Reduced expression of connective tissue growth factor (CTGF/CCN2) mediates collagen loss in chronologically aged human skin

Abstract

Reduced production of type I procollagen is a prominent feature of chronologically aged human skin. Connective tissue growth factor (CTGF/CCN2), a downstream target of the transforming growth factor-beta (TGF-beta)/Smad pathway, is highly expressed in numerous fibrotic disorders, in which it is believed to stimulate excessive collagen production. CTGF is constitutively expressed in normal human dermis in vivo, suggesting that CTGF is a physiological regulator of collagen expression. We report here that the TGF-beta/Smad/CTGF axis is significantly reduced in dermal fibroblasts, the major collagen-producing cells, in aged (> or = 80 years) human skin in vivo. In primary human skin fibroblasts, neutralization of endogenous TGF-beta or knockdown of CTGF substantially reduced the expression of type I procollagen mRNA, protein, and promoter activity. In contrast, overexpression of CTGF stimulated type I procollagen expression, and increased promoter activity. Inhibition of TGF-beta receptor kinase, knockdown of Smad4, or overexpression of inhibitory Smad7 abolished CTGF stimulation of type I procollagen expression. However, CTGF did not stimulate Smad3 phosphorylation or Smad3-dependent transcriptional activity. These data indicate that in human skin fibroblasts, type I procollagen expression is dependent on endogenous production of both TGF-beta and CTGF, which act through interdependent yet distinct mechanisms. Downregulation of the TGF-beta/Smad/CTGF axis likely mediates reduced type I procollagen expression in aged human skin in vivo.

Figure 1

TGF-β1 mRNA and protein expression are reduced in the dermis of aged human skin in vivo. (A) Total RNA was prepared from dermis, obtained by dissection of full thickness young (21–30 years) or aged (80 or greater years of age) human skin. TGF-β1 and 36B4 (internal reference for normalization) mRNA were quantified in the dermis of young (21–30 years of age) and aged (80 and greater years of age) human skin, by real time RT-PCR. Data are means+SEM, N=6 young and 6 aged subjects, *p<0.05. (B) Localization of TGF-β1 mRNA in the dermis of young and aged human skin was determined by antisense riboprobe in situ hybridization. Open arrows indicate vascular cell. Solid arrows indicate fibroblasts. Results are representative five young and five aged individuals. Bar=300 m. (C) Localization of TGF-β1 protein, in the dermis of young and aged human skin, was determined by immunohistology. Results are representative of five young and five aged individuals. Bar=300 m. (D) Dermal fibroblasts were obtained by laser capture microdissection from frozen sections of young (21–30 years) and aged (80+ years) human skin. TGF-β1 and 36B4 (internal reference) mRNA levels were quantified by real-time RT-PCR. Data are means SEM, N=6 young and 6 aged subjects, *p<0.05.

Figure 2

CTGF mRNA and protein expression are reduced in the dermis of aged human skin in vivo. (A) Total RNA was prepared from the dermis, obtained by dissection of full thickness young (21–30 years) and aged (80 or greater years of age) human skin. CTGF and 36B4 (internal reference for normalization) mRNA levels were quantified by real-time RT-PCR. Data are means + SEM, N=5 young and 5 aged subjects, *p<0.05. (B) Localization of CTGF mRNA in the dermis of young and aged human skin was determined by antisense riboprobe in situ hybridization. Results are representative five young and five aged individuals. Bar=300 μm. Open arrows indicate vascular cell. Solid arrows indicate fibroblasts. (C) Localization of CTGF protein, in the dermis of young and aged human skin, was determined by immunohistology. Results are representative of five young and five aged individuals. Bar=300 μm. (D) Dermal fibroblasts were obtained by laser capture microdissection from frozen sections of young (21–30 years) and aged (80+ years) human skin. CTGF and 36B4 (internal reference for normalization) mRNA levels were quantified by real-time RT-PCR. Data are means + SEM, N=8 young and 8 aged subjects, *p<0.05.

Figure 3

Type I procollagen mRNA and protein expression are reduced in the dermis of aged human skin in vivo. (A) Total RNA was prepared from the dermis, obtained by dissection of full thickness young (21–30 years) and aged (80 or greater years of age) human skin. Type I procollagen and 36B4 (internal reference) mRNA levels were quantified by real-time RT-PCR. Data are means + SEM, N=6 young and 6 aged subjects, *p<0.05. (B) Localization of type I procollagen mRNA, in the dermis of young and aged human skin, was determined by antisense riboprobe in situ hybridization. Results are representative five young and five aged individuals. Bar=300 μm. (C) Localization of type I procollagen protein, in the dermis of young and aged human skin, was determined by immunohistology. Results are representative of five young and five aged individuals. Bar=300 μm. (D) Dermal fibroblasts were obtained by laser capture microdissection from frozen sections of young (21–30 years) and aged (80+ years) human skin. Type I procollagen and 36B4 (internal reference for normalization) mRNA levels were quantified by real-time RT-PCR. Data are means + SEM, N=13 young and 13 aged subjects, *p<0.05.

Figure 4

Neutralization of endogenous TGF-β reduces expression of CTGF and type I procollagen in primary adult human dermal fibroblasts. Fibroblasts were cultured in the presence of the indicated amounts of pan TGF-β neutralizing antibody for 24 hours. (A) CTGF and (B) type I procollagen mRNA were quantified by real-time RT-PCR. 36B4 mRNA levels were used as internal reference. Data are means + SEM, N=3, *p<0.05. (C) Fibroblasts were cultured in the presence of the indicated amounts of pan TGF-β neutralizing antibody for the indicated times. Type I procollagen protein secreted into the culture media was quantified by ELISA. Data are means + SEM, N=3, *p<0.05.

Figure 5

Knockdown of endogenous TGF-β reduces expression of CTGF and type I procollagen in primary adult human dermal fibroblasts. (A) Fibroblasts were transfected with the indicated non-specific (Ctrl) or TGF-β isoform-specific siRNA. TGF-β1, TGF-β2, and TGF-β3 mRNA levels were determined by real-time RT-PCR, 48 hours after transfection. 36B4 mRNA levels were used as internal reference. Data are means + SEM, N=3, *p<0.05. (B–E) Fibroblasts were transfected with pooled TGF-β1, TGF-β2, and TGF-β3 siRNA. Cells were harvested 48 hours after transfection and analyzed for CTGF or type I procollagen mRNA or protein levels. Transcript levels were determined by real-time RT-PCR (36B4 mRNA was used as internal reference), and protein levels were determined by Western analyses (β-actin was used as internal control). (B) CTGF mRNA levels. Data are means + SEM, N=3, *p<0.05. (C) CTGF protein levels. Insets shows representative Western blots. Data are means + SEM, N=3, *p<0.05. (D) Type I procollagen mRNA levels. Data are means + SEM, N=3, *p<0.05. (E) Type I procollagen protein levels. Insets shows representative Western blots. Data are means + SEM, N=3, *p<0.05.

Figure 6

Inhibition of type I TGF-β receptor reduces expression of CTGF and type I procollagen in primary adult human dermal fibroblasts. Fibroblasts were treated with DMSO vehicle (Ctrl) or type I TGF-β receptor inhibitor SB431542 (10μM) for 24 hours. CTGF and type I procollagen mRNA and protein levels were determined by real-time RT-PCR and Western analyses, respectively. (A) CTGF mRNA levels were normalized to 36B4 mRNA levels, used as internal reference. Data are means+SEM, N=3, *p<0.05. (B) CTGF protein levels were normalized to s-actin used as internal control. Insets shows representative Western blots. Data are means+SEM, N=3, *p<0.05. (C) Type I procollagen mRNA levels were normalized to 36B4 mRNA levels, used as internal reference. Data are means+SEM, N=3, *p<0.05. (D) Type I procollagen protein levels were normalized to s-actin used as internal control. Insets shows representative Western blots. Data are means+SEM, N=3, *p<0.05.

Figure 7

Knockdown of CTGF reduces expression of type I procollagen in primary adult human dermal fibro-blasts. Fibroblasts were transfected with non-specific siRNA (Ctrl) or CTGF siRNA. Total RNA or protein was prepared 48 hours after transfection. (A) CTGF and 36B4 (internal reference for normalization) mRNA levels were quantified by real-time RT-PCR. Data are means + SEM, N=3, *p<0.05. (B) CTGF and s-actin (internal reference for normalization) protein levels were quantified by Western analyses. Insets show representative Western blots. Data are means + SEM, N=3, *p<0.05. (C) Type I procollagen and 36B4 (internal reference for normalization) mRNA levels were quantified by real-time RT-PCR. Data are means + SEM, N=3, *p<0.05. (D) Type I procollagen and s-actin (internal reference) protein levels were quantified by Western analyses. Insets show representative Western blots. Data are means + SEM, N=3, *p<0.05. (E) Fibroblasts were transfected with type I procollagen α2 promoter (COL1A2) CAT reporter and Lac Z reporter (internal control for normalization), with empty vector (pCDNA3.1) and non-specific siRNA (Ctrl), or CTGF expression vector, or CTGF siRNA. CAT and β-galactosidase activities were determined 48 hours after transfection. Data are means + SEM, N=3, *p<0.05.

Figure 8

CTGF regulation of type I procollagen is dependent on TGF-β/Smad signaling pathway in primary adult human dermal fibroblasts. (A) Fibroblasts were transfected with empty vector or CTGF expression vector. Thirty-two hours after transfection, cells were treated for 16 hours with vehicle or type I TGF-β receptor inhibitor SB431542 (10μM). Type I procollagen (PROCOL-I), CTGF, and s-actin (internal reference for normalization) protein levels were quantified by Western analyses. Insets show representative Western blots. Data are means + SEM, N=3, *p<0.05. (B) Fibroblasts were transfected with empty vector (−) (pCDNA3.1) and non-specific siRNA (Ctrl siRNA), or CTGF expression vector with Smad4 siRNA, or Smad7 expression vector. Whole cell protein extracts were prepared 48 hours after transfection. Type I procollagen, CTGF, Smad4, Smad7, and s-actin (internal reference) protein levels were quantified by Western analyses. Insets show representative Western blots. Data are means + SEM, N=3, *p<0.05.

Figure 9

CTGF does not modulate TGF-β/Smad signaling pathway in primary adult human dermal fibro-blasts. (A) Fibroblasts were transfected with CTGF siRNA or CTGF expression vector or their appropriate controls (CTRL), non-specific siRNA, orempty vector pCDNA3.1, respectively. Whole cell lysates were prepared 48 hours after transfection. Phosphorylated Smad2 (p-Smad2), phosphorylated Smad3 (p-Smad3), total Smad2, total Smad3, CTGF, and s-actin (internal reference for normalization) protein levels were determined by Western analyses. Inset shows representative Western blots. N=3. (B) Fibroblasts were co-transfected with Smad3 luciferse reporter construct (SBEX4) and LacZ reporter (internal control for normalization), or CTGF siRNA, or CTGF expression vector, or their appropriate controls (CTRL), empty luciferase reporter plasmid (pGL3), or non-specific siRNA, or empty expression vector, respectively. Whole cell extracts were prepared 48 hours after transfection, and assayed for luciferase and β-galactosidase activities. Data are means + SEM, N=3.