Cytoplasmic localization of tagged hAgo2 proteins and the RL-5BoxB reporter mRNA. HeLa cells were cotransfected with pRL-5BoxB and plasmids expressing either N-HA-hAgo2 or HA-hAgo2. Cells were lysed in a buffer containing the different indicated concentrations of NP-40 and fractionated into nuclear (N) and cytoplasmic (C) fractions. Equivalent amounts of the fractions were analyzed for RL-5BoxB mRNA by Northern blotting and for the indicated proteins by Western blotting. The U2B″ protein and α-tubulin served as markers for nuclear and cytoplasmic proteins, respectively.

Tethered hAgo2 and hAgo4, but not hAgo2 mutants and Hiwi, induce the repression. (A) RL activity in extracts from HeLa cells cotransfected with plasmids expressing RL-5BoxB reporter and N-HA-tagged mutant hAgo2 fusions. Western analysis of fusion protein expression is shown below the histograms. Schematic representation of hAgo2 and its deletion mutants is shown in lower panel. (B) Tethering of hAgo4 but not Hiwi represses translation. For other details, see legend to Figure 1 ▶.

Tethered hAgo2 down-regulates protein synthesis. (A) Scheme of the basic RL mRNA reporter, containing five 19-nt BoxB hairpins, interacting with N-HA-hAgo2. (B) RL activity detected in extracts from HeLa cells expressing the indicated fusion proteins. Cells were cotransfected with constructs expressing the RL-5BoxB reporter, FL, and indicated fusion proteins. Histograms in panels B through D represent normalized mean values (±SD) of RL/FL activities from a minimum of three experiments. RL activity values seen in the presence of HA-hAgo2 were set as 1. Expression levels of fusion proteins, as determined by Western analysis using anti-HA mAb, are shown below the histograms. The aliquot of the N-HA-LacZ-expressing extract applied to the gel was 10 times smaller than aliquots of other extracts. Generally, the N-HA-LacZ protein is expressed at an ~10-fold higher level than N-HA-hAgo2 or HA-hAgo2. Increased RL activity in extracts expressing N-HA-LacZ is likely due to the effect of the protein on the stability of mRNA reporters containing BoxB sequences (see also Fig. 5 ▶). (C) Activity of reporter RL mRNAs containing different numbers of BoxB hairpins. Relative activities of different reporters in cells coexpressing HA-hAgo2 were as follows: No BoxB, 1.0; 1BoxB, 1.0; 2BoxB, 0.7; 3BoxB, 1.4; and 5BoxB, 1.2. (D) Activity of RL-5BoxB mRNAs with hairpins spaced (sp) 36, 253, 469, or 684 nt downstream of the translation termination codon. Relative activities of different reporters in cells coexpressing HA-hAgo2 were as follows: sp36, 1.0; sp253, 1.6; sp469, 1.7; and sp684, 1.0. (E) Activity of the FL-5BoxB reporter mRNA in extracts from cells expressing the indicated fusion proteins. Histograms represent normalized mean values of FL/RL activities from three experiments. FL activity seen in the presence of HA-hAgo2 was set to 1.

N-HA-hAgo2 protein is active in the let-7-directed RNA cleavage and interacts with endogenous Gemin 4. (A) Processing of perfectly complementary and bulged substrates uniformly labeled with [α-³²P]-UTP. The sequence of the RNA substrate is shown at the top, with the sequence complementary to let-7 RNA underlined. The corresponding sequence of the mutant RNA substrate, forming a bulged duplex on base-pairing to let-7 RNA, is shown below, with mutated nucleotides indicated in bold. (Lanes N-HA-LacZ and N-HA-hAgo2) Anti-HA immunoprecipitates (IPs) of extracts expressing N-HA-LacZ and N-HA-hAgo2 proteins, respectively. (Lane “Buffer”) Incubations performed in the absence of anti-HA IPs. Positions of the 5′ and 3′ cleavage products and 27- and 21-nt RNA markers are indicated. Note that the 3′ cleavage product contains 2.5 times more label than does the 5′ product. (B) Mapping of the cleavage site in the 5′-terminally labeled fully complementary RNA. (Lanes T1 and Alkaline) RNase T1 and alkaline ladders prepared with the same substrate RNA. Note that the ladder oligoribonucleotides contain a 3′-terminal phosphate, which is absent in the let-7-mediated 5′-cleavage product. Position of the cleavage between C and U residues is indicated by an arrow. Low processing activity seen in N-HA-LacZ control lanes in panels A and B is due to unspecific retention of endogenous let-7 miRNP on the HA-antibody-containing beads (data not shown). (C) Interaction of N-HA-Ago2 protein with endogenous Gemin 4. Anti-HA IPs from extracts expressing N-HA-LacZ and N-HA-hAgo2 fusion proteins were analyzed by Western for Gemin 4 and HA-tagged proteins. Input fraction represents 2% of the N-HA-hAgo2 extract used for IP.

Repression by N-HA-hAgo2 and N-HA-hAgo4 occurs without changes in reporter mRNA level. Northern analysis (middle panels) was performed with total RNA isolated from transfected cells, using probes specific for RL-5BoxB mRNA and green fluorescence protein (GFP) mRNA, expressed from the cotransfected plasmid. The RL activity in extracts from the same transfected cells is shown in the upper panel. PhosphorImaging quantification of the RL-5BoxB mRNA, normalized to GFP mRNA, is shown in the lower panel; values are means (±SD) from three independent experiments. The RL-5BoxB mRNA level in cells cotransfected with HA-hAgo2 is set to 1.