- Email: [email protected]
Vg1 RNA localization in oocytes in the absence of xVICKZ3 RNA-binding activity
r 2007, Copyright the Authors Differentiation (2007) 75:566–574 DOI: 10.1111/j.1432-0436.2007.00197.x Journal compilation r 2007, International Society of Differentiation
OR IGI N A L A R T IC L E
Kinneret Rand . Joel K. Yisraeli
Vg1 RNA localization in oocytes in the absence of xVICKZ3 RNA-binding activity
Received February 13, 2007; accepted in revised form April 25, 2007
Abstract Vg 1 RNA becomes localized at the vegetal cortex of Xenopus oocytes in a process requiring both intact microtubules (MT) and microfilaments. This localization occurs during a narrow window of oogenesis, when a number of RNA-binding proteins associate with the RNA. xVICKZ3 (Vg1 RBP/Vera), the first Vg1 RNA-binding protein identified, helps mediate the association of Vg1 RNA with MT and is co-localized with the RNA at the vegetal cortex. Given the complexity of the Vg1 RNA ribonucleoprotein (RNP) complex, it has remained unclear how xVICKZ3 functions in Vg1 RNA localization. Here, we have taken a closer look at the process of xVICKZ3 localization in oocytes. We have made use of deletion constructs to perform a structure–function analysis of xVICKZ3. The ability of xVICKZ3–GFP constructs to vegetally localize correlates with their association to MT but not with Vg1 RNA-binding ability. We find that when the ability of xVICKZ3 to bind Vg1 RNA is inhibited by the injection of a construct that dominantly inhibits RNA binding, both the construct and Vg1 RNA still localize, apparently through their continued association with a Vg1 RNA-containing RNP complex. These results emphasize the importance of protein–protein interactions in both xVICKZ3 and Vg1 RNA localization.
. ) Kinneret Rand1 Joel K. Yisraeli (* Hebrew University—Hadassah Medical School Institute for Medical Research Department of Anatomy and Cell Biology Jerusalem, Israel Tel: 1972 2 675 8434 Fax: 1972 2 675 7451 E-mail: [email protected] 1 Present address: Laboratory of Molecular Genetics, NICHD National Institute of Health, Bethesda MD, USA. E-mail: [email protected]
Key words Vg1 RNA Vg1 RBP/Vera intracellular RNA localization ZBP1 hnRNP I
Introduction VICKZ proteins are a family of RNA-binding proteins highly conserved among vertebrates (Yisraeli, 2005). Three paralogs, also very similar to each other, exist in humans, mice, and chicken, while only one paralog has been found in Xenopus (Git and Standart, 2002). In addition to playing roles in RNA stability and translational control (Doyle et al., 1998; Nielsen et al., 1999; Huttelmaier et al., 2005), these proteins have been shown to undergo intracellular sorting and to co-localize with at least some RNAs that are asymmetrically distributed (Ross et al., 1997; Bassell et al., 1998; Deshler et al., 1998; Havin et al., 1998; Oleynikov and Singer, 2003; Leung et al., 2006; Vikesaa et al., 2006; Yao et al., 2006). In chick embryo fibroblasts, b-actin mRNA, a target of cVICKZ1 (ZBP1), localizes to the leading edge of migrating cells in a process that requires intact microfilaments (Lawrence and Singer, 1986; Sundell and Singer, 1991; Ross et al., 1997). cVICKZ1 is both required for, and co-localizes with, b-actin mRNA at the leading edge, although its localization there is dependent upon binding its target RNA (Oleynikov and Singer, 2003). VICKZ proteins also help target b-actin mRNA to both dendritic spines and axonal growth cones in neurons, although in these cells, microtubules (MTs) are required for this localization (Zhang et al., 2001; Eom et al., 2003; Tiruchinapalli et al., 2003; Leung et al., 2006; Yao et al., 2006). In Xenopus oocytes, xVICKZ3 is co-localized with Vg1 RNA at the vegetal cortex of stage IV–VI oocytes, in a process that requires intact MTs for the translocation, and intact microfilaments for cortical anchoring (Yisraeli et al., 1990; Elisha et al., 1995; Havin et al., 1998). Although it does not contain an obvious
567
MT-binding domain, xVICKZ3 co-precipitates with MTs and helps mediate the association of Vg1 RNA to MTs (Elisha et al., 1995). In addition to xVICKZ3, a number of proteins (VgRBP60/hnRNP I, Prrp, 40LoVe) have been shown to bind Vg1 RNA in its 3 0 UTR, within the cis-acting element that mediates vegetal localization (vegetal localization element [VLE]) (Cote et al., 1999; Zhao et al., 2001; Czaplinski et al., 2005). Kinesin I and II, XStau, and Vg1RBP71/KSRP have also been found to be in Vg1 ribonucleoprotein (RNP) complexes, and appear to help mediate Vg1 RNA localization to the vegetal cortex (Kroll et al., 2002; Allison et al., 2004; Betley et al., 2004; Yoon and Mowry, 2004). Given the plethora of proteins that interact with Vg1 RNA, as well as the potential for protein–protein interactions among them (not necessarily dependent on RNA binding (Kress et al., 2004), it has been difficult to assign precise functions to these factors. We have recently described the generation of an xVICKZ3 construct that functions in a dominant negative fashion by forming a heterodimer with wild-type xVICKZ3 and thereby preventing binding to Vg1 and other target RNAs (Oberman et al., 2007). When injected into Xenopus embryos at the two-cell stage, this construct prevents the migration of neural crest cells,
Construct name RRM2-KH1-3 (283-1455) RRM2-KH12 (283-1131) RRM2-KH1-4 (283-1734) KH2-4 (814-1734) KH4 (1435-1734) KH34 (1114-1734) xVICKZ3
function analysis of the protein. Through the use of chimeric xVICKZ3–GFP deletion constructs, we find that only those constructs that associate with MTs are capable of localizing to the vegetal cortex, irrespective of their ability to bind target RNA. Making use of the dominant negative xVICKZ3 construct (xVICKZ3D), we find that Vg1 RNA localization does not require xVICKZ3 binding. Nevertheless, xVICKZ3D is still localized to the vegetal cortex and remains in a complex with Vg1 RNA, apparently as a result of protein–protein interactions. These results thus emphasize the importance of both RNA–protein and protein–protein interactions in Vg1 RNA localization.
Methods Plasmids and cloning xVICKZ3–GFP truncation variants or full-length xVICKZ3 were amplified from clone D (Havin et al., 1998) using a combination of the upstream and downstream oligonucleotide primers listed below (all oligonucleotides are listed in the 5 0 –3 0 direction). The amplified full-length xVICKZ3 ORF (Accession number AF064634) was cloned into pET21d (Novagen, San Diego, CA) with a C-terminal fusion of a 6xHis tag epitope. The xVICKZ3 fragments were cloned into pSP73 (Promega, Madison, WI) containing an A23C30 sequence, with a C-terminal fusion of a GFP.
Forward primer CTCGAGACCATGGAGGTACTGGACAGCC CTCGAGACCATGGAGGTACTGGACAGCC CTCGAGACCATGGAGGTACTGGACAGCC CTCGAGGCTAGCACCATGGAGATCATGCAGAAGGAAGC CTCGAGGCTAGCACCATGGCTCAAGGAAGGATCTATGG CTCGAGGCTAGCACCATGGCACATTTGATTCCTGG CATGCCATGGACAAGCTGTATATT
phenocopying the effect of xVICKZ3 knock-down in embryos via antisense morpholino oligonucleotide injection (Yaniv et al., 2003). The homologous human VICKZ construct similarly functions to dominantly inhibit RNA binding of endogenous hVICKZ proteins and concomitantly reduces cell migration in human prostate carcinoma cells. These results indicate that in both Xenopus embryos and human carcinoma cells, VICKZ protein function is mediated via the ability to bind target RNAs. To explore the role of xVICKZ3 in Vg1 RNA localization in oocytes, we have undertaken a structure–
Reverse primer AAGCTTGCTAGCACCATAGATCCTTCCTTGAGC AAGCTTGCTAGCACCAGGAATCAAATGTGC AAGCTTCTGCTGCTGTCTTCTTACC AAGCTTCTGCTGCTGTCTTCTTACC AAGCTTCTGCTGCTGTCTTCTTACC AAGCTTCTGCTGCTGTCTTCTTACC ACCGCTCGAGTTTTCTTCTCGGTTG
xVICKZ3D and xVICKZ3D–GFP were constructed by deleting the entire KH4 domain minus the last a helix in order to create a recombinant protein missing all amino acids between His504 and Gly554, inclusive (Oberman et al., 2007) (termed xVICKZ3D4 and xVICKZ3D4–GFP in that paper).
Xenopus oocytes, microinjection, protein extracts, and MT precipitation Stage III or early stage IV oocytes were manually isolated from ovaries of Xenopus laevis, and microinjected with 1 ng of RNA in 10 nl of DEPC-treated water. Injected oocytes were cultured for 4 days in oocyte culture media containing 10% of frog serum
568 containing vitellogenin, as previously described (Havin et al., 1998). In order to compare localization results of the different constructs, the percent of oocytes showing localization for each construct was normalized to that of xVICKZ3, in the same trial, using a metaanalysis approach (Comprehensive Meta-analysis, 1.0.17), which uses the inverse of the variance of the estimate provided by each participating trial for the weights (Cooper and Hedges, 1994). Protein extracts were prepared from Xenopus oocytes by homogenizing defolliculated oocytes in TGKED buffer and spinning at 100,000 g, as described (Elisha et al., 1995). MT extracts and precipitations were performed as described (Elisha et al., 1995), using 40 mM Taxol and 1 mM guanosine triphosphate (GTP).
Whole-mount immunostaining Oocytes were fixed in MEMFA (0.1 M MOPS, pH 7.4, 2 mM ethylene glycol tetraacetic acid [EGTA], 1 mM MgSO4, and 3.7% Formaldehyde) for 30 min, blocked and immunostained overnight at 41C with either rabbit anti-VICKZ antibody (Zhang et al., 1999) at a 1:100 dilution, or rabbit anti-GFP (Invitrogen, Carlsbad, CA) at a 1:50 dilution. After washing, the oocytes were incubated with affinity-purified goat anti-rabbit IgG conjugated to rhodamine (Jackson Immunoresearch, West Grove, PA) at a 1:100 dilution. Oocytes were cleared with 2:1 BB/BA (benzyl benzoate/benzyl alcohol).
Immunoprecipitation and Western blot analysis Immunoprecipitation was performed by incubating, overnight, 2 mg anti-VICKZ with 30 ml Protein A-Sepharose beads (Amersham, Piscataway, NJ) in immunoprecipitation buffer (150 mM NaCl, 10 mM Tris, pH 7, 1 mM EDTA, 1 mM EGTA, 0.5% NP-40 and 1 Protease Inhibitor Cocktail). After washing the beads in the above buffer, equivalent amounts of oocyte extracts were added and incubated with the beads. After a second set of washing, the beads were resuspended in Laemmli dye, and run on a 10% SDSPAGE gel for Western blotting with the anti-VICKZ antibody (1:20,000 dilution) or anti-hnRNP I (a gift from Dr. Kim Mowry). For Western blotting, proteins were resolved using SDS-PAGE and blotted onto Protran BA 85 Nitrocellulosenitrate(E) membranes (Whatman, Atlanta, GA). Membranes were blocked in 5% dry milk. xVICKZ3 and xVICKZ3D were detected with rabbit polyclonal antisera at a 1:20,000 dilution, and xVICKZ3–GFP constructs were detected by rabbit anti-GFP polyclonal antisera at a 1:5000 dilution, in 0.5% dry milk, followed by HRP-coupled goat anti-rabbit antibody (Jackson Immunoresearch), at a 1:20,000 dilution. Reverse-transcriptase polymerase chain reactions (RT-PCRs) of immunoprecipitated RNA were performed as described in the Reverse-iT kit (AB Gene, Rochester, NY).
In vitro RNA synthesis and UV crosslinking RNAs for microinjection were transcribed using the Cap Scribe kit (Roche, Indianapolis, IN), and purified on Sephadex G-50 columns (Sigma, St. Louis, MO). VLE–fluorescein-labeled RNA was transcribed in a 20 ml reaction in the presence of 1 transcription buffer, rNTP mix (1 mM rATP, 1 mM rGTP, 1 mM rCTP, 0.7 mM UTP), 0.5 mM UTP–fluorescein, 10 mM DTT, 40 u RNasin, 1 mM Cap-GTP, and RNA polymerase. UV crosslinking was performed with a radioactive VLE probe, as described (Havin et al., 1998).
Results xVICKZ3 contains redundant localization sequences To begin to decipher how xVICKZ3 is localized to the vegetal cortex of stage III/IV oocytes, we made use of an xVICKZ3–GFP fusion that rescues embryos in which xVICKZ3 levels have been reduced by antisense morpholino oligonucleotides (AMO) (Yaniv et al., 2003). In stage III/IV oocytes, this construct is fully functional as well, as demonstrated by its ability to localize, upon injection and oocyte culture, to the vegetal cortex in a pattern identical to endogenous xVICKZ3 (Fig. 1). The vertebrate VICKZ proteins all contain two RNA recognition motifs (RRMs) at their N-terminus, and four hnRNA K-homology (KH) motifs at the C-terminus. Based on theoretical and experimental data, it has been suggested that the KH motifs are organized in VICKZ proteins into domain pairs, or didomains, consisting of KH1 and 2, and KH3 and 4 (Git and Standart, 2002). A series of deletions were generated in the xVICKZ3–GFP construct that removed various domains from the protein, and the ability of these fusion proteins to undergo localization was assessed when RNAs encoding them were injected into stage III/IV oocytes (under culturing conditions that induce Vg1 RNA and xVICKZ3 localization to the vegetal cortex). The distribution of several of the constructs is shown in Fig. 1, and the complete list of constructs, with their relative localization efficiencies calculated using a metaanalysis approach, is shown in Fig. 3. These results indicate that there is no one region that solely confers vegetal localization ability upon the chimeric protein; instead, there appears to be a redundancy in the localization signals present. Thus, both RRM2-KH12 and KH34 are each sufficient to mediate localization, although there is no overlap between their sequences; no statistically significant differences in the ability to confer vegetal localization were found between the two regions. The RRM domains are completely dispensable for localization. MT binding required for localization of xVICKZ3 constructs Vg1 RNA localization in stage III/IV oocytes requires the presence of intact MTs (Yisraeli et al., 1990). Furthermore, xVICKZ3 has been shown to precipitate with MTs and mediate the specific association of Vg1 RNA with MTs reconstituted in vitro (Elisha et al., 1995). To determine which sequences in xVICKZ3 mediate MT association, the same constructs used above to identify localization elements were evaluated for their ability to associate with MTs in vivo. MTs and associated proteins were precipitated from oocytes injected with RNAs encoding the various constructs, and the distribution of the chimeric proteins to either the pellet or supernatant
569 Fig. 1 Localization of endogenous xVICKZ3 and xVICKZ3–GFP chimeras to the vegetal cortex of Xenopus oocytes. Capped RNA encoding several xVICKZ3–GFP chimeras (see Fig. 2) was injected into stage III/IV oocytes and cultured for 4 days. Injected and control, uninjected oocytes were stained with either anti-GFP (xVICKZ3, RRM2KH1-4, KH2-4, KH34, and KH4) or anti-VICKZ (endogenous xVICKZ3) antibody and visualized using confocal microscopy. Representative pictures from groups of oocytes are shown. Injected GFP tagged xVICKZ3 localizes at the vegetal cortex in the same pattern as endogenous xVICKZ3. The RRM2-KH14, KH2-4, and the didomain chimera KH34 are all localized to the vegetal cortex in a similar fashion as well. In contrast, KH4 is expressed in a diffuse homogenous pattern throughout the oocyte. The vegetal cortex is oriented downwards. GFP, green fluorescent protein.
fraction was compared with that of endogenous xVICKZ3 using Western blot analysis of the electrophoresed samples (Fig. 2). Taxol and GTP stabilize MTs, and MT-associated proteins (such as endogenous xVICKZ3) pellet with polymerized MTs in these extracts. Extracts made without Taxol and GTP, however, contain very few stable MTs; MT-associated proteins, as well as free tubulin, are found in the supernatants, even after high-speed centrifugation. Thus, the ability of Taxol and GTP to cause a protein to pellet from extracts is a strong indication that the protein is associated with MTs. A strong correlation is observed between the ability of a construct to associate with MTs and its ability to localize to the vegetal cortex of Xenopus oocytes (sum-
marized in Fig. 3). Constructs that show no ability to localize (such as KH4 or GFP alone) are found predominantly in the supernatant fractions, regardless of the presence or absence of Taxol and GTP. As seen with the ability to localize to the vegetal cortex, there is a redundancy in the sequences that mediate MT association, with a KH didomain appearing to be a minimal requirement for association. These results, in conjunction with the localization results above, suggest that MT association predicts the ability of an xVICKZ3 construct to localize.
xVICKZ3 RNA binding is not required for VLE RNA localization in oocytes As discussed above, a number of RNA-binding proteins in oocytes have been shown to bind to Vg1 RNA at the VLE sequence (Deshler et al., 1998; Havin et al., 1998;
Fig. 2 MT association assay. Stage III/IV oocytes were injected with full-length xVICKZ3–GFP (A), RRM2-KH1-3–GFP (B), KH4–GFP (C), or GFP alone (D), and cultured for 4 days. MTs were allowed to polymerize from oocyte extracts in the absence ( ) or presence (1) of taxol and guanosine triphosphate (which enhance and stabilize MT polymerization) and were then centrifuged. The proteins in the pellet (P) were compared with those in the supernatant (S) by Western blot analysis, using either anti-VICKZ (A and B) or anti-GFP (C and D) antisera. Preferential association with MTs is observed for endogenous xVICKZ3, xVICKZ3–GFP, and RRM2-KH1-3–GFP, but not for KH4–GFP and GFP alone. A full list of all the constructs tested and their relative associations with MTs is shown in Fig. 3. GFP, green fluorescent protein; MT, microtubules.
570
Fig. 3 Vegetal localization and MT association of the xVICKZ3– GFP chimeric constructs. A schematic representation of the different domains (white rectangles) of full-length xVICKZ3 and the GFP chimeras is shown (to scale). GFP (gray rectangle, not shown to scale) is fused to the C 0 terminus. The localization scores shown for each construct represent the percent of oocytes with localized RNA normalized to the number showing localization with fulllength xVICKZ3 (in the same injection experiment) using metaanalysis (see Methods). For samples with p-values o0.05, the
p-value is listed. The relative association of each protein with MTs was determined by the fraction of the protein associating, in the presence of taxol and guanosine triphosphate, with the MT pellet as compared with the supernatant (see Fig. 2 for some examples). RRM, RNA recognition motif; KH, hnRNP K homology domain. 111, strongest MT association; , no MT association (above background). GFP, green fluorescent protein; MT, microtubules; RNP, ribonucleoprotein.
Cote et al., 1999; Zhao et al., 2001; Czaplinski et al., 2005). Given the apparently complex nature of these RNA–protein interactions (Kress et al., 2004), it has been difficult to determine the role of any one specific protein in the localization process. Overexpression of a dominant negative construct of xVICKZ3, lacking most of the KH4 domain (termed xVICKZ3D) prevents xVICKZ3-mediated cell migration in vivo in embryos through the formation of a heterodimer with endogenous xVICKZ3. The heterodimer is inactive in UVcrosslinking not only to VLE RNA but other RNAs as well, such as vegetally localized TGFb-5 mRNA (Oberman et al., 2007). We decided to use the same construct to assay whether the RNA-binding activity of xVICKZ3 is required for the process of VLE RNA localization to the vegetal cortex of oocytes. To test whether xVICKZ3D would also inhibit endogenous xVICKZ3 RNA binding when expressed in oocytes, RNA encoding xVICKZ3D was injected into stage III/IV oocytes, incubated for 5 days, and then protein extracts were used in UV crosslinking assays. As seen in Fig. 4A, expression of exogenous xVICKZ3D was essentially equimolar to that of the endogenous xVICKZ3 and did not reduce endogenous xVICKZ3 levels relative to uninjected, control oocytes. The presence of the deleted construct in the oocytes, however, caused essentially a complete loss of the UV-crosslinking activity of the endogenous xVICKZ3. Notably,
the ability of xVICKZ3D to associate with MTs remained the same as full-length xVICKZ3 (Fig. 4B). Because of the heterogeneous nature of Xenopus oocytes, even those of similar diameter often show a wide range in the degree of endogenous Vg1 RNA localization (Schwartz, 1996). For this reason, we chose to assay the effect of xVICKZ3D on exogenous VLE RNA localization. Thus, stage III/IV oocytes were first injected with RNA encoding xVICKZ3D and incubated overnight in medium to allow for accumulation of the exogenous protein. Fluorescein-labeled VLE RNA was then injected into these oocytes and allowed to undergo localization for an additional 4 days. As seen in Fig. 4C, the presence of the xVICKZ3D construct had no effect on the ability of VLE RNA to undergo localization, even though the presence of this protein completely inhibited the ability of xVICKZ3 to UV-crosslink VLE RNA. Thus, the RNA-binding activity of xVICKZ3, even though required in embryos and in somatic cells for cell migration, is not required in oocytes for VLE RNA localization.
xVICKZ3D localizes to the vegetal cortex as part of an RNP complex In oocytes injected with the xVICKZ3D construct, the ability of VLE RNA to localize to the vegetal cortex
571
suggested that either xVICKZ3 is not required for RNA localization, or, alternatively, that it need not bind the RNA in order to mediate vegetal RNA localization. The latter option would imply that xVICKZ3D remains part of the VLE RNP complex that translocates to the vegetal cortex, even while not binding RNA directly. Work on the chick cVICKZ1 homolog, ZBP1, however, has shown that when RNA binding is inhibited (either through deletions in ZBP1 or by antisense DNA oligonucleotides directed against the zipcode localization elements of b-actin RNA), ZBP1 does not localize to lamellopodia (Farina et al., 2003; Oleynikov and Singer, 2003). To begin to understand how VLE RNA is correctly localized in the absence of normal xVICKZ3 binding, we first analyzed whether xVICKZ3D localizes to the vegetal cortex in oocytes. In contrast to what was observed with ZBP1, RNA binding appears not to be required for xVICKZ3 vegetal localization. xVICKZ3D– GFP, despite its inability to bind VLE RNA sequences, undergoes localization as well as the full-length construct (Fig. 4C). Because of the potential of protein– protein interactions among the different VLE-binding proteins, we wondered whether disrupting xVICKZ3 binding to VLE RNA was enough to dissociate the protein from the VLE RNP complex. xVICKZ3-containing complexes were immunoprecipitated (IP 0 d) from xVICKZ3D-injected oocytes using anti-VICKZ antibody bound to beads, under conditions that preserved RNA–protein and protein–protein interactions. Associated RNAs were assayed by RT-PCR (Fig. 4D). De-
Fig. 4 VLE RNA localization in oocytes is not dependent on its binding to xVICKZ3 but rather on its association with the VLE RNP complex. (A) Stage III/IV oocytes were injected with xVICKZ3D and cultured for five days. Proteins were extracted from uninjected (control oocyte) or xVICKZ3D-injected (xVICKZ3D injected oocyte) oocytes and used for both Western blot analysis with an anti-VICKZ antibody (top panel, one oocyte equivalent/lane) and UV crosslinking assays with a radioactive VLE probe (bottom panel, two oocyte equivalents/lane). Because xVICKZ3D is 51 amino acids shorter than xVICKZ3, it can be observed running slightly ahead of the full-length endogenous protein on the Western blot, where it is expressed in approximately equimolar amounts as compared with the endogenous protein (in both control and injected oocytes). (B) In parallel to the experiments in A, the ability of injected xVICKZ3D to associate with MTs in oocytes was assayed by a microtubules pull down, as described in Fig. 2. xVICKZ3D appears to behave just like the endogenous xVICKZ3, both in the absence and presence of Taxol and guanosine triphosphate. (C) Stage III/IV oocytes were injected with either fluoresceinated VLE RNA (VLE-Fl) alone or first with xVICKZ3D–GFP RNA and then, 24 hr later, fluoresceinated VLE RNA (VLE-Fl1xVICKZ3D). After VLE RNA injection, oocytes were cultured for 4 days. Localization of the (exogenous) VLE RNA was visualized through the use of an anti-fluorescine antibody (VLE RNA), while xVICKZ3D–GFP localization was visualized
with an anti-GFP antibody (xVICKZ3D). As seen in the left two oocytes, VLE RNA is capable of vegetal localization, whether or not xVICKZ3D protein is present in the cell. Although this truncated protein inhibits binding between endogenous xVICKZ3 and VLE RNA (A), and does not bind VLE RNA itself (Oberman et al., 2007), xVICKZ3D–GFP undergoes vegetal localization as well (right oocyte). (D) xVICKZ3D RNA was injected into stage III/IV oocytes alone (xVICKZ3D) or with VLE RNA (xVICKZ3D1VLE RNA) and cultured for 5 days. Protein extracts from injected and uninjected oocytes were immunoprecipitated (IP) with either antiVICKZ antibody (aVICKZ) or preimmune serum (PI), and RNAs were isolated and used for RT-PCR in order to determine the RNAs associated with xVICKZ3 RNP complexes. Endogenous Vg1 RNA (left lane) and VLE RNA (second lane from the left) remain part of the xVICKZ3 complex, even in the presence of xVICKZ3D and are not pulled down with preimmune serum. Vg1 RNA is also part of the xVICKZ3 complex in uninjected oocytes (right lane). (E) Stage III/IV oocytes were injected with xVICKZ3D–GFP and cultured for 4 days. Protein extracts from the injected oocytes were immunoprecipitated with either antiVICKZ antibody (aVICKZ) or preimmune serum (Preimmune) and Western blotted with anti-VICKZ antibody (aVICKZ). The filter was then stripped and incubated with an anti-hnRNP I antibody (ahnRNP I). A band of 60 kDa, clearly present in the anti-VICKZ IP lane and absent in the Preimmune lane, is observed when the stripped filter is incubated with anti-hnRNP I antibody, suggesting that xVICKZ3 and hnRNP I are part of the same RNP complex. GFP, green fluorescent protein; VLE, Vegetal localization element; RNP, ribonucleoprotein.
572
spite the inability of xVICKZ3 in the xVICKZ3D-injected oocytes to crosslink VLE RNA, endogenous Vg1 RNA and exogenous VLE RNA both immunoprecipitate with xVICKZ3, suggesting that they remain part of the same RNP complex even in the absence of direct binding. Vg1 RNA was not detected in samples immunoprecipitated with pre-immune serum, indicating that the RNA does not adventitiously associate with serum or beads during centrifugation. Mowry and colleagues have shown that hnRNP I (also referred to as VgRBP60 or PTB) interacts with xVICKZ3 both in vitro and in vivo, and they have suggested that this association can occur in the absence of RNA (Kress et al., 2004). In accordance with these results, we find that when extracts from oocytes injected with xVICKZ3D are immunoprecipitated with antiVICKZ, hnRNP I is indeed co-precipitated (Fig. 4E). Thus, taken together, these data indicate that VLE RNA can undergo normal localization even in the absence of direct xVICKZ3-VLE RNA binding. Disruption of RNA binding, however, does not cause xVICKZ3 to be completely dislodged from the VLE RNP complex; xVICKZ3 remains indirectly associated with Vg1 RNA.
Discussion We have analyzed the mechanism involved in localizing xVICKZ3 to the vegetal cortex of oocytes through the use of deletion constructs and found that vegetal targeting correlates with the ability of the construct to associate with MTs. Even the ability of the construct to bind RNA is not required, inasmuch as the dominant negative construct, xVICKZ3D, that inhibits wild-type xVICKZ3 RNA binding, also localizes. Consistent with this finding is the fact that the RRM-KH12 construct is also capable of localization, even though it does not dimerize and binds VLE RNA very poorly (Git and Standart, 2002). In oocytes, xVICKZ3 is part of an RNP complex that is assembled first in the nucleus, where it associates with hnRNP I in an interaction that is not mediated by RNA (Kress et al., 2004). In the cytoplasm, where the Vg1 RNP complex is reorganized and additional proteins are recruited, our results suggest that xVICKZ3 is maintained in the complex by protein–protein interactions, even in the absence of its direct RNA binding; thus, both hnRNP I and Vg1 and VLE RNAs are brought down by anti-VICKZ co-immunoprecipitation, even in the presence of xVICKZ3D. Given that the pull down of hnRNP I with antibody directed against xVICKZ3 is prevented in cytoplasmic, but not nuclear, oocyte extracts by RNAse treatment (Kress et al., 2004), these interactions may be stabilized and/or mediated by additional proteins present in the complex. Other proteins known to co-
immunoprecipitate with xVICKZ3 in oocytes include Prrp, XStau, Kinesin I and II, Vg1RBP71/KSRP, and 40LoVe (Zhao et al., 2001; Kroll et al., 2002; Allison et al., 2004; Betley et al., 2004; Yoon and Mowry, 2004; Czaplinski et al., 2005). Given the high degree of conservation among the VICKZ family homologs, it is informative to compare the results of the structure–function analysis of xVICKZ3 in Xenopus oocytes to that of cVICKZ1 in chick embryo fibroblasts (Farina et al., 2003) and hVICKZ1 in mouse NIH 3T3 cells (Nielsen et al., 2002). In both VICKZ1 studies, both KH didomains are required for lamellipodia localization, and neither of the KH didomains alone localized. In the case of cVICKZ1, sequences between the two RRM domains are also required for peripheral localization; in the case of hVICKZ1, the entire N-terminus, including both RRMs and intervening sequences, appears dispensable for localization. In xVICKZ3, however, either KH didomain can mediate, independently, xVICKZ3 localization to the vegetal cortex, and the RRM domains are neither necessary nor sufficient for localization. It is interesting to note that the minimal hVICKZ1 construct found to undergo peripheral localization (KH1-4) was also shown to associate with MTs and associated proteins (Nielsen et al., 2002). In the case of xVICKZ3, the minimal construct that mediates vegetal localization was each of the KH didomains, and each of these in fact associates with MTs and associated proteins. At least some of these differences may reflect the distinct nature of Xenopus oocytes, which can be 50–100 times larger in diameter than cells in culture, and have a radial MT array that extends, in stage III/IV oocytes, from the germinal center to the cortex (Gard, 1991, 1999). The large distances that need to be transversed likely dictate the requirement for an MT-based RNA transport system (as, for example, in neuronal axons), and the rich MT network that develops places a premium on proteins that can associate with it. In addition, transported RNP complexes may well require a more elaborate array of proteins and protein–protein interactions in order to maintain the integrity of the complex; indeed, an abundance of oocyte proteins have been identified as part of the Vg1 RNP complex. In cells in culture, where RNAs are translocated over much shorter distances, transport mechanisms are generally actin based. Although certainly a number of different proteins are present in the cVICKZ1 RNP in chick embryo fibroblasts (Ross et al., 1997; Gu et al., 2002), it will be interesting to see if the requirement of RNA binding for cVICKZ1 lamellipodial localization in these cells (Oleynikov and Singer, 2003), as opposed to what we observe with xVICKZ3 vegetal localization, reflects a less complex RNP structure. The exact role of xVICKZ3 in the Vg1 RNA localization in oocytes remains unclear. Injection of xVICKZ3D into embryos has a clear, and direct, effect
573
on neural crest migration, and essentially phenocopies the effect of knocking down xVICKZ3 expression through the use of AMO (Oberman et al., 2007). The large amount of stable xVICKZ3 protein present in oocytes, however, made AMO knock-down impractical, and hence we decided to make use of the xVICKZ3D construct to inhibit xVICKZ3 activity. In tissue culture cells, VICKZD3 specifically inhibits the RNA-binding activity of endogenous VICKZ3 but not other RNAbinding proteins (Oberman et al., 2007). In oocytes, this construct indeed inhibits xVICKZ3 binding of VLE RNA while not affecting the localization of RNA, demonstrating the dispensability of xVICKZ3 RNA binding in localization. Unlike the AMO knock-down in embryos, however, expression of the dominant negative construct did not lead to loss of the full-length protein from the RNP complex. Thus, we cannot conclude whether VLE RNA localization to the vegetal cortex, in the presence of xVICKZ3D is a result of xVICKZ3’s continued help in directing localization, perhaps by mediating MT association, or if other proteins present in the complex provide the directional information needed for vegetal localization. Kwon et al. (2002) have shown that injection of anti-VICKZ antibody into stage III oocytes reduces localization of both VLE and VegT RNAs by 50%. Nevertheless, their results do not distinguish between a physical hindrance phenomenon, where the presence of bound antibody may affect association of the RNP to either localization or anchoring machinery, or a neutralization effect of xVICKZ3, where the protein is either removed from the complex or effectively prevented from functioning. A clear answer will only be possible once xVICKZ3 can be definitively removed from the complex. Acknowledgments We would like to thank Dr. Froma Oberman for the xVICKZ3D–GFP construct, Dr. Kim Mowry for the antihnRNP I antibody, Dr. Leah Rosen for help with the meta-analysis, and Dr. Mark Tarshish for help with the confocal microscopy. Different aspects of this work were supported by the Israel Cancer Research Fund (ICRF), Israel Science Foundation (ISF), and the Association for International Cancer Research (AICR).
References Allison, R., Czaplinski, K., Git, A., Adegbenro, E., Stennard, F., Houliston, E. and Standart, N. (2004) Two distinct Staufen isoforms in Xenopus are vegetally localized during oogenesis. Rna 10:1751–1763. Bassell, G.J., Zhang, H., Byrd, A.L., Femino, A.M., Singer, R.H., Taneja, K.L., Lifshitz, L.M., Herman, I.M. and Kosik, K.S. (1998) Sorting of beta-actin mRNA and protein to neurites and growth cones in culture. J Neurosci 18:251–265. Betley, J.N., Heinrich, B., Vernos, I., Sardet, C., Prodon, F. and Deshler, J.O. (2004) Kinesin II mediates Vg1 mRNA transport in Xenopus oocytes. Curr Biol 14:219–224. Cooper, H. and Hedges, L. (1994) The handbook of research synthesis. Russell Sage Foundation, New York.
Cote, C.A., Gautreau, D., Denegre, J.M., Kress, T.L., Terry, N.A. and Mowry, K.L. (1999) A Xenopus protein related to hnRNP I has a role in cytoplasmic RNA localization. Mol Cell 4:431–437. Czaplinski, K., Kocher, T., Schelder, M., Segref, A., Wilm, M. and Mattaj, I.W. (2005) Identification of 40LoVe, a Xenopus hnRNP D family protein involved in localizing a TGF-beta-related mRNA during oogenesis. Dev Cell 8:505–515. Deshler, J.O., Highett, M.I., Abramson, T. and Schnapp, B.J. (1998) A highly conserved RNA-binding protein for cytoplasmic mRNA localization in vertebrates. Curr Biol 8:489–496. Doyle, G.A., Betz, N.A., Leeds, P.F., Fleisig, A.J., Prokipcak, R.D. and Ross, J. (1998) The c-myc coding region determinant-binding protein: a member of a family of KH domain RNA-binding proteins. Nucleic Acids Res 26:5036–5044. Elisha, Z., Havin, L., Ringel, I. and Yisraeli, J.K. (1995) Vg1 RNA binding protein mediates the association of Vg1 RNA with microtubules in Xenopus oocytes. EMBO J 14:5109–5114. Eom, T., Antar, L.N., Singer, R.H. and Bassell, G.J. (2003) Localization of a beta-actin messenger zipcode-binding protein modulates the density of filopodial synapses. J Neurosci 23:10433–10444. Farina, K.L., Huttelmaier, S., Musunuru, K., Darnell, R. and Singer, R.H. (2003) Two ZBP1 KH domains facilitate beta-actin mRNA localization, granule formation, and cytoskeletal attachment. J Cell Biol 160:77–87. Gard, D.L. (1991) Organization, nucleation, and acetylation of microtubules in Xenopus laevis oocytes: a study by confocal immunofluorescence microscopy. Dev Biol 143:346–362. Gard, D.L. (1999) Confocal microscopy and 3-D reconstruction of the cytoskeleton of Xenopus oocytes. Microsc Res Tech 44:388– 414. Git, A. and Standart, N. (2002) The KH domains of Xenopus Vg1RBP mediate RNA binding and self-association. RNA 8:1319–1333. Gu, W., Pan, F., Zhang, H., Bassell, G.J. and Singer, R.H. (2002) A predominantly nuclear protein affecting cytoplasmic localization of beta-actin mRNA in fibroblasts and neurons. J Cell Biol 156:41–51. Havin, L., Git, A., Elisha, Z., Oberman, F., Yaniv, K., Schwartz, S.P., Standart, N. and Yisraeli, J.K. (1998) RNA-binding protein conserved in both microtubule- and microfilament-based RNA localization. Genes Dev 12:1593–1598. Huttelmaier, S., Zenklusen, D., Lederer, M., Dictenberg, J., Lorenz, M., Meng, X., Bassell, G.J., Condeelis, J. and Singer, R.H. (2005) Spatial regulation of beta-actin translation by Src-dependent phosphorylation of ZBP1. Nature 438:512–515. Kress, T.L., Yoon, Y.J. and Mowry, K.L. (2004) Nuclear RNP complex assembly initiates cytoplasmic RNA localization. J Cell Biol 165:203–211. Kroll, T.T., Zhao, W.M., Jiang, C. and Huber, P.W. (2002) A homolog of FBP2/KSRP binds to localized mRNAs in Xenopus oocytes. Development 129:5609–5619. Kwon, S., Abramson, T., Munro, T.P., John, C.M., Kohrmann, M. and Schnapp, B.J. (2002) UUCAC- and vera-dependent localization of VegT RNA in Xenopus oocytes. Curr Biol 12: 558–564. Lawrence, J.B. and Singer, R.H. (1986) Intracellular localization of messenger RNAs for cytoskeletal proteins. Cell 45:407–415. Leung, K.M., van Horck, F.P., Lin, A.C., Allison, R., Standart, N. and Holt, C.E. (2006) Asymmetrical beta-actin mRNA translation in growth cones mediates attractive turning to netrin-1. Nat Neurosci 9:1247–1256. Nielsen, F.C., Nielsen, J., Kristensen, M.A., Koch, G. and Christiansen, J. (2002) Cytoplasmic trafficking of IGF-II mRNAbinding protein by conserved KH domains. J Cell Sci 115: 2087–2097. Nielsen, J., Christiansen, J., Lykke-Andersen, J., Johnsen, A.H., Wewer, U.M. and Nielsen, F.C. (1999) A family of insulin-like growth factor II mRNA-binding proteins represses translation in late development. Mol Cell Biol 19:1262–1270.
574 Oberman, F., Rand, K., Maizels, Y., Rubinstein, A. and Yisraeli, J.K. (2007) VICKZ proteins mediate cell migration via their RNA binding activity. RNA (in press). Oleynikov, Y. and Singer, R.H. (2003) Real-time visualization of ZBP1 association with beta-actin mRNA during transcription and localization. Curr Biol 13:199–207. Ross, A.F., Oleynikov, Y., Kislauskis, E.H., Taneja, K.L. and Singer, R.H. (1997) Characterization of a beta-actin mRNA zipcode-binding protein. Mol Cell Biol 17:2158–2165. Schwartz, S.P. (1996) Localization of Vg1—a maternal mRNA. Hebrew University, Jerusalem, Israel. Sundell, C.L. and Singer, R.H. (1991) Requirement of microfilaments in sorting of actin messenger RNA. Science 253:1275–1277. Tiruchinapalli, D.M., Oleynikov, Y., Kelic, S., Shenoy, S.M., Hartley, A., Stanton, P.K., Singer, R.H. and Bassell, G.J. (2003) Activity-dependent trafficking and dynamic localization of zipcode binding protein 1 and beta-actin mRNA in dendrites and spines of hippocampal neurons. J Neurosci 23: 3251–3261. Vikesaa, J., Hansen, T.V., Jonson, L., Borup, R., Wewer, U.M., Christiansen, J. and Nielsen, F.C. (2006) RNA-binding IMPs promote cell adhesion and invadopodia formation. EMBO J 25:1456–1468. Yaniv, K., Fainsod, A., Kalcheim, C. and Yisraeli, J.K. (2003) The RNA-binding protein Vg1 RBP is required for cell migration during early neural development. Development 130: 5649–5661.
Yao, J., Sasaki, Y., Wen, Z., Bassell, G.J. and Zheng, J.Q. (2006) An essential role for beta-actin mRNA localization and translation in Ca21-dependent growth cone guidance. Nat Neurosci 9:1265–1273. Yisraeli, J.K. (2005) VICKZ proteins: a multi-talented family of regulatory RNA-binding proteins. Biol Cell 97:87–96. Yisraeli, J.K., Sokol, S. and Melton, D.A. (1990) A two-step model for the localization of a maternal mRNA in Xenopus oocytes: involvement of microtubules and microfilaments in translocation and anchoring of Vg1 mRNA. Development 108:289–298. Yoon, Y.J. and Mowry, K.L. (2004) Xenopus Staufen is a component of a ribonucleoprotein complex containing Vg1 RNA and kinesin. Development 131:3035–3045. Zhang, H.L., Eom, T., Oleynikov, Y., Shenoy, S.M., Liebelt, D.A., Dictenberg, J.B., Singer, R.H. and Bassell, G.J. (2001) Neurotrophin-induced transport of a beta-actin mRNP complex increases beta-actin levels and stimulates growth cone motility. Neuron 31:261–275. Zhang, Q., Yaniv, K., Oberman, F., Wolke, U., Git, A., Fromer, M., Taylor, W.L., Meyer, D., Standart, N., Raz, E. and Yisraeli, J.K. (1999) Vg1 RBP intracellular distribution and evolutionarily conserved expression at multiple stages during development. Mech Dev 88:101–106. Zhao, W.M., Jiang, C., Kroll, T.T. and Huber, P.W. (2001) A proline-rich protein binds to the localization element of Xenopus Vg1 mRNA and to ligands involved in actin polymerization. EMBO J 20:2315–2325.
OR IGI N A L A R T IC L E
Kinneret Rand . Joel K. Yisraeli
Vg1 RNA localization in oocytes in the absence of xVICKZ3 RNA-binding activity
Received February 13, 2007; accepted in revised form April 25, 2007
Abstract Vg 1 RNA becomes localized at the vegetal cortex of Xenopus oocytes in a process requiring both intact microtubules (MT) and microfilaments. This localization occurs during a narrow window of oogenesis, when a number of RNA-binding proteins associate with the RNA. xVICKZ3 (Vg1 RBP/Vera), the first Vg1 RNA-binding protein identified, helps mediate the association of Vg1 RNA with MT and is co-localized with the RNA at the vegetal cortex. Given the complexity of the Vg1 RNA ribonucleoprotein (RNP) complex, it has remained unclear how xVICKZ3 functions in Vg1 RNA localization. Here, we have taken a closer look at the process of xVICKZ3 localization in oocytes. We have made use of deletion constructs to perform a structure–function analysis of xVICKZ3. The ability of xVICKZ3–GFP constructs to vegetally localize correlates with their association to MT but not with Vg1 RNA-binding ability. We find that when the ability of xVICKZ3 to bind Vg1 RNA is inhibited by the injection of a construct that dominantly inhibits RNA binding, both the construct and Vg1 RNA still localize, apparently through their continued association with a Vg1 RNA-containing RNP complex. These results emphasize the importance of protein–protein interactions in both xVICKZ3 and Vg1 RNA localization.
. ) Kinneret Rand1 Joel K. Yisraeli (* Hebrew University—Hadassah Medical School Institute for Medical Research Department of Anatomy and Cell Biology Jerusalem, Israel Tel: 1972 2 675 8434 Fax: 1972 2 675 7451 E-mail: [email protected] 1 Present address: Laboratory of Molecular Genetics, NICHD National Institute of Health, Bethesda MD, USA. E-mail: [email protected]
Key words Vg1 RNA Vg1 RBP/Vera intracellular RNA localization ZBP1 hnRNP I
Introduction VICKZ proteins are a family of RNA-binding proteins highly conserved among vertebrates (Yisraeli, 2005). Three paralogs, also very similar to each other, exist in humans, mice, and chicken, while only one paralog has been found in Xenopus (Git and Standart, 2002). In addition to playing roles in RNA stability and translational control (Doyle et al., 1998; Nielsen et al., 1999; Huttelmaier et al., 2005), these proteins have been shown to undergo intracellular sorting and to co-localize with at least some RNAs that are asymmetrically distributed (Ross et al., 1997; Bassell et al., 1998; Deshler et al., 1998; Havin et al., 1998; Oleynikov and Singer, 2003; Leung et al., 2006; Vikesaa et al., 2006; Yao et al., 2006). In chick embryo fibroblasts, b-actin mRNA, a target of cVICKZ1 (ZBP1), localizes to the leading edge of migrating cells in a process that requires intact microfilaments (Lawrence and Singer, 1986; Sundell and Singer, 1991; Ross et al., 1997). cVICKZ1 is both required for, and co-localizes with, b-actin mRNA at the leading edge, although its localization there is dependent upon binding its target RNA (Oleynikov and Singer, 2003). VICKZ proteins also help target b-actin mRNA to both dendritic spines and axonal growth cones in neurons, although in these cells, microtubules (MTs) are required for this localization (Zhang et al., 2001; Eom et al., 2003; Tiruchinapalli et al., 2003; Leung et al., 2006; Yao et al., 2006). In Xenopus oocytes, xVICKZ3 is co-localized with Vg1 RNA at the vegetal cortex of stage IV–VI oocytes, in a process that requires intact MTs for the translocation, and intact microfilaments for cortical anchoring (Yisraeli et al., 1990; Elisha et al., 1995; Havin et al., 1998). Although it does not contain an obvious
567
MT-binding domain, xVICKZ3 co-precipitates with MTs and helps mediate the association of Vg1 RNA to MTs (Elisha et al., 1995). In addition to xVICKZ3, a number of proteins (VgRBP60/hnRNP I, Prrp, 40LoVe) have been shown to bind Vg1 RNA in its 3 0 UTR, within the cis-acting element that mediates vegetal localization (vegetal localization element [VLE]) (Cote et al., 1999; Zhao et al., 2001; Czaplinski et al., 2005). Kinesin I and II, XStau, and Vg1RBP71/KSRP have also been found to be in Vg1 ribonucleoprotein (RNP) complexes, and appear to help mediate Vg1 RNA localization to the vegetal cortex (Kroll et al., 2002; Allison et al., 2004; Betley et al., 2004; Yoon and Mowry, 2004). Given the plethora of proteins that interact with Vg1 RNA, as well as the potential for protein–protein interactions among them (not necessarily dependent on RNA binding (Kress et al., 2004), it has been difficult to assign precise functions to these factors. We have recently described the generation of an xVICKZ3 construct that functions in a dominant negative fashion by forming a heterodimer with wild-type xVICKZ3 and thereby preventing binding to Vg1 and other target RNAs (Oberman et al., 2007). When injected into Xenopus embryos at the two-cell stage, this construct prevents the migration of neural crest cells,
Construct name RRM2-KH1-3 (283-1455) RRM2-KH12 (283-1131) RRM2-KH1-4 (283-1734) KH2-4 (814-1734) KH4 (1435-1734) KH34 (1114-1734) xVICKZ3
function analysis of the protein. Through the use of chimeric xVICKZ3–GFP deletion constructs, we find that only those constructs that associate with MTs are capable of localizing to the vegetal cortex, irrespective of their ability to bind target RNA. Making use of the dominant negative xVICKZ3 construct (xVICKZ3D), we find that Vg1 RNA localization does not require xVICKZ3 binding. Nevertheless, xVICKZ3D is still localized to the vegetal cortex and remains in a complex with Vg1 RNA, apparently as a result of protein–protein interactions. These results thus emphasize the importance of both RNA–protein and protein–protein interactions in Vg1 RNA localization.
Methods Plasmids and cloning xVICKZ3–GFP truncation variants or full-length xVICKZ3 were amplified from clone D (Havin et al., 1998) using a combination of the upstream and downstream oligonucleotide primers listed below (all oligonucleotides are listed in the 5 0 –3 0 direction). The amplified full-length xVICKZ3 ORF (Accession number AF064634) was cloned into pET21d (Novagen, San Diego, CA) with a C-terminal fusion of a 6xHis tag epitope. The xVICKZ3 fragments were cloned into pSP73 (Promega, Madison, WI) containing an A23C30 sequence, with a C-terminal fusion of a GFP.
Forward primer CTCGAGACCATGGAGGTACTGGACAGCC CTCGAGACCATGGAGGTACTGGACAGCC CTCGAGACCATGGAGGTACTGGACAGCC CTCGAGGCTAGCACCATGGAGATCATGCAGAAGGAAGC CTCGAGGCTAGCACCATGGCTCAAGGAAGGATCTATGG CTCGAGGCTAGCACCATGGCACATTTGATTCCTGG CATGCCATGGACAAGCTGTATATT
phenocopying the effect of xVICKZ3 knock-down in embryos via antisense morpholino oligonucleotide injection (Yaniv et al., 2003). The homologous human VICKZ construct similarly functions to dominantly inhibit RNA binding of endogenous hVICKZ proteins and concomitantly reduces cell migration in human prostate carcinoma cells. These results indicate that in both Xenopus embryos and human carcinoma cells, VICKZ protein function is mediated via the ability to bind target RNAs. To explore the role of xVICKZ3 in Vg1 RNA localization in oocytes, we have undertaken a structure–
Reverse primer AAGCTTGCTAGCACCATAGATCCTTCCTTGAGC AAGCTTGCTAGCACCAGGAATCAAATGTGC AAGCTTCTGCTGCTGTCTTCTTACC AAGCTTCTGCTGCTGTCTTCTTACC AAGCTTCTGCTGCTGTCTTCTTACC AAGCTTCTGCTGCTGTCTTCTTACC ACCGCTCGAGTTTTCTTCTCGGTTG
xVICKZ3D and xVICKZ3D–GFP were constructed by deleting the entire KH4 domain minus the last a helix in order to create a recombinant protein missing all amino acids between His504 and Gly554, inclusive (Oberman et al., 2007) (termed xVICKZ3D4 and xVICKZ3D4–GFP in that paper).
Xenopus oocytes, microinjection, protein extracts, and MT precipitation Stage III or early stage IV oocytes were manually isolated from ovaries of Xenopus laevis, and microinjected with 1 ng of RNA in 10 nl of DEPC-treated water. Injected oocytes were cultured for 4 days in oocyte culture media containing 10% of frog serum
568 containing vitellogenin, as previously described (Havin et al., 1998). In order to compare localization results of the different constructs, the percent of oocytes showing localization for each construct was normalized to that of xVICKZ3, in the same trial, using a metaanalysis approach (Comprehensive Meta-analysis, 1.0.17), which uses the inverse of the variance of the estimate provided by each participating trial for the weights (Cooper and Hedges, 1994). Protein extracts were prepared from Xenopus oocytes by homogenizing defolliculated oocytes in TGKED buffer and spinning at 100,000 g, as described (Elisha et al., 1995). MT extracts and precipitations were performed as described (Elisha et al., 1995), using 40 mM Taxol and 1 mM guanosine triphosphate (GTP).
Whole-mount immunostaining Oocytes were fixed in MEMFA (0.1 M MOPS, pH 7.4, 2 mM ethylene glycol tetraacetic acid [EGTA], 1 mM MgSO4, and 3.7% Formaldehyde) for 30 min, blocked and immunostained overnight at 41C with either rabbit anti-VICKZ antibody (Zhang et al., 1999) at a 1:100 dilution, or rabbit anti-GFP (Invitrogen, Carlsbad, CA) at a 1:50 dilution. After washing, the oocytes were incubated with affinity-purified goat anti-rabbit IgG conjugated to rhodamine (Jackson Immunoresearch, West Grove, PA) at a 1:100 dilution. Oocytes were cleared with 2:1 BB/BA (benzyl benzoate/benzyl alcohol).
Immunoprecipitation and Western blot analysis Immunoprecipitation was performed by incubating, overnight, 2 mg anti-VICKZ with 30 ml Protein A-Sepharose beads (Amersham, Piscataway, NJ) in immunoprecipitation buffer (150 mM NaCl, 10 mM Tris, pH 7, 1 mM EDTA, 1 mM EGTA, 0.5% NP-40 and 1 Protease Inhibitor Cocktail). After washing the beads in the above buffer, equivalent amounts of oocyte extracts were added and incubated with the beads. After a second set of washing, the beads were resuspended in Laemmli dye, and run on a 10% SDSPAGE gel for Western blotting with the anti-VICKZ antibody (1:20,000 dilution) or anti-hnRNP I (a gift from Dr. Kim Mowry). For Western blotting, proteins were resolved using SDS-PAGE and blotted onto Protran BA 85 Nitrocellulosenitrate(E) membranes (Whatman, Atlanta, GA). Membranes were blocked in 5% dry milk. xVICKZ3 and xVICKZ3D were detected with rabbit polyclonal antisera at a 1:20,000 dilution, and xVICKZ3–GFP constructs were detected by rabbit anti-GFP polyclonal antisera at a 1:5000 dilution, in 0.5% dry milk, followed by HRP-coupled goat anti-rabbit antibody (Jackson Immunoresearch), at a 1:20,000 dilution. Reverse-transcriptase polymerase chain reactions (RT-PCRs) of immunoprecipitated RNA were performed as described in the Reverse-iT kit (AB Gene, Rochester, NY).
In vitro RNA synthesis and UV crosslinking RNAs for microinjection were transcribed using the Cap Scribe kit (Roche, Indianapolis, IN), and purified on Sephadex G-50 columns (Sigma, St. Louis, MO). VLE–fluorescein-labeled RNA was transcribed in a 20 ml reaction in the presence of 1 transcription buffer, rNTP mix (1 mM rATP, 1 mM rGTP, 1 mM rCTP, 0.7 mM UTP), 0.5 mM UTP–fluorescein, 10 mM DTT, 40 u RNasin, 1 mM Cap-GTP, and RNA polymerase. UV crosslinking was performed with a radioactive VLE probe, as described (Havin et al., 1998).
Results xVICKZ3 contains redundant localization sequences To begin to decipher how xVICKZ3 is localized to the vegetal cortex of stage III/IV oocytes, we made use of an xVICKZ3–GFP fusion that rescues embryos in which xVICKZ3 levels have been reduced by antisense morpholino oligonucleotides (AMO) (Yaniv et al., 2003). In stage III/IV oocytes, this construct is fully functional as well, as demonstrated by its ability to localize, upon injection and oocyte culture, to the vegetal cortex in a pattern identical to endogenous xVICKZ3 (Fig. 1). The vertebrate VICKZ proteins all contain two RNA recognition motifs (RRMs) at their N-terminus, and four hnRNA K-homology (KH) motifs at the C-terminus. Based on theoretical and experimental data, it has been suggested that the KH motifs are organized in VICKZ proteins into domain pairs, or didomains, consisting of KH1 and 2, and KH3 and 4 (Git and Standart, 2002). A series of deletions were generated in the xVICKZ3–GFP construct that removed various domains from the protein, and the ability of these fusion proteins to undergo localization was assessed when RNAs encoding them were injected into stage III/IV oocytes (under culturing conditions that induce Vg1 RNA and xVICKZ3 localization to the vegetal cortex). The distribution of several of the constructs is shown in Fig. 1, and the complete list of constructs, with their relative localization efficiencies calculated using a metaanalysis approach, is shown in Fig. 3. These results indicate that there is no one region that solely confers vegetal localization ability upon the chimeric protein; instead, there appears to be a redundancy in the localization signals present. Thus, both RRM2-KH12 and KH34 are each sufficient to mediate localization, although there is no overlap between their sequences; no statistically significant differences in the ability to confer vegetal localization were found between the two regions. The RRM domains are completely dispensable for localization. MT binding required for localization of xVICKZ3 constructs Vg1 RNA localization in stage III/IV oocytes requires the presence of intact MTs (Yisraeli et al., 1990). Furthermore, xVICKZ3 has been shown to precipitate with MTs and mediate the specific association of Vg1 RNA with MTs reconstituted in vitro (Elisha et al., 1995). To determine which sequences in xVICKZ3 mediate MT association, the same constructs used above to identify localization elements were evaluated for their ability to associate with MTs in vivo. MTs and associated proteins were precipitated from oocytes injected with RNAs encoding the various constructs, and the distribution of the chimeric proteins to either the pellet or supernatant
569 Fig. 1 Localization of endogenous xVICKZ3 and xVICKZ3–GFP chimeras to the vegetal cortex of Xenopus oocytes. Capped RNA encoding several xVICKZ3–GFP chimeras (see Fig. 2) was injected into stage III/IV oocytes and cultured for 4 days. Injected and control, uninjected oocytes were stained with either anti-GFP (xVICKZ3, RRM2KH1-4, KH2-4, KH34, and KH4) or anti-VICKZ (endogenous xVICKZ3) antibody and visualized using confocal microscopy. Representative pictures from groups of oocytes are shown. Injected GFP tagged xVICKZ3 localizes at the vegetal cortex in the same pattern as endogenous xVICKZ3. The RRM2-KH14, KH2-4, and the didomain chimera KH34 are all localized to the vegetal cortex in a similar fashion as well. In contrast, KH4 is expressed in a diffuse homogenous pattern throughout the oocyte. The vegetal cortex is oriented downwards. GFP, green fluorescent protein.
fraction was compared with that of endogenous xVICKZ3 using Western blot analysis of the electrophoresed samples (Fig. 2). Taxol and GTP stabilize MTs, and MT-associated proteins (such as endogenous xVICKZ3) pellet with polymerized MTs in these extracts. Extracts made without Taxol and GTP, however, contain very few stable MTs; MT-associated proteins, as well as free tubulin, are found in the supernatants, even after high-speed centrifugation. Thus, the ability of Taxol and GTP to cause a protein to pellet from extracts is a strong indication that the protein is associated with MTs. A strong correlation is observed between the ability of a construct to associate with MTs and its ability to localize to the vegetal cortex of Xenopus oocytes (sum-
marized in Fig. 3). Constructs that show no ability to localize (such as KH4 or GFP alone) are found predominantly in the supernatant fractions, regardless of the presence or absence of Taxol and GTP. As seen with the ability to localize to the vegetal cortex, there is a redundancy in the sequences that mediate MT association, with a KH didomain appearing to be a minimal requirement for association. These results, in conjunction with the localization results above, suggest that MT association predicts the ability of an xVICKZ3 construct to localize.
xVICKZ3 RNA binding is not required for VLE RNA localization in oocytes As discussed above, a number of RNA-binding proteins in oocytes have been shown to bind to Vg1 RNA at the VLE sequence (Deshler et al., 1998; Havin et al., 1998;
Fig. 2 MT association assay. Stage III/IV oocytes were injected with full-length xVICKZ3–GFP (A), RRM2-KH1-3–GFP (B), KH4–GFP (C), or GFP alone (D), and cultured for 4 days. MTs were allowed to polymerize from oocyte extracts in the absence ( ) or presence (1) of taxol and guanosine triphosphate (which enhance and stabilize MT polymerization) and were then centrifuged. The proteins in the pellet (P) were compared with those in the supernatant (S) by Western blot analysis, using either anti-VICKZ (A and B) or anti-GFP (C and D) antisera. Preferential association with MTs is observed for endogenous xVICKZ3, xVICKZ3–GFP, and RRM2-KH1-3–GFP, but not for KH4–GFP and GFP alone. A full list of all the constructs tested and their relative associations with MTs is shown in Fig. 3. GFP, green fluorescent protein; MT, microtubules.
570
Fig. 3 Vegetal localization and MT association of the xVICKZ3– GFP chimeric constructs. A schematic representation of the different domains (white rectangles) of full-length xVICKZ3 and the GFP chimeras is shown (to scale). GFP (gray rectangle, not shown to scale) is fused to the C 0 terminus. The localization scores shown for each construct represent the percent of oocytes with localized RNA normalized to the number showing localization with fulllength xVICKZ3 (in the same injection experiment) using metaanalysis (see Methods). For samples with p-values o0.05, the
p-value is listed. The relative association of each protein with MTs was determined by the fraction of the protein associating, in the presence of taxol and guanosine triphosphate, with the MT pellet as compared with the supernatant (see Fig. 2 for some examples). RRM, RNA recognition motif; KH, hnRNP K homology domain. 111, strongest MT association; , no MT association (above background). GFP, green fluorescent protein; MT, microtubules; RNP, ribonucleoprotein.
Cote et al., 1999; Zhao et al., 2001; Czaplinski et al., 2005). Given the apparently complex nature of these RNA–protein interactions (Kress et al., 2004), it has been difficult to determine the role of any one specific protein in the localization process. Overexpression of a dominant negative construct of xVICKZ3, lacking most of the KH4 domain (termed xVICKZ3D) prevents xVICKZ3-mediated cell migration in vivo in embryos through the formation of a heterodimer with endogenous xVICKZ3. The heterodimer is inactive in UVcrosslinking not only to VLE RNA but other RNAs as well, such as vegetally localized TGFb-5 mRNA (Oberman et al., 2007). We decided to use the same construct to assay whether the RNA-binding activity of xVICKZ3 is required for the process of VLE RNA localization to the vegetal cortex of oocytes. To test whether xVICKZ3D would also inhibit endogenous xVICKZ3 RNA binding when expressed in oocytes, RNA encoding xVICKZ3D was injected into stage III/IV oocytes, incubated for 5 days, and then protein extracts were used in UV crosslinking assays. As seen in Fig. 4A, expression of exogenous xVICKZ3D was essentially equimolar to that of the endogenous xVICKZ3 and did not reduce endogenous xVICKZ3 levels relative to uninjected, control oocytes. The presence of the deleted construct in the oocytes, however, caused essentially a complete loss of the UV-crosslinking activity of the endogenous xVICKZ3. Notably,
the ability of xVICKZ3D to associate with MTs remained the same as full-length xVICKZ3 (Fig. 4B). Because of the heterogeneous nature of Xenopus oocytes, even those of similar diameter often show a wide range in the degree of endogenous Vg1 RNA localization (Schwartz, 1996). For this reason, we chose to assay the effect of xVICKZ3D on exogenous VLE RNA localization. Thus, stage III/IV oocytes were first injected with RNA encoding xVICKZ3D and incubated overnight in medium to allow for accumulation of the exogenous protein. Fluorescein-labeled VLE RNA was then injected into these oocytes and allowed to undergo localization for an additional 4 days. As seen in Fig. 4C, the presence of the xVICKZ3D construct had no effect on the ability of VLE RNA to undergo localization, even though the presence of this protein completely inhibited the ability of xVICKZ3 to UV-crosslink VLE RNA. Thus, the RNA-binding activity of xVICKZ3, even though required in embryos and in somatic cells for cell migration, is not required in oocytes for VLE RNA localization.
xVICKZ3D localizes to the vegetal cortex as part of an RNP complex In oocytes injected with the xVICKZ3D construct, the ability of VLE RNA to localize to the vegetal cortex
571
suggested that either xVICKZ3 is not required for RNA localization, or, alternatively, that it need not bind the RNA in order to mediate vegetal RNA localization. The latter option would imply that xVICKZ3D remains part of the VLE RNP complex that translocates to the vegetal cortex, even while not binding RNA directly. Work on the chick cVICKZ1 homolog, ZBP1, however, has shown that when RNA binding is inhibited (either through deletions in ZBP1 or by antisense DNA oligonucleotides directed against the zipcode localization elements of b-actin RNA), ZBP1 does not localize to lamellopodia (Farina et al., 2003; Oleynikov and Singer, 2003). To begin to understand how VLE RNA is correctly localized in the absence of normal xVICKZ3 binding, we first analyzed whether xVICKZ3D localizes to the vegetal cortex in oocytes. In contrast to what was observed with ZBP1, RNA binding appears not to be required for xVICKZ3 vegetal localization. xVICKZ3D– GFP, despite its inability to bind VLE RNA sequences, undergoes localization as well as the full-length construct (Fig. 4C). Because of the potential of protein– protein interactions among the different VLE-binding proteins, we wondered whether disrupting xVICKZ3 binding to VLE RNA was enough to dissociate the protein from the VLE RNP complex. xVICKZ3-containing complexes were immunoprecipitated (IP 0 d) from xVICKZ3D-injected oocytes using anti-VICKZ antibody bound to beads, under conditions that preserved RNA–protein and protein–protein interactions. Associated RNAs were assayed by RT-PCR (Fig. 4D). De-
Fig. 4 VLE RNA localization in oocytes is not dependent on its binding to xVICKZ3 but rather on its association with the VLE RNP complex. (A) Stage III/IV oocytes were injected with xVICKZ3D and cultured for five days. Proteins were extracted from uninjected (control oocyte) or xVICKZ3D-injected (xVICKZ3D injected oocyte) oocytes and used for both Western blot analysis with an anti-VICKZ antibody (top panel, one oocyte equivalent/lane) and UV crosslinking assays with a radioactive VLE probe (bottom panel, two oocyte equivalents/lane). Because xVICKZ3D is 51 amino acids shorter than xVICKZ3, it can be observed running slightly ahead of the full-length endogenous protein on the Western blot, where it is expressed in approximately equimolar amounts as compared with the endogenous protein (in both control and injected oocytes). (B) In parallel to the experiments in A, the ability of injected xVICKZ3D to associate with MTs in oocytes was assayed by a microtubules pull down, as described in Fig. 2. xVICKZ3D appears to behave just like the endogenous xVICKZ3, both in the absence and presence of Taxol and guanosine triphosphate. (C) Stage III/IV oocytes were injected with either fluoresceinated VLE RNA (VLE-Fl) alone or first with xVICKZ3D–GFP RNA and then, 24 hr later, fluoresceinated VLE RNA (VLE-Fl1xVICKZ3D). After VLE RNA injection, oocytes were cultured for 4 days. Localization of the (exogenous) VLE RNA was visualized through the use of an anti-fluorescine antibody (VLE RNA), while xVICKZ3D–GFP localization was visualized
with an anti-GFP antibody (xVICKZ3D). As seen in the left two oocytes, VLE RNA is capable of vegetal localization, whether or not xVICKZ3D protein is present in the cell. Although this truncated protein inhibits binding between endogenous xVICKZ3 and VLE RNA (A), and does not bind VLE RNA itself (Oberman et al., 2007), xVICKZ3D–GFP undergoes vegetal localization as well (right oocyte). (D) xVICKZ3D RNA was injected into stage III/IV oocytes alone (xVICKZ3D) or with VLE RNA (xVICKZ3D1VLE RNA) and cultured for 5 days. Protein extracts from injected and uninjected oocytes were immunoprecipitated (IP) with either antiVICKZ antibody (aVICKZ) or preimmune serum (PI), and RNAs were isolated and used for RT-PCR in order to determine the RNAs associated with xVICKZ3 RNP complexes. Endogenous Vg1 RNA (left lane) and VLE RNA (second lane from the left) remain part of the xVICKZ3 complex, even in the presence of xVICKZ3D and are not pulled down with preimmune serum. Vg1 RNA is also part of the xVICKZ3 complex in uninjected oocytes (right lane). (E) Stage III/IV oocytes were injected with xVICKZ3D–GFP and cultured for 4 days. Protein extracts from the injected oocytes were immunoprecipitated with either antiVICKZ antibody (aVICKZ) or preimmune serum (Preimmune) and Western blotted with anti-VICKZ antibody (aVICKZ). The filter was then stripped and incubated with an anti-hnRNP I antibody (ahnRNP I). A band of 60 kDa, clearly present in the anti-VICKZ IP lane and absent in the Preimmune lane, is observed when the stripped filter is incubated with anti-hnRNP I antibody, suggesting that xVICKZ3 and hnRNP I are part of the same RNP complex. GFP, green fluorescent protein; VLE, Vegetal localization element; RNP, ribonucleoprotein.
572
spite the inability of xVICKZ3 in the xVICKZ3D-injected oocytes to crosslink VLE RNA, endogenous Vg1 RNA and exogenous VLE RNA both immunoprecipitate with xVICKZ3, suggesting that they remain part of the same RNP complex even in the absence of direct binding. Vg1 RNA was not detected in samples immunoprecipitated with pre-immune serum, indicating that the RNA does not adventitiously associate with serum or beads during centrifugation. Mowry and colleagues have shown that hnRNP I (also referred to as VgRBP60 or PTB) interacts with xVICKZ3 both in vitro and in vivo, and they have suggested that this association can occur in the absence of RNA (Kress et al., 2004). In accordance with these results, we find that when extracts from oocytes injected with xVICKZ3D are immunoprecipitated with antiVICKZ, hnRNP I is indeed co-precipitated (Fig. 4E). Thus, taken together, these data indicate that VLE RNA can undergo normal localization even in the absence of direct xVICKZ3-VLE RNA binding. Disruption of RNA binding, however, does not cause xVICKZ3 to be completely dislodged from the VLE RNP complex; xVICKZ3 remains indirectly associated with Vg1 RNA.
Discussion We have analyzed the mechanism involved in localizing xVICKZ3 to the vegetal cortex of oocytes through the use of deletion constructs and found that vegetal targeting correlates with the ability of the construct to associate with MTs. Even the ability of the construct to bind RNA is not required, inasmuch as the dominant negative construct, xVICKZ3D, that inhibits wild-type xVICKZ3 RNA binding, also localizes. Consistent with this finding is the fact that the RRM-KH12 construct is also capable of localization, even though it does not dimerize and binds VLE RNA very poorly (Git and Standart, 2002). In oocytes, xVICKZ3 is part of an RNP complex that is assembled first in the nucleus, where it associates with hnRNP I in an interaction that is not mediated by RNA (Kress et al., 2004). In the cytoplasm, where the Vg1 RNP complex is reorganized and additional proteins are recruited, our results suggest that xVICKZ3 is maintained in the complex by protein–protein interactions, even in the absence of its direct RNA binding; thus, both hnRNP I and Vg1 and VLE RNAs are brought down by anti-VICKZ co-immunoprecipitation, even in the presence of xVICKZ3D. Given that the pull down of hnRNP I with antibody directed against xVICKZ3 is prevented in cytoplasmic, but not nuclear, oocyte extracts by RNAse treatment (Kress et al., 2004), these interactions may be stabilized and/or mediated by additional proteins present in the complex. Other proteins known to co-
immunoprecipitate with xVICKZ3 in oocytes include Prrp, XStau, Kinesin I and II, Vg1RBP71/KSRP, and 40LoVe (Zhao et al., 2001; Kroll et al., 2002; Allison et al., 2004; Betley et al., 2004; Yoon and Mowry, 2004; Czaplinski et al., 2005). Given the high degree of conservation among the VICKZ family homologs, it is informative to compare the results of the structure–function analysis of xVICKZ3 in Xenopus oocytes to that of cVICKZ1 in chick embryo fibroblasts (Farina et al., 2003) and hVICKZ1 in mouse NIH 3T3 cells (Nielsen et al., 2002). In both VICKZ1 studies, both KH didomains are required for lamellipodia localization, and neither of the KH didomains alone localized. In the case of cVICKZ1, sequences between the two RRM domains are also required for peripheral localization; in the case of hVICKZ1, the entire N-terminus, including both RRMs and intervening sequences, appears dispensable for localization. In xVICKZ3, however, either KH didomain can mediate, independently, xVICKZ3 localization to the vegetal cortex, and the RRM domains are neither necessary nor sufficient for localization. It is interesting to note that the minimal hVICKZ1 construct found to undergo peripheral localization (KH1-4) was also shown to associate with MTs and associated proteins (Nielsen et al., 2002). In the case of xVICKZ3, the minimal construct that mediates vegetal localization was each of the KH didomains, and each of these in fact associates with MTs and associated proteins. At least some of these differences may reflect the distinct nature of Xenopus oocytes, which can be 50–100 times larger in diameter than cells in culture, and have a radial MT array that extends, in stage III/IV oocytes, from the germinal center to the cortex (Gard, 1991, 1999). The large distances that need to be transversed likely dictate the requirement for an MT-based RNA transport system (as, for example, in neuronal axons), and the rich MT network that develops places a premium on proteins that can associate with it. In addition, transported RNP complexes may well require a more elaborate array of proteins and protein–protein interactions in order to maintain the integrity of the complex; indeed, an abundance of oocyte proteins have been identified as part of the Vg1 RNP complex. In cells in culture, where RNAs are translocated over much shorter distances, transport mechanisms are generally actin based. Although certainly a number of different proteins are present in the cVICKZ1 RNP in chick embryo fibroblasts (Ross et al., 1997; Gu et al., 2002), it will be interesting to see if the requirement of RNA binding for cVICKZ1 lamellipodial localization in these cells (Oleynikov and Singer, 2003), as opposed to what we observe with xVICKZ3 vegetal localization, reflects a less complex RNP structure. The exact role of xVICKZ3 in the Vg1 RNA localization in oocytes remains unclear. Injection of xVICKZ3D into embryos has a clear, and direct, effect
573
on neural crest migration, and essentially phenocopies the effect of knocking down xVICKZ3 expression through the use of AMO (Oberman et al., 2007). The large amount of stable xVICKZ3 protein present in oocytes, however, made AMO knock-down impractical, and hence we decided to make use of the xVICKZ3D construct to inhibit xVICKZ3 activity. In tissue culture cells, VICKZD3 specifically inhibits the RNA-binding activity of endogenous VICKZ3 but not other RNAbinding proteins (Oberman et al., 2007). In oocytes, this construct indeed inhibits xVICKZ3 binding of VLE RNA while not affecting the localization of RNA, demonstrating the dispensability of xVICKZ3 RNA binding in localization. Unlike the AMO knock-down in embryos, however, expression of the dominant negative construct did not lead to loss of the full-length protein from the RNP complex. Thus, we cannot conclude whether VLE RNA localization to the vegetal cortex, in the presence of xVICKZ3D is a result of xVICKZ3’s continued help in directing localization, perhaps by mediating MT association, or if other proteins present in the complex provide the directional information needed for vegetal localization. Kwon et al. (2002) have shown that injection of anti-VICKZ antibody into stage III oocytes reduces localization of both VLE and VegT RNAs by 50%. Nevertheless, their results do not distinguish between a physical hindrance phenomenon, where the presence of bound antibody may affect association of the RNP to either localization or anchoring machinery, or a neutralization effect of xVICKZ3, where the protein is either removed from the complex or effectively prevented from functioning. A clear answer will only be possible once xVICKZ3 can be definitively removed from the complex. Acknowledgments We would like to thank Dr. Froma Oberman for the xVICKZ3D–GFP construct, Dr. Kim Mowry for the antihnRNP I antibody, Dr. Leah Rosen for help with the meta-analysis, and Dr. Mark Tarshish for help with the confocal microscopy. Different aspects of this work were supported by the Israel Cancer Research Fund (ICRF), Israel Science Foundation (ISF), and the Association for International Cancer Research (AICR).
References Allison, R., Czaplinski, K., Git, A., Adegbenro, E., Stennard, F., Houliston, E. and Standart, N. (2004) Two distinct Staufen isoforms in Xenopus are vegetally localized during oogenesis. Rna 10:1751–1763. Bassell, G.J., Zhang, H., Byrd, A.L., Femino, A.M., Singer, R.H., Taneja, K.L., Lifshitz, L.M., Herman, I.M. and Kosik, K.S. (1998) Sorting of beta-actin mRNA and protein to neurites and growth cones in culture. J Neurosci 18:251–265. Betley, J.N., Heinrich, B., Vernos, I., Sardet, C., Prodon, F. and Deshler, J.O. (2004) Kinesin II mediates Vg1 mRNA transport in Xenopus oocytes. Curr Biol 14:219–224. Cooper, H. and Hedges, L. (1994) The handbook of research synthesis. Russell Sage Foundation, New York.
Cote, C.A., Gautreau, D., Denegre, J.M., Kress, T.L., Terry, N.A. and Mowry, K.L. (1999) A Xenopus protein related to hnRNP I has a role in cytoplasmic RNA localization. Mol Cell 4:431–437. Czaplinski, K., Kocher, T., Schelder, M., Segref, A., Wilm, M. and Mattaj, I.W. (2005) Identification of 40LoVe, a Xenopus hnRNP D family protein involved in localizing a TGF-beta-related mRNA during oogenesis. Dev Cell 8:505–515. Deshler, J.O., Highett, M.I., Abramson, T. and Schnapp, B.J. (1998) A highly conserved RNA-binding protein for cytoplasmic mRNA localization in vertebrates. Curr Biol 8:489–496. Doyle, G.A., Betz, N.A., Leeds, P.F., Fleisig, A.J., Prokipcak, R.D. and Ross, J. (1998) The c-myc coding region determinant-binding protein: a member of a family of KH domain RNA-binding proteins. Nucleic Acids Res 26:5036–5044. Elisha, Z., Havin, L., Ringel, I. and Yisraeli, J.K. (1995) Vg1 RNA binding protein mediates the association of Vg1 RNA with microtubules in Xenopus oocytes. EMBO J 14:5109–5114. Eom, T., Antar, L.N., Singer, R.H. and Bassell, G.J. (2003) Localization of a beta-actin messenger zipcode-binding protein modulates the density of filopodial synapses. J Neurosci 23:10433–10444. Farina, K.L., Huttelmaier, S., Musunuru, K., Darnell, R. and Singer, R.H. (2003) Two ZBP1 KH domains facilitate beta-actin mRNA localization, granule formation, and cytoskeletal attachment. J Cell Biol 160:77–87. Gard, D.L. (1991) Organization, nucleation, and acetylation of microtubules in Xenopus laevis oocytes: a study by confocal immunofluorescence microscopy. Dev Biol 143:346–362. Gard, D.L. (1999) Confocal microscopy and 3-D reconstruction of the cytoskeleton of Xenopus oocytes. Microsc Res Tech 44:388– 414. Git, A. and Standart, N. (2002) The KH domains of Xenopus Vg1RBP mediate RNA binding and self-association. RNA 8:1319–1333. Gu, W., Pan, F., Zhang, H., Bassell, G.J. and Singer, R.H. (2002) A predominantly nuclear protein affecting cytoplasmic localization of beta-actin mRNA in fibroblasts and neurons. J Cell Biol 156:41–51. Havin, L., Git, A., Elisha, Z., Oberman, F., Yaniv, K., Schwartz, S.P., Standart, N. and Yisraeli, J.K. (1998) RNA-binding protein conserved in both microtubule- and microfilament-based RNA localization. Genes Dev 12:1593–1598. Huttelmaier, S., Zenklusen, D., Lederer, M., Dictenberg, J., Lorenz, M., Meng, X., Bassell, G.J., Condeelis, J. and Singer, R.H. (2005) Spatial regulation of beta-actin translation by Src-dependent phosphorylation of ZBP1. Nature 438:512–515. Kress, T.L., Yoon, Y.J. and Mowry, K.L. (2004) Nuclear RNP complex assembly initiates cytoplasmic RNA localization. J Cell Biol 165:203–211. Kroll, T.T., Zhao, W.M., Jiang, C. and Huber, P.W. (2002) A homolog of FBP2/KSRP binds to localized mRNAs in Xenopus oocytes. Development 129:5609–5619. Kwon, S., Abramson, T., Munro, T.P., John, C.M., Kohrmann, M. and Schnapp, B.J. (2002) UUCAC- and vera-dependent localization of VegT RNA in Xenopus oocytes. Curr Biol 12: 558–564. Lawrence, J.B. and Singer, R.H. (1986) Intracellular localization of messenger RNAs for cytoskeletal proteins. Cell 45:407–415. Leung, K.M., van Horck, F.P., Lin, A.C., Allison, R., Standart, N. and Holt, C.E. (2006) Asymmetrical beta-actin mRNA translation in growth cones mediates attractive turning to netrin-1. Nat Neurosci 9:1247–1256. Nielsen, F.C., Nielsen, J., Kristensen, M.A., Koch, G. and Christiansen, J. (2002) Cytoplasmic trafficking of IGF-II mRNAbinding protein by conserved KH domains. J Cell Sci 115: 2087–2097. Nielsen, J., Christiansen, J., Lykke-Andersen, J., Johnsen, A.H., Wewer, U.M. and Nielsen, F.C. (1999) A family of insulin-like growth factor II mRNA-binding proteins represses translation in late development. Mol Cell Biol 19:1262–1270.
574 Oberman, F., Rand, K., Maizels, Y., Rubinstein, A. and Yisraeli, J.K. (2007) VICKZ proteins mediate cell migration via their RNA binding activity. RNA (in press). Oleynikov, Y. and Singer, R.H. (2003) Real-time visualization of ZBP1 association with beta-actin mRNA during transcription and localization. Curr Biol 13:199–207. Ross, A.F., Oleynikov, Y., Kislauskis, E.H., Taneja, K.L. and Singer, R.H. (1997) Characterization of a beta-actin mRNA zipcode-binding protein. Mol Cell Biol 17:2158–2165. Schwartz, S.P. (1996) Localization of Vg1—a maternal mRNA. Hebrew University, Jerusalem, Israel. Sundell, C.L. and Singer, R.H. (1991) Requirement of microfilaments in sorting of actin messenger RNA. Science 253:1275–1277. Tiruchinapalli, D.M., Oleynikov, Y., Kelic, S., Shenoy, S.M., Hartley, A., Stanton, P.K., Singer, R.H. and Bassell, G.J. (2003) Activity-dependent trafficking and dynamic localization of zipcode binding protein 1 and beta-actin mRNA in dendrites and spines of hippocampal neurons. J Neurosci 23: 3251–3261. Vikesaa, J., Hansen, T.V., Jonson, L., Borup, R., Wewer, U.M., Christiansen, J. and Nielsen, F.C. (2006) RNA-binding IMPs promote cell adhesion and invadopodia formation. EMBO J 25:1456–1468. Yaniv, K., Fainsod, A., Kalcheim, C. and Yisraeli, J.K. (2003) The RNA-binding protein Vg1 RBP is required for cell migration during early neural development. Development 130: 5649–5661.
Yao, J., Sasaki, Y., Wen, Z., Bassell, G.J. and Zheng, J.Q. (2006) An essential role for beta-actin mRNA localization and translation in Ca21-dependent growth cone guidance. Nat Neurosci 9:1265–1273. Yisraeli, J.K. (2005) VICKZ proteins: a multi-talented family of regulatory RNA-binding proteins. Biol Cell 97:87–96. Yisraeli, J.K., Sokol, S. and Melton, D.A. (1990) A two-step model for the localization of a maternal mRNA in Xenopus oocytes: involvement of microtubules and microfilaments in translocation and anchoring of Vg1 mRNA. Development 108:289–298. Yoon, Y.J. and Mowry, K.L. (2004) Xenopus Staufen is a component of a ribonucleoprotein complex containing Vg1 RNA and kinesin. Development 131:3035–3045. Zhang, H.L., Eom, T., Oleynikov, Y., Shenoy, S.M., Liebelt, D.A., Dictenberg, J.B., Singer, R.H. and Bassell, G.J. (2001) Neurotrophin-induced transport of a beta-actin mRNP complex increases beta-actin levels and stimulates growth cone motility. Neuron 31:261–275. Zhang, Q., Yaniv, K., Oberman, F., Wolke, U., Git, A., Fromer, M., Taylor, W.L., Meyer, D., Standart, N., Raz, E. and Yisraeli, J.K. (1999) Vg1 RBP intracellular distribution and evolutionarily conserved expression at multiple stages during development. Mech Dev 88:101–106. Zhao, W.M., Jiang, C., Kroll, T.T. and Huber, P.W. (2001) A proline-rich protein binds to the localization element of Xenopus Vg1 mRNA and to ligands involved in actin polymerization. EMBO J 20:2315–2325.