Figure 1.
Expression of Periostin and CK6 during cutaneous wound healing.
H&E: Representative histological sections of cutaneous incisional wounds. (A) Morphology of the wounded site shows a thin edge of epithelial cells migrating across the wound bed, termed the epithelial tongue and (B) intact and normal skin adjacent to the wounded site were stained with Hematoxylin and Eosin (H&E). (C) Epithelial cells at the epithelial tongue express intracellular Periostin. (D) In normal adjacent skin, Periostin is in the connective tissue at the basal lamina, which is juxtaposed to the epithelial basal layer. (E) Note that basal and parabasal layers of the epithelial tongue have a large number of proliferating BrDU positive cells. (F) As expected, the epithelial basal layer of adjacent skin has few proliferating cells. (G) Upregulation of the epithelia stress/tension marker CK6 is depicted in the epithelial tongue compared to normal adjacent skin observed in (H). Scale bars represent 50 μm.
Figure 2.
Periostin-driven migration and proliferation.
(A) Total cell lysates and conditioned medium (cond. medium) from NOK-SI and hPDL cells were blotted for Periostin. New cell culture medium supplemented with 10% FBS was used as a negative control (control). Intracellular Periostin is detected in epithelial cell lysate. However, conditioned medium from NOK-SI shows that keratinocytes did not secrete Periostin, as the same band was observed in the negative control media. hPDL cells have low levels of the intracellular Periostin isoform as observed in the cell lysate. Increased levels of secreted Periostin were found in the conditioned medium from hPDL. (B) hPDL conditioned medium induces keratinocyte proliferation compared to vehicle alone (***p<0.001), which is reduced upon administration of anti-Periostin antibody (*p<0.05). (C) Representative pictures of NOK-SI migration following treatment with recombinant Periostin (50 ng/ml), EGF (100 ng/ml) as the positive control, or vehicle. Scale bars represent 50 μm. (D) Graphic represents the quantification of the wound areas at indicated times (n=4; mean ± S.E.M). (E) Periostin enhances proliferation of keratinocytes compared to vehicle treated cells (***p<0.001). EGF treatment was used as positive control (*p<0.05) (n=6; mean ± S.E.M).
Figure 3.
F-actin polarization and PI3K/mTOR signaling activation by Periostin-induced epithelial cell migration.
(A) Phalloidin detection shows cells with polarized F-actin (white arrow) following treatment with recombinant Periostin compared to vehicle control. Scale bars represent 10 μm. (B) Graphic represents percentage of cells with stress fiber formation (polarized F-actin) after periostin or vehicle stimuli. Results were determined by measuring fields using independent triplicates (**p<0.01) (C) Activation of PI3K and mTOR signaling is triggered by Periostin treatment in a dose-dependent manner, as detected by phosphorylated AKT at Threonine 308 (pAKTThr308) and Serine 473 (pAKTSer473) and phosphorylated S6 (pS6). Note that 50 ng/ml of Periostin is the optimal concentration for PI3K activation. GAPDH was used as a loading control.
Figure 4.
Co-expression of Periostin and mTOR during cellular migration and mechanical stress induced by tension.
(A) A representative wound healing scratch assay shows keratinocytes stained for Periostin (TRITC-red), pS6 (FITC-green) and DNA (Hoechst-blue). Note colocalization of Periostin and pS6 staining (on merge and insert) in the migratory area. Scale bars represent 50 μm. (B) Quantification of positive cells co-expressing Periostin and pS6 are depicted. Note increased number of positive cells co-expressing Periostin and pS6. Most of these cells are in the migratory area (***p<0.001). (C) NOK-SI cells were subjected to biomechanical stimulation (load of 14% stretching at 6 cycles/min) at the indicated time points. (D) Western blot analysis for Periostin and pS6 expression in NOK-SI subjected to load assay. Non-stimulated cells (no load force) served as a control, and GAPDH was used as protein loading control.
Figure 5.
Periostin-driven cellular migration requires mTOR signaling.
(A) Representative pictures of the NOK-SI cell scratch assay following treatment with vehicle, recombinant Periostin (50 ng/ml), and rapamycin (50 nM). Scale bars represent 50 μm. (B) Quantitative analysis of open-wounded area over time (n=4; mean ± S.E.M.). Note that rapamycin abrogates the Periostin migratory activity of epithelial cells (***p<0.001). (C) Proliferation assay using keratinocytes treated with rapamycin and/or Periostin. Note that Periostin alone induced significant cellular proliferation at 50 ng/ml (*p<0.05). Treatment with rapamycin blocked periostin-induced cell proliferation (ns p>0.05). (D) Representative immunoblot depicting knockdown of Raptor and Rictor after siRNA treatment. Scramble siRNA oligonucleotides sequences were used as controls. GAPDH was used as loading controls. (E) Graphic shows the quantitative analyses of open-wounded areas using NOK-SI cells over time (n=4; mean ± S.E.M.). Note that siRNA targeting Raptor abrogates Periostin-induced cellular migratory resulting on complete wound closure by 48 hours (**p<0.05). siRNA targeting Rictor did not change the Periostin induced accelerated cellular migration resulting on wound closure by 24 hours (ns p>0.05). (F) Proliferation assay using NOK-SI cells treated with siRNA for Raptor, Rictor, or siRNA scramble, and/or Periostin. Note that Periostin induced significant cellular proliferation (*p<0.05). Treatment with siRNA for Raptor or Rictor resulted in disruption of Periostin induced cellular proliferation (***p<0.001).
Figure 6.
Proposed mechanism of Periostin-mediated accelerated epithelial healing through the mTOR pathway.
During wound healing, activation of the Periostin signaling circuitry is initiated by the extracellular accumulation of Periostin secreted by fibroblasts, and by intracellular periostin originated after mechanical stress. Following, activation of the mTOR pathway occurs. Notably, the mTORC1 is required for Periostin-driven accelerated epithelial migration, while activation of mTORC1 and mTORC2 is required for epithelial proliferation.