A novel EST-derived RNAi screen reveals a critical role for farnesyl diphosphate synthase in β2-adrenergic receptor internalization and down-regulation

Abstract

The β2-adrenergic receptor (β2AR) plays important physiological roles in the heart and lung and is the primary target of β-agonists, the mainstay asthma drugs. Activation of β2AR by β-agonists is attenuated by receptor down-regulation, which ensures transient stimulation of the receptor but reduces the efficacy of β-agonists. Here we report the identification, through a functional genome-wide RNA interference (RNAi) screen, of new genes critically involved in β2AR down-regulation. We developed a lentivirus-based RNAi library consisting of 26-nt short-hairpin RNAs (shRNAs). The library was generated enzymatically from a large collection of expressed sequence tag (EST) DNAs corresponding to ∼20,000 human genes and contains on average ∼6 highly potent shRNAs (>75% knockdown efficiency) for each gene. Using this novel shRNA library, together with a robust cell model for β2AR expression, we performed fluorescence-activated cell sorting and isolated cells that, as a consequence of shRNA-mediated gene inactivation, exhibited defective agonist-induced down-regulation. The screen discovered several previously unrecognized β2AR regulators, including farnesyl diphosphate synthase (FDPS). We showed that inactivation of FDPS by shRNA, small interfering RNA, or the highly specific pharmaceutical inhibitor alendronate inhibited β2AR down-regulation. Notably, in human airway smooth muscle cells, the physiological target of β-agonists, alendronate treatment functionally reversed agonist-induced endogenous β2AR loss as indicated by an increase in cAMP production. FDPS inactivation interfered with β2AR internalization into endosomes through disrupting the membrane localization of the Rab5 small GTPase. Furthermore, Rab5 overexpression reversed the deficient receptor down-regulation induced by alendronate, suggesting that FDPS regulates receptor down-regulation in a Rab5-dependent manner. Together, our findings reveal a FDPS-dependent mechanism in the internalization and down-regulation of β2AR, identify FDPS as a potential target for improving the therapeutic efficacy of β-agonists, and demonstrate the utility of the unique EST-derived shRNA library for functional genetics studies.

Figure 1.

Development of an EST-derived 26-nt shRNA library. A) Schematic outline of the enzymatic procedure to convert double-stranded DNA fragments into structures expressing 26-nt shRNAs. Details are given in Materials and Methods. B) Lentiviral shRNA-expressing vector pLentiSuper2. 5′LTR, 5′ long terminal repeat; 3′ΔLTR, promoter-deleted 3′ long terminal repeat; H1, RNA polymerase III H1 promoter; SV40, SV40 viral prompter; BSD, blasticidin selection marker; eshRNA, enzymatic prepared shRNA; Amp, ampicillin selection marker. C) Knockdown efficiency of shRNAs from a single gene library. HEK293T cells were cotransfected with 800 ng DNA of 60 individual shRNA-expressing clones or empty vector pLentiSuper2, together with 200 ng DNA of target gene (ARRDC1-GFP) and 50 ng DNA of nontarget control. Expression of both target and control genes were detected by immunoblotting using an anti-GFP antibody. Expression level of ARRDC1-GFP was normalized to that of GFP. Knockdown efficiency by shRNA was calculated as percentage relative to empty vector controls. D) Pooling and conversion of a large collection of ESTs into the shRNA library.

Figure 2.

A FACS-based shRNA screen identified new genes required for β2AR down-regulation. A) Efficient down-regulation of β2AR in 293β2AR cells. Cells were treated with 10 μM ISO for 16 h, and the receptor expression was determined by immunoblotting. B) Flow cytometry analysis of β2AR level at the cell surface. 293β2AR cells were mock-treated or treated with 10 μM ISO for 16 h and stained with FITC-conjugated anti-Flag antibody. Cells not stained with the antibody were used as a negative control. C) Schematic of a FACS-based shRNA screen for genes required for β2AR down-regulation. D) FACS enrichment of 293β2AR cells with a high level of surface β2AR. Cells were infected with the EST-derived shRNA lentiviral library, treated with 10 μM ISO for 16 h, and stained with FITC-conjugated anti-Flag antibody. Four rounds of sorting were performed to enrich FITC-bright cells. Graphs show the flow cytometry analyses of cell populations prior to the first or the fourth round of sorting with gating conditions indicated. See Table 1 for list of validated gene hits from the RNAi screen.

Figure 3.

FDPS is required for β2AR down-regulation. A) Suppression of FDPS expression by siRNA increased cell surface β2AR level on ISO stimulation. 293β2AR cells were transfected with FDPS-specific siRNA or nontargeting siRNA (NT), and 72 h later cells were treated with 10 μM ISO for 3 h. Amount of surface β2AR was determined by flow cytometry. Three independent analyses were done. *P < 0.05. B) Inhibition of β2AR degradation by FDPS siRNA. 293β2AR cells were treated as in A, and the corresponding lysates were subjected to immunoblotting. C) SiRNA-mediated knockdown of FDPS inhibited endogenous β2AR down-regulation. Primary HASM cells transfected with control or FDPS siRNAs were cultured in the absence or the presence of 30 μM ISO for 3 d. Immunoblotting was done using the indicated antibodies.

Figure 4.

Effects of FDPS inhibitor ALN on β2AR down-regulation and β-agonist-induced cAMP formation. A) Effect of ALN on down-regulation of the fast-degrading β2AR. 293β2AR cells were treated with 50 μM ALN for 48 h and then exposed to 10 μM ISO for 3 h, and cell lysates were subjected to immunoblotting. B) Effects of ALN on full-length β2AR down-regulation. HEK293 cells stably expressing Flag-tagged full-length β2AR were treated with 50 μM ALN and 10 μM ISO for 48 h. β2AR expression was determined by immunoblotting. C) Concentration-dependent effects of ALN on β2AR down-regulation. HASM cells were treated with indicated concentrations of ALN in the absence and presence of 30 μM ISO for 3 d. Cell lystaes were analyzed by immunoblotting using indicated antibodies. D) Time-dependent effects of ALN on β2AR down-regulation. HASM cells were treated with 10 μM ALN for the indicated time in the absence and presence of 30 μM ISO for 3 d. Cell lysates were analyzed by immunoblotting. E) Effects of ALN on agonist-stimulated cAMP production. HASM cells were treated with 3 d ISO in the absence or presence of ALN (10 μM, 3 d), washed with PBS, and cultured in ISO-free medium for 4 h before subjected to acute ISO stimulation for 10 min. cAMP level was measured as described in Materials and Methods. *P < 0.05; **P < 0.01.

Figure 5.

FDPS modulates β2AR internalization and down-regulation in a Rab5-dependent manner. A) Effect of ALN on β2AR internalization into early endosomes. 293β2AR cells were treated with 50 μM ALN for 48 h and then stimulated with 10 μM ISO for 30 min. Cells were then fixed, permeablized, and stained with primary antibodies (polyclonal anti-Flag antibody for Flag-β2AR and a monoclonal antibody for EEA1) and respective TRIC- or FITC-conjugated secondary antibodies. Images were visualized under confocal microscopy. B) Effect of ALN on β2AR internalization. 293β2AR were mock-treated or treated with ALN and stimulated with 10 μM ISO for 30 min to induce internalization. Flow cytometry was used to measure amount of β2AR at the cell surface. Percentages of β2AR remained at the cell surface on ISO stimulation were calculated relative to surface receptor level without ISO treatment. ***P < 0.001. C) Effect of ALN on β2AR trafficking into late endosomes. 293β2AR cells were treated as in A and then stimulated with 10 μM ISO for indicated time. Immunostaining was performed to detect Flag-β2AR and LAMP2. D) Effect of wild-type or mutant Rab5 expression on β2AR down-regulation. 293β2AR cells were transfected with wild-type or mutant (S34N) GFP-Rab5 expression plasmids and after 48 h, cells were treated with 10 μM ISO for 3 h. Cell extracts were subjected to immunoblotting to detect receptor levels. Anti-GFP antibody was used to detect the expression of GFP-Rab5 fusion protein. E) Effect of ALN on Rab5 membrane association. 293β2AR cells were treated with 50 μM ALN for 48 h followed by 10 μM ISO stimulation for 30 min. Cells were lysed, and membrane-associated fractions were isolated using the Mem-PER Mammalian Membrane Protein Extraction Kit (Pierce). Immunoblotting was done on the membrane-associated fraction (MAF) and the whole cell extract (WCE) using indicated antibodies. F) Effect of ALN on Rab5 localization. 293β2AR cells transfected with GFP-Rab5 were mock-treated or treated with ALN for 48 h and then subjected to 10 μM ISO stimulation (or control vehicle treatment) for 30 min. Cells were fixed, permeabilized, and stained with anti-Flag antibody. Images were visualized with a confocal fluorescence microscope. G) Effect of Rab5 overexpression on the inhibition by ALN on β2AR down-regulation. 293β2AR cells were transfected with Rab5 expression plasmid or control plasmid pCDNA3.1, and after 48 h cells were treated with 25 μM ALN or control vehicle for 24 h. Cell were then subjected to 10 μM ISO stimulation (or control vehicle treatment) for 2 h, and extracts were analyzed by immunoblotting.