The cytoplasm of Xenopus oocytes contains a factor that protects double-stranded RNA from adenosine-to-inosine modification.

Abstract

Here we describe studies of double-stranded RNA (dsRNA) adenosine deaminase in Xenopus laevis, in particular during meiotic maturation, the period during which a stage VI oocyte matures to an egg. We show that dsRNA adenosine deaminase is in the nuclei of stage VI oocytes. Most importantly, we demonstrate that the cytoplasm of stage VI oocytes contains a factor that protects microinjected dsRNA from deamination when dsRNA adenosine deaminase is released from the nucleus during meiotic maturation. Our data suggest that the protection factor is a cytoplasmic dsRNA-binding protein or proteins that bind to dsRNA in a sequence-independent manner to occlude dsRNA from binding to dsRNA adenosine deaminase. The cytoplasmic double-stranded RNA-binding protein(s) does not bind to other nucleic acids and can be titrated at high concentrations of dsRNA. These studies raise the question of whether all dsRNA-binding proteins share endogenous substrates and also suggest potential means of regulating dsRNA adenosine deaminase in vivo.The double-stranded RNA (dsRNA) adenosine deaminase (dsRAD), initially called the unwinding/modifying activity, was detected first in Xenopus laevis (4, 25) and subsequently in organisms throughout the animal kingdom (34; for reviews, see references 1, 2, and 17). This enzyme deaminates adenosines within dsRNA to produce inosines (24). Recently dsRAD has been purified from Xenopus eggs; this protein is a single subunit that migrates at 120 kDa by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (14). In vitro studies demonstrate that dsRAD can deaminate adenosines within intermolecular and intramolecular RNA duplexes (21, 29) but cannot act on adenosines within DNA or single-stranded RNA (ssRNA) (4, 33). Strong evidence indicating that dsRAD is responsible for RNA editing of mRNAs encoding the B subunit of a specific type of L-glutamate-activated ion channel has been presented (13). In addition to RNA editing, other biological roles for dsRAD have been proposed (for a review, see reference 2), and it is possible that dsRAD plays multiple roles in vivo.Initial experiments in X laevis showed that dsRAD could easily be detected by assaying the fate of dsRNA microinjected into the cytoplasm of eggs and early embryos (.8 cells; 4, 25).However, the activity could not be observed in similar microinjections into the cytoplasm of oocytes or embryos at later stages of development, specifically, subsequent to the midblastula transition (post-MBT). Further experiments revealed that dsRAD was present in post-MBT embryos but was localized to the nucleus and thus not detected in early experiments involving cytoplasmic microinjections (5). dsRAD has also been shown to be nuclear in the somatic cells of mammals (34). Taken together, these results raised the possibility that dsRAD is in the nuclei of stage VI oocytes and exists in the cytoplasm of eggs and early embryos because it is released from the nucleus during meiotic maturation, the process by which a stage VI oocyte becomes an egg (for a review, see ...

A minor fraction of basic fibroblast growth factor mRNA is deaminated in Xenopus stage VI and matured oocytes

“…Using three different assays we observed that the majority of bFGF mature mRNA was not deaminated in Xenopus stage VI oocytes+ We also found that the bFGF mRNA population isolated after oocyte maturation remained predominantly unmodified, despite its exposure to the ADAR released into the cytoplasm following GVBD+ The latter observation agrees with our earlier report that predicted that the factor cyto-dsRBP would protect cytoplasmically located dsRNA from deamination (Saccomanno & Bass, 1994)+ With all three assays we detected A-to-G transitions in a small fraction of the bFGF mRNA, whether the RNA was isolated from stage FIGURE 4. Restriction-enzyme assays of mixed cDNA populations+ A PhosphorImager image of restriction digests electrophoresed on a 1% agarose gel is shown+ A and B are two independent experiments+ bFGF RNA was amplified by RT-PCR using RNA isolated from either oocytes (lanes 1 and 2) or from matured oocytes (lanes 3 and 4)+ The 900-base-pair PCR products, corresponding to the sense strand of the overlap region, were incubated in the absence (Ϫ) or presence (ϩ) of restriction enzymes (DraIII, ApaI, StuI, and BpmI)+ The percent product digested was obtained by subtracting a background value obtained in the absence of restriction enzymes (Ϫ)+ Background values ranged from 0+6% to a maximum of 1+4%+ The percent digestion was determined with a Molecular Dynamics PhosphorImager using ImageQuaNT software+ VI oocytes or matured oocytes; this suggested that at least a small fraction of the bFGF mRNA from both stages was deaminated by ADAR+ Since ADAR activity is located in the nucleus in stage VI oocytes, our observation of a low level of deaminated bFGF mRNA at this stage indicates that a fraction of the bFGF sense and antisense RNAs are hybridized in the nucleus and acted on by ADARs+ As shown at the top of Figure 1, the bFGF premRNA transcripts can form a much longer duplex prior to splicing that might serve as an ADAR substrate in the nucleus+ However, we did not find any A-to-G changes in the sequence outside the region of complementarity between the mature RNAs, suggesting adenosine deamination occurred on the mature, spliced mRNA, rather than the pre-mRNA (see Fig+ 2 legend)+ Of course, further analysis will be required to prove this definitively+ Within the sensitivity limits of our experiments, the percentage of deaminated bFGF RNA molecules observed before and after oocyte maturation was similar+ We speculate that the small percentage of deaminated bFGF RNA molecules detected after oocyte maturation were deaminated in the oocyte nucleus and persisted after maturation+ However, previous studies indicate that a significant population of the bFGF sense and antisense mRNAs are hybridized in the oocyte cytoplasm (Kimelman & Kirschner, 1989)+ Therefore, we cannot rule out the possibility that a low level of cytoplasmically located bFGF duplex molecules escape protection from cyto-dsRBP and are deaminated by ADAR released at oocyte maturation+ In previous studies hypermutated bFGF RNA transcripts were not observed in the RNA isolated from stage IV or stage VI oocytes, or mixed oocyte populations from total ovary (Kimelman & Kirschner, 1989)+ However, our current data indicate that the percentage of deaminated bFGF transcripts in the RNA population from stage VI oocytes is low, and suggest that analysis of many cDNA clones would be required to identify those which are deaminated+ We believe that we identified deaminated cDNA clones derived from the stage VI oocyte RNA because we analyzed large populations of cDNA molecules utilizing three different methods+…”

“…One study suggests that a region of complementarity between the sense and antisense transcripts from the basic fibroblast growth factor (bFGF) gene are hypermutated according to the above scenario (Kimelman & Kirschner, 1989)+ In X. laevis the bFGF gene is transcribed bidirectionally to produce a 4+2-kb sense transcript and a 1+35-kb antisense transcript (Kimelman et al+, 1988;Volk et al+, 1989)+ A 900-base-pair overlap region exists between the mature sense and antisense transcripts, at each of their 39 ends (see Fig+ 1)+ Previously it was reported that, although the bFGF duplex was not deaminated in stage VI oocytes, it became deaminated after oocyte maturation, and was subsequently degraded (Kimelman & Kirschner, 1989)+ This report was consistent with our observation that ADAR activity is located in the nucleus in stage VI oocytes, and is released into the cytoplasm during oocyte maturation, when the nuclear membrane breaks down (germinal vesicle breakdown (GVBD); Saccomanno & Bass, 1994)+ That is, assuming the majority of the bFGF duplex is cytoplasmic, it should not be exposed to ADAR activity until after GVBD+ However, further analysis of ADAR activity in our laboratory identified a factor(s) in the cytoplasm of stage VI oocytes, cytoplasmic dsRNA binding protein (cyto-dsRBP), that specifically protects dsRNA from ADAR activity released at oocyte maturation (Saccomanno & Bass, 1994) + Although this factor was characterized by microinjection of an exogenous dsRNA, the presence of cyto-dsRBP was in conflict with the apparent deamination of endogenous bFGF RNA (Kimelman & Kirschner, 1989) and suggested the bFGF duplex should be protected from adenosine deamination in the stage VI oocyte and throughout oocyte maturation+ Discussions between the two laboratories provided a possible solution to the apparent conflicts+ In particular, although many laboratories induce maturation in vitro by incubating oocytes with progesterone, the protocols used by Kimelman and Kirschner involved microinjecting a crude preparation of maturation promoting factor (MPF) into stage VI oocytes+ Experiments in our laboratory demonstrated that the crude preparation of MPF contained a significant concentration of ADAR activity (M+ Paul & B+ Bass, unpubl+ data)+ Possibly, the injection of exogenous ADAR into the cytoplasm displaced the cyto-dsRBP bound to the bFGF duplex, leading to its aberrant deamination+ Nonetheless, since ADARs require base-paired substrates, the work by Kimelman and Kirschner indicates that the bFGF sense and antisense RNAs are hybridized in the cytoplasm of stage VI oocytes+…”

“…Stage VI oocytes and matured oocytes were isolated and prepared as described previously (Saccomanno & Bass, 1994), except that individual oocytes were removed from follicle cells with MBS (10 mM HEPES, pH 7+6, 88 mM NaCl, 1 mM KCl, 2+4 mM NaHCO 3 , 0+7 mM CaCl 2 , and 50 mg/mL of gentamicin sulfate) that contained 1 mg/mL of collagenase D+ Stage VI oocytes were matured by incubating in MBS containing 10 mg of progesterone per milliliter+ Matured oocytes were collected 6 h after ;70% of the oocytes had undergone GVBD as determined visually by the appearance of a white spot at the center of the animal hemisphere+ GVBD occurred within 3-4 h after progesterone addition+…”

Dramatic Increase of the RNA Editing for Glutamate Receptor Subunits During Terminal Differentiation of Clonal Human Neurons

“…Indeed, the presence of inhibitors of DRADA in the form of nonsequence-specific dsRNA binding proteins has been reported in Xenopus oocytes and also in silk moth embryos. These inhibitors may regulate the action of DRADA developmentally by suhcellular compartmentalization (Skeiky and Iatrou, 1991;Saccomanno and Bass, 1994). Alternatively, this negative regulator could be a dsRNA for which DRADA enzymes have a similar affinity as f'or GluR RNAs.…”

Editor meets silencer: crosstalk between RNA editing and RNA interference

“…In support of this hypothesis, A→I RNA editing of repeat RNAs occurs only at low levels in ovaries and testes 8 . Furthermore, the nuclear versus cytoplasmic localization and activation of ADARs are regulated during maturation of oocytes and early embryos of X. laevis 93 .…”