Published online before print January 24, 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Critical Care Medicine Department, Clinical Center, National Institutes of Health, Bethesda, Maryland, USA
1 Correspondence: Critical Care Medicine Department, Bldg. 10, Rm. 2C145, Clinical Center, National Institutes of Health, Bethesda, MD 20892, USA. E-mail: rdanner{at}cc.nih.gov
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, and p21/Waf1 3'UTRs were inserted into a luciferase (LUC) reporter gene system and found to decrease LUC activity and mRNA half-life in transfected THP-1 cells. The inhibitory effect of these 3'UTRs on LUC expression inversely correlated with the number of AUUUA motifs. Sequence truncation of the IL-8 3'UTR revealed that two segments, one with AURE sites and another without, contributed to mRNA destabilization. NO• activation of p38 MAPK increased LUC activity and mRNA half-life for reporter constructs that contained either of these IL-8 3'UTR segments. AURE-dependent and -independent NO• effects were blocked by p38 MAPK inhibition, and AURE-dependent effects were also blocked by site-directed mutagenesis of AUUUA sites. Two proteins, HuR and heterogeneous nuclear ribonucleoprotein A0, were identified, which bound to the AURE-containing region of exogenous and endogenous IL-8 mRNA in a NO•-p38 MAPK-dependent manner. These results demonstrate that NO•-p38 MAPK signaling can stabilize mRNA via AURE-dependent and -independent mechanisms.
Key Words: signal transduction • post-transcriptional gene regulation • HuR • hnRNP A0
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The core sequence of AUREs is the AUUUA motif. AURE-containing transcripts have been grouped into three classes based on the distribution of this motif [8
, 10
, 11
]. Class I AURE transcripts, such as p21/Waf1, a cell-cycle master regulatory gene, contain multiple isolated AUUUA motifs. Class II AURE transcripts, which include IL-8 and TNF-, contain at least two overlapping AUUUA motifs. Class III AURE transcripts (e.g., c-jun) contain no canonical AUUUA motifs but rather general U-rich or AU-rich regions that are nonetheless functionally active [8
, 10
, 11
]. To date, a large number of AURE-binding proteins have been identified, including human protein R (HuR), AU factor 1 (AUF1), tristetraproline (TTP), butyrate response factor 1 (BRF1), K homology-type splicing regulatory protein (KSRP), and other members of the heterogeneous nuclear ribonucleoprotein (hnRNP) family of proteins [11
]. Although all of these proteins bind to mRNA and interact with AUREs, they differ in function, transcript and cell specificity, and responsiveness to various signaling pathways. HuR stabilizes [12
], but TTP [13
], BRF1 [11
], and KSRP [14
] destabilize mRNA. AUF1 can stabilize or destabilize AURE-containing transcripts, depending on the relative abundance of different AUF1 isoforms [15
]. Various extracellular signals have been shown to target one or more of these AURE-binding proteins and thereby regulate mRNA degradation or translation [3
, 4
, 9
, 16
17
18
].
NO• is an important signaling molecule that regulates a wide range of cellular activities, including gene expression. It has been shown that NO• regulates transcription through specificity protein 1 (Sp1) [19
, 20
], NF-
B [21
], AP-1 [20
, 22
], peroxisome proliferator-activated receptor-
[23
], early growth response gene-1 [24
], and hypoxia-inducible factor-1 (HIF-1) [25
]. Using oligonucleotide microarrays in LPS-stimulated human monocytic THP-1 cells, we recently demonstrated that NO• stabilized a large set of mRNA transcripts by activation of Erk1/2 or p38 MAPK [4
]. Sequence analysis revealed that cytosine-uridine-rich elements (CURE), another post-transcriptional regulatory motif found in the 3'UTR of mRNA, were over-represented in the mRNA cluster stabilized by NO•-Erk signaling [4
]. Further experiments demonstrated that this signaling pathway induced the translocation of hnRNP K and hnRNP E2/E1 to the cytoplasm, where these proteins bound to CURE sites, thereby stabilizing affected transcripts while inhibiting their translation [4
]. NO•-p38 MAPK signaling was found to stabilize a different set of genes that had a preponderance of AURE sites in their 3'UTRs [4
, 26
]. Consistent with this microarray result, p38 MAPK signaling has been implicated in regulating the mRNA half-life of more than 40 AURE-containing genes [18
], including TNF- [27
], p21/Waf1 [26
, 28
], and IL-8 [27
, 29
]. Importantly, NO• was shown in LPS-stimulated THP-1 cells [4
, 30
] and PMA-differentiated U937 cells [26
, 28
] to stabilize mRNA through p38 MAPK activation. However, the specific 3'UTR sequences and configuration of elements that transduce NO• signaling to individual transcripts as well as the identity of the binding proteins involved remain unknown.
In this investigation, NO•-induced mRNA stabilization was examined at the level of the 3'UTR sequence, AURE number and location, and protein-binding interactions in human THP-1 cells. A firefly luciferase (LUC) reporter gene system was used to first determine how AURE class affects the impact of NO• on LUC activity and mRNA half-life. Sequence truncation and site-directed mutagenesis were then used to further delimit the IL-8 3'UTR elements that control mRNA decay and the response to NO•-p38 MAPK signaling. RNA EMSA (REMSA), UV cross-link assays, and RNA immunoprecipitation were finally used to identify the proteins that transduce NO•-p38 signaling to specific 3'UTR elements within IL-8 transcripts and thereby regulate their stability.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-tubulin was purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). RNasin, the plasmid pGL3 that expresses a firefly LUC reporter gene through the SV40 promoter, and the plasmid pRL-TK expressing Renilla LUC were purchased from Promega (Madsion, WI, USA). THP-1 cells, a human monocytic line obtained from American Type Culture Collection (Manassas, VA, USA), were maintained in RPMI, supplemented with 10% FCS (Cellgro, Herndon, VA, USA) as described previously [4 ].
Construction of 3'UTR plasmids
As naïve THP-1 cells do not constitutively express IL-8 and TNF-, we first stimulated cells with LPS (1 µg/mL) for 1 h to induce the transcription of these genes. Total RNA was then extracted from cells using the RNeasy mini kit (Qiagen Sciences, Germantown, MD, USA). First-strand cDNA was synthesized from 3 µg total RNA and then used to amplify the 3'UTRs of IL-8 (nt 937–1352 of GenBank sequence Y00787), TNF- (nt 1235–1481 of GenBank sequence X01394), and p21/Waf1 (nt 557–857 of GenBank sequence U03106) using the SuperScriptTM first-strand synthesis system for a RT-PCR kit (Invitrogen Life Technologies, Carlsbad, CA, USA). These 3'UTR PCR products were finally cloned into pGL3 at the XbaI site residing just downstream of LUC to make pGL3-IL8, pGL3-TNF, and pGL3-p21/Waf1. The construct pGL3-IL8, containing the wild-type IL-8 3'UTR, was used as a template for PCR to generate various truncation mutants of IL-8 3'UTR. These mutants were then inserted into the Xbal I site of pGL3 to create pGL3-IL8A, pGL3-IL8B, pGL3-IL8B2, pGL3-IL8C, and pGL3-IL8D. Plasmid pGL3-IL8M was generated from pGL3-IL8B by mutating all five AUUUA motifs to AUGUA using the QuickChange® multi-site-directed mutagenesis kit (Stragene, La Jolla, CA, USA). Antisense controls of p21/Waf1, IL-8, and TNF- 3'UTR had no AURE, or AUUUA motifs were mutated. Sequences of all 3'UTR constructs were confirmed by DNA sequencing. The sequences of PCR primers are shown in Table 1
.
|
View this table: [in a new window] |
Table 1. Sequences of PCR Primers
|
Quantitative real-time RT-PCR (qRT-PCR)
Total RNA was isolated from THP-1 cells using RNeasy mini kits (Qiagen Inc., Valencia, CA, USA). TaqMan® qRT-PCR (ABI, Rockville, MD, USA) was used to quantify mRNA levels. Gene-specific probes and PCR primers for LUC and Renilla LUC were described previously [4
]. Human IL-8 mRNA-specific probes and primers were obtained from ABI. The high-capacity cDNA archive kit (ABI) was used to prepare cDNA from 2 µg total RNA. Resulting cDNA was used for RT-PCR in triplicate, according to the standard ABI protocol. LUC mRNA was normalized to Renilla LUC mRNA.
Preparation of 3'UTR riboprobes
The plasmid pGL3-IL8 containing wild-type IL-8 3'UTR was used as a template for PCR amplification of different IL-8 3'UTR regions: IL8A and IL8B. All PCR primers contained the T7 promoter (GGCCAGTGAATTGTAATACGACTCACTATAGGGAGGCG) at the 5' end, followed by a target cDNA sequence at the 3' end (see Table 1
). These resulting PCR products were then used for in vitro transcription to generate biotin-labeled IL8A and IL8B riboprobes. The in vitro transcription was carried out using Riboprobe® in vitro transcription systems (Promega). Unlabeled RNA was synthesized with 1 mM UTP instead of biotin-16-UTP.
REMSA
Cytoplasmic fractions were extracted using the Nu-CLEARTM extraction kit (Sigma-Aldrich). Binding reactions between labeled riboprobe (70 ng) and cytoplasmic extract (10 µg) were performed in 20 µL buffer (15 mM HEPES, pH 7.4, 10 mM KCl, 5 mM MgCl2, 5% glycerol, 0.2 mM DTT, and 50 µg/mL yeast tRNA) for 30 min at room temperature, followed by digestion with RNase T1 (150 U) for 20 min at 37°C. RNA-protein complexes were separated in 4–20% polyacrylamide gel with 0.5x Tris/boric acid/EDTA buffer, transferred onto a nylon membrane, and detected using the LightShiftTM chemiluminescent EMSA kit (Pierce, Rockford, IL, USA). In competition experiments, a 100-fold molar excess of unlabeled probe was added to the incubation mixture. For antibody supershift assays, 2 µg specific polyclonal antibody or control rabbit serum was preincubated with cytoplasmic proteins for 20 min at room temperature prior to addition of probes.
For UV cross-link studies, after the RNase T1 digestion, reaction mixtures were irradiated in a UV oven (Stratalinker 1800, Stragene) at 254 nm for 5 min on an automatic setting. Binding reactions were then solubilized in SDS sample buffer and boiled for 3 min. RNA-protein complexes were then applied to a 4–20% SDS-PAGE gel for Western blotting.
RNA immunoprecipitation
Naive THP-1 cells do not constitutively express IL-8, so cells were activated with LPS (1 µg/mL) for 1 h to induce mRNA transcription. After treatment with or without the p38 MAPK inhibitor SB (0.1 µM) for 30 min, cells were further incubated for 3 h with fresh or decomposed NONO (300 µM), followed by isolation of cytoplasmic extracts in the presence of RNasin (100 units/mL) and vanadyl ribonucleoside complexes (2 mM). The extract for each condition was divided into three equal aliquots. The first was used to purify cytoplasmic RNA for measurement of the housekeeping gene β-actin. The remaining two aliquots were subjected to immunoprecipitation with anti-HuR and anti-hnRNP A0 anitbodies, as described previously [31
, 32
]. RNA was extracted from the precipitates using the RNeasy mini kit. IL-8 mRNA recovered by immunoprecipitation with the two different antibodies was measured by qRT-PCR and normalized to β-actin mRNA for each of the experimental conditions.
Data analysis
LUC mRNA decay was analyzed using a three-way ANOVA procedure (the first factor was 3'UTR construct or treatment, the second factor was time, the third factor was experiment), followed by appropriate post hoc tests for the comparisons of interest. LUC activities and IL-8 mRNA levels recovered by immunoprecipitation were compared between experimental conditions using paired t-tests. The correlation of LUC activity with number of 3'UTR AUUUA motifs was calculated using the Pearson method. Data are presented as mean ± SEM of at least three independent experiments. Differences were considered significant when two-sided P values were equal to or less than 0.05.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, and p21/Waf1 into an expression plasmid just downstream from the reporter gene LUC. IL-8 and TNF- have five and eight class II AURE motifs, respectively, and p21/Waf1 has three class I AURE motifs (Fig. 1A
). Compared with antisense sequences of identical length but with no AURE sites, insertion of these AURE-containing sense 3'UTRs representing two different classes all significantly decreased LUC activity (Fig. 1B
; P<0.045 for all). Likewise, these 3'UTRs also decreased mRNA stability compared with the parental plasmid pGL3 (Fig. 1C
; P<0.001 for all). More interestingly, LUC activity inversely correlated with the number of AUUUA motifs (r=–0.955; P=0.044). With an increasing number of AURE sites, LUC activity fell (Fig. 1B
for sense 3'UTR), and mRNA half-life became progressively shorter (Fig. 1C)
. Variations in 3'UTR length appeared somewhat less important for these genes, as pGL3-p21/Waf1 had the longest mRNA half-life, although its 3'UTR length (300 nt) was intermediate between pGL3-TNF (246 nt) and pGL3-IL8 (413 nt).
 View larger version (18K): [in a new window] |
Figure 1. LUC reporter gene activity and mRNA stability correlate with AURE number in 3'UTRs. THP-1 cells were transfected with pGL3 or pGL3 constructs containing one of the following sense or antisense 3'UTR sequences: p21/Waf1 3'UTR (pGL3-p21/Waf1), IL-8 3'UTR (pGL3-IL8), or TNF- 3'UTR (pGL3-TNF). (A) Schematic of the sense 3'UTRs of human p21/Waf1, IL-8, and TNF-. (B) After 24 h incubation in culture media, cells were collected to measure LUC activities. (C) After 30 min pretreatment with ActD (2.5 µg/mL), cells were further incubated for 0–4 h to measure LUC mRNA stability by qRT-PCR for the pGL3 constructs containing the sense sequence of 3'UTR. Values represent the mean ± SEM of three experiments each performed in triplicate.
|
, and p21/Waf1 all have multiple AURE motifs in their 3'UTRs (Fig. 1A)
. To further dissect the functions of differently placed AURE motifs, we focused on the 3'UTR of IL-8. Different reporter gene constructs that contain the wild-type 3'UTR of IL-8 (nt 937–nt 1352) or various mutants (Fig. 2A
) were generated and transfected into THP-1 cells. As seen in Figure 2B
, transfection with pGL3-IL8, pGL3-IL8A, and pGL3-IL8B resulted in LUC activity that was only 33 ± 2%, 62 ± 2%, and 56 ± 3%, respectively, of the level produced by parental pGL3 (P<0.004 for all three comparisions). These results suggest that Region A (nt 937–nt 1045) without AURE sites and Region B (nt 1046–nt 1325) with five AURE motifs significantly contribute to reduced mRNA stability, and these effects are additive. As in Figure 1
comparing the 3'UTRs of different genes, mRNA stability tended to increase as the AURE motif number decreased in various versions of the IL-8 3'UTR. For pGL3-IL8B, pGL3-IL8B2, pGL3-IL8C, and pGL3-IL8D, which have 5, 3, 1, and 0 AURE motifs (Fig. 2A)
, respectively, relative LUC activity increased from 56 ± 3% to 89 ± 2% (Fig. 2B
; P<0.001). Site-directed mutagenesis of all five AURE motifs (AUUUA to AUGUA) in pGL3-IL8B significantly increased LUC activity (Fig. 2B
; P=0.0004, pGL3-IL8M vs. pGL3-IL8B), confirming the mRNA destabilizing effect of AURE sites.
 View larger version (14K): [in a new window] |
Figure 2. Differential effects of AUREs and various IL-8 3'UTR segments on mRNA stability. (A) Schematic of the 3'UTR of the human IL-8 and various mutants: wild-type, truncation mutants A, B, B2, C, and D, and site-directed mutagenesis mutant M with disrupted AUUUA sites. Each ellipse indicates an AURE site or a mutated AURE site (strike-through). (B) THP-1 cells were transfected with pGL3 or pGL3 constructs containing one of the following IL-8 3'UTR sequences: wild-type IL8 (pGL3-IL8), truncation mutant A (pGL3-IL8A), B (pGL3-IL8B), B2 (pGL3-IL8B2), C (pGL3-IL8C), or D (pGL3-IL8D), or AURE site-directed mutagenesis mutant M (pGL3-IL8M). After 24 h incubation in culture media, cells were collected to measure reporter gene LUC activities. Values represent the mean ± SEM of three experiments, each performed in triplicate (*, P 0.01, for each construct vs. pGL3; **, P=0.0004, for pGL3-IL8M vs. pGL3-IL8B).
|
were shown to destabilize the mRNA of an artificial construct, an effect closely associated with the presence of AUUUA sequences. Next, NO•-p38 MAPK stabilization of IL8 mRNA was investigated at the level of 3'UTRs. As shown in Figure 3A
, NO• donor dose-dependently increased p38 MAPK phosphorylation in THP-1 cells, and SB, a specific p38 MAPK inhibitor, blocked this effect. NO• also increased LUC activity in cells transfected with pGL3-IL8 and pGL3-IL8B, which have AURE sites in their 3'UTR (Fig. 3B
; P<0.02 for NONO vs. Control), an effect abolished by SB (Fig. 3B
; P<0.05 for SB/NONO vs. NONO). Consistent with these results, site-directed mutagenesis of AURE sites in pGL3-IL8B not only increased LUC activity (Fig. 3B
; P=0.008, pGL3-IL8M vs. pGL3-IL8B within Control) but also eliminated the up-regulatory effect of NO• (Fig. 3B
; P=0.4575, NONO vs. Control for pGL3-IL8M). Unexpectedly, similar to its effect on pGL3-IL8- and pGL3-IL8B-transfected cells, NO• increased LUC activity after pGL3-IL8A transfection, although this plasmid construct does not have recognized AURE sites in its 3'UTR (Fig. 3B
; P=0.02, NONO vs. Control for pGL3-IL8A). Associated with increased LUC activity, NO•-p38 MAPK signaling also significantly increased the stability of LUC mRNA for the pGL3-IL8A and pGL3-IL8B constructs (Fig. 3C
; P<0.001 for NONO vs. Control), an effect blocked by SB, a specific p38 MAPK inhibitor (Fig. 3C
; P<0.001 for SB/NONO vs. NONO).
 View larger version (25K): [in a new window] |
Figure 3. NO•-p38 MAPK signaling regulates LUC reporter gene activity and mRNA stability through inserted 3'UTRs derived from IL-8. (A) NO• increases p38 MAPK phosphorylation, an effect blocked by the p38 MAPK inhibitor SB (0.1 µM). THP-1 cells were pretreated with or without SB for 30 min. After further incubation with NONO (0–500 µM) for 30 min, cells were then lysed for Western blotting. The experiment was repeated twice with similar results. (B) THP-1 cells were transfected with pGL3 or pGL3 constructs containing one of the following 3'UTR sequences of IL-8: wild-type IL-8 (pGL3-IL8), truncation mutant A (pGL3-IL8A) or B (pGL3-IL8B), or AURE site-directed mutagenesis mutant M (pGL3-IL8M). After treatment with or without the p38 MAPK inhibitor SB (0.1 µM) for 30 min, cells were further incubated for 24 h with NONO (300 µM) or decomposed NONO (Control), followed by the measurement of LUC activity (*, P0.02, for NONO vs. Control; **, P<0.05, for NONO vs. SB/NONO). (C) THP-1 cells were transfected as in B. After 30 min pretreatment with ActD (2.5 µg/mL), in the absence or presence of SB (0.1 µM), cells were further incubated for 0–8 h with NONO (300 µM) or decomposed NONO [Control (Contl)], followed by the measurement of LUC mRNA levels by qRT-PCR. Values represent the mean ± SEM of three experiments each performed in triplicate.
|
 View larger version (74K): [in a new window] |
Figure 4. NO· induces binding of cytoplasmic proteins to the 3'UTR of IL-8. (A) The 3'UTR sequence of IL-8 showing the full-length wild-type strand (IL8Wt) and the truncation mutants A (IL8A) and B (IL8B). (B) REMSA with riboprobe IL8A and (C) riboprobe IL8B. THP1 cells were treated with NONO (300 µM) or decomposed NONO (Control) for 3 h, followed by the isolation of cytoplasmic protein for REMSA. Normal serum or antibodies against HuR or hnRNP A0 were preincubated with the samples as indicated in supershift experiments. Arrows indicate specific complexes formed between riboprobes and cytoplasmic protein. NS, Free riboprobe or nonspecific-binding complexes.
|
-tubulin validated equal loading of cytoplamic extracts (Fig. 5A
and 5B)
. Collectively, these results indicate that NO• increases the binding of HuR and hnRNP A0 to AURE sites in the 3'UTR of IL-8, an event that is associated with mRNA stabilization.
 View larger version (19K): [in a new window] |
Figure 5. UV-cross-link assay of HuR and hnRNP A0 with the 3'UTR of IL-8. THP-1 cells were treated with NONO (300 µM) or decomposed NONO (NONO Control) for 3 h prior to the preparation of cytoplasmic proteins. Riboprobe IL8B, which covers region B of the 3'UTR of IL-8, was incubated with cytoplasmic proteins and digested with RNase T1. Thereafter, the mixtures were irradiated by UV on ice and resolved by SDS-PAGE (4–20% gel) for Western blotting to detect (A) HuR and (B) hnRNP A0. The same blots were stripped and incubated with -tubulin antibody to show the relative amount of protein loading.
|
 View larger version (30K): [in a new window] |
Figure 6. NO•-p38 MAPK signaling induces binding of cytoplasmic HuR and hnRNP A0 to endogenous IL-8 mRNA. THP-1 cells were activated with LPS (1 µg/mL) for 1 h to induce mRNA transcription. After treatment with or without the p38 MAPK inhibitor SB (0.1 µM) for 30 min, cells were further incubated for 3 h with NONO (300 µM) or decomposed NONO (Control), followed by isolation of cytoplasmic extracts. (A) Cytoplasmic extracts were immunoprecipitated with anti-HuR and anti-hnRNP A0, followed by measurement of associated IL-8 mRNA by qRT-PCR. Values represent the mean ± SEM of three experiments each performed in triplicate (*, P<0.02, for NONO vs. Control; **, P<0.04, for NONO vs. SB/NONO). (B) Western blotting of cytoplasmic HuR and hnRNP A0 across the four experimental conditions.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
are potent cytokines that play major roles in propagating the innate immune response [30
]. These three genes have class I or II AURE sites within their 3'UTRs [8
, 11
]. In the present study, we demonstrated that the 3'UTRs of p21/Waf1, TNF-, and IL-8 destabilized the mRNA of a heterologous reporter, and mRNA half-life was inversely correlated with the number of AURE motifs. This is consistent with the results of a recent microarray study in which mRNAs with more class I/II AUUUA motifs were shown to have a shorter half-life than those with fewer motifs [18
]. Approximately 4000 AURE-containing mRNAs have been computationally mapped to the human genome in the 3'UTR AURE database (http://rc.kfshrc.edu.sa/ared/) [8
]. Of these, AURE-containing mRNAs, genes involved in cell proliferation and immune response, are significantly over-represented when compared with the overall genome [8
]. Therefore, AURE-mediated mRNA degradation appears to be a major mechanism by which cells regulate the expression of genes with rapid turnover that control the cell cycle and inflammation, such as p21/Waf1, TNF-, and IL-8. The rapid degradation of genes that are typically induced quickly and strongly by extracellular signals via transcription likely allows for the precise control of cell cycle progression and limits the collateral damage that can result from an excessive immune response. The wild-type 3'UTR of IL-8 studied here spans 416 nt (nt 937–1352 of GenBank sequence Y00787), which were divided into two main regions, A and B, for the purpose of analysis. Region B (307 nt; nt 1046–1352), containing five AUUUA motifs, substantially destabilized the mRNA of an artificial reporter construct. Each AURE motif appeared to independently contribute to this effect. Consistent with our findings, another study identified a 29-nt Region B sequence from nt 1048 to 1076 with four AUUUA motifs that functioned as a "core domain" for mRNA destabilization [33 ]. We also found that the 150 nt Region D (nt 1203–1352) at the 3' end of IL8B showed a slight ability to suppress reporter gene expression, as did a mutated Region B, which had no AUUUA motifs but still contained an intact D sequence. These results suggest that a minor destabilizing element may exist within Region D. A search of UTRscan database (http://www.ba.itb.cnr.it/BIG/UTRScan/) [34 ] with the D sequence found a putative internal ribosome entry site (IRES) at nt 1259–1352. IRES, often present within the 5' UTR of virus RNA and known to mediate cap-independent protein translation of virus-derived RNA, has also been associated with heterologous mRNA degradation [35 ]. Therefore, it is possible that this putative IRES element was responsible for the modest suppression of reporter gene constructs containing mutated AURE sites but an intact IL8D region (pGL3-IL8M).
The 109 nt region A (nt 937–1045) of the IL-8 3'UTR has no recognized AURE motifs. Unexpectedly, this region alone also demonstrated a relatively strong destabilizing effect on reporter mRNA. The 31-nt sequence (nt 1017–1047) at the 3' end of Region A was shown previously to lack destabilizing activity but was proposed to function as an "auxiliary domain" that might enhance adjacent AURE effects [33 ]. Our work clearly shows that Region A as defined here (nt 937–1045) has substantial mRNA-destabilizing activity in the absence of adjacent AURE sites. The possible presence of noncanonical AURE sites does not appear to explain these effects, as Region A did not complex with the same RNA-binding proteins that interact with Region B. Further investigations will be required to identify these previously unrecognized, destabilizing elements. The wild-type 3'UTR of IL-8, comprised of Regions A and B together, has a stronger effect on mRNA degradation than Regions A and B alone. Therefore, Regions A and B within the wild-type 3'UTR of IL-8 appear to collectively control basal mRNA decay.
Transcripts of p21/Waf1 [26
], TNF- [5
], and IL-8 [31
, 33
] have been shown to be stabilized by p38 MAPK in various cell lines. We have also shown that NO•-induced p38 MAPK activation in THP-1 cells stabilizes the mRNA transcripts of p21/Waf1 and IL-8 [26
, 30
]. Consistent with these results, we further found here that NO• activation of p38 MAPK stabilized a heterologous mRNA reporter construct, pGL3-IL8, which contained the 416-nt wild-type 3'UTR of IL-8 (nt 937–1352). Moreover, NO•-p38 MAPK signaling also stabilized LUC mRNA containing the IL8A or IL8B 3'UTRs. Mutation of the AURE sites in IL8B (IL8M) made LUC mRNA more stable and unresponsive to NO•. These results indicate that AUUUA motifs mediate, at least in part, the ability of NO• to stabilize IL-8 mRNA. In support of this conclusion, REMSA, using cytoplamic extracts from THP-1 cells, demonstrated that NO• induces the formation of an IL8B protein complex. Moreover, the importance of AURE sites in NO• modulation of mRNA stability is underscored by our previous microarray findings that AURE-containing mRNAs are over-represented among NO•-stabilized but p38 MAPK inhibitor-destabilized transcripts [4
, 26
].
Notably, Region A, within the full-length of 3'UTR of IL-8, as discussed above, does not contain any discernible AURE sequences. Unexpectedly, NO•-p38 MAPK signaling also stabilized LUC mRNA, which only contained the IL8A region. Thus, it seems likely that unidentified, non-AURE elements in Region A can also mediate NO•-p38 MAPK mRNA stabilization. The full-length 3'UTR of IL-8 destabilized our artificial mRNA construct more than Region A or Region B alone. The effects of Regions A and B were roughly equal and appeared additive. Further, NO• induced protein binding to both regions; although consistent with the dissimilar sequences, the complexes were not found to contain the same proteins.
A number of AURE-binding proteins have been identified, of which HuR, KSRP, and TIAR have been reported to bind to the 3'UTR of IL-8 in HeLa or breast cancer cells [31 , 33 ]. In THP-1 cells here, REMSA, UV cross-linking, and RNA immunoprecipitation demonstrated that HuR and hnRNP A0 bind to exogenous and endogenous IL8 mRNA. NO• treatment of cells increased this binding, and a p38 MAPK inhibitor blocked the NO• effect. These protein-binding effects were associated with a NO•-p38 MAPK-mediated increase in the mRNA half-life. Signaling through p38 MAPK has been shown to increase cytoplasmic accumulation of HuR via MAPK-activated protein kinase 2 (MK2) [36 ], and HuR overexpression induced stabilization of several AURE-containing genes [12 ]. In contrast, we could not document NO•-induced HuR translocation to the cytoplasm by Western blotting. Nonetheless, with the results reported here, HuR appears to be an important target of NO•-p38 MAPK signaling and a mediator of mRNA stability through its interaction with AUUUA motifs.
In contrast to our findings, another group reported in HeLa cells that HuR bound AUUUA sites in vitro yet did not stabilize reporter mRNA containing an IL-8 AURE sequence [33 ]. These disparate results could be a result of differences in the length of the 3'UTRs from IL-8 that were studied. The IL-8 3'UTRs containing AURE sites investigated here, IL8Wt (416 nt) and IL8B (307 nt), were much longer than the 60-nt AURE sequences used in early experiments [33 ]. The affinity of HuR for these short fragments may have been low, or alternatively, flanking sequences and complex secondary structures may be required for full functionality as well as the recruitment of other proteins such as hnRNP A0. Besides HuR, KSRP and TIAR have also been shown to bind the 3'UTR of IL-8 in breast cancer cells [31 ]. Unlike these previous results, antibodies against KSRP and TIAR failed to super-shift the proteins that complexed with IL8B riboprobe in our REMSA (data not shown). Given that KSRP and TIAR are ubiquitously expressed [37 ], the cause of these disparate results is not clear but may be related to differences in antibodies or the biological context of the experiments. Likewise, antibodies to TTP, AUF1, and BRF1, another important group of AURE-binding proteins, typically associated with mRNA destabilization, did not produce supershifts by REMSA in our experiments.
Another MK2 substrate, hnRNP A0, has been shown to bind to AURE sites in the 3'UTR of TNF- and MIP-2 in response to p38 MAPK activation in mouse macrophages [17
]. Consistent with this work, here, we found that hnRNP A0 also bound to the 3'UTR of IL-8 in THP-1 cells in response to NO•. Moreover, hnRNP A0 as well as HuR bound only to Region B, which contains 5 AUUUA motifs, and not Region A, which lacks discernible AURE sequences. To date, the identity of the NO•-responsive element in Region A or its cognate RNA-binding proteins have not yet been defined. Nonetheless, these results taken together demonstrate that NO• can regulate mRNA stability via AURE-dependent and -independent mechanisms.
Received September 20, 2007; revised December 12, 2007; accepted December 26, 2007.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
production by tristetraprolin Science 281,1001-1005 promoter functions as a nitric oxide response element J. Biol. Chem. 274,33190-33193B J. Biol. Chem. 272,30969-30974 through a p38 MAPK signaling pathway FASEB J. 21,950-961 mRNA stability Mol. Cell. Biol. 21,6461-6469This article has been cited by other articles:
 |
|
 |
|
  W. Hu, F. Li, S. Mahavadi, and K. S. Murthy Upregulation of RGS4 expression by IL-1{beta} in colonic smooth muscle is enhanced by ERK1/2 and p38 MAPK and inhibited by the PI3K/Akt/GSK3{beta} pathway Am J Physiol Cell Physiol, June 1, 2009; 296(6): C1310 - C1320. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Kuwano, A. Rabinovic, S. Srikantan, M. Gorospe, and B. Demple Analysis of Nitric Oxide-Stabilized mRNAs in Human Fibroblasts Reveals HuR-Dependent Heme Oxygenase 1 Upregulation Mol. Cell. Biol., May 15, 2009; 29(10): 2622 - 2635. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
