- Email: [email protected]
Splice variant-specific silencing of angiotensin II type 1a receptor messenger RNA by RNA interference in vascular smooth muscle cells
BBRC Biochemical and Biophysical Research Communications 339 (2006) 499–505 www.elsevier.com/locate/ybbrc
Splice variant-specific silencing of angiotensin II type 1a receptor messenger RNA by RNA interference in vascular smooth muscle cells Ali Hassan *, Hong Ji, Yinghua Zhang, Kathryn Sandberg Center for the Study of Sex Differences in Health, Aging and Disease, Department of Medicine, Georgetown University Medical Center, Washington, DC 20057, USA Received 31 October 2005 Available online 15 November 2005
Abstract In the rat, two distinct angiotensin II type 1a (rAT1a) receptor mRNAs are synthesized from a single rAT1a receptor gene by alternative splicing. These two transcripts are comprised of exons 1, 2, and 3 (E1,2,3) or exons 1 and 3 (E1,3). Since exon 3 contains the entire coding region, both transcripts encode identical rAT1a receptors. Real-time PCR revealed that in rat aortic smooth muscle cells (RASMC), E1,2,3 mRNA accounted for 69.5 ± 0.9% of total rAT1a receptor mRNA. The aim of this study was to use RNA interference (RNAi) to selectively silence the rAT1a receptor splice variants. Forty-eight hour treatment of RASMC with E1,3-targeting siRNA (10 nM; S1E1,3) resulted in a 91.2 ± 0.5% (n = 3, P < 0.001) reduction in E1,3 mRNA and a 19.0 ± 3.0% (n = 4, P < 0.05) reduction in AT1 receptor specific binding compared with cells treated with a non-silencing control siRNA; under these conditions, no effect was observed on levels of E1,2,3 mRNA. Conversely, treatment with E1,2,3-targeting siRNA (S2E2) had no effect on E1,3 mRNA while reducing E1,2,3 mRNA by 73.9 ± 4.2% (n = 3, P < 0.001), and AT1 receptor binding by 39.4 ± 5.4% (n = 4, P < 0.001) compared with control. These data show that the majority of functional AT1 receptor expression in RASMC derives from the E1,2,3 splice variant. These data also demonstrate that rAT1a receptor mRNA can be silenced in a splice-variant specific manner using siRNA in RASMC, thus providing an excellent model system for investigating the role of alternative splicing in the regulation of rAT1a receptor expression. 2005 Elsevier Inc. All rights reserved. Keywords: siRNA; Gene silencing; Angiotensin type 1a receptor; RNAi; Alternative splicing; Vascular smooth muscle
Blood pressure and electrolyte balance in mammals are tightly regulated by the renin–angiotensin system (RAS) [1]. Alterations in the RAS are implicated in human hypertensive disorders, congestive heart failure, and with associated cardiac and renal damage [1–3]. The active hormone, angiotensin II (Ang II), exerts its multiple physiological effects, including vasoconstriction, aldosterone production, and catecholamine release, by binding to high affinity receptors located in key target tissues such as the vasculature, adrenal cortex, kidney, heart, and brain [4]. There are two major classes of angiotensin receptors, AT1 and AT2, which exhibit distinct tissue distribution patterns, sig-
*
Corresponding author. Fax: +1 202 687 7278 . E-mail address: [email protected] (A. Hassan).
0006-291X/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2005.11.043
nal transduction pathways, and mechanisms of receptor regulation. Both receptor subtypes are members of the G protein-coupled receptor superfamily [4]. While relatively little is known regarding the physiological role of AT2 receptors, a large body of evidence demonstrates that the classic actions of Ang II on blood pressure and fluid homeostasis are mediated via AT1 receptors [5,6]. Altered regulation of AT1 receptor expression and signaling has been linked to hypertension and the subsequent development of cardiovascular and renal disease [3,7,8]. Thus, the regulation of AT1 receptors is of marked physiological and pathophysiological interest. We and others have been investigating the role of alternative splicing in AT1 receptor expression [8–12]. Genomic analysis suggests that all angiotensin receptor mRNAs are alternatively spliced [4,12]. Interestingly, in all mammalian
500
A. Hassan et al. / Biochemical and Biophysical Research Communications 339 (2006) 499–505
AT1 receptors thus far cloned, splicing does not occur in the open reading frame that encodes the AT1 receptor; instead, alternative splicing is confined to the 5 0 leader sequence (5 0 LS). In the rat, there are two subtypes of AT1 receptors: rAT1a and rAT1b. The rAT1b receptor is enriched in the pituitary and adrenal, whereas the rAT1a receptor is more widely expressed in tissues including vascular smooth muscle, heart, lung, and brain [13–16]. There are 3 exons in the rAT1a receptor gene. Removal of exon 2 by alternative splicing results in the production of two distinct mRNA transcripts, exons 1 and 3 (E1,3), and exons 1, 2, and 3 (E1,2,3). Exon 3 contains the entire open reading frame for the rAT1a receptor and the 3 0 UTR [17,18]. Thus, the two RNA transcripts differ only in the length of their 5 0 LS and the receptor protein produced from both is identical (Fig. 1A). We have shown that alternative splicing of the rAT1a receptor occurs in a tissue-specific manner and contributes to tissue-specific differences in AT1 receptor expression [19]. The aim of this study was to determine the role of the two rAT1a receptor mRNA splice variants in functional AT1 receptor expression in cells endogenously expressing A
AT1 receptors. We chose rat aortic smooth muscle cells (RASMC) because they endogenously express rAT1a receptors in high abundance and are a key target for Ang II action [20]. Furthermore, RASMC do not express rAT1b receptors at levels that could confound experimental results [14,21,22] In this study, we developed conditions in which splice variant-specific small interfering RNA (siRNA)-mediated RNA interference (RNAi) [23,24] selectively silences either the E1,3 transcript, the E1,2,3 transcript or both rAT1a receptor transcripts and determined the effect of specifically silencing these variants on functional AT1 receptor expression, as assessed by 125I-[Sar1,Ile8]Ang II radioligand binding. Materials and methods Small interfering RNA constructs. All siRNAs were chemically synthesized by Qiagen (Valencia, CA). Each siRNA was dissolved in suspension buffer (100 mM KOAc, 30 mM Hepes–KOH, and 2 mM MgOAc, pH 7.4) at a concentration of 20 lM and stored in aliquots at 20 C until use. siRNA transfections. Chinese hamster ovary (CHO) cells stably expressing either E1,3 (CHO-E1,3) or E1,2,3 (CHO-E1,2,3) rAT1a recep-
S2E2
S3E3 AUG
rAT1a E1,2,3 mRNA
5'-
Exon 1
Exon 2
Exon 3 rAT1a receptor open reading frame
5'UTR
S1E1,3
-3'
3'UTR
S3E3 AUG
rAT1a E1,3 mRNA
5'-
Exon 1 5'UTR
B
Exon 3 rAT1a receptor open reading frame
-3'
3'UTR
S1E1,3 (E1,3 mRNA targeting)
5’-UUCCCUGGUCAAGUGGAUUUU-3’ 3’-GUAAGGGACCAGUUCACCUAA-5’ S2E2 (E1,2,3 mRNA targeting)
5’-GAGAUUGGUAUAAAAUGGCUU-3’ 3’-CUCUCUAACCAUAUUUUACCG-5’ S3E3 (E1,3 and E1,2,3 mRNA targeting)
5’-CCUCUACGCCAGUGUGUUCdTdT-3’ 3’-dTdTGGAGAUGCGGUCACACAAG-5’ Control (CHO)
5’-AUCGUCGGUAGAUUGCCUUUU-3’ 3’-GUUAGCAGCCAUCUAACGGAA-5’ Control (RASMC)
5’-UUCUCCGAACGUGUCACGUdTdT-3’ 3’-dTdTAAGAGGCUUGCACAGUGCA-5’ Fig. 1. Design of siRNA duplexes targeting rAT1a receptor splice variants. (A) There are two rAT1a receptor splice variants. The first (E1,2,3) contains all three exons while the second contains only exons 1 and 3 (E1,3). (B) S1E1,3 siRNA was designed to target the junction between exons 1 and 3 and thus was specific for the E1,3 splice variant. S2E2 siRNA was designed to target exon 2 and thus was specific for the E1,3 splice variant. S3E3 siRNA was designed to target exon 3 and thus was directed against both E1,3 and E1,2,3 mRNA. Additionally two control siRNAs (which did not match any known rat or Chinese hamster sequences) were selected for use as controls.
A. Hassan et al. / Biochemical and Biophysical Research Communications 339 (2006) 499–505 tor mRNA were generated and cultured as previously described [11]. For transfection, cells were sub-cultured with trypsin/versene and plated at a density of either 1.25 · 104 cells/cm2 (CHO-E1,3) or 2.5 · 104 cells/cm2 (CHO-E1,2,3) in DulbeccoÕs modified EagleÕs medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 2 mM L-glutamine. Twenty-four hours after plating, the medium was replaced and the cells were transfected with siRNA complexed with the cationic liposomal transfection reagent, lipofectamine 2000 (Invitrogen, Carlsbad CA), according to the manufacturerÕs protocol. Immediately before complexing with lipofectamine, the siRNA was annealed by heating for 1 min at 90 C and then incubated for 1 h at 37 C. The final siRNA and lipofectamine concentrations were 100 nM and 1667 ng/ll, respectively. RASMC (Cell Applications, San Diego, CA) from male Sprague– Dawley rats were cultured in DMEM/F-12 supplemented with 10% FBS and 2 mM L-glutamine. For transfection, cells were sub-cultured with trypsin/versene and plated at a density of 1 · 104 cells/cm2 in phenol redfree DMEM. Forty-eight hours after plating, the medium was replaced and the cells were transfected in DMEM with siRNA complexed with the transfection reagent oligofectamine (Invitrogen). The final siRNA and oligofectamine concentrations were 10 nM and 2 ll/ml, respectively. Real-time PCR. Total RNA was extracted from RASMC using Trizol reagent (Molecular Research Center, Cincinnati, OH). The total RNA was quantitated by the RiboGreen RNA Quantitation system (Molecular Probes, Eugene, OR). First-strand cDNA was prepared from total RNA using the iScript cDNA Synthesis method (Bio-Rad) with random hexamer primers. Quantitation of rAT1a receptor splice variant mRNA and 18S rRNA (for control) was performed by real-time PCR using the ABI Prism 7700 Sequence Detection System (Applied Biosystems Inc., Foster City, CA). The rAT1a receptor splice variant-specific primers (300 nM) sequences were as follows: forward primers: 119F (E1,2,3), 5 0 -CCA CAT TCC CTG AGT TAA CAT ATG A-3 0 and 114F (E1,3), 5 0 -CTC TGC CAC ATT CCC TGG TC-3 0 ; reverse primers: 310R (for both E1,2,3 and E1,3), 5 0 -TCT TTT GAT ACC ATC TTC AGC AGA A-3 0 ; and probe (10 lM): 232T (E1,2,3 and E1,3), 6 FAM-TCG AAT AGT GTC TGA GAC CAA CTC AAC CCA-TAMRA. PCR conditions were optimized for the probe (232T) and both sets of primers (119F and 310R and 114F and 310R) using control cDNAs. PCRs without reverse transcription were included to control for contamination by genomic DNA. The standard curves for 18S rRNA and rAT1a receptor mRNA were made by a series of five-times dilutions (53, 54, 55, 56, 57, and 58) of the cDNA. Levels of rAT1a receptor splice variant mRNA in samples were calculated from the standard curve values. AT1 receptor radioligand binding assay. Whole cell AT1 receptor binding was measured as previously described [25]. Briefly, 48 h after transfection, the cell medium was aspirated and replaced with monoiodinated 125I[Sar1,Ile8]Ang II (2–3 · 105 cpm; Peptide Radioiodination Service, Oxford, MS) in DMEM + 0.1% BSA. After incubation at room temperature for 90 min, unbound ligand was removed by washing each well twice with 1 ml ice-cold phosphate-buffered saline (PBS). Bound ligand was recovered by dissolving the protein in each well with 1 ml 0.01 M NaOH. The quantity of 125 I-[Sar1,Ile8]Ang II present in each sample was determined using a Cobra c-spectrophotometer (Packard Bell, Palo Alto, CA). Protein content in wells was assessed using the DC protein assay (Bio-Rad). Statistical analysis. All data are reported as means ± SEM. When comparisons were made between two different groups, statistical significance was determined using StudentÕs t test. When multiple comparisons were made, statistical significance was determined using one-way ANOVA followed by TukeyÕs post-test. All statistical analysis was performed using the software package Prism 4.0b (GraphPad Software, San Diego, CA).
Results Specificity of rAT1a receptor splice variant-targeting siRNA designs RNAi was induced by transfection of cells with chemically synthesized siRNA. To selectively silence the E1,3
splice variant, an siRNA duplex (S1E1,3) was selected that targeted the junction between exons 1 and 3 (Fig. 1A). Similarly, an siRNA duplex (S2E2) was selected that selectively targeted exon 2. A third siRNA (S3E3) that targeted a sequence in exon 3 was used as a positive control. BLAST searching showed that none of the selected siRNA sequences (Fig. 1B) had any significant matches to any other known mammalian genes. Two siRNA sequences were chosen for use as non-silencing controls in RASMC and CHO cells, respectively. To evaluate the specificity of the siRNA constructs, CHO cells stably expressing either the E1,3 or E1,2,3 splice variant of the rAT1a receptor mRNA were transfected with S1E1,3, S2E2, S3E3 or non-silencing siRNA. Forty-eight hours after transfection, 125I-[Sar1,Ile8]Ang II binding to the cells was measured. Transfection of CHO-E1,3 cells with S1E1,3 induced a 76.6 ± 1.5% (n = 3, P < 0.001, TukeyÕs test) reduction in AT1 receptor binding compared to the control siRNA (Fig. 2A). In contrast, S1E1,3 siRNA treatment had no effect on AT1 receptor binding in CHO-E1,2,3 cells (n = 4, NS). Conversely, transfection with S2E2 caused a significant reduction in binding in CHO-E1,2,3 cells (71.0 ± 1.5%, n = 4, P < 0.001, TukeyÕs), but no effect in CHO-E1,3 cells (n = 4, NS) (Fig. 2B). Treatment with S3E3 markedly reduced AT1 receptor binding in both CHO-E1,3 [93.2 ± 0.5, n = 3, P < 0.001 (TukeyÕs)] and CHO-E1,2,3 cells (88.3 ± 2.7%, n = 4, P < 0.001, TukeyÕs). E1,2,3 mRNA is the predominant rAT1a receptor splice variant in RASMC Measurement of E1,3 and E1,2,3 mRNA levels by quantitative real-time PCR revealed that the major rAT1a receptor splice variant expressed in RASMC is E1,2,3; E1,2,3 mRNA accounted for 69.5 ± 0.9% (n = 3) of the total rAT1a receptor mRNA population, while the E1,3 transcript accounted for the remaining 30.5% of the total rAT1a receptor mRNA. Mock transfection (oligofectamine only) or transfection with the non-silencing control siRNA had no effect on the relative proportions of E1,3 and E1,2,3 mRNA in RASMC; 48 h after mock transfection (i.e., oligofectamine only) or treatment with non-silencing siRNA, the E1,2,3 transcript remained at 70.0 ± 3.3% or 69.7 ± 0.7% of the total rAT1a receptor mRNA population, respectively. Time- and dose-dependency of siRNA-mediated silencing of rAT1a receptor expression In order to determine the optimal siRNA concentration for silencing of rAT1a receptor expression in RASMC, cells were treated with either non-silencing or S3E3 siRNA at concentrations between 0.01 and 30 nM. AT1 receptor binding was measured 48 h after transfection. Transfection with 10 nM S3E3 induced the greatest reduction in AT1 receptor binding compared with control (92.4 ± 0.2%, n = 3, P < 0.001, t test; Fig. 3A). Similarly, the time-course of siRNA-mediated silencing of rAT1a receptor expression
502
A. Hassan et al. / Biochemical and Biophysical Research Communications 339 (2006) 499–505
Fig. 2. RNAi of rAT1a receptor splice variants stably expressed in Chinese hamster ovary (CHO) cells. (A) Treatment with both S1E1,3 and S3E3 siRNAs significantly reduced AT1 receptor binding compared with the control in CHO cells stably expressing the E1,3 rAT1a receptor splice variant. Treatment with S2E2 siRNA did not reduce AT1 receptor binding in these cells. (B) Treatment with S2E2 siRNA and S3E3 siRNA, but not S1E1,3 siRNA, induced a significant reduction in binding compared with the control in CHO cells stably expressing the E1,2,3 splice variant. Data are expressed as means ± SEM. ***P < 0.01.
was determined by transfecting cells with either non-silencing or S3E3 siRNA (both at a concentration of 10 nM) and measuring AT1 receptor binding between 6 h and 12 days after transfection. A significant reduction in binding was observed 24 h after transfection and this significant reduction was maintained for 7 days. Maximal reductions in binding occurred at 48 and 96 h after transfection [91.6 ± 1.9%, (n = 3, P < 0.001, t test) and 92.4 ± 0.6 (n = 3, P < 0.001, t test), respectively; Fig. 3B]. rAT1a receptor splice variant-specific RNAi in RASMC Forty-eight hours after treatment of RASMC with the siRNA duplexes shown in Fig. 1 at a concentration of 10 nM, RNA was extracted from cells and assayed by quantitative real-time RT-PCR. Fig. 4A shows the effect of siRNA treatment on E1,3 mRNA levels. Treatment with S1E1,3 induced a 91.2 ± 0.5% reduction in E1,3 mRNA (n = 3, P < 0.001, TukeyÕs test) compared to treatment with
Fig. 3. Time- and dose-dependency of siRNA-mediated silencing of AT1 receptor expression in RASMC. (A) Treatment with S3E3 siRNA for 48 h (closed symbols) induced a dose-dependent reduction in AT1 receptor binding compared with treatment with control (i.e., non-silencing) siRNA (open symbols) for a similar duration. (B) Treatment with 10 nM S3E3 siRNA for 48 h (closed symbols) induced a significant reduction in AT1 receptor binding compared with treatment with control siRNA (open symbols) after 24 h of treatment. This significant reduction was maintained for 7 days. Data are expressed as means ± SEM. **P < 0.01; ***P < 0.001.
the non-silencing siRNA. Treatment with S2E2 did not significantly alter levels of E1,3 mRNA while treatment with S3E3 resulted in an 89.3 ± 0.6% reduction in E1,3 mRNA (n = 3, P < 0.001, TukeyÕs). In contrast, treatment with S1E1,3 had no effect on E1,2,3 mRNA, but treatment with both S2E2 and S3E3 caused reductions in E1,2,3 mRNA of 73.9 ± 4.2 (n = 3, P < 0.001, TukeyÕs) and 91.6 ± 0.9% (n = 3, P < 0.001, TukeyÕs), respectively (Fig. 4B). While S1E1,3 and S2E2 potently induced inhibition of their respective targets because they silenced only one of the rAT1a receptor transcripts (E1,3 and E1,2,3 mRNA, respectively), their effect on total rAT1a receptor mRNA was less marked. Treatment with S1E1,3 reduced rAT1a receptor mRNA by 26.8 ± 5.8% (n = 3, P < 0.05, TukeyÕs) compared with control, whereas treatment with S2E2 reduced receptor mRNA by 55.0 ± 3.5% (n = 3, P < 0.001, TukeyÕs). Because S3E3 targeted both transcripts, treatment with this siRNA caused a profound reduction in rAT1a receptor mRNA (90.9 ± 0.8%, n = 3, P < 0.001, TukeyÕs; Fig. 4C).
A. Hassan et al. / Biochemical and Biophysical Research Communications 339 (2006) 499–505
binding) compared to cells treated with the non-silencing siRNA. Treatment with S1E1,3 reduced binding by 19.0 ± 3.0% (n = 4, P < 0.05, TukeyÕs), while S2E2 reduced binding by 39.4 ± 5.4% (n = 4, P < 0.001, TukeyÕs). Treatment with S3E3 reduced binding by 87.2 ± 2.0% (n = 4, P < 0.001, TukeyÕs) compared with control (Fig. 5). Discussion
Fig. 4. Splice variant-specific RNAi of rAT1a receptor transcripts in RASMC. Forty-eight hours after siRNA treatment, cells treated with S1E1,3 siRNA showed marked reductions in E1,3 mRNA (A) but not E1,2,3 mRNA (B). Conversely, cells treated with S2E2 siRNA showed marked reductions in E1,2,3 mRNA but not E1,3 mRNA. Treatment with S3E3 siRNA caused significant reductions in levels of both E1,3 and E1,2,3 mRNA. The effect of siRNA treatment on total rAT1a receptor mRNA is shown in (C). Data are expressed as means ± SEM. *P < 0.05; ***P < 0.001.
siRNA-mediated silencing of rAT1a receptor splice variants reduced AT1 receptor binding in RASMC Treatment with all three rAT1a receptor mRNA targeting siRNAs for 48 h at a concentration of 10 nM reduced AT1 receptor binding (as measured by 125I-[Sar1,Ile8]Ang II
The aim of this study was to determine the role of the two rAT1a receptor mRNA splice variants in functional AT1 receptor expression in cells endogenously expressing AT1 receptors. We developed RNAi technology to selectively silence rAT1a receptor splice variants in RASMC and measured the effect of silencing these variants on AT1 receptor binding. The first step was to design siRNA sequences which could selectively mediate the degradation of the E1,3 transcript (i.e., S1E1,3), the E1,2,3 transcript (S2E2) or both the E1,3 and E1,2,3 transcripts (S3E3). Testing of these siRNA designs in CHO cells expressing either the E1,3 or E1,2,3 rAT1a receptor splice variants demonstrated that the chosen siRNA designs induced a potent and highly selective degradation of their intended targets in this cell type (Fig. 2). Next, we investigated whether these siRNA constructs were similarly selective in RASMC, a primary cell line that endogenously expresses both rAT1a receptor splice variants. Analysis of rAT1a receptor mRNA levels following S1E1,3, S2E2, and S3E3 treatment confirmed that the siRNA designs mediated highly specific degradation of their intended targets in cells endogenously expressing both transcripts. It is important to note that the E1,2,3 splice variant is predominant in RASMC, making up 70% of the total rAT1a receptor mRNA pool in these cells, with E1,3 mRNA accounting for the remainder. Thus, although S1E1,3 siRNA treatment induced a marked and significant reduction in E1,3 mRNA, this reduction amounted to only a small reduction in total rAT1a receptor mRNA. Measurement of AT1 receptor binding 48 h after transfection demonstrated that, in addition to their effects on rAT1a receptor mRNA, treatment with each of these siRNA duplexes induced a subsequent significant reduction in AT1 receptor binding compared with the non-silencing control, indicating that not only did siRNA treatment reduce expression of rAT1a receptor transcripts but it also reduced the level of functional AT1 receptors at the cell surface. The present data also demonstrate that, because of the highly sequence-specific nature of the interaction between the siRNA guide and its target sequence, it is possible to use RNAi to selectively reduce the expression of individual mRNA splice variants while leaving other splice variants intact in mammalian cells. RNAi of transcripts containing particular exons has previously been demonstrated in cultured Drosophila cells [26] and MDA231 breast cancer cells [27], respectively. The data described in this study extend these findings by demonstrating that it is possible to selec-
504
A. Hassan et al. / Biochemical and Biophysical Research Communications 339 (2006) 499–505
human AT1 (hAT1) receptor have also revealed that splicing in the 5 0 LS can affect receptor expression [9]. Furthermore, studies of the hAT1 receptor suggest that alterations in splicing contribute to cardiovascular pathophysiology. Analysis of hAT1 receptor transcripts in endomyocardial biopsies found that in failing hearts, the proportion of exon 2-containing transcripts in the total hAT1 receptor mRNA pool was significantly increased in the atria and left ventricle by 22% and 16%, respectively [32]. While alternative splicing may play a critical role in regulation of the RAS, the mechanisms by which alternative splicing of the AT1 receptor is regulated remain unclear. Splice variant-specific silencing of rAT1a receptor transcripts provides a powerful tool with which to investigate these regulatory processes. Fig. 5. Treatment with siRNA targeting rAT1a receptor splice variants reduces AT1 receptor binding in RASMC. Forty-eight hours treatment with siRNA targeting either the E1,3 splice variant (S1E1,3), E1,2,3 splice variant (S2E2), or both splice variants (S3E3) induced significant reductions in binding of 125I-[Sar1,Ile8]Ang II. Data are expressed as means ± SEM. *P < 0.05; ***P < 0.001.
tively silence transcripts that do not contain a particular exon by targeting the junction between two exons. Since alternative splicing makes a highly important contribution to the complexity of the proteome and is regulated developmentally and in response to cellular stimuli [28,29], it will be important to investigate the role of alternative splicing in many physiological processes. Use of splice variant-specific RNAi is likely to be an invaluable tool in these investigations. Finally, these data demonstrate that it is possible to potently reduce expression of the rAT1a receptor using siRNA-mediated RNAi in RASMC. While evidence from in vitro and cell culture systems suggests that any mRNA can be silenced using RNAi techniques [23], and a previous study has demonstrated silencing of the rAT1a receptor in transiently transfected CHO cells [30], it was unclear whether selective silencing could be achieved using siRNA in cells endogenously expressing the receptor since the ability to induce a reduction in protein expression may be limited by the rate of turnover of the specific mRNA and protein—two factors which are likely to be cell type-specific. Furthermore, it was uncertain what length of time silencing of the rAT1a receptor could be maintained in RASMC. Using S3E3 siRNA (which targeted both rAT1a receptor transcripts), we demonstrated that siRNA mediated silencing occurs rapidly, with a marked and significant reduction in AT1 receptor binding observed by 24 h after transfection and that this reduction could be maintained for at least 7 days post-transfection. Silencing occurred at relatively low siRNA concentrations (peak silencing was observed at 10 nM). We have previously shown that alternative splicing of the rAT1 receptor contributes to tissue-specific differences in rAT1 receptor expression at least in part by altering rates of mRNA translation [19,22,31]. Studies of the
Acknowledgments This work was supported by the American Heart Association (Mid-Atlantic Affiliate) Jocelyn Beard Moran Fellowship (0425548U) and a National Kidney Foundation (National Capital Area) Grant-in-Aid to A.H. and by NIH Grant HL57502 to K.S. References [1] P. Corvol, X. Jeunemaitre, A. Charru, Y. Kotelevtsev, F. Soubrier, Role of the renin–angiotensin system in blood pressure regulation and in human hypertension: new insights from molecular genetics, Recent Prog. Horm. Res. 50 (1995) 287–308. [2] J.R. Teerlink, Neurohumoral mechanisms in heart failure: a central role for the renin–angiotensin system, J. Cardiovasc. Pharmacol. 27 (Suppl. 2) (1996) S1–S8. [3] K. Sandberg, H. Ji, Kidney angiotensin receptors and their role in renal pathophysiology, Semin. Nephrol. 20 (2000) 402–416. [4] K. Sandberg, H. Ji, Comparative analysis of amphibian and mammalian angiotensin receptors, Comp. Biochem. Physiol. A Mol. Integr. Physiol. 128 (2001) 53–75. [5] K.K. Griendling, B. Lassegue, R.W. Alexander, Angiotensin receptors and their therapeutic implications, Annu. Rev. Pharmacol. Toxicol. 36 (1996) 281–306. [6] U. Stroth, T. Unger, The renin–angiotensin system and its receptors, J. Cardiovasc. Pharmacol. 33 (Suppl. 1) (1999) S21–S28, discussion S41-23. [7] R.M. Touyz, E.L. Schiffrin, Signal transduction mechanisms mediating the physiological and pathophysiological actions of angiotensin II in vascular smooth muscle cells, Pharmacol. Rev. 52 (2000) 639– 672. [8] G. Nickenig, D.G. Harrison, The AT(1)-type angiotensin receptor in oxidative stress and atherogenesis: Part II: AT(1) receptor regulation, Circulation 105 (2002) 530–536. [9] M.M. Martin, B.M. Willardson, G.F. Burton, C.R. White, J.N. McLaughlin, S.M. Bray, J.W. Ogilvie Jr., T.S. Elton, Human angiotensin II type 1 receptor isoforms encoded by messenger RNA splice variants are functionally distinct, Mol. Endocrinol. 15 (2001) 281–293. [10] N.P. Martin, R.J. Lefkowitz, S.K. Shenoy, Regulation of V2 vasopressin receptor degradation by agonist-promoted ubiquitination, J. Biol. Chem. 278 (2003) 45954–45959. [11] Y. Zhang, H. Ji, M. Elena Fabucci, C. Falconetti, W. Zheng, K. Sandberg, Translational control of the rat angiotensin type 1a receptor by alternative splicing, Gene 341 (2004) 93–100.
A. Hassan et al. / Biochemical and Biophysical Research Communications 339 (2006) 499–505 [12] T.S. Elton, M.M. Martin, Alternative splicing: a novel mechanism to fine-tune the expression and function of the human AT1 receptor, Trends Endocrinol. Metab. 14 (2003) 66–71. [13] S.S. Kakar, K.K. Riel, J.D. Neill, Differential expression of angiotensin II receptor subtype mRNAs (AT-1A and AT-1B) in the brain, Biochem. Biophys. Res. Commun. 185 (1992) 688–692. [14] S.S. Kakar, J.C. Sellers, D.C. Devor, L.C. Musgrove, J.D. Neill, Angiotensin II type-1 receptor subtype cDNAs: differential tissue expression and hormonal regulation, Biochem. Biophys. Res. Commun. 183 (1992) 1090–1096. [15] Y. Kitami, T. Okura, K. Marumoto, R. Wakamiya, K. Hiwada, Differential gene expression and regulation of type-1 angiotensin II receptor subtypes in the rat, Biochem. Biophys. Res. Commun. 188 (1992) 446–452. [16] K. Sandberg, H. Ji, A.J. Clark, H. Shapira, K.J. Catt, Cloning and expression of a novel angiotensin II receptor subtype, J. Biol. Chem. 267 (1992) 9455–9458. [17] K. Langford, K. Frenzel, B.M. Martin, K.E. Bernstein, The genomic organization of the rat AT1 angiotensin receptor, Biochem. Biophys. Res. Commun. 183 (1992) 1025–1032. [18] T.J. Murphy, R.W. Alexander, K.K. Griendling, M.S. Runge, K.E. Bernstein, Isolation of a cDNA encoding the vascular type-1 angiotensin II receptor, Nature 351 (1991) 233–236. [19] Y. Zhang, H. Ji, M. Elena Fabucci, C. Falconetti, W. Zheng, K. Sandberg, Translational control of the rat angiotensin type 1a receptor by alternative splicing, Gene 341 (2004) 93–100. [20] B.C. Berk, Angiotensin II signal transduction in vascular smooth muscle: pathways activated by specific tyrosine kinases, J. Am. Soc. Nephrol. 10 (Suppl. 11) (1999) S62–S68. [21] V. Briand, L. Riva, A.M. Galzin, Characterization of the angiotensin II AT1 receptor subtype involved in DNA synthesis in cultured vascular smooth muscle cells, Br. J. Pharmacol. 112 (1994) 1195–1201. [22] S. Lee, H. Ji, Z. Wu, W. Zheng, A. Hassan, K. Sandberg, Translational regulation of angiotensin Ii type 1 receptors in proliferating vascular smooth muscle cells, Am. J. Physiol. Regul. Integr. Comp. Physiol. (2005).
[23] S.M. Elbashir, J. Harborth, K. Weber, T. Tuschl, Analysis of gene function in somatic mammalian cells using small interfering RNAs, Methods 26 (2002) 199–213. [24] N.J. Caplen, S. Parrish, F. Imani, A. Fire, R.A. Morgan, Specific inhibition of gene expression by small double-stranded RNAs in invertebrate and vertebrate systems, Proc. Natl. Acad. Sci. USA 98 (2001) 9742–9747. [25] H. Ji, W. Zheng, Y. Zhang, K.J. Catt, K. Sandberg, Genetic transfer of a nonpeptide antagonist binding site to a previously unresponsive angiotensin receptor, Proc. Natl. Acad. Sci. USA 92 (1995) 9240–9244. [26] A.M. Celotto, B.R. Graveley, Exon-specific RNAi: a tool for dissecting the functional relevance of alternative splicing, RNA 8 (2002) 718–724. [27] B. Ge, X. Xiong, Q. Jing, J.L. Mosley, A. Filose, D. Bian, S. Huang, J. Han, TAB1beta (transforming growth factor-beta-activated protein kinase 1-binding protein 1beta), a novel splicing variant of TAB1 that interacts with p38alpha but not TAK1, J. Biol. Chem. 278 (2003) 2286–2293. [28] J. Xie, D.L. Black, A CaMK IV responsive RNA element mediates depolarization-induced alternative splicing of ion channels, Nature 410 (2001) 936–939. [29] S.M. Rao, R.D. Howells, cis-acting elements in the 5 0 -untranslated region of rat testis proenkephalin mRNA regulate translation of the precursor protein, J. Biol. Chem. 268 (1993) 22164–22169. [30] J. Vazquez, M.F. Correa de Adjounian, C. Sumners, A. Gonzalez, C. Diez-Freire, M.K. Raizada, Selective silencing of angiotensin receptor subtype 1a (AT1aR) by RNA interference, Hypertension 45 (2005) 115–119. [31] H. Ji, Y. Zhang, W. Zheng, Z. Wu, S. Lee, K. Sandberg, Translational regulation of angiotensin type 1a receptor expression and signaling by upstream AUGs in the 5 0 leader sequence, J. Biol. Chem. 279 (2004) 45322–45328. [32] C. Warnecke, D. Surder, R. Curth, E. Fleck, V. Regitz-Zagrosek, Analysis and functional characterization of alternatively spliced angiotensin II type 1 and 2 receptor transcripts in the human heart, J. Mol. Med. 77 (1999) 718–727.
Splice variant-specific silencing of angiotensin II type 1a receptor messenger RNA by RNA interference in vascular smooth muscle cells Ali Hassan *, Hong Ji, Yinghua Zhang, Kathryn Sandberg Center for the Study of Sex Differences in Health, Aging and Disease, Department of Medicine, Georgetown University Medical Center, Washington, DC 20057, USA Received 31 October 2005 Available online 15 November 2005
Abstract In the rat, two distinct angiotensin II type 1a (rAT1a) receptor mRNAs are synthesized from a single rAT1a receptor gene by alternative splicing. These two transcripts are comprised of exons 1, 2, and 3 (E1,2,3) or exons 1 and 3 (E1,3). Since exon 3 contains the entire coding region, both transcripts encode identical rAT1a receptors. Real-time PCR revealed that in rat aortic smooth muscle cells (RASMC), E1,2,3 mRNA accounted for 69.5 ± 0.9% of total rAT1a receptor mRNA. The aim of this study was to use RNA interference (RNAi) to selectively silence the rAT1a receptor splice variants. Forty-eight hour treatment of RASMC with E1,3-targeting siRNA (10 nM; S1E1,3) resulted in a 91.2 ± 0.5% (n = 3, P < 0.001) reduction in E1,3 mRNA and a 19.0 ± 3.0% (n = 4, P < 0.05) reduction in AT1 receptor specific binding compared with cells treated with a non-silencing control siRNA; under these conditions, no effect was observed on levels of E1,2,3 mRNA. Conversely, treatment with E1,2,3-targeting siRNA (S2E2) had no effect on E1,3 mRNA while reducing E1,2,3 mRNA by 73.9 ± 4.2% (n = 3, P < 0.001), and AT1 receptor binding by 39.4 ± 5.4% (n = 4, P < 0.001) compared with control. These data show that the majority of functional AT1 receptor expression in RASMC derives from the E1,2,3 splice variant. These data also demonstrate that rAT1a receptor mRNA can be silenced in a splice-variant specific manner using siRNA in RASMC, thus providing an excellent model system for investigating the role of alternative splicing in the regulation of rAT1a receptor expression. 2005 Elsevier Inc. All rights reserved. Keywords: siRNA; Gene silencing; Angiotensin type 1a receptor; RNAi; Alternative splicing; Vascular smooth muscle
Blood pressure and electrolyte balance in mammals are tightly regulated by the renin–angiotensin system (RAS) [1]. Alterations in the RAS are implicated in human hypertensive disorders, congestive heart failure, and with associated cardiac and renal damage [1–3]. The active hormone, angiotensin II (Ang II), exerts its multiple physiological effects, including vasoconstriction, aldosterone production, and catecholamine release, by binding to high affinity receptors located in key target tissues such as the vasculature, adrenal cortex, kidney, heart, and brain [4]. There are two major classes of angiotensin receptors, AT1 and AT2, which exhibit distinct tissue distribution patterns, sig-
*
Corresponding author. Fax: +1 202 687 7278 . E-mail address: [email protected] (A. Hassan).
0006-291X/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2005.11.043
nal transduction pathways, and mechanisms of receptor regulation. Both receptor subtypes are members of the G protein-coupled receptor superfamily [4]. While relatively little is known regarding the physiological role of AT2 receptors, a large body of evidence demonstrates that the classic actions of Ang II on blood pressure and fluid homeostasis are mediated via AT1 receptors [5,6]. Altered regulation of AT1 receptor expression and signaling has been linked to hypertension and the subsequent development of cardiovascular and renal disease [3,7,8]. Thus, the regulation of AT1 receptors is of marked physiological and pathophysiological interest. We and others have been investigating the role of alternative splicing in AT1 receptor expression [8–12]. Genomic analysis suggests that all angiotensin receptor mRNAs are alternatively spliced [4,12]. Interestingly, in all mammalian
500
A. Hassan et al. / Biochemical and Biophysical Research Communications 339 (2006) 499–505
AT1 receptors thus far cloned, splicing does not occur in the open reading frame that encodes the AT1 receptor; instead, alternative splicing is confined to the 5 0 leader sequence (5 0 LS). In the rat, there are two subtypes of AT1 receptors: rAT1a and rAT1b. The rAT1b receptor is enriched in the pituitary and adrenal, whereas the rAT1a receptor is more widely expressed in tissues including vascular smooth muscle, heart, lung, and brain [13–16]. There are 3 exons in the rAT1a receptor gene. Removal of exon 2 by alternative splicing results in the production of two distinct mRNA transcripts, exons 1 and 3 (E1,3), and exons 1, 2, and 3 (E1,2,3). Exon 3 contains the entire open reading frame for the rAT1a receptor and the 3 0 UTR [17,18]. Thus, the two RNA transcripts differ only in the length of their 5 0 LS and the receptor protein produced from both is identical (Fig. 1A). We have shown that alternative splicing of the rAT1a receptor occurs in a tissue-specific manner and contributes to tissue-specific differences in AT1 receptor expression [19]. The aim of this study was to determine the role of the two rAT1a receptor mRNA splice variants in functional AT1 receptor expression in cells endogenously expressing A
AT1 receptors. We chose rat aortic smooth muscle cells (RASMC) because they endogenously express rAT1a receptors in high abundance and are a key target for Ang II action [20]. Furthermore, RASMC do not express rAT1b receptors at levels that could confound experimental results [14,21,22] In this study, we developed conditions in which splice variant-specific small interfering RNA (siRNA)-mediated RNA interference (RNAi) [23,24] selectively silences either the E1,3 transcript, the E1,2,3 transcript or both rAT1a receptor transcripts and determined the effect of specifically silencing these variants on functional AT1 receptor expression, as assessed by 125I-[Sar1,Ile8]Ang II radioligand binding. Materials and methods Small interfering RNA constructs. All siRNAs were chemically synthesized by Qiagen (Valencia, CA). Each siRNA was dissolved in suspension buffer (100 mM KOAc, 30 mM Hepes–KOH, and 2 mM MgOAc, pH 7.4) at a concentration of 20 lM and stored in aliquots at 20 C until use. siRNA transfections. Chinese hamster ovary (CHO) cells stably expressing either E1,3 (CHO-E1,3) or E1,2,3 (CHO-E1,2,3) rAT1a recep-
S2E2
S3E3 AUG
rAT1a E1,2,3 mRNA
5'-
Exon 1
Exon 2
Exon 3 rAT1a receptor open reading frame
5'UTR
S1E1,3
-3'
3'UTR
S3E3 AUG
rAT1a E1,3 mRNA
5'-
Exon 1 5'UTR
B
Exon 3 rAT1a receptor open reading frame
-3'
3'UTR
S1E1,3 (E1,3 mRNA targeting)
5’-UUCCCUGGUCAAGUGGAUUUU-3’ 3’-GUAAGGGACCAGUUCACCUAA-5’ S2E2 (E1,2,3 mRNA targeting)
5’-GAGAUUGGUAUAAAAUGGCUU-3’ 3’-CUCUCUAACCAUAUUUUACCG-5’ S3E3 (E1,3 and E1,2,3 mRNA targeting)
5’-CCUCUACGCCAGUGUGUUCdTdT-3’ 3’-dTdTGGAGAUGCGGUCACACAAG-5’ Control (CHO)
5’-AUCGUCGGUAGAUUGCCUUUU-3’ 3’-GUUAGCAGCCAUCUAACGGAA-5’ Control (RASMC)
5’-UUCUCCGAACGUGUCACGUdTdT-3’ 3’-dTdTAAGAGGCUUGCACAGUGCA-5’ Fig. 1. Design of siRNA duplexes targeting rAT1a receptor splice variants. (A) There are two rAT1a receptor splice variants. The first (E1,2,3) contains all three exons while the second contains only exons 1 and 3 (E1,3). (B) S1E1,3 siRNA was designed to target the junction between exons 1 and 3 and thus was specific for the E1,3 splice variant. S2E2 siRNA was designed to target exon 2 and thus was specific for the E1,3 splice variant. S3E3 siRNA was designed to target exon 3 and thus was directed against both E1,3 and E1,2,3 mRNA. Additionally two control siRNAs (which did not match any known rat or Chinese hamster sequences) were selected for use as controls.
A. Hassan et al. / Biochemical and Biophysical Research Communications 339 (2006) 499–505 tor mRNA were generated and cultured as previously described [11]. For transfection, cells were sub-cultured with trypsin/versene and plated at a density of either 1.25 · 104 cells/cm2 (CHO-E1,3) or 2.5 · 104 cells/cm2 (CHO-E1,2,3) in DulbeccoÕs modified EagleÕs medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 2 mM L-glutamine. Twenty-four hours after plating, the medium was replaced and the cells were transfected with siRNA complexed with the cationic liposomal transfection reagent, lipofectamine 2000 (Invitrogen, Carlsbad CA), according to the manufacturerÕs protocol. Immediately before complexing with lipofectamine, the siRNA was annealed by heating for 1 min at 90 C and then incubated for 1 h at 37 C. The final siRNA and lipofectamine concentrations were 100 nM and 1667 ng/ll, respectively. RASMC (Cell Applications, San Diego, CA) from male Sprague– Dawley rats were cultured in DMEM/F-12 supplemented with 10% FBS and 2 mM L-glutamine. For transfection, cells were sub-cultured with trypsin/versene and plated at a density of 1 · 104 cells/cm2 in phenol redfree DMEM. Forty-eight hours after plating, the medium was replaced and the cells were transfected in DMEM with siRNA complexed with the transfection reagent oligofectamine (Invitrogen). The final siRNA and oligofectamine concentrations were 10 nM and 2 ll/ml, respectively. Real-time PCR. Total RNA was extracted from RASMC using Trizol reagent (Molecular Research Center, Cincinnati, OH). The total RNA was quantitated by the RiboGreen RNA Quantitation system (Molecular Probes, Eugene, OR). First-strand cDNA was prepared from total RNA using the iScript cDNA Synthesis method (Bio-Rad) with random hexamer primers. Quantitation of rAT1a receptor splice variant mRNA and 18S rRNA (for control) was performed by real-time PCR using the ABI Prism 7700 Sequence Detection System (Applied Biosystems Inc., Foster City, CA). The rAT1a receptor splice variant-specific primers (300 nM) sequences were as follows: forward primers: 119F (E1,2,3), 5 0 -CCA CAT TCC CTG AGT TAA CAT ATG A-3 0 and 114F (E1,3), 5 0 -CTC TGC CAC ATT CCC TGG TC-3 0 ; reverse primers: 310R (for both E1,2,3 and E1,3), 5 0 -TCT TTT GAT ACC ATC TTC AGC AGA A-3 0 ; and probe (10 lM): 232T (E1,2,3 and E1,3), 6 FAM-TCG AAT AGT GTC TGA GAC CAA CTC AAC CCA-TAMRA. PCR conditions were optimized for the probe (232T) and both sets of primers (119F and 310R and 114F and 310R) using control cDNAs. PCRs without reverse transcription were included to control for contamination by genomic DNA. The standard curves for 18S rRNA and rAT1a receptor mRNA were made by a series of five-times dilutions (53, 54, 55, 56, 57, and 58) of the cDNA. Levels of rAT1a receptor splice variant mRNA in samples were calculated from the standard curve values. AT1 receptor radioligand binding assay. Whole cell AT1 receptor binding was measured as previously described [25]. Briefly, 48 h after transfection, the cell medium was aspirated and replaced with monoiodinated 125I[Sar1,Ile8]Ang II (2–3 · 105 cpm; Peptide Radioiodination Service, Oxford, MS) in DMEM + 0.1% BSA. After incubation at room temperature for 90 min, unbound ligand was removed by washing each well twice with 1 ml ice-cold phosphate-buffered saline (PBS). Bound ligand was recovered by dissolving the protein in each well with 1 ml 0.01 M NaOH. The quantity of 125 I-[Sar1,Ile8]Ang II present in each sample was determined using a Cobra c-spectrophotometer (Packard Bell, Palo Alto, CA). Protein content in wells was assessed using the DC protein assay (Bio-Rad). Statistical analysis. All data are reported as means ± SEM. When comparisons were made between two different groups, statistical significance was determined using StudentÕs t test. When multiple comparisons were made, statistical significance was determined using one-way ANOVA followed by TukeyÕs post-test. All statistical analysis was performed using the software package Prism 4.0b (GraphPad Software, San Diego, CA).
Results Specificity of rAT1a receptor splice variant-targeting siRNA designs RNAi was induced by transfection of cells with chemically synthesized siRNA. To selectively silence the E1,3
splice variant, an siRNA duplex (S1E1,3) was selected that targeted the junction between exons 1 and 3 (Fig. 1A). Similarly, an siRNA duplex (S2E2) was selected that selectively targeted exon 2. A third siRNA (S3E3) that targeted a sequence in exon 3 was used as a positive control. BLAST searching showed that none of the selected siRNA sequences (Fig. 1B) had any significant matches to any other known mammalian genes. Two siRNA sequences were chosen for use as non-silencing controls in RASMC and CHO cells, respectively. To evaluate the specificity of the siRNA constructs, CHO cells stably expressing either the E1,3 or E1,2,3 splice variant of the rAT1a receptor mRNA were transfected with S1E1,3, S2E2, S3E3 or non-silencing siRNA. Forty-eight hours after transfection, 125I-[Sar1,Ile8]Ang II binding to the cells was measured. Transfection of CHO-E1,3 cells with S1E1,3 induced a 76.6 ± 1.5% (n = 3, P < 0.001, TukeyÕs test) reduction in AT1 receptor binding compared to the control siRNA (Fig. 2A). In contrast, S1E1,3 siRNA treatment had no effect on AT1 receptor binding in CHO-E1,2,3 cells (n = 4, NS). Conversely, transfection with S2E2 caused a significant reduction in binding in CHO-E1,2,3 cells (71.0 ± 1.5%, n = 4, P < 0.001, TukeyÕs), but no effect in CHO-E1,3 cells (n = 4, NS) (Fig. 2B). Treatment with S3E3 markedly reduced AT1 receptor binding in both CHO-E1,3 [93.2 ± 0.5, n = 3, P < 0.001 (TukeyÕs)] and CHO-E1,2,3 cells (88.3 ± 2.7%, n = 4, P < 0.001, TukeyÕs). E1,2,3 mRNA is the predominant rAT1a receptor splice variant in RASMC Measurement of E1,3 and E1,2,3 mRNA levels by quantitative real-time PCR revealed that the major rAT1a receptor splice variant expressed in RASMC is E1,2,3; E1,2,3 mRNA accounted for 69.5 ± 0.9% (n = 3) of the total rAT1a receptor mRNA population, while the E1,3 transcript accounted for the remaining 30.5% of the total rAT1a receptor mRNA. Mock transfection (oligofectamine only) or transfection with the non-silencing control siRNA had no effect on the relative proportions of E1,3 and E1,2,3 mRNA in RASMC; 48 h after mock transfection (i.e., oligofectamine only) or treatment with non-silencing siRNA, the E1,2,3 transcript remained at 70.0 ± 3.3% or 69.7 ± 0.7% of the total rAT1a receptor mRNA population, respectively. Time- and dose-dependency of siRNA-mediated silencing of rAT1a receptor expression In order to determine the optimal siRNA concentration for silencing of rAT1a receptor expression in RASMC, cells were treated with either non-silencing or S3E3 siRNA at concentrations between 0.01 and 30 nM. AT1 receptor binding was measured 48 h after transfection. Transfection with 10 nM S3E3 induced the greatest reduction in AT1 receptor binding compared with control (92.4 ± 0.2%, n = 3, P < 0.001, t test; Fig. 3A). Similarly, the time-course of siRNA-mediated silencing of rAT1a receptor expression
502
A. Hassan et al. / Biochemical and Biophysical Research Communications 339 (2006) 499–505
Fig. 2. RNAi of rAT1a receptor splice variants stably expressed in Chinese hamster ovary (CHO) cells. (A) Treatment with both S1E1,3 and S3E3 siRNAs significantly reduced AT1 receptor binding compared with the control in CHO cells stably expressing the E1,3 rAT1a receptor splice variant. Treatment with S2E2 siRNA did not reduce AT1 receptor binding in these cells. (B) Treatment with S2E2 siRNA and S3E3 siRNA, but not S1E1,3 siRNA, induced a significant reduction in binding compared with the control in CHO cells stably expressing the E1,2,3 splice variant. Data are expressed as means ± SEM. ***P < 0.01.
was determined by transfecting cells with either non-silencing or S3E3 siRNA (both at a concentration of 10 nM) and measuring AT1 receptor binding between 6 h and 12 days after transfection. A significant reduction in binding was observed 24 h after transfection and this significant reduction was maintained for 7 days. Maximal reductions in binding occurred at 48 and 96 h after transfection [91.6 ± 1.9%, (n = 3, P < 0.001, t test) and 92.4 ± 0.6 (n = 3, P < 0.001, t test), respectively; Fig. 3B]. rAT1a receptor splice variant-specific RNAi in RASMC Forty-eight hours after treatment of RASMC with the siRNA duplexes shown in Fig. 1 at a concentration of 10 nM, RNA was extracted from cells and assayed by quantitative real-time RT-PCR. Fig. 4A shows the effect of siRNA treatment on E1,3 mRNA levels. Treatment with S1E1,3 induced a 91.2 ± 0.5% reduction in E1,3 mRNA (n = 3, P < 0.001, TukeyÕs test) compared to treatment with
Fig. 3. Time- and dose-dependency of siRNA-mediated silencing of AT1 receptor expression in RASMC. (A) Treatment with S3E3 siRNA for 48 h (closed symbols) induced a dose-dependent reduction in AT1 receptor binding compared with treatment with control (i.e., non-silencing) siRNA (open symbols) for a similar duration. (B) Treatment with 10 nM S3E3 siRNA for 48 h (closed symbols) induced a significant reduction in AT1 receptor binding compared with treatment with control siRNA (open symbols) after 24 h of treatment. This significant reduction was maintained for 7 days. Data are expressed as means ± SEM. **P < 0.01; ***P < 0.001.
the non-silencing siRNA. Treatment with S2E2 did not significantly alter levels of E1,3 mRNA while treatment with S3E3 resulted in an 89.3 ± 0.6% reduction in E1,3 mRNA (n = 3, P < 0.001, TukeyÕs). In contrast, treatment with S1E1,3 had no effect on E1,2,3 mRNA, but treatment with both S2E2 and S3E3 caused reductions in E1,2,3 mRNA of 73.9 ± 4.2 (n = 3, P < 0.001, TukeyÕs) and 91.6 ± 0.9% (n = 3, P < 0.001, TukeyÕs), respectively (Fig. 4B). While S1E1,3 and S2E2 potently induced inhibition of their respective targets because they silenced only one of the rAT1a receptor transcripts (E1,3 and E1,2,3 mRNA, respectively), their effect on total rAT1a receptor mRNA was less marked. Treatment with S1E1,3 reduced rAT1a receptor mRNA by 26.8 ± 5.8% (n = 3, P < 0.05, TukeyÕs) compared with control, whereas treatment with S2E2 reduced receptor mRNA by 55.0 ± 3.5% (n = 3, P < 0.001, TukeyÕs). Because S3E3 targeted both transcripts, treatment with this siRNA caused a profound reduction in rAT1a receptor mRNA (90.9 ± 0.8%, n = 3, P < 0.001, TukeyÕs; Fig. 4C).
A. Hassan et al. / Biochemical and Biophysical Research Communications 339 (2006) 499–505
binding) compared to cells treated with the non-silencing siRNA. Treatment with S1E1,3 reduced binding by 19.0 ± 3.0% (n = 4, P < 0.05, TukeyÕs), while S2E2 reduced binding by 39.4 ± 5.4% (n = 4, P < 0.001, TukeyÕs). Treatment with S3E3 reduced binding by 87.2 ± 2.0% (n = 4, P < 0.001, TukeyÕs) compared with control (Fig. 5). Discussion
Fig. 4. Splice variant-specific RNAi of rAT1a receptor transcripts in RASMC. Forty-eight hours after siRNA treatment, cells treated with S1E1,3 siRNA showed marked reductions in E1,3 mRNA (A) but not E1,2,3 mRNA (B). Conversely, cells treated with S2E2 siRNA showed marked reductions in E1,2,3 mRNA but not E1,3 mRNA. Treatment with S3E3 siRNA caused significant reductions in levels of both E1,3 and E1,2,3 mRNA. The effect of siRNA treatment on total rAT1a receptor mRNA is shown in (C). Data are expressed as means ± SEM. *P < 0.05; ***P < 0.001.
siRNA-mediated silencing of rAT1a receptor splice variants reduced AT1 receptor binding in RASMC Treatment with all three rAT1a receptor mRNA targeting siRNAs for 48 h at a concentration of 10 nM reduced AT1 receptor binding (as measured by 125I-[Sar1,Ile8]Ang II
The aim of this study was to determine the role of the two rAT1a receptor mRNA splice variants in functional AT1 receptor expression in cells endogenously expressing AT1 receptors. We developed RNAi technology to selectively silence rAT1a receptor splice variants in RASMC and measured the effect of silencing these variants on AT1 receptor binding. The first step was to design siRNA sequences which could selectively mediate the degradation of the E1,3 transcript (i.e., S1E1,3), the E1,2,3 transcript (S2E2) or both the E1,3 and E1,2,3 transcripts (S3E3). Testing of these siRNA designs in CHO cells expressing either the E1,3 or E1,2,3 rAT1a receptor splice variants demonstrated that the chosen siRNA designs induced a potent and highly selective degradation of their intended targets in this cell type (Fig. 2). Next, we investigated whether these siRNA constructs were similarly selective in RASMC, a primary cell line that endogenously expresses both rAT1a receptor splice variants. Analysis of rAT1a receptor mRNA levels following S1E1,3, S2E2, and S3E3 treatment confirmed that the siRNA designs mediated highly specific degradation of their intended targets in cells endogenously expressing both transcripts. It is important to note that the E1,2,3 splice variant is predominant in RASMC, making up 70% of the total rAT1a receptor mRNA pool in these cells, with E1,3 mRNA accounting for the remainder. Thus, although S1E1,3 siRNA treatment induced a marked and significant reduction in E1,3 mRNA, this reduction amounted to only a small reduction in total rAT1a receptor mRNA. Measurement of AT1 receptor binding 48 h after transfection demonstrated that, in addition to their effects on rAT1a receptor mRNA, treatment with each of these siRNA duplexes induced a subsequent significant reduction in AT1 receptor binding compared with the non-silencing control, indicating that not only did siRNA treatment reduce expression of rAT1a receptor transcripts but it also reduced the level of functional AT1 receptors at the cell surface. The present data also demonstrate that, because of the highly sequence-specific nature of the interaction between the siRNA guide and its target sequence, it is possible to use RNAi to selectively reduce the expression of individual mRNA splice variants while leaving other splice variants intact in mammalian cells. RNAi of transcripts containing particular exons has previously been demonstrated in cultured Drosophila cells [26] and MDA231 breast cancer cells [27], respectively. The data described in this study extend these findings by demonstrating that it is possible to selec-
504
A. Hassan et al. / Biochemical and Biophysical Research Communications 339 (2006) 499–505
human AT1 (hAT1) receptor have also revealed that splicing in the 5 0 LS can affect receptor expression [9]. Furthermore, studies of the hAT1 receptor suggest that alterations in splicing contribute to cardiovascular pathophysiology. Analysis of hAT1 receptor transcripts in endomyocardial biopsies found that in failing hearts, the proportion of exon 2-containing transcripts in the total hAT1 receptor mRNA pool was significantly increased in the atria and left ventricle by 22% and 16%, respectively [32]. While alternative splicing may play a critical role in regulation of the RAS, the mechanisms by which alternative splicing of the AT1 receptor is regulated remain unclear. Splice variant-specific silencing of rAT1a receptor transcripts provides a powerful tool with which to investigate these regulatory processes. Fig. 5. Treatment with siRNA targeting rAT1a receptor splice variants reduces AT1 receptor binding in RASMC. Forty-eight hours treatment with siRNA targeting either the E1,3 splice variant (S1E1,3), E1,2,3 splice variant (S2E2), or both splice variants (S3E3) induced significant reductions in binding of 125I-[Sar1,Ile8]Ang II. Data are expressed as means ± SEM. *P < 0.05; ***P < 0.001.
tively silence transcripts that do not contain a particular exon by targeting the junction between two exons. Since alternative splicing makes a highly important contribution to the complexity of the proteome and is regulated developmentally and in response to cellular stimuli [28,29], it will be important to investigate the role of alternative splicing in many physiological processes. Use of splice variant-specific RNAi is likely to be an invaluable tool in these investigations. Finally, these data demonstrate that it is possible to potently reduce expression of the rAT1a receptor using siRNA-mediated RNAi in RASMC. While evidence from in vitro and cell culture systems suggests that any mRNA can be silenced using RNAi techniques [23], and a previous study has demonstrated silencing of the rAT1a receptor in transiently transfected CHO cells [30], it was unclear whether selective silencing could be achieved using siRNA in cells endogenously expressing the receptor since the ability to induce a reduction in protein expression may be limited by the rate of turnover of the specific mRNA and protein—two factors which are likely to be cell type-specific. Furthermore, it was uncertain what length of time silencing of the rAT1a receptor could be maintained in RASMC. Using S3E3 siRNA (which targeted both rAT1a receptor transcripts), we demonstrated that siRNA mediated silencing occurs rapidly, with a marked and significant reduction in AT1 receptor binding observed by 24 h after transfection and that this reduction could be maintained for at least 7 days post-transfection. Silencing occurred at relatively low siRNA concentrations (peak silencing was observed at 10 nM). We have previously shown that alternative splicing of the rAT1 receptor contributes to tissue-specific differences in rAT1 receptor expression at least in part by altering rates of mRNA translation [19,22,31]. Studies of the
Acknowledgments This work was supported by the American Heart Association (Mid-Atlantic Affiliate) Jocelyn Beard Moran Fellowship (0425548U) and a National Kidney Foundation (National Capital Area) Grant-in-Aid to A.H. and by NIH Grant HL57502 to K.S. References [1] P. Corvol, X. Jeunemaitre, A. Charru, Y. Kotelevtsev, F. Soubrier, Role of the renin–angiotensin system in blood pressure regulation and in human hypertension: new insights from molecular genetics, Recent Prog. Horm. Res. 50 (1995) 287–308. [2] J.R. Teerlink, Neurohumoral mechanisms in heart failure: a central role for the renin–angiotensin system, J. Cardiovasc. Pharmacol. 27 (Suppl. 2) (1996) S1–S8. [3] K. Sandberg, H. Ji, Kidney angiotensin receptors and their role in renal pathophysiology, Semin. Nephrol. 20 (2000) 402–416. [4] K. Sandberg, H. Ji, Comparative analysis of amphibian and mammalian angiotensin receptors, Comp. Biochem. Physiol. A Mol. Integr. Physiol. 128 (2001) 53–75. [5] K.K. Griendling, B. Lassegue, R.W. Alexander, Angiotensin receptors and their therapeutic implications, Annu. Rev. Pharmacol. Toxicol. 36 (1996) 281–306. [6] U. Stroth, T. Unger, The renin–angiotensin system and its receptors, J. Cardiovasc. Pharmacol. 33 (Suppl. 1) (1999) S21–S28, discussion S41-23. [7] R.M. Touyz, E.L. Schiffrin, Signal transduction mechanisms mediating the physiological and pathophysiological actions of angiotensin II in vascular smooth muscle cells, Pharmacol. Rev. 52 (2000) 639– 672. [8] G. Nickenig, D.G. Harrison, The AT(1)-type angiotensin receptor in oxidative stress and atherogenesis: Part II: AT(1) receptor regulation, Circulation 105 (2002) 530–536. [9] M.M. Martin, B.M. Willardson, G.F. Burton, C.R. White, J.N. McLaughlin, S.M. Bray, J.W. Ogilvie Jr., T.S. Elton, Human angiotensin II type 1 receptor isoforms encoded by messenger RNA splice variants are functionally distinct, Mol. Endocrinol. 15 (2001) 281–293. [10] N.P. Martin, R.J. Lefkowitz, S.K. Shenoy, Regulation of V2 vasopressin receptor degradation by agonist-promoted ubiquitination, J. Biol. Chem. 278 (2003) 45954–45959. [11] Y. Zhang, H. Ji, M. Elena Fabucci, C. Falconetti, W. Zheng, K. Sandberg, Translational control of the rat angiotensin type 1a receptor by alternative splicing, Gene 341 (2004) 93–100.
A. Hassan et al. / Biochemical and Biophysical Research Communications 339 (2006) 499–505 [12] T.S. Elton, M.M. Martin, Alternative splicing: a novel mechanism to fine-tune the expression and function of the human AT1 receptor, Trends Endocrinol. Metab. 14 (2003) 66–71. [13] S.S. Kakar, K.K. Riel, J.D. Neill, Differential expression of angiotensin II receptor subtype mRNAs (AT-1A and AT-1B) in the brain, Biochem. Biophys. Res. Commun. 185 (1992) 688–692. [14] S.S. Kakar, J.C. Sellers, D.C. Devor, L.C. Musgrove, J.D. Neill, Angiotensin II type-1 receptor subtype cDNAs: differential tissue expression and hormonal regulation, Biochem. Biophys. Res. Commun. 183 (1992) 1090–1096. [15] Y. Kitami, T. Okura, K. Marumoto, R. Wakamiya, K. Hiwada, Differential gene expression and regulation of type-1 angiotensin II receptor subtypes in the rat, Biochem. Biophys. Res. Commun. 188 (1992) 446–452. [16] K. Sandberg, H. Ji, A.J. Clark, H. Shapira, K.J. Catt, Cloning and expression of a novel angiotensin II receptor subtype, J. Biol. Chem. 267 (1992) 9455–9458. [17] K. Langford, K. Frenzel, B.M. Martin, K.E. Bernstein, The genomic organization of the rat AT1 angiotensin receptor, Biochem. Biophys. Res. Commun. 183 (1992) 1025–1032. [18] T.J. Murphy, R.W. Alexander, K.K. Griendling, M.S. Runge, K.E. Bernstein, Isolation of a cDNA encoding the vascular type-1 angiotensin II receptor, Nature 351 (1991) 233–236. [19] Y. Zhang, H. Ji, M. Elena Fabucci, C. Falconetti, W. Zheng, K. Sandberg, Translational control of the rat angiotensin type 1a receptor by alternative splicing, Gene 341 (2004) 93–100. [20] B.C. Berk, Angiotensin II signal transduction in vascular smooth muscle: pathways activated by specific tyrosine kinases, J. Am. Soc. Nephrol. 10 (Suppl. 11) (1999) S62–S68. [21] V. Briand, L. Riva, A.M. Galzin, Characterization of the angiotensin II AT1 receptor subtype involved in DNA synthesis in cultured vascular smooth muscle cells, Br. J. Pharmacol. 112 (1994) 1195–1201. [22] S. Lee, H. Ji, Z. Wu, W. Zheng, A. Hassan, K. Sandberg, Translational regulation of angiotensin Ii type 1 receptors in proliferating vascular smooth muscle cells, Am. J. Physiol. Regul. Integr. Comp. Physiol. (2005).
[23] S.M. Elbashir, J. Harborth, K. Weber, T. Tuschl, Analysis of gene function in somatic mammalian cells using small interfering RNAs, Methods 26 (2002) 199–213. [24] N.J. Caplen, S. Parrish, F. Imani, A. Fire, R.A. Morgan, Specific inhibition of gene expression by small double-stranded RNAs in invertebrate and vertebrate systems, Proc. Natl. Acad. Sci. USA 98 (2001) 9742–9747. [25] H. Ji, W. Zheng, Y. Zhang, K.J. Catt, K. Sandberg, Genetic transfer of a nonpeptide antagonist binding site to a previously unresponsive angiotensin receptor, Proc. Natl. Acad. Sci. USA 92 (1995) 9240–9244. [26] A.M. Celotto, B.R. Graveley, Exon-specific RNAi: a tool for dissecting the functional relevance of alternative splicing, RNA 8 (2002) 718–724. [27] B. Ge, X. Xiong, Q. Jing, J.L. Mosley, A. Filose, D. Bian, S. Huang, J. Han, TAB1beta (transforming growth factor-beta-activated protein kinase 1-binding protein 1beta), a novel splicing variant of TAB1 that interacts with p38alpha but not TAK1, J. Biol. Chem. 278 (2003) 2286–2293. [28] J. Xie, D.L. Black, A CaMK IV responsive RNA element mediates depolarization-induced alternative splicing of ion channels, Nature 410 (2001) 936–939. [29] S.M. Rao, R.D. Howells, cis-acting elements in the 5 0 -untranslated region of rat testis proenkephalin mRNA regulate translation of the precursor protein, J. Biol. Chem. 268 (1993) 22164–22169. [30] J. Vazquez, M.F. Correa de Adjounian, C. Sumners, A. Gonzalez, C. Diez-Freire, M.K. Raizada, Selective silencing of angiotensin receptor subtype 1a (AT1aR) by RNA interference, Hypertension 45 (2005) 115–119. [31] H. Ji, Y. Zhang, W. Zheng, Z. Wu, S. Lee, K. Sandberg, Translational regulation of angiotensin type 1a receptor expression and signaling by upstream AUGs in the 5 0 leader sequence, J. Biol. Chem. 279 (2004) 45322–45328. [32] C. Warnecke, D. Surder, R. Curth, E. Fleck, V. Regitz-Zagrosek, Analysis and functional characterization of alternatively spliced angiotensin II type 1 and 2 receptor transcripts in the human heart, J. Mol. Med. 77 (1999) 718–727.