During spermatogenesis, extensive restructuring of blood–testis barrier takes place to facilitate the migration of preleptotene/leptotene spermatocytes from the basal to the adluminal compartment in the seminiferous epithelium. However, the biochemical mechanisms involved in this event remain elusive. Recent studies have shown that pro-inflammatory cytokine TNFα (tumour necrosis factor α) plays a crucial role in this event by inhibiting the expression of tight junction proteins in Sertoli cells. In the present study, we sought to examine the detailed mechanism on how TNFα affects the expression of CLMP (coxsackie- and adenovirus-receptor-like membrane protein), a newly identified tight junction transmembrane protein, in the testis. Addition of TNFα (10 ng/ml) to Sertoli cell culture (TM4 cells) significantly reduced the steady-state CLMP mRNA and protein levels. In an mRNA stability assay, it was shown that the rate of CLMP mRNA degradation was significantly increased when cells treated with TNFα were compared with vehicle. Blockage of the JNK (c-Jun N-terminal kinase) signalling pathway by SP600125 significantly abolished the TNFα-mediated destabilization of CLMP mRNA. Activation of the JNK signalling pathway by TNFα up-regulated the expression of an RNA-binding protein, TTP (tristetraprolin). TTP was present in the RNA–protein complex in the RNA EMSA (electrophoretic mobility shift assay) and decreased the CLMP 3′-UTR (untranslated region)-dependent luciferase activity. Taken together, these results suggest that the TNFα-mediated mRNA degradation of the CLMP gene is controlled by TTP through the JNK signalling cascade.

INTRODUCTION

Spermatogenesis is one of the most complicated cellular processes by which spermatogonia (diploid, 2n) develop into mature spermatozoa (haploid, 1n). This process is highly dependent upon the timely restructuring of TJs (tight junctions) at the BTB (blood–testis barrier), which facilitates the passage of preleptotene and leptotene spermatocytes from the basal to the adluminal compartment of the seminiferous epithelium (for a review, see [1]). Recent studies have shown that TNFα (tumour necrosis factor α), a 17 kDa secretary product of Sertoli and germ cells in the testis, is emerging as a critical pro-inflammatory cytokine known to disrupt the BTB and impair Sertoli–germ cell interactions that are pertinent to spermatogenesis [2]. Local administration of recombinant TNFα to adult rat testes at 0.5 and 2 μg/testis significantly perturbed the BTB, leading to germ cell loss from the epithelium in vivo [2]. However, the mechanism involved in the BTB disruption remains unexplored. Indeed, accumulating evidence has shown that the disruption of the TJ barrier by TNFα often coincides with the reduction in the steady-state mRNA and protein level of different TJ components. For instance, the mRNA level of claudin-11 was significantly reduced by TNFα in cultured Sertoli cells [3]. Moreover, TNFα-mediated Sertoli cell TJ disruption was also associated with a significant fall in occludin expression [4], and such results are confirmed further by immunohistochemistry and immunofluorescence microscopy [2]. Therefore understanding the mechanisms by which TJ components are regulated would provide important information for better understanding of junction restructuring during spermatogenesis.

It is known that TNFα-mediated TJ disruption involves the activation of several MAPKs (mitogen-activated protein kinases) [5]; however, the molecular targets and mechanisms of the MAPK activation in such events remain entirely unknown. Except that TNFα could reduce the expression of occludin via suppression of occludin promoter activity in HT-29/B6 cells [6], it is not known whether TNFα exerts its regulatory effects on gene transcription of other TJ proteins, post-transcriptional modification of their mRNA transcripts or post-translational modification at the protein level, which in turn modulates TJ dynamics. In the present study, we sought to examine the detailed mechanism on how TNFα regulates the expression of CLMP (coxsackie- and adenovirus-receptor-like membrane protein), a novel TJ transmembrane protein, in the testis [7]. We report that the expression of CLMP, in response to TNFα, is mediated via the JNK (c-Jun N-terminal kinase) pathway at the post-transcriptional level. These findings not only unravel the mechanism and signalling pathways by which TNFα suppresses CLMP expression, but also provide the first report in the literature showing that cytokine regulates the expression level of TJ protein via post-transcriptional modification in testicular cells.

MATERIALS AND METHODS

Biochemicals and plasmids

Actinomycin D, p42/p44 kinase inhibitor (U0126), p38 kinase inhibitor (SB202190), JNK inhibitor (SP600125) and recombinant mouse TNFα were obtained from Calbiochem. Scrambled and JNK2 siRNA (small interfering RNA) were purchased from Dharmacon. Rabbit polyclonal anti-SAPK (stress-activated protein kinase)/JNK and anti-phospho-SAPK/JNK (Thr183/Tyr185) antibodies were purchased from Cell Signaling Technology. Goat polyclonal anti-TTP (tristetraprolin) (N-18) and rabbit polyclonal anti-actin antibodies were from Santa Cruz Biotechnology and Sigma–Aldrich respectively. Rabbit polyclonal anti-ZO-1 (zonula occludens 1) and FITC-conjugated goat anti-rabbit IgG were obtained from Zymed Laboratories. Rabbit polyclonal anti-CLMP antibody was synthesized against a peptide sequence GTHTEIKRVAEEKVTL corresponding to the N-terminus of mouse CLMP according to previous studies [7]. The specificity of this antibody has been characterized previously in Caco-1 cells and TM4 cells [8].

Construction of plasmids

The 5′-flanking region of the CLMP gene was obtained by overlap-extension-PCR amplification of mouse genomic DNA using gene-specific primers (Table 1). The PCR product was cloned into the MluI and BglII sites upstream of the luciferase gene in the promoterless pGL3-Basic vector (Promega) and designated p[−946/−1]Luc (Figure 4B, upper panel). To generate the reporter construct p[−946/−1]Luc-3′-UTR665, the mouse CLMP 3′-UTR (untranslated region) was amplified by PCR using the RT (reverse transcription) product of TM4 cells and further cloned downstream of the luciferase gene in the XbaI site of the p[−946/−1]Luc (Figure 4B, lower panel). The two constructs were used parallel in luciferase assays to test the influence of the 3′-UTR of CLMP on the stability of luciferase mRNA. All plasmids were prepared using the Plasmid Midi kit (Promega) and confirmed by ABI DNA sequencing.

Table 1
Nucleotide sequences of primers used in plasmid construction, in vitro transcription and RT–PCR

S, sense; AS, antisense. Italic bases indicate the linker sequence.

Primer name Location Orientation Sequence (5 ′→3′) Purpose 
1064 451/469 CAGTGTGAGTCTGCCTCTG RT–PCR 
1065 670/653 AS CAGTCACCCGTACCACAC RT–PCR 
1118 −1/−15 AS CTCTCGAGCCCGATCTCCGGACA 5′-UTR 
1129 −946/−930 CTACGCGTGATTTCATACCGACC 5′-UTR 
1213 1123/1140 CATCTAGACTTAGAGTGGACTTGACT 3′-UTR 
1214 1787/1770 AS GCTCTAGATTAGGGTTCGACGGAAAT 3′-UTR / in vitro transcription 
1256 T7 promoter 5′ linker GAGAATTCTAATACGACTCACTATAGGGCCGCGG In vitro transcription 
1257 1123/1138 GGCCGCGGCTTAGAGTGGACTTGA In vitro transcription 
1258 1342/1325 AS ACAGTGAGGTGCCTTTAG In vitro transcription 
1259 1562/1544 AS AAGGCGTGGGCTTCTGTT In vitro transcription 
Primer name Location Orientation Sequence (5 ′→3′) Purpose 
1064 451/469 CAGTGTGAGTCTGCCTCTG RT–PCR 
1065 670/653 AS CAGTCACCCGTACCACAC RT–PCR 
1118 −1/−15 AS CTCTCGAGCCCGATCTCCGGACA 5′-UTR 
1129 −946/−930 CTACGCGTGATTTCATACCGACC 5′-UTR 
1213 1123/1140 CATCTAGACTTAGAGTGGACTTGACT 3′-UTR 
1214 1787/1770 AS GCTCTAGATTAGGGTTCGACGGAAAT 3′-UTR / in vitro transcription 
1256 T7 promoter 5′ linker GAGAATTCTAATACGACTCACTATAGGGCCGCGG In vitro transcription 
1257 1123/1138 GGCCGCGGCTTAGAGTGGACTTGA In vitro transcription 
1258 1342/1325 AS ACAGTGAGGTGCCTTTAG In vitro transcription 
1259 1562/1544 AS AAGGCGTGGGCTTCTGTT In vitro transcription 

Cell culture and transfection

Mouse TM4 cells (American Type Culture Collection, Manassas, VA, U.S.A.) were maintained in DMEM (Dulbecco's modified Eagle's medium) (Gibco) supplemented with 10% (v/v) fetal bovine serum. A total of 0.5×105 cells were seeded in a 12-well plate 24 h before transfection. Cells were transfected with 0.5 μg of each luciferase reporter construct and 0.1 μg of the control plasmid, pEGFP, using GeneJuice® Transfection Reagent (Novagen). Luciferase activity was measured using a Lumat LB9507 luminometer (EG&G) and normalized against the GFP (green fluorescent protein) signalling. For mRNA stability experiments, cells were cultured with either vehicle (PBS/0.1% BSA) or TNFα (10 ng/ml) for 2 h before addition of actinomycin D (5 μg/ml) to block further transcription. Total RNA was isolated at 0, 3, 6, 9 or 12 h following actinomycin D treatment. In some experiments, kinase inhibitors were pre-treated for 20 min to 1 h before the addition of TNFα.

RT–PCR

Total RNA was isolated from cells by TRIzol® reagent (Invitrogen) and reverse-transcribed as described previously [9]. RT products were used as templates for subsequent PCR co-amplification using S16 and CLMP primer pairs (Table 1). PCR products were analysed on a 5% (w/v) agarose gel and visualized by ethidium bromide staining. S16 and CLMP amplifications were in their exponential phases, and the authenticity of PCR product was confirmed by ABI DNA sequencing.

Real-time PCR

The CLMP mRNA level was measured by real-time PCR with a SYBR Green PCR kit (Applied Biosystems). The S16 gene was used as an internal standard for normalization, and transcript levels of each gene were quantified using the 7300 Real Time PCR system (Applied Biosystems) according to the manufacturer's instructions. Fluorescence signals were measured at the extension step throughout the amplification process. The specificity of the fluorescence signal was confirmed by melting curve analysis and agarose gel electrophoresis. The expression level of the target gene was determined using the 2−ΔΔCt method [10].

Preparation of cytoplasmic extracts

Cytoplasmic lysates were prepared as described in [11] with modification. Cells were rinsed once with ice-cold PBS and harvested with a hypotonic buffer [10 mM Hepes (pH 7.6) containing 40 mM KCl, 3 mM MgCl2, 2 mM dithiothreitol, 5% glycerol, 0.5% Nonidet P40 and protease inhibitors (0.5 mM PMSF, 1 mM sodium orthovanadate, aprotinin and pepstatin A)]. After incubation on ice for 15 min, nuclei were removed by centrifugation at 600 g for 10 min. The supernatant was divided into 40 μl aliquots and snap-frozen at −70 °C. Protein concentration was determined using Bradford protein assay (Bio-Rad Laboratories).

In vitro transcription

RNA probes were prepared using a MAXIscript® kit (Ambion) following the manufacturer's instructions. To generate a DNA template for in vitro transcription, a T7 phage promoter sequence was added to the 5′-end of the CLMP 3′-UTR. In the first amplification step, an upstream sense PCR primer #1257, containing a linker sequence (GGCCGCGG) at its 5′-end, was used together with the antisense downstream primer (#1214, #1258 or #1259). In the second amplification step, a universal promoter primer composed of the T7 promoter consensus sequence and the linker sequence (#1256) was used with the antisense downstream primer to generate the template for in vitro transcription (Table 1). The PCR fragments were resolved by electrophoresis to confirm the authenticity. Riboprobes were synthesized using T7 RNA polymerase in the presence of 50 μCi of [α-32]UTP (400 Ci/mmol) (Amersham Biosciences). Reactions were stopped by addition of 1 unit of Turbo DNase I. Unincorporated nucleotides were removed using G-25 spin columns (Amersham Biosciences). Specific radioactivities of probes were determined by scintillation counting.

RNA EMSA (electrophoretic mobility-shift assay)

RNA EMSA was performed as described [12] in a RNA-binding buffer [20 mM Hepes (pH 7.6) containing 40 mM KCl, 3 mM MgCl2, 2 mM dithiothreitol and 5% glycerol]. A 10 μg amount of TM4 cytoplasmic protein was incubated on ice for 20 min with 500000 c.p.m. of α-32P-labelled RNA probe. RNase T1 and heparan sulfate were added to final concentrations of 50 units/ml and 5 mg/ml respectively, and the reaction was allowed to continue for a further 20 min on ice. Quantities (5 μl) of loading buffer (90% glycerol and 0.025% Bromophenol Blue) were added to the samples, which were then resolved by electrophoresis on a non-denaturing 4% polyacrylamide gel at 200 V at 4 °C for 4 h. Gels were dried and autoradiographed. For competition studies, unlabelled RNA probes were added to the binding reactions for 20 min before the addition of labelled probes. For antibody supershift assays, anti-TTP antibody (2 μg) was added simultaneously with the cell extract.

Western blotting

Cytoplasmic extracts were resolved on an SDS/7.5% polyacrylamide gel, transferred on to nitrocellulose membrane (Amersham Biosciences) and blocked with 5% (w/v) non-fat dried milk powder in PBS/Tris with 0.1% Tween 20 for 2 h at room temperature (20–22 °C). The membrane was then incubated with primary antibody overnight at 4 °C. After 1 h of incubation with the corresponding secondary antibody, proteins were visualized using ECL® (enhanced chemiluminescence) reagent (Amersham Biosciences).

Statistical analysis

For all transfection assays, data were shown as means±S.D. for duplicate assays in three independent experiments. Statistical analysis was performed using Student's t test using Prism software (GraphPad). For RT–PCR, Western blotting and RNA EMSA, all studies were repeated at least three times and consistent results were obtained.

RESULTS

TNFα reduces CLMP mRNA and protein levels

Recent study has shown that intratesticular administration of TNFα could transiently down-regulate the expression of several TJ proteins such as occludin and ZO-1 at the BTB [2]. However, it is not known whether TNFα can affect the expression of CLMP. To investigate the effect of TNFα on the regulation of CLMP expression, TM4 cells were treated with different concentrations of TNFα (0–15 ng/ml) for 24 h followed by RT–PCR and immunoblot analyses. It was found that the steady-state CLMP mRNA level was reduced dose-dependently by TNFα when compared with the control (Figure 1A). Addition of TNFα at concentrations of 10 and 15 ng/ml could reduce CLMP mRNA levels by 33%. In a parallel experiment, it was found that TNFα could reduce CLMP levels (Figure 1B). The level of 48 kDa CLMP protein was reduced more significantly than that of 44 kDa upon TNFα treatment. These observations could be attributed to differential regulation of mRNA variants and post-translational modification of the protein isoforms. Northern blot analysis has shown that two splicing variants of CLMP with sizes of ∼1.8 and 4.3 kb are detected in the testis [2]. Two variants might share a common UTR sequence (e.g. at the 3′-UTR) that can be regulated by a particular cytokine or signalling pathway, but magnitudes of regulation can be different. The overall mRNA turnover rate is controlled in concert by both 3′- and 5′-UTRs. Apart from differential regulation of mRNA, it is possible that TNFα might also act on the protein turnover via different post-translational processes such as ubiquitination and endocytosis.

TNFα reduces CLMP mRNA and protein levels by affecting CLMP mRNA stability

Figure 1
TNFα reduces CLMP mRNA and protein levels by affecting CLMP mRNA stability

RT–PCR (A) and immunoblot (B) analysis of CLMP expression in TM4 cells after TNFα treatment. Total RNA and protein lysate were prepared from TM4 cells treated with vehicle (PBS/0.1% BSA) or TNFα (1–15 ng/ml) for 24 h. CLMP mRNA levels were quantified by RT–PCR and normalized to S16. The empty vector pGL3-Basic or p[−946/−1]Luc was transfected into TM4 cells and subjected to TNFα treatment 24 h before termination (C). The relative luciferase activity is represented as the fold induction when compared with the pGL3-Basic vector treated with vehicle after normalization by GFP illumination. Actinomycin D chase studies of CLMP mRNA in the absence or presence of TNFα by RT–PCR (D) and real-time PCR (E). TM4 cells were treated with either vehicle (PBS/0.1% BSA) or TNFα (10 ng/ml) for 2 h. Cells were subsequently harvested at zero time (as control) or treated further with actinomycin D (5 μg/ml) for 3, 6, 9 or 12 h. The CLMP mRNA half-life was estimated by extrapolation from the graph and corrected to the nearest 1 h (D). Results are the means±S.D. for at least three independent experiments. Numerical values of S.D. are omitted from the table beneath the histogram (E). *P<0.01 and **P<0.05 compared with vehicle control; ns, not significant. ActD, actinomycin D; Veh, vehicle.

Figure 1
TNFα reduces CLMP mRNA and protein levels by affecting CLMP mRNA stability

RT–PCR (A) and immunoblot (B) analysis of CLMP expression in TM4 cells after TNFα treatment. Total RNA and protein lysate were prepared from TM4 cells treated with vehicle (PBS/0.1% BSA) or TNFα (1–15 ng/ml) for 24 h. CLMP mRNA levels were quantified by RT–PCR and normalized to S16. The empty vector pGL3-Basic or p[−946/−1]Luc was transfected into TM4 cells and subjected to TNFα treatment 24 h before termination (C). The relative luciferase activity is represented as the fold induction when compared with the pGL3-Basic vector treated with vehicle after normalization by GFP illumination. Actinomycin D chase studies of CLMP mRNA in the absence or presence of TNFα by RT–PCR (D) and real-time PCR (E). TM4 cells were treated with either vehicle (PBS/0.1% BSA) or TNFα (10 ng/ml) for 2 h. Cells were subsequently harvested at zero time (as control) or treated further with actinomycin D (5 μg/ml) for 3, 6, 9 or 12 h. The CLMP mRNA half-life was estimated by extrapolation from the graph and corrected to the nearest 1 h (D). Results are the means±S.D. for at least three independent experiments. Numerical values of S.D. are omitted from the table beneath the histogram (E). *P<0.01 and **P<0.05 compared with vehicle control; ns, not significant. ActD, actinomycin D; Veh, vehicle.

TNFα reduces CLMP mRNA levels by affecting its mRNA stability

The reduction of CLMP mRNA levels upon TNFα treatment could be mediated through gene transcription and/or post-transcriptional modification. To examine whether TNFα alters CLMP mRNA levels via gene transcription, TM4 cells were transfected with the functional promoter construct, p[−946/−1]Luc, and transfected cells were subjected to TNFα treatment. The p[−946/−1]Luc containing the core promoter region has been shown previously to drive basal CLMP gene transcription [8]. As shown in Figure 1(C), transfection of p[−946/−1]Luc caused a 4.8-fold increase in promoter activity in TM4 cells when compared with pGL3 control. However, there was no significant change in promoter activity in the presence or absence of TNFα. These results suggest that the reduction of CLMP level by TNFα is not due to transcriptional regulation. To examine whether TNFα modulates CLMP mRNA stability, actinomycin D assays were performed. TM4 cells were cultured with TNFα (10 ng/ml) or vehicle for 2 h before the addition of actinomycin D (5 μg/ml), a transcription inhibitor, which is commonly used to block transcription. Cells were harvested at 0, 3, 6, 9 or 12 h, and the change of CLMP mRNA level was examined by RT–PCR (Figure 1D) and real-time PCR (Figure 1E). As shown in Figures 1(D) and 1(E), the mRNA stability of CLMP was significantly reduced upon TNFα treatment. In the absence of TNFα stimulation, ∼65% of the initial amount of CLMP mRNA remained after 12 h of incubation with actinomycin D in both RT–PCR and real-time PCR analyses. In contrast, only 33.3% of the original amount of CLMP mRNA was left after 12 h of incubation in the presence of TNFα. In particular, the half-life of CLMP mRNA was significantly shortened by 41%, from 17 h in untreated cells to 10 h upon TNFα treatment (extrapolated from Figure 1D, lower panel), indicating that TNFα could speed up CLMP mRNA degradation.

TNFα controls CLMP mRNA stability via the JNK signalling pathway

It has been established previously that TNFα activates three classical MAPK cascades, namely p42/p44 MAPK, p38 MAPK and JNK/SAPK in cultured Sertoli cells [13]. To identify the signalling pathway that is involved in the regulation of CLMP mRNA stability, inhibitors of three MAPKs, namely U0126 (a p42/p44 kinase inhibitor), SB202190 (a p38 kinase inhibitor) and SP600125 (a JNK inhibitor), were used. TM4 cells were pre-treated with MAPK inhibitors followed by TNFα treatment, and the effects on CLMP mRNA stability after TNFα exposure were assessed. From the RT–PCR results, pre-treatment of the cells with U0126 and SB202190 could not abolish TNFα-mediated CLMP mRNA destabilization (Figures 2A and 2B). However, pre-treatment with the JNK inhibitor, SP600125, could completely abolish the effect of TNFα-mediated destability on CLMP transcripts and the rate of mRNA degradation was restored to a level comparable with cells treated with vehicle (Figures 2A and 2C). Negative controls using vehicle (PBS/0.1% BSA), DMSO or SP600125 alone were included to confirm the specificity of the kinase inhibitors (Figures 2A and 2D). These results suggest that the destabilization of CLMP mRNA in response to TNFα is mediated through the JNK cascade.

TNFα-mediated CLMP mRNA turnover requires the activation of JNK

Figure 2
TNFα-mediated CLMP mRNA turnover requires the activation of JNK

(A) TM4 cells were pre-incubated with DMSO or different kinase inhibitors [U0126 (10 μM; 60 min), SB202190 (1 μM; 20 min) or SP600125 (20 μM; 30 min)], followed by the addition of actinomycin D (5 μg/ml). After 1 h, vehicle (PBS/0.1% BSA) or TNFα (10 ng/ml) was added. Cells were harvested at 0, 3, 6, 9 or 12 h after treatment. CLMP mRNA levels were quantified by RT–PCR and normalized to S16 levels. Three additional internal controls, including SP600125, PBS/0.1% BSA and DMSO alone, were also included. Graphical data representations of the experiments performed in (A) are shown in (BD). The normalized CLMP/S16 level at the time of the addition of TNFα (zero time) was set as 100%. Results are means±S.D. for four independent experiments. *P<0.01 compared with DMSO+vehicle; ns, not significant. Veh, vehicle.

Figure 2
TNFα-mediated CLMP mRNA turnover requires the activation of JNK

(A) TM4 cells were pre-incubated with DMSO or different kinase inhibitors [U0126 (10 μM; 60 min), SB202190 (1 μM; 20 min) or SP600125 (20 μM; 30 min)], followed by the addition of actinomycin D (5 μg/ml). After 1 h, vehicle (PBS/0.1% BSA) or TNFα (10 ng/ml) was added. Cells were harvested at 0, 3, 6, 9 or 12 h after treatment. CLMP mRNA levels were quantified by RT–PCR and normalized to S16 levels. Three additional internal controls, including SP600125, PBS/0.1% BSA and DMSO alone, were also included. Graphical data representations of the experiments performed in (A) are shown in (BD). The normalized CLMP/S16 level at the time of the addition of TNFα (zero time) was set as 100%. Results are means±S.D. for four independent experiments. *P<0.01 compared with DMSO+vehicle; ns, not significant. Veh, vehicle.

To investigate the involvement of the JNK pathway in controlling CLMP mRNA stability, the expression of JNK protein was analysed by immunoblotting using cytoplasmic proteins prepared from TNFα-treated TM4 cells. As shown in Figure 3(A), addition of TNFα to the cell culture did not alter the expression of total JNK for both isoforms. However, a significant increase (2.83-fold) in the expression of the phospho-p54 JNK isoform, but not the phospho-p46 JNK isoform, was observed when normalized against total JNK after TNFα treatment. The effect of TNFα on p54 JNK phosphorylation was confirmed further using various concentrations of TNFα (0–15 ng/ml) for 24 h followed by immunoblot analyses. As shown in Figure 3(B), the p54 JNK phosphorylation was increased dose-dependently after TNFα stimulation. In addition, the p54 JNK activation was fully inhibited in cells pre-treated with SP600125 (Figure 3C). To confirm further the functional role of p54 JNK on the expression of CLMP, specific JNK2 siRNA was transfected in TM4 cells. RT–PCR analyses showed that the mRNA level of CLMP was significantly increased in JNK2 siRNA-transfected cells compared with the control. These results together support the notion that cytoplasmic p54 JNK is the predominant JNK isoform that is involved and activated in TNFα-mediated CLMP mRNA degradation.

Phosphorylation of the p54 isoform, but not the p46 isoform, of JNK is responsible for TNFα-mediated CLMP mRNA degradation in cultured TM4 cells

Figure 3
Phosphorylation of the p54 isoform, but not the p46 isoform, of JNK is responsible for TNFα-mediated CLMP mRNA degradation in cultured TM4 cells

(A) TM4 cells were treated with either vehicle (PBS/0.1% BSA) or TNFα (10 ng/ml) for 24 h. Cells were harvested in a SDS lysis buffer containing phosphatase inhibitors. Cell extracts were resolved by SDS/PAGE, and immunoblots were probed with anti-phospho-JNK (top panel) or anti-(total JNK) (bottom panel) antibodies. Results are expressed as the ratio of the phospho-JNK to the total JNK levels where cells treated with vehicle alone was designated as 1. Results in the histogram are means±S.D. for three independent experiments. (B) Cells were treated with an increasing dose of TNFα for 24 h, and the level of p54 JNK activation was monitored by immunoblot analysis. (C) Cells were pre-treated with either SP600125 or DMSO followed by TNFα treatment. The level of p54 JNK activation was examined by immunoblot analysis. (D) TM4 cells were transfected with 50 pmol of siControl (control siRNA) or JNK2 siRNA and were harvested at 48 h post-transfection. CLMP mRNA levels were quantified by RT–PCR and normalized to S16 levels. *P<0.01 compared with vehicle control; ns, not significant. Veh, vehicle.

Figure 3
Phosphorylation of the p54 isoform, but not the p46 isoform, of JNK is responsible for TNFα-mediated CLMP mRNA degradation in cultured TM4 cells

(A) TM4 cells were treated with either vehicle (PBS/0.1% BSA) or TNFα (10 ng/ml) for 24 h. Cells were harvested in a SDS lysis buffer containing phosphatase inhibitors. Cell extracts were resolved by SDS/PAGE, and immunoblots were probed with anti-phospho-JNK (top panel) or anti-(total JNK) (bottom panel) antibodies. Results are expressed as the ratio of the phospho-JNK to the total JNK levels where cells treated with vehicle alone was designated as 1. Results in the histogram are means±S.D. for three independent experiments. (B) Cells were treated with an increasing dose of TNFα for 24 h, and the level of p54 JNK activation was monitored by immunoblot analysis. (C) Cells were pre-treated with either SP600125 or DMSO followed by TNFα treatment. The level of p54 JNK activation was examined by immunoblot analysis. (D) TM4 cells were transfected with 50 pmol of siControl (control siRNA) or JNK2 siRNA and were harvested at 48 h post-transfection. CLMP mRNA levels were quantified by RT–PCR and normalized to S16 levels. *P<0.01 compared with vehicle control; ns, not significant. Veh, vehicle.

Induction of TTP expression decreases CLMP mRNA stability via its interaction with CLMP 3′-UTR

Regulation of mRNA stability is often mediated through the interaction of RNA-binding proteins with the 3′-UTR of the transcript. Previous studies have shown that TNFα could activate the expression of a RNA-binding protein, TTP, in macrophages and TTP could bind to the 3′-UTR of TNFα, resulting in TNFα mRNA destabilization [14]. TTP not only is expressed in macrophages, but also is one of the major RNA-binding proteins found in the testis [15]. It is logical to speculate that TTP may be the corresponding RNA-binding protein involved in TNFα-mediated CLMP mRNA degradation. To test this hypothesis, we first investigated whether TNFα stimulates the expression of TTP in TM4 cells via the activation of the JNK cascade. As shown in Figure 4(A), there was a significant increase in TTP protein level in TM4 cells upon TNFα stimulation compared with the vehicle control. In addition, blockage of the JNK cascade by the JNK inhibitor, SP600125, could attenuate TNFα-mediated TTP expression. These results suggest that TNFα induces TTP expression via activation of the JNK pathway.

TNFα induces the expression of TTP, which reduces CLMP mRNA stability

Figure 4
TNFα induces the expression of TTP, which reduces CLMP mRNA stability

(A) TM4 cells were pre-treated with SP600125 or DMSO followed by vehicle or TNFα treatment (10 ng/ml) for 24 h. Cells were harvested, lysed and subjected to immunoblot analysis for TTP protein levels. (B) Schematic representation of CLMP promoter self-driven luciferase reporter constructs used in the transfection experiment. Transcription (bent arrow) was driven by the CLMP promoter upstream of the luciferase gene. Numbering is relative to the CLMP ATG start codon. The 3′-UTR of CLMP mRNA was cloned into the XbaI site downstream of the luciferase gene in sense orientation. A 0.5 μg amount of either p[−946/−1]Luc or p[−946/−1]Luc-3′-UTR665 was transfected into TM4 cells. (C) p[−946/−1]Luc or p[−946/−1]Luc-3′-UTR665 was transfected into TM4 cells and subjected to either PBS/0.1% BSA (vehicle) or TNFα treatment for 24 h. (D) p[−946/−1]Luc or p[−946/−1]Luc-3′-UTR665 was co-transfected into TM4 cells along with 0.5 μg of either the empty vector pCMV or the TTP overexpression plasmid. The relative luciferase activity is represented as the fold induction when compared with the control vectors after normalization by GFP illumination. Results are means±S.D. for three independent experiments. *P<0.01 compared with vehicle control (A, C), p[−946/−1]Luc (B) or pCMV empty vector (D); ns, not significant. Veh, vehicle.

Figure 4
TNFα induces the expression of TTP, which reduces CLMP mRNA stability

(A) TM4 cells were pre-treated with SP600125 or DMSO followed by vehicle or TNFα treatment (10 ng/ml) for 24 h. Cells were harvested, lysed and subjected to immunoblot analysis for TTP protein levels. (B) Schematic representation of CLMP promoter self-driven luciferase reporter constructs used in the transfection experiment. Transcription (bent arrow) was driven by the CLMP promoter upstream of the luciferase gene. Numbering is relative to the CLMP ATG start codon. The 3′-UTR of CLMP mRNA was cloned into the XbaI site downstream of the luciferase gene in sense orientation. A 0.5 μg amount of either p[−946/−1]Luc or p[−946/−1]Luc-3′-UTR665 was transfected into TM4 cells. (C) p[−946/−1]Luc or p[−946/−1]Luc-3′-UTR665 was transfected into TM4 cells and subjected to either PBS/0.1% BSA (vehicle) or TNFα treatment for 24 h. (D) p[−946/−1]Luc or p[−946/−1]Luc-3′-UTR665 was co-transfected into TM4 cells along with 0.5 μg of either the empty vector pCMV or the TTP overexpression plasmid. The relative luciferase activity is represented as the fold induction when compared with the control vectors after normalization by GFP illumination. Results are means±S.D. for three independent experiments. *P<0.01 compared with vehicle control (A, C), p[−946/−1]Luc (B) or pCMV empty vector (D); ns, not significant. Veh, vehicle.

In order to examine the effect of TNFα and TTP on CLMP 3′-UTR, two reporter constructs (p[−946/−1]Luc and p[−946/−1]Luc-3′-UTR665) were transfected into TM4 cells. As depicted in Figure 4(B), transfection of p[−946/−1]Luc-3′-UTR reduced the luciferase activity by half when compared with the p[−946/−1]Luc construct. Since the two reporter constructs were different only in the CLMP 3′-UTR sequence inserted in the XbaI site downstream of the luciferase reporter gene, the change in luciferase activity will be contributed to only by the inserted 3′-UTR. This indicates that the additional 665 nt at the 3′-UTR region of CLMP are responsible for mRNA degradation. In addition, administration of TNFα to p[−946/−1]Luc-3′-UTR665 reduced the luciferase activity by 30%, but not when added to the p[−946/−1]Luc construct (Figure 4C). These results suggest that the 3′-UTR of CLMP is responsible for the TNFα-mediated CLMP mRNA degradation.

In order to examine the effect of TTP on CLMP 3′-UTR, TTP expression plasmid was co-transfected with a luciferase reporter construct, p[−946/−1]Luc or p[−946/−1]Luc-3′-UTR665 (Figure 4D). The change in TTP protein levels was examined using immunoblotting (see inset). When TTP was overexpressed with the construct p[−946/−1]Luc, there was no significant change in luciferase activity in comparison with that of pCMV. In contrast, when TTP was overexpressed with the construct p[−946/−1]Luc-3′-UTR665 carrying the CLMP 3′-UTR, the luciferase activity was reduced by 40%. These results indicate that TTP destabilizes CLMP mRNA via its action on the CLMP 3′-UTR.

Binding of TTP to the CLMP 3′-UTR was demonstrated by RNA EMSA using three radiolabelled RNA probes corresponding to the first 220, 440 and 665 nt of the CLMP 3′-UTR. As depicted in Figure 5(A) (lanes 1 and 2), both the 665 and 440 nt RNA probes containing an AUUUA motif were able to form a prominent RNA–protein complex with TNFα-treated TM4 cytoplasmic proteins. In contrast, the 220 nt RNA probe lacking the AUUUA motif failed to generate the RNA–protein complex, as there was no difference in the mobility of the 220 nt RNA probe in the presence or absence of cytoplasmic proteins (Figure 5A, lanes 3 and 4). Since the 665 nt RNA probe formed a complex at a lower resolution, which is a frequent problem in RNA EMSA when a relatively long probe is used, the 440 nt probe was used in all subsequent experiments. Apparently, an RNA–protein complex was formed when the 440 nt RNA probe was incubated with cytoplasmic proteins (Figure 5B). The mobility of the RNA–protein complex decreased when an increasing amount of cytosolic proteins was used in this binding reaction. The fact is that cytosolic protein is the limiting factor in this binding reaction. Therefore an increase in the amount in cytosolic proteins could ensure effective binding of all essential proteins to each RNA probe.

TTP is responsible for the formation of RNA–protein complex upon CLMP 3′-UTR

Figure 5
TTP is responsible for the formation of RNA–protein complex upon CLMP 3′-UTR

(A) Binding of cytoplasmic proteins to CLMP 3′-UTR riboprobes of different lengths. Numbers represent the relative lengths of the sequence downstream of the TGA stop codon. The relative position of the AUUUA motif is boxed (upper panel). Binding of the labelled RNA probe was analysed in the presence of 10 μg of cytoplasmic proteins prepared from TNFα-treated TM4 cells. EMSAs were performed using the 440 nt RNA probe (BE). (B) A 1–10 μg amount of CP was added to the binding reaction with the labelled RNA probe. (C) A 10–100-fold excess of the unlabelled competitor was added to the binding reaction mixtures 20 min before the addition of labelled probe. (D) A 10 μg amount of cytoplasmic proteins prepared from vehicle- or TNFα-treated cells was used in the RNA EMSAs. (E) Cytoplasmic proteins were pre-incubated with anti-TTP antibody or pre-immune serum before the addition of the radiolabelled probe. The asterisk indicates the large antibody–RNA–protein complex. CP, cytoplasmic proteins.

Figure 5
TTP is responsible for the formation of RNA–protein complex upon CLMP 3′-UTR

(A) Binding of cytoplasmic proteins to CLMP 3′-UTR riboprobes of different lengths. Numbers represent the relative lengths of the sequence downstream of the TGA stop codon. The relative position of the AUUUA motif is boxed (upper panel). Binding of the labelled RNA probe was analysed in the presence of 10 μg of cytoplasmic proteins prepared from TNFα-treated TM4 cells. EMSAs were performed using the 440 nt RNA probe (BE). (B) A 1–10 μg amount of CP was added to the binding reaction with the labelled RNA probe. (C) A 10–100-fold excess of the unlabelled competitor was added to the binding reaction mixtures 20 min before the addition of labelled probe. (D) A 10 μg amount of cytoplasmic proteins prepared from vehicle- or TNFα-treated cells was used in the RNA EMSAs. (E) Cytoplasmic proteins were pre-incubated with anti-TTP antibody or pre-immune serum before the addition of the radiolabelled probe. The asterisk indicates the large antibody–RNA–protein complex. CP, cytoplasmic proteins.

The specificity of protein interaction with the CLMP 3′-UTR was analysed further using competition assays. As shown in Figure 5(C), the RNA probe–protein interaction could be competed specifically by the addition of 10- and 100-fold molar excess of unlabelled competitors.

To determine whether TNFα would enhance the binding between the RNA-binding proteins and the RNA probe, cytosolic proteins were prepared from vehicle- or TNFα-treated TM4 cells. As shown in Figure 5(D), when cytosolic protein from TNFα-treated TM4 cells was incubated with the 440 nt probe, a larger RNA–protein complex with low mobility was formed when compared with the vehicle control. This result indicates that more RNA-binding proteins can bind to the 440 nt probe after TNFα stimulation. Antibody supershift assays were then conducted to verify whether the RNA-binding activity could be attributed to TTP (Figure 5E). With the addition of anti-TTP antibody, the formation of the RNA–protein complex was partially abolished when compared with the control. In addition, an intense complex was generated and trapped at the well (marked with an asterisk). This was probably due to the formation of a larger antibody–RNA–protein complex, making it difficult to enter the gel. The less intense RNA–protein complex left behind indicated that other RNA-binding proteins are also involved in the RNA–protein complex formation. Taken together, these results confirmed that TTP is one of the major RNA-binding proteins responsible for this RNA–protein complex formation.

DISCUSSION

Despite the recognized importance of cytokines in TJ barrier function, little information is available on the mechanism by which cytokines act directly on gene regulation of TJ proteins, particularly at the mRNA level. In the present study, we have shown that TNFα induces destabilization of CLMP mRNA and that this event is mediated via the activation of the JNK pathway and the interaction between TTP with the CLMP 3′-UTR. These findings demonstrated for the first time the direct connection between cytokine, the signal transduction pathway and the action of RNA-binding protein. Such a connection is established in regulating the expression of a TJ protein at the post-transcriptional level.

The p54 JNK isoform reduces TJ protein expression by inducing mRNA instability

CLMP is a newly identified TJ integral protein expressed in the testis [7]. By the use of immunofluorescence microscopy, CLMP was detected in areas of cell–cell contact and was co-localized with the TJ marker ZO-1, suggesting its importance in contributing to the TJ barrier function in mouse testis. In the testis, TNFα is mainly secreted by germ cells and testicular macrophages and acts on the Sertoli cell TNFα receptors [16]. Administration of TNFα on TM4 cells could therefore efficiently mimic the exogenous stimulation of Sertoli cells in in vivo situations. Using mRNA decay assay, TNFα was shown to decrease the expression of CLMP by reducing CLMP mRNA stability via the JNK signalling pathway. Previous study using the CdCl2 model has proved that the intrinsic JNK activity was activated concomitantly with a significant decrease in occludin and ZO-1 expression during BTB disruption [17]. However, it is not known whether TNFα down-regulates those TJ proteins via transcription modification or protein degradation. In this study, the destabilization of mRNA transcripts of CLMP was shown to be one of the molecular mechanisms utilized by TNFα and JNK in regulating TJ protein expression. It is of interest to note that TNFα could differentially activate the cytosolic p54 JNK isoform, but not the p46 JNK isoform, in TNFα-mediated TJ disruption. This is not unusual, as similar differential expression and activation of the p54 JNK isoform were observed in cytosolic fraction of heart tissue upon protein kinase C activation, whereas the phosphorylated state of the p46 JNK isoform in the cytosolic fraction remained unchanged [18]. It is apparent that the p54 JNK isoform is the predominant activated isoform in the cytosolic fraction. Since the regulation of mRNA stability occurs in cytoplasm, there is no doubt that the cytosolic p54 JNK isoform, rather than the p46 JNK isoform, is activated upon TNFα stimulation.

Role of TTP in mouse testis and its effect on CLMP 3′-UTR

TTP has been described as a CCCH (Cys-Cys-Cys-His) tandem zinc-finger protein that binds to mRNA transcripts containing AREs (AU-rich elements) and promotes mRNA degradation [14]. Although TTP mRNA is expressed in a variety of adult mouse tissues, TTP protein expression is highly restricted to liver and testis in male mice, suggesting a potential tissue-specific function [15]. Despite the extensive studies on the role of TTP in moderating inflammation in macrophages, the functional significance of TTP in the testis is entirely unknown. This is so far the first report to study the function of TTP in the testis and to unravel its role in modulating the expression of TJ proteins by alteration of CLMP mRNA stability.

Previous study has showed that TTP is transiently induced in macrophage in response to TNFα and, in return, inhibits TNFα production by destabilizing its mRNA [14]. As in macrophages, the expression of TTP in TM4 cells could also be elevated upon TNFα stimulation. To investigate the role of TTP in regulating CLMP mRNA stability, a reporter system generated by subcloning the CLMP 3′-UTR into the reporter construct driven by the CLMP promoter p[−946/−1]Luc was used. It was demonstrated that TTP could induce CLMP mRNA degradation, since co-transfection of pCMV-TTP with the p[−946/−1]Luc-3′-UTR construct significantly reduced the luciferase activity, whereas no significant change was observed when empty pCMV was co-transfected.

CLMP 3′-UTR contains a single copy of the AUUUA motif which has been known as a binding site for TTP (for a review, see [19]). It is located 387 nt downstream of the TGA stop codon. In RNA EMSA analysis, a large RNA–protein complex was formed when cytosolic proteins were incubated with either a 440 nt or a 665 nt RNA probe, but not with the 220 nt RNA probe, seemingly suggesting that TTP could bind to the region between nucleotides 220 and 440 of the CLMP 3′-UTR where the AUUUA motif is located. In the antibody supershift experiment, addition of anti-TTP antibody could partially abolish the formation of RNA–protein complex, and a supershifted band was formed and trapped at the sample well region. These suggest that the binding of TTP to the CLMP 3′-UTR is involved in the regulation of CLMP mRNA gene expression. Taken together, we have demonstrated for the first time that TTP functions as a negative regulator in CLMP gene regulation by interacting with the 3′-UTR of the CLMP transcripts to speed up their degradation.

Post-transcriptional gene regulation, particularly mRNA turnover, is increasingly recognized as an important means of controlling gene expression [20]. During spermatogenesis, the TJ at the BTB must open periodically to allow the migration of germ cells from the basal to the adluminal compartment without compromising the junction integrity. As such, different TJ proteins should be expressed on a timely regulated basis so that the BTB could ‘open’ and ‘close’ at specific stages of germ cell development. Altering mRNA transcript stability of different TJ proteins can be one of the most efficient mechanisms that leads to the precise expression of different TJ proteins. In the present study, we have identified a signalling pathway involving the p54 JNK isoform and TTP that leads to CLMP mRNA degradation. These findings are particularly important, because it is the first study showing that mRNA stability is involved in regulating TJ protein expression, and in connection with the TNFα-dependent TJ barrier impairment. It is expected that additional studies on the control of mRNA turnover and other post-transcriptional events on TJ proteins will be valuable for a thorough understanding of the BTB dynamics.

We thank Dr P. J. Blackshear for providing the TTP expression vector. This work was supported by grants from the Hong Kong Research Grant Council (HKU7536/05M, HKU7599/06M and HKU7609/06M) and The University of Hong Kong (CRCG Seed Funding).

Abbreviations

     
  • BTB

    blood–testis barrier

  •  
  • CLMP

    coxsackie- and adenovirus-receptor-like membrane protein

  •  
  • EMSA

    electrophoretic mobility shift assay

  •  
  • GFP

    green fluorescent protein

  •  
  • JNK

    c-Jun N-terminal kinase

  •  
  • MAPK

    mitogen-activated protein kinase

  •  
  • RT

    reverse transcription

  •  
  • SAPK

    stress-activated protein kinase

  •  
  • siRNA

    small interfering RNA

  •  
  • TJ

    tight junction

  •  
  • TNFα

    tumour necrosis factor α

  •  
  • TTP

    tristetraprolin

  •  
  • UTR

    untranslated region

  •  
  • ZO-1

    zonula occludens 1

References

References
1
Cheng
C. Y.
Mruk
D. D.
Cell junction dynamics in the testis: Sertoli–germ cell interactions and male contraceptive development
Physiol. Rev.
2002
, vol. 
82
 (pg. 
825
-
874
)
2
Li
M. W.
Xia
W.
Mruk
D. D.
Wang
C. Q.
Yan
H. H.
Siu
M. K.
Lui
W. Y.
Lee
W. M.
Cheng
C. Y.
Tumor necrosis factor α reversibly disrupts the blood-testis barrier and impairs Sertoli–germ cell adhesion in the seminiferous epithelium of adult rat testes
J. Endocrinol.
2006
, vol. 
190
 (pg. 
313
-
329
)
3
Hellani
A.
Ji
J.
Mauduit
C.
Deschildre
C.
Tabone
E.
Benahmed
M.
Developmental and hormonal regulation of the expression of oligodendrocyte-specific protein/claudin 11 in mouse testis
Endocrinology
2000
, vol. 
141
 (pg. 
3012
-
3019
)
4
Siu
M. K.
Lee
W. M.
Cheng
C. Y.
The interplay of collagen IV, tumor necrosis factor-α, gelatinase B (matrix metalloprotease-9), and tissue inhibitor of metalloproteases-1 in the basal lamina regulates Sertoli cell–tight junction dynamics in the rat testis
Endocrinology
2003
, vol. 
144
 (pg. 
371
-
387
)
5
Patrick
D. M.
Leone
A. K.
Shellenberger
J. J.
Dudowicz
K. A.
King
J. M.
Proinflammatory cytokines tumor necrosis factor-α and interferon-γ modulate epithelial barrier function in Madin–Darby canine kidney cells through mitogen activated protein kinase signaling
BMC Physiol.
2006
, vol. 
6
 pg. 
2
 
6
Mankertz
J.
Tavalali
S.
Schmitz
H.
Mankertz
A.
Riecken
E. O.
Fromm
M.
Schulzke
J. D.
Expression from the human occludin promoter is affected by tumor necrosis factor α and interferon γ
J. Cell Sci.
2000
, vol. 
113
 (pg. 
2085
-
2090
)
7
Raschperger
E.
Engstrom
U.
Pettersson
R. F.
Fuxe
J.
CLMP, a novel member of the CTX family and a new component of epithelial tight junctions
J. Biol. Chem.
2004
, vol. 
279
 (pg. 
796
-
804
)
8
Sze
K.-L.
Lee
W.
Lui
W.-Y.
Expression of CLMP, a novel tight junction protein, is mediated via the interaction of GATA with the Kruppel family proteins, KLF4 and Sp1, in mouse TM4 Sertoli cells
J. Cell. Physiol.
2008
, vol. 
214
 (pg. 
334
-
344
)
9
Lui
W. Y.
Lee
W. M.
cAMP perturbs inter-Sertoli tight junction permeability barrier in vitro via its effect on proteasome-sensitive ubiquitination of occludin
J. Cell. Physiol.
2005
, vol. 
203
 (pg. 
564
-
572
)
10
Livak
K. J.
Schmittgen
T. D.
Analysis of relative gene expression data using real-time quantitative PCR and the 2−ΔΔCt method
Methods
2001
, vol. 
25
 (pg. 
402
-
408
)
11
Lasa
M.
Mahtani
K. R.
Finch
A.
Brewer
G.
Saklatvala
J.
Clark
A. R.
Regulation of cyclooxygenase 2 mRNA stability by the mitogen-activated protein kinase p38 signaling cascade
Mol. Cell. Biol.
2000
, vol. 
20
 (pg. 
4265
-
4274
)
12
Hel
Z.
Skamene
E.
Radzioch
D.
Two distinct regions in the 3′ untranslated region of tumor necrosis factor α mRNA form complexes with macrophage proteins
Mol. Cell. Biol.
1996
, vol. 
16
 (pg. 
5579
-
5590
)
13
De Cesaris
P.
Starace
D.
Riccioli
A.
Padula
F.
Filippini
A.
Ziparo
E.
Tumor necrosis factor-α induces interleukin-6 production and integrin ligand expression by distinct transduction pathways
J. Biol. Chem.
1998
, vol. 
273
 (pg. 
7566
-
7571
)
14
Carballo
E.
Lai
W. S.
Blackshear
P. J.
Feedback inhibition of macrophage tumor necrosis factor-α production by tristetraprolin
Science
1998
, vol. 
281
 (pg. 
1001
-
1005
)
15
Lu
J. Y.
Schneider
R. J.
Tissue distribution of AU-rich mRNA-binding proteins involved in regulation of mRNA decay
J. Biol. Chem.
2004
, vol. 
279
 (pg. 
12974
-
12979
)
16
De
S. K.
Chen
H. L.
Pace
J. L.
Hunt
J. S.
Terranova
P. F.
Enders
G. C.
Expression of tumor necrosis factor-α in mouse spermatogenic cells
Endocrinology
1993
, vol. 
133
 (pg. 
389
-
396
)
17
Wong
C. H.
Mruk
D. D.
Siu
M. K.
Cheng
C. Y.
Blood–testis barrier dynamics are regulated by α2-macroglobulin via the c-Jun N-terminal protein kinase pathway
Endocrinology
2005
, vol. 
146
 (pg. 
1893
-
1908
)
18
Ping
P.
Zhang
J.
Huang
S.
Cao
X.
Tang
X. L.
Li
R. C.
Zheng
Y. T.
Qiu
Y.
Clerk
A.
Sugden
P.
, et al. 
PKC-dependent activation of p46/p54 JNKs during ischemic preconditioning in conscious rabbits
Am. J. Physiol.
1999
, vol. 
277
 (pg. 
H1771
-
H1785
)
19
Carrick
D. M.
Lai
W. S.
Blackshear
P. J.
The tandem CCCH zinc finger protein tristetraprolin and its relevance to cytokine mRNA turnover and arthritis
Arthritis Res. Ther.
2004
, vol. 
6
 (pg. 
248
-
264
)
20
Ross
J.
mRNA stability in mammalian cells
Microbiol. Rev.
1995
, vol. 
59
 (pg. 
423
-
450
)