Advertisement

Age-Associated Increase in Skin Fibroblast–Derived Prostaglandin E2 Contributes to Reduced Collagen Levels in Elderly Human Skin

      Production of type I collagen declines during aging, leading to skin thinning and impaired function. Prostaglandin E2 (PGE2) is a pleiotropic lipid mediator that is synthesized from arachidonic acid by the sequential actions of cyclooxygenases (COX) and PGE synthases (PTGES). PGE2 inhibits collagen production by fibroblasts in vitro. We report that PTGES1 and COX2 progressively increase with aging in sun-protected human skin. PTGES1 and COX2 mRNA were increased 3.4-fold and 2.7-fold, respectively, in the dermis of elderly (>80 years) versus young (21–30 years) individuals. Fibroblasts were the major cell source of both enzymes. PGE2 levels were increased 70% in elderly skin. Fibroblasts in aged skin display reduced spreading due to collagen fibril fragmentation. To investigate the relationship between spreading and PGE2 synthesis, fibroblasts were cultured on micropost arrays or hydrogels of varying mechanical compliance. Reduced spreading/mechanical force resulted in increased expression of both PTGES1 and COX2 and elevated levels of PGE2. Inhibition of PGE2 synthesis by diclofenac enhanced collagen production in skin organ cultures. These data suggest that reduced spreading/mechanical force of fibroblasts in aged skin elevates PGE2 production, contributing to reduced collagen production. Inhibition of PGE2 production may be therapeutically beneficial for combating age-associated collagen deficit in human skin.

      Abbreviations:

      COX2 (cyclooxygenase 2), ECM (extracellular matrix), EIA (enzyme immunoassay), PGE2 (prostaglandin E2), PTGES (prostaglandin E synthase)

      Introduction

      Elderly skin is typically thin and fragile, with increased susceptibility to bruising and impaired wound healing (
      • Fisher G.J.
      • Varani J.
      • Voorhees J.J.
      Looking older: fibroblast collapse and therapeutic implications.
      ). These alterations largely reflect fragmentation and reduction in type I collagen fibrils, which comprise the bulk of the dermal extracellular matrix (ECM) (
      • Varani J.
      • Dame M.K.
      • Rittie L.
      • et al.
      Decreased collagen production in chronologically aged skin: roles of age-dependent alteration in fibroblast function and defective mechanical stimulation.
      ). Dermal fibroblasts secrete type I procollagen, which is converted into collagen via proteolytic modifications. Mature collagen self-assembles into collagen fibrils, which are stabilized by inter- and intra-fibrillar cross-linking (
      • Canty E.G.
      • Kadler K.E.
      Procollagen trafficking, processing and fibrillogenesis.
      ). Although mature cross-linked collagen fibrils are very stable, with a half-life of approximately 15 years (
      • Verzijl N.
      • DeGroot J.
      • Thorpe S.R.
      • et al.
      Effect of collagen turnover on the accumulation of advanced glycation end products.
      ), gradual cleavage, mediated by matrix metalloproteases, occurs during aging (
      • Fisher G.J.
      • Quan T.
      • Purohit T.
      • et al.
      Collagen fragmentation promotes oxidative stress and elevates matrix metalloproteinase-1 in fibroblasts in aged human skin.
      ). This cleavage results in the accumulation of cross-linked fragments, owing to resistance of cross-links to proteolysis. Thus, during aging, the dermal ECM becomes progressively degraded and disorganized, which deleteriously alters the function of resident fibroblasts (
      • Fisher G.J.
      • Varani J.
      • Voorhees J.J.
      Looking older: fibroblast collapse and therapeutic implications.
      ).
      Fibroblasts attach to surrounding collagen fibrils to form adhesion complexes, which act through the cytoskeleton to exert contractile forces. Resistance to this contraction generates mechanical forces within fibroblasts, which largely determine morphology, cytoskeletal organization, signal transduction, and gene expression (
      • Tarutani M.
      • Cai T.
      • Dajee M.
      • et al.
      Inducible activation of Ras and Raf in adult epidermis.
      ;
      • Hegedus L.
      • Cho H.
      • Xie X.
      • et al.
      Additional MDA-MB-231 breast cancer cell matrix metalloproteinases promote invasiveness.
      ;
      • Mammoto T.
      • Mammoto A.
      • Ingber D.E.
      Mechanobiology and developmental control.
      ). Collagen fragmentation causes loss of attachment sites, resulting in reduced fibroblast spreading, which is observed in aged human skin. This contracted state, with concomitant reduced mechanical force, is associated with downregulation of collagen fibril production in human skin (
      • Varani J.
      • Dame M.K.
      • Rittie L.
      • et al.
      Decreased collagen production in chronologically aged skin: roles of age-dependent alteration in fibroblast function and defective mechanical stimulation.
      ;
      • Fisher G.J.
      • Varani J.
      • Voorhees J.J.
      Looking older: fibroblast collapse and therapeutic implications.
      ;
      • Xia W.
      • Hammerberg C.
      • Li Y.
      • et al.
      Expression of catalytically active matrix metalloproteinase-1 in dermal fibroblasts induces collagen fragmentation and functional alterations that resemble aged human skin.
      ).
      PGE2 is a pleiotropic lipid signaling molecule produced by multiple cell types (
      • Humphrey J.D.
      • Dufresne E.R.
      • Schwartz M.A.
      Mechanotransduction and extracellular matrix homeostasis.
      ). PGE2 is synthesized from arachidonic acid by the sequential actions of cyclooxygenases (COX1 and COX2) and prostaglandin E synthases (PTGES1, 2, and 3) (
      • Liu X.
      • Wu H.
      • Byrne M.
      • et al.
      A targeted mutation at the known collagenase cleavage site in mouse type I collagen impairs tissue remodeling.
      ;
      • Iskratsch T.
      • Wolfenson H.
      • Sheetz M.P.
      Appreciating force and shape-the rise of mechanotransduction in cell biology.
      ). Elevated COX2 often coincides with PTGES1 induction in a wide variety of tumor lesions and in response to inflammatory stimuli (
      • Verrecchia F.
      • Chu M.L.
      • Mauviel A.
      Identification of novel TGF-beta /Smad gene targets in dermal fibroblasts using a combined cDNA microarray/promoter transactivation approach.
      ;
      • Pickup M.
      • Novitskiy S.
      • Moses H.L.
      The roles of TGFbeta in the tumour microenvironment.
      ;
      • Quan T.
      • Fisher G.
      Role of age-associated alterations of the dermal extracellular matrix microenvironment in human skin aging.
      ). COX1, PTGES2, and PTGES3 are not typically inducible or involved in excess PGE2 production seen in abnormal conditions. PGE2 is a major prostaglandin in human skin and is normally synthesized at low levels but is markedly increased in skin squamous cell carcinoma and inflammatory conditions, such as sunburn (
      • Muller-Decker K.
      Cyclooxygenase-dependent signaling is causally linked to non-melanoma skin carcinogenesis: pharmacological, genetic, and clinical evidence.
      ). Non-steroidal anti-inflammatory drugs are a class of commonly used drugs including aspirin, ibuprofen, indomethacin, and diclofenac, which inhibit the activities of both COX1 and COX2, thereby inhibiting PGE2 production (
      • Qin Z.
      • Voorhees J.J.
      • Fisher G.J.
      • et al.
      Age-associated reduction of cellular spreading/mechanical force up-regulates matrix metalloproteinase-1 expression and collagen fibril fragmentation via c-Jun/AP-1 in human dermal fibroblasts.
      ). COX2 and PTGES1 induction in response to acute inflammatory stimuli has been extensively investigated (
      • Verrecchia F.
      • Chu M.L.
      • Mauviel A.
      Identification of novel TGF-beta /Smad gene targets in dermal fibroblasts using a combined cDNA microarray/promoter transactivation approach.
      ;
      • Samuelsson B.
      • Morgenstern R.
      • Jakobsson P.J.
      Membrane prostaglandin E synthase-1: a novel therapeutic target.
      ;
      • Pickup M.
      • Novitskiy S.
      • Moses H.L.
      The roles of TGFbeta in the tumour microenvironment.
      ;
      • Quan T.
      • Fisher G.
      Role of age-associated alterations of the dermal extracellular matrix microenvironment in human skin aging.
      ); however, the potential effects of cellular mechanical forces on COX2 or PTGES1 expression are unknown.
      In cell cultures, PGE2 inhibits fibroblast collagen production, in part through impeding actions of transforming growth factor-β (
      • Sandulache V.C.
      • Parekh A.
      • Li-Korotky H.
      • et al.
      Prostaglandin E2 inhibition of keloid fibroblast migration, contraction, and transforming growth factor (TGF)-beta1-induced collagen synthesis.
      ). The inhibitory effect of PGE2 on collagen synthesis has been shown to have a protective role against lung fibrosis, where impaired PGE2 production and action are implicated in excessive fibroblast collagen deposition (
      • Huang S.
      • Wettlaufer S.H.
      • Hogaboam C.
      • et al.
      Prostaglandin E(2) inhibits collagen expression and proliferation in patient-derived normal lung fibroblasts via E prostanoid 2 receptor and cAMP signaling.
      ).
      This study reports that PTGES1 mRNA expression progressively increases during aging and investigates the role of PGE2 in the age-related decline of type I collagen production in human skin. Taken together, the data support the concept that the dermal microenvironment raises PTGES1 and PGE2 levels, which contributes to reduced collagen in aged skin. Given that PGE2 synthesis can be effectively inhibited by a wide range of agents (
      • Qin Z.
      • Voorhees J.J.
      • Fisher G.J.
      • et al.
      Age-associated reduction of cellular spreading/mechanical force up-regulates matrix metalloproteinase-1 expression and collagen fibril fragmentation via c-Jun/AP-1 in human dermal fibroblasts.
      ), targeting PGE2 is an appealing strategy to combat age-associated skin collagen deficiency.

      Results

      PTGES1 mRNA expression progressively increases during aging in human skin in vivoReferences

      In order to investigate age-related alterations in gene expression, total RNA was extracted from full-thickness skin samples obtained from sun-protected buttock skin of persons between the ages of 18 and 75 (n=62) (
      • Swindell W.R.
      • Johnston A.
      • Sun L.
      • et al.
      Meta-profiles of gene expression during aging: limited similarities between mouse and human and an unexpectedly decreased inflammatory signature.
      ). Global gene expression analysis was conducted using an in situ oligonucleotide array platform (Affymetrix Human Genome U133 Plus 2.0 array). Out of 19,851 human genes, 268 exhibited statistically significant age-associated changes in expression levels (false discovery rate <0.10). PTGES1 expression positively correlated with age, and this correlation was the most statistically significant. PTGES1 expression was assessed by two probes targeting two different regions of the PTGES1 transcript and yielded similar correlation coefficients and yearly rates of increase, as calculated by linear regression. The results obtained from one probe, 210367_s_at, are shown in Figure 1a. The linear correlation between increased PTGES1 expression and increased age was highly statistically significant (P=2.6 × 10−7, n=62), and the degree of linearity, expressed as Pearson’s correlation coefficient, was r=0.6 (r can vary from 0 to 1, with 0 representing no correlation and 1 representing perfect linear correlation). When compared with young skin (18 years of age), elderly skin (75 years of age) had a 1.6-fold overall increase in PTGES1 expression.
      Figure thumbnail gr1
      Figure 1PTGES1 mRNA expression progressively increases in human skin during chronological aging. (a) Relative PTGES1 mRNA expression in human buttock skin was determined by cDNA microarray analysis. Results show positive correlation of the PTGES1 mRNA level and age of donors (n=62, P=2.6 × 10−7, r=0.60). (b) PTGES1 mRNA expression in buttock skin was determined by qPCR. (n=40, P=1.38 10−7, r=0.73). The subjects used in microarray analysis were different from the subjects used in qPCR analysis. cDNA, complementary DNA.
      In order to substantiate the age-associated increase in PTGES1, we acquired buttock skin samples from an additional 40 subjects aged 21–94. Consistent with our microarray data, PTGES1 expression progressively increased with age as determined by quantitative PCR (qPCR) and increased 2.92-fold overall in 94-year- versus 21-year-old skin (n=40, P=1.38 × 10−7, r=0.73) (Figure 1b). Taken together, these data demonstrate an intriguing correlation between aging and skin PTGES1 gene expression.

      Dermal fibroblasts are the primary source of increased expression for both PTGES1 and COX2 mRNA in aged human skin

      Elevation of COX2 expression often occurs concomitantly with PTGES1 induction. However, COX2 mRNA levels in whole skin preparations did not significantly differ between young and aged individuals (data not shown). We hypothesized that dermal expression of PTGES1 and COX2 mRNA may account for age-dependent differences. In order to determine dermal and epidermal PTGES1 and COX2 expression, we used laser capture microdissection of skin sections, followed by qPCR. Consistent with our hypothesis, dermal, but not epidermal, PTGES1 and COX2 gene expression were significantly different in young versus aged individuals. PTGES1 and COX2 levels were significantly increased 3.4-fold (P<0.05) and 2.7-fold, respectively (P<0.05), in aged dermis (Figure 2a and b). This finding suggests that aging dermal, but not epidermal, cells express more PTGES1 and COX2.
      Figure thumbnail gr2
      Figure 2Dermal fibroblasts are a major cell source for increased expression of PTGES1 and COX2 in aged (>80 years) versus young (21–30 years) skin. The dermis and the epidermis of frozen skin sections were separated and collected using laser capture microscopy. Relative (a) PTGES1 and (b) COX2 mRNA levels in the epidermis and dermis of young and aged individuals were determined by qPCR normalized to 36B4 mRNA. n=6, *P<0.05. (c and d) Cells were released from fresh sun-protected skin samples by collagenase digestion. Released cells were fractionated by immuno-affinity magnetic beads to separate fibroblasts for other cell types. Total RNA was extracted from isolated cells. (c) PTGES1 mRNA levels were determined in fibroblast-enriched and fibroblast-depleted cells (n=4, *P<0.05). (d) PTGES1 mRNA levels were determined in fibroblast-enriched cells isolated from skin samples from young and aged individuals. (n=5–7, *P<0.05).
      In order to determine whether fibroblasts are responsible for elevated dermal PTGES1 gene expression, we separated fibroblasts from other dermal cells in skin samples using anti-fibroblast antibody-coated magnetic microbeads. Fibroblast enrichment was validated by mRNA quantification of several cell markers (Supplementary Table S1 online). Isolated cells were directly analyzed without culturing. PTGES1 mRNA levels were approximately 10-fold (P<0.05) higher in fibroblast-enriched cells than in fibroblast-depleted dermal cells (Figure 2c). These data indicate that dermal fibroblasts are responsible for the majority of dermal PTGES1 expression.
      In order to determine whether dermal fibroblasts are responsible for the changes to PTGES1 expression in aging, we isolated fibroblasts from young (21–30 years) and aged (>80 years) skin. Consistent with our data, PTGES1 mRNA expression was 2.2-fold higher in fibroblasts from aged skin (P<0.05) (Figure 2d). We also quantified COX2 mRNA expression in isolated fibroblasts; however, COX2 mRNA levels in freshly isolated fibroblasts were markedly elevated by the isolation process. Taken together, the data suggest that dermal fibroblasts are the primary cell source of elevated PTGES1 and COX2 mRNA expression in aged skin.

      PTGES1 protein expression is elevated in skin fibroblasts of elderly (>80 years) versus young (21–30 years) individuals

      In order to determine whether PTGES1 protein levels were increased in aged dermal fibroblasts, we performed immunohistochemistry on both young and aged skin samples. Positive staining that overlapped with or was immediately adjacent to the nuclei was most prominently seen in dermal stromal cells in both young and aged skin. In addition, PTGES1-positive dermal cells displayed morphology characteristic of fibroblasts, including being embedded in collagenous ECM without direct contact with other cells or structures, elongated cell bodies, and oval shaped nuclei (Figure 3). Statistical analysis showed that the percentage of positively stained dermal stromal cells was increased 2.1-fold (75 vs. 35%, P<0.01) in aged skin. These data suggest that PTGES1 protein is preferentially expressed by fibroblasts, and fibroblast-derived PTGES1 protein expression is increased in the aged dermis, consistent with increased PTGES1 mRNA expression (Figure 2).
      Figure thumbnail gr3
      Figure 3PTGES1 protein expression is increased in fibroblasts in aged human skin. Immunohistochemistry of PTGES1 protein in young (21–30 years old) and aged (>80 years old) sun-protected buttocks human skin. Left panels: representative immunostaining in upper and lower dermis. PTGES1 protein staining is brownish, and nucleus counterstaining with hematoxylin is blue. Arrows point to PTGES1 positively stained cells. Right panel: Percentage of total dermal stromal cells that were PTGES1-positive. (n=6, *P<0.01) Scale bar=50 μm.
      COX2 protein expression was also examined by immunohistochemistry but did not yield specific staining, consistent with previous studies showing that COX2 protein expression in normal human skin is below the limit of immunohistochemistry detection (
      • Hoot K.E.
      • Oka M.
      • Han G.
      • et al.
      HGF upregulation contributes to angiogenesis in mice with keratinocyte-specific Smad2 deletion.
      ).

      PGE2 levels are higher in elderly versus young skin

      Age-associated increases in COX2 and PTGES1 likely result in enhanced PGE2 production. To examine this possibility, we incubated fresh samples of buttock skin from young (21–30 years) and aged (>80 years) persons in culture medium for 5 minutes and quantified PGE2 levels in the conditioned media. PGE2 levels were 1.7-fold higher (P<0.05) in skin specimens from elderly individuals (Figure 4). Taken together, these data indicate that increased COX2 and PTGES1 expression in aged dermal fibroblasts is associated with increased PGE2 production in human skin in vivo.
      Figure thumbnail gr4
      Figure 4Prostaglandin E2 (PGE2) levels are increased in the skin of elderly compared with young individuals. Skin biopsies (2 mm) were obtained from young (21–30 years) and elderly (>80 years) sun-protected buttock skin. PGE2 levels were quantified by substrate enzyme immunoassay and normalized to DNA content of corresponding skin samples. (n=10, *P<0.05).

      Enhanced ECM compliance elevates COX2 and PTGES1 expression and PGE2 synthesis in primary adult human dermal fibroblasts

      Reduced fibroblast spreading/mechanical force is considered a driving force of fibroblast dysfunction in elderly skin (
      • Fisher G.J.
      • Varani J.
      • Voorhees J.J.
      Looking older: fibroblast collapse and therapeutic implications.
      ). We therefore examined whether PTGES1 and COX2 expression were regulated by changes to ECM compliance. For these studies, we utilized micropost arrays and hydrogel assays. Fibroblasts were cultured on a micropost array with two different compliances––low compliance (15.3 kPa) and high compliance (1.3 kPa).
      Electron micrographs (Figure 5a) of fibroblasts adhered to type I collagen-coated micropost arrays revealed bending of the microposts through cytoskeleton contractile forces. More compliant microposts (1.3 kPa) were more deformed, resulting in reduced fibroblast spreading, resembling fibroblasts seen in fragmented collagenous ECM of aged dermis in vivo (
      • Fisher G.J.
      • Varani J.
      • Voorhees J.J.
      Looking older: fibroblast collapse and therapeutic implications.
      ). Less compliant microposts (15.3 kPa) resisted deformation and fibroblasts had a spread morphology, resembling the stretched fibroblasts seen in the intact collagenous ECM of young dermis in vivo (Figure 5a). On average, fibroblasts cultured on 1.3 kPa microposts displayed a 7.9-fold lower traction force compared with those cultured on 15.3-kPa microposts (Figure 5b), as measured by calculating the number of posts deformed by a cell, as well as the degree of deformation. These data confirm that fibroblasts grown on a less compliant ECM experience more mechanical force.
      Figure thumbnail gr5
      Figure 5Reduced spreading/mechanical force elevates COX2 and PTGES1 expression and prostaglandin E2 (PGE2) levels in human skin fibroblasts. (a–d) Fibroblasts obtained from young (21–30 years) individuals seeded on micropost arrays. (a) Scanning electron microscopy of fibroblasts. Micrographs are representative of three experiments examining >60 fibroblasts. (b) Quantification of traction force (nanonewton, nN) for 10 fibroblasts/condition (n=3, *P<0.001). (c) COX2 and PTGES1 mRNA levels were determined by qPCR (n=3, *P<0.05). (d) PGE2 levels in conditioned medium were quantified by EIA and normalized to cell number (n=3, *P<0.05). (e and f) Fibroblasts cultured from young skin were seeded on type I collagen-coated hydrogels with low or high compliance. (e) PTGES1 and COX2 mRNA levels were quantified by qPCR. (n=3, *P<0.05). (f) PGE2 levels in conditioned media quantified by EIA. (n=3, *P<0.05). EIA, enzyme immunoassay.
      To determine the effect of mechanical force on PGE2 production, PTGES1 and COX2 mRNA expression were measured in fibroblasts cultured on 15.3 kPa and 1.3 kPa microposts. PTGES1 and COX2 expression increased by 1.7-fold and 3.0-fold, respectively, when cultured on more compliant (1.3 kPa) compared with less compliant (15.3 kPa) microposts (Figure 5c). In addition, PGE2 levels in the conditioned media increased by 1.9-fold (Figure 5d), indicating that increased ECM compliance results in increased PGE2 production.
      To further confirm the stimulatory effects of high ECM compliance on PGE2 synthesis, we analyzed fibroblasts cultured on type I collagen-coated polyacrylamide hydrogels. Fibroblasts cultured on a more compliant hydrogel (7 kPa) exhibited increased PTGES1 (2.0-fold) and COX-2 (3.8-fold) mRNA expression and enhanced PGE2 levels (4.5-fold) compared with fibroblasts cultured on a less compliant 30 kPa hydrogel (Figure 5e and f). In concert with changes of PGE2, but in an opposite direction, procollagen mRNA and protein expression were increased by 2.2- and 2.8-fold, respectively, in 30 kPa hydrogels compared with 7 kPa hydrogels (Supplementary Figure S1 online). Taken together, these data indicate that increased ECM compliance results in reduced fibroblast spreading/mechanical force, which leads to enhanced PGE2 production and reduction in procollagen expression.

      PGE2 inhibits procollagen production in human dermal fibroblasts

      We next assessed the effects of PGE2 on collagen production of fibroblasts from young (21–30 years) skin. Fibroblasts were cultured in serum-free medium in the presence of PGE2 or vehicle (DMSO) for 24 hours. Type I procollagen protein was examined by western blot. PGE2 treatment reduced procollagen protein levels by approximately 66% (P<0.05) (Figure 6a).
      Figure thumbnail gr6
      Figure 6Prostaglandin E2 (PGE2) inhibits procollagen production in adult human skin fibroblasts. (a) Fibroblasts were cultured with the addition of PGE2 (10 nM) or vehicle (DMSO) for 24 hours. Type I procollagen protein levels in cell lysates were determined by western blot, normalized to β-actin. Inset shows representative western blots. (n=5, *P<0.05). (b) Skin samples were incubated in serum-free α-MEM media with the addition of diclofenac (10 μM) or vehicle (DMSO) for 16 hours. Type I collagen mRNA levels were quantified by qPCR and normalized to 36B4 mRNA levels (n=6). Type I procollagen protein levels in conditioned media were quantified by EIA. (n=3, *P<0.05).
      In order to determine whether inhibition of PGE2 production can enhance collagen production, we examined the effect of the general COX inhibitor, diclofenac, and the COX2-specific inhibitor, celecoxib, in organ culture of skin samples obtained from buttock skin. Fresh skin biopsies were cultured in the presence of inhibitor or vehicle (DMSO) for 16 hours. Type I procollagen protein in conditioned medium was quantified by enzyme immunoassay (EIA). Biopsies also underwent RNA extraction for quantification of type I procollagen expression. Diclofenac significantly reduced PGE2 levels by 92±7% (n=3, P<0.05, data not shown) and enhanced procollagen mRNA and protein expression by 2.3-fold and 1.9-fold, respectively (P<0.05) (Figure 6b). Celecoxib reduced PGE2 levels by approximately 80%, and enhanced procollagen mRNA expression by 1.9-fold (Supplementary Figure S2 online), suggesting that PGE2 derived from COX2 is largely responsible for inhibition of procollagen expression in human skin organ cultures. The results are consistent with the notion that increased COX-2 expression in aged dermis can lead to increased PGE2 production and consequently reduced collagen expression.

      Discussion

      Chronological aging is a progressive process driven by the accumulation and synergy of subtle alterations, resulting in declining organ structure and function. Because of its accessibility, the skin is an ideal organ for investigating the cumulative nature of aging and its underlying causalities in humans (
      • Fisher G.J.
      • Varani J.
      • Voorhees J.J.
      Looking older: fibroblast collapse and therapeutic implications.
      ). Age-related increase in PTGES1 mRNA and protein expression was consistently revealed by examining whole skin, isolated dermis, and freshly isolated fibroblasts, using different methods and different groups of donors. Fibroblasts in aged human skin displayed comparable increases in mRNA (2.2-fold) and protein (2.1-fold) levels, although direct quantitative comparisons between results obtained from two different techniques should be interpreted with caution.
      Although we found that epidermal COX2 and PTGES1 mRNA expression do not differ between young and aged, the results do not exclude the possibility that aged epidermal cells may also contribute to enhanced PGE2 levels observed in aged skin. The relative contribution of dermal and epidermal cells to enhanced PGE2 levels observed in whole skin specimens remains to be determined by future investigations.
      We have proposed that age-associated molecular and cellular alterations in the dermis create a self-perpetuating cycle that drives the progression of ECM degeneration (
      • Fisher G.
      • Sachs D.
      • Voorhees J.
      Ageing: collagenase-mediated collagen fragmentation as a rejuvenation target.
      ). Age-associated fragmentation of collagen fibrils impairs interactions of fibroblasts with the collagenous ECM, leading to suppression of collagen production and enhanced production of matrix metalloproteases, which collectively cause further deterioration of the ECM. Impaired ECM causes further fibroblast dysfunction in a self-sustaining loop (
      • Varani J.
      • Dame M.K.
      • Rittie L.
      • et al.
      Decreased collagen production in chronologically aged skin: roles of age-dependent alteration in fibroblast function and defective mechanical stimulation.
      ;
      • Fisher G.J.
      • Varani J.
      • Voorhees J.J.
      Looking older: fibroblast collapse and therapeutic implications.
      ;
      • Xia W.
      • Hammerberg C.
      • Li Y.
      • et al.
      Expression of catalytically active matrix metalloproteinase-1 in dermal fibroblasts induces collagen fragmentation and functional alterations that resemble aged human skin.
      ). This study implicates PGE2 in this self-perpetuating cycle. Impaired ECM causes reduced mechanical force within fibroblasts, which in turn elevates COX2 and PTGES1 gene expression, likely through activations of transcription factors such as NF-kB, which are known to regulate expression of these two genes (
      • Bage T.
      • Lindberg J.
      • Lundeberg J.
      • et al.
      Signal pathways JNK and NF-kappaB, identified by global gene expression profiling, are involved in regulation of TNFalpha-induced mPGES-1 and COX-2 expression in gingival fibroblasts.
      ).
      Several age-associated alterations found in skin fibroblasts in vivo may conspire to cause reduced collagen production in the context of collagen fibril fragmentation. In addition to PGE2, previous studies have found reduced expression of connective tissue growth factor/CCN2 and enhanced expression of CCN1 (cysteine-rich protein 61) in aged human skin fibroblasts. These alterations inhibit collagen production, at least in part, through blockade of transforming growth factor-β signaling (
      • Quan T.
      • He T.
      • Shao Y.
      • et al.
      Elevated cysteine-rich 61 mediates aberrant collagen homeostasis in chronologically aged and photoaged human skin.
      ;
      • Quan T.
      • Shao Y.
      • He T.
      • et al.
      Reduced expression of connective tissue growth factor (CTGF/CCN2) mediates collagen loss in chronologically aged human skin.
      ;
      • He T.
      • Quan T.
      • Shao Y.
      • et al.
      Oxidative exposure impairs TGF-beta pathway via reduction of type II receptor and SMAD3 in human skin fibroblasts.
      ). Therapeutics targeting these pathways may retard or reverse the decline of skin collagen production during aging, thereby improving skin health. Elevated PGE2 is an appealing target for intervention, as PGE2 production in human skin can be effectively reduced by topical non-steroidal anti-inflammatory drugs (
      • Qin Z.
      • Voorhees J.J.
      • Fisher G.J.
      • et al.
      Age-associated reduction of cellular spreading/mechanical force up-regulates matrix metalloproteinase-1 expression and collagen fibril fragmentation via c-Jun/AP-1 in human dermal fibroblasts.
      ). This study provides rationale for investigation into the potential beneficial effects of non-steroidal anti-inflammatory drugs for collagen deficiency in elderly skin. Mice models that display elevated skin PGE2 levels and resultant collagen reduction would provide opportunity for both further mechanistic studies and evaluation of agents that could retard age-related collagen deficiency via the inhibition of PGE2 synthesis and actions.
      In addition to a possible contribution to collagen deficiency, fibroblast-derived PGE2 may be a risk factor for other prevalent age-associated skin disorders. PGE2 may impair functions of adjacent immune cells and epidermal keratinocytes via a paracrine mechanism. Given the critical role of PGE2 in promoting cutaneous squamous cell carcinoma and inhibiting functions of antigen-presenting cells required for allergenic contact dermatitis, elevated fibroblast-derived PGE2 may contribute to the observed increase in skin cancer and reduced skin immunity in the elderly (
      • Muller-Decker K.
      Cyclooxygenase-dependent signaling is causally linked to non-melanoma skin carcinogenesis: pharmacological, genetic, and clinical evidence.
      ;
      • Scott N.M.
      • Ng R.L.
      • Gorman S.
      • et al.
      Prostaglandin E2 imprints a long-lasting effect on dendritic cell progenitors in the bone marrow.
      ). In addition to local effects, skin-derived PGE2 has been demonstrated to have a systemic impact. This effect is mediated by transit of PGE2 and PGE2-induced cytokines into the circulation (
      • Soontrapa K.
      • Honda T.
      • Sakata D.
      • et al.
      Prostaglandin E2-prostaglandin E receptor subtype 4 (EP4) signaling mediates UV irradiation-induced systemic immunosuppression.
      ;
      • Ullrich S.E.
      • Byrne S.N.
      The immunologic revolution: photoimmunology.
      ;
      • Scott N.M.
      • Ng R.L.
      • Gorman S.
      • et al.
      Prostaglandin E2 imprints a long-lasting effect on dendritic cell progenitors in the bone marrow.
      ). Whether constitutively elevated PGE2 in chronologically aged, sun-protected, skin impacts other organ systems is an intriguing question for future investigations.

      Materials and Methods

      Human tissue procurement

      Full-thickness skin biopsies (2, 4, and 6 mm in diameter) were taken from buttocks of healthy human subjects as previously described (
      • Li Y.
      • Xia W.
      • Liu Y.
      • et al.
      Solar ultraviolet irradiation induces decorin degradation in human skin likely via neutrophil elastase.
      ). All procedures involving human subjects were approved by the University of Michigan Institutional Review Board. Informed written consent was obtained from all human subjects.

      Complementary DNA microarray analysis

      Affymetrix Human Genome U133 Plus 2.0 arrays (54,675 probe-sets of 19,851 genes) were used for expression profiling (Santa Clara, CA). Normalized expression values were calculated using the robust multichip average algorithm (
      • Irizarry R.A.
      • Hobbs B.
      • Collin F.
      • et al.
      Exploration, normalization, and summaries of high density oligonucleotide array probe level data.
      ). One probe-set was analyzed for each gene. The representative probe-set was chosen by the highest average robust mulitichip average expression intensity among all 62 samples. Raw robust mulitichip average intensities were adjusted to remove gender and batch variation. Simple linear regression was then used to identify genes significantly changed with age. Raw P-values were adjusted using the Benjamini-Hochberg method (
      • Benjamini Y.
      • Hochberg Y.
      Controlling the false discovery rate: a practical and powerful approach to multiple testing.
      ). False discovery rate -adjusted P-value threshold of 0.10 was used.

      Laser-capture microdissection

      Laser capture microdissection was performed as previously described (
      • Li Y.
      • Xia W.
      • Liu Y.
      • et al.
      Solar ultraviolet irradiation induces decorin degradation in human skin likely via neutrophil elastase.
      ). Briefly, 30 tissue sections were prepared from each skin biopsy embedded in optimal cutting temperature. Appendage-free dermis and interfollicular-epidermis were separately collected using laser capture microdissection (Leica ASLMD System; Leica Microsystems, Wetzlar, Germany).

      Fibroblast isolation

      Skin biopsies were digested with collagenase (5 mg ml−1 in DMEM) for 16 hours at 37 °C as described previously (
      • Li Y.
      • Xia W.
      • Liu Y.
      • et al.
      Solar ultraviolet irradiation induces decorin degradation in human skin likely via neutrophil elastase.
      ). Fibroblasts that were released from the dermis were separated from other cell types using anti-fibroblast Microbeads (Miltenyi Biotec, San Diego, CA) according to the manufacturer’s instruction.

      Cell culture

      Fibroblasts were cultured from punch biopsies (6 mm) from buttock skin (aged 21–30 years) and maintained as previously described (
      • Fisher G.J.
      • Quan T.
      • Purohit T.
      • et al.
      Collagen fragmentation promotes oxidative stress and elevates matrix metalloproteinase-1 in fibroblasts in aged human skin.
      ). Fibroblasts between passages 3 and 8 were used in this study.

      RNA extraction and qPCR

      RNA was extracted using an RNA extraction kit (Qiagen, Chatsworth, CA). qPCR reagents were from Applied Biosystems (Foster City, CA). Ribosomal protein 36B4 was used to normalize gene expression (
      • Li Y.
      • Xia W.
      • Liu Y.
      • et al.
      Solar ultraviolet irradiation induces decorin degradation in human skin likely via neutrophil elastase.
      ).

      Micropost arrays and hydrogels

      Polydimethylsiloxane micropost arrays were fabricated, analyzed by scanning electron microscopy, and used for fibroblast traction force measurements, as previously described (
      • Fu J.
      • Wang Y.K.
      • Yang M.T.
      • et al.
      Mechanical regulation of cell function with geometrically modulated elastomeric substrates.
      ;
      • Bonnans C.
      • Chou J.
      • Werb Z.
      Remodelling the extracellular matrix in development and disease.
      ). Hydrogels were fabricated on glass coverslips as previously described (
      • Caulin C.
      • Nguyen T.
      • Lang G.A.
      • et al.
      An inducible mouse model for skin cancer reveals distinct roles for gain- and loss-of-function p53 mutations.
      ).

      Protein extraction and western blot analysis

      Protein extraction from primary fibroblasts and western analysis were performed as previously described (
      • Li Y.
      • Xia W.
      • Liu Y.
      • et al.
      Solar ultraviolet irradiation induces decorin degradation in human skin likely via neutrophil elastase.
      ).

      Immunohistochemistry

      Immunohistochemistry was performed with anti-PTGES1 mouse monoclonal antibody (Santa Cruz Biotech, Santa Cruz, CA) on frozen skin sections as previously described (
      • Rittie L.
      • Stoll S.W.
      • Kang S.
      • et al.
      Hedgehog signaling maintains hair follicle stem cell phenotype in young and aged human skin.
      ). Fifteen microscopic fields of each section were used for quantitation of cell staining.

      PGE2 enzyme immunoassay (EIA)

      After excision, skin buttock biopsies (2 mm) were immediately incubated in 200 μl α-MEM medium for 5 minutes. PGE2 levels in the conditioned media were quantified using the EIA assay (Cayman, Ann Arbor, MI). DNA was extracted from skin biopsies using the Micro-DNA Extraction Kit (Qiagen) and quantified by ultraviolet spectrometry. PGE2 levels were normalized to DNA content in corresponding biopsies. This method was validated by quantifying increased PGE2 levels in skin biopsies from acute ultraviolet-irradiated skin, compared with non-irradiated skin.

      Type I procollagen EIA from organ cultures

      Fresh skin biopsies were cultured in the presence of diclofenac or vehicle (DMSO) for 16 hours. Type I procollagen in organ culture conditioned medium was determined by an EIA kit (Takara Bio, Otsu Shiga, Japan) and normalized to total protein based on the manufacturer’s instruction.

      Statistical analysis

      The correlations between PTGES1 levels and age were analyzed using linear regression. Differences between two groups of samples were analyzed using the two-tailed t-test. Differences were considered significant when P<0.05. Data are presented as mean± standard error of mean. N numbers represent sample size of each group.

      ACKNOWLEDGMENTS

      We thank Diane Fiolek for her help with graphics and manuscript editing. We thank Suzan Rehbine, LPN, for human biopsy procurement. This research was in part supported by Dermatology Foundation Career Development Award (YL) Doris Duke Charitable Foundation, (JG), and by NIH awards T32-AR0007197 (CW), K08-AR060802 (JG), and R13-AR009431-48.

      Supplementary Material

      Supplementary material is linked to the online version of the paper at http://www.nature.com/jid

      REFERENCES

        • Bage T.
        • Lindberg J.
        • Lundeberg J.
        • et al.
        Signal pathways JNK and NF-kappaB, identified by global gene expression profiling, are involved in regulation of TNFalpha-induced mPGES-1 and COX-2 expression in gingival fibroblasts.
        BMC Genomics. 2010; 11: 241
        • Benjamini Y.
        • Hochberg Y.
        Controlling the false discovery rate: a practical and powerful approach to multiple testing.
        J R Stat Soc Series B Methodol. 1995; 57: 289-300
        • Bonnans C.
        • Chou J.
        • Werb Z.
        Remodelling the extracellular matrix in development and disease.
        Nat Rev Mol Cell Biol. 2014; 15: 786-801
        • Canty E.G.
        • Kadler K.E.
        Procollagen trafficking, processing and fibrillogenesis.
        J Cell Sci. 2005; 118: 1341-1353
        • Caulin C.
        • Nguyen T.
        • Lang G.A.
        • et al.
        An inducible mouse model for skin cancer reveals distinct roles for gain- and loss-of-function p53 mutations.
        J Clin Invest. 2007; 117: 1893-1901
        • Fisher G.
        • Sachs D.
        • Voorhees J.
        Ageing: collagenase-mediated collagen fragmentation as a rejuvenation target.
        Brit J Dermatol. 2014; 171: 446-449
        • Fisher G.J.
        • Quan T.
        • Purohit T.
        • et al.
        Collagen fragmentation promotes oxidative stress and elevates matrix metalloproteinase-1 in fibroblasts in aged human skin.
        Am J Pathol. 2009; 174: 101-114
        • Fisher G.J.
        • Varani J.
        • Voorhees J.J.
        Looking older: fibroblast collapse and therapeutic implications.
        Arch Dermatol. 2008; 144: 666-672
        • Fu J.
        • Wang Y.K.
        • Yang M.T.
        • et al.
        Mechanical regulation of cell function with geometrically modulated elastomeric substrates.
        Nat Methods. 2010; 7: 733-736
        • He T.
        • Quan T.
        • Shao Y.
        • et al.
        Oxidative exposure impairs TGF-beta pathway via reduction of type II receptor and SMAD3 in human skin fibroblasts.
        Age (Dordr). 2014; 36: 9623
        • Hegedus L.
        • Cho H.
        • Xie X.
        • et al.
        Additional MDA-MB-231 breast cancer cell matrix metalloproteinases promote invasiveness.
        J Cell Physiol. 2008; 216: 480-485
        • Hoot K.E.
        • Oka M.
        • Han G.
        • et al.
        HGF upregulation contributes to angiogenesis in mice with keratinocyte-specific Smad2 deletion.
        J Clin Invest. 2010; 120: 3606-3616
        • Huang S.
        • Wettlaufer S.H.
        • Hogaboam C.
        • et al.
        Prostaglandin E(2) inhibits collagen expression and proliferation in patient-derived normal lung fibroblasts via E prostanoid 2 receptor and cAMP signaling.
        Am J Physiol Lung Cell Mol Physiol. 2007; 292: L405-L413
        • Humphrey J.D.
        • Dufresne E.R.
        • Schwartz M.A.
        Mechanotransduction and extracellular matrix homeostasis.
        Nat Rev Mol Cell Biol. 2014; 15: 802-812
        • Irizarry R.A.
        • Hobbs B.
        • Collin F.
        • et al.
        Exploration, normalization, and summaries of high density oligonucleotide array probe level data.
        Biostatistics. 2003; 4: 249-264
        • Iskratsch T.
        • Wolfenson H.
        • Sheetz M.P.
        Appreciating force and shape-the rise of mechanotransduction in cell biology.
        Nat Rev Mol Cell Biol. 2014; 15: 825-833
        • Li Y.
        • Xia W.
        • Liu Y.
        • et al.
        Solar ultraviolet irradiation induces decorin degradation in human skin likely via neutrophil elastase.
        PLoS One. 2013; 8: e72563
        • Liu X.
        • Wu H.
        • Byrne M.
        • et al.
        A targeted mutation at the known collagenase cleavage site in mouse type I collagen impairs tissue remodeling.
        J Cell Biol. 1995; 130: 227-237
        • Mammoto T.
        • Mammoto A.
        • Ingber D.E.
        Mechanobiology and developmental control.
        Annu Rev Cell Dev Biol. 2013; 29: 27-61
        • Muller-Decker K.
        Cyclooxygenase-dependent signaling is causally linked to non-melanoma skin carcinogenesis: pharmacological, genetic, and clinical evidence.
        Cancer Metastasis Rev. 2011; 30: 343-361
        • Pickup M.
        • Novitskiy S.
        • Moses H.L.
        The roles of TGFbeta in the tumour microenvironment.
        Nat Rev Cancer. 2013; 13: 788-799
        • Qin Z.
        • Voorhees J.J.
        • Fisher G.J.
        • et al.
        Age-associated reduction of cellular spreading/mechanical force up-regulates matrix metalloproteinase-1 expression and collagen fibril fragmentation via c-Jun/AP-1 in human dermal fibroblasts.
        Aging Cell. 2014; 13: 1028-1037
        • Quan T.
        • Fisher G.
        Role of age-associated alterations of the dermal extracellular matrix microenvironment in human skin aging.
        Gerontology. 2015; (e-pub ahead of print 4 February 2015)
        • Quan T.
        • He T.
        • Shao Y.
        • et al.
        Elevated cysteine-rich 61 mediates aberrant collagen homeostasis in chronologically aged and photoaged human skin.
        Am J Pathol. 2006; 169: 482-490
        • Quan T.
        • Shao Y.
        • He T.
        • et al.
        Reduced expression of connective tissue growth factor (CTGF/CCN2) mediates collagen loss in chronologically aged human skin.
        J Invest Dermatol. 2010; 130: 415-424
        • Rittie L.
        • Stoll S.W.
        • Kang S.
        • et al.
        Hedgehog signaling maintains hair follicle stem cell phenotype in young and aged human skin.
        Aging Cell. 2009; 8: 738-751
        • Samuelsson B.
        • Morgenstern R.
        • Jakobsson P.J.
        Membrane prostaglandin E synthase-1: a novel therapeutic target.
        Pharmacol Rev. 2007; 59: 207-224
        • Sandulache V.C.
        • Parekh A.
        • Li-Korotky H.
        • et al.
        Prostaglandin E2 inhibition of keloid fibroblast migration, contraction, and transforming growth factor (TGF)-beta1-induced collagen synthesis.
        Wound Repair Regen. 2007; 15: 122-133
        • Scott N.M.
        • Ng R.L.
        • Gorman S.
        • et al.
        Prostaglandin E2 imprints a long-lasting effect on dendritic cell progenitors in the bone marrow.
        J Leukoc Biol. 2014; 95: 225-232
        • Soontrapa K.
        • Honda T.
        • Sakata D.
        • et al.
        Prostaglandin E2-prostaglandin E receptor subtype 4 (EP4) signaling mediates UV irradiation-induced systemic immunosuppression.
        Proc Natl Acad Sci USA. 2011; 108: 6668-6673
        • Swindell W.R.
        • Johnston A.
        • Sun L.
        • et al.
        Meta-profiles of gene expression during aging: limited similarities between mouse and human and an unexpectedly decreased inflammatory signature.
        PLoS One. 2012; 7: e33204
        • Tarutani M.
        • Cai T.
        • Dajee M.
        • et al.
        Inducible activation of Ras and Raf in adult epidermis.
        Cancer Res. 2003; 63: 319-323
        • Ullrich S.E.
        • Byrne S.N.
        The immunologic revolution: photoimmunology.
        J Invest Dermatol. 2012; 132: 896-905
        • Varani J.
        • Dame M.K.
        • Rittie L.
        • et al.
        Decreased collagen production in chronologically aged skin: roles of age-dependent alteration in fibroblast function and defective mechanical stimulation.
        Am J Pathol. 2006; 168: 1861-1868
        • Verrecchia F.
        • Chu M.L.
        • Mauviel A.
        Identification of novel TGF-beta /Smad gene targets in dermal fibroblasts using a combined cDNA microarray/promoter transactivation approach.
        J Biol Chem. 2001; 276: 17058-17062
        • Verzijl N.
        • DeGroot J.
        • Thorpe S.R.
        • et al.
        Effect of collagen turnover on the accumulation of advanced glycation end products.
        J Biol Chem. 2000; 275: 39027-39031
        • Xia W.
        • Hammerberg C.
        • Li Y.
        • et al.
        Expression of catalytically active matrix metalloproteinase-1 in dermal fibroblasts induces collagen fragmentation and functional alterations that resemble aged human skin.
        Aging Cell. 2013; 12: 661-671