The EMBO Journal Vol.16 No.7 pp.1751-1758, 1997
RNA-RNA interaction is required for the formation of specific bicoid mRNA 3' UTR-STAUFEN ribonucleoprotein particles
Dominique Ferrandon1, Iris Koch, Eric Westhof2 and Christiane NUsslein-Volhard
Max-Planck-Institut fiir Entwicklungsbiologie, Spemannstrasse 35/III, D-72076 Tiibingen, Germany and 2Institut de Biologie Moieculaire et Cellulaire, 15, rue Rene Descartes, F-67084 Strasbourg Cedex, France
'Present address: Institut de Biologie Moleculaire et Cellulaire, 15, rue Rene Descartes, F-67084 Strasbourg Cedex, France
The formation of the anterior pattern of the Drosophila embryo is dependent on the localization of the mRNA of the morphogen Bicoid (bed) to the anterior pole of the egg cell. Staufen protein (STAU) is required in a late step of the localization to anchor the bed mRNA in the anterior cytoplasm. We have shown previously that endogenous STAU associates specifically with injected bed mRNA 3'-untranslated region (UTR), resulting in the formation of characteristic RNA-protein particles that are transported along micro-tubules of the mitotic spindles in a directed manner. The regions recognized by STAU in this in vivo assay are predicted to form three stem-loop structures involving large double-stranded stretches. Here, we show that the STAU interaction requires a double-stranded conformation of the stems within the RNA localization signal. In addition, base pairing between two single-stranded loops plays a major role in particle formation. This loop-loop interaction is intermolecular, not intramolecular; thus dimers or multimers of the RNA localization signal must be associated with STAU in these particles. The bed mRNA 3' UTR can also
dimerize in vitro in the absence of STAU. Thus, in addition to RNA-protein interactions, RNA-RNA interaction might be involved in the formation of ribonucleoprotein particles for transport and localization.
Keywords: Drosophila/dsRNA/RNA dimerization/RNA localization and transport/RNA-protein interactions
Introduction
In Drosophila, two RNAs localized at opposite poles of the egg determine the antero-posterior polarity of the embryo, the bicoid (bed) mRNA anteriorly, and oskar (osk) mRNA posteriorly (St Johnston and Nusslein-Volhard, 1992). These localized RNAs provide sources of protein gradients that ultimately control the transcription of zygotic target genes in a concentration-dependent manner. Localization of RNA is also observed in other germ cells such as Xenopus eggs (reviewed in St Johnston, 1995). Although the subcellular localization of mRNAs is best understood in these systems, it is not limited to germ cells
and is probably used as a general mechanism for protein targeting in many polarized cell types (St Johnston, 1995).
The localization of bcd mRNA is a multistep process that requires three known genes (Frohnhofer and Niisslein-Volhard, 1987; Berleth et al., 1988; Stephenson et al., 1988; St Johnston et al., 1989). Mutations in exuperantia and swallow disrupt the localization of bcd mRNA to the anterior cortex of the oocyte during oogenesis. This premature mislocalization leads to an almost homogeneous distribution (exuperantia), or to a shallow gradient (swallow)ofbcd mRNA in the embryo. In embryos from staufen females, bed mRNA is distributed in a steep anterior gradient, indicating that STAU is required after egg activation to anchor the bcd mRNA, after its release from the cortex, in the anterior cytoplasm of the egg (St Johnston etal., 1991). STAU contains five double-stranded (ds) RNA-binding motifs common to a family of dsRNA-binding proteins (Green and Mathews, 1992; St Johnston et al., 1992; Gatignol et al., 1993; Bass et al., 1994; Gibson and Thompson, 1994; Kim et al., 1994); a peptide containing one STAU dsRNA-binding domain has been shown to bind dsRNA in vitro, albeit without sequence specificity (St Johnston et al., 1992). STAU is present in excess throughout the Drosophila egg but is associated specifically with bed mRNA at the anterior pole. Endogenous STAU also associates with bcd mRNA 3' UTR when injected into the early embryo, resulting in the formation of characteristic RNA-protein particles. Although these particles remain at the site of injection, they associate with and migrate along astral microtubules during mitosis and thus form bipolar patterns that can be visualized at metaphase with the STAU antibody (Ferrandon et al., 1994). These observations led us to propose a simple model in which the binding of STAU to its cognate RNA induces a conformational change in STAU that allows it to associate, either directly or indirectly, with a motor that transports the RNA complex.
The cis-acting sequences required for all steps of the localization process have been mapped by transgenic analysis to a 625 nucleotide region in the 3' UTR of the bcd mRNA that is predicted to form a complex secondary structure (Macdonald and Struhl, 1988; Macdonald, 1990; Seeger and Kaufman, 1990; Ephrussi and Lehmann, 1992; Gavis and Lehmann, 1992; Macdonald et al., 1993). The transgenic approach has been successful in identifying elements involved in the localization pathway. However, as most mutations already disrupt the early steps in the localization pathway, the transgenes are not suitable for the analysis of later events such as the interaction with STAU (Macdonald et al., 1993; D.Ferrandon and C.Niisslein-Volhard, in preparation). To circumvent this problem, we have developed an in vivo assay based on the injection of in vitro synthesized transcripts into the embryo and monitoring their ability to recruit STAU to form RNA-
© Oxford University Press
1751
D.Ferrandon et al.
protein particles that are transported along microtubules (Ferrandon et al., 1994). This assay appears faithfully to reproduce the specificity of the interaction of STAU with endogenous RNA since only bed mRNA 3' UTR shows a strong recruitment of the protein, while a number of control RNAs fail to do so. Further, bed mRNA 3' UTR
particles are not formed in mutant embryos lacking the STAU protein. Using a linker-scanning strategy, we mapped the regions recognized by STAU within the bed mRNA 3' UTR. They are predicted to form three noncontiguous stem-loop structures involving long double-stranded regions (Ferrandon et al., 1994). Here, we show that STAU binding requires the double-stranded conformation of the stems within the RNA localization signal. In addition, base pairing between two single-stranded loops plays a major role in transport particle formation. We show that this loop-loop interaction is intermolecular, and not intramolecular, indicating that dimers or multimers of the RNA localization signal associate with STAU. Thus, the formation of large nucleoprotein transport granules may be dependent on RNA-RNA interactions as well as RNA-protein interactions.
Results
The secondary structure of the bed mRNA 3 UTR is required for its specific association with STAU
The regions within the bed mRNA 3' UTR recognized by STAU were mapped using a linker-scanning strategy (Ferrandon et al., 1994). They are predicted (Zuker, 1989) to form three stem-loop structures involving large double-stranded regions: stem-loop III, and the distal parts of stem-loops IV (IVc) and V (Vb) (Figure 1). The same structure is predicted for the bed mRNA 3' UTR of other species of Drosophila (Macdonald, 1990; Seeger and Kaufman, 1990). However, the sequence divergence between those species is relatively low; as a result, the phylogenic approach yields too few compensatory base changes to demonstrate convincingly the base pairing structure (data not shown). To test the relevance of the predicted secondary structure to STAU binding, we designed sets of compensatory mutations on both strands of putative helices. We injected the corresponding RNAs into the early embryo and monitored the association with STAU. The sequences of the mutants are presented in Figure 1 and a typical result is shown in Figure 2a. Single-strand mutations DF503 or DF504 in stem III are each predicted to disrupt base pairing in this region; either of these two mutations prevents the formation of STAU-containing particles by the corresponding injected RNA (Figure 2b and c). The introduction of both mutations into the same transcript results in an RNA in which the base pairs in stem IIIb can still form, but with an altered primary sequence. The injection of this double mutant RNA leads to formation of STAU-containing granules (Figure 2d) (Table I). Thus, the double-stranded helix of stem IIIb is required for the interaction between bed mRNA 3' UTR and STAU, but the primary sequence is not important. Similar results were obtained with almost all sets of compensatory mutations within predicted helical stems (Table I). In the case of loop IVc, the double-strand mutation DF508 does not restore the ability to interact with STAU, suggesting a sequence-specific interaction of
this structure with STAU or with a cofactor. Alternatively, the secondary structure prediction of stem-loop IVc, or that of the double mutant, may be incorrect. In summary, in five out of six cases tested, the preservation of the base complementarity is sufficient for specific association with STAU, despite the substantial changes in the primary sequence of the RNA. This suggests that STAU, containing five RNA-binding motifs, directly binds to several double-stranded regions within the bed mRNA 3' UTR and that the double-stranded conformation of the RNA within the stem-loops is essential for this binding.
Loop-loop interactions between two bcd mRNA molecules are required for bcd mRNA 3 UTR-STAU particle formation
Six conserved nucleotides in the distal loop of helix III are perfectly complementary to a side loop of the same helix (Macdonald, 1990) (Figure 1). To examine a possible base pairing between these two loops, we tested a set of compensatory mutations that affect four ofthe nucleotides in each loop (DF495, DF498). RNA containing the double compensatory mutations (DF496) associates specifically with STAU in the embryo, whereas RNAs containing mutations on either single-stranded region do not (Figure 2e-g). These results suggest that the two loops contact each other through base pairing. One possibility is that this pairing occurs within the molecule, thus forming an element of tertiary structure called a pseudoknot (Westhof and Jaeger, 1992). Attempts to model the putative pseudoknot were unsuccessful because it is not possible to bend stem IIIb sufficiently to allow these two single-stranded loops to base-pair with each other. Moreover, the specific in vivo interaction between bed mRNA 3' UTR and STAU is not affected by the DF497 mutation, which rigidities this stem further by removing the two asymmetric bulges that form flexible joints. These observations suggest that the two loops cannot base-pair with each other within the same RNA molecule. To test the possibility that base pairing occurs between the loops of different RNA molecules, we injected embryos with a 1:1 mixture of the single loop mutant RNAs, DF495 and DF498. Although neither mutant RNA can interact with STAU on its own, this mixture recruits STAU into particles that are localized (Figure 2h, Table I). Thus, base pairing between the distal and side loops of stem III of two different molecules appears to be required for STAU binding. Indeed, the formation of tetramers or higher order oligomers may account for these results as well as the formation of dimers.
The bcd mRNA 3 UTR oligomerizes in vitro
To investigate the intermolecular association of the RNA in vitro, we incubated bed mRNA 3' UTR under low or high salt conditions, which respectively hinder or promote base pairing, and separated the RNA by non-denaturing agarose gel electrophoresis (Marquet et al., 1991; Tounekti et al., 1992). Incubation under low salt conditions results in only one species of the bed mRNA 3' UTR, whereas under high salt conditions additional bands are observed (Figure 3a). The lowest band most probably corresponds to the monomeric form, whereas the upper bands are likely to correspond to dimers, trimers and tetramers. Similar results are obtained with the 480 nucleotide long DF525 RNA that represents the minimum STAU-binding
1752
RNA-RNA interactions and bed 3' UTR-STAUFEN particles
Fig. 1. The proposed secondary structure of the bed mRNA 3' UTR and the mutations. The core region of the bed mRNA 3' UTR predicted by the MFOLD program of Zuker (1989) is shown (nucleotides 181-720 of the bed 3' UTR; G of the UAG stop codon is position 0) (Ferrandon et al., 1994). The STAU-binding site is shown in blue (Ferrandon et al., 1994). We designed sets of compensatory mutations on both strands of putative helices (designated by Roman numerals) to test directly for such structures. Each set of mutations and its associated name is color coded: the wild-type nucleotides are colored; the corresponding mutations are shown next to them (in black). The name of single-strand mutations is not underlined whereas the name of the corresponding double compensatory mutant is. The nucleotides altered by mutation DF497 are boxed in red; the corresponding mutant nucleotides are also indicated in red. The two connected arrows show the sequence complementarity of six nucleotides of the distal loop of helix III to six nucleotides of its side loop.
site or with a 210 nucleotide long RNA that forms helix III (RNA III) (Figure 3b). In contrast, when a 180 nucleotide long RNA spanning helices Vb and IVc was used, only one band could be detected under both low and high salt conditions, suggesting that this part of the bed mRNA 3' UTR does not dimerize on its own under our in vitro conditions. To test whether the upper bands on the gel do correspond to multimers of the RNA and not to altered conformations of RNA monomers, we mixed equimolar amounts of bed mRNA 3' UTR and of the short RNA III. Under high salt conditions, this results in a shift of the major band (Figure 3b, lane III+wt), suggesting that this band corresponds to a heterodimer between RNA III and the wild-type RNA. The upper band observed in the mix migrates more slowly than the 3' UTR dimer band and therefore probably represents heteromultimer forms of these two RNAs. These data
indicate that the bed mRNA 3' UTR can oligomerize in vitro and that stem-loop III is involved in the process.
Modeling of the dimer structure reveals a new dimerization motif
The structure of a dimer constituted of two helices III base paired via their side and terminal loops could be successfully assembled and modeled (Figure 4). The looploop interaction contains six central Watson-Crick base pairs (A289-C294 to U834-G829) with two tentative non-Watson-Crick pairs at each extremity (A288...A835, and G295...U828), thus forming a mini-helix. The helical domain between the two interacting loops, helix IIIb, is slightly bent at the large asymmetric bulge and its axis is in rough continuity with that of the loop-loop helix. The helical axis of the supporting helix, the end of helix IIIa, is located between those of the two helices IIIb. The dimer
1753
D.Ferrandon et al.
Fig. 2. The specific interaction of the bed mRNA 3' UTR with STAU in vivo. Synthetic bed mRNA 3' UTR was injected in a small region of cytoplasm of freshly-laid eggs. One hour after injection, the embryos were fixed and stained with a STAU-specific antibody that gives a red signal by indirect immunofluorescence. All RNA batches were encoded before injection and the key was revealed only after scoring of the injected embryos. The position of metaphase plates on the same focal plane is revealed by DAPI (4',6'-diamidinophenyl indole) staining (blue signal). The left-hand side of each panel represents schematically the structure of helix Illb of the injected RNA. Nucleotides whose pairing is tested are shown in green. Mutated nucleotides are shown in red. The disruption of helix Illb induced by mutations DF503 and DF504 is predicted to be much more extensive than schematized here. The STAU-bed mRNA 3' UTR particles observed with the wild-type (wt) RNA have been shown to associate with astral microtubules (Ferrandon et al., 1994). With single-strand mutation RNAs, a non-microtubule-associated signal similar to that produced by the non-specific interaction of STAU with dsRNA was often observed at the injection site, suggesting that STAU recognizes the remaining helices in the structure but is not activated to associate with microtubules (Ferrandon et al., 1994). This signal is not easily observed in this figure since the complex remains at the injection point outside of the focal plane of the pictures.
possesses a 2-fold axis of symmetry located half-way on the line connecting the helical axis of helices Ilia. This model suggests that the helices IIIb may be structurally required to orient the loops for efficient base pairing.
Discussion
By testing mutations which change the primary sequence, as well as double mutations restoring base pairing within the stems of the bed mRNA 3' UTR, we have shown that in most instances the double-stranded conformation and not the primary sequence is crucial for the specific association with the STAU protein. In addition, dimers or oligomers of the RNA must be present in the ribonucleo-protein particles observed in vivo. These findings support the notion that STAU, containing five dsRNA-binding motifs, binds directly to the bed mRNA 3' UTR. We propose that the specificity of this binding is provided by
the spatial correspondence between the dsRNA-binding motifs within STAU and the helices found in oligomers of the bed mRNA 3' UTR. Several proteins contain more than one dsRNA-binding motif, each of which displays little sequence specificity (Green and Mathews, 1992; St Johnston et al., 1992; Gatignol et al., 1993; Bass et al., 1994; Gibson and Thompson, 1994; Kim et al., 1994; Bycroft et al., 1995). However, in a few instances, sequence-specific interactions with mRNAs have been reported (Polson and Bass, 1994; Davis and Watson, 1996). It is possible that a particular RNA conformation is the basis for recognition in these cases. So far, STAU is the only protein with more than three copies of the dsRNA-binding motif, and this might explain its exquisite structural requirements for RNA recognition.
RNA loop-loop interactions have been reported to play an important role, intermolecularly in the control of DNA replication (Tomizawa, 1993; Marino et al., 1995; Predki
1754
RNA-RNA interactions and bed 3' UTR-STAUFEN particles
Table I. Results of the injection experiments
RNAa
Stem-loopb
In vivo STAU-bcd 3'UTRbindingc
DF503 DF504 DF505
DF503 + DF504
DF521
DF522
DF523
DF515
DF516
DF517
DF518
DF519
DF520
DF512
DF513
DF514
DF506
DF507
DF508
DF495
DF498
DF496
DF495 + DF498
DF497
(Illb)
(Illb) (Ilia) (Ilia) (Vb) (IVc)
(IIIb) (IIIb)
+d+ +
+ +/+ + +
++
-/+
++ ++
aInjected RNAs described in Figure 1. Double compensatory mutants are underlined. For DF503 + DF504 (or DF495 + DF498), a one:one mix of each single-strand mutant was injected. bThe number of the stem-loop structure affected by the mutations. cQualitative scoring of the specific formation of STAU-containing particles observed in ~50-150 injected embryos (Ferrandon et al., 1994). Injections where most embryos show a signal similar to that observed in Figure 2a, d and h were scored as + + + .A ++ score was given when, in addition to specific particles, a diffuse STAU background staining was detected in most embryos (Figure 2g). For some constructs, a few particles were detected next to nuclei in <10 embryos and were thus given a -/+ score. Finally, injection series where no STAU punctate signal was observed were scored as -(Figure 2b, c, e, and f).
dStem mutations DF503 or DF504 can only be rescued in cis (DF505), and not in trans (DF503 + DF504), suggesting that the helix is recognized directly by STAU or is required to orient the side and terminal loops so that they can base pair easily.
Fig. 3. The bed mRNA 3' UTR oligomerizes in vitro. Synthetic RNAs were incubated under low (l) or high salt conditions (h) which respectively hinder or favor intermolecular association of the RNA (Marquet et al., 1991). The two samples were then analyzed side by side by non-denaturing agarose gel electrophoresis. (a) wt RNA; several bands are observed under high salt conditions which represent the dimer form (D) (see below) and most likely the trimer (Tr) and tetramer (Te) forms. (b) RNA 525 corresponds to the 480 nucleotide long minimum binding site of STAU on the bcd mRNA. RNA III corresponds to 210 nucleotides of helix III whereas RNA IV+V contains helices Vb and IVc [DF500 and DF501 respectively (Ferrandon et al., 1994)]. Note that dimerization occurs even under low salt conditions for RNAs 525 and III. In the low salt condition lane (l) of the mixing experiment III+ wt, one can distinguish successively the following bands from bottom to top: monomer of RNA III, dimer of RNA III, monomer of wt RNA, and finally a very faint band corresponding to a heterodimer between RNA III and wt RNA. This band which is higher than the wt RNA monomer form is seen much more clearly in the high salt condition lane (h). All of the wt monomer RNA form has been converted to this heterodimer band or to a higher heteromultimer band, and not to a wt dimer RNA form, thus demonstrating that the higher bands observed under high salt conditions are indeed dimer or higher multimer forms.
et al., 1995) and in the dimerization of HIV RNA (Paillart et al, 1994, 1996; Skripkin et al, 1994), and intramolecularly in the autocatalytic group I and group II introns (Michel and Ferat, 1995; Jaeger et al., 1996). In our model, helix III appears to be the main element involved in the intermolecular association. However, other parts of the RNA may be required to stabilize the inter-molecular RNA interaction. Our data indicate that the bed mRNA 3' UTR under high salt conditions can form dimers and higher order oligomers even in the absence of STAU, and that helix III is also involved in dimer formation in vitro. Modeling studies show that trimers or tetramers can be formed easily by base pairing between the two helix III loops. However, for the macroscopic particles that are observed in vivo, the interaction of the RNA with STAU is required. One possibility is that STAU functions as a linker between RNA dimers or short oligomers, either through its five dsRNA-binding motifs or by dimerization itself. Alternatively, STAU could serve to stabilize the RNA-RNA interactions, allowing the formation of multimer chains.
STAU is also required to localize osk mRNA to the
posterior pole of the oocyte (Ephrussi et al., 1991; Kim-Ha et al, 1991; St Johnston et al, 1991). osk mRNA 3' UTR does not show any homology to the bed mRNA 3' UTR (Kim-Ha et al., 1993). In the case of osk, large particles, the polar granules, are formed at the posterior pole of the oocyte that contain, in addition to STAU, a number of identified proteins as well as RNAs (St Johnston, 1993). In contrast to bed mRNA, the interaction of STAU with osk mRNA takes place only during oogenesis; in our injection assay, osk mRNA 3' UTR does not associate with STAU (Ferrandon et al., 1994). This suggests that other proteins not present in the egg cytoplasm are required for the association of osk mRNA with STAU. It will be interesting to see whether the osk mRNA 3' UTR also multimerizes upon association with STAU.
Nucleoprotein particles have been observed in many instances of mRNA transport, both at the optical and electron microscopy level, suggesting that particle formation may play an important role in packaging the RNA for efficient localization (Ainger et al., 1993; Trembleau et al, 1994, 1995; Forristal et al, 1995; Racca et al,
1997). The large particles of bcd mRNA 3' UTR and
1755
D.Ferrandon et al.
Fig. 4. Model of the dimer interaction between two helices III. Right-hand panel: view of a modeled 3D dimer of hairpin Illb and the distal part of helix IIIa interacting via loop-loop interactions (one molecule in green, the other in red). The apical loops of both molecules are in yellow and the complementary side loops in dark blue as shown on the 2D model in the left-hand panel. Helix IIIb is 25 bp long (with two mismatches) and, despite the presence of two asymmetric bulges which can constitute flexible joints, it is not possible to bend helix IIIb sufficiently so that the two side loops would base-pair with each other while maintaining the rest of the secondary structure with a correct geometry and stereochemistry of the RNA. This and the results obtained with the injection of DF497 that rigidifies this stem (Figure 1, Table I) make it unlikely that the wild-type RNA forms a pseudoknot. In addition to the dimer, trimer and tetramer models can also be envisaged.
STAU that are detected in the injected embryos are not observed with the endogenous RNA or with injected full-length transcripts. However, smaller particles are observed with full-length transcripts when the cell cycle lengthens at cycle 14 (Ferrandon et al., 1994). These transcripts contain large open reading frames, and it may be steric hindrance caused by translation or their greater size that prevents the formation of larger particles. As these full-length bed mRNA molecules also recruit STAU, migrate to the cortex and eventually induce the formation of small particles, it seems likely that they interact with STAU in the same way as the 3' UTR alone.
The present results suggest that both RNA-RNA and RNA-protein interactions may provide the basis for the formation of very large ribonucleoprotein particles such as the polar granules of Drosophila (50 times the size of a ribosome) (St Johnston, 1993), P-granules of Caenor-habditis elegans (Seydoux and Fire, 1994) and germinal granules of Xenopus laevis (Kloc et al., 1993; Forristal et al., 1995; Kloc and Etkin, 1995). In conclusion, through its role in mRNA sorting and transport, RNA quaternary structure might turn out to be fundamental to the generation of cell polarity.
Materials and methods
Plasmids and RNA
The mutations were introduced in the bcd mRNA 3' UTR by a two-step PCR procedure as described (Higuchi et al., 1988) using DF400 (Ferrandon et al., 1994) as a template and the following external primers for all mutant constructs (bcd sequences are in capitals): 5' sense primer, ataagaatgcggccgCCTGGACGAGAGGCGTGT (the first nucleotide is nucleotide 1 of the bcd mRNA 3' UTR) and 3' antisense primer,
ccggaattcAAGGGACGGAAATATGGG (the last nucleotide is nucleotide 870 of the bcd mRNA 3' UTR) as well as the following primers for mutagenesis (mutated nucleotides are shown in bold; only the sense primer is indicated since the antisense primers are exactly complementary to the sense): DF503, TATTCGCCTTAGCATCTTCTGGGTGGC; DF504, TGCAACCAGTAGATGTTGAGGCCATTTG; DF506, AACA-TTTGCGCTAAGATTGACCAAGAA; DF507, TTGACCATCTTAA-
GTCAGCAAATTGTA; DF508, ATTTGCTGACTTAAGATGGTCA-ATCTT; DF509, TCAAAATGAAAATACAATTCTCTTGGGCGT;
DF510, CCCTTAAAGATCGTTGTATTAAACAATAATA; DF512, TTCTCTTGGGCCATTACTCATACAAATG; DF513, CTCATACAAT-GTAATGCCTTAAAGATC;   DF515, CCAGTTAACTCTATACAAA-
GGCTGCAATACGC; DF516, GCGTATTCCATCCTTTGTTATCG-TCCC; DF518, TTTCCCTGCAAGCTAGTATTCGCCTTA; DF519, TTTGGGCTTAACTAGCTTCCATGGAAA; DF521, CGCCTTAGAT-GTATGACCATCACTGCTCCACTAAAG; DF522, CCGGGAATA-
TGTAAGGTCTTACATTTGAG.
Double-mutant DNA fragments were obtained by repeating this procedure using one single-strand mutant as a template and the mutagenic primers of the second single-strand mutation. The only exception was DF508, where a specific primer was designed since the stem-loop is too short. After purification and digestion with NotI and KcoRI, the PCR bands were cloned into NotI-KcoRI-cut pBNMB vector (Ferrandon et al., 1994). The mutations were confirmed by sequencing. Some mutant constructs contain an additional G®T mutation at nucleotide 211. This mutation alone has no effect in the injection assay (data not shown). RNAs were prepared as previously described (Ferrandon et al., 1994).
Injection experiments
Experiments were carried out as described (Ferrandon et al., 1994). Pictures were taken with a Zeiss Axiophot microscope. The black and white negatives corresponding to the Cy3 channel (STAU) and the DAPI channel were scanned, assigned false colors, and superimposed upon each other using Adobe Photoshop 3.0.
In vitrodimerizationexperiments
These experiments were carried out as previously described (Marquet et al., 1991). Briefly, RNA dissolved in 8 ml of MilliQ-water (Millipore) was heated for 2 min at 90°C and was chilled on ice. Two ml of the
1756
RNA-RNA interactions and bed 3' UTR-STAUFEN particles
appropriate buffer were then added. For samples incubated under high salt conditions, the tubes were incubated at 37°C and subsequently were chilled on ice. The samples were loaded on 1.1% non-denaturing agarose gels. Electrophoresis was carried out at room temperature in 1X TBM buffer (90 mM Tris-borate pH 8.3, 0.1 mM MgCl2) at 7 V/cm. Gels were then stained in 0.5 mg/ml ethidium bromide.
Incubations were carried out in buffer containing 50 mM sodium cacodylate pH 7.5, and either 40 mM KCl, 0.1 mM MgCl2 (standard monomer buffer: low salt) or 300 mM KCl, 5 mM MgCl2 (standard dimer buffer: high salt). RNA final concentration was usually 0.6 mg/ml. In mixing experiments, equimolar amounts of each RNA were added to a final concentration of 0.6 mg/ml.
RNA modeling
The structure was refined with NUCLIN-NUCLSQ (Westhof et al., 1985) and the drawing made with DRAWNA (Massire et al., 1994).
Acknowledgements
We thank Georg Kalhammer for excellent technical help, the Tubingen Fly group, Anne Ephrussi, Gail Habicht, Jean-Luc Imler, Roland Marquet and Daniel St Johnston for discussions and for critically reading previous versions of the manuscript.
References
Ainger,K., Avossa,D., Morgan,F., Hill,S., Barry,C., Barbarese,E. and Carson,J. (1993) Transport and localization of exogenous myelin basic protein mRNA microinjected into oligodendrocytes. J. Cell Biol., 123, 431-441.
Bass,B.L., Hurst,S.R. and Singer,J.D. (1994) Binding properties of newly identified Xenopus proteins containing dsRNA-binding motifs. Curr. Biol., 4, 301-314.
Berleth,T., Burri,M., Thoma,G., Bopp,D., Richstein,S., Frigerio,G., Noll,M. and Niisslein-Volhard,C. (1988) The role of localization of bicoid RNA in organizing the anterior pattern of the Drosophila embryo. EMBO J., 7, 1749-1756.
Bycroft,M., Grunert,S., Murzin,A.G., Procter,M. and St Johnston,D. (1995) NMR solution structure of a double-stranded RNA-binding domain from Drosophila Staufen protein reveals homology to the N-terminal domain of ribosomal protein S5. EMBO J., 14, 3563-3571.
Davis,S. and Watson,J.C. (1996) In vitro activation of the interferon-induced, double-stranded RNA-dependent protein kinase PKR by RNA from the 3' untranslated regions of human alpha-tropomyosin. Proc. Natl Acad. Sci. USA, 93, 508-513.
Ephrussi,A. and Lehmann,R. (1992) Induction of germ cell formation by oskar. Nature, 358, 387-392.
Ephrussi,A., Dickinson,L.K. and Lehmann,R. (1991) oskar organizes the germ plasm and directs localization of the posterior determinant nanos. Cell, 66, 37-50.
Ferrandon,D., Elphick,L., Nusslein-Volhard,C. and St Johnston,D. (1994) Staufen protein associates with the 3' UTR of bicoid mRNA to form particles that move in a microtubule-dependent manner. Cell, 79,
1221-1232.
Forristal,C., Pondel,M., Chen,L. and King,M.L. (1995) Patterns of localization and cytoskeletal association of two vegetally localized RNAs, Vgl and Xcat-2. Development, 121, 201-208.
Frohnhofer,H.G. and Nusslein-Volhard,C. (1987) Maternal genes required for the anterior localization of bicoid activity in the embryo of Drosophila. Genes Dev., 1, 880-890.
Gatignol,A., Buckler,C. and Jeang,K.T. (1993) Relatedness of an RNA-binding motif in human-immunodeficiency-virus type-1 TAR RNA-binding protein TRBP to human P1/DSI kinase and Drosophila Staufen. Mol. Cell. Biol., 13, 2193-2202.
Gavis,E.R. and Lehmann,R. (1992) Localization of nanos RNA controls embryonic polarity. Cell, 71, 310-313.
Gibson,T.J. and Thompson,J.D. (1994) Detection of dsRNA-binding domains in RNA helicase A and Drosophila maleless: implications for monomeric RNA helicases. Nucleic Acids Res., 22, 2552-2556.
Green,S.R. and Mathews,M.B. (1992) Two RNA-binding motifs in the double-stranded RNA-activated protein-kinase, DAI. Genes Dev. , 6,
2478-2490.
Higuchi,R., Krummel,B. and Saiki,R.K. (1988) A general method of in vitro preparation and specific mutagenesis of DNA fragments: study of protein and DNA interactions. Nucleic Acids Res., 16, 7351-7364.
Jaeger,L., Michel,F. and Westhof,E. (1996) The structure of group I ribozymes. In Eckstein,F. and Lilley,D.M.J. (eds), Nucleic Acids and Molecular Biology. Springer-Verlag, Berlin, pp. 33-51.
Kim,J.L., Wang,Y., Sanford,T., Zeng,Y. and Nishikura,K. (1994) Molecular cloning of cDNA for double-stranded RNA adenosine deaminase, a candidate enzyme for nuclear RNA editing. Proc. Natl Acad. Sci. USA, 91, 11457-11461.
Kim-Ha,J., Smith,J.L. and Macdonald,P.M. (1991) oskar mRNA is localized to the posterior pole of the Drosophila oocyte. Cell, 66, 23-35.
Kim-Ha,J., Webster,P., Smith,J. and Macdonald,P. (1993) Multiple RNA regulatory elements mediate distinct steps in localization of oskar mRNA. Development, 119, 169-178.
Kloc,M. and Etkin,L.D. (1995) Two distinct pathways for the localization of RNAs at the vegetal cortex in Xenopus oocytes. Development, 121,
287-297.
Kloc,M., Spohr,G. and Etkin,L.D. (1993) Translocation of repetitive RNA sequences with the germ plasm in Xenopus oocytes. Science,
262, 1712-1714.
Macdonald,P.M. (1990) bicoid mRNA localization signal: phylogenetic conservation of function and RNA secondary structure. Development, 110, 161-71.
Macdonald,P.M. and Struhl,G. (1988) Cis-acting sequences responsible
for anterior localization of bicoid mRNA in Drosophila embryos.
Nature, 336, 595-598. Macdonald,P.M.,   Kerr,K.,   Smith,J.L.   and  Leask,A.   (1993) RNA
regulatory element Ble1 directs the early steps of bicoid messenger-
RNA localization. Development, 118, 1233-1243. Marino,J.P., Gregorian,R.S.,Jr, Csankovszki,G. and Crothers,D.M. (1995)
Bent helix formation between RNA hairpins with complementary
loops. Science, 268, 1448-1454.
Marquet,R., Baudin,F., Gabus,C., Darlix,J.-L., Mougel,M., Ehresmann,C.
and Ehresmann,B. (1991) Dimerization of human immunodeficiency
virus (type 1) RNA: stimulation by cations and possible mechanism.
Nucleic Acids Res., 19, 2349-2357. Massire,C., Gaspin,C. and Westhof,E. (1994) DRAWNA: a program for
drawing schematic views of nucleic acids. J. Mol. Graphics, 12,
201-206.
Michel,F. and Ferat,J.L. (1995) Structure and activities of group II introns. Annu. Rev. Biochem., 64, 435-461.
Paillart,J.-C., Marquet,R., Skripkin,E., Ehresmann,B. and Ehresmann,C. (1994) Mutational analysis of the bipartite dimer linkage structure of human immunodeficiency virus type 1 genomic RNA. J. Biol. Chem.,
269, 27486-27493.
Paillart,J.-C., Skripkin,E., Ehresmann,B., Ehresmann,C. and Marquet,R. (1996) A loop-loop 'kissing' complex is the essential part of the dimer linkage of genomic HIV-1 RNA. Proc. Natl Acad. USA, 93,
5572-5577.
Polson,A.G. and Bass,B.L. (1994) Preferential selection ofadenosines for modification by double-stranded RNA adenosine deaminase. EMBO J., 13, 5701-5711.
Predki,P.F., Nayak,M., Gottlieb,M.B.C. and Regan,L. (1995) Dissecting RNA-protein interactions: RNA-RNA recognition by Rop. Cell, 80,
41-50.
Racca,C., Gardiol,A. and Triller,A. (1997) Dendritic post-synaptic localization of alpha-glycin receptor subunit mRNAs. J. Neurosci. , 17, in press.
Seeger,M.A. and Kaufman,T.C. (1990) Molecular analysis of the bicoid gene from Drosophila pseudoobscura: identification of conserved domains within coding and noncoding regions of the bicoid mRNA.
EMBO J., 9, 2977-2987.
Seydoux,G. and Fire,A. (1994) Soma-germline asymmetry in the distributions   of embryonic   RNAs   in  Caenorhabditis elegans.
Development, 120, 2823-2834.
Skripkin,E., Paillart,J.-C., Marquet,R., Ehresmann,B. and Ehresmann,C. (1994) Identification of the primary site of the human immunodeficiency virus type 1 RNA dimerization in vitro. Proc. Natl
Acad. Sci. USA, 91, 4945-4949.
St Johnston,D. (1993) Pole plasm and the posterior group of genes. In Bate,M. and Martinez-Arias,A. (eds), The Development of Drosophila melanogaster. Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, NY, Vol. I, pp. 325-363.
St Johnston,D. (1995) The intracellular localization of mRNAs. Cell, 81, 161-170.
St Johnston,D. and Nüsslein-Volhard,C. (1992) The origin of pattern and polarity in the Drosophila embryo. Cell, 68, 201-219.
1757
D.Ferrandon et al.
St Johnston,D., Driever,W., Berleth,T., Richstein,S. and Nusslein-Volhard,C. (1989) Multiple steps in the localization of bicoid RNA to the anterior pole of the Drosophila oocyte. Development Suppl., 107, 13-19.
St Johnston,D., Beuchle,D. and Nusslein-Volhard,C. (1991) Staufen,a gene required to localize maternal RNAs in the Drosophila egg. Cell, 66, 51-63.
St Johnston,D.,  Brown,N.H.,  Gall,J.G.  and Jantsch,M.  (1992) A
conserved double-stranded RNA-binding domain. Proc. Natl Acad.
Sci. USA, 89, 10979-10983. Stephenson,E.C., Chao,Y.C.  and Fackenthal,J.D.  (1988) Molecular
analysis of the swallow gene of Drosophila melanogaster. Genes Dev.,
2, 1655-1665.
Tomizawa,J.-I. (1993) Evolution of functional structures of RNA. In Gesteland,R.F. and Atkins,J.F. (eds), The RNA World. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 419-445.
Tounekti,N., Mougel,M., Roy,C., Marquet,C., Darlix,J.-L., Paoletti,J., Ehresmann,B. and Ehresmann,C. (1992) Effect of dimerization on the conformation of the encapsidation Psi domain of Moloney murine leukemia virus RNA. J. Mol. Biol., 223, 205-220.
Trembleau,A., Morales,M. and Bloom,F.E. (1994) Aggregation of vasopressin mRNA in a subset of axonal swellings of the median eminence and posterior pituitary: light and electron microscopy evidence. J. Neurosci., 14, 39-53.
Trembleau,A., Melia,K.R. and Bloom,F.E. (1995) BC1 RNA and vasopressin mRNA in rat neurohypophysis: axonal compartmentaliz-ation and differential regulation during dehydration and rehydration. Eur. J. Neurosci., 7, 2249-2260.
Westhof,E. and Jaeger,L. (1992) RNA pseudoknots. Curr. Opin. Struct.
Biol., 2, 327-333.
Westhof,E., Dumas,P. and Moras,D. (1985) Crystallographic refinement
of yeast aspartic acid transfer RNA. J. Mol. Biol., 184, 119-145. Zuker,M. (1989) On finding all suboptimal foldings of an RNA molecule.
Science, 244, 48-52.
Received on October 1, 1996; revised on November 28, 1996
1758