Induction of the Hyaluronic Acid-Binding Protein, Tumor Necrosis Factor-Stimulated Gene-6, in Cervical Smooth Muscle Cells by Tumor Necrosis Factor-α and Prostaglandin E₂

Abstract

Immediately before parturition the cervix undergoes striking changes in structure (ripening) that facilitate dilatation and effacement. Cervical ripening shares many features in common with inflammation-associated tissue remodeling, making it a valuable process to explore with respect to the biochemical events in extracellular matrix restructuring. Cervical ripening can be pharmacologically induced with prostaglandin E₂ (PGE₂). Among the biochemical changes in the cervix at parturition is a marked increase in the hyaluronic acid (HA) content. HA and HA-binding proteins have been implicated in tissue hydration, release of collagenase, and leukocyte migration, but their roles in cervical ripening have not been explored. In the present study we examined the ability of PGE₂ to induce expression of the HA-binding protein, tumor necrosis factor-stimulated gene (TSG)-6, in human cervical smooth muscle cells (hCSMCs) and compared the PGE₂ response to that of tumor necrosis factor-α (TNF-α), an established inducer of TSG-6. TNF-α stimulated TSG-6 mRNA accumulation in a dose- and time-dependent manner, with the maximal response observed at 10 ng/ml after 6 hours of incubation. PGE₂ stimulated TSG-6 mRNA expression, but the magnitude of response was substantially less than that produced by TNF-α, and it was maximal only after 24 hours of incubation. Quantitative real-time polymerase chain reaction was performed to assess the induction of TSG-6 mRNA and nascent transcripts at 24 hours of treatment. Induction of TSG-6 mRNA and nascent transcripts in response to 10 μmol/L of PGE₂ was 5.7-fold and 6.3-fold greater than control values, respectively, whereas TNF-α (10 ng/ml) induced TSG-6 mRNA and nascent transcripts by 80-fold and 134-fold, respectively. TNF-α and PGE₂ stimulated secretion of TSG-6 into the culture medium as detected by Western blotting. The effects of PGE₂ on secretion of TSG-6 were delayed compared to TNF-α. A 1.3-kb fragment of the human TSG-6 proximal promoter drove luciferase expression in transfected hCSMCs. PGE₂ increased TSG-6 promoter activity 1.75-fold. Paradoxically, TNF-α reduced TSG-6 promoter activity by 50%. We conclude that hCSMCs express the hyaladherin TSG-6; that TSG-6 expression in these cells is regulated by PGE₂ as well as proinflammatory cytokines; responses of hCSMCs to TNF-α and PGE₂ are distinct in terms of magnitude and the time course; and PGE₂ and TNF-α exert different effects on the TSG-6 proximal promoter.

The ripening of the human cervix at term has been likened to an inflammatory process. The study of the biochemical events underlying this process could reveal important features of normal and pathological tissue remodeling. A substantial increase in the cervical content of hyaluronic acid (HA) at the onset of labor is thought to contribute to the inflammation-like remodeling of the cervix. The accumulation of HA promotes tissue hydration, 1-5 release of collagenase, 6,7 and leukocyte migration. 8,9 These actions of HA are probably mediated and/or modulated by hyaladherins, HA-binding proteins. Hyaladherins share a common motif, the link module, that confers binding affinity to HA. 10-12 Relatively little is known about the hyaladherins produced in the cervix during the ripening process. The inflammation-associated protein TSG-6 (tumor necrosis factor stimulated gene-6) is a hyaladherin expressed by fibroblasts, chondrocytes, monocytes, vascular smooth muscle cells, and cumulus-oocyte complexes in response to inflammatory mediators and growth factors. 12-15 The function of TSG-6 is not well understood, but it has been suggested to be an endogenous inhibitor of neutrophil migration, an anti-inflammatory molecule, and a modulator of extracellular matrix remodeling. 16,17 These activities are potentially important in the cervical ripening process that is characterized by infiltration of leukocytes and extensive reorganization of the cervical extracellular matrix.

In the present study we examined the expression of TSG-6 by human cervical smooth muscle cells (hCSMCs) in culture, focusing on the effects of molecules that are known to promote cervical ripening, proinflammatory cytokines, and prostaglandin E₂ (PGE₂).

Materials and Methods

Cell Culture

Proliferating human cervical smooth muscle cells (hCSMCs) (Clonetics, San Diego, CA) that we have previously demonstrated to retain the molecular phenotype of smooth muscle cells, 18 were grown in Smooth Muscle Cell Growth Medium-2 basal medium (Clonetics) supplemented with 5% fetal bovine serum. This medium contains human epidermal growth factor (0.5 ng/ml), human fibroblast growth factor (1.0 ng/ml), insulin (5 μg/ml), gentamicin (50 μg/ml), and amphotericin B (50 μg/ml). hCSMCs were cultured at 37°C under an atmosphere of 5% CO₂ in air. Subcultures of hCSMCs from passages 3 to 6 were used in all of the experiments. Each experiment was reproduced on at least two occasions with similar results.

Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR) and TSG-6 Probe Preparation

Five μg of total RNA extracted from hCSMCs was reverse-transcribed to produce cDNA using reverse transcriptase and oligo dT (Promega, Madison, WI) as a primer. The TSG-6 cDNA was amplified using 10% of the RT reaction in 100 μl containing 50 pmol of forward primer, 50 pmol of reverse primer, 5 U of Taq polymerase (Life Technologies, Inc., Grand Island, NY) with 0.2 mmol/L of dNTPs and 1.5 mmol/L of MgCl₂. The sequences of the primers used to generate the TSG-6 probe were: forward primer 5′-ATTTGTGAGCAGCCCCTAAC-3′ and reverse primer 5′-AGTGAGATCAAAGGAGTTCC-3′. The expected size of the PCR product is 951 bp. PCR was performed in a 9600 GeneAmp PCR thermal cycler using the following conditions: 94°C (1 minute) for one cycle, 94°C (1 minute), 57°C (1 minute), 72°C (1.5 minutes) for 35 cycles, and a final incubation at 72°C for 7 minutes. The PCR product was subcloned into PCR 2.1 vector (Invitrogen, Carlsbad, CA) and sequenced to verify that it encoded the TSG-6 cDNA. To obtain the insert for Northern blotting, the plasmid containing the sequence-verified PCR product was digested with EcoRI and the insert was purified using a gel extraction kit (Qiagen, Valencia, CA) before labeling by the random primer method.

RNA Isolation and Northern Blot Analysis

Subconfluent cultures of hCSMCs were grown in medium supplemented with 1% fetal bovine serum for 24 hours before treatment with tumor necrosis factor (TNF)-α (0.01 to 20 ng/ml) (R&D System, Minneapolis, MN) or PGE₂ (5 to 40 μmol/L) (Sigma Chemical Co., St. Louis, MO) for up to 72 hours. Total RNA was extracted from the cultures with Trizol reagent (Life Technologies, Inc.) using procedures recommended by the manufacturer. Equal amounts of RNA (30 to 50 μg/lane) were separated on 1% agarose-formaldehyde-denaturing gels and transferred to nylon membranes. Membranes were hybridized with ³²P-labeled TSG-6 cDNA at 42°C for 16 to 18 hours, followed by two sequential washings for 15 minutes in 2× sodium chloride/sodium phosphate ethylenediaminetetraacetic acid (SSPE), 0.1% sodium dodecyl sulfate (SDS) at 37°C, and two washings in 0.1× SSPE, 0.1% SDS at 55°C. Blots were analyzed using a phosphoimager and then exposed to Kodak X-Omat AR film (Eastman-Kodak, Rochester, NY). The relative abundance of TSG-6 mRNA was normalized to 28S rRNA.

Effects of Actinomycin D (Act D) and Cycloheximide (CHX) on TNF-α- and PGE₂-Induced mRNA Expression

Subconfluent hCSMCs were cultured with 2 μg/ml of Act D or 50 μg/ml of CHX with 10 ng/ml of TNF-α or 10 μmol/L of PGE₂ for 6 or 24 hours. Total RNA was subjected to Northern blotting as described above.

Western Blot Analysis

hCSMCs were cultured in serum-free medium for 24 hours, and then treated with TNF-α (10 ng/ml) for 24 hours or PGE₂ (5 to 40 μmol/L) for up to 72 hours. The conditioned medium was subjected to 10% SDS-polyacrylamide gel electrophoresis and then Western blotting. The polyclonal antiserum used in Western blotting was raised as described in Carrette and colleagues 15 except that a 17-amino acid peptide (CTSTGNKNFLAGRFSHL) corresponding to the final 16-amino acid residues of human TSG-6 19 and a nonauthentic N-terminal Cys residue was used as the antigen. Supersignal West Femto Maximum Sensitivity Substrate (Pierce, Rockford, IL) was used to detect antibody bound to antigen.

TSG-6 Promoter Analysis

The promoter region of the human TSG-6 gene was amplified by PCR using a forward primer spanning bp −1320 to −1301 with a KpnI linker (5′-CGAGGTACCTCACTAACCCTATCTGTGAA-3′) and a reverse primer spanning bp −20 to −1 with a NheI linker (3′-CTACACCTTTGGTCTACAAACGATCGAGC-5′). The 1.32-kb fragment spanning bp −1320 to −1 of the human TSG-6 gene was cloned into the pGL3 basic vector (Promega), which contains firefly luciferase as a reporter gene. 5′-deletion constructs were prepared by the subcloning of PCR products generated using various forward TSG-6-specific primers with attached linkers and the reverse primer described above. The forward primer used to generate the −756 TSG-6 promoter fragment spanned from bp −756 to −737 with a KpnI linker (5′-CGAGGTACCCCTTGATCGTCTTCTTCAAA-3′); the forward primer for the −160 TSG-6 promoter fragment spanned from bp −160 to −141 with a KpnI linker (5′-CGCGGTACCATTCTATCTCCTTAGTTTTG-3′); the forward primer for the −100 TSG-6 promoter fragment spanned from bp −100 to −81 with a KpnI linker (5′-CGCGGTACCTGAGATAATTGTGTAACTCT-3′). The sequences of all of these PCR-generated promoter fragments were confirmed to be identical to the corresponding TSG-6 promoter sequence previously reported by Lee and colleagues 20 with the following exceptions that insertions of A, G, C, T, A, C, and TG between nucleotides −1292 and −1291, −867 and −866, −681 and −680, −678 and −677, −631 and −630, −472 and −471, and −53 and −52, respectively, were consistently found. These same sequence variations were identified independently in Dr. Day’s laboratory and have been deposited in EMBL with accession codes of AJ413948 and AJ413949.

Transfection and Cell Stimulation

hCSMCs (1 × 10⁵) were seeded in each well of 12-well plastic culture plates in medium supplemented with 1% fetal bovine serum. Cells were transfected using FuGENE 6 Transfection Reagent (Roche, Indianapolis, IN) with 0.5 μg/ml of each of the promoter constructs, 0.5 μg/ml of the CMV-β galactosidase plasmid, and 2 μl/ml of FuGENE. After 24 hours, hCSMCs were incubated in 0.1% ethanol (control), 10 ng/ml TNF-α, or 10 μmol/L PGE₂ in medium supplemented with 1% fetal bovine serum for 24 hours to determine the effect on TSG-6 promoter activity. At the end of the culture period, cells were harvested and the cell lysates were assayed for luciferase and β-galactosidase activity as described below.

Luciferase and β-Galactosidase Assay

Luciferase activity was determined in a LUMAT LB 9507 luminometer (EG&G Berthold, Gaithersburg, MD) with Promega luciferin as substrate. β-galactosidase activity was determined by a standard colorimetric assay using 2-nitrophenyl β-d-galactopyranoside as substrate. The luciferase assay results were normalized to β-galactosidase activity to correct for variations in transfection efficiency. Each treatment group contained triplicate cultures and each experiment was repeated three to four times. Relative luciferase units (RLU) defined as luciferase light units/β-galactosidase activity are presented as means ± SE.

Reverse Transcription Reaction and Quantitative Real-Time PCR for TSG-6 mRNA

Proliferating hCSMCs were incubated with 0.1% ethanol vehicle or 10 μmol/L of PGE₂ for 24 hours. Total RNA was extracted using Trizol reagent (Life Technologies, Inc.). To limit the possibility of detection of genomic DNA, total RNA was treated with RQ1-RNase-free DNase (Promega) for 30 minutes at 37°C before reverse transcription with Moloney murine leukemia virus-reverse transcriptase (Promega) as described by the manufacturer. Five μg of total RNA was reverse-transcribed to single-strand cDNA using 1 μl of oligo(dT)15 primer (Promega), 0.5 μl of RNase inhibitor from human placenta (Roche Molecular Biochemicals), and 1 μl of Moloney murine leukemia virus-reverse transcriptase (Promega) as described by the manufacturer.

Quantitative real-time PCR was performed to assess the induction of TSG-6 mRNA in hCSMCs in response to 10 μmol/L of PGE₂. Primers for the analysis of the human TSG-6 gene elements were designed in the third exon: forward, 5′-TCATGTCTGTGCTGCTGGATG-3′; and reverse, 5′-GGGCCCTGGCTTCACAA-3′. The resulting cDNA was diluted 10-fold in sterile water and aliquots were subjected to quantitative real-time PCR. The real-time PCR reaction used 90 nmol/L of each primer and 12.5 μl of SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA). Agarose gel electrophoresis indicated the presence of a 67-bp single band for the TSG-6 PCR product. To account for differences in starting material, the human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) primers and probe reagents from Applied Biosystems were used as described by the manufacturer. The target and GAPDH PCR reactions were performed in separate tubes in triplicate and the average threshold cycle for the triplicate was used in all subsequent calculations.

Quantitative Real-Time PCR of TSG-6 Nascent Transcripts

To limit the possibility of detection of genomic DNA, 5 μg of total RNA was treated with RQ1-RNase-free DNase (Promega) as described above. For the human TSG-6 and human GAPDH nascent transcript expression, primers for reverse transcription were designed in the third and fourth intron of each gene, respectively; TSG-6: 5′-CAACCTCTTAGCAGCATGGAACTGT-3′, GAPDH: 5′-TAGTTGCCTCCCCAAAGCAC-3′. Five μg of total RNA was reverse-transcribed to single-stranded cDNA as described above. The resulting amplicons derived from the specific TSG-6 and GAPDH intronic primers were diluted 10-fold in sterile water and aliquots were subjected to quantitative real-time PCR. The forward and reverse PCR primers for TSG-6 nascent RNA expression were the same as those described above for TSG-6 mRNA expression. For GAPDH nascent transcript expression, PCR primers were designed in the third exon (forward, 5′-GATTCCACCCATGGCAAATT-3′; reverse, 5′-AAGATGGTGATGGGATTTCCATT-3′). Optimization studies for detection of the nascent GAPDH transcript indicated that 90 nmol/L of each primer should be used in each 25-μl reaction. Agarose gel electrophoresis revealed the presence of a 83-bp single band for the nascent GAPDH transcript PCR product. PCR reactions for nascent TSG-6 and GAPDH transcripts were performed in separate tubes in triplicate and the average threshold cycle for the triplicates was used in all subsequent calculations.

Results

TNF-α and PGE₂ Increase TSG-6 Gene Expression by hCSMCs

TNF-α is an established inducer of TSG-6 expression. We first defined the response of the hCSMCs to this proinflammatory cytokine. TNF-α had a dose-dependent stimulatory effect on expression of the 18S TSG-6 mRNA (Figure 1) ▶ . As little as 0.1 ng/ml of TNF-α raised TSG-6 message abundance, and a maximal response was observed with 10 ng/ml. Time-course studies conducted with 10 ng/ml of TNF-α (Figure 2) ▶ revealed a rapid increase in TSG-6 mRNA, detectable after 0.5 hours of treatment, the first time point we examined, and reaching a maximum at 6 hours, with mRNA levels declining modestly thereafter.

Figure 1.

Effect of TNF-α on expression of TSG-6 mRNA. A Northern blot analysis of total RNA (40 μg/lane) extracted from hCSMCs cultured in the absence (control) or presence of different concentrations of TNF-α for 48 hours is shown. The blot was sequentially probed to detect 28S rRNA.

Figure 2.

Time course of TNF-α-induced expression of TSG-6 mRNA. A: Northern blot analysis of total RNA (40 μg/lane) extracted from hCSMCs cultured in the absence (control) or presence of TNF-α (10 ng/ml) for the indicated time periods is shown. The blot was also probed for 28S rRNA. B: Histogram showing induction of TSG-6 mRNA normalized to 28S rRNA relative to the control value that was arbitrarily set to 1.0. The results are representative of two separate studies.

PGE₂ also caused a dose-dependent increase in TSG-6 mRNA expression at 5 to 20 μmol/L; a lesser response was observed with higher PGE₂ concentrations of 30 and 40 μmol/L (Figure 3) ▶ that may reflect desensitization or cell demise. In contrast to TNF-α, which rapidly induced TSG-6 mRNA, the response of hCSMCs to PGE₂ was more sluggish. A modest response was seen at 6 hours and a peak in TSG-6 mRNA was observed at 24 hours of treatment (Figure 4) ▶ . Notably, TSG-6 mRNA induction by PGE₂ after 24 hours of treatment was substantially lower than that observed with 10 ng/ml of TNF-α based on the signal strength of simultaneously hybridized blots and blots containing RNA from TNF-α- and PGE₂-treated cells.

Figure 3.

Effect of PGE₂ on expression of TSG-6 mRNA. A Northern blot analysis of total RNA (40 μg/lane) extracted from hCSMCs cultured in the absence (control) or presence of different concentrations of PGE₂ or TNF-α (10 ng/ml) for 24 hours is shown. The blot was sequentially probed with 28S rRNA.

Figure 4.

Time course of PGE₂-induced TSG-6 mRNA expression. A: Northern blot analysis of total RNA (40 μg/lane) extracted from hCSMCs cultured in the absence (control) or presence of PGE₂ (10 μmol/L) for up to 72 hours is shown. The blots were sequentially probed for 28S rRNA. B: Histogram showing induction of TSG-6 mRNA normalized to 28S rRNA relative to the control value that was arbitrarily set to 1.0. Values are means ± SE from five separate experiments.

Quantitative Real-Time PCR for TSG-6 mRNA and Nascent Transcripts

We performed quantitative real-time PCR to assess the induction of TSG-6 mRNA and nascent transcripts, the latter being an index of TSG-6 gene transcription, in hCSMCs treated with 10 μmol/L of PGE₂ for 24 hours (Table 1) ▶ . The increase in TSG-6 mRNA and nascent transcripts in response to 10 μmol/L of PGE₂ was 5.7-fold and 6.3-fold compared to the 0.1% ethanol control group, respectively. In simultaneously conducted experiments TNF-α treatment increased TSG-6 mRNA and nascent transcript levels by ∼80-fold and 134-fold, respectively. These observations are consistent with the significant differences in TSG-6 induction in response to the prostanoid and the cytokine demonstrated by Northern blotting. We also examined the effects of 8-Br-cAMP because PGE₂ is thought to act on hCSMCs via the EP₄ receptor, which is coupled to adenylate cyclase. 21 The response to 8-Br-Br-cAMP was basically similar to that produced by PGE₂.

Table 1.

Effect of PGE₂ on TSG-6 mRNA and Nascent Transcripts

	Fold increase of TSG-6 mRNA and nascent transcripts ratio
		PGE2/ control	TNF-α/ control	8-Br-cAMP/ control
mRNA	5.7 ± 1.2	79.5 ± 16.5	4.3 ± 0.6
Nascent transcripts	6.3 ± 2.6	133.7 ± 50.4	2.9 ± 0.5

Quantitative real-time PCR was used to measure the fold increase in TSG-6 mRNA and nascent transcripts in hCSMCs treated with PGE₂ (10 μM), TNF-α (10 ng/ml), or 8-Br-cAMP (1 mmol/L) or the ethanol vehicle. The values presented are from three independent experiments, each conducted in triplicate.

We performed several experiments to determine whether PGE₂ (10 μmol/L) acts synergistically with TNF-α at either low concentrations (0.1 ng/ml) or concentrations that maximally stimulate TSG-6 expression (10 ng/ml). Cells were treated with each agent alone or in combination for 24 hours. We found no evidence for PGE₂ augmentation of the TSG-6 mRNA response to TNF-α at either low or high TNF-α concentrations (data not shown).

TNF-α and PGE₂ Stimulate TSG-6 Secretion, But with Different Temporal Patterns

Exposure of hCSMCs to TNF-α at 10 ng/ml for 24 hours resulted in the release of the 39-kd TSG-6 protein into the culture medium as detected by Western blotting (Figure 5) ▶ . In contrast, there was no detectable release of TSG-6 by hCSMCs exposed to 10 μmol/L of PGE₂ for 24 hours. However, when hCSMCs were cultured with 10 μmol/L of PGE₂ for up to 72 hours, detectable TSG-6 was present in the conditioned medium by 48 hours of incubation (Figure 6) ▶ . The effects of PGE₂ on secretion of TSG-6 after 72 hours of culture were dose-dependent with maximal secretion observed with 10 and 20 μmol/L of PGE_2, and as with TSG-6 mRNA expression, lower production of TSG-6 was observed with concentrations of 30 and 40 μmol/L (Figure 7) ▶ .

Figure 5.

TSG-6 production by hCSMCs treated with TNF-α or PGE₂. hCSMCs were incubated with TNF-α at 10 ng/ml or PGE₂ at 10 μmol/L for 24 hours in serum-free medium and aliquots of conditioned media (80 μl/lane) were subjected to SDS-polyacrylamide gel electrophoresis and then Western blotting.

Figure 6.

Time course of TSG-6 production by hCSMCs treated with PGE₂. hCSMCs were incubated without or with PGE₂ (10 μmol/L) for up to 72 hours in serum-free medium and aliquots of conditioned media (80 μl/lane) were subjected to SDS-polyacrylamide gel electrophoresis and then Western blotting.

Figure 7.

TSG-6 production by hCSMCs treated with different concentrations of PGE₂. hCSMCs were incubated in the absence or presence of different concentrations of PGE₂ for 72 hours in serum-free medium and aliquots of conditioned media (80 μl/lane) were subjected to SDS-polyacrylamide gel electrophoresis and then Western blotting.

TSG-6 Expression in Response to TNF-α and PGE₂ Requires On-Going RNA But Not Protein Synthesis

The TNF-α- and PGE₂-stimulated rise in TSG-6 mRNA was not dependent on on-going protein synthesis as it occurred in the presence of the inhibitor, CHX that blocks de novo protein synthesis by >95% at the concentration used in this study (50 μg/ml) (Figure 8) ▶ . In fact, CHX alone caused a modest increase in TSG-6 mRNA (2.4-fold greater than control, normalized to 28S rRNA) and augmented the effects of PGE₂ (4.3-fold at 6 hours of treatment and 14.67-fold at 24 hours), but not those of TNF-α. However, TNF-α and PGE₂-stimulated TSG-6 mRNA expression required on-going RNA synthesis as Act D completely blocked the rise in TSG-6 mRNA in both TNF-α- and PGE₂-treated hCSMCs without affecting basal levels at 6 hours of treatment and reducing basal mRNA abundance normalized to 28S rRNA by ∼60% after 24 hours of treatment.

Figure 8.

Effect of CHX and Act D on TNF-α and PGE₂ induction of TSG-6 mRNA in hCSMCs. A: Northern blot analysis of total RNA (50 μg/lane) extracted from hCSMCs treated with 10 ng/ml of TNF-α, 10 μmol/L of PGE₂, 50 μg/ml of CHX, 2 μg/ml of Act D, or the combination of these for 6 hours is shown. The blot was sequentially probed for 28S rRNA. B: hCSMCs were exposed to the same treatments for a period of 24 hours and Northern blot analysis was conducted as described above.

PGE₂ Stimulates TSG-6 Promoter Activity

To examine the regulation of TSG-6 gene transcription, we cloned the TSG-6 promoter from hCSMC genomic DNA. A 1.3-kb fragment of the human TSG-6 proximal promoter drove luciferase expression in transfected hCSMCs. Treatment of the cells with 10 μmol/L of PGE₂ after 24 hours of incubation increased TSG-6 promoter activity 1.75-fold (Table 2) ▶ . Paradoxically, TNF-α (10 ng/ml) reduced TSG-6 promoter activity by 64%, indicating that PGE₂ and TNF-α regulate TSG-6 transcription by different mechanisms. We also examined four 5′-deletion constructs (Figure 9) ▶ . Ten μmol/L of PGE₂ increased the activity of the −756 and −160 constructs, but not the −100 construct that had only modest promoter activity compared to the empty pGL3 vector. Notably removal of sequences between positions −1320 bp and −756 bp tended to enhance promoter function. A similar result has been reported by Lee and colleagues 20 who studied TSG-6 promoter activity in human FS-4 foreskin fibroblasts.

Table 2.

Transcriptional Activation of TSG-6 Promoter in Response to TNF-α and PGE₂

Treatment	RLU × 10−3 (mean ± SE)
Control	35.71 ± 2.1
TNF-α (10 ng/ml)	22.71 ± 2.9
PGE2 (10 μM)	64.42 ± 3.6

Effect of TNF-α and PGE₂ on 1.32-kb TSG-6 promoter activity in hCSMCs. Transfection and cell stimulation were performed as described in Materials and Methods. The relative luciferase units (RLU) presented are from three independent experiments, each conducted with triplicate cultures. Values with different superscripts are significantly different (P < 0.05) by the Tukey-Kramer test.

Figure 9.

Analysis of TSG-6 promoter function. The indicated promoter constructs were transfected into hCSMCs and incubated with 0.1% ethanol or 10 μmol/L of PGE₂ as described in Materials and Methods. The relative light units (RLU) presented are means ± SE from four independent experiments, each conducted with triplicate cultures.

Discussion

Cervical ripening during parturition involves biochemical and cellular changes that are similar to those occurring in tissue inflammation. 22 The ripening process is associated with a marked increase in the level of HA. 23-25 Proinflammatory cytokines and PGE₂ induce glycosaminoglycan production by cervical fibroblasts. 26-28 HA promotes tissue hydration, 1-5 release of collagenase, 6,7 and leukocyte migration, all components of the cervical ripening process. 8,9 Collectively, these observations suggest that HA plays an important role in the process of cervical ripening.

Although the functions of HA in the cervix during parturition are becoming clear, there has been little attention given to the HA-binding proteins that are thought to mediate and/or modulate the actions of HA. TSG-6 is one of the HA-binding proteins containing a single link module composed of two α-helices and two triple-stranded β-sheets arranged around a large hydrophobic core. 11 TSG-6 interacts with HA via its link module and amino acid residues involved in ligand binding have recently been determined. 17,29 TSG-6 has been implicated in the regulation of leukocyte migration and its pattern of expression suggests that it may be involved in extracellular matrix remodeling. 16,30 Elevated TSG-6 protein levels have been found in the synovial fluid of arthritic patients, 31 and recombinant TSG-6 has potent anti-inflammatory activity in vivo. 16 Based on these results, it was suggested that TSG-6 constitutes part of a cytokine-initiated feedback loop that operates to down-regulate the inflammatory response. 16 TSG-6 has been shown to form a stable, probably covalent, 120-kd complex with the serine protease inhibitor, inter-α inhibitor (IαI). 32 It has been reported that this complex has significantly increased anti-plasmin activity over IαI alone. 16 Thus TSG-6 may be important in regulating leukocyte migration and matrix remodeling because plasmin enhances the activation of latent metalloproteinases involved in extracellular matrix breakdown. 13,16,33

Our observations demonstrate that hCSMCs express TSG-6, and that TSG-6 expression in these cells is regulated, as expected, by proinflammatory cytokines. However, we have also established for the first time that PGE₂ controls the TSG-6 gene. Notably, the responses of hCSMCs to TNF-α and PGE₂ are distinct in terms of both magnitude and time course. We found that TNF-α and PGE₂ treatment of hCSMCs increased nascent TSG-6 transcripts demonstrating that the rise in TSG-6 mRNA is at least in part the result of increased TSG-6 transcription. This observation is consonant with the fact that Act D blocked the induction of TSG-6 mRNA. Yet, the effect of PGE₂ and TNF-α on TSG-6 proximal promoter activity suggest that the regulatory elements responsible for increased transcription are different. This is not unanticipated because PGE₂ acts on cervical fibroblasts through the EP₄ receptor that couples to adenylate cyclase, 21 whereas proinflammatory cytokines activate receptors that do not immediately engage the cAMP signaling cascade. The ability of cAMP-mediated signaling pathways to influence TSG-6 expression was verified by demonstrating that 8-Br-cAMP (1 mmol/L for 24 hours) increased TSG-6 mRNA and nascent transcript accumulation. The modest increase in TSG-6 mRNA abundance in CHX-treated cells and the augmentation of the response to PGE₂ may suggest an effect of CHX on TSG-6 mRNA stability or possibly inhibition of synthesis of a transcriptional silencer. The fact that CHX did not block the effects of PGE₂ on TSG-6 gene expression argues against the involvement of a prostanoid-induced protein autocrine factor in the TSG-6 response.

Lee and colleagues 20 reported that sequences between −165 to +78 in the TSG-6 promoter confer regulation by proinflammatory cytokines on this gene. This region contains a potential binding site for AP-1 as well as other transcription factors. However, the transcriptional response of the TSG-6 promoter construct reported by Lee and colleagues 20 in FS-4 cells to TNF-α treatment was modest at best compared to the response to interleukin-1. The magnitude of stimulated promoter activity in the study of Lee and colleagues 20 also did not correspond to the change in TSG-6 mRNA or the results of nuclear run-on assays. Our findings suggest that elements in the TSG-6 gene outside of those included in our promoter constructs and those of Lee and colleagues 20 must contribute to the transcriptional response to TNF-α as well as the maximal response to PGE₂.

The production of TSG-6 in the cervix driven by proinflammatory cytokines and PGE₂, molecules that promote cervical ripening, implicates TSG-6 in the cervical ripening process. Based on existing information, we speculate that TSG-6 modulates the inflammation-like biochemical and cellular changes in the cervix by retaining leukocytes in the cervix while at the same time exerting anti-inflammatory actions and restricting proteolytic activity so that alterations in the extracellular matrix are spatially and temporally regulated.

TSG-6 can be detected in human cervical fluid of women who present with symptoms of preterm labor (T Fujimoto and JF Strauss, unpublished observations). This finding is consistent with the notion that the cervix produces TSG-6 in situ. It also raises the possibility that detection of TSG-6 could serve as marker for impending cervical changes leading to preterm birth.