|
Figure 1. Schematic representation of Xenopus ADAR1 and mutant constructs used in this paper. The 1,271–amino acid Xenopus ADAR1 protein is depicted to scale at the top with mutant constructs shown underneath. Subregions of the protein are indicated as follows: REP, 11-aa peptide repeats; ZBD, Z-DNA binding domain; NLS, nuclear localization signal; dsRBDs, double-stranded RNA-binding domains; and Deaminase, catalytic deaminase domain. The ability of constructs to label chromosomes (Chr) and enrich at the special loop (Sp. loop) is shown on the right and indicated either as positive (+), negative (−), or patchy (+/−). Deletion of the deaminase domain (Δ deaminase) had no effect on chromosomal labeling and special loop enrichment. Expression of the central part of the protein from the end of the ZBD up to the end of dsRBD3 (dsRBD1-2-3) resulted in the same localization pattern as the wild-type protein, which is indicated by chromosomal association and special loop enrichment. Conversely, removal of the three dsRBDS from the full-length protein (Δ dsRBDs) resulted in a loss of all nuclear staining. Replacement of the endogenous dsRBDs with single dsRBDs (dsRBD2), duplicated dsRBDs (dsRBD2-2) or triplicated dsRBDs (dsRBD2-2-2) failed to restore normal chromosomal association. All constructs gave patchy labeling (+/−) and none enriched at the special loop (for simplicity, only dsRBD2 containing constructs are shown).
|
|
Figure 2. Deletion of the deaminase domain does not affect chromosomal localization. (A) RNA transcribed from full-length myc-tagged ADAR1 and a myc-tagged construct where the deaminase domain had been removed was injected individually into Xenopus oocytes. Lampbrush chromosome preparations were made and in vivo translation of each RNA was followed using the anti–myc mAb 9E10 and a secondary FITC-labeled antibody (FITC). Simultaneously, all preparations were stained for endogenous ADAR1 using the SAT3 antisera that was detected with a secondary TRITC-labeled antibody (TRITC). Injection of full-length ADAR1 (top) resulted in normal chromosomal localization compared with the endogenous protein. This was also the case when the deaminase domain was removed (bottom, Δ Deaminase), indicating at least at the chromosomal level that this domain does not play a role in targeting ADAR1 to nascent transcripts. Preparations were also stained with DAPI, and images of chromosomes were taken by differential interference contrast (NOM). Bar, 20 μm. (B) Western blot analysis of oocyte germinal vesicles (GV) and cytoplasms (CYT) from uninjected oocytes and oocytes expressing myc-tagged ADAR1 constructs to verify expression. Both myc-tagged versions express well and accumulate in the nucleus and are easily detected with the mAb 9E10. Cytoplasmic signals are seen 24 h after injection, but diminish over time as the protein is transported into the nucleus. In comparison, no signal is detected in uninjected oocytes indicating the specificity of mAb 9E10 for the injected constructs. Translated products correspond nicely to the predicted molecular masses of 175 kD (full-length ADAR1) and 130 kD (Δ Deaminase) indicated by arrows. Breakdown products can be observed at 120 kD (full-length) and 100 kD (Δ Deaminase). However, molecular mass calculations indicate that proteolytic cleavage has to occur upstream of the dsRBDs and, therefore, does not affect the RNA-binding capacity of the resulting fragments.
|
|
Figure 3. The dsRBDs are necessary and sufficient for chromosomal association. Lampbrush chromosome preparations were made from oocytes injected with RNA transcribed from either of two myc-tagged constructs. In vivo translation of both constructs was followed using mAb 9E10 and a secondary FITC-labeled antibody (FITC). Expression of the central part of the protein from the end of the Z-β domain up to the end of dsRBD3 (top, dsRBD1-2-3) results in normal chromosomal association and special loop enrichment. In contrast, removal of the three dsRBDs from the full-length protein (bottom, Δ dsRBDs) causes loss of chromosomal and intranuclear staining. All preparations were costained for endogenous ADAR1 using the SAT3 antiserum and detected with a secondary TRITC-labeled antibody (TRITC). Enrichment at the special loop on the third chromosome is marked by arrows. Preparations were also stained with DAPI, and images of chromosomes were taken by differential interference contrast (NOM). Bar, 20 μm.
|
|
Figure 4. Northwestern analysis of dsRBDs constructs. The RNA-binding assay of bacterially expressed dsRBD-GST fusion proteins was performed with rI/rC (A). In parallel, the same extracts were run on a gel and stained with Coomassie brilliant blue to allow quantification of the recombinant protein (B). The ratio of signal to protein was quantified by laser densitometry and is depicted graphically (C). The number and type of dsRBDs are indicated. Also shown is the empty pGEX vector used as a negative control, whereas the second dsRBD of XlrbpA is included as a positive control. Of the single domains, dsRBD2 and dsRBD3 were found to be the best RNA binders. Constructs containing duplications of individual dsRBDs showed a higher affinity for rI/rC, whereas triplicating individual dsRBDs further enhanced this affinity.
|
|
Figure 5. Individual dsRBDs can lead to ADAR1 enrichment on different transcription units. RNA transcribed from full-length myc-tagged ADAR1 constructs with either a single copy of dsRBD1 (A) or dsRBD2 (B) in place of the three endogenous dsRBDs was injected into oocytes. Translation of both constructs was followed using mAb 9E10 and a secondary FITC-labeled antibody. Simultaneously, all preparations were stained for endogenous ADAR1 using the SAT3 antisera (endogenous). Preparations were also stained with DAPI, and images of chromosomes were taken by differential interference contrast (NOM). Both dsRBD constructs were able to restore at least some level of chromosomal labeling, but at reduced levels compared with the endogenous ADAR1 staining. Interestingly, labeling often appeared at the same position along the arms of each homologue (A and B, arrows) indicating specific targeting to the same chromosomal loops. (C) To determine the influence of overall RNA–binding strength, constructs containing a duplication (dsRBD 2-2 GFP) or a triplication (dsRBD 2-2-2 MYC) of dsRBD were tagged with GFP and myc, respectively, injected, and detected within the same oocyte. Both constructs did label the same chromosomal sites (C, arrows), thus leading to a homogeneous yellow labeling in the overlay (merge). (D) To directly compare two individual dsRBDs on the same chromosome, RNA, transcribed from GFP-tagged ADAR1 containing a single dsRBD1, was coinjected with RNA made from a myc-tagged ADAR1 construct containing a single dsRBD2. Translation of the GFP-tagged construct was followed using appropriate antibodies. The dsRBD2-containing construct shows specific enrichment at a few chromosomal sites (D, top arrows), whereas the dsRBD1-containing construct shows a more homogeneous chromosomal staining. However, a few sites are specifically highlighted by the dsRBD1-containing constructs and, thus, appear green in the merged image (D, bottom arrows). Merged images of dsRBD1 and dsRBD2 labeling are shown (merge) as well as images of the chromosomes taken by differential interference contrast (NOM). Bar, 20 μm
|
|
Figure 6. Individual dsRBDs can target ADAR1 to different subsets of loops. Double injections of RNA from different myc- and GFP triplication constructs are shown as indicated. Translation of the myc constructs was detected using mAb 9E10 and a secondary TRITC-labeled antibody (TRITC), whereas the GFP constructs were detected using an anti–rabbit GFP antibody and a secondary FITC-labeled antibody (FITC). Merged enlargements are shown between (merge). (A) Injection of identical constructs label the same chromosomal sites. Shown is a double injection of RNA made from dsRBD1 triplications that were either myc- or GFP-tagged. Both tagged versions label the chromosomes equally well and target to the same sites and loops (e.g., arrows). Such a complete overlap becomes even more apparent when similar regions of the chromosome from each channel (boxed) are enlarged, colored, and merged. (B and C) Different dsRBDs target ADAR1 to different chromosomal sites. RNA made from GFP-tagged dsRBD1 triplication was injected into oocytes followed by injection of RNA either from myc-tagged dsRBD2 triplication (B) or myc-tagged dsRBD3 triplication (C). Two different representative chromosomal sites are shown in B. Each construct labels chromosomes differently with some sites being labeled by one construct not being labeled by the other (arrows). Consequently, in the merged images, there is no complete overlap of staining. Together, this demonstrates the capability of individual dsRBDs specifically targeting different transcription units. Images of chromosomes taken by differential interference contrast (NOM) are also shown. Bar, 10 μm. (D) Western blot analysis of oocyte germinal vesicles (GV) and cytoplasms (CYT) from uninjected oocytes and oocytes expressing myc- and GFP-tagged ADAR1 constructs shown in A–C to verify expression. Translated products correspond nicely to the predicted molecular masses of ∼180 kD (arrows).
|