We analyzed the response of uterine smooth muscle cells to interleukin-1β (IL-1β). We first showed that PHM1-31 myometrial cells, our cellular model, are contractile. To determine the molecular mechanisms of uterine smooth muscle cell activation by proinflammatory cytokines, we performed genechip expression array profiling studies of PHM1-31 cells in the absence and the presence of IL-1β. In total, we identified 198 known genes whose mRNA levels are significantly modulated (> 2.0-fold change) following IL-1β exposure. We confirmed the expression changes for selected genes by independent mRNA and protein analysis. The group of genes induced by IL-1β includes transcription factors and inflammatory response genes such as nuclear factor of κ light polypeptide gene enhancer in B-cells (NFκB), pentraxin-related gene (PTX3), and tumor necrosis factor α-induced protein 3/A20 (TNFAIP3/A20). We also found up-regulation of chemokines like C-X-C motif ligand 3 (CXCL3) and extracellular matrix remodeling signaling molecules like tenascin C (TNC). Our data suggest that IL-1β elicits the rapid activation of a cellular network of genes particularly implicated in inflammatory response that may create a cellular environment favorable for myometrial cell contraction. Our results provide novel insights into the mechanisms of uterine smooth muscle cell regulation and possibly infection-induced preterm labor.
Proinflammatory cytokines play important roles in a variety of reproductive processes including ovulation, implantation, placentation, cervical dilation, and parturition (Saji et al. 2000, Huang 2006, Romero et al. 2006). Recent evidence suggests that cytokines are involved in events that lead to preterm labor and delivery, particularly in association with intrauterine infection (Goldenberg et al. 2000, Park et al. 2005). Preterm birth, defined as infants who are delivered before 37 weeks of gestation, accounts for 6–10% of all births and is a major risk factor for infant morbidity and mortality (Slattery & Morrison 2002). Cervical and intrauterine infection and inflammation in the mother are present in ~30–50% of preterm labor cases (Gomez et al. 1997, Challis et al. 2002). Bacterial infection leads to the activation of macrophages and lymphocytes. The subsequent release of proinflammatory cytokines such as interleukin-1β (IL-1β), IL6, tumor necrosis factor α (TNFα), and the chemotactic protein IL8 attracts additional macrophages as well as eosinophils and neutrophils. The release of proinflammatory cytokines induces phospholipid metabolizing enzymes and the release of prostaglandins (Zaragoza et al. 2006). It has been reported that expression of prostaglandin-endoperoxide synthase 2 gene (PTGS2), a major regulator of parturition, is induced by IL-1β, and this induction is accompanied by a threefold increase in prostaglandin E2 (PGE2; Bartlett et al. 1999, Rauk & Chiao 2000). In addition, cytokine signaling also activates extracellular matrix (ECM) remodeling enzymes and matrix metalloproteinases. Proinflammatory cytokines are present at the maternal–fetal interface throughout human pregnancy and labor (Lappas et al. 2002). Under normal conditions, elevated levels of IL-1β are observed in amniotic fluid only late in pregnancy. In contrast, upon infection, the levels of IL-1β, IL6, TNFα, and IL8 are significantly increased, possibly leading to the early onset of labor (Romero et al. 1989, Saji et al. 2000, Suzuki et al. 2006). Recent data in nonhuman primate and mouse models further support the notion that specifically IL-1β and TNFα play a pivotal role in triggering preterm labor (Hirsch et al. 2006, Sadowsky et al. 2006).
In this study, we wanted to further explore the link between the presence of proinflammatory cytokines and myometrial cell function. In particular, we were interested in uncovering the early response of uterine smooth muscle cells to IL-1β exposure. First, we demonstrated that PHM1-31 uterine smooth cells are able to contract using a recently developed collagen lattice-based retraction assay. We then performed genome-wide expression profiling of PHM1-31 cells following exposure to IL-1β to identify the network of cytokine-activated genes in uterine smooth muscle cells. Our data suggest that myometrial cells specifically respond to IL-1β resulting in gene expression changes that create an endocrine environment that might promote uterine smooth muscle cell contraction.
Induction of contraction in PHM1-31 myometrial cells by oxytocin
PHM1-31 myometrial cells were originally derived from the upper edge of the incision of the uterine segment in a non-labor pregnant woman (Monga et al. 1996). We and others have shown that PHM1-31 cells possess morphological and molecular characteristics typical of myometrial cells (Monga et al. 1996, Massrieh et al. 2006). Recently, a novel in vitro system has been developed to determine whether cells have the ability to contract in response to oxytocin, a potent uterotonic agent (Devost & Zingg 2007). We used this functional assay to examine the PHM1-31 cell model (Fig. 1A). As observed previously for M11 and hTERT-C3 myometrial cells (Devost & Zingg 2007), PHM1-31 cells exhibit ~50% basal contraction. This contraction is specific to muscle cells since we did not detect any contraction with human embryonic kidney HEK 293 cells (data not shown). Furthermore, we found that treatment with oxytocin (100 nM), a uterotonic agent, led to an additional 10% increase in the contraction of PHM1-31 cells.
Genechip array analysis of IL-1β-treated PHM1-31 cells
To better understand the effect of IL-1β on smooth muscle cells, we performed genechip expression array analysis. RNA extracted from control PHM1-31 cells and PHM1-31 cells treated with 1 ng/ml IL-1β for 1 h was converted into cRNA and then hybridized to a HG-U133_Plus_2 Affymetrix genechip. We performed three independent experiments to obtain statistically significant data. A minus versus average plot and principal component analysis plot were done confirming that the three chips analyzed behaved similarly (data not shown). This quality control step is essential to avoid normalization problems in the data set. We performed a robust multi-array average (RMA) analysis and found a set of 198 genes (186 up-regulated and 12 down-regulated genes) showing a ≥2.0-fold change in transcript levels in untreated versus IL-1β-treated PHM1-31 cells. These genes are primarily involved in inflammatory response, transcriptional regulation, cell adhesion, and signal transduction.
Table 1 lists all the genes that are induced at least threefold, whereas Table 2 lists all the genes down-regulated by IL-1β treatment. A cutoff of twofold has been chosen for the down-regulated genes since we did not detect any genes reduced more than threefold upon IL-1β treatment in PHM1-31 cells. Overall, in our microarray, the number of up-regulated genes by IL-1β is greater than the number of down-regulated genes. This was not unexpected since previous studies in other cell lines have shown a similar pattern (Rossi et al. 2005, Ha et al. 2006). This data could be explained by the fact that IL-1β is an activator of pathways more than a repressor. The heat map in Fig. 2 shows the transcript levels in untreated and IL-1β-treated cells for the three independent sets of experiments to underscore the reproducibility of the data.
Analysis of genes modulated by IL-1β
We selected a series of genes identified in the genechip expression array studies for further analysis. To this end, we used several selection criteria: a transcript-level change of at least threefold, a significant basal gene expression level when compared with background level, as well as a possible link between the particular gene and myometrial cell function, gestation, or preterm labor. Using these criteria, we chose to investigate seven genes indicated in bold in Table 1: chemokine C-X-C motif ligand 3 (CXCL3), pentraxin-related gene (PTX3), tenascin C (TNC), TNFα-induced protein 3/A20 (TNFAIP3/A20), p105 subunit of the nuclear factor of κ light polypeptide gene enhancer in B-cells (NFκB1), v-maf musculoaponeurotic fibrosarcoma oncogene homolog F (MAFF), and plasminogen activator urokinase (PLAU). We confirmed by Northern blot experiments the expression-level changes observed for CXCL3, PTX3, TNC, and PLAU in the microarray studies for 1 h of 1 ng/ml IL-1β treatment (Fig. 3A). We also performed a time course experiment with 1 ng/ml of IL-1β for 0.5–24 h. For all the genes studied, we observed an increase in mRNA levels 24 h after the initial IL-1β treatment with an induction peak at 3 h for the PTX3 gene, at 8 h for the CXCL3 and PLAU genes, and at 24 h for the TNC gene (Fig. 3). For CXCL3, the induction is rapid since it is observed 30 min after IL-1β exposure of PHM1-31 cells (Fig. 3A).
By immunoblot analysis, we observed an induction of NFκB p50 protein as early as 30 min after IL-1β stimulation of PHM1-31 cells (Fig. 4A). This induction is maintained 24 h following IL-1β treatment. We also studied TNFAIP3/A20, a target gene of NFκB (Krikos et al. 1992). After 30 min of IL-1β exposure, TNFAIP3/A20 protein level is already induced and this induction is gradually increased until 24 h (Fig. 4A).
We also analyzed MAFF, a transcription factor belonging to the basic leucine zipper family. Previously, we showed that MAFF mRNA levels are induced within 30 min with IL-1β in PHM1-31 cells (Massrieh et al. 2006). In this study, we confirmed that this induction also occurs rapidly at the protein level (Fig. 4B). MAFF protein-level induction persists 24 h following IL-1β stimulation.
In this report, we analyzed the response of PHM1-31 myometrial cells to the proinflammatory cytokine IL-1β. We and others have previously characterized PHM1-31 uterine smooth muscle cells (Monga et al. 1996, Madsen et al. 2004, Massrieh et al. 2006). Their morphology is highly similar to that of regular proliferating uterine smooth muscle cells as they are long and spindle shaped with a central nucleus. Using an in vitro collagen lattice-based assay (Wilson et al. 2001), we showed that PHM1-31 cells are able to contract in the absence of an inducing agent (Fig. 1). Upon stimulation of PHM1-31 cells with oxytocin, a posterior pituitary hormone, we observed an additional 10% increase of contraction. Oxytocin has been shown to promote contraction during parturition and is widely used as a uterotonic agent (Gimpl & Fahrenholz 2001). Our novel data further establish PHM1-31 cells as a suitable model to dissect uterine smooth muscle cell function.
Previously, various functional genomic approaches have been used to identify events occurring during the parturition process. These studies examined rodent or human models, comparing the gene expression profiles before and after the onset of labor (Girotti & Zingg 2003, Havelock et al. 2005) in laboring and non-laboring myometrium specimens (Aguan et al. 2000, Bethin et al. 2003, Girotti & Zingg 2003) and in preterm and term myometrium samples (Charpigny et al. 2003, Girotti & Zingg 2003).
It is well established that cytokines are involved in the events leading to preterm labor with intrauterine infection (Hirsch et al. 2006, Sadowsky et al. 2006). Thus, using a genechip array approach, we identified a large number of IL-1β responsive genes in human PHM1-31 myometrial cells suggesting a complex cellular response to cytokine stimulation (Tables 1 and 2). The predominant changes in gene expression are those associated with inflammatory and/or immune response (≈ 15% of the induced genes). This observation corroborates previous microarray studies showing an important role of the inflammatory response in the uterus during labor (Girotti & Zingg 2003, Havelock et al. 2005). Among the inflammatory response genes, we identified PTX3 (or TNFAIP5) as a gene up-regulated by IL-1β. PTX3 has been initially identified as an IL-1β-inducible gene in endothelial cells (Breviario et al. 1992). It has been demonstrated recently that PTX3 expression is higher in the maternal plasma of women with preterm delivery compared with women delivering at term (Assi et al. 2007). Hence, our data may add support to the speculation that PTX3 is implicated in the labor process in uterine smooth muscle cells, in particular during an inflammatory response.
We also identified TNFAIP3/A20 as an inflammatory response gene strongly induced by IL-1β in PHM1-31 cells. It is well established that TNFAIP3/A20 is an NFκB target gene (Krikos et al. 1992). More recently, it has been shown that TNFAIP3/A20 blocks NFκB signaling via a negative feedback loop (Wertz et al. 2004). We found that TNFAIP3/A20 transcript and protein levels are rapidly induced in PHM1-31 cells by 30 min and its expression is kept at a high level for at least 24 h. Previously, it has been reported that the binding of the general transcriptional machinery and particularly the transcription factor specificity protein 1 (Sp1) to the TNFAIP3/A20 promoter is essential for a rapid induction of this gene through the NFκB pathway in response to TNFα (Ainbinder et al. 2002). We hypothesize that a similar mechanism occurs in myometrial cells to regulate TNFAIP3/A20. This is of interest as NFκB has been shown to a play a major role in cytokine signaling in uterine smooth muscle cells (Zaragoza et al. 2006) and possibly in parturition (Condon et al. 2005). Indeed, many inflammatory response genes such as TNFAIP3/A20, superoxide dismutase 2 (SOD2), IL6, and leukemia inhibitory factor (LIF) have been demonstrated to be NFκB-dependent (Xu et al. 1999, Legrand-Poels et al. 2000, Ainbinder et al. 2002, Fan et al. 2004).
In addition to NFκB protein, the MAFF transcription factor, a member of the Maf family of basic leucine zipper transcription factors, may be of particular interest for IL-1β response in myometrial cells. We confirmed data of our earlier study with respect to the induction of the MAFF by IL-1β (Fig. 2 and Table 1; Massrieh et al. 2006). Originally, human MAFF had been identified in a one-hybrid assay as a factor binding to regulatory sequences of the oxytocin receptor gene (Kimura et al. 1999). Interestingly, MAFF is highly expressed in term myometrium but is not present in early gestation (14 weeks) and nonpregnant myometrial tissue (Kimura et al. 1999, Bethin et al. 2003). In accordance with earlier data (Massrieh et al. 2006), we also showed in our microarray studies that the transcript levels of the highly homologous MAFG and MAFK genes are not significantly altered (1.16- and 1.19-fold changes respectively) in 1 h following IL-1β treatment (data not shown). Our results suggest that MAFF, but not MAFG and MAFK, may function as a transcriptional mediator of the early inflammatory response in myometrial cells.
Cytokine stimulation of PHM1-31 cells also results in an increased chemotaxis as shown, for example, by induction of CXCL3 gene expression. Chemotaxis could be defined as a cellular response leading to migration and activation of monocytes and macrophages in the site of inflammation. In this process, chemotactic cytokines, also called chemokines, play a major role and are considered as secondary pro-inflammatory mediators (Dimitriadis et al. 2005). Thus, elevated mRNA level of CXCL3, a gene encoding a small cytokine belonging to the CXC chemokine family, in PHM1-31 cells 1 h following IL-1β treatment can be considered as an early indicator of inflammatory response in those cells. Other chemokines (listed in Tables 1 and 2 in the inflammatory response section) are also significantly induced by IL-1β clearly establishing the myometrium as a tissue mediating a strong inflammatory response.
We have also noted that IL-1β treatment enhances expression of several ECM remodeling enzymes such as TNC and PLAU. It has been shown that TNC, an ECM glycoprotein, is up-regulated in the myometrium of pregnant rabbits compared with non-pregnant myometrium (Cario-Toumaniantz et al. 2003). Nevertheless, further experiments are required to uncover the role of TNC as well as the function of PLAU in myometrial cells and particularly during preterm labor.
Uterine activation results from the coordinated expression of a cassette of contraction-associated proteins including oxytocin receptor, connexin-43, and prostaglandin receptors (Cook et al. 2000). Prostaglandins play a major role in preterm and term labor, particularly in response to cytokines (Mitchell et al. 1993, Keelan et al. 2003, Park et al. 2005). It is well established that IL-1β is a major activator of the crucial and rate-determining enzyme in prostaglandin biosynthesis (PTGS2, also called COX-2; Rauk & Chiao 2000). This action is likely mediated by the nuclear factor NFκB (Ackerman et al. 2004, Soloff et al. 2004). In agreement with these data, expression of PTGS2 and NFκB genes is strongly induced by IL-1β (7.9- and 10.9-fold increase for PTGS2 and NFκB respectively).
Based on our data, we propose a novel hypothetical model of IL-1β signaling in myometrial cells as depicted in Fig. 5. In PHM1-31 cells, IL-1β induces an inflammatory and immune response and increases the expression of key transcription factors like NFκB. Consequently, chemotaxis and synthesis of prostaglandins occur. In parallel, IL-1β increases the expression of the ECM remodeling enzymes. Our genechip data indicate that 1 h exposure of IL-1β has no effect on gene expression of key molecules required for contraction (i.e., oxytocin receptor and connexin 43) in PHM1-31 cells (data not shown). Hence, we hypothesize that PHM1-31 cells exposed to IL-1β for 1 h are in an intermediate phase between the quiescent state and an activated state leading to uterine activation. Interestingly, it has recently been demonstrated that preterm labor in monkeys can be provoked through an intraamniotic infusion of IL-1β or TNFα (Sadowsky et al. 2006). In addition, it has been recently shown that absence of both IL-1β and TNFα in knockout mice clearly delays labor (Hirsch et al. 2006). Hence, the action of multiple cytokines may be required to convert myometrial cells into a fully contractile state. In agreement with these data, we observed a significant induction of interleukins IL6 (8.6-fold induction; Schmid et al. 2001) and IL8 (7.3-fold induction; Arntzen et al. 1998) mRNA levels in response to IL-1β. In addition, it has been demonstrated that PHM1-31 cells retain the ability to respond to oxytocin and thapsigargin with an increase in intracellular calcium (Shlykov et al. 2003). Nevertheless, PHM1-31 cells appear to lack significant L-type Ca2+ channel expression, which suggest that the entry of Ca2+ into the cells may be through a passive leak or another mechanism to be elucidated (Sanborn 2001, Shlykov et al. 2003). Moreover, the role of calcium mobilization in the contraction of myometrium remains to be clarified. In this respect, it has been previously demonstrated that the sarcoplasmic reticulum calcium ATPase 2b (SERCA) is increased in primary-cultured human myometrial smooth muscle cells exposed to IL-1β for 24 h (Tribe et al. 2003). We did not observe any change in SERCA gene transcript levels (data not shown), probably due to the early timepoint examined.
In summary, this study and our previous data (Massrieh et al. 2006) underline the importance of cytokine signaling in uterine smooth muscle cells (Fig. 5). This is of interest because cytokine production is highly induced after bacterial infection of the choriodecidual space (Zaga et al. 2004). Our results validate the myometrium as a strong inflammatory response tissue. Combined with previous published studies, our results reiterate the notion that parturition is a complex and concerted process associated with both temporal and spatial expression changes. Hence, our studies provide the basis for further studies into the mechanisms leading to infection-induced preterm labor.
Materials and Methods
Cells and culture conditions
PHM1-31 myometrial cells (provided by Dr Barbara Sanborn (Monga et al. 1996, Madsen et al. 2004)) were cultured at 37 °C in high-glucose DMEM containing 0.1 mg/ml geneticin, 10% fetal calf serum (heat inactivated at 57 °C), 2 mM l-glutamine, 50 U/ml penicillin, and 50 μg/ml streptomycin. Cytokine induction experiments were performed with cells up to passage 24 reaching 80–90% confluency. Both control and IL-1β-treated cells were fetal bovine serum (FBS) starved for 16–24 h and then treated for 0.5–24 h with vehicle (H2O) or with 1 ng/ml human recombinant IL-1β (Research Diagnostics Inc., Concord, MA, USA) in the absence of serum.
The collagen lattice-based retraction assay was performed as described previously (Devost & Zingg 2007). Shortly, collagen type 1 was prepared from rat tail and resuspended overnight in 0.01 M HCl to prepare a 5 mg/ml stock solution and kept at 4 °C until use. To maintain the pH between 7.0 and 7.5, 5×PBS and 0.1 M NaOH were added and the preparation was subsequently diluted to 1.5 mg/ml with DMEM/F-12 supplemented with 0.5% FBS. Ice-cold collagen solution (0.5 ml) was added to individual wells of a 24-well plate and incubated at 37 °C for 1 h to allow gelling of the collagen. The collagen lattice was then covered with 1 ml DMEM/ F-12 medium supplemented with 0.5% FBS containing 2.5×104 PHM1-31 myometrial cells. The cells were left to settle for 2 h at 37 °C. The collagen lattice was detached from the bottom of the well with a spatula, and left overnight at 37°C in the absence or presence of 100 nM oxytocin (Sigma). The lattices were fixed overnight at 4 °C by adding 1 ml 8% (w/v) paraformaldehyde in PBS at pH 7.4 to stop the contraction and were stored at 4 °C until analysis. To quantify the surface area of the lattices, the liquid of each well was aspirated and the plate was photographed using the Alpha Innotech Imaging System (Alpha Innotech, San Leandro, CA, USA) with an Olympus C-5060 digital camera. The surface area was quantified using the AlphaEase 5.5 densitometry program (Alpha Innotech). The contraction assay was done at least in quadruplicates and the data were expressed as percentage contraction. Percentage contraction was taken as the percentage of lattice size diminution relative to the size area of the well. The contraction assay was repeated in four independent experiments.
Genechip expression array analysis
Total RNA from untreated PHM1 cells and PHM1-31 cells treated with IL-1β for 1 h was prepared using Trizol reagent (Invitrogen). The quality of the purified RNA was verified using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA). Probe synthesis, hybridization, and scanning were done according to standard Affymetrix protocol. The detailed protocol is available on the following website: http://www.affymetrix.com/support/technical/manual/expression_manual.affx. The microarray used was a HG-U133_Plus_2 genechip expression array (Affymetrix, Santa Clara, CA, USA) covering ~38 500 genes of the human genome. The microarray analysis was repeated in three independent experiments.
Robust multi-array average (RMA) analysis was used to analyze the microarray data. It involves three steps: a model-based background correction, quantile normalization at the probe level, and median polish to produce a robust average of probe intensities for expression summary. Intensity values in log (base 2) scale were then converted into fold change, allowing the identification of differentially expressed genes in control and IL-1β-treated cells.
A heat map was generated with the TIGR MultiExperiment Viewer (MeV 4.0) software (http://www.tm4.org/mev.html).
Northern blot hybridization
Total RNA was extracted using the Trizol reagent (Invitrogen) according to the manufacturer’s instructions. RNA was quantified using standard spectrophotometry. Ten micrograms of RNA were denatured, subjected to electrophoresis on a 1% (w/v) agarose gel containing 6% (v/v) formaldehyde, transferred overnight onto a HYBOND-XL nylon membrane (Amersham Biosciences), fixed by u.v. cross-linking, and was hybridized (Church & Gilbert 1984) with 32P-labeled probes. The cDNA probes were prepared from PHMI-31 total RNA by a standard RT-PCR procedure. Forward and reverse PCR primers used to generate cDNA probes were as follows (GenBank accession numbers are provided in parentheses): CXCL3 (BC065743), 5′-GCAGGGAATTCACCTCAAGA-3′ and 5′-GGTGCTCC-CCTTGTTCAGTA-3′ ; TNC (NM_002160), 5′-AGAGAAC-CAGCCAGTGGTGT-3′ and 5′-GCCTGCTCCTGCAGTACATT-3′; PLAU (BC013575), 5′-TCACCACCAAAATGCTGTGT-3′ and 5′-AGGCCATTCTCTTCCTTGGT-3′; PTX3 (BC039733), 5′-TGCGATTCTGTTTTGTGCTC-3′ and 5′-TGAAGAGCTTGT-CCCATTCC-3′ .
To generate the TNFAIP3/A20 probe, a HindIII digestion of the FLAG-A20 plasmid, kindly provided by Dr R. Lin (Lady Davis Institute for Medical Research, Montreal, Canada), was performed and the resulting 340 bp fragment was used for labeling (Lin et al. 2006). The cDNA fragments were radiolabeled with [α-32P]dCTP by random priming according to the instructions of the manufacturer (Roche). Hybridization and washes were performed using standard procedures. Membranes were exposed to a phosphorscreen and analyzed by Phosphor-Imager (Molecular Dynamics, Sunnyvale, CA, USA).
Whole cell extracts were prepared by scraping the PHM1-31 cells using 1× PBS followed by centrifugation. The pellets were then resuspended in lysis buffer (250 mM sucrose, 420 mM NaCl, 10 mM Tris/HCl, 2 mM MgCl2, 1 mM CaCl2, 1% (v/v) Triton X-100) containing protease inhibitors (Roche) supplemented with 0.5 mM dithiothreitol and 0.2 mM phenylmethylsulphonyl fluoride. After 10 min swelling on ice, samples were briefly centrifuged, and the supernatant was collected. Protein concentrations were determined using a protein assay kit (Bio-Rad). Thirty micrograms of the lysate were loaded on a gel and electrophoresed in NuPAGE 4–12% Bis-Tris gels (Invitrogen). Resolved proteins were transferred electrophoretically to a polyvinylidene difluoride membrane (Millipore, Billerica, MA, USA). After membrane saturation at room temperature for 1 h with TBS (0.1 M Tris–HCl (pH 8.0), 0.15 M NaCl) containing 0.05% (v/v) Tween 20 and 5% (v/v) milk, the blots were incubated in the same solution overnight with antibodies specific for either human actin (1:10 000; Sigma, #A-5441), human TNFAIP3/A20 (5 μg/ml; ab13597; Abcam, Cambridge, MA, USA), human NFκB p50 (1:400; sc-7178; Santa Cruz Biotechnology, Santa Cruz, CA, USA), or human MAFF (1:5000; Massrieh et al. 2006). After two washes in TBS containing 0.05% (v/v) Tween 20, primary antibody was detected using 1-h incubation with secondary HRP-conjugated antibodies. A goat anti-mouse antibody (1:30 000; 31430; Pierce, Rockford, IL, USA) was used for actin and TNFAIP3/A20 detection and a goat anti-rabbit antibody (1:30 000; 31 460; Pierce) for NFκB and MAFF detection. Subsequently, the proteins were visualized by exposure to X-ray films (Kodak) using Immobilon Western (Millipore) as a source of chemiluminescent HRP substrate. Membranes were stripped for 15 min at room temperature in Restore buffer (Pierce) and washed four times in TBS before a new hybridization was performed.
Quantification and statistical analysis
Quantification of Northern blot experiments was performed using PhosphorImager and Image Quant software (version 5.2; Molecular Dynamics). Immunoblots were quantified by densitometry using Genetools software (Syngene, Frederick, MD, USA). The results represent a mean±s.e.m. of at least three independent experiments. To evaluate the significance of differences among means, a Student’s t-test was performed (*P<0.05, **P<0.01, ***P<0.001).
Ninety-eight genes are significantly up-regulated (fold change > 3) in human PHM1-31 myometrial cells after interleukin-1β (IL-1β) stimulation (1 ng/ml) for 1 h.
|Gene symbol||Fold change||Gene ID||Description|
|Inflammatory response/immune response|
|TNFAIP3/A20||28.73||202643_s_at||Tumor necrosis factor, α-induced protein 3|
|CXCL3||21.55||207850_at||Chemokine (C-X-C motif) ligand 3|
|SOD2||19.21||215078_at||Superoxide dismutase 2, mitochondrial|
|CCL20||17.86||205476_at||Chemokine (C-C motif) ligand 20|
|CXCL2||15.89||209774_x_at||Chemokine (C-X-C motif) ligand 2|
|LIF||10.06||205266_at||Leukemia inhibitory factor (cholinergic differentiation factor)|
|IL6||8.61||205207_at||Interleukin 6 (interferon, β2)|
|PTGS2||7.89||204748_at||Prostaglandin-endoperoxide synthase 2 (prostaglandin G/H synthase and cyclooxygenase)|
|PTX3||5.97||206157_at||Pentraxin-related gene, rapidly induced by IL-1β|
|CXCL1||5.84||204470_at||Chemokine (C-X-C motif) ligand 1 (melanoma growth-stimulating activity, α)|
|CSF3||4.20||207442_at||Colony-stimulating factor 3 (granulocyte)|
|TNFAIP6||3.67||206025_s_at||Tumor necrosis factor, α-induced protein 6|
|OLR1||3.63||242397_at||Oxidised low-density lipoprotein (lectin-like) receptor 1|
|GEM||3.12||204472_at||GTP-binding protein overexpressed in skeletal muscle|
|GBP2||3.06||242907_at||Guanylate-binding protein 2, interferon inducible|
|Cell adhesion/cell motility|
|TNC||20.71||216005_at||Tenascin C (hexabrachion)|
|IFNAR2||11.87||230735_at||Interferon (α, β, and ω) receptor 2|
|SERPINE2||7.54||236599_at||Serpin peptidase inhibitor, clade E (nexin, plasminogen activator inhibitor type 1), member 2|
|ICAM1||6.44||202638_s_at||Intercellular adhesion molecule 1 (CD54), human rhinovirus receptor|
|CSF2||6.29||210229_s_at||Colony-stimulating factor 2 (granulocyte-macrophage)|
|VCAM1||3.11||203868_s_at||Vascular cell adhesion molecule 1|
|BIRC3||15.04||210538_s_at||Baculoviral IAP repeat-containing 3|
|C11orf17||12.33||220987_s_at||Chromosome 11 open reading frame 17|
|NFKBIA||9.72||201502_s_at||Nuclear factor of κ light polypeptide gene enhancer in B-cells inhibitor, α|
|IER3||6.83||201631_s_at||Immediate early response 3|
|SOCS3||6.53||227697_at||Suppressor of cytokine signaling 3|
|TNFAIP8||5.37||242827_x_at||Tumor necrosis factor, α-induced protein 8|
|AXUD1||5.04||225557_at||AXIN1 up-regulated 1|
|SERPINB2||4.72||204614_at||Serpin peptidase inhibitor, clade B (ovalbumin), member 2|
|IFNB1||3.60||208173_at||Interferon, β 1, fibroblast|
|PPP1R15A||3.42||202014_at||Protein phosphatase 1, regulatory (inhibitor) subunit 15A|
|TNFAIP8||3.29||210260_s_at||Tumor necrosis factor, α-induced protein 8|
|TNFAIP2||5.60||202510_s_at||Tumor necrosis factor, α-induced protein 2|
|RHOB||3.25||212099_at||Ras homolog gene family, member B|
|DNA repair/regulation of transcription|
|EGR3||20.53||206115_at||Early growth response 3|
|EGR2||14.61||205249_at||Early growth response 2 (Krox-20 homolog, Drosophila)|
|NFKB1||10.89||239876_at||Nuclear factor of κ light polypeptide gene enhancer in B-cells 1 (p105)|
|TNF||10.70||207113_s_at||Tumor necrosis factor (TNF superfamily, member 2)|
|IRF1||7.09||202531_at||Interferon regulatory factor 1|
|EGR1||6.29||201693_s_at||Early growth response 1|
|JUNB||5.70||201473_at||Jun B proto-oncogene|
|ATF3||5.38||202672_s_at||Activating transcription factor 3|
|JUN||5.29||201466_s_at||v-jun sarcoma virus 17 oncogene homolog (avian)|
|NKX3-1||5.22||209706_at||NK3 transcription factor related, locus 1 (Drosophila)|
|REL||4.98||206036_s_at||v-rel reticuloendotheliosis viral oncogene homolog (avian)|
|NR4A2||4.45||216248_s_at||Nuclear receptor subfamily 4, group A, member 2|
|BTG2||4.42||201236_s_at||BTG family, member 2|
|CEBPD||4.11||213006_at||CCAAT/enhancer-binding protein (C/EBP), δ|
|MSC||4.03||209928_s_at||Musculin (activated B-cell factor-1)|
|NR4A3||3.98||209959_at||Nuclear receptor subfamily 4, group A, member 3|
|BHLHB2||3.96||201170_s_at||Basic helix-loop-helix domain containing, class B, 2|
|MAFF||3.65||36711_at||v-maf musculoaponeurotic fibrosarcoma oncogene homolog F (avian)|
|NFKBIE||3.53||203927_at||Nuclear factor of κ light polypeptide gene enhancer in B-cells inhibitor, epsilon|
|KLF10||3.43||202393_s_at||Kruppel-like factor 10|
|ETS1||3.23||241435_at||V-ets erythroblastosis virus E26 oncogene homolog 1 (avian)|
|SMURF2||3.20||241900_at||SMAD-specific E3 ubiquitin protein ligase 2|
|NFKB2||3.02||207535_s_at||Nuclear factor of κ light polypeptide gene enhancer in B-cells 2 (p49/p100)|
|RASD1||5.79||223467_at||RAS, dexamethasone-induced 1|
|DUSP2||5.67||204794_at||Dual specificity phosphatase 2|
|PBEF1||5.42||243296_at||Pre-B-cell colony-enhancing factor 1|
|HBEGF||4.17||203821_at||Heparin-binding EGF-like growth factor|
|DSCR1||3.86||208370_s_at||Down syndrome critical region gene 1|
|TRAF1||3.26||205599_at||TNF receptor-associated factor 1|
|SLC39A14||6.23||1561886_a_at||Solute carrier family 39 (zinc transporter), member 14|
|ZFP36||7.02||201531_at||Zinc finger protein 36, C3H type, homolog (mouse)|
|Regulation of translation|
|SAMD4A||6.68||230503_at||Sterile α motif domain containing 4A|
|Protein amino acid phosphorylation|
|MAP3K8||14.06||205027_s_at||MAP kinase kinase kinase 8|
|DUSP1||7.30||201041_s_at||Dual specificity phosphatase 1|
|EGFR||4.92||232925_at||Epidermal growth factor receptor (erythroblastic leukemia viral (v-erb-b) oncogene homolog, avian)|
|IRAK2||4.27||231779_at||Interleukin-1 receptor-associated kinase 2|
|DUSP5||3.59||209457_at||Dual specificity phosphatase 5|
|CCL2||3.41||216598_s_at||Chemokine (C-C motif) ligand 2|
|PTPN1||3.30||240260_at||Protein tyrosine phosphatase, non-receptor type 1|
|PAPPA||3.63||232748_at||Pregnancy-associated plasma protein A, pappalysin 1|
|CRIM1||3.04||233073_at||Cysteine-rich transmembrane BMP regulator 1 (chordin-like)|
|NFKBIZ||13.41||223217_s_at||Nuclear factor of κ light polypeptide gene enhancer in B-cells inhibitor, ζ|
|ZC3H12A||8.90||218810_at||Zinc finger CCCH-type containing 12A|
|WTAP||6.56||244219_at||Wilms tumor 1-associated protein|
|C20orf175||5.47||240410_at||Chromosome 20 open reading frame 175|
|TA-NFKBH||5.43||230052_s_at||T-cell activation NFκB-like protein|
|TJP2||5.24||1565863_at||Tight junction protein 2 (zona occludens 2)|
|FAM43A||4.90||227410_at||Family with sequence similarity 43, member A|
|PRG1||4.77||1554676_at||Proteoglycan 1, secretory granule|
|TJP2||4.18||232017_at||Tight junction protein 2 (zona occludens 2)|
|IER2||3.89||202081_at||Immediate early response 2|
|JMJD3||3.87||213146_at||Jumonji domain containing 3|
|COL4A2||3.50||237624_at||Collagen, type IV, α2|
|GCH1||3.03||204224_s_at||GTP cyclohydrolase 1 (dopa-responsive dystonia)|
|PPP1R15B||3.00||224692_at||Protein phosphatase 1, regulatory (inhibitor) subunit 15B|
Twelve genes are significantly down-regulated (fold change down > 2) in human PHM1-31 myometrial cells after interleukin-1β (IL-1β) stimulation (1 ng/ml) for 1 h.
|Gene symbol||Fold change||Gene ID||Description|
|ARHGAP5||0.48||242110_at||Rho GTPase-activating protein 5|
|Regulation of transcription|
|ID1||0.41||208937_s_at||Inhibitor of DNA binding 1, dominant negative helix-loop-helix protein|
|YEATS2||0.41||236462_at||YEATS domain containing 2|
|SLC16A4||0.48||228455_at||Solute carrier family 16 (monocarboxylic acid transporters), member 4|
|DKK3||0.43||232947_at||Dickkopf homolog 3 (Xenopus laevis)|
|RGS2||0.42||202388_at||Regulator of G-protein signaling 2, 24 kDa|
|GSTA4||0.47||235405_at||Glutathione S-transferase A4|
|FBXL11||0.49||215191_at||F-box and leucine-rich repeat protein 11|
|HPS3||0.49||241036_at||Hermansky–Pudlak syndrome 3|
|ZFP64||0.40||242759_at||Zinc finger protein 64 homolog (mouse)|
Received 20 June 2007 First decision 24 August 2007 Accepted 7 September 2007
We would like to thank Barbara Sanborn and Chun-Ying Ku for providing the PHM1-31 cells, and Zaynab Nouhi, Jadwiga Gasiorek, Lucie Carrière, and Florence Doualla-Bell for helpful discussions and reading of the manuscript. We are grateful to Andre Ponton for help with genechip array analysis and to Rongtuan Lin for the gift of FLAG-TNFAIP3 plasmid. Wael Massrieh and Anne Pierre Charlot helped in the early phase of this project. Volker Blank acknowledges the receipt of a Charles O Monat Award from McGill University. Grégory Chevillard was supported by a Fonds Québecois de Recherche sur la Nature et les Technologies (FQRNT) fellowship. This research was funded by grants from the Hospital for Sick Children Foundation/CIHR Institute for Child Health and Development and Cancer Research Society Inc. to Volker Blank. The authors declare that there is no conflict of interest that would prejudice the impartiality of this scientific work.