Neuronal PAS Domain 2 (Npas2)-Deficient Fibroblasts Accelerate Skin Wound Healing and Dermal Collagen Reconstruction

Abstract

The circadian clock, which consists of endogenous self-sustained and cell-autonomous oscillations in mammalian cells, is known to regulate a wide range of peripheral tissues. The unique upregulation of a clock gene, neuronal PAS domain protein 2 (Npas2), observed along with fibroblast aging prompted us to investigate the role of Npas2 in the homeostasis of dermal structure using in vivo and in vitro wound healing models. Time-course healing of a full-thickness skin punched wound exhibited significantly faster wound closure in Npas2-/- mice than wild-type (WT) C57Bl/6J mice. Dorsal skin fibroblasts isolated from WT, Npas2+/-, and Npas2-/- mice exhibited consistent profiles of core clock gene expression except for Npas2 and Per2. In vitro behavioral characterizations of dermal fibroblasts revealed that Npas2-/- mutation was associated with increased proliferation, migration, and cell contraction measured by floating collagen gel contraction and single-cell force contraction assays. Npas2 knockout fibroblasts carrying sustained the high expression level of type XII and XIV FAICT collagens and synthesized dermis-like thick collagen fibers in vitro. Confocal laser scanning microscopy demonstrated the reconstruction of dermis-like collagen architecture in the wound healing area of Npas2-/- mice. This study indicates that the induced Npas2 expression in fibroblasts may interfere with skin homeostasis, wound healing, and dermal tissue reconstruction, providing a basis for novel therapeutic target and strategy. Anat Rec, 2019. © 2019 Wiley Periodicals, Inc.

Keywords: Npas2; circadian rhythm; collagen; fibroblast; wound healing.

Fig. 1.

Full-thickness skin punch wound healing. (A) A standardized photograph of the skin wound was obtained from 0 to 12 days after surgery, depicting the progressive wound closure and contraction. (B) The relative wound area was calculated at 2, 4, 6, and 12 days. Npas2 KO mice showed a significantly smaller wound area than that of WT mice at day 12 (**P < 0.01). (C) Histological observation of wounds at Day 7 showed the formation of granulation tissue (GT) and the restoration of epithelial integrity (EP); however, the wound margin (dotted line) was clearly observed. At Day 14, the wound margin highlighted by hair follicles (HFs) was less clear and approached toward the granulation tissue (GT).

Fig. 2.

Characterization of WT, Npas2+/−, and Npas2−/− skin fibroblasts. (A) The genotype of each fibroblast batch was determined by genomic DNA PCR. The WT Npas2 allele generated a 250 bp PCR product, whereas the mutant allele generated a 350 bp PCR product. (B) The WST-1 assay demonstrated the increased cell proliferation rate in Npas2 KO fibroblasts (**P < 0.01, significant difference compared with WT at the time points via the Tukey analysis). (C) The expression of core clock genes and the LacZ reporter gene was determined by RT-PCR every 6 for 48 hr (P value in the figure: two-way ANOVA for the interaction between the time and genotype factors. *P < 0.05, **P < 0.01, significant difference compared with WT at the time points via the Tukey analysis) (C).

Fig. 3.

In vitro wound healing experiment using WT, Npas2+/− and Npas2−/− fibroblasts. (A) Images of time-lapse micrographs captured the progressive scratch wound healing assay. The number of migrated cells within the scratched area was significantly larger in the Npas2 KO groups at 6 hr and 12 hr (**P < 0.01). (C) Standardized images of floating collagen gel depicted an increased collagen gel contraction in the Npas2 KO fibroblast groups. (D) The area of collagen gels decreased over time. The gel contraction speed was faster in Npas2 KO fibroblasts (**P < 0.01, significant difference shown only compared with WT). (E) Schematic presentation of the FLECS-based single-cell contraction. (F) The ratio of contracted cells was increased in Npas2 KO fibroblasts. (G) Npas2 KO mutation did not affect the gene expression of β-actin (Actb) and α-SMA (Acta2) in dermal fibroblasts. (H) The steady state gene expression level of integrin subunits αV (ItgaV), β3 (Itgb3), and β5 (Itgb5) in dermal fibroblasts was not affected by Npas2 KO mutation.

Fig. 4.

Collagen synthesis by WT, Npas2+/− and Npas2−/− fibroblasts in vitro. (A) Gene expression of collagen type I (Collal and Colla2), type III (Col3a1), type XII (Coll2a1), and type XIV (Coll4a1) (**P < 0.01, *P < 0.05, significant difference shown only compared with WT). FACIT collagen type XII and type XIV showed significantly increased steady-state mRNA levels in Npas2+/− and Npas2−/− fibroblasts. (B) Images for cultured fibroblasts with picrosirius red staining highlighted the synthesis of collagen fibers. (C) The in vitro collagen fiber deposition was measured by picrosirius red staining (**P < 0.01 by one-way ANOVA with post hoc Holm test).

Fig. 5.

Evaluation of collagen fiber structure in the wound healing area. (A) Confocal laser scanning microscopy depicted the collagen fiber architecture stained with picrosirius red at 14 days after surgery. (B) Measurement of the wound closure ratio using the wound closure area (WCA) calculated as the width of the ISA (a) between panniculus carnosus (PC) subtracted by the granulation tissue area (GT: b), which was normalized by ISA. (C) The wound closure ratio was greater in Npas2+/− and Npas2−/− mice at Day 14, albeit statistical significance was achieved only between the WT and Npas2−/− groups.