TGF-β Suppresses COX-2 Expression by Tristetraprolin-Mediated RNA Destabilization in A549 Human Lung Cancer Cells

1. Cell culture and TGF-β treatment

Human lung adenocarcinoma A549 cells authenticated using short tandem repeat analysis were purchased from the American Type Culture Collection (ATCC, Manassas, VA). Cells were stored in liquid nitrogen, passaged for less than 6 months before use in this study, and cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum (FBS) and gentamicin (10 μg/mL) at 37°C under a 5% CO₂ atmosphere. Human recombinant TGF-β1 (PeproTech, Rocky Hill, NJ) was rehydrated in 4 mM HCl with 1 mg/mL bovine serum albumin and used at a final concentration of 5 ng/mL for all experiments.

2. Western blot analysis

Cells were washed with ice-cold phosphate buffered saline and incubated in extraction buffer (20 mM Tris-Cl [pH 7.4], 100 mM NaCl, 1% NP40, 0.5% sodium deoxycholate, 5 mM MgCI₂, 0.1 mM phenylmethylsulfonyl fluoride, 0.1 mM pepstatin A, 0.1 mM antipain, 0.1 mM chemostatin, 0.2 mM leupeptin, 10 μg aprotinin, 0.5 mg/mL soybean trypsin inhibitor, and 1 mM benzamidine) on ice for 15 minutes. Equal amounts of total cell protein (100 μg) were loaded onto 10% sodium dodecyl sulfate–polyacrylamide denaturing gels and transferred to nitrocellulose membranes. The blots were then probed with anti-human COX-2, anti-Smad2, and anti-Smad3 antibodies (Santa Cruz Biotechnology, Santa Cruz, CA). The membranes were also incubated with antibody specific for α-tubulin (Sigma Chemical Co., St. Louis, MO) as a loading control. Antibody binding was detected using an ECL system (Amersham Pharmacia Biotech, Piscataway, NJ).

3. RNA extraction and Northern blot analysis

A TRI Reagent kit (Molecular Research Center, Cincinnati, OH) was used to isolate total RNA from the A549 cells. Total cellular RNA (20 μg) was separated on 1% agarose gels containing formaldehyde in MOPS (0.2 M 3-N-morpholino-propanesulfonic acid, 0.05 M sodium acetate, and 0.01 M EDTA). The mRNA was then transferred onto nylon membranes (Schleicher and Schuell, Keene, NH) by capillary blotting and crosslinked to the membranes using ultraviolet light. Finally, the blots were hybridized with COX-2 cDNA to evaluate RNA expression. Equal loading in lanes was confirmed using 18S RNA as an internal control.

4. Actinomycin D chase experiment (mRNA stability assay)

Cells were grown in media with TGF-β. Cells were incubated for the indicated times following the addition of 5 μg/mL actinomycin D (Sigma Chemical Co.). Total RNA was subsequently extracted and real-time polymerase chain reaction (PCR) was conducted to quantify the levels of COX-2 mRNA.

5. Reverse transcription PCR and real-time PCR

cDNA was synthesized from 1 μg of total RNA using 1 U of ImProm-II Reverse Transcriptase (Promega, Fitchburg, WI) and random hexamers. The TTP gene was amplified by PCR using the following primers as described by Ogawa et al. [19]: 5′-TCA TCC ACA ACC CTA GCG AA-3′ (sense) and 5′-GAT GCG ATT GAA GAT GGG GA-3′ (anti-sense). The COX-2 gene was amplified with 5′-ATC ACA GGC TTC CAT TGA CC-3′ (sense) and 5′-TAT CAT CTA GTC CGG AGC GG-3′ (anti-sense) primers. PCR was performed in 20 μL reactions by subjecting the samples to 94°C for 5 minutes followed by 30 cycles of 94°C for 30 seconds and 60°C for 30 seconds (for TTP) or 94°C for 30 seconds and 55°C for 30 seconds (for COX-2) followed by 72°C for 30 seconds. To estimate the efficiency of cDNA synthesis for each cell line, β-actin was used as an internal control. The PCR products were resolved in 1.5% SeaKem agarose gels, stained with ethidium bromide, and photographed under ultraviolet light.

Real-time reverse transcription PCR (RT-PCR) was performed using an iCycler Real-Time PCR Detection System (Bio-Rad, Hercules, CA). DNA melting curves were generated to monitor product specificities. mRNA was quantified according to fluorescence resonance energy transfer using a SYBR Green 490 fluorophore. PCR amplification and melting curves were analyzed using the iCycler Optical System Software ver. 3.0 (Bio-Rad,). A SYBR Green stock (1:2,000 dilution) was prepared in water. Primers were diluted to 10 pM in water. Real-time PCR was performed in a total volume of 20 μL using the following protocol: 5 minutes at 95°C followed by 40 cycles of 30 seconds at 95°C and 30 seconds at 60°C (for TTP) or 55°C (for COX-2). All samples were run in triplicate. RNA was quantified using the ΔΔCt method and normalized relative to that of glyceraldehyde 3-phosphate dehydrogenase as an internal control.

6. Transient transfection and luciferase assays

Cells were seeded at a density of 8×10⁴/well in 12-well dishes and grown to 70%-80% confluence in complete growth media containing 10% FBS. Various COX-2 promoter reporter constructs generously provided by Drs. Hiroyasu Inoue and Tadashi Tanabe at the National Cardiovascular Center Research Institute, Japan (400 ng/well) and 100 ng of pSV-β galactosidase vector were used to co-transfect the cells with Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. After 4 hours, complete medium was added and the cells were incubated for an additional 18 hours. After 24 hours of transfection, the cells were treated with or without TGF-β for 6 hours. Luciferase activity was measured using a TR717 microplate luminometer (Applied Biosystems, Foster City, CA) and a Bioluminescent Reporter Gene Assay System (Tropix, Inc., Bedford, MA) according to the manufacturer’s instructions. A pGL2 basic control vector without the insert was used as a negative control for the transfection experiments. The levels of luciferase activity were normalized to that of β-galactosidase.

7. Smad2, Smad3, and TTP knockdown using small interfering RNA (siRNA)

For siRNA knockdown, 80 nM of control, Smad2, or Smad3 oligomer were used for transfection in a 60-mm dish using Lipofectamine 2000 (Invitrogen) for 4 hours, after which the transfection medium was replaced with normal medium. The target sequences of the siRNA molecules specific for Smad2, Smad3, and TTP were as follows: siRNA-control (non-specific), 5′-TTC TCC GAA CGT GTC AGC T-3′; siRNA-TTP, 5′-GTT GTG GAT GAA GTG GCA GCG-3′. The siRNA molecules specific for Smad2 and Smad3 were synthetic siRNA duplexes of 21 nucleotides each that were guaranteed to silence the expression of human and mouse Smad2 and Smad3 (Cellogenetics, CLG-1107 for siSmad2 and CLG-1108 for siSmad3).

8. Statistic analysis

Results are representative of at least three independent experiments and are presented as the mean±standard deviation. Comparisons between groups were made using a Student’s paired t test. A p < 0.05 was considered to indicate statistical significance.

1. TGF-β blocks spontaneous induction of COX-2 expression in A549 cells

Due to the well-known effects of serum on COX-2 expression, we first evaluated the basal expression of COX-2 protein in the presence or absence of serum in A549 cells prior to assessing the effects of TGF-β on COX-2 expression. We demonstrated that COX-2 protein levels spontaneously increased in cells grown with or without serum, although COX-2 expression was more strongly augmented under serum-starvation conditions (Fig. 1A). Interestingly, spontaneous induction of COX-2 expression was inhibited upon TGF-β stimulation, regardless of the presence or absence of serum in the medium (Fig. 1A). Northern blot analysis showed that TGF-β reduced COX-2 mRNA levels in a time-dependent manner, with initial changes observed within 1.5 to 2 hours of TGF-β treatment (Fig. 1B). Moreover, COX-2 protein levels decreased in parallel with reduced COX-2 mRNA expression (Fig. 1C), suggesting that TGF-β down-regulates COX-2 protein expression in A549 cells by lowering COX-2 mRNA levels.

Fig. 1.

Expression of cyclooxygenase 2 (COX-2) mRNA and protein is suppressed by transforming growth factor β (TGF-β) treatment in A549 cells. (A) COX-2 protein levels were spontaneously elevated after cell seeding. This increase was more marked with serum starvation. TGF-β blocked the induction of COX-2 expression in the presence or absence of serum. A549 cells were incubated with or without 5 ng/mL TGF-β in the presence or absence of serum. Total cellular proteins were isolated at the indicated times, after which Western blotting was performed with anti–COX-2 or anti– α-tubulin antibodies. (B) COX-2 mRNA expression was down-regulated by TGF-β. The cells were treated with 5 ng/mL TGF-β for 1.5 to 2 hours, after which the total RNA was isolated at the indicated times and subjected to Northern blot analysis (20 μg per lane) to measure COX-2 mRNA expression. (C) COX-2 protein levels were down-regulated by TGF-β in A549 cells. The cells were treated with 5 ng/mL TGF-β. Total cellular proteins were isolated at the indicated times and analyzed by Western blotting as described above.

2. TGF-β does not affect COX-2 promoter activity in A549 cells

To determine whether the down-regulation of COX-2 mRNA expression by TGF-β is mediated at the transcription level, we investigated the effects of TGF-β on COX-2 promoter activity. To accomplish this, we used various COX-2 promoter-driven reporter constructs containing the regions spanning –1432 to +59 bp or –124 to +59 (Fig. 2A). Relative levels of COX-2 promoter activity were measured by a luciferase activity assay. None of the constructs assessed showed significantly altered expression in response to TGF-β (Fig. 2B), suggesting that regulation of COX-2 expression by TGF-β is not mediated by altered transcription rates in A549 cells.

Fig. 2.

Transforming growth factor β (TGF-β) has no effect on cyclooxygenase 2 (COX-2) promoter activity, but affects COX-2 mRNA half-life. (A, B) Cells were transfected with a pGL2 basic control vector or the indicated reporter constructs, then treated with or without 5 ng/mL TGF-β for 6 hours. Levels of luciferase activity were measured as described in the Materials and Methods section. Data were normalized relative to β-galactosidase activity. No significant difference between the luciferase activities of TGF-β–treated and untreated cells was observed. (C) Cells were stimulated with TGF-β (5 ng/mL) for 1 hour, then treated with actinomycin D (5 μg/mL). The cells were harvested at the indicated times after actinomycin D was added. Total RNA was isolated, and real-time polymerase chain reaction was performed to measure the expression of COX-2 and glyceraldehyde 3-phosphate dehydrogenase (GAPDH). The levels of COX-2 mRNA were normalized relative to GAPDH mRNA and presented as a percentage of that found in the untreated control. NF-kB, nuclear factor kB; IL-6, interleukin 6; CRE, cAMP response element.

3. COX-2 mRNA is post-transcriptionally regulated via destabilization by TGF-β

mRNA decay is critical for controlling the expression of many genes associated with inflammation and cancer via regulation of promoter activities. Since TGF-β did not affect COX-2 promoter activity in our system, we next examined the effects of TGF-β on COX-2 mRNA stability by an actinomycin D chase experiment. The rates of COX-2 mRNA decay were determined by quantifying the levels of COX-2 mRNA that remained in the presence of actinomycin D using real-time PCR. COX-2 mRNA decayed more rapidly in the presence of TGF-β (Fig. 2C). These results suggest that TGF-β triggers the down-regulation of COX-2 expression by decreasing mRNA stability rather than suppressing transcription.

4. COX-2 expression is down-regulated by TGF-β in a Smad3-dependent manner

Since TGF-β triggers various cellular responses through Smad2 and Smad3 [20], the primary intracellular mediators of TGF-β signaling, or a Smad-independent pathway, we determined whether the suppression of COX-2 expression by TGF-β is dependent on Smad2 and/or Smad3. To accomplish this, the effects of Smad2 and Smad3 knockdown on the TGF-β–associated inhibition of COX-2 expression were evaluated using RNA interference. No significant change in TGF-β–induced down-regulation of COX-2 expression was observed in the presence of Smad2-specific siRNA (Fig. 3). In contrast, Smad3-specific siRNA inhibited TGF-β–induced COX-2 down-regulation, resulting in COX-2 expression levels similar to basal levels (Fig. 3). These results suggest that the suppression of COX-2 expression by TGF-β is achieved via mRNA destabilization mediated by a Smad3-dependent pathway.

Fig. 3.

Smad3 knockdown restores the transforming growth factor β (TGF-β)–mediated down-regulation of cyclooxygenase 2 (COX-2) expression. A549 cells were transiently transfected with 80 nM of Smad2- or Smad3-specific siRNA. After transfection, the cells were treated with vehicle or 5 ng/mL of TGF-β for 12 or 24 hours, after which cell lysates were obtained for Western blot analysis. Total cell proteins (70 μg) were subjected to Western blotting and the membranes were probed with the indicated antibodies.

5. TTP expression is rapidly and transiently induced by TGF-β

TTP has recently been described as playing a pivotal role in the destabilization of COX-2 mRNA by directly binding to the 3′-UTR [13,15]. It was also reported that TGF-β transiently and rapidly induces TTP transcription by stimulating the binding of Smad3 to the TTP promoter in human T cells [19]. Based on these previous findings, we investigated whether TGF-β–induced COX-2 mRNA destabilization is mediated by TTP binding to the 3′-UTR of COX-2 mRNA. RT-PCR (Fig. 4A) and real-time PCR (Fig. 4B) analyses demonstrated that TTP expression was rapidly increased within 1 hour of TGF-β stimulation and abruptly decreased thereafter. These data suggest that TGF-β rapidly and transiently induced TTP expression in A549 cells, which is consistent with previous findings in other cell types [19].

Fig. 4.

Tristetraprolin (TTP)-mediated cyclooxygenase 2 (COX-2) mRNA destabilization is induced by transforming growth factor β (TGF-β). (A, B) Total RNA was recovered from A549 cells treated with TGF-β (5 ng/mL) for different periods of time and Reverse transcribed. Reverse transcription–polymerase chain reaction (RT-PCR) and real-time PCR were then performed using primers specific for TTP, β-actin, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as described in the Materials and Methods section. The columns represent the mean of three independent experiments and are shown with error bars (±SE). ^*p < 0.001. (C) A549 cells were transfected with control or TTP-specific siRNA. After 24 hours of transfection, the cells were treated with TGF-β for 1 hour. RT-PCR and quantitative real-time PCR were used to measure the levels of TTP mRNA. Columns, the mean of three independent experiments; Bars, ±SE; ^*p < 0.001. (D) Cells transfected with control or TTP-specific siRNA were stimulated with or without TGF-β for 1 hour, after which actinomycin D (5 μg/mL) was added. Total RNA was isolated at different times, and COX-2 mRNA levels were quantified by real-time PCR. COX-2 mRNA expression was analyzed as described for Fig. 2C. (E) Cells transfected with control or TTP-specific siRNA were stimulated with TGF-β for the indicated times. Total cell proteins were then recovered and analyzed by Western blotting as described in the Materials and Methods.

6. TTP plays an essential role in TGF-β-mediated COX-2 mRNA destabilization

We next determined whether the rapid and transient increase in TTP expression helps reduce COX-2 mRNA stability. Using RNA interference, we eliminated TGF-β–induced expression of TTP and assessed subsequent changes in COX-2 mRNA stability. After confirming the successful knockdown of TTP by siRNA transfection and lack of response to TGF-β (Fig. 4C), we measured COX-2 mRNA stability. COX-2 mRNA stability (Fig. 4D) was unaffected by treatment with TGF-β in A549 cells transfected with TTP-specific siRNA. In contrast, COX-2 mRNA stability decreased in response to TGF-β in A549 cells transfected with the control siRNA. These results highlight the essential role of TTP in the TGF-β–induced destabilization of COX-2 mRNA. Interestingly, we found that the decay rate of COX-2 mRNA in TTP-depleted cells was greatly reduced and even lower than that observed in untreated cells (Fig. 4D). Thus, it is possible that TTP is also involved in regulation of the basal stability of COX-2 mRNA. Accordingly, the level of COX-2 protein was higher in TTP-depleted cells than control cells (Fig. 4E), even without TGF-β stimulation (Fig. 4E).

Taken together, these data are the first to indicate that the TGF-β–Smad3 pathway acts as a negative regulator of COX-2 expression via induction of TTP expression and subsequent TTP-dependent destabilization of COX-2 mRNA.