HSP90 stabilizes unloaded Argonautes. (A and B) Increasing the level of loaded hAgo2 makes it more resistant to geldanamycin treatment. HeLa cells were transfected with increasing amount of siRNA targeting the Firefly luciferase without (A) and with (B) the cognate target plasmid followed by treatment of either DMSO or geldanamycin (10 μM for 16 h). The level of endogenous hAgo2 was assayed with Western blotting. β-Actin was used as a loading control. (C and D) Argonaute that is impaired in small RNA binding is more sensitive to HSP90 inhibition. (C) HeLa cells were transiently transfected (24 h) with FLAG-tagged hAgo2 (FLAG:Ago2) or FLAG-tagged hAgo2 with a series of mutations within its PAZ domain (FLAG:Ago2PAZ10) and were subsequently treated with either DMSO or geldanamycin (10 μM for 16 h). Expressions of the overexpressed plasmids and endogenous hAgo2 were assayed with Western blotting. β-Actin was used as a loading control. (D) A similar experiment to that desribed in C, only the FLAG-tagged Ago2s were cotransfected with 10 nM nontargeting siRNA.

Inhibition of HSP90 does not affect Argonaute–Dicer interaction but impairs siRNA efficiency (A) Geldanamycin treatment does not change the amount of Dicer associated with hAgo2. hAgo2 was immunoprecipitated from mock- and geldanamycin-treated (10 μM for 16 h) Hela cells, and the bound fractions were assayed for the presence of hAgo2 and Dicer with Western blotting. Sup, supernatant. Ten percent of the total lysate and 25% of the total IP were loaded. A longer exposure of the Dicer Western blot is presented to show input Dicer levels. (B and C) Geldanamycin treatment decreases siRNA efficiency. Mock- and geldanamycin-treated (10 μM for 16 h) HeLa cells were cotransfected with renilla and firefly luciferase plasmids and with increasing concentrations of siRNA targeting the firefly luciferase. Dual luciferase data at each concentration were normalized to data obtained with a nontargeting siRNA (10 nM) and fitted to a sigmoid curve using a Hill coefficient of 1. The concentration of siRNA required for half-maximal inhibition (IC50) was calculated and separately plotted (C).

Geldanamycin treatments disperse microscopic P-bodies and decrease the level of Argonautes and GW182 proteins. (A and B) Geldanamycin decreases the size and number of P-bodies in human cells. HeLa (A) and T-Rex 293 (B) cells were treated with geldanamycin (7 μM for 16 h), and P-bodies were visualized with immunostaining for endogenous GW182 (A) or overexpressed FLAG-tagged TNRC6C (B). DAPI staining was used to visualize cell nuclei. (C) Geldanamycin dampens endogenous Argonautes and GW182 expression. HeLa cells were treated with geldanamycin (10 μM for 16 h), and the expression level of proteins involved in the miRNA pathway as well as the let-7 target Ras were followed by Western blotting. β-Actin was used as a loading control. The numbers on the top of each panel indicate the relative abundance of the respective proteins. (D) Overexpressed Ago2 and TNRC6C are sensitive to HSP90 inhibition. 293 T-Rex cells stably expressing tetracycline-inducible (Tet) FLAG-tagged hAgo2, and FLAG-tagged TNRC6C were treated with geldanamycin (GD; 10 μM for 16 h), and the expressions of the transgenes were followed by Western blot using a FLAG antibody. The efficiencies of the geldanamycin treatments were confirmed by Western blotting of endogenous hAgo2. α-Tubulin was used as the loading control. The numbers on the top of each panel indicate the relative abundance of the respective proteins.

Geldanamycin treatment does not alter miRNA level and miRNA function in human cells. (A) Geldanamycin decreases Ago2 and GW182 levels after 8 h treatment. HeLa cells were treated with DMSO or with geldanamycin (10 μM) for the indicated times. The protein level of endogenous hAgo2 and GW182 were followed by Western blotting. β-Actin was used as loading control for Western blotting. (B) miRNA level is unaltered after up to 24 h of geldanamycin treatment. HeLa cells were treated with DMSO or geldanamycin (10 μM) for the indicated times shown in A. The relative let-7 level between the DMSO and geldanamycin-treated cells was calculated at each time point using U6 as a loading control. The graph shows the mean of three independent experiments; error bars, ±SE. (C) Inhibition of HSP90 activity does not impair miRNA functions. HeLa cells were pretreated with geldanamycin (10 μM for 16 h) and transfected with luciferase reporters that measure translational repression and RNAi-mediated by endogenous let-7 (left panel) and reporters that contain 3′UTRs of known let-7–targeted mRNAs (right panel). The dual luciferase data obtained from the geldanamycin-treated cells (GD) were normalized to the luciferase reading of the mock-treated cells (M). Error bars, ±SE of three experiments. Left, no let-7, Renilla luciferase containing no let-7 sites; let-7X8, Renilla luciferase containing eight let-7 sites that mediate translational repression; let-7X3, Renilla luciferase contains three perfect complementary let-7a sites; and let-7X3m, similar to let-7X3 but carrying three point mutations in the let-7a seeds and one at the cleavage sites. Right, nRAS, Firefly luciferase carrying the 3′UTR of nRAS; kRAS, Firefly luciferase carrying the 3′UTR of kRAS. (D) HSP90 association with complexes containing mature miRNAs is not stoichiometric. Affinity purification was carried out with control (c) and 2′-O-methyl oligo complementary to human let-7a (let-7) in T-Rex cells inducible expressing FLAG::Ago2. The bound fractions were assayed for the presence of FLAG::Ago2, hAgo2, and HSP90 with Western blotting (left panel); 10% of total lysate was loaded and half of the bound fraction. FLAG IP was carried out in the same lysate using identical conditions. The bound fraction was assayed for FLAG::Ago2, and HSP90 with Western blotting (right panel). Empty bead (e) was used as a negative control in the IP experiment. Ten percent of the total lysate was loaded, and half of the IP for the HSP90 panel and 10% of the total IP for the Flag::Ago2 panel was loaded.