RNA regulatory elements mediate control of Drosophila body pattern by the posterior morphogen nanos

Diversity of cytoplasmic functions for the 3' untranslated region of eukaryotic transcripts

The 3' untranslated region (3' UTR) can control gene expression by affecting the localization, stability and translation of mRNAs. The recent finding that 3' UTRs can control the decapping rate of mRNAs, in combination with their ability to influence the initiation of translation, suggests that 3' UTRs act through a direct or indirect interaction between the 3' and 5' ends of mRNAs.

nanos is an evolutionarily conserved organizer of anterior-posterior polarity

In Drosophila melanogaster, nanos functions as a localized determinant of posterior pattern. Nanos RNA is localized to the posterior pole of the maturing egg cell and encodes a protein that emanates from this localized source. Nanos acts as a translational repressor and thereby establishes a gradient of the morphogen Hunchback. Here we show that the mechanism by which nanos acts in Drosophila is a common developmental strategy in Dipteran insects. We used cytoplasmic transplantation assays to demonstrate that nanos activity is found in posterior poleplasm of five diverse Dipteran species. Genes homologous to nanos were identified from Drosophila virilis, the housefly Musca domestica, and the midge Chironomus samoensis. These genes encode RNAs that are each localized, like nanos, to the embryonic posterior pole. Most importantly, we demonstrate that these homologues can functionally substitute for nanos in D. melanogaster. These results suggest that nanos acts in a similar pathway for axis determination in other insects. Comparison of the Nanos sequences reveals only 19% overall protein sequence similarity; high conservation of a novel zinc finger near the carboxy terminus of the protein defines a region critical for nanos gene function.

Degradation of maternal string mRNA is controlled by proteins encoded on maternally contributed transcripts

In Drosophila, maternal string mRNAs are stable for the first few hours of development, but undergo specific timed degradation at the cellularisation stage. To determine whether the proteins that control this degradation are maternally or zygotically transcribed, we have used in situ hybridisation to determine the fate of maternal string transcripts in mutant embryos which lack individual chromosome arms. Our data indicate that maternal string mRNA persists for the normal period in all these mutants. Using alpha-amanitin to inhibit zygotic transcription we have shown that degradation of maternal mRNA is unaffected. Therefore, the proteins required to activate the degradation of string mRNA are encoded on a maternally contributed mRNA. We discuss possible models to explain the degradation pathway.

Serum induction of MEF2/RSRF expression in vascular myocytes is mediated at the level of translation

Vascular smooth muscle cells (VSMCs) reversibly coordinate the expression of VSMC-specific genes and the genes required for cell cycle progression. Here we demonstrate that isoforms of the MEF2/RSRF transcription factor are expressed in VSMCs and in vascular tissue. The MEF2A DNA-binding activity was upregulated when quiescent VSMCs were stimulated to proliferate with serum mitogens. The serum-induction of MEF2A DNA-binding activity occurred approximately 4 h following serum activation, and this correlated with an increase in the level of MEF2A protein without changes in the level of MEF2A mRNA or protein stability. These results indicate that MEF2A induction by serum is regulated at the level of translation.

Caenorhabditis elegans cyclin A- and B-type genes: a cyclin A multigene family, an ancestral cyclin B3 and differential germline expression

We have cloned cDNAs for Caenorhabditis elegans cyclins A1, B and B3. While cyclins A1 and B are most closely related to either A- or B-type cyclins of other species, cyclin B3 is less related to these cyclins. However, this cyclin is most similar to the recently identified chicken cyclin B3. Our identification of a Caenorhabditis homolog demonstrates that cyclin B3 has been conserved in evolution. Cyclin A1 is a member of an A-type multigene family; however the cyclin A1 cDNA only recognizes a single band on northern blots. A single-sized RNA is also observed for the cyclin B3 cDNA. In contrast, three different transcripts are observed for the cyclin B cDNA. Based on our analyses using RNAs from germline-defective mutants and from populations enriched for males, one cyclin B transcript is specific to the paternal germline. The two other cyclin B transcripts, as well as the cyclin A1 and cyclin B3 transcripts, are most abundant in the maternal germline and are only present at low levels in other tissues. Moreover, the 3′ untranslated regions of each Caenorhabditis cyclin cDNA possess several copies of potential translational control elements shown in Xenopus and Drosophila maternal cyclin mRNAs to function during oogenesis and early embryogenesis.

Rolling replication of short DNA circles

Natural genes and proteins often contain tandemly repeated sequence motifs that dramatically increase physiological specificity and activity. Given the selective value of such repeats, it is likely that several different mechanisms have been responsible for their generation. One mechanism that has been shown to generate relatively long tandem repeats (in the kilobase range) is rolling circle replication. In this communication, we demonstrate that rolling circle synthesis in a simple enzymatic system can produce tandem repeats of monomers as short as 34 bp. In addition to suggesting possible origins for natural tandem repeats, these observations provide a facile means for constructing libraries of repeated motifs for use in "in vitro evolution" experiments designed to select molecules with defined biological or chemical properties.

Translational regulation of oskar mRNA by bruno, an ovarian RNA-binding protein, is essential

Oskar (osk) protein directs the deployment of nanos (nos), the posterior body-patterning morphogen in Drosophila. To avoid inappropriate activation of nos, osk activity must appear only at the posterior pole of the oocyte, where the osk mRNA becomes localized during oogenesis. Here, we show that translation of osk mRNA is, and must be, repressed prior to its localization; absence of repression allows osk protein to accumulate throughout the oocyte, specifying posterior body patterning throughout the embryo. Translational repression is mediated by an ovarian protein, bruno, that binds specifically to bruno response elements (BREs), present in multiple copies in the osk mRNA 3'UTR. Addition of BREs to a heterologous mRNA renders it sensitive to translational repression in the ovary.

Spnr, a murine RNA-binding protein that is localized to cytoplasmic microtubules

Previous studies in transgenic mice have established the importance of the 3' untranslated region (UTR) of the spermatid-specific protamine-1 (Prm-1) mRNA in its translational control during male germ cell development. To clone genes that mediate the translational repression or activation of the Prm-1 mRNA, we screened cDNA expression libraries made with RNA from pachytene spermatocytes and round spermatids, with an RNA probe corresponding to the 3' UTR of Prm-1. We obtained six independent clones that encode Spnr, a spermatid perinuclear RNA-binding protein. Spnr is a 71-kD protein that contains two previously described RNA binding domains. The Spnr mRNA is expressed at high levels in the testis, ovary, and brain, and is present in multiple forms in those tissues. Immunolocalization of the Spnr protein within the testis shows that it is expressed exclusively in postmeiotic germ cells and that it is localized to the manchette, a spermatid-specific microtubular array. Although the Spnr protein is expressed too late to be directly involved in the translational repression of Prm-1 specifically, we suggest that the Spnr protein may be involved in other aspects of spermatid RNA metabolism, such as RNA transport or translational activation.

The intracellular localization of messenger RNAs

As I hope this review has made clear, mRNA localization plays an important role in directing specific proteins to their correct position within a cell. Although the study of this process is still in its infancy, it is already apparent that there are several ways that mRNAs can be targeted to particular subcellular sites. However, the molecular mechanisms responsible for these different localization pathways are still largely obscure, and their elucidation must await the identification of the specific factors that mediate the interactions between the localized mRNAs and more general components such as the cytoskeleton. Most examples of localized mRNAs are likely to share several common features. First, the site of localization will be determined by the preexisting polarity of the cell, and this will most often depend on the organization of the cytoskeleton, either directly, in the case of active transport, or indirectly, when localization is mediated by localized anchoring sites or stability factors. Second, mRNA localization is likely to be tightly coupled to translational control. If it is important for a cell to synthesize a protein in a particular place, then the translation of the mRNA must be repressed until it is localized. Indeed, there are already several examples where the direct linkage between translational control and localization has been demonstrated, and these are discussed in the accompanying review by Curtis et al. (1995).

Polyadenylation of c-mos mRNA as a control point in Xenopus meiotic maturation

c-mos protein, encoded by a proto-oncogene, is essential for the meiotic maturation of frog oocytes. Polyadenylation of c-mos messenger RNA is shown here to be a pivotal regulatory step in meiotic maturation. Maturation is prevented by selective amputation of polyadenylation signals from c-mos mRNA. Injection of a prosthetic RNA, which restores c-mos polyadenylation signals by base pairing to the amputated mRNA, rescues maturation and can stimulate translation in trans. Prosthetic RNAs may provide a general strategy by which to alter patterns of mRNA expression in vivo.

Novel Drosophila melanogaster genes encoding RRM-type RNA-binding proteins identified by a degenerate PCR strategy

We are interested in identifying Drosophila melanogaster RNA-binding proteins involved in important developmental decisions made at the level of mRNA processing, stability, localization or translational control. A large subset of the proteins known to interact with specific RNA sequences shares an evolutionarily conserved 80-90-amino-acid (aa) domain referred to as an RNA-recognition motif (RRM), including two ribonucleoprotein identifier sequences known as RNP-1 and RNP-2. Hence, we have herein applied degenerate polymerase chain reaction (PCR) methodology to clone three additional members (termed rox2, rox8 and rox21) of the D. melanogaster RRM-protein gene superfamily encoding putative trans-acting regulatory factors. Representative cDNA clones were isolated, the conceptual aa sequences of the candidate Rox proteins were inferred from their nucleotide sequences, and database searches were conducted. Rox2 displays extensive aa sequence similarities to putative RNA-binding proteins encoded by the genomes of the plants Oryza sativa and Arabidopsis thaliana; Rox21 resembles essential metazoan pre-mRNA splicing factors; as described elsewhere, Rox8 is likely a fly homolog of the two human TIA-1-type nucleolysins [Brand and Bourbon, Nucleic Acids Res. 21 (1993) 3699-3704].

Binding of pumilio to maternal hunchback mRNA is required for posterior patterning in Drosophila embryos

Posterior patterning in Drosophila embryos is governed by nanos (nos), which acts by repressing the translation of maternal transcripts of the hunchback (hb) gene. Sites in hb mRNA that mediate this repression, named nanos response elements (NREs), have been identified. However, we know of no evidence of a direct interaction between nos, or any other protein, and the NRE. Here, we show that two proteins present in embryonic extracts, neither one nos, bind specifically to the NRE in vitro. Furthermore, we show that binding in vitro correlates with NRE function in vivo. One of the NRE-binding factors is encoded by pumilio (pum), a gene that, like nos, is essential for abdominal segmentation. These and other observations suggest that pum acts by recognizing the NRE and then recruiting nos. Presumably, the resulting complex inhibits some component of the translation machinery.

Regulation of beta-tubulin function and expression in Drosophila spermatogenesis

In this study we examined two aspects of beta-tubulin function in Drosophila spermatogenesis: 1) beta-tubulin structural requirements for assembly of different categories of microtubules and 2) regulatory requirements for production of the correct tubulin protein level. In normal Drosophila spermatogenesis, the testis-specific beta 2-tubulin isoform supports multiple microtubule functions. Our previous work showed that another Drosophila isoform, beta 3, cannot support spermatogenesis, whereas a carboxyl-truncated form of beta 2, beta 2 delta C, can at least to some extent provide all of beta 2's normal functions, save one: beta 2 delta C cannot support organization of axonemal microtubules into the supramolecular architecture of the axoneme. Here, to test whether beta 2 carboxyl sequences can rescue the functional failure of the beta 3 isoform in spermatogenesis, we constructed a gene encoding a chimeric protein, beta 3 beta 2C, in which beta 3 sequences in the carboxyl region are replaced with those of beta 2. Unlike either beta 3 or beta 2 delta C, beta 3 beta 2C can provide partial function for both assembly of axonemal microtubules and their organization into the supramolecular architecture of the axoneme. In particular, the beta 2 carboxyl sequences mediate morphogenesis of the axoneme doublet tubule complex, including accessory microtubule assembly and attachment of spokes and linkers. However, our data also reveal aspects of beta 2-specific function that require structural features other than the primary sequence of the isotype-defining variable regions, the C terminus and the internal variable region. Tests of fecundity in males that coexpress beta 2 and the chimeric beta 3 beta 2C protein showed that in Drosophila there are differential requirements for sperm motility in the male and in the female reproductive tract. Since some aspects of microtubule function in spermatogenesis are sensitive to the tubulin pool size, we examined the mechanisms for control of tubulin protein levels in the male germ cells. We found that both beta 2-tubulin mRNA accumulation and protein synthesis are dependent on gene dose, and that the level of expression is regulated by 3' noncoding sequences in the beta 2 gene. Our data show that the regulatory mechanisms that control tubulin pool levels in the Drosophila male germ line differ from those observed in cultured animal somatic cells. Finally, expression of transgenic constructs is consistent with early cessation of X chromosome expression in Drosophila spermatogenesis.

Translational control of activin in Xenopus laevis embryos

Activin is a potent mesoderm inducing factor present in embryos of Xenopus laevis. Recent evidence has implicated activin in the inhibition of neural development in addition to the well-established induction of mesoderm in ectodermal explants. These diverse effects are critically dependent on the concentration of activin yet little is known about the mechanisms regulating the level of activin in the embryo. We report that the 3' untranslated region (3' UTR) of activin beta B mRNA inhibits the translation of activin in embryos. Micro-injection of activin mRNA from which the 3' UTR has been deleted is 8-10-fold more potent in inducing mesoderm than mRNA containing the 3' UTR. Truncation of the 3' UTR also leads to a marked enhancement of activin protein levels in embryos but has no effect when the truncated mRNA is translated in vitro. The 3' UTR also confers translational inhibition on a heterologous mRNA. These data show that a maternal factor(s) present in X. laevis regulates the translation of injected activin beta B mRNA. This factor(s) could be responsible for regulating the levels of endogenous activin beta B protein during mesoderm induction and the specification of ectodermal derivatives such as neural and epidermal tissues.

Coordinate initiation of Drosophila development by regulated polyadenylation of maternal messenger RNAs

Pattern formation in Drosophila depends initially on the translational activation of maternal messenger RNAs (mRNAs) whose protein products determine cell fate. Three mRNAs that dictate anterior, dorsoventral, and terminal specification--bicoid, Toll, and torso, respectively--showed increases in polyadenylate [poly(A)] tail length concomitant with translation. In contrast, posteriorly localized nanos mRNA, although also translationally activated, was not regulated by poly(A) status. These results implicate at least two mechanisms of mRNA activation in flies. Studies with bicoid mRNA showed that cytoplasmic polyadenylation is necessary for translation, establishing this pathway as essential for embryogenesis. Combined, these experiments identify a regulatory pathway that can coordinate initiation of maternal pattern formation systems in Drosophila.

CPEB is a specificity factor that mediates cytoplasmic polyadenylation during Xenopus oocyte maturation

The translational activation of several maternal mRNAs during Xenopus oocyte maturation is stimulated by cytoplasmic poly(A) elongation, which requires the uridine-rich cytoplasmic polyadenylation element (CPE) and the hexanucleotide AAUAAA. Here, we have enriched a CPE-binding protein (CPEB) by single-step RNA affinity chromatography, have obtained a CPEB cDNA, and have assessed the role of CPEB in cytoplasmic polyadenylation. The 62 kDa CPEB contains two RNA recognition motifs, and within this region, it is 62% identical to orb, an oocyte-specific RNA-binding protein from Drosophila. CPEB mRNA and protein are abundant in oocytes and are not detected in embryos beyond the gastrula stage. During oocyte maturation, CPEB is phosphorylated at a time that corresponds with the induction of polyadenylation. Immunodepletion of CPEB from polyadenylation-proficient egg extracts renders them incapable of adenylating exogenous RNA. Partial restoration of polyadenylation in depleted extracts is achieved by the addition of CPEB, thus demonstrating that this protein is required for cytoplasmic polyadenylation.

Selection of a subset of mRNAs from combinatorial 3' untranslated region libraries using neuronal RNA-binding protein Hel-N1

Hel-N1, a human RNA-binding protein, shares significant homology with Drosophila protein ELAV, which is essential for fly neuronal development. Hel-N1 has been shown to bind in vitro to 3' untranslated regions of mRNAs encoding c-myc, c-fos, granulocyte/macrophage colony-stimulating factor, and transcriptional repressor, Id. We report that Hel-N1 and a related form, Hel-N2, are expressed in human medulloblastoma cells, but their ratio differs significantly from that in adult brain and fetal brain. Selection of RNA targets from randomized combinatorial libraries yielded (A+U)-rich consensus sequences for both Hel-N1 and Hel-N2. As a means to identify cellular RNA targets for these proteins, we devised combinatorial shape libraries representing naturally derived 3' untranslated regions and were able to select a structurally related subset of transcripts that bound to Hel-N1. Approximately 10% of the proteins encoded by these subset mRNAs were identifiable in the data bases and most are implicated in cell growth regulation. This approach provides a means to gain access to novel genes expressed in various cell types by partitioning mRNAs containing common sequence elements using RNA-binding proteins.

Sequences responsible for intracellular localization of beta-actin messenger RNA also affect cell phenotype

We have characterized the structure and function of RNA sequences that direct beta-cytoplasmic actin mRNA to the cell periphery were mapped to two segments of 3'-untranslated region by expression of LacZ/beta-actin chimeric mRNAs in chicken embryo fibroblasts (CEFs). A 54-nt segment, the "RNA zipcode," and a homologous but less active 43-nt segment each localized beta-galactosidase activity to the leading lamellae. This zipcode contains the full activity, and mutations or deletions within it reduce, but do not eliminate, its activity, indicating that several motifs contribute to the activity. Two of these motifs, when multimerized, can regenerate almost full activity. These sequences are highly conserved in evolution, since the human beta-actin zipcode, positioned identically in the 3'UTR localizes equally well in chicken cells. Complementary phosphorothioate oligonucleotides against the zipcode delocalized endogenous beta-actin mRNA, whereas those complementary to the region just outside the zipcode, or sense oligonucleotides, did not. Actin mRNA or protein levels were unaffected by the antisense treatments, but a dramatic change in lamellipodia structure, and actin stress fiber organization was observed using the same antizipcode oligonucleotides which delocalized the mRNA. Hence, discrete 3'UTR sequences direct beta-actin isoform synthesis to the leading lamellae and affect cell morphology, presumably through the actin cytoskeleton.

Translational repression as a conserved mechanism for the regulation of embryonic polarity

The mechanisms used to establish embryonic polarity are still largely unknown. A recent paper(1) describes the expression pattern of the gene glp-1, which is required for induction events during development of the nematode Caenorhabditis elegans. Although glp-1 RNA is found throughout the early embryo, Glp-1 protein is only expressed in anterior cells. This negative translational regulation in posterior cells is shown to be mediated through sequences in the glp-1 3'untranslated region (3'UTR). Thus in nematodes, as in Drosophila, translational repression is one mechanism used to establish the embryonic anterior-posterior axis.

Germ cell-less encodes a cell type-specific nuclear pore-associated protein and functions early in the germ-cell specification pathway of Drosophila

The maternally supplied plasm at the posterior pole of a Drosophila embryo contains determinants that specify both the germ-cell precursors (pole cells) and the posterior axis. One pole plasma component, the product of the germ cell-less gene, has been found to be required for specification of pole cells, but not posterior somatic cells. Mothers with reduced levels of gcl give rise to progeny that lack pole cells, but are otherwise normal. Mothers overexpressing gcl, on the other hand, produce progeny exhibiting a transient increase of pole cells. Ectopic localization of gcl to the anterior pole of the embryo causes nuclei at that location to adopt characteristics of pole cell nuclei, with concurrent loss of somatic cells. We also present evidence indicating that the gcl protein associates specifically with the nuclear pores of the pole cell nuclei. This localization suggests a novel mechanism in the specification of cell fate for the germ line.

Synergy between the hunchback and bicoid morphogens is required for anterior patterning in Drosophila

Anterior patterning of the Drosophila embryo is specified by the localized expression of the gap genes, which is controlled by the gradient of the maternal morphogen bicoid (bcd). Another maternal component, hunchback (hb), can substitute for bcd in the thorax and abdomen. Here we show that hb is required for bcd to execute all of its functions. Removal of both maternal and zygotic hb produces embryos with disrupted polarity that fail to express all known bcd target genes correctly. Proper expression of hb and the head gap genes requires synergistic activation by hb and bcd. We propose that it is the combined activity of bcd and hb, and not bcd alone, that forms the morphogenetic gradient that specifies polarity along the embryonic axis and patterns the embryo. bcd may be a newly acquired Drosophila gene, which is gradually replacing some of the functions performed by maternal hb in other species.

Molecular biology of embryonic development: how far have we come in the last ten years?

The successes of molecular developmental biology over the last ten years have been particularly impressive in those directions favored by its major paradigms. New technologies have both guided and been guided by the progress of the field. I review briefly some of the major insights into embryonic development that have derived from research in four specific areas: early embryogenesis of various forms; 'pattern formation'; evolutionary conservation of regulatory elements; and spatial mechanisms of gene regulation. There remain many major problem areas, some of which may require new orientations to solve.

Induction of mesoderm in Xenopus laevis embryos by translation initiation factor 4E

The microinjection of messenger RNA encoding the eukaryotic translation initiation factor 4E (eIF-4E) into early embryos of Xenopus laevis leads to the induction of mesoderm in ectodermal explants. This induction occurs without a stimulation of overall protein synthesis and is blocked by the co-expression of a dominant negative mutant of the proto-oncogene ras or a truncated activin type II receptor. Although other translation factors have been studied in vertebrate and invertebrate embryos, none have been shown to play a direct role in development. The results here suggest a mechanism for relaying and amplifying signals for mesoderm induction.

Drosophila virilis oskar transgenes direct body patterning but not pole cell formation or maintenance of mRNA localization in D. melanogaster

The Drosophila melanogaster gene oskar is required for both posterior body patterning and germline formation in the early embryo; precisely how oskar functions is unknown. The oskar transcript is localized to the posterior pole of the developing oocyte, and oskar mRNA and protein are maintained at the pole through early embryogenesis. The posterior maintenance of oskar mRNA is dependent upon the presence of oskar protein. We have cloned and characterized the Drosophila virilis oskar homologue, virosk, and examined its activity as a transgene in Drosophila melanogaster flies. We find that the cis-acting mRNA localization signals are conserved, although the virosk transcript also transiently accumulates at novel intermediate sites. The virosk protein, however, shows substantial differences from oskar: while virosk is able to rescue body patterning in a D. melanogaster oskar- background, it is impaired in both mRNA maintenance and pole cell formation. Furthermore, virosk induces a dominant maternal-effect lethality when introduced into a wild-type background, and interferes with the posterior maintenance of the endogenous oskar transcript in early embryogenesis. Our data suggest that virosk protein is unable to anchor at the posterior pole of the early embryo; this defect could account for all of the characteristics of virosk mentioned above. Our observations support a model in which oskar protein functions both by nucleating the factors necessary for the activation of the posterior body patterning determinant and the germ cell determinant, and by anchoring these factors to the posterior pole of the embryo. While the posterior body patterning determinant need not be correctly localized to provide body patterning activity, the germ cell determinant may need to be highly concentrated adjacent to the cortex in order to direct pole cell formation.

Genetic control of development in Xenopus laevis

In this paper we address the question of how genes can control development by using Xenopus as a model system, since it combines the classical advantages of amphibian embryology with advanced molecular techniques. Several developmental regulator genes have been shown to encode for transcription factors which trigger the activation of downstream genes, thus resulting in a cascade of regulatory events. In the first two examples, we deal with regulatory events that underlie early body patterning in vertebrates, and with the role of homeobox transcription factors in deciphering positional information along the body axis. In the third example, we address the question of the role of post-transcriptional regulation in development by studying the possible regulatory role of a cytoplasmic zinc finger protein, presumably acting through RNA-protein interactions. The general idea is that understanding how genes can control development will hopefully lead to understanding the construction of a shape, and eventually of an organism.

Components acting in localization of bicoid mRNA are conserved among Drosophila species

Substantial insights into basic strategies for embryonic body patterning have been obtained from genetic analyses of Drosophila melanogaster. This knowledge has been used in evolutionary comparisons to ask if genes and functions are conserved. To begin to ask how highly conserved are the mechanisms of mRNA localization, a process crucial to Drosophila body patterning, we have focused on the localization of bcd mRNA to the anterior pole of the embryo. Here we consider two components involved in that process: the exuperantia (exu) gene, required for an early step in localization; and the cis-acting signal that directs bcd mRNA localization. First, we use the cloned D. melanogaster exu gene to identify the exu genes from Drosophila virilis and Drosophila pseudoobscura and to isolate them for comparisons at the structural and functional levels. Surprisingly, D. pseudoobscura has two closely related exu genes, while D. melanogaster and D. virilis have only one each. When expressed in D. melanogaster ovaries, the D. virilis exu gene and one of the D. pseudoobscura exu genes can substitute for the endogenous exu gene in supporting localization of bcd mRNA, demonstrating that function is conserved. Second, we reevaluate the ability of the D. pseudoobscura bcd mRNA localization signal to function in D. melanogaster. In contrast to a previous report, we find that function is retained. Thus, among these Drosophila species there is substantial conservation of components acting in mRNA localization, and presumably the mechanisms underlying this process.

Nuclear polyadenylation factors recognize cytoplasmic polyadenylation elements

In the cytoplasm of oocytes and early embryos, addition of poly(A) to mRNAs can activate their translation. We demonstrate that despite many differences between poly(A) addition in the cytoplasm and nucleus, these two forms of polyadenylation may involve identical trans-acting factors. Nuclear polyadenylation requires the sequence AAUAAA, the AAUAAA-binding cleavage and polyadenylation specificity factor (CPSF), and a poly(A) polymerase (PAP). We show that CPSF and PAP, purified from calf thymus, exhibit the same sequence specificity observed in the cytoplasm during frog oocyte maturation, requiring both AAUAAA and a proximal U-rich sequence. The enhanced polyadenylation of RNAs containing U-rich sequences is caused by their increased affinity for CPSF. Frog nuclear polyadenylation factors display cytoplasmic sequence specificity when dilute, suggesting that a difference in their concentrations in the nucleus and cytoplasm underlies the different sequence specificities in the two compartments. Because polyadenylation in extracts prepared from oocytes before maturation is stimulated by addition of CPSF, the onset of polyadenylation during early development may be attributable to the activation or synthesis of a CPSF-like factor. We suggest that sequences upstream of AAUAAA that are required for cleavage and polyadenylation of certain pre-mRNAs in the nucleus may be functionally equivalent to the upstream, U-rich sequences that function in the cytoplasm, enhancing CPSF binding. We propose that CPSF and PAP comprise a core polyadenylation apparatus in the cytoplasm of oocytes and early embryos.

The 3'-untranslated regions of c-mos and cyclin mRNAs stimulate translation by regulating cytoplasmic polyadenylation

Early in the development of many animals, before transcription begins, any change in the pattern of protein synthesis is attributable to a change in the translational activity or stability of an mRNA in the egg. As a result, translational control is critical for a variety of developmental decisions, including axis formation in Drosophila and sex determination in Caenorhabditis elegans. Previous work demonstrated that increases in poly(A) length can activate translation, whereas removal of poly(A) can prevent it. In this report we focus on the control of c-mos and cyclin A1, B1, and B2 mRNAs during meiotic maturation and after fertilization of frog eggs. We show that addition and removal of poly(A) from these mRNAs is extensively regulated: The time at which each mRNA receives or loses poly(A), as well as the number of adenosines it gains or loses, differ substantially. Signals in the 3'-untranslated region (UTR) of each mRNA are sufficient to reconstitute both the temporal and quantitative control of poly(A) addition: Chimeric mRNAs in which a luciferase-coding region is joined to the 3' UTRs of cyclin A1, cyclin B1, or c-mos mRNA, receive poly(A) of the same length and at the same time as do the endogenous mRNAs. Moreover, each 3' UTR also regulates translation of the chimeric mRNAs, determining when and how much translation of the luciferase reporter is stimulated during maturation. The magnitude of stimulation in luciferase activity varies from 5- to 100-fold, depending on the 3' UTR. Translational stimulation by each 3' UTR requires poly(A) lengthening, as it is prevented by mutations that prevent that process. These results suggest that the 3' UTRs of cyclin and c-mos mRNAs control not only whether or not an mRNA is turned on during maturation, but when that activation occurs and to what extent. Translational control of c-mos mRNA, which may be achieved through regulation of poly(A) length, may be critical in the activation of maturation, and in the onset of cleavage divisions. Our findings, as well as those of others, suggest that even quite complex patterns of translational activation in the early embryo can be attained through the differential control of a common mechanism.

Developmental timekeeping. Marking time with antisense

Post-transcriptional regulation of the heterochronic gene lin-14 by the binding of a natural antisense RNA controls the temporal pattern of larval development in the nematode Caenorhabditis elegans.

The deadenylation conferred by the 3' untranslated region of a developmentally controlled mRNA in Xenopus embryos is switched to polyadenylation by deletion of a short sequence element

The maternal Xenopus Eg mRNAs are adenylated and translated in the mature oocyte and then, after fertilization, are deadenylated and released from polysomes. Therefore, after fertilization, a change occurs in the cellular mechanisms that control mRNA adenylation. In the study reported here, we show that the 3' untranslated region of Eg2 mRNA contains a cis-acting element that is required for the deadenylation of chimeric RNAs after fertilization. This cis-acting element is contained within a single 17-nucleotide portion of the Eg2 mRNA. Disruption of this deadenylation element allows adenylation of the chimeric transcripts in the embryo. Therefore, this cis-acting element is part of the sequence information required for the developmental switch from adenylation to deadenylation of the maternal Eg2 mRNA in Xenopus embryos.

The deadenylation conferred by the 3' untranslated region of a developmentally controlled mRNA in Xenopus embryos is switched to polyadenylation by deletion of a short sequence element

The maternal Xenopus Eg mRNAs are adenylated and translated in the mature oocyte and then, after fertilization, are deadenylated and released from polysomes. Therefore, after fertilization, a change occurs in the cellular mechanisms that control mRNA adenylation. In the study reported here, we show that the 3' untranslated region of Eg2 mRNA contains a cis-acting element that is required for the deadenylation of chimeric RNAs after fertilization. This cis-acting element is contained within a single 17-nucleotide portion of the Eg2 mRNA. Disruption of this deadenylation element allows adenylation of the chimeric transcripts in the embryo. Therefore, this cis-acting element is part of the sequence information required for the developmental switch from adenylation to deadenylation of the maternal Eg2 mRNA in Xenopus embryos.

Arabidopsis homologs of the shaggy and GSK-3 protein kinases: molecular cloning and functional expression in Escherichia coli

The conservation in evolution of fundamental signal transduction modules offers a means of isolating genes likely to be involved in plant development. We have amplified by PCR Arabidopsis cDNA and genomic sequences related to the product of the shaggy/zeste-white 3 (sgg) segment polarity gene of Drosophila. This regulatory protein is functionally homologous to glycogen synthase kinase-3 in mammals (GSK-3), which regulates, among others, the DNA-binding activity of the c-jun/AP1 transcription factor. Analysis of PCR products led to the identification of five genes; for two of which, corresponding full-length cDNAs, ASK-alpha and gamma (for Arabidopsis shaggy-related protein kinase), were characterized. The encoded proteins were 70% identical to GSK-3 and sgg over the protein kinase catalytic domain and, after production in Escherichia coli, autophosphorylated mainly on threonine and serine residues, but phosphotyrosine was also detected. ASK-alpha and ASK-gamma also phosphorylated phosphatase inhibitor-2 and myelin basic protein, on threonine and serine, respectively. The high conservation of the protein kinases of GSK-3 family, and their action at the transcriptional level, suggest that the ASK proteins have important functions in higher plants.

Genetics of nanos localization in Drosophila

The Drosophila gene nanos is required for two processes. During oogenesis, nanos function is required for the continued production of egg chambers, and nanos is expressed in the early germarium. During embryogenesis, nanos is required maternally to specify abdominal segmentation. Nanos shares this latter function with nine other genes, collectively known as the posterior group. Of this group, nanos encodes a determinant, and is localized as an RNA to the posterior pole of early embryos. This RNA is translated to form a gradient of nanos protein with highest concentrations at the posterior. Analysis of the distribution of nanos gene products in embryos mutant for posterior group genes shows that eight of these genes are required for localization, but not stability, of the nanos RNA. Embryos mutant for posterior group alleles which produce weak abdominal phenotypes show reduced amounts of localized nanos RNA. This correlation between nanos RNA localization and abdominal phenotype suggests that nanos acts as a localization-dependent posterior determinant. Localization of nanos is not affected by mutations in bicoid or torso, confirming that the three maternal systems of anterior-posterior determination initially act independently.

Regulation of translation by specific protein/mRNA interactions

This review will focus on cases of specific translational control by protein/RNA interactions in the 5'- or 3'-UTR of eukaryote mRNA where either the cis-acting RNA determinant or the trans-acting protein (or preferably both) have been identified with fair certainty. Examples of messages that are regulated by 5' motifs, which are proposed to occlude ribosome binding when bound by their specific factors, include ferritin and ribosomal protein mRNAs and the autoregulated thymidylate synthase and poly(A)-binding mRNAs. However, it has become increasingly evident recently that 3' UTR determinants and their specific binding proteins also regulate translation efficiency either directly, or indirectly via an influence on the polyadenylation status of the mRNA. It is still unclear how events at the 3' end of mRNA influence ribosome binding. Most, if not all, of the mRNAs known to be regulated by 3' UTR motifs are subject to regulation during early development or during differentiation such as several spermatocyte and oocyte mRNAs and erythroid lipoxygenase mRNA. To date, in all cases where translation is controlled directly by specific protein/mRNA interactions, the protein seems to act as a negative regulator, a translational repressor, whose binding to the specific site on the mRNA results in inhibition of initiation. The only cases of translational activation known so far concern internal initiation of translation of picornaviral RNAs, but this topic is beyond the scope of this review.

Germ-line determination in Caenorhabditis and Ascaris: Will a helicase begin to unravel the mystery?

How cell lineages are established during development in higher eukaryotes is being addressed by geneticists and by developmental and molecular biologists. In Drosophila melanogaster, a gene corresponding to a germ-line-specific RNA helicase, vasa, has been shown to be a component o f the posteriorly localized germ granules o f the developing embryo. A putative RNA helicase, glh-I r which appears germ-line specific in its expression, has recently been reported from the free-living nematode Caenorhabditis elegans. Parasitologists studying the nematode Ascaris lumbricoides var. suum have found it to be a useful complement to Caenorhabditis. Deborah Roussell, Michael Gruidl and Karen Bennett predict that Ascaris will be valuable in determining the role played by germ-line helicases in development.

Posttranscriptional regulation of the heterochronic gene lin-14 by lin-4 mediates temporal pattern formation in C. elegans

During C. elegans development, the temporal pattern of many cell lineages is specified by graded activity of the heterochronic gene Lin-14. Here we demonstrate that a temporal gradient in Lin-14 protein is generated posttranscriptionally by multiple elements in the lin-14 3'UTR that are regulated by the heterochronic gene Lin-4. The lin-14 3'UTR is both necessary and sufficient to confer lin-4-mediated posttranscriptional temporal regulation. The function of the lin-14 3'UTR is conserved between C. elegans and C. briggsae. Among the conserved sequences are seven elements that are each complementary to the lin-4 RNAs. A reporter gene bearing three of these elements shows partial temporal gradient activity. These data suggest a molecular mechanism for Lin-14p temporal gradient formation: the lin-4 RNAs base pair to sites in the lin-14 3'UTR to form multiple RNA duplexes that down-regulate lin-14 translation.

The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14

lin-4 is essential for the normal temporal control of diverse postembryonic developmental events in C. elegans. lin-4 acts by negatively regulating the level of LIN-14 protein, creating a temporal decrease in LIN-14 protein starting in the first larval stage (L1). We have cloned the C. elegans lin-4 locus by chromosomal walking and transformation rescue. We used the C. elegans clone to isolate the gene from three other Caenorhabditis species; all four Caenorhabditis clones functionally rescue the lin-4 null allele of C. elegans. Comparison of the lin-4 genomic sequence from these four species and site-directed mutagenesis of potential open reading frames indicated that lin-4 does not encode a protein. Two small lin-4 transcripts of approximately 22 and 61 nt were identified in C. elegans and found to contain sequences complementary to a repeated sequence element in the 3' untranslated region (UTR) of lin-14 mRNA, suggesting that lin-4 regulates lin-14 translation via an antisense RNA-RNA interaction.

Translational regulation of tra-2 by its 3' untranslated region controls sexual identity in C. elegans

C. elegans hermaphrodites make sperm and then oocytes in an otherwise female animal. Gain-of-function mutations in the sex-determining gene tra-2 (tra-2(gf)) transform hermaphrodites into females (spermless hermaphrodites). The tra-2(gf) mutations map to a perfect direct repeat in the 3' untranslated region; each repeat is called a direct repeat element (DRE). Three experiments demonstrate that DREs repress tra-2 at the translational level. First, tra-2(gf) mRNAs are associated with larger polysomes than are their wild-type counterparts. Second, translation of a reporter RNA is inhibited by DREs. Third, disruption of DREs does not increase tra-2 mRNA levels. An RNA binding activity specifically associates with the DREs. We propose that tra-2 translation is inhibited by association of an RNA binding-factor with the DREs and that this translational control is essential for development of C. elegans as a hermaphrodite/male species.

glh-1, a germ-line putative RNA helicase from Caenorhabditis, has four zinc fingers

We have cloned a family of putative RNA helicases from the free-living nematode Caenorhabditis elegans. One of these, a cDNA that we call glh-1, most closely matches in sequence and expression the previously described germ-line helicases PL10 from mouse and vasa from Drosophila. The amino terminus of the predicted protein of glh-1 contains a set of glycine-rich repeats similar in location and sequence to those in the predicted vasa protein. However, unlike all other putative RNA helicases, glh-1 also contains four retroviral-type zinc fingers. The RNA expression pattern of this Caenorhabditis helicase correlates with the presence of germ-line tissue in the parasitic nematode Ascaris lumbricoides var. suum and with the presence of germ cells in wild type and several germ-line mutants of Caenorhabditis. In the germ-line mutants glp-4 and glp-1, additional larger species of glh-1 RNA exist, which correspond to different adenylylated forms of the glh-1 transcript; these may be specified by motifs in the 3' untranslated region of glh-1 that are similar to adenylylation control elements and nos response elements.

Localized RNAs and their functions

The eukaryotic cell is partitioned by membranes into spatially and functionally discrete subcellular organelles. In addition, the cytoplasm itself is partitioned into discrete subregions that carry out specific functions. Such compartmentation can be achieved by localizing proteins and RNAs to different subcellular regions. This review will focus on localized RNAs, with a particular emphasis on RNA localization mechanisms and on the possible biological functions of localization of these RNAs. In recent years, an increasing number of localized RNAs have been identified in a variety of cell types among many animal species. Emphasis here will be on localized RNAs in the most intensively studied systems-Drosophila and Xenopus eggs and early embryos.

Distribution of tudor protein in the Drosophila embryo suggests separation of functions based on site of localization

Mutations in the tudor locus of Drosophila affect two distinct determinative processes in embryogenesis; segmentation of the abdomen and determination of the primordial germ cells. The distribution of tudor protein during embryogenesis, and the effect of various mutations on its distribution, suggest that tudor protein may carry out these functions separately, based on its location in the embryo. The protein is concentrated in the posterior pole cytoplasm (germ plasm), where it is found in polar granules and mitochondria. Throughout the rest of the embryo, tudor protein is associated with the cleavage nuclei. Mutations in all maternal genes known to be required for the normal functioning of the germ plasm eliminate the posterior localization of tudor protein, whereas mutations in genes required for the functioning of the abdominal determinant disrupt the localization around nuclei. Analysis of embryos of different maternal genotypes indicates that the average number of pole cells formed is correlated with the amount of tudor protein that accumulates in the germ plasm. Our results suggest that tudor protein localized in the germ plasm is instrumental in germ cell determination, whereas nuclear-associated tudor protein is involved in determination of segmental pattern in the abdomen.

Formation of germ cells in Drosophila

In Drosophila, germ-cell formation is directed by the posterior pole plasm of the embryo. Genes acting to provide the germ-cell determinant have been identified and characterized, and some are now being manipulated to test their roles. Although recent results largely support the notion of a simple pathway for assembly of pole plasm, complexities are becoming apparent.

Segmentation of the Drosophila embryo

Segmentation in Drosophila is a sequential process directed by at least 30 genes that encode various types of proteins, including: many transcription factors; a putative RNA-binding protein; a membrane-associated receptor kinase; several intracellular protein kinases; a number of secreted signaling molecules; and others of unknown function. Although the detailed molecular reactions used to generate the metameric subdivisions of the embryo are not yet understood, a general outline of the processes involved has been described. The manner in which spatial relations in the developing embryo are established can now be described.

The torso receptor localizes as well as transduces the spatial signal specifying terminal body pattern in Drosophila

Specification of the end portions of the Drosophila body depends on the torso (tor) protein, a receptor tyrosine kinase that accumulates uniformly along the entire surface of the embryo but is activated only in the vicinity of the poles. Several genes are normally required for activating tor and appear to define a system in which a gene product tethered to the extracellular vitelline membrane at each end of the egg provides a local source for an extracellular tor ligand. This ligand would have to diffuse from the membrane to the cell surface of the embryo without losing its spatial localization. Here we report that the failure to accumulate tor protein at one or both poles leads to spatially inappropriate activity of more centrally located receptor. This ectopic activity depends on the same gene functions normally required for activating tor; thus we infer that it reflects inappropriate diffusion of the ligand to more central regions of the body. We conclude that the receptor not only transduces the spatial signal imparted by the tor ligand, but also ensures its correct localization by sequestering the ligand. Ligand trapping by receptor may also localize spatial signals in other patterning systems, including specification of the dorsal-ventral axis in Drosophila and of vulval cell fates in Caenorhabditis elegans.