BIOLOGICAL OVERVIEW

For many years, interest in Squid was limited to Drosophila biologists interested in determining which maternal genes were involved in the origin of dorsal/ventral patterning during early embryogenesis. Gurken mRNA localization to the anterior, dorsal region of the oocyte triggers a response in the overlying follicle cell that leads to the development of D/V patterning in the fertilized egg. Gurken mRNA localization and translation are coupled by an interaction between two RNA-binding proteins: Sqd and Bruno, a translational repressor protein. Because Sqd protein binds GRK mRNA directly and belongs to the class of heterogeneous nuclear ribonucleoproteins (hnRNPs) implicated in nuclear mRNA export, it seems likely that Sqd functions in the nuclear export of GRK mRNA. The function of the Sqd protein appears to be associated with the regulated nuclear export of GRK mRNA: it may be that Sqd is responsible for delivering the GRK message to a cytoplasmic protein involved in GRK anchoring, events that are possibly coupled to GRK translation. Thus, one might expect that Sqd protein should interact with some cytoplasmic ovarian proteins (Norvell, 1999).

Heterogeneous nuclear ribonucleoproteins (hnRNPs) are predominantly nuclear RNA-binding proteins that form complexes with RNA polymerase II transcripts. These proteins function in a staggering array of cellular activities, ranging from transcription and pre-mRNA processing in the nucleus to cytoplasmic mRNA translation and turnover (for a recent review, see Krecic, 1998). Squid is most closely related to several classes of hnRNPs, and is included in the D1/D2 class. Of the approximately 25 classes of known hnRNPs, the D1/D2 class, along with the A1/A1 class and A2/B1 class, each have two RNA-binding domains (RBDs), also termed RNA recognition motifs (RRMs) and an arginine/glycine-rich box (RGG). This is the same configuration as several other Drosophila hnRNPs, including Musashi (Nakamura, 1994) and Hrp48 (Hammand, 1997). Since the homology of Squid to all of these proteins falls in the 42% range, the diverse functions of each of these classes of hnRNPs will be dealt with in the squid Evolutionary homologs section.

Two recent publications have redefined the role of maternal Squid and suggest a wider biological function for the gene than previously thought: Norvell (1999) examines the transport of Gurken mRNA and the work of Lall (1999) shows that Squid promotes apical cytoplasmic transport and localization of Fushi tarazu transcripts.

The Norvell (1999) study analyzes the involvement of different Squid isoforms in the process of Gurken mRNA localization. The interactions of Gurken protein during oogenesis include not only the protein-protein interaction with Bruno, and the protein-mRNA interaction with Gurken mRNA, but a protein-protein interaction with the nuclear import protein Transportin, and in addition, a protein-protein interaction with Fs(1)K10, a protein identified by the effects of a female sterile mutation. This study reveals that three different Squid isoforms have different subcellular distributions and play distinct roles in Gurken mRNA localization. The relationship between K10 and the distribution of Sqd protein demonstrates the importance of Sqd accumulation within the oocyte nucleus. Although the nuclear import of Sqd protein is most likely driven by its association with Transportin, K10 function is required for the stable accumulation of Sqd in the oocyte nucleus. This places K10 function upstream of Sqd in the germ line. Accumulation of Sqd protein in the nurse cells is not affected, and in addition, Sqd is detectable within the oocyte cytoplasm of K10 mutants. The major effect of K10 must be on the nuclear retention of one of the Squid isoforms (Norvell, 1999).

This overview will treat the zygotic functions of Squid, leaving discussion of the effects of Squid on Gurken mRNA accumulation to other sections of this gene report (see Protein interactions, transport and localization of transcripts). One class of localized transcripts is encoded by the zygotic Drosophila pair-rule genes [including fushi tarazu (ftz) and hairy (h)], which are required to establish reiterated (segmental) embryonic pattern. Pair-rule transcripts are transcribed in the syncytial blastoderm embryo as seven distinct transverse stripes along the anterior/posterior axis. At this stage, a monolayer of nuclei at the embryonic surface subdivides the cortical cytoplasm ('periplasm') into apical and basal compartments. Most zygotic transcripts do not localize to a specific periplasmic compartment in blastoderm embryos, but pair-rule transcripts accumulate exclusively in the apical periplasm. The function of this localization is not yet established, although it may serve to restrict protein diffusion within the syncytial embryo. Several lines of evidence have argued against pair-rule transcripts localizing by cytoplasmic transport. (1) Pair-rule mRNAs are not detectable in the basal cytoplasm, even after extensive overstaining or stabilization of transcripts. (2) Pair-rule transcripts are extremely unstable (t1/2 6.5 min), so they are more likely to localize directly and rapidly. Stabilizing the transcripts does not prevent localization, showing that their selective accumulation is not due to selective degradation of basal transcripts. These observations have led to the proposal that Drosophila pair-rule transcripts localize directly (i.e., by selective [vectorial] export through one side of the nuclear envelope) (Francis-Lang, 1996). (3) Further evidence for this model comes from experiments examining aneuploid blastoderm embryos, in which pair-rule transcripts localize apically to displaced internalized nuclei (Francis-Lang, 1996): this indicates linkage between the nucleus and sites of transcript localization. Evidence connecting transcription and localization comes from studies showing that bcd transcripts localize apically only in cells where they are being synthesized. bcd transcripts are made in nurse cells, before being transported into the adjacent oocyte where they localize anteriorly. Maternal bcd transcripts localize apically to nuclei in nurse cells but not in the mature egg. Zygotic transcripts with a bcd 3'-untranslated region (3'UTR) localize apically in blastoderm embryos. Together, these results indicate that the bcd apical localization signal only operates when coupled with synthesis (Lall, 1999 and references).

Previous studies have shown that the 3'UTRs of pair-rule transcripts are necessary and sufficient to target transcripts apically (Davis, 1991). In the case of ftz, the localization signal resides within a 1.3 kb region of the ftz gene, which includes the ftz 3'UTR (Davis, 1991). This signal has been defined more precisely using germline transformation and it has been found that hybrid lacZ-ftz 3'UTR transcripts with 205 bp of ftz 3' genomic sequence are apically localized (LTf2). Transcripts that lack the last 53 bases of the 3'UTR fail to localize apically, showing that the ftz 3'UTR is necessary and sufficient to target a heterologous transcript apically (Lall, 1999).

Maternal Drosophila and Xenopus transcripts that localize via cytoplasmic mechanisms also have localization signals in the 3'UTR. Thus, endogenous pair-rule mRNAs might localize similarly, despite previous indirect evidence for a nuclear mechanism. This possibility was examined directly by testing whether transcripts injected into the blastoderm cytoplasm can localize specifically: if nuclear events are essential for apical accumulation of pair-rule transcripts, ftz mRNAs that have not been exposed to such an environment should not localize. To check localization, visualized transcripts that had been labeled with fluorescent tags were visualized directly. Capped, polyadenylated transcripts incorporating aminoallyl-UTP were synthesized in vitro and chemically labeled with either fluorescein (FITC) or Rhodamine (Rh). FITC-labeled ftz and Rh-labeled ftzDelta3' transcripts were injected into the basal periplasm during nuclear cleavage cycles 13 or 14, when endogenous pair-rule transcripts are restricted to the apical periplasm. Localization of the RNAs was examined by confocal microscopy 0-30 min after injection. Initial experiments have shown that both transcripts are short lived, in accord with in vivo measurements of ftz transcript half-life. Thus, the transcripts do not persist long enough to test for selective localization. To overcome this problem, cycloheximide was coinjected. This has previously been shown to stabilize endogenous pair-rule transcripts without affecting their localization. Under these conditions, injected ftz and ftzDelta3' transcripts are still readily detectable 30 min or more after injection. No evidence of selective localization of the full-length ftz transcripts is found. In essentially all embryos (>99%; n > 400), full-length ftz transcripts fail to accumulate selectively in the apical periplasm; both injected transcripts diffuse out from the site of injection and colocalize for at least 30 min. Thus, injected ftz transcripts are unable to mimic the apical localization shown by endogenous pair-rule and reporter transcripts, indicating that purely cytoplasmic mechanisms are insufficient to account for pair-rule transcript localization (Lall, 1999).

Naked pair-rule transcripts could be unable to localize either because localization depends on vectorial nuclear export, or because transcripts need prior exposure to a nuclear environment for apical targeting (e.g., to recruit nuclear proteins that are subsequently required for localization in the cytoplasm). To distinguish between these alternatives, a test was performed to see whether nuclear proteins might promote cytoplasmic localization of pair-rule transcripts. FITC-ftz and Rh-ftzDelta3' were exposed to Drosophila embryonic nuclear extracts and coinjected the 'preincubated transcripts' with cycloheximide into the basal periplasm of cycle 14 blastoderm embryos. Strikingly, preincubated FITC-ftz transcripts specifically accumulate in the apical periplasm within 10 min of injection, whereas ftzDelta3' transcripts, lacking the 3'UTR, remain unlocalized. Nuclear extract is much more active than control proteins in promoting apical transcript localization. Together, these results indicate that nuclear extracts include specific factors required to localize pair-rule transcripts in blastoderm embryos (Lall, 1999).

Preincubated ftz transcripts localize in apical caps above the nuclei, thereby differing slightly from endogenous transcripts that localize as a continuous stripe domain. A further difference is that the accumulations of preincubated transcripts appear more particulate than those of endogenous transcripts, although this may reflect differences in expression levels or detection methods. In any case, the efficiency of localization is high, and little residual transcript remains at the site of injection. Thus, nuclear extracts include factors that specifically promote apical localization of preincubated ftz transcripts. Endogenous ftz transcripts are never observed in the basal periplasm, indicating that localization by a cytoplasmic mechanism should be extremely rapid. Efficient localization occurs in all embryos 4-5 min after injection, and 50% of embryos display apical localized transcripts 2-2.5 min after injection. Thus, preincubated transcripts also localize rapidly (Lall, 1999).

Localization of preincubated ftz is dependent on microtubules but independent of microfilaments. Early cycle 14 embryos were injected with 2 µg/ml of colcemid and 10 min later, injected with preincubated FITC-ftz transcript mixture. Localization is almost completely disrupted: the behavior of FITC-ftz resembles that of coinjected Rh-ftzDelta3'. This inhibition of localization indicates that pair-rule transcript localization depends on an intact microtubule cytoskeleton. By contrast, preincubated FITC-ftz transcripts still localize apically in embryos that have been coinjected with Cytochalasin B, which disrupts actin-dependent processes such as anchoring of nuclei to the cortex and, indeed, causes displacement of nuclei into the basal periplasm. Thus, processes disrupted by Cytochalasin B are not required for apical localization of preincubated FITC-ftz transcripts (Lall, 1999).

Nuclear factors promoting cytoplasmic transcript targeting are evolutionarily conserved. A human nuclear extract from TIG-3 cells (human fetal lung fibroblasts) can also promote pair-rule transcript localization. Preexposed FITC-ftz transcripts specifically localize apically (21/21 at 50 ng/µl protein), indicating that human nuclear extracts indeed promote pair-rule transcript localization. This degree of activity is higher than that of an equivalent Drosophila extract, although the proportion of localized transcripts within each embryo appears lower with human extracts. In any case, an activity that promotes pair-rule transcript localization is conserved between flies and humans (Lall, 1999).

For several reasons, the possibility was considered that the nuclear factors facilitating localization are hnRNP proteins: (1) hnRNP's are well conserved between flies and humans; (2) the nuclear factors that facilitate localization appear to function in the cytoplasm and therefore must shuttle between nucleus and cytoplasm, as do hnRNP's (Piñol-Roma, 1992 and Visa, 1996; reviewed in Piñol-Roma, 1997 and Mattaj, 1998); (3) the Drosophila Sqd hnRNP protein, a homolog of the mammalian hnRNP-A/B proteins, is required for gurken (grk) transcripts to localize during oogenesis, and (4) hnRNP-A2 has been shown to bind to a 3'UTR sequence required for the localization of MBP transcripts in rat oligodendrocytes (Hoek, 1998, as cited in Lall, 1999).

To test whether Sqd protein promotes cytoplasmic transport, transcripts were preincubated with each of the three Sqd protein isoforms and injected into blastoderm embryos. All three Sqd isoform fusion proteins are active in promoting apical localization. Thus, preexposure to any of the Sqd isoforms leads to localization of labeled ftz but not ftzDelta3' transcripts. The slightly lower activity of SqdB is probably not significant, although it is notable that expression of this isoform is incapable of supporting GRK transcript localization during oogenesis (Norvell, 1999). Association of ftz transcripts with Sqd is very rapid, being essentially complete within the 2 min period required to establish injections. A test was also performed to see whether the ability of hnRNP proteins to promote transcript localization has been evolutionarily conserved. ftz transcripts preincubated with each of the human hnRNP-A1, -A2 and -B proteins localize apically, showing that the activity of this class of A/B hnRNP proteins is conserved between Drosophila and humans (Lall, 1999).

Immunostaining of early embryos shows that Sqd is indeed present in blastoderm nuclei, as expected if Sqd is the major in vivo localizing activity. Unfortunately, a direct test for whether Sqd is required to localize pair-rule transcripts could not be made because strong sqd mutant eggs are not fertile. However, it was determined whether Sqd selectively recognizes the ftz 3'UTR by examining protein extracts from ovaries, the major source of Sqd in blastoderm embryos. Proteins that bind to ftz-3'UTR transcripts were labeled by UV cross-linking to ³²P-labeled transcripts and visualized following gel electrophoresis. ftz-3'UTR transcripts label a predominant 42 kDa protein in Drosophila ovary extracts, the same size as Sqd and as the activity labeled by the grk-3'UTR. The 42 kDa protein is indeed Sqd, being immunoprecipitated by anti-Sqd antibodies and only weakly labeled by a control nanos-3'UTR transcript. Thus, Sqd binds specifically to the ftz-3'UTR and represents a major such activity in oocyte extracts (Lall, 1999).

The ability of human hnRNP proteins to promote localization and the uniform expression of Sqd within blastoderm nuclei argue against a vectorial export mechanism. Localization of preincubated pair-rule transcripts requires an intact microtubule cytoskeleton and is likely to be mediated by microtubule-dependent motors. In wild-type blastoderm embryos, each nucleus is indeed capped by an apical bundle of microtubules that could serve as a framework for transcript transport. Apical transport along these microtubules would require transport by a dynein-like motor. However, pair-rule transcripts can localize elsewhere in aneuploid blastoderm embryos. In 3L- embryos, some nuclei become internalized and lack associated apical microtubules (Francis-Lang, 1996); nevertheless, transcripts accumulate adjacent and apical to these nuclei, raising the possibility that pair-rule transcripts run along a different, minority class of microtubules. After transport, pair-rule transcripts are anchored to the cytoskeleton, as shown by their lack of diffusion in blastoderm embryos. However, the transcripts still localize apically to nuclei displaced from the periphery of the embryo following Cytochalasin B treatment, indicating that transcript attachment sites differ from those of nuclei and are perhaps not microfilament based. In the latter case, they would differ from those already implicated in transcript localization (Lall, 1999 and references).

GENE STRUCTURE

The sqd cDNA clones contain one of two alternative 3' exons. All messages are derived from the same gene by differential RNA processing. 2.3, 1.9 and 1.7 kb transcripts all contain the same 3' exon. The 2.3 kb transcript encodes SqdS. SqbB uses an alternative 3' exon (Kelley, 1993).

Exons - 3

Bases in 3' UTR - 610 (SqdA) and 256 (SqdB)

PROTEIN STRUCTURE

Amino Acids - 308 (SqdB) and 321 (SqdA)

Structural Domains

To better understand the role(s) of hnRNP proteins in the process of mRNA formation, the major nuclear proteins that interact with hnRNAs in Drosophila melanogaster have been identified and characterized. cDNA clones of several D. melanogaster hnRNP proteins have been isolated and sequenced, and the genes encoding these proteins have been mapped cytologically on polytene chromosomes. These include the hnRNP proteins hrp36, hrp40, and hrp48, which together account for the major proteins of hnRNP complexes in D. melanogaster. All of the proteins described here contain two amino-terminal RNP consensus sequence RNA-binding domains and a carboxyl-terminal glycine-rich domain. This configuration, which is also found in the hnRNP A/B proteins of vertebrates, is referred to as 2 x RBD-Gly. The sequences of the D. melanogaster hnRNP proteins help define both highly conserved and variable amino acids within each RBD and support the suggestion that each RBD in multiple RBD-containing proteins has been conserved independently and has a different function. Although 2 x RBD-Gly proteins from evolutionarily distant organisms are conserved in their general structure, a surprising diversity among the members of this family of proteins is found. A mAb to the hrp40 proteins crossreacts with the human A/B and G hnRNP proteins and detects immunologically related proteins in divergent organisms from yeast to man. These data establish 2 x RBD-Gly as a prevalent hnRNP protein structure across eukaryotes. This information about the composition of hnRNP complexes and about the structure of hnRNA-binding proteins will facilitate studies of the functions of these proteins (Matunis, 1992).

The glycine rich domain continues until almost the end of the SqbB polypeptide, but in SquA the last 22 residues are replaced with 35 residues that lack glycine and are 95% hydrophilic (Kelley, 1993).

A 38-amino-acid motif, termed M9, mediates the nuclear import of hnRNPs through an interaction with the nuclear import protein Transportin (Siomi and Dreyfuss 1995; Pollard, 1996). Sequence analysis of the Sqd protein isoforms reveals that such an M9 motif is present in Sqd. A comparison of the human hnRNPA1 M9 sequence with the Sqd isoforms shows that M9 is differentially present in the isoforms. Specifically, within the alternatively spliced region of SqdS, there is a region (amino acids 300-338) that aligns with the human hnRNPA1 M9 sequence (16 of 38 amino acids identical) (Siomi et al. 1998). Interestingly, this sequence is not present in either the SqdA or the SqdB isoforms. However, there is another M9-like sequence present in all three Sqd isoforms (amino acids 215-254) that aligns with the human hnRNPA1 M9 motif (13 of 38 amino acids identical) (Norvell, 1999).

date revised: 2 February 2023

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.