY-5Y cells were exposed to control MSCs-medium (=control med), unstimulated D-MSCs conditioned medium (=SCAP, DPSC, or PDLSC med), stimulation medium alone (containing neuregulin1-β1, basic fibroblast growth factor, platelet derived growth factor and forskolin, =stim med), or stimulated D-MSCs conditioned medium (D-MSCs cultured for 14 days in MSCs-medium supplemented with neuregulin1-β1, basic fibroblast growth factor, platelet derived growth factor and forskolin, =stim-SCAP, stim-DPSC, or stim-PDLSC med). Exposure to various control or conditioned media was for additional 48 h (a–c) or 72 h (d–g). Quantitative analyses (n = 400 cells/group) of βIII-tubulin immunostained SHSY-5Y neurons exposed to various media are shown for the percentage of cells with neurite outgrowth (a), total neurite outgrowth (b), and longest neurite length (c). SCAP-groups were further analysed for possible effects of secreted factors on neurite outgrowth, by utilizing neutralizing antibodies (n-ab) against BDNF or VEGF-A (d–g). Representative βIII-tubulin immunostaining (400X magnification) of SHSY-5Y neurons showed decrease of neurite outgrowth upon blockage of BDNF, but not for VEGF-A (e–g). Quantitative analyses of SHSY-5Y neurites/cell (n = 38–41 cells/group) from (d) are shown in (f), and from (e) are presented in (g). Mean values with error bars showing SD (a–c) and SEM (f,g). Statistical significance are indicated with *P < 0.05; **P < 0.01; ***P < 0.001 or if not significant (ns). Scale bar: 50 μm.

Figure 5

The effect of transplanted human D-MSCs on nerve regeneration in vivo. βIII-tubulin staining of longitudinal sections through the nerve conduit showing regeneration from the proximal stump into the 10 mm gap after 2 weeks. Representative stainings are shown for empty conduits (Empty: n = 5), rat Schwann cells (r-SC: n = 5), human-SCAP (h-SCAP: n = 4 donor I, n = 4 donor II), human DPSC (h-DPSC: n = 3 donor I, n = 3 donor II), or human PDLSC (h-PDLSC: n = 3 donor I, n = 3 donor II) (a). The presence of human nuclei-specific antigen (h-nuclei; green) showed transplanted cells (shown for donor II) in the conduit after 2 weeks. Note that there are no co-localizations with S100 antigen (red) (b). The distance of the axon regeneration in the conduit (c) and Caspase-3 expression in DRG (d). Axon regeneration was measured in control empty conduits (Empty: n = 5), conduits with rat SC (n = 5) and conduits with unstimulated D-MSCs (h-SCAP: n = 4 donor I, n = 4 donor II; h-DPSC: n = 3 donor I, n = 3 donor II; or h-PDLSC: n = 3 donor I, n = 3 donor II) (c). Significance levels of *P < 0.05; **P < 0.01; ***P < 0.001 are indicated. Caspase-3 gene expression in L4-L6 DRG from rats with no injury (Control), or after nerve injury and repair (Empty, rSC, h-SCAP, h-DPSC, or h-PDLSC). Relative expression levels are shown with regard to control samples (value = 1) (d). Connecting lines show relative significance; *P < 0.05, **P < 0.01, and ***P < 0.001, n.s., not significantly different. n = 5 from pooled animal RNA samples (5–8 animals per group). Mean values + /−SD are indicated. Scale bar: 250 μm.