Extracellular Signal-Regulated Kinase–Dependent Stabilization of Hepatic Low-Density Lipoprotein Receptor mRNA by Herbal Medicine Berberine
http://www.100md.com
动脉硬化血栓血管生物学 2005年第10期
From the VA Palo Alto Health Care System (P.A., Y.Z., J.L.), Palo Alto, Calif; and Institute of Medicinal Biotechnology (J.-D.J.), Chinese Academy of Medical Sciences, Beijing, China.
Correspondence to Jingwen Liu, PhD, VA Palo Alto Health Care System, 3801 Miranda Ave, Palo Alto, CA 94304. E-mail Jingwen.Liu@med.va.gov
Abstract
Objective— Our recent studies identified berberine (BBR) as a novel cholesterol-lowering drug that upregulates low-density lipoprotein (LDL) receptor expression through mRNA stabilization. Here, we investigated mechanisms underlying regulatory effects of BBR on LDL receptor (LDLR) messenger.
Methods and Results— We show that the extracellular signal-regulated kinase (ERK) signaling pathway is used primarily by BBR to attenuate the decay of LDLR mRNA in HepG2 cells. Using different reporter constructs, we demonstrate that BBR affects LDLR mRNA stability entirely through 3' untranslated region (UTR) in an ERK-dependent manner, and this stabilizing effect is more prominent in liver-derived cells than nonhepatic cell lines. In contrast to BBR, the mRNA stabilizing effect of bile acid chenodeoxycholic acid is mediated through the LDLR coding sequence, whereas the 5'UTR, 3'UTR, and the coding sequence of LDLR mRNA are all implicated in the action of phorbol 12-myristate 13-acetate. By performing UV cross-linking and SDS-PAGE, we identify 2 cytoplasmic proteins of 52 and 42 kDa that specifically bind to the LDLR 3'UTR in BBR-inducible and ERK-dependent manners.
Conclusions— These new findings demonstrate that the BBR-induced stabilization of LDLR mRNA is mediated by the ERK signaling pathway through interactions of cis-regulatory sequences of 3'UTR and mRNA binding proteins that are downstream effectors of this signaling cascade.
Our recent studies identified berberine (BBR) as a novel cholesterol-lowering drug that upregulates low-density lipoprotein (LDL) receptor expression through mRNA stabilization. Here, we investigated mechanisms underlying regulatory effects of BBR on LDL receptor (LDLR) messenger. These new findings demonstrate that the BBR-induced stabilization of LDLR mRNA is mediated by the ERK signaling pathway through interactions of cis-regulatory sequences of 3'UTR and mRNA binding proteins that are downstream effectors of this signaling cascade.
Key Words: LDL receptor ? berberine ? mRNA stability ? extracellular signal-regulated kinase ? 3' untranslated region
Introduction
The expression level of liver low-density lipoprotein (LDL) receptor is one of most important regulators of human plasma LDL cholesterol (LDL-c).1 Increased hepatic LDL receptor (LDLR) expression results in an improved clearance of LDL-c from circulation, hence directly reducing the risk of coronary heart disease.2–4 Currently, 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins) are the most widely used cholesterol-lowering drugs that effectively decrease the plasma concentration of LDL-c and reduce mortality and morbidity from coronary artery disease.5,6
Our research group is interested in the discovery of new LDLR modulators from natural resources that potentially use regulatory mechanisms different but complementary to statins. Through a rationalized screening of compounds derived from Chinese herbs, we recently identified berberine (BBR), an alkaloid isolated from the herb Huanglian as a novel upregulator of hepatic LDLR expression.7 Through studies in HepG2 and Bel-7402 hepatoma cells, we showed that BBR strongly increases LDLR mRNA expression. This action mode is totally independent of the intracellular concentration of cholesterol and the activation process of sterol regulatory element binding protein. However, BBR causes strong and sustained activation of extracellular signal-regulated kinase (ERK). Blocking ERK activation by U0126, the inhibitor of ERK upstream kinase totally prohibited the elevation of LDLR mRNA level in BBR-treated cells. Additional studies further demonstrate that BBR does not stimulate the LDLR promoter activity; instead, it increases LDLR expression by extending the half-life of LDLR mRNA through sequences present in the proximal section of the LDLR mRNA 3' untranslated region (UTR). In addition to cultured hepatoma cells, BBR was shown to increase liver LDLR in hypercholesterolemic hamsters and to reduce serum LDL-c in hyperlipidemic patients and hamsters. These studies together strongly suggest that BBR is a promising novel hypolipidemic drug, and it elevates hepatic LDLR expression through a unique post-transcriptional mechanism.
Before our studies on BBR, 3 compounds, namely phorbol 12-myristate 13-acetate (PMA), a bile acid chenodeoxycholic acid (CDCA), and a derivative of fibric acid gemfibrozil have been reported to regulate LDLR expression by mechanisms involving post-transcriptional events. In PMA-treated HepG2 cells, the mRNA level of LDLR was rapidly elevated. Increased stability of the mRNA partially accounted for this elevation.8 Analysis of the LDLR mRNA 3'UTR revealed that stabilization of LDLR mRNA in the presence of PMA required sequences in the distal 3'UTR.9 Subsequently, CDCA was shown to increase the mRNA level of LDLR in HepG2 cells through mechanisms that affect the turnover rate of LDLR mRNA.10 However, it is unknown whether 3'UTR of LDLR mRNA is involved in the stabilizing effect of CDCA. The activity of gemfibrozil in regulating LDLR mRNA stability was demonstrated by the extended half-life of the LDLR mRNA in cells treated with 40 μmol/L of gemfibrozil.11
In this study, we further characterized the molecular mechanisms by which BBR regulates LDLR mRNA stability in hepatoma-derived cell lines and nonliver cell types. We also compared effects of BBR on native and heterologous LDLR mRNA expressions with CDCA, PMA, and gemfibrozil to determine whether BBR shares similar regulatory mechanisms with these compounds that possess diversified structures and different properties.
Methods
The Methods section is available online at http://atvb.ahajournals.org.
Results
Requirement of ERK Activation for BBR-Mediated LDLR mRNA Stabilization Through 3'UTR
Our previous studies showed that the activity of BBR in increasing LDLR mRNA expression in HepG2 cells requires ERK activation.7 To investigate whether other signaling pathways are also involved in the action of BBR, HepG2 cells were treated with BBR at a concentration of 10 μg/mL for 8 hours in the absence or presence of inhibitors to different kinases, including p38 kinase inhibitor SB-203580, phosphatidylinositol 3-kinase inhibitor wortmannin, protein kinase C inhibitor calphostin, c-Jun N-terminal kinase inhibitor curcumin, and the mitogen-activated protein kinase kinase-1 (MEK1) inhibitor U0126. Total RNA was isolated, and the level of LDLR mRNA was determined by quantitative real-time RT-PCR assays. BBR induced a 3.1-fold increase in LDLR mRNA level, and this activity was only prohibited when U0126 was present during the treatment, whereas all other kinase inhibitors at their effective or maximal concentrations failed to prevent the raising of LDLR mRNA levels on BBR stimulation (Figure I, available online at http://atvb.ahajournals.org).
Because the ERK signaling pathway has been linked to transcriptional activations of LDLR gene,12–18 it is important to verify that in the case of BBR-mediated regulation, ERK activation is truly required for the mRNA stabilization. To this end, we performed 2 different experiments. First, HepG2 cells were untreated or treated with actinomycin D alone or combined with U0126 for 30 minutes before the addition of BBR. Total RNA was isolated after 2 or 4 hours of BBR treatment. Northern blot analysis showed that inhibition of transcription by actinomycin D markedly reduced the abundance of LDLR mRNA to undetectable levels in cells without BBR. Under the same condition of transcriptional suppression, LDLR mRNA was readily detected in cells that were treated with BBR for 2 hours and became even more abundant in 4-hour BBR-treated cells. However, BBR effects were completely prevented when cells were cotreated with BBR and U0126 (Figure IIA, available online at http://atvb.ahajournals.org). Second, cells were untreated or treated with U0126 in the absence or presence of BBR for 12 hours. Actinomycin D was then added to cells, and total RNA was isolated at indicated time points. The results show that BBR treatment significantly attenuated the rapid decay of LDLR mRNA only in cells that were not treated with U0126. With U0126, the LDLR mRNA turnover rate in BBR-treated cells was very similar to that in control cells without or with the MEK1 inhibitor (Figure IIB).
pLuc-UTR-2 reporter contains the BBR-responsible 905-nt segment of LDLR mRNA 3'UTR at the end of Luc coding sequence (cds).7 To further link the ERK signaling pathway with BBR-induced mRNA stabilization, HepG2 cells were transfected with pLuc-UTR-2. Twenty-four hours after transfection, cells were trypsinized and reseeded into 4 dishes. The next day, cells were treated with DMSO as the solvent control, U0126, BBR, or BBR plus U0126 for 8 hours. The effect of U0126 on BBR-induced stabilization of Luc-UTR-2 chimeric mRNA was determined by measuring Luc mRNA levels using the real-time quantitative RT-PCR assay (Figure IIC). U0126 did not inhibit the basal level of Luc mRNA in cells without BBR treatment, but it totally abolished the BBR-induced elevation of Luc-UTR-2 mRNA expression. These results (Figure IIA through IIC) provide the direct support for a key role of ERK signaling pathway in BBR-mediated LDLR mRNA stabilization through 3'UTR.
Construction of Cytomegalovirus Promoter–Driven LDLR Plasmids Minus or Plus 3'UTR and Their Responsiveness to BBR Stimulation
Human LDLR mRNA consists of a short 5'UTR (93 nt), a cds of 2582 nt, and a long 2.5-kb 3'UTR.19 Using the program MFOLD,20 we found that multiple secondary structures and putative regulatory sequences are present in 3'UTR as well as in cds. There are precedents in literature that some regulatory sequences for mRNA stability exist in cds.21,22 To determine whether 3'UTR exclusively mediates the BBR activity on LDLR mRNA stability, we separately cloned the LDLR cds alone and cds plus entire 3'UTR (cds-UTR) into a cytomegalovirus promoter–driven vector pcDNA3.1 (Figure 1A). Plasmids were transfected into HepG2 cells. One day after transfection, cells were trypsinized and reseeded evenly into 2 or 3 dishes for each plasmid transfection. Before BBR treatment, all dishes were incubated overnight in a medium containing 10 μg/mL cholesterol plus 1 μg/mL 25-hydroxycholesterol to repress the expression of endogenous LDLR mRNA. Total RNA was isolated from transfected cells untreated or treated with BBR for 8 hours and used in quantitative real-time PCR assays to determine LDLR mRNA abundance. The level of LDLR mRNA was increased 150-fold in pcDNA/LDLR-cds–transfected cells compared with control vector–transfected cells, and BBR treatment did not increase the abundance of LDLR-cds transcript. Inclusion of 3'UTR in the construct significantly reduced the amount of exogenously expressed LDLR mRNA by 3.4-fold. This is consistent with previous reports9,23 and confirmed the functionality of 3'UTR in destabilizing the messenger in this experimental system. Importantly, inclusion of 3'UTR adjacent to the cds regenerated a marked response to BBR stimulation in cells transfected with pcDNA/LDLR-cds-UTR. Moreover, the strong induction of LDLR-cds-UTR mRNA expression was not detected in transfected cells that were cotreated with BBR and U0126 (Figure 1A). To further demonstrate the response of 3'UTR to ERK activation, plasmid pcDNA/LDLR-cds-UTR was either cotransfected with the empty vector pcDNA4 or the vector pcDNA4-MEKR4F that encodes a constitutively active form of MEK124,25 into HepG2 cells, and LDLR mRNA levels were determined by real-time PCR assays. Figure 1B shows that pLDLR-cds-UTR responded to the active form of MEK1 with a 2-fold increase in LDLR mRNA levels, resembling the situation of BBR induction.
Figure 1. Construction and examination of reporters containing 3'UTR, cds, and 5'UTR of LDLR mRNA. A, Diagram of LDLR reporter constructs and results of transfection. HepG2 cells were transfected with pcDNA, pcDNA/LDLR-cds, or pcDNA/LDLR-cds-UTR. The next day, transfected cells were split, reseeded, and incubated in medium supplemented with sterols overnight before BBR treatment of 8 hours. The relative amount of LDLR mRNA was measured by real-time RT-PCR using the LDLR-specific fluorgenic probe. After correction with GAPDH, the amount of LDLR mRNA in pcDNA-transfected cells without BBR treatment was defined as 1, and the amount of LDLR mRNA in other samples was plotted relative to that value. The data shown represent 3 separate experiments with consistent results. B, The plasmid pcDNA/LDLR-cds-UTR was cotransfected with the control vector or the pcDNA4-MEKR4F into HepG2 cells. The relative amount of LDLR mRNA was measured by real-time RT-PCR using the LDLR-specific fluorgenic probe as described in A. Data shown represent 3 separate experiments with consistent results. C, Cells were transfected with pLuc-5'UTR, and responses of transfected cells to BBR treatment were determined by Northern blot. Fusion transcripts were detected using a 32P-labeled Luc cDNA probe. The membrane was then sequentially probed with LDLR and GAPDH probes. The figure shown represents resultsfrom 3 separate experiments.
There are examples for regulatory roles of 5'UTR in mRNA stability.26,27 To absolutely ascertain the exclusive role of 3'UTR in the action of BBR, we also cloned the 93-nt short sequence of 5'UTR of LDLR mRNA into the pLuc vector, and the response of pLuc-5'UTR to BBR treatment was examined in HepG2 cells by Northern blot analysis using a 32P-labeled probe specific to the coding region of Luc. Inclusion of the 5'UTR sequence to Luc coding sequence did not alter the Luc mRNA stability (data not shown). Treatment of transfected cells with BBR at 2 different doses for 8 hours failed to increase levels of Luc-5'UTR chimeric transcript but clearly elevated the endogenous LDLR mRNA expression (Figure 1C), thereby demonstrating a lack of response of 5'UTR to BBR stimulation. Collectively, based on these results, we conclude that the stabilizing activity of BBR on LDLR mRNA is predominantly mediated through 3'UTR in an ERK-dependent manner.
Liver-Specific Regulation of LDLR Expression by BBR
It has been recognized that the regulation of LDLR expression in liver is different from other tissues.1 However, mechanisms controlling the hepatic-specific regulation of LDLR are largely unknown. We sought to determine whether the upregulatory activity of BBR in LDLR expression is specific to hepatocytes. To test this, effects of BBR were examined in HepG2, Bel-7402, and Hepa-1c1c7, along with 3 nonhepatic cell lines, Chinese hamster ovary (CHO), human embryonic kidney 293 (HEK293), and human primary fibroblasts, with the quantitative real-time RT-PCR method using LDLR and GAPDH probes specific to each species (Figure 2A). BBR significantly increased LDLR mRNA expression in HepG2, Bel-7402, and Hepa-1c1c7 cells. In contrast, expression levels of LDLR mRNA were not elevated in CHO, HEK293, or fibroblasts on BBR stimulation. We further measured the 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine (DiI)–LDL uptake activities of LDLR in HepG2, Hepa-1c, and CHO cells untreated or treated with BBR for 18 hours at indicated doses. After treatment, DiI-LDL at a dose of 6 μg/mL was added to medium, and cells were trypsinized 4 hours later. Uptakes of DiI-LDL were measured by FACScan with 2x104 cells per sample. BBR induced dose-dependent increases in DiI-LDL uptake by 3.7-fold and 2.5-fold at the maximum dose in HepG2 and Hepa-1c, whereas no significant increase of LDLR uptake activity was observed in CHO cells (Figure 2B).
Figure 2. Effects of BBR on LDLR expression in hepatic and nonhepatic cell lines. A, Quantitative real-time RT-PCR analysis of LDLR mRNA levels in different cell lines without or with BBR treatment. HPF indicates human primary fibroblasts. B, Uptake of DiI-LDL was measured by FACScan with 2x104 cells per sample. The mean fluorescence value (MFV) of untreated cells is expressed as 100%. The data shown are representative of 3 separate assays. C, Northern blot analysis of Luc-UTR-1 fusion transcript in cells transfected with pLuc-UTR-1 untreated or treated with BBR for 8 hours.
To determine whether LDLR mRNA 3'UTR contributes to the liver-specific regulation of LDLR by BBR, wild type (wt) (pLuc) and the chimeric plasmid pLuc-UTR-1, which contains the entire 3'UTR region of LDLR mRNA, were transfected into HepG2, HEK293, and CHO cells. The basal and BBR-regulated expressions of wt and chimeric Luc mRNAs were examined by Northern blot analysis using the 32P-labeled Luc probe. Comparing with HepG2 cells, destabilizing effects of LDLR 3'UTR on Luc mRNA in HEK293 and CHO cells were moderate, resulting in 50% to 60% decreases, whereas in HepG2 cells, inclusion of LDLR 3'UTR to Luc mRNA lowered the Luc mRNA by 90%. Importantly, treating cells with BBR increased the Luc-UTR-1 mRNA level only in HepG2 cells (Figure 2C). We further examined effects of BBR on ERK activation in CHO and HEK293 cells. Whereas PMA increased ERK phosphorylation in HEK293 cells, levels of phosphorylated ERK were not increased by BBR in HEK293 or CHO cells (Figure III, available online at http://atvb.ahajournals.org). Together, results in Figure 2 imply that the lack of ERK activation is attributable to the absence of stabilizing effects of BBR on LDLR mRNA in nonhepatic cells.
Involvement of Different Regulatory Sequences of LDLR Transcript in Actions of BBR, Bile Acid, and PMA to Modulate the mRNA Stability
In addition to BBR, bile acid CDCA, tumor-promoter PMA, and gemfibrozil were reported to increase the half-life of LDLR mRNA. We were interested in knowing whether BBR shares similar regulatory mechanisms with these compounds. In the top panel of Figure 3A, HepG2 cells were treated with BBR (10 μg/mL), CDCA (100 μmol/L), and gemfibrozil (40 μmol/L) for indicated time points, and LDLR mRNA expressions were determined by Northern blot. BBR elevated LDLR mRNA levels with kinetics similar to CDCA and gemfibrozil. A magnitude of 2- to 2.5-fold increase was detected by the exposure time of 24 hours in cells treated with each compound. PMA induced a rapid increase in LDLR mRNA expression with an 8-fold induction by 2 hours, and that was declined to 3-fold by 8-hour exposure (Figure 3A, bottom). This different action mode from BBR, CDCA, or gemfibrozil likely reflects the dual effects of PMA on transcriptional activation as well as on the mRNA stabilization of LDLR gene. These data were in good agreement with previously published results in different studies.8,10,11 We next examined the involvement of the ERK signaling pathway in regulations of LDLR expression by these compounds. Western blotting with antiphosphorylated ERK antibody demonstrated that ERK was activated in HepG2 cells treated with each of the compound (Figure 3B). U0126 treatment reduced LDLR mRNA levels nearly to or below the baseline value of control in cells treated with BBR, CDCA, or gemfibrozil but partially reduced the mRNA elevation in PMA-treated cells (Figure 3C). These data clearly indicate that the ERK signaling pathway is used exclusively or partially in regulating LDLR mRNA expression by these compounds.
Figure 3. Comparisons of effects of BBR with CDCA, gemfibrozil (GEM), and PMA on expressions of native and heterologous LDLR transcripts. A, Northern blot (top) and quantitative real-time RT-PCR (bottom) analyses of LDLR mRNA levels in HepG2 cells treated for indicated time points with BBR (10 μg/mL), GEM (40 μmol/L), CDCA (100 μmol/L), and PMA (50 nmol/L). B, Western blot analysis of phosphorylated ERK. C, Quantitative real-time RT-PCR assay of LDLR mRNA levels in HepG2 cells treated with individual compounds for 8 hours in the absence and presence of U0126. D, Northern blot analyses of Luc-UTR-1 and Luc-5'UTR fusion transcripts in HepG2 cells transfected with pLuc-UTR-1 or pLuc-5'UTR plasmids. Transfected cells were treated for 8 hours with CDCA (100 μmol/L) or PMA (25 nmol/L). E, Quantitative real-time RT-PCR assay of LDLR mRNA levels in HepG2 cells transfected with pcDNA/LDLR-cds or pcDNA/LDLR-cds-UTR. After transfection, cells were treated for 8 hours with indicated compounds. The amount of LDLR mRNA in untreated cells was defined as 1, and the amount of LDLR mRNA in treated cells was plotted relative to that value. The data shown represent 3 separate experiments with consistent results.
PMA was demonstrated to stabilize the chimeric mRNA through the distal region of LDLR 3'UTR.9 However, it is unknown whether the stabilizing activity of CDCA is also mediated through 3'UTR. HepG2 cells were transfected with pLuc-UTR-1 and pLuc-5'UTR vectors. Twenty-four hours after transfection, cells were split and reseeded equally into 3 dishes for control, CDCA, or PMA treatment of 8 hours. Expressions of Luc-UTR chimeric transcripts were examined by Northern blot analysis using the 32P-labeled Luc cDNA probe (Figure 3D). Surprisingly, expression levels of Luc-UTR-1 and Luc-5'UTR mRNAs were not increased at all by CDCA but were both increased by PMA treatment. Reprobing the membrane with a 32P-labeled LDLR probe clearly showed that CDCA and PMA induced a strong elevation of the endogenous LDLR mRNA up to 3-fold. To explore the possible role of LDLR cds in the action of CDCA, HepG2 cells were transfected with pcDNA/LDLR-cds and pcDNA/LDLR-cds-UTR. Effects of CDCA, PMA, and BBR on stabilities of LDLR-cds or LDLR-cds-UTR mRNAs in transfected cells were assessed by measuring LDLR mRNA levels using the quantitative real-time RT-PCR method. Figure 3E showed that BBR had no effect on LDLR-cds but increased the level of LDLR-cds-UTR by 2-fold. In contrast, CDCA increased the level of LDLR-cds and LDLR-cds-UTR by 3.9-fold and 2.3-fold, respectively, indicating that CDCA increased the stability of LDLR mRNA through the coding sequence. PMA caused a 2-fold increase in LDLR-cds and 3-fold increase in LDLR-cds-UTR, suggesting that coding sequence and 3'UTR contain PMA-responsive stabilizing elements. These results together clearly demonstrate that stabilizations of LDLR mRNA by BBR, CDCA, and PMA are mediated through different regulatory sequences located either in the 3'UTR, the cds, or the 5'UTR of the LDLR transcript.
Induction of Protein Bindings to the LDLR 3'UTR by BBR Treatment
Our previous studies using pLuc-UTR-2 deletion vectors have suggested that the proximal section of LDLR 3'UTR (nt 3038 to 3582) containing the UCAU motif and Au-rich elements is responsive to BBR stimulation.7 In this study, we performed UV cross-linking experiments to characterize the BBR-regulated mRNA binding proteins that interact with this region of UTR. In UV cross-linking, a covalent bond is formed by UV irradiation between the 32P-labeled RNA and the closely interacting proteins.28 After UV irradiation, the RNA–protein complexes are digested by RNases. This produces proteins with an associated short oligo ribonucleotide that corresponds to the protected RNA fragment. A total of 50 μg of cytoplasmic extracts prepared from HepG2 cells untreated or treated with BBR in the absence or presence of U0126 was incubated with a 545-nt in vitro transcribed, 32P-labeled LDLR UTR riboprobe. After the binding reaction, samples were subjected to UV irradiation and treated with RNases A and T1 to remove the exogenous unprotected RNA before electrophoresis on SDS-PAGE. As shown in Figure 4, 2 distinct proteins with molecular masses of 52 and 42 kDa of untreated cytoplasmic extract cross-linked to the LDLR 3'UTR (lane 2). The intensities of these 2 bands were significantly increased by BBR treatment (lane 1); cotreatment of BBR with U0126 substantially reduced the binding of these proteins to the LDLR UTR probe (lane 3). No labeled bands were detected when UV light irradiation was omitted (lanes 5 to 7), thereby indicating the specific formation of the ribonucleoprotein (RNP) complexes.
Figure 4. Identification of cytoplasmic proteins interacting with the BBR-responsive LDLR 3'UTR. HepG2 cytoplasmic extracts from untreated (lanes 2 and 6), BBR 12 hour–treated without U0126 (lanes 1 and 5), and with U0126 (lanes 3 and 7) were incubated with 32P-labeled LDLR RNA probe (nt 3038 to 3582) for 30 minutes at 30°C. One set of sample was exposed to UV for 15 minutes on ice (lanes 1 to 3), and another set of samples did not receive UV exposure (lanes 5 to 7). After digestion with RNases A and T1, followed by SDS-PAGE, analysis was performed using a PhosphorImager. 14C-labeled molecular mass markers are shown in lane 4. Arrowheads indicate the labeled proteins, and the asterisk indicates undigested free probe. The figure shown is a representative of 3 separate assays with consistent results.
Discussion
Messenger RNA stability plays a major role in gene expression in mammalian cells. It is an important mechanism used by cells to achieve a rapid change in gene expression in response to a changing environment. With regard to the regulation of LDLR expression in liver cells, altering the mRNA stability may provide a means to fine-tune the expression level of LDLR to better control the circulating LDL-c. In this study, by conducting investigations on BBR-induced mRNA stabilization, we obtained a number of new findings that significantly increased our current understanding of how the turnover rate of hepatic LDLR mRNA is regulated with regard to the involving signaling pathway and the downstream events at the mRNA sequence level.
Comparing the effect of MEK1 inhibitor U0126 with other inhibitors directed to several critical kinases, we show that ERK activation is solely required for BBR to increase LDLR mRNA expression in liver-derived cells; this finding rules out the possibility of cross-talk between ERK and other signaling pathways. The requirement of ERK activation is completely for attenuating the rapid turnover of the LDLR transcript because the ability of BBR to increase LDLR mRNA levels under the condition of transcriptional suppression can be totally taken away by blocking the MEK1 activity with U0126. Our previous studies using different deletion constructs indicated that the proximal section of the 3'UTR region (pLuc-UTR-2) mediates the stabilizing effect of BBR.7 Here, by using the compound U0126, we demonstrate that the ERK signaling pathway is integrated by regulatory sequences residing in this region, because blocking this cellular pathway totally abrogated the increased stability of the heterologous Luc-UTR-2 mRNA.
It is recognized that the regulation of LDLR transcription in liver cells is different from that in other cell types.1 It is of importance to know whether the mechanisms controlling the messenger stability in liver cells are uniquely different as well. By transfection of the wt pLuc and the chimeric pLuc-UTR-1 vectors into HepG2, CHO, and HEK293 cells, we made an interesting observation that adjunction of LDLR 3'UTR to Luc cds produced differential effects on the mRNA stability in different cell types. In HepG2 cells, the Luc mRNA stability is greatly reduced by 3'UTR, whereas in CHO and HEK293 cells, the presence of 3'UTR only moderately reduced Luc mRNA levels. By examination of the actions of BBR in different cell lines, we found that BBR has much stronger activity in liver-derived cell lines than in other cell types. This could be explained by the differential responses of these cells to BBR stimulation with regard to activations of the ERK signaling pathway. We found that unlike the strong and immediate induction of ERK phosphorylation in HepG2 cells, BBR treatment in CHO and 293 cells did not activate ERK. Importantly, the chimeric Luc-UTR-1 transcript is also not stabilized in these cells on BBR treatment. These data provided the first example to demonstrate that the stability of LDLR mRNA controlled by the 3'UTR in liver-derived cells is regulated differently from that in nonhepatic cell lines.
A previous study has shown that ERK activation is required for CDCA to stabilize the LDLR mRNA,10 resembling the situation of BBR. Interestingly, by using different reporter constructs, we found that CDCA had no stabilizing effect on 3'UTR but greatly increased the level of LDLR mRNA that only has the cds. On the other hand, PMA shows stimulating activities on the reporter constructs of 5'UTR, 3'UTR, as well as cds, likely reflecting its ability in inducing multisignaling pathways in cells.18,29,30 Additional studies are needed to fine-map these regions of LDLR mRNA to identify the regulatory elements that respond to BBR, CDCA, or PMA treatment. Nevertheless, these data provide the first evidence to demonstrate that the ERK signaling cascade can be integrated at various cis-regulatory sequences residing in different regions of the LDLR mRNA to modulate its stability in an inducer-dependent manner.
The stability of a particular mRNA in most cases is regulated by interactions of mRNA binding proteins and cis-regulatory sequences within the 3'UTR in responding to extracellular signals. To determine which mRNA binding proteins are involved in transmitting the BBR-activated ERK signaling to alter the LDLR mRNA stability, we used the method of UV cross-linking with connection to SDS-PAGE to detect the RNP complexes that are formed with the BBR-responsive LDLR 3'UTR region. We found that 2 proteins of 52 and 42 kDa of HepG2 cytoplasmic extract bind to this UTR probe in BBR-inducible and ERK-dependent manners. Thus, these 2 proteins are likely the candidates for mediating the effect of BBR in LDLR mRNA stabilization and are downstream effectors of the ERK signaling pathway. Further studies to reveal the identities of the proteins and their precise binding sequences of LDLR 3'UTR are currently undertaken in our laboratory.
In summary, we demonstrated that the BBR-induced stabilization of LDLR mRNA is mediated by the ERK signaling pathway through interactions of cis-regulatory sequences of 3'UTR and mRNA binding proteins that are downstream effectors of this signaling cascade. Our studies also reveal that various regions of the LDLR mRNA are involved in the stabilization process to respond to different environmental stimuli. These findings provide insight for developing new therapeutic interventions that lower plasma LDL-c by increases of liver LDLR expression through mRNA stabilization.
Acknowledgments
This study was supported by the Department of Veterans of Affairs (Office of Research and Development, Medical Research Service).
References
Spady DK. Hepatic clearance of plasma low density lipoproteins. Semin Liver Dis. 1996; 12: 373–385.
Grundy SM. Statin trials and goals of cholesterol-lowering therapy. Circulation. 1998; 97: 1436–1439.
Ansell BJ, Watson KE, Fogelman AM. An evidence-based assessment of the NCEP adult treatment panel II guidelines. National cholesterol education program. J Am Med Assoc. 1999; 282: 2051–2057.
Bays H, Stein EA. Pharmacotherapy for dyslipidemia-current therapies and future agents. Expert Opin Pharmacother. 2003; 4: 1901–1938.
Moghadasian MH. A safety look at currently available statins. Expert Opin Drug Saf. 2002; 1: 269–274.
Klotz U. Pharmacological comparison of the statins. Arzneimittelforschung. 2003; 53: 605–611.
Kong W, Wei J, Abidi P, Lin M, Inaba S, Li C, Wang Y, Wang Z, Si S, Pan H, Wang S, Wu J, Yue W, Liu J, Jiang J. Berberine is a novel cholesterol-lowering drug working through a unique mechanism distinct from statins. Nat Med. 2004; 10: 1344–1352.
Wilson GM, Roberts EA, Deeley RG. Modulation of LDL receptor mRNA stability by phorbol esters in human liver cell culture models. J Lipid Res. 1997; 38: 437–446.
Wilson GM, Vasa MZ, Deeley RG. Stabilization and cytoskeletal-association of LDL receptor mRNA are mediated by distinct domains in its 3' untranslated region. J Lipid Res. 1998; 39: 1025–1032.
Nakahara M, Fujii H, Maloney PR, Shimizu M, Sato R. Bile acids enhance low density lipoprotein receptor gene expression via a MAPK cascade-mediated stabilization of mRNA. J Biol Chem. 2002; 277: 37229–37234.
Goto D, Okimoto T, Ono M, Shimotsu H, Abe K, Tsujita Y, Kuwano M. Upregulation of low density lipoprotein receptor by gemfibrozil, a hypolipidemic agent, in human hepatoma cells through stabilization of mRNA transcripts. Arterioscler Thromb Vasc Biol. 1997; 17: 2707–2712.
Li C, Kraemer FB, Ahlborn TE, Liu J. Induction of low density lipoprotein receptor (LDLR) transcription by oncostatin M is mediated by the extracellular signal-regulated kinase signaling pathway and the repeat 3 element of the LDLR promoter. J Biol Chem. 1999; 274: 0001–0002.
Kumar A, Chambers TC, Cloud-Heflin BA, Mehta KD. Phorbol ester-induced low density lipoprotein receptor gene expression in HepG2 cells involves protein kinase C-mediated p42/44 MAP kinase activation. J Lipid Res. 1997; 38: 2240–2248.
Kumar A, Middleton A, Chambers TC, Mehta KD. Differential roles of extracellular signal-regulated kinase-1/2 and p38MAPK in interleukin-1beta- and tumor necrosis factor-alpha-induced low density lipoprotein receptor expression in HepG2 cells. J Biol Chem. 1998; 273: 15742–15748.
Kapoor GS, Atkins B, Mehta KD. Activation of Raf-1/MEK-1/2/p42/p44 MAPK cascade alone is sufficient to uncouple LDL receptor expression from cell growth. Mol Cell Biochem. 2002; 236: 13–22.
Kotzka J, Muller-Wieland D, Roth G, Kremer L, Munck M, Schurmann S, Knebel B, Krone W. Sterol regulatory element binding proteins (SREBP)-1a and SREBP-2 are linked to the MAP-kinase cascade. J Lipid Res. 2000; 41: 99–108.
Roth G, Kotzka J, Kremer L, Lehr S, Lohaus C, Meyer HE, Krone W, Müller-Weiland D. MAP kinases Erk1/2 phosphorylate sterol regulatory element-binding protein (SREBP)-1a at serine 117 in vitro. J Biol Chem. 2000; 275: 33302–33307.
Kapoor GS, Golden C, Atkins B, Metha KD. pp90RSK- and protein kinase C-dependent pathway regulates p42/44MAPK-induced LDL receptor transcription in HepG2 cells. J Lipid Res. 2003; 44: 584–593.
Yamamoto T, Davis CG, Brown MS, Schneider WJ, Casey ML, Goldstein JL, Russell DW. The human LDL receptor: a cysteine-rich protein with multiple Alu sequence in its mRNA. Cell. 1984; 39: 27–38.
Zuker M. Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res. 2003; 31: 3406–3415.
Paste M, Huez G, Kruys V. Deadenylation of interferon-beta mRNA is mediated by both the AU-rich element in the 3'-untranslated region and an instability sequence in the coding region. Eur J Biochem. 2003; 270: 1590–1597.
Bufler P, Gamboni-Robertson F, Azam T, Kim SH, Dinarello CA. Interleukin-1 homologues IL-1F7b and IL-18 contain functional mRNA instability elements within the coding region responsive to lipopolysaccharide. Biochem J. 2004; 381: 503–510.
Knouff C, Malloy S, Wilder J, Altenburg MK, Maeda N. Doubling expression of the low density lipoprotein receptor by truncation of the 3'-untranslated region sequence ameliorates type III hyperlipoproteinemia in mice expressing the human apoE2 isoform. J Biol Chem. 2001; 276: 3856–3862.
Mansour SJ, Matten WT, Hermnn AS, Candia JM, Rong S, Fukasawa K, Vande GF, Ahn NG. Transformation of mammalian cells by constitutively active MAP kinase kinase. Science. 1994; 265: 966–969.
Zhang F, Lin M, Abidi P, Thiel G, Liu J. Specific interaction of Egr1 and c/EBP ? leads to the transcriptional activation of the human low density lipoprotein receptor gene. J Biol Chem. 2003; 278: 44246–44254.
Paste M, Huez G, Kruys V. Deadenylation of interferon-beta mRNA is mediated by both the AU-rich element in the 3'-untranslated region and an instability sequence in the coding region. Mol Genet Genomics. 2003; 269: 340–349.
Davies WL, Vandenberg JI, Sayeed RA, Trezise AE. Post-transcriptional regulation of the cystic fibrosis gene in cardiac development and hypertrophy. Biochem Biophys Res Commun. 2004; 319: 410–418.
Hartley R, Meuth-Metzinger VL, Osborne HB. Screening for sequence-specific RNA-BPs by comprehensive UV cross-linking. BMC Mol Biol. 2002; 3: 1–8.
Mehta KD, Radominska-Pandya A, Kapoor G, Dave B, Atkins B. Critical role of diacylglycerol-and phospholipid-regulated protein kinase C-epsilon in induction of low density lipoprotein receptor transcription in response to depletion of cholesterol. Mol Cell Biol. 2002; 22: 3783–3793.
Mehta KD. Role of mitogen-activated protein kinase and protein kinase C in regulating low density lipoprotein receptor expression. Gene Expr. 2002; 10: 153–164.(Parveen Abidi; Yue Zhou; )
Correspondence to Jingwen Liu, PhD, VA Palo Alto Health Care System, 3801 Miranda Ave, Palo Alto, CA 94304. E-mail Jingwen.Liu@med.va.gov
Abstract
Objective— Our recent studies identified berberine (BBR) as a novel cholesterol-lowering drug that upregulates low-density lipoprotein (LDL) receptor expression through mRNA stabilization. Here, we investigated mechanisms underlying regulatory effects of BBR on LDL receptor (LDLR) messenger.
Methods and Results— We show that the extracellular signal-regulated kinase (ERK) signaling pathway is used primarily by BBR to attenuate the decay of LDLR mRNA in HepG2 cells. Using different reporter constructs, we demonstrate that BBR affects LDLR mRNA stability entirely through 3' untranslated region (UTR) in an ERK-dependent manner, and this stabilizing effect is more prominent in liver-derived cells than nonhepatic cell lines. In contrast to BBR, the mRNA stabilizing effect of bile acid chenodeoxycholic acid is mediated through the LDLR coding sequence, whereas the 5'UTR, 3'UTR, and the coding sequence of LDLR mRNA are all implicated in the action of phorbol 12-myristate 13-acetate. By performing UV cross-linking and SDS-PAGE, we identify 2 cytoplasmic proteins of 52 and 42 kDa that specifically bind to the LDLR 3'UTR in BBR-inducible and ERK-dependent manners.
Conclusions— These new findings demonstrate that the BBR-induced stabilization of LDLR mRNA is mediated by the ERK signaling pathway through interactions of cis-regulatory sequences of 3'UTR and mRNA binding proteins that are downstream effectors of this signaling cascade.
Our recent studies identified berberine (BBR) as a novel cholesterol-lowering drug that upregulates low-density lipoprotein (LDL) receptor expression through mRNA stabilization. Here, we investigated mechanisms underlying regulatory effects of BBR on LDL receptor (LDLR) messenger. These new findings demonstrate that the BBR-induced stabilization of LDLR mRNA is mediated by the ERK signaling pathway through interactions of cis-regulatory sequences of 3'UTR and mRNA binding proteins that are downstream effectors of this signaling cascade.
Key Words: LDL receptor ? berberine ? mRNA stability ? extracellular signal-regulated kinase ? 3' untranslated region
Introduction
The expression level of liver low-density lipoprotein (LDL) receptor is one of most important regulators of human plasma LDL cholesterol (LDL-c).1 Increased hepatic LDL receptor (LDLR) expression results in an improved clearance of LDL-c from circulation, hence directly reducing the risk of coronary heart disease.2–4 Currently, 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins) are the most widely used cholesterol-lowering drugs that effectively decrease the plasma concentration of LDL-c and reduce mortality and morbidity from coronary artery disease.5,6
Our research group is interested in the discovery of new LDLR modulators from natural resources that potentially use regulatory mechanisms different but complementary to statins. Through a rationalized screening of compounds derived from Chinese herbs, we recently identified berberine (BBR), an alkaloid isolated from the herb Huanglian as a novel upregulator of hepatic LDLR expression.7 Through studies in HepG2 and Bel-7402 hepatoma cells, we showed that BBR strongly increases LDLR mRNA expression. This action mode is totally independent of the intracellular concentration of cholesterol and the activation process of sterol regulatory element binding protein. However, BBR causes strong and sustained activation of extracellular signal-regulated kinase (ERK). Blocking ERK activation by U0126, the inhibitor of ERK upstream kinase totally prohibited the elevation of LDLR mRNA level in BBR-treated cells. Additional studies further demonstrate that BBR does not stimulate the LDLR promoter activity; instead, it increases LDLR expression by extending the half-life of LDLR mRNA through sequences present in the proximal section of the LDLR mRNA 3' untranslated region (UTR). In addition to cultured hepatoma cells, BBR was shown to increase liver LDLR in hypercholesterolemic hamsters and to reduce serum LDL-c in hyperlipidemic patients and hamsters. These studies together strongly suggest that BBR is a promising novel hypolipidemic drug, and it elevates hepatic LDLR expression through a unique post-transcriptional mechanism.
Before our studies on BBR, 3 compounds, namely phorbol 12-myristate 13-acetate (PMA), a bile acid chenodeoxycholic acid (CDCA), and a derivative of fibric acid gemfibrozil have been reported to regulate LDLR expression by mechanisms involving post-transcriptional events. In PMA-treated HepG2 cells, the mRNA level of LDLR was rapidly elevated. Increased stability of the mRNA partially accounted for this elevation.8 Analysis of the LDLR mRNA 3'UTR revealed that stabilization of LDLR mRNA in the presence of PMA required sequences in the distal 3'UTR.9 Subsequently, CDCA was shown to increase the mRNA level of LDLR in HepG2 cells through mechanisms that affect the turnover rate of LDLR mRNA.10 However, it is unknown whether 3'UTR of LDLR mRNA is involved in the stabilizing effect of CDCA. The activity of gemfibrozil in regulating LDLR mRNA stability was demonstrated by the extended half-life of the LDLR mRNA in cells treated with 40 μmol/L of gemfibrozil.11
In this study, we further characterized the molecular mechanisms by which BBR regulates LDLR mRNA stability in hepatoma-derived cell lines and nonliver cell types. We also compared effects of BBR on native and heterologous LDLR mRNA expressions with CDCA, PMA, and gemfibrozil to determine whether BBR shares similar regulatory mechanisms with these compounds that possess diversified structures and different properties.
Methods
The Methods section is available online at http://atvb.ahajournals.org.
Results
Requirement of ERK Activation for BBR-Mediated LDLR mRNA Stabilization Through 3'UTR
Our previous studies showed that the activity of BBR in increasing LDLR mRNA expression in HepG2 cells requires ERK activation.7 To investigate whether other signaling pathways are also involved in the action of BBR, HepG2 cells were treated with BBR at a concentration of 10 μg/mL for 8 hours in the absence or presence of inhibitors to different kinases, including p38 kinase inhibitor SB-203580, phosphatidylinositol 3-kinase inhibitor wortmannin, protein kinase C inhibitor calphostin, c-Jun N-terminal kinase inhibitor curcumin, and the mitogen-activated protein kinase kinase-1 (MEK1) inhibitor U0126. Total RNA was isolated, and the level of LDLR mRNA was determined by quantitative real-time RT-PCR assays. BBR induced a 3.1-fold increase in LDLR mRNA level, and this activity was only prohibited when U0126 was present during the treatment, whereas all other kinase inhibitors at their effective or maximal concentrations failed to prevent the raising of LDLR mRNA levels on BBR stimulation (Figure I, available online at http://atvb.ahajournals.org).
Because the ERK signaling pathway has been linked to transcriptional activations of LDLR gene,12–18 it is important to verify that in the case of BBR-mediated regulation, ERK activation is truly required for the mRNA stabilization. To this end, we performed 2 different experiments. First, HepG2 cells were untreated or treated with actinomycin D alone or combined with U0126 for 30 minutes before the addition of BBR. Total RNA was isolated after 2 or 4 hours of BBR treatment. Northern blot analysis showed that inhibition of transcription by actinomycin D markedly reduced the abundance of LDLR mRNA to undetectable levels in cells without BBR. Under the same condition of transcriptional suppression, LDLR mRNA was readily detected in cells that were treated with BBR for 2 hours and became even more abundant in 4-hour BBR-treated cells. However, BBR effects were completely prevented when cells were cotreated with BBR and U0126 (Figure IIA, available online at http://atvb.ahajournals.org). Second, cells were untreated or treated with U0126 in the absence or presence of BBR for 12 hours. Actinomycin D was then added to cells, and total RNA was isolated at indicated time points. The results show that BBR treatment significantly attenuated the rapid decay of LDLR mRNA only in cells that were not treated with U0126. With U0126, the LDLR mRNA turnover rate in BBR-treated cells was very similar to that in control cells without or with the MEK1 inhibitor (Figure IIB).
pLuc-UTR-2 reporter contains the BBR-responsible 905-nt segment of LDLR mRNA 3'UTR at the end of Luc coding sequence (cds).7 To further link the ERK signaling pathway with BBR-induced mRNA stabilization, HepG2 cells were transfected with pLuc-UTR-2. Twenty-four hours after transfection, cells were trypsinized and reseeded into 4 dishes. The next day, cells were treated with DMSO as the solvent control, U0126, BBR, or BBR plus U0126 for 8 hours. The effect of U0126 on BBR-induced stabilization of Luc-UTR-2 chimeric mRNA was determined by measuring Luc mRNA levels using the real-time quantitative RT-PCR assay (Figure IIC). U0126 did not inhibit the basal level of Luc mRNA in cells without BBR treatment, but it totally abolished the BBR-induced elevation of Luc-UTR-2 mRNA expression. These results (Figure IIA through IIC) provide the direct support for a key role of ERK signaling pathway in BBR-mediated LDLR mRNA stabilization through 3'UTR.
Construction of Cytomegalovirus Promoter–Driven LDLR Plasmids Minus or Plus 3'UTR and Their Responsiveness to BBR Stimulation
Human LDLR mRNA consists of a short 5'UTR (93 nt), a cds of 2582 nt, and a long 2.5-kb 3'UTR.19 Using the program MFOLD,20 we found that multiple secondary structures and putative regulatory sequences are present in 3'UTR as well as in cds. There are precedents in literature that some regulatory sequences for mRNA stability exist in cds.21,22 To determine whether 3'UTR exclusively mediates the BBR activity on LDLR mRNA stability, we separately cloned the LDLR cds alone and cds plus entire 3'UTR (cds-UTR) into a cytomegalovirus promoter–driven vector pcDNA3.1 (Figure 1A). Plasmids were transfected into HepG2 cells. One day after transfection, cells were trypsinized and reseeded evenly into 2 or 3 dishes for each plasmid transfection. Before BBR treatment, all dishes were incubated overnight in a medium containing 10 μg/mL cholesterol plus 1 μg/mL 25-hydroxycholesterol to repress the expression of endogenous LDLR mRNA. Total RNA was isolated from transfected cells untreated or treated with BBR for 8 hours and used in quantitative real-time PCR assays to determine LDLR mRNA abundance. The level of LDLR mRNA was increased 150-fold in pcDNA/LDLR-cds–transfected cells compared with control vector–transfected cells, and BBR treatment did not increase the abundance of LDLR-cds transcript. Inclusion of 3'UTR in the construct significantly reduced the amount of exogenously expressed LDLR mRNA by 3.4-fold. This is consistent with previous reports9,23 and confirmed the functionality of 3'UTR in destabilizing the messenger in this experimental system. Importantly, inclusion of 3'UTR adjacent to the cds regenerated a marked response to BBR stimulation in cells transfected with pcDNA/LDLR-cds-UTR. Moreover, the strong induction of LDLR-cds-UTR mRNA expression was not detected in transfected cells that were cotreated with BBR and U0126 (Figure 1A). To further demonstrate the response of 3'UTR to ERK activation, plasmid pcDNA/LDLR-cds-UTR was either cotransfected with the empty vector pcDNA4 or the vector pcDNA4-MEKR4F that encodes a constitutively active form of MEK124,25 into HepG2 cells, and LDLR mRNA levels were determined by real-time PCR assays. Figure 1B shows that pLDLR-cds-UTR responded to the active form of MEK1 with a 2-fold increase in LDLR mRNA levels, resembling the situation of BBR induction.
Figure 1. Construction and examination of reporters containing 3'UTR, cds, and 5'UTR of LDLR mRNA. A, Diagram of LDLR reporter constructs and results of transfection. HepG2 cells were transfected with pcDNA, pcDNA/LDLR-cds, or pcDNA/LDLR-cds-UTR. The next day, transfected cells were split, reseeded, and incubated in medium supplemented with sterols overnight before BBR treatment of 8 hours. The relative amount of LDLR mRNA was measured by real-time RT-PCR using the LDLR-specific fluorgenic probe. After correction with GAPDH, the amount of LDLR mRNA in pcDNA-transfected cells without BBR treatment was defined as 1, and the amount of LDLR mRNA in other samples was plotted relative to that value. The data shown represent 3 separate experiments with consistent results. B, The plasmid pcDNA/LDLR-cds-UTR was cotransfected with the control vector or the pcDNA4-MEKR4F into HepG2 cells. The relative amount of LDLR mRNA was measured by real-time RT-PCR using the LDLR-specific fluorgenic probe as described in A. Data shown represent 3 separate experiments with consistent results. C, Cells were transfected with pLuc-5'UTR, and responses of transfected cells to BBR treatment were determined by Northern blot. Fusion transcripts were detected using a 32P-labeled Luc cDNA probe. The membrane was then sequentially probed with LDLR and GAPDH probes. The figure shown represents resultsfrom 3 separate experiments.
There are examples for regulatory roles of 5'UTR in mRNA stability.26,27 To absolutely ascertain the exclusive role of 3'UTR in the action of BBR, we also cloned the 93-nt short sequence of 5'UTR of LDLR mRNA into the pLuc vector, and the response of pLuc-5'UTR to BBR treatment was examined in HepG2 cells by Northern blot analysis using a 32P-labeled probe specific to the coding region of Luc. Inclusion of the 5'UTR sequence to Luc coding sequence did not alter the Luc mRNA stability (data not shown). Treatment of transfected cells with BBR at 2 different doses for 8 hours failed to increase levels of Luc-5'UTR chimeric transcript but clearly elevated the endogenous LDLR mRNA expression (Figure 1C), thereby demonstrating a lack of response of 5'UTR to BBR stimulation. Collectively, based on these results, we conclude that the stabilizing activity of BBR on LDLR mRNA is predominantly mediated through 3'UTR in an ERK-dependent manner.
Liver-Specific Regulation of LDLR Expression by BBR
It has been recognized that the regulation of LDLR expression in liver is different from other tissues.1 However, mechanisms controlling the hepatic-specific regulation of LDLR are largely unknown. We sought to determine whether the upregulatory activity of BBR in LDLR expression is specific to hepatocytes. To test this, effects of BBR were examined in HepG2, Bel-7402, and Hepa-1c1c7, along with 3 nonhepatic cell lines, Chinese hamster ovary (CHO), human embryonic kidney 293 (HEK293), and human primary fibroblasts, with the quantitative real-time RT-PCR method using LDLR and GAPDH probes specific to each species (Figure 2A). BBR significantly increased LDLR mRNA expression in HepG2, Bel-7402, and Hepa-1c1c7 cells. In contrast, expression levels of LDLR mRNA were not elevated in CHO, HEK293, or fibroblasts on BBR stimulation. We further measured the 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine (DiI)–LDL uptake activities of LDLR in HepG2, Hepa-1c, and CHO cells untreated or treated with BBR for 18 hours at indicated doses. After treatment, DiI-LDL at a dose of 6 μg/mL was added to medium, and cells were trypsinized 4 hours later. Uptakes of DiI-LDL were measured by FACScan with 2x104 cells per sample. BBR induced dose-dependent increases in DiI-LDL uptake by 3.7-fold and 2.5-fold at the maximum dose in HepG2 and Hepa-1c, whereas no significant increase of LDLR uptake activity was observed in CHO cells (Figure 2B).
Figure 2. Effects of BBR on LDLR expression in hepatic and nonhepatic cell lines. A, Quantitative real-time RT-PCR analysis of LDLR mRNA levels in different cell lines without or with BBR treatment. HPF indicates human primary fibroblasts. B, Uptake of DiI-LDL was measured by FACScan with 2x104 cells per sample. The mean fluorescence value (MFV) of untreated cells is expressed as 100%. The data shown are representative of 3 separate assays. C, Northern blot analysis of Luc-UTR-1 fusion transcript in cells transfected with pLuc-UTR-1 untreated or treated with BBR for 8 hours.
To determine whether LDLR mRNA 3'UTR contributes to the liver-specific regulation of LDLR by BBR, wild type (wt) (pLuc) and the chimeric plasmid pLuc-UTR-1, which contains the entire 3'UTR region of LDLR mRNA, were transfected into HepG2, HEK293, and CHO cells. The basal and BBR-regulated expressions of wt and chimeric Luc mRNAs were examined by Northern blot analysis using the 32P-labeled Luc probe. Comparing with HepG2 cells, destabilizing effects of LDLR 3'UTR on Luc mRNA in HEK293 and CHO cells were moderate, resulting in 50% to 60% decreases, whereas in HepG2 cells, inclusion of LDLR 3'UTR to Luc mRNA lowered the Luc mRNA by 90%. Importantly, treating cells with BBR increased the Luc-UTR-1 mRNA level only in HepG2 cells (Figure 2C). We further examined effects of BBR on ERK activation in CHO and HEK293 cells. Whereas PMA increased ERK phosphorylation in HEK293 cells, levels of phosphorylated ERK were not increased by BBR in HEK293 or CHO cells (Figure III, available online at http://atvb.ahajournals.org). Together, results in Figure 2 imply that the lack of ERK activation is attributable to the absence of stabilizing effects of BBR on LDLR mRNA in nonhepatic cells.
Involvement of Different Regulatory Sequences of LDLR Transcript in Actions of BBR, Bile Acid, and PMA to Modulate the mRNA Stability
In addition to BBR, bile acid CDCA, tumor-promoter PMA, and gemfibrozil were reported to increase the half-life of LDLR mRNA. We were interested in knowing whether BBR shares similar regulatory mechanisms with these compounds. In the top panel of Figure 3A, HepG2 cells were treated with BBR (10 μg/mL), CDCA (100 μmol/L), and gemfibrozil (40 μmol/L) for indicated time points, and LDLR mRNA expressions were determined by Northern blot. BBR elevated LDLR mRNA levels with kinetics similar to CDCA and gemfibrozil. A magnitude of 2- to 2.5-fold increase was detected by the exposure time of 24 hours in cells treated with each compound. PMA induced a rapid increase in LDLR mRNA expression with an 8-fold induction by 2 hours, and that was declined to 3-fold by 8-hour exposure (Figure 3A, bottom). This different action mode from BBR, CDCA, or gemfibrozil likely reflects the dual effects of PMA on transcriptional activation as well as on the mRNA stabilization of LDLR gene. These data were in good agreement with previously published results in different studies.8,10,11 We next examined the involvement of the ERK signaling pathway in regulations of LDLR expression by these compounds. Western blotting with antiphosphorylated ERK antibody demonstrated that ERK was activated in HepG2 cells treated with each of the compound (Figure 3B). U0126 treatment reduced LDLR mRNA levels nearly to or below the baseline value of control in cells treated with BBR, CDCA, or gemfibrozil but partially reduced the mRNA elevation in PMA-treated cells (Figure 3C). These data clearly indicate that the ERK signaling pathway is used exclusively or partially in regulating LDLR mRNA expression by these compounds.
Figure 3. Comparisons of effects of BBR with CDCA, gemfibrozil (GEM), and PMA on expressions of native and heterologous LDLR transcripts. A, Northern blot (top) and quantitative real-time RT-PCR (bottom) analyses of LDLR mRNA levels in HepG2 cells treated for indicated time points with BBR (10 μg/mL), GEM (40 μmol/L), CDCA (100 μmol/L), and PMA (50 nmol/L). B, Western blot analysis of phosphorylated ERK. C, Quantitative real-time RT-PCR assay of LDLR mRNA levels in HepG2 cells treated with individual compounds for 8 hours in the absence and presence of U0126. D, Northern blot analyses of Luc-UTR-1 and Luc-5'UTR fusion transcripts in HepG2 cells transfected with pLuc-UTR-1 or pLuc-5'UTR plasmids. Transfected cells were treated for 8 hours with CDCA (100 μmol/L) or PMA (25 nmol/L). E, Quantitative real-time RT-PCR assay of LDLR mRNA levels in HepG2 cells transfected with pcDNA/LDLR-cds or pcDNA/LDLR-cds-UTR. After transfection, cells were treated for 8 hours with indicated compounds. The amount of LDLR mRNA in untreated cells was defined as 1, and the amount of LDLR mRNA in treated cells was plotted relative to that value. The data shown represent 3 separate experiments with consistent results.
PMA was demonstrated to stabilize the chimeric mRNA through the distal region of LDLR 3'UTR.9 However, it is unknown whether the stabilizing activity of CDCA is also mediated through 3'UTR. HepG2 cells were transfected with pLuc-UTR-1 and pLuc-5'UTR vectors. Twenty-four hours after transfection, cells were split and reseeded equally into 3 dishes for control, CDCA, or PMA treatment of 8 hours. Expressions of Luc-UTR chimeric transcripts were examined by Northern blot analysis using the 32P-labeled Luc cDNA probe (Figure 3D). Surprisingly, expression levels of Luc-UTR-1 and Luc-5'UTR mRNAs were not increased at all by CDCA but were both increased by PMA treatment. Reprobing the membrane with a 32P-labeled LDLR probe clearly showed that CDCA and PMA induced a strong elevation of the endogenous LDLR mRNA up to 3-fold. To explore the possible role of LDLR cds in the action of CDCA, HepG2 cells were transfected with pcDNA/LDLR-cds and pcDNA/LDLR-cds-UTR. Effects of CDCA, PMA, and BBR on stabilities of LDLR-cds or LDLR-cds-UTR mRNAs in transfected cells were assessed by measuring LDLR mRNA levels using the quantitative real-time RT-PCR method. Figure 3E showed that BBR had no effect on LDLR-cds but increased the level of LDLR-cds-UTR by 2-fold. In contrast, CDCA increased the level of LDLR-cds and LDLR-cds-UTR by 3.9-fold and 2.3-fold, respectively, indicating that CDCA increased the stability of LDLR mRNA through the coding sequence. PMA caused a 2-fold increase in LDLR-cds and 3-fold increase in LDLR-cds-UTR, suggesting that coding sequence and 3'UTR contain PMA-responsive stabilizing elements. These results together clearly demonstrate that stabilizations of LDLR mRNA by BBR, CDCA, and PMA are mediated through different regulatory sequences located either in the 3'UTR, the cds, or the 5'UTR of the LDLR transcript.
Induction of Protein Bindings to the LDLR 3'UTR by BBR Treatment
Our previous studies using pLuc-UTR-2 deletion vectors have suggested that the proximal section of LDLR 3'UTR (nt 3038 to 3582) containing the UCAU motif and Au-rich elements is responsive to BBR stimulation.7 In this study, we performed UV cross-linking experiments to characterize the BBR-regulated mRNA binding proteins that interact with this region of UTR. In UV cross-linking, a covalent bond is formed by UV irradiation between the 32P-labeled RNA and the closely interacting proteins.28 After UV irradiation, the RNA–protein complexes are digested by RNases. This produces proteins with an associated short oligo ribonucleotide that corresponds to the protected RNA fragment. A total of 50 μg of cytoplasmic extracts prepared from HepG2 cells untreated or treated with BBR in the absence or presence of U0126 was incubated with a 545-nt in vitro transcribed, 32P-labeled LDLR UTR riboprobe. After the binding reaction, samples were subjected to UV irradiation and treated with RNases A and T1 to remove the exogenous unprotected RNA before electrophoresis on SDS-PAGE. As shown in Figure 4, 2 distinct proteins with molecular masses of 52 and 42 kDa of untreated cytoplasmic extract cross-linked to the LDLR 3'UTR (lane 2). The intensities of these 2 bands were significantly increased by BBR treatment (lane 1); cotreatment of BBR with U0126 substantially reduced the binding of these proteins to the LDLR UTR probe (lane 3). No labeled bands were detected when UV light irradiation was omitted (lanes 5 to 7), thereby indicating the specific formation of the ribonucleoprotein (RNP) complexes.
Figure 4. Identification of cytoplasmic proteins interacting with the BBR-responsive LDLR 3'UTR. HepG2 cytoplasmic extracts from untreated (lanes 2 and 6), BBR 12 hour–treated without U0126 (lanes 1 and 5), and with U0126 (lanes 3 and 7) were incubated with 32P-labeled LDLR RNA probe (nt 3038 to 3582) for 30 minutes at 30°C. One set of sample was exposed to UV for 15 minutes on ice (lanes 1 to 3), and another set of samples did not receive UV exposure (lanes 5 to 7). After digestion with RNases A and T1, followed by SDS-PAGE, analysis was performed using a PhosphorImager. 14C-labeled molecular mass markers are shown in lane 4. Arrowheads indicate the labeled proteins, and the asterisk indicates undigested free probe. The figure shown is a representative of 3 separate assays with consistent results.
Discussion
Messenger RNA stability plays a major role in gene expression in mammalian cells. It is an important mechanism used by cells to achieve a rapid change in gene expression in response to a changing environment. With regard to the regulation of LDLR expression in liver cells, altering the mRNA stability may provide a means to fine-tune the expression level of LDLR to better control the circulating LDL-c. In this study, by conducting investigations on BBR-induced mRNA stabilization, we obtained a number of new findings that significantly increased our current understanding of how the turnover rate of hepatic LDLR mRNA is regulated with regard to the involving signaling pathway and the downstream events at the mRNA sequence level.
Comparing the effect of MEK1 inhibitor U0126 with other inhibitors directed to several critical kinases, we show that ERK activation is solely required for BBR to increase LDLR mRNA expression in liver-derived cells; this finding rules out the possibility of cross-talk between ERK and other signaling pathways. The requirement of ERK activation is completely for attenuating the rapid turnover of the LDLR transcript because the ability of BBR to increase LDLR mRNA levels under the condition of transcriptional suppression can be totally taken away by blocking the MEK1 activity with U0126. Our previous studies using different deletion constructs indicated that the proximal section of the 3'UTR region (pLuc-UTR-2) mediates the stabilizing effect of BBR.7 Here, by using the compound U0126, we demonstrate that the ERK signaling pathway is integrated by regulatory sequences residing in this region, because blocking this cellular pathway totally abrogated the increased stability of the heterologous Luc-UTR-2 mRNA.
It is recognized that the regulation of LDLR transcription in liver cells is different from that in other cell types.1 It is of importance to know whether the mechanisms controlling the messenger stability in liver cells are uniquely different as well. By transfection of the wt pLuc and the chimeric pLuc-UTR-1 vectors into HepG2, CHO, and HEK293 cells, we made an interesting observation that adjunction of LDLR 3'UTR to Luc cds produced differential effects on the mRNA stability in different cell types. In HepG2 cells, the Luc mRNA stability is greatly reduced by 3'UTR, whereas in CHO and HEK293 cells, the presence of 3'UTR only moderately reduced Luc mRNA levels. By examination of the actions of BBR in different cell lines, we found that BBR has much stronger activity in liver-derived cell lines than in other cell types. This could be explained by the differential responses of these cells to BBR stimulation with regard to activations of the ERK signaling pathway. We found that unlike the strong and immediate induction of ERK phosphorylation in HepG2 cells, BBR treatment in CHO and 293 cells did not activate ERK. Importantly, the chimeric Luc-UTR-1 transcript is also not stabilized in these cells on BBR treatment. These data provided the first example to demonstrate that the stability of LDLR mRNA controlled by the 3'UTR in liver-derived cells is regulated differently from that in nonhepatic cell lines.
A previous study has shown that ERK activation is required for CDCA to stabilize the LDLR mRNA,10 resembling the situation of BBR. Interestingly, by using different reporter constructs, we found that CDCA had no stabilizing effect on 3'UTR but greatly increased the level of LDLR mRNA that only has the cds. On the other hand, PMA shows stimulating activities on the reporter constructs of 5'UTR, 3'UTR, as well as cds, likely reflecting its ability in inducing multisignaling pathways in cells.18,29,30 Additional studies are needed to fine-map these regions of LDLR mRNA to identify the regulatory elements that respond to BBR, CDCA, or PMA treatment. Nevertheless, these data provide the first evidence to demonstrate that the ERK signaling cascade can be integrated at various cis-regulatory sequences residing in different regions of the LDLR mRNA to modulate its stability in an inducer-dependent manner.
The stability of a particular mRNA in most cases is regulated by interactions of mRNA binding proteins and cis-regulatory sequences within the 3'UTR in responding to extracellular signals. To determine which mRNA binding proteins are involved in transmitting the BBR-activated ERK signaling to alter the LDLR mRNA stability, we used the method of UV cross-linking with connection to SDS-PAGE to detect the RNP complexes that are formed with the BBR-responsive LDLR 3'UTR region. We found that 2 proteins of 52 and 42 kDa of HepG2 cytoplasmic extract bind to this UTR probe in BBR-inducible and ERK-dependent manners. Thus, these 2 proteins are likely the candidates for mediating the effect of BBR in LDLR mRNA stabilization and are downstream effectors of the ERK signaling pathway. Further studies to reveal the identities of the proteins and their precise binding sequences of LDLR 3'UTR are currently undertaken in our laboratory.
In summary, we demonstrated that the BBR-induced stabilization of LDLR mRNA is mediated by the ERK signaling pathway through interactions of cis-regulatory sequences of 3'UTR and mRNA binding proteins that are downstream effectors of this signaling cascade. Our studies also reveal that various regions of the LDLR mRNA are involved in the stabilization process to respond to different environmental stimuli. These findings provide insight for developing new therapeutic interventions that lower plasma LDL-c by increases of liver LDLR expression through mRNA stabilization.
Acknowledgments
This study was supported by the Department of Veterans of Affairs (Office of Research and Development, Medical Research Service).
References
Spady DK. Hepatic clearance of plasma low density lipoproteins. Semin Liver Dis. 1996; 12: 373–385.
Grundy SM. Statin trials and goals of cholesterol-lowering therapy. Circulation. 1998; 97: 1436–1439.
Ansell BJ, Watson KE, Fogelman AM. An evidence-based assessment of the NCEP adult treatment panel II guidelines. National cholesterol education program. J Am Med Assoc. 1999; 282: 2051–2057.
Bays H, Stein EA. Pharmacotherapy for dyslipidemia-current therapies and future agents. Expert Opin Pharmacother. 2003; 4: 1901–1938.
Moghadasian MH. A safety look at currently available statins. Expert Opin Drug Saf. 2002; 1: 269–274.
Klotz U. Pharmacological comparison of the statins. Arzneimittelforschung. 2003; 53: 605–611.
Kong W, Wei J, Abidi P, Lin M, Inaba S, Li C, Wang Y, Wang Z, Si S, Pan H, Wang S, Wu J, Yue W, Liu J, Jiang J. Berberine is a novel cholesterol-lowering drug working through a unique mechanism distinct from statins. Nat Med. 2004; 10: 1344–1352.
Wilson GM, Roberts EA, Deeley RG. Modulation of LDL receptor mRNA stability by phorbol esters in human liver cell culture models. J Lipid Res. 1997; 38: 437–446.
Wilson GM, Vasa MZ, Deeley RG. Stabilization and cytoskeletal-association of LDL receptor mRNA are mediated by distinct domains in its 3' untranslated region. J Lipid Res. 1998; 39: 1025–1032.
Nakahara M, Fujii H, Maloney PR, Shimizu M, Sato R. Bile acids enhance low density lipoprotein receptor gene expression via a MAPK cascade-mediated stabilization of mRNA. J Biol Chem. 2002; 277: 37229–37234.
Goto D, Okimoto T, Ono M, Shimotsu H, Abe K, Tsujita Y, Kuwano M. Upregulation of low density lipoprotein receptor by gemfibrozil, a hypolipidemic agent, in human hepatoma cells through stabilization of mRNA transcripts. Arterioscler Thromb Vasc Biol. 1997; 17: 2707–2712.
Li C, Kraemer FB, Ahlborn TE, Liu J. Induction of low density lipoprotein receptor (LDLR) transcription by oncostatin M is mediated by the extracellular signal-regulated kinase signaling pathway and the repeat 3 element of the LDLR promoter. J Biol Chem. 1999; 274: 0001–0002.
Kumar A, Chambers TC, Cloud-Heflin BA, Mehta KD. Phorbol ester-induced low density lipoprotein receptor gene expression in HepG2 cells involves protein kinase C-mediated p42/44 MAP kinase activation. J Lipid Res. 1997; 38: 2240–2248.
Kumar A, Middleton A, Chambers TC, Mehta KD. Differential roles of extracellular signal-regulated kinase-1/2 and p38MAPK in interleukin-1beta- and tumor necrosis factor-alpha-induced low density lipoprotein receptor expression in HepG2 cells. J Biol Chem. 1998; 273: 15742–15748.
Kapoor GS, Atkins B, Mehta KD. Activation of Raf-1/MEK-1/2/p42/p44 MAPK cascade alone is sufficient to uncouple LDL receptor expression from cell growth. Mol Cell Biochem. 2002; 236: 13–22.
Kotzka J, Muller-Wieland D, Roth G, Kremer L, Munck M, Schurmann S, Knebel B, Krone W. Sterol regulatory element binding proteins (SREBP)-1a and SREBP-2 are linked to the MAP-kinase cascade. J Lipid Res. 2000; 41: 99–108.
Roth G, Kotzka J, Kremer L, Lehr S, Lohaus C, Meyer HE, Krone W, Müller-Weiland D. MAP kinases Erk1/2 phosphorylate sterol regulatory element-binding protein (SREBP)-1a at serine 117 in vitro. J Biol Chem. 2000; 275: 33302–33307.
Kapoor GS, Golden C, Atkins B, Metha KD. pp90RSK- and protein kinase C-dependent pathway regulates p42/44MAPK-induced LDL receptor transcription in HepG2 cells. J Lipid Res. 2003; 44: 584–593.
Yamamoto T, Davis CG, Brown MS, Schneider WJ, Casey ML, Goldstein JL, Russell DW. The human LDL receptor: a cysteine-rich protein with multiple Alu sequence in its mRNA. Cell. 1984; 39: 27–38.
Zuker M. Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res. 2003; 31: 3406–3415.
Paste M, Huez G, Kruys V. Deadenylation of interferon-beta mRNA is mediated by both the AU-rich element in the 3'-untranslated region and an instability sequence in the coding region. Eur J Biochem. 2003; 270: 1590–1597.
Bufler P, Gamboni-Robertson F, Azam T, Kim SH, Dinarello CA. Interleukin-1 homologues IL-1F7b and IL-18 contain functional mRNA instability elements within the coding region responsive to lipopolysaccharide. Biochem J. 2004; 381: 503–510.
Knouff C, Malloy S, Wilder J, Altenburg MK, Maeda N. Doubling expression of the low density lipoprotein receptor by truncation of the 3'-untranslated region sequence ameliorates type III hyperlipoproteinemia in mice expressing the human apoE2 isoform. J Biol Chem. 2001; 276: 3856–3862.
Mansour SJ, Matten WT, Hermnn AS, Candia JM, Rong S, Fukasawa K, Vande GF, Ahn NG. Transformation of mammalian cells by constitutively active MAP kinase kinase. Science. 1994; 265: 966–969.
Zhang F, Lin M, Abidi P, Thiel G, Liu J. Specific interaction of Egr1 and c/EBP ? leads to the transcriptional activation of the human low density lipoprotein receptor gene. J Biol Chem. 2003; 278: 44246–44254.
Paste M, Huez G, Kruys V. Deadenylation of interferon-beta mRNA is mediated by both the AU-rich element in the 3'-untranslated region and an instability sequence in the coding region. Mol Genet Genomics. 2003; 269: 340–349.
Davies WL, Vandenberg JI, Sayeed RA, Trezise AE. Post-transcriptional regulation of the cystic fibrosis gene in cardiac development and hypertrophy. Biochem Biophys Res Commun. 2004; 319: 410–418.
Hartley R, Meuth-Metzinger VL, Osborne HB. Screening for sequence-specific RNA-BPs by comprehensive UV cross-linking. BMC Mol Biol. 2002; 3: 1–8.
Mehta KD, Radominska-Pandya A, Kapoor G, Dave B, Atkins B. Critical role of diacylglycerol-and phospholipid-regulated protein kinase C-epsilon in induction of low density lipoprotein receptor transcription in response to depletion of cholesterol. Mol Cell Biol. 2002; 22: 3783–3793.
Mehta KD. Role of mitogen-activated protein kinase and protein kinase C in regulating low density lipoprotein receptor expression. Gene Expr. 2002; 10: 153–164.(Parveen Abidi; Yue Zhou; )