Skin-punch biopsies were obtained with informed consent under an Institutional Review Board-approved protocol at the University of Pittsburgh from the clinically affected and unaffected skin of six patients with SSc and five healthy twins from an existing twin cohort 1112. Healthy twins were used as controls since they share the genetic background as the SSc patients. All SSc patients had diffuse skin thickening and met the American College of Rheumatology preliminary criteria for classification as SSc 1.

Biopsies were performed on the leading edge of dermal thickening and clinically normal skin. The skin samples were minced, placed in 60 mm tissue culture dishes, and cultured at 37°C in a humidified atmosphere in DMEM (Cellgro, Herndon, VA, USA) supplemented with 10% fetal bovine serum (Sigma-Aldrich, St Louis, MO, USA), 100 IU/ml penicillin, and 100 μg/ml streptomycin (Invitrogen, Carlsbad, CA, USA).

Serum was obtained from postmenopausal patients with diffuse cutaneous SSc (n = 68) and from age-matched and sex-matched healthy controls (n = 35). Both groups had no exposure to HRT. The average age of the SSc patients was 67.6 ± 5.2 years and that of controls 66 ± 0.84 years (not significant). Patients with SSc had disease duration <3 years, with onset defined as the time of the first symptom attributable to SSc.

Skin fibroblasts (2×10⁵ cells per well) were seeded in 35 mm cell culture dishes in DMEM/10% fetal bovine serum. The following day, the medium was replaced with phenol-red free DMEM (Cellgro) without serum for 24 hours to deprive the cells of estrogen. Fresh phenol red-free DMEM plus 10% charcoal-stripped fetal bovine serum (Hyclone, Logan, UT, USA) was added with one of the following: ethanol as vehicle control (0.1%) or E2 (10 nM; Sigma-Aldrich) for 24 hours (for RNA extraction) or 48 hours (for protein extraction). Transforming growth factor beta (10 ng/ml; R&D Systems, Minneapolis, MN, USA) was used as a positive control. ICI 182,780 (100 nM; Tocris, Ballwin, MO, USA), a pure ER antagonist, and signaling inhibitors (MEK inhibitor U0126, inositol-1,4,5-triphosphate (PI3K) inhibitor LY294002, and p38 mitogen-activated protein kinase (MAPK) inhibitor SB202190, 10 μM each; Cell Signaling Technology, Beverly, MA, USA) were added where indicated. To determine the role of ERα and ERβ on FN individually, cells were cultured with propyl-pyrazole-triol (PPT), an ERα specific ligand (100 nM; Tocris) 13, and genistein, an ERβ selective ligand (100 nM; Sigma-Aldrich), under similar conditions to those used for E2 treatment.

ECM was extracted as we have described previously 14. Briefly, cells were rinsed with PBS and incubated with 8 M urea in PBS for 20 minutes. Cells were aspirated and the ECM was rinsed three times with PBS. ECM from an equal number of cells was scraped in 100 μl sample buffer (20 mM dithiothreitol, 6% SDS, 0.25 M Tris, pH 6.8, 10% glycerol, 10 mM NaF, and bromophenyl blue) and analyzed by western blot. Equal volumes of ECM were loaded in each lane.

Skin fibroblasts in early passage (passages 3 and 4) were harvested and RNA was extracted using TRIzol (Invitrogen). mRNA was reverse transcribed using Superscript II (Invitrogen) following the manufacturer's recommendations. The cDNA generated was used as a template for amplification by PCR with primers specific for FN, 5'-ACCGTGTGGGTACAGGTG-3' and 5'-GTCACAGAGGCTACTAT-3', and β-actin, 5'-ATGTTTGAGACCTTCAACAC-3' and 5'-CACGTCACACTTCATGATGG-3'. PCR amplification was performed in a 50 μl reaction containing Taq DNA polymerase (Invitrogen), 10× PCR buffer (750 mM Tris-HCl, pH 8.8, 200 mM (NH₄)₂SO₄ and 0.1% Tween 20), 1.5 mM MgSO₄, and 1 mM of each deoxynucleotide triphosphate in a Peltier Thermal Cycler-200 (MJ Research, Waltham, MA, USA).

Conditions were an initial denaturation at 95°C for 4 minutes, followed by 35 cycles of 94°C for 45 seconds, 55°C for 30 seconds, and 68°C for 2 minutes. Final extension was at 68°C for 5 minutes. Then 20 μl each reaction was electrophoresed on a 1% agarose gel in 1× Tris/acetate/ethylenediamine tetraacetic acid buffer and products were visualized following staining with ethidium bromide. The molecular weights of the PCR products were: FN 513 bp and β-actin 494 bp.

Cells were grown to confluency in 35 mm culture dishes. Cells were rinsed with 1× PBS and scraped in sample buffer (20 mM dithiothreitol, 6% SDS, 0.25 M Tris, pH 6.8, 10% glycerol, 10 mM NaF, and bromophenyl blue). Samples were separated by electrophoresis on 8% SDS-polacrylamide gels and transferred to nitrocellulose membranes. Membranes were blocked with 5% nonfat milk in 1× TBS-Tween 20 (0.2 M Tris, 0.14 M NaCl, 0.1% Tween 20), followed by incubation with mouse monoclonal anti-human EDA-FN antibody, rabbit polyclonal anti-human FN antibody, rabbit polyclonal anti-ERα antibody, rabbit polyclonal anti-ERβ antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA), mouse monoclonal anti-human vitronectin, mouse monoclonal anti-β-actin (Sigma-Aldrich), or mouse monoclonal anti-GAPDH (Ambion, Austin, TX, USA) in 1× TBS-Tween 20. Membranes were then incubated with horseradish peroxide-conjugated donkey anti-rabbit IgG (Amersham, Piscataway, NJ, USA) or donkey anti-mouse IgG (Amersham). Immunoreactive proteins were detected by chemiluminescence (PerkinElmer Life Sciences, Boston, MA, USA), followed by autoradiography.

Human abdominal skin was obtained from cosmetic plastic surgery. All tissues were obtained according to the guidelines of the University of Pittsburgh and under a protocol approved by the Institutional Review Board of the University of Pittsburgh. As described previously 15, subcutaneous fat tissue was removed uniformly and samples composed of complete epidermal and dermal strata were cut into 1.5 cm×1.5 cm sections. Skin was maintained in organ culture in the presence of the indicated factors, E2 (10 nM), ICI 182,780 (100 nM), PPT (100 nM), and genistein (100 nM). Skin was harvested, fixed in 10% formalin, and embedded in paraffin.

Dermal and collagen bundle thickness were measured in skin sections stained with H & E. Dermal thickness was defined as the distance from the granular layer to the junction between the dermis and subcutaneous fat. Images were taken on a Nikon Eclipse 800 microscope (Nikon Instruments, Inc., Huntley, IL, USA) using identical camera settings, and ImageJ (National Institutes of Health, Bethesda, MD, USA) was used to measure thickness. Thickness was measured in five random fields in each sample.

Sections (6 μm) of paraffin-embedded skin tissues were de-paraffinized, endogenous peroxidase was quenched using 10% H₂O₂, and endogenous biotin was blocked using the biotin blocking kit (Dakocytomation, Carpinteria, CA, USA). The sections were blocked with 5% serum and incubated with anti-FN antibody followed by secondary antibody. Bound secondary antibody was detected using the aminoethyl carbazole Red kit (Invitrogen, Carlsbad, CA, USA). A light hematoxylin counterstain was used to identify nuclei. Images were taken on a Nikon Eclipse 800 microscope.

Serum levels of E2 and estrone were measured using liquid chromatography-tandem mass spectrometry in the Small Biomolecule Core Facility in the School of Pharmacy at the University of Pittsburgh. The liquid chromatography-tandem mass spectrometry method employs liquid-liquid extraction, derivatization, and detection with a triple quad mass spectrometer using 0.5 ml serum.

For the in vitro and ex vivo data, statistical comparisons were performed using the Mann-Whitney U test. For the comparison of serum levels of E2 and estrone, two separate sets of analyses were performed: case versus control comparisons of estrone and E2; and case-only comparisons of clinical manifestations based on high, intermediate, and low estrone or E2. For these comparisons, the Wilcoxon rank-sum test, the chi-square test of proportions, and Fisher's exact test were used where appropriate.

The effect of E2 on FN expression was examined using RT-PCR and western blot analysis. In untreated samples, FN mRNA and protein levels in SSc patient fibroblasts were higher than those in their healthy twins. E2 increased FN mRNA and protein levels in healthy twin and SSc fibroblasts (Figure 1A,B). E2 increased FN mRNA and protein levels in a time-dependent and dose-dependent manner in cell supernatants and ECM (Figure 1C,D). E2 induced production of total FN and EDA-domain-containing matrix FN (Figure 1C) and the increase in secreted FN was significant (Figure 1E). The ER antagonist ICI 182,780 blocked the effect of E2 on FN mRNA and protein expression but did not affect transforming growth factor beta-induced FN levels (Figure 1F).

To investigate the mechanism mediating E2 induction of FN, we pretreated skin fibroblasts with vehicle, MEK inhibitor, PI3K inhibitor, or p38 MAPK inhibitor for 1 hour prior to the addition of E2. FN protein levels were assessed by western blot analysis 48 hours post treatment. PI3K inhibitor and p38 MAPK inhibitor attenuated the E2-mediated increase of FN (Figure 2A). MEK inhibitor had a more modest effect on E2 induction of FN. We also examined the effect of the chemical inhibitors on ERα and ERβ. ERα was increased by E2 and this increase was blocked by PI3K inhibitor, p38 MAPK inhibitor, and MEK inhibitor. There was no significant difference in the expression of ERβ under the same conditions (Figure 2A).

To assess the individual effects of ERα and/or ERβ on FN expression, we used PPT, an ERα ligand, and genistein, an ERβ ligand. Primary fibroblasts were treated with vehicle, E2, PPT, or genistein for 48 hours. ECM was harvested and analyzed by western blot. Vitronectin was detected as an ECM loading control. E2 and PPT increased FN protein levels in the ECM (Figure 2B). Genistein modestly increased FN protein levels (Figure 2B). Vitronectin levels were not altered by any of the treatments.

To further examine the effect of E2 in skin tissues, the dermal and collagen bundle thicknesses in dermis were assessed using an ex vivo organ culture system. Explanted skin tissues on 35 mm well plates were treated with E2, ERα or ERβ agonists (PPT or genistein, respectively), or vehicle (ethanol for E2 and PPT, and dimethylsulfoxide for genistein) for 7 days, and skin sections were stained with H & E. As shown in Figure 3, E2 and PPT induce increase of dermal thickness (Figure 3A) and collagen bundle thickness (Figure 3B) compared with vehicle (dermis: 1.61 ± 0.12, P <0.05 and 1.54 ± 0.05, P <0.05, respectively; collagen bundles: 2.62 ± 0.18, P <0.05 and 1.84 ± 0.15, P <0.05, respectively), and ICI 182,780 blocked the effect of E2. On the contrary, genistein did not induce thickening of dermis or collagen bundles (Figure 4A,B). We also assessed the extent of deposition of FN using immunohistochemistry. As shown in Figure 4, the results of FN deposition in collagen bundles were similar to those for thickness of skin and collagen bundles. E2 thus induces skin fibrosis, and this effect is mediated by ERα.

Patient and control E2 serum samples were divided into low (<5 pg/ml), intermediate (5 to 10 pg/ml), and high (>10 pg/ml) levels (Table 1). Similarly, patient and control estrone serum samples were divided into low (<15 pg/ml), intermediate (15 to 75 pg/ml), and high (>75 pg/ml) levels (Table 1). There was a significant difference between SSc patient and control E2 and estrone levels (P = 0.04 and P = 0.006, respectively). The frequency of the data points is shown in the dot plots of Figure 5. Levels of E2 and estrone were also analyzed by disease specific clinical manifestations occurring at any time during the illness. Although the associations did not reach statistical significance, a larger proportion of patients with high estrone levels (42%) had gastrointestinal involvement compared with those patients with low estrone levels (17%, P = 0.07).