SPARC Is Highly Expressed in Young Skin and Promotes Extracellular Matrix Integrity in Fibroblasts via the TGF-β Signaling Pathway

Abstract

The matricellular secreted protein acidic and rich in cysteine (SPARC; also known as osteonectin), is involved in the regulation of extracellular matrix (ECM) synthesis, cell-ECM interactions, and bone mineralization. We found decreased SPARC expression in aged skin. Incubating foreskin fibroblasts with recombinant human SPARC led to increased type I collagen production and decreased matrix metalloproteinase-1 (MMP-1) secretion at the protein and mRNA levels. In a three-dimensional culture of foreskin fibroblasts mimicking the dermis, SPARC significantly increased the synthesis of type I collagen and decreased its degradation. In addition, SPARC also induced receptor-regulated SMAD (R-SMAD) phosphorylation. An inhibitor of transforming growth factor-beta (TGF-β) receptor type 1 reversed the SPARC-induced increase in type I collagen and decrease in MMP-1, and decreased SPARC-induced R-SMAD phosphorylation. Transcriptome analysis revealed that SPARC modulated expression of genes involved in ECM synthesis and regulation in fibroblasts. RT-qPCR confirmed that a subset of differentially expressed genes is induced by SPARC. These results indicated that SPARC enhanced ECM integrity by activating the TGF-β signaling pathway in fibroblasts. We inferred that the decline in SPARC expression in aged skin contributes to process of skin aging by negatively affecting ECM integrity in fibroblasts.

Keywords: MMP-1; SPARC; TGF-β; extracellular matrix; fibroblast; skin aging; type I collagen.

Figure 1

Analysis of SPARC expression in young (ages 24, 21, and 28 y) and elderly (ages 87, 79, and 76 y) human skin tissues. (A) Messenger RNA level of SPARC in human skin tissues determined by quantitative RT-PCR. Graph shows comparison of SPARC mRNA levels between elderly and young human skin tissues. Values are shown as means ± standard deviation (SD) of eight independent experiments. * p < 0.05 vs. young tissues. (B) IHC detection of SPARC polypeptide in skin tissues from young and elderly participants. Specimens were incubated with SPARC antibody, followed by horseradish peroxidase-conjugated secondary antibody with 3-amino-9-ethylcarbazole, and counterstained with hematoxylin. Graphs show the relative levels of SPARC protein in the dermis and epidermis of elderly and young human skin tissues. Values are shown as means ± SD of three independent experiments. * p < 0.05 and ** p < 0.01 vs. young tissues. Magnification, ×200. Bar, 100 µm.

Figure 2

Effects of SPARC on secretion of type I collagen and MMP-1 in fibroblasts. Subconfluent human foreskin fibroblasts were incubated in serum-free DMEM with SPARC (0–8 μg/mL) for (A), (2 μg/mL) for (B), or TGF-β1 (3 ng/mL) or without (Con) for 24 h. Expression of type I collagen and MMP-1 in conditioned media and of GAPDH in fibroblast lysates was evaluated by western blotting using antibodies against pN-COL1A1, MMP-1, and GAPDH. Graphs show relative levels of COL1A1 and MMP-1 proteins. Values represent means ± SD of three independent experiments. * p < 0.05 and ** p < 0.01 vs. Con.

Figure 3

Effects of SPARC on mRNA levels of COL1A1, COL1A2, and MMP-1 in fibroblasts. Serum-starved foreskin fibroblasts were incubated for 12 h with SPARC (2 µg/mL) or TGF-β1 (3 ng/mL) or without (Con). Levels of COL1A1, COL1A2, and MMP-1 mRNAs were evaluated using conventional (A) and quantitative (B) RT-PCR analyses. Values are shown as means ± SD of five independent experiments. *** p < 0.001 vs. Con.

Figure 4

Effects of SPARC on biosynthesis and degradation of type I collagen in 3D cultured fibroblasts. Foreskin fibroblasts were embedded in a type I collagen matrix to mimic the dermis in vivo, then incubated in serum-free and phenol-red-free DMEM without (Con) or with SPARC (2 µg/mL) for 24 h. Foreskin fibroblasts were stained with mouse anti-pN-COL1A1 and Rhodamine Red-conjugated anti-mouse IgG antibodies to detect nascent type I collagen, and with rabbit anti-type I collagen cleavage-site and Alexa Fluor 488-conjugated anti-rabbit IgG antibodies to detect cleavage site of type I collagen. Nuclei were stained with Hoechst 33258. Type I collagen synthesis and degradation were analyzed by confocal fluorescence microscopy. Boxed areas are enlarged to show subcellular images of type I collagen synthesis and type I collagen degradation. Staining intensity of synthesized and degraded type I collagen was quantified using ImageJ software and normalized to nuclear staining intensity. Values are shown as means ± SD of three independent experiments. * p < 0.05 and ** p < 0.01 vs. Con. Magnification, ×200; bar, 200 µm.

Figure 5

Analysis of TGF-β receptor-regulated R-SMADs activation by SPARC in fibroblasts. (A) Foreskin fibroblasts were starved for 12 h, then incubated with SPARC (2 µg/mL) or TGF-β1 (3 ng/mL) for 30 min. Cell lysates were analyzed by western blotting using antibodies against phospho-SMAD2 (p-SMAD2), SMAD2, phospho-SMAD3 (p-SMAD3), SMAD3, and GAPDH. (B) Serum-starved foreskin fibroblasts were incubated with SB431542 (10 µM) for 10 min then stimulated with SPARC (2 µg/mL) or TGF-β1 (3 ng/mL) for 30 min. Levels of p-SMAD2, SMAD2, and GAPDH were examined by western blotting using p-SMAD2, SMAD2, and GAPDH antibodies. (C) Serum-starved foreskin fibroblasts were incubated with SB431542 (10 µM) and stimulated with SPARC (2 µg/mL) or TGF-β1 (3 ng/mL) for 24 h. Expression of type I collagen and MMP-1 in conditioned media and of GAPDH in cell lysates was evaluated by western blotting using antibodies against pN-COL1A1, MMP-1, and GAPDH.

Figure 6

Time course effects of SPARC on SMAD2 phosphorylation in fibroblasts. Serum-starved foreskin fibroblasts were incubated with vehicle (Con), SPARC (2 μg/mL), or TGF-β1 (3 ng/mL) for 0, 30, 90, and 270 min. Cell lysates were analyzed by western blotting using antibodies against p-SMAD2, SMAD2 and GAPDH. Graph shows relative intensity of p-SMAD2/GAPDH in fibroblasts incubated with SPARC or TGF-β1, normalized to that of fibroblasts incubated with TGF-β1 for 30 min. Values are shown as means ± SD of three independent experiments. * p < 0.05, ** p < 0.01, and *** p < 0.001 vs. 0 min.

Figure 7

Identification of genes differentially expressed by SPARC in fibroblasts via RNA sequencing. We sequenced mRNA derived from foreskin fibroblasts incubated with SPARC (2 μg/mL) or without (Con) for 12 h. (A) DEGs were identified by analyzing three independent sets of RNA mixtures obtained from foreskin fibroblasts incubated without (Con) or with SPARC (2 μg/mL) for 12 h. DEGs were selected based on criteria of absolute (log₂ (fold change)) ≥ 1.0 and p-value < 0.05. (B) DEGs were classified into three GO categories: BP, CC, and MF. Significance of GO terms in each category is indicated as -log₁₀ (adjusted p) values. (C) Nine genes selected based on absolute (log₂ (fold change)) ≥ 1.5 and p-value < 0.05. Messenger RNA levels of nine genes, COL1A1, MMP-1, and GAPDH, were evaluated using conventional (left) and quantitative (right) RT-PCR analyses. Values are shown as means ± SD of three independent experiments. * p < 0.05, ** p < 0.01, and *** p < 0.001 vs. Con.