Many mRNAs specifically localize within the cytoplasm and are present in RNA-protein complexes. It is generally assumed that localization and complex formation of these RNAs are controlled by trans-acting proteins encoded by genes different than the RNAs themselves. Here, we analyze slow as molasses (slam) mRNA that prominently colocalizes with its encoded protein at the basal cortical compartment during cellularization. The functional implications of this striking colocalization have been unknown. Here, we show that slam mRNA translation is spatiotemporally controlled. We found that translation was largely restricted to the onset of cellularization when Slam protein levels at the basal domain sharply increase. slam mRNA was translated locally, at least partially, as not yet translated mRNA transiently accumulated at the basal region. Slam RNA accumulated at the basal domain only if Slam protein was present. Furthermore, a slam RNA with impaired localization but full coding capacity was only weakly translated. We detected a biochemical interaction of slam mRNA and protein as demonstrated by specific co-immunoprecipitation from embryonic lysate. The intimate relationship of slam mRNA and protein may constitute a positive feedback loop that facilitates and controls timely and rapid accumulation of Slam protein at the prospective basal region.
While proteins and their encoding messenger RNAs share the same intracellular space during the translation process, thereafter they are usually spatially and biochemically separated. RNA localization follows a specific subcellular pattern—such as apical or basal—and is thought to have important physiological implications during development, which are generally independent from the protein function. Here, we investigate the potential mutual dependencies between slam mRNA and its encoded protein during cellularization in early Drosophila embryos. slam RNA and protein are known to colocalize and are essential for epithelial compartmentalization and timely invagination of the plasma membrane between adjacent nuclei. We now show that Slam protein is required for RNA localization at the basal domain and that this event is needed for efficient translation. In addition to the functional interactions, we find that slam RNA and protein are both present in a specific molecular complex. Our findings indicate that slam is locally translated and that the interaction between Slam protein and RNA constitutes a self-enhancing mechanism leading to the fast accumulation of Slam protein at the basal domain during the first minutes of cellularization.
Citation: Yan S, Acharya S, Gröning S, Großhans J (2017) Slam protein dictates subcellular localization and translation of its own mRNA. PLoS Biol 15(12): e2003315. https://doi.org/10.1371/journal.pbio.2003315
Academic Editor: Simon Bullock, MRC Laboratory of Molecular Biology, United Kingdom of Great Britain and Northern Ireland
Received: June 16, 2017; Accepted: November 14, 2017; Published: December 4, 2017
Copyright: © 2017 Yan et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files. Numericals of quantitative data shown in the figures can be found in the file S1 Data. The data from Next Generation Sequencing are accessible through GEO Series accession number GSE99761 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE99761).
Funding: Göttingen Center for Molecular Biology (grant number). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Deutsche Forschungsgemeinschaft www.dfg.de (grant number SFB860/TP A9 and INST1525/16-1 FUGG). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: ACU, alternative codon usage; CDS, coding DNA sequence; cncRNA, RNA with coding and noncoding functions; CS, vector plasmid CS2; CyHx, cycloheximide; FC, furrow canal; FRAP, fluorescence recovery after photobleaching; Gal4, galactose-responsive transcription factor; GFP, green fluorescent protein; kD, kilodalton; MW, molecular weight; PCP, bacteriophage PP7 coat protein; PP7, bacteriophage PP7; qPCR, quantitative polymerase chain reaction; S2 cells, Drosophila melanogaster Schneider 2 cells; slam, slow as molasses; TEV, Tobacco Etch Virus nuclear-inclusion-a endopeptidase; UAS, Upstream Activation Sequence; UTR, untranslated region; WT, wild-type
Subcellular RNA localization is a widespread phenomenon [1–4]. A large-scale survey of RNA localization in Drosophila embryos by RNA in situ hybridization with fluorescent probes revealed that about 70% of all tested transcripts were distributed in a specific subcellular pattern, such as apical or basal localization . The physiological relevance of RNA localization is unknown for most of these transcripts, however. The function and mechanism of RNA localization have been studied in detail in numerous cases. Specific localization is usually mediated by cis-acting elements within the transcript and trans-acting factors [6–9]. Trans-acting factors are encoded by transcripts different than the localizing RNA, in most cases.
We reported previously that slow as molasses (slam) RNA and protein strikingly colocalize during early Drosophila development . However, the function and underlying mechanism of slam mRNA-protein colocalization have so far not been analyzed. slam is a key player in the change from syncytial to cellular development [11–14]. slam is specifically expressed during this stage and is required for furrow invagination and separation of cortical domains during cellularization and, later, for germ cell migration [10, 15, 16]. slam mRNA and protein strongly accumulate at the basal cortical domain, which forms the so-called furrow canal (FC). For the specification of the basal domain, slam functions redundantly with nullo. Markers of the FC such as Dia are present in slam mutants but are uniformly distributed in slam nullo double mutants .
An important feature of the transition from syncytial to cellular development is the rapid and coordinated change of several processes within a few minutes only. As zygotic gene expression gradually increases during the course of many minutes, additional posttranscriptional mechanisms potentially control the temporally and spatially regulated activity of the key players such as slam to ensure a switch-like change in behavior of the cellular processes from nuclear cycle 13 to cycle 14.
Here, we investigate the functional and molecular interactions of slam mRNA and protein and address the function of slam RNA localization and RNA-protein colocalization. In addition to functional interactions, we identified a specific biochemical interaction, in that Slam protein specifically coprecipitated with slam mRNA.
Temporal and spatial regulation of Slam protein expression
slam RNA and protein (about 4,000 nucleotides and 1,173 amino acid residues, S1 Fig) mark the basal region of the cellularization furrow, the FC (Fig 1A and 1B) . Morphologically visible furrows emerge within a few minutes after the last mitosis by invagination of the plasma membrane between adjacent nuclei (Fig 1A). slam RNA and protein colocalization was also observed when Slam was ectopically localized (Fig 1B). We achieved ectopic Slam protein localization by employing embryos from nuf females, which are impaired for the recycling endosomes [17, 18]. In these embryos, we detected slam RNA together with Slam protein at the apical membrane. These data show that slam RNA and protein colocalize also in situations of ectopic localization and indicate that the mechanisms controlling slam RNA and protein localization are interconnected.
(A) Schematic drawing of cellularization. slam RNA (red) and protein (green) mark the basal cortical domain at the tip of the metaphase furrow and the cellularization furrow. Orange marks RNA protein colocalization. The metaphase furrow transforms into the cellularization furrow. At the onset of cellularization, a new furrow emerges between daughter nuclei and becomes indistinguishable from the old furrows soon after. Furrow invagination is absent in slam mutants. (B) Fixed wild type and embryos from nuf females stained for slam RNA (grey/red), Slam protein (grey/green), and DNA (blue). Yellow arrows point to mislocalized slam RNA and protein in nuf embryos. Insets are images with increased contrast. Scale bars = 10 μm. (C) Quantification of fluorescence at the FC after fixation and staining for slam RNA and protein. Embryos were manually staged according to morphology, i.e., furrow depth and nuclear shape/length and its corresponding time relative to mitosis 13. Error bars indicate standard error of the mean. N = 3 embryos. About 10 furrows were scored in each embryo. The underlying data for this figure can be found in S1 Data. a. u., arbitrary units.
Previous expression and histological analysis indicated an up-regulation of slam RNA and protein during the blastoderm stage [10, 15, 16]. Expression analysis by NanoString technology defined the window of strong up-regulation to nuclear cycle 13 and early cellularization (cell cycle 14) . We confirmed these data by measuring total RNA and protein levels. We found an about 10-fold up-regulation of total RNA levels during cellularization (2–3 h) by quantitative PCR (S2A Fig) and a peak of total protein levels in extracts from embryos in cellularization (S2B Fig). Slam is required for formation of the cellularization furrows [15, 16] and identity of the basal domain . As slam RNA and protein are strongly enriched at the FC, we next analyzed accumulation of slam RNA and protein specifically at the FC in fixed and stained embryos, which were staged by nuclear cycle and by the length of the furrow and nuclei (S2C Fig). By quantification of the fluorescence signal at the FC, we observed a strong (6-fold) up-regulation of both localized RNA and protein within a short period of a few minutes at the transition from mitosis 13 to interphase 14 (Fig 1C). Following this up-regulation, protein levels remained constant, whereas RNA levels gradually decreased after 30 min during the second half of cellularization (Fig 1C).
Based on the colocalization of RNA and protein, we hypothesized that translation of slam might be linked to RNA localization. To address this hypothesis, we first defined Slam protein stability and the timing of translation. slam RNA may be translated with a constant rate throughout cellularization or specifically during onset of cellularization. To separate the contribution of translation and degradation to the steady state levels of Slam protein, we first measured the stability of green fluorescent protein (GFP)-slam protein during cellularization (Fig 2A, S1 Fig). GFP-slam serves as a proxy for untagged Slam, because GFP-slam can rescue the cellularization phenotype of slam mutants. Furthermore, the dynamics of GFP-slam largely reflect the dynamics of Slam protein . To measure the half-life of GFP-slam, we injected cycloheximide into embryos in cellularization to stop new translation (S3A–S3C Fig) and recorded the persistence of GFP-slam fluorescence. From the measured decay of GFP-slam during periods of 10 min, we extrapolated an estimated half-life of approximately 40 min (Fig 2A and 2B). A half-life of 40 min is in the range of the length of cellularization and thus indicates that GFP-slam is quite a stable protein. Together with the relatively stable expression levels of GFP-slam as well as endogenous Slam protein (Fig 1C, S2B Fig), the long half-life indicates a low translation rate of slam mRNA after the initial phase of cellularization. We could not measure the role of GFP-slam synthesis for the dynamics of GFP-slam during the onset of cellularization, as injection of cycloheximide would induce a mitotic arrest.
Wild-type embryos (A—C) injected with buffer or CyHx or embryos (D) from shibire heterozygous females shifted to nonpermissive temperature in mid-cellularization, expressing GFP-slam. (A) Images from time-lapse recordings, pre-bleach, post-bleach, and 10 min after bleach. Scale bars = 10 μm. (B) Time course of GFP-slam fluorescence following injection of CyHx. Corresponds to upper images in (A). Approximate half-life (t1/2) of GFP-slam during the measurement period was calculated by assuming an exponential decay. (C) Time course of GFP-slam fluorescence following injection and photobleaching within the bleached area (circle in yellow). Corresponds to middle and lower row of images in (A). (D) Time course of GFP-slam fluorescence following photobleaching in the bleached area (circle in yellow). (C, D) Recovery rate (m) was calculated by linear fitting. (D) Fluorescent trace was normalized to unbleached area. The underlying data for this figure can be found in S1 Data. a. u., arbitrary units.
Secondly, we analyzed the mobility of GFP-slam by quantifying fluorescence recovery after photobleaching (FRAP). Previously, we found that the recovery of GFP-slam fluorescence dramatically changes from fast and complete during the onset of cellularization to slow and incomplete recovery during the course of cellularization . The slow fluorescence recovery after the initial phase of cellularization may be due to the exchange of bleached and unbleached molecules by mobile GFP-slam molecules. Alternatively, fluorescence recovery may be due to translation of new GFP-slam molecules. We distinguished these 2 options by FRAP experiments in embryos, in which new GFP-slam synthesis was blocked by cycloheximide. We observed a comparable recovery rate with and without cycloheximide. Thus, new GFP-slam translation does not contribute to fluorescence recovery during cellularization (Fig 2A and 2C). These data are consistent with the long half-life of GFP-slam during cellularization. Furthermore, the slow exchange of Slam molecules did not require vesicle budding, because a mutation in Dynamin (shibire)  had little influence on the recovery rate (Fig 2D). In summary, our data suggest that Slam protein is largely synthesized during a short period of a few minutes at the transition from mitosis 13 to interphase 14. Following this initial phase, Slam protein is subject to low turnover during the remainder of cellularization.
Not yet translated slam mRNA transiently accumulates at the prospective basal domain
Next, we addressed the spatial and temporal dimension of Slam protein accumulation at the basal domain. Protein accumulation may be due to recruitment of Slam protein from the cytoplasm to the basal compartment or, alternatively, to local translation of slam RNA. To distinguish these 2 options, we employed a recently developed method to fluorescently label not yet translated mRNA molecules [21, 22]. We generated genomic transgenes with a bacteriophage PP7 (PP7) hairpin loop inserted in the slam coding sequence (Fig 3A, S1 Fig). We inserted the PP7 sites close to the stop codon of the mRNA in order to assay all slam transcripts, including the mRNAs, which have initiated but not completed translation. Corresponding transcripts bind a bacteriophage PP7 coat protein (PCP)-GFP marker protein only until a ribosome has moved over the PP7 site [21, 22]. Thus, only not yet translated transcripts are labelled by PCP-GFP fluorescence. The slamPP7 RNA shows a spatial and temporal expression pattern similar to endogenous slam RNA, persisting until the second half of cellularization (S2D Fig). In fixed embryos, we detected a dotted PCP-GFP signal (Fig 3B). PCP-GFP staining largely colocalized with Slam protein at the basal region. The staining was dynamic and only visible in embryos in early cellularization. Time-lapse imaging allowed a precise timing of PCP-GFP dynamics (Fig 3C). The dotted signal was largely observed during the onset of cellularization and quickly disappeared during the following minutes. These data show that a proportion of slam mRNA molecules reaches the basal domain before the first ribosomes have passed 3′-located PP7 sites and the first round of translation has been completed. Correspondingly, these data suggest that the synthesis of at least a fraction of the Slam protein molecules is completed at the FC. As the localized PCP signal quickly disappears, much fewer or no new and not yet fully translated slam RNA molecules reach the FC during cellularization.
(A) Experimental scheme. Not yet translated slam mRNA is labeled by PCP-GFP binding to a PP7 site within the coding sequence prior to the first round of translation. During the first round of translation, PCP-GFP dissociates from slamPP7 RNA. (B) Fixed embryos zygotically expressing slamPP7 stained for PCP-GFP (green), Slam protein (red), and DNA (blue). Sagittal sections (Stage) allow staging. Scale bar = 10 μm. Inset at 2.25× magnification. (C) Images from time-lapse recording for PCP-GFP in embryos with zygotic expression of slamPP7. Time 0:00 min is approximately at mitosis 13. Scale bar = 5 μm.
Slam protein is required for slam RNA localization
Given the peculiar temporal and spatial restrictions of slam translation, we functionally analyzed the role of RNA protein colocalization. We established a complementation assay either with injection of in vitro transcribed RNA or with zygotic expression from a transgene (Fig 4A). Injected slam RNA with a fluorescent label accumulated at the basal region as revealed by time-lapse imaging (Fig 4B, S1 Movie). Localization of the injected RNA and its derived protein was confirmed by staining for RNA and protein in fixed embryos. We did not observe an influence of GFP or myc tags on RNA and protein localization (S4A Fig). RNA and protein localization was also reconstituted in embryos lacking any endogenous slam RNA and protein (S4A Fig). slam deficient embryos were recognized by the absence of endogenous slam RNA and protein and staining at the injection site. These experiments demonstrate that the localization of slam RNA and colocalization of slam RNA and protein can be reconstituted in vivo.
(A) Experimental scheme. Wild-type embryos were injected with synthetic slam mRNAs. After incubation, embryos were imaged or fixed and stained. (B) Fluorescently (schematically indicated by *) labeled slam RNA. Images from time-lapse recording. Arrows point to RNA at the FC. (C) Crossing scheme for generation of slam deficient (Δslam) embryos. Half of the embryos receive the Δslam allele and are maternally and zygotically deficient for slam; the other half receives a wild-type allele from the male and is zygotically rescued. (D) GFP-stop-slam RNA was injected into embryos from slam germ line clones. Fixed embryos were stained for the injected GFP-stop-slam RNA (grey/red), Slam protein (grey/green), and DNA (blue). Inset: areas marked by dashed box are shown at 2.25× magnification. Scale bars = 10 μm.
In a simple model, slam RNA would accumulate independently of Slam protein at the basal region. Alternatively, Slam protein may be involved in the localization of its mRNA. To distinguish these 2 options, slam RNA with an early stop codon (GFP-stop-slam RNA) (Fig 4C) was introduced into embryos that are maternally and zygotically deficient for the slam locus (“m−z−”). Such embryos were derived from females with slam germ line clones crossed to slam heterozygous males. Fifty percent of the embryos from such a cross are slam (”m−z−“) deficient and 50% of the embryos zygotically express slam (“m−z+,” zygotic rescue) (Fig 4C). These 2 genotypes were easily distinguished by staining for endogenous Slam protein and by their morphology. The zygotically rescued embryos served as an internal reference for the injection and staining procedure. We could score for localization of the injected RNA also in slam deficient embryos, because the FC/basal domain is specified even in the absence of slam . We did not detect localizing RNA in slam deficient embryos. The signal for GFP-stop-slam RNA was uniformly distributed at the apical surface. In contrast, in control embryos, the injected RNA colocalized with endogenous Slam protein (Fig 4D). The GFP-stop-slam RNA localization in control embryos may be due to an interaction with Slam protein or with the endogenous slam RNA. We favor the first option, because injected slam RNA encoding functional Slam protein is sufficient for slam RNA and protein localization in embryos lacking endogenous slam RNA (S4A Fig). Taken together, these experiments show that Slam protein is required for slam RNA localization at the FC.
Slam protein has an intrinsic RNA-independent affinity for the FC
Next, we asked whether slam RNA localization is required for protein localization. For this, we mapped multiple regions within the 5′ untranslated region and the coding sequence, which are sufficient for localization at the FC (S4 Fig). Given the complexity of multiple parts contributing to RNA localization, we generated a novel slam gene, slam alternative codon usage ([ACU]), in which a majority of the codons were replaced by synonymous codons (Fig 5A, S1 and S5 Figs). slam[ACU] RNA lost the ability to localize to the FC (Fig 5A and 5B). slam[ACU] RNA was uniformly distributed in the cytoplasm in the presence or absence of endogenous slam, when expressed from a transgene (myc-GFP-slam[ACU]) or injected as synthetic RNA (myc-slam[ACU]) (Fig 5A and 5B). These data indicate that our mutagenesis strategy successfully impaired the FC localization of slam RNA.
(A—C) Fixed embryos stained for slam or slam[ACU] RNA (grey/red), Slam protein (grey/green), and DNA (blue), as indicated. (A) Myc-GFP-slam[ACU] was expressed in zygotically rescued and slam deficient embryos from a transgene driven by maternal Gal4. (B) Zygotically rescued and slam deficient embryos from slam germ line clones injected with myc-slam[ACU] RNA. Arrows point to Slam protein localization at the FC. (C, E) slam[ACU] genomic transgene was expressed in zygotically rescued (m−z+) and slam deficient embryos (m−z−) from slam germ line clones. (C) Arrows point to slam[ACU] RNA and Slam protein localizing at the FC. WT embryos were stained for endogenous slam RNA and protein in a separate tube. (D) Quantification of slam WT (endogenous gene) and ACU RNA (from transgene) abundance by reverse transcription and qPCR within the same embryos (2 copies of slam[WT] and 2 transgenic copies of slam[ACU]. Expression of slam[ACU] was normalized to slam[WT]. (E) Quantification of fluorescence staining for Slam protein at the FC for images shown in panel C. Control (m−z+), ACU (m−z−with ACU transgene). Three embryos with 10 furrows each, student t test P = 3.5 × 10−29. Arbitrary units. (F) Western blot of extracts from manually staged and genotyped embryos from slam germ line clones with or without expression of slam[ACU] from the genomic transgene. Zygotically rescued embryos express Slam from the endogenous gene. Scale bars = 10 μm. Error bars indicate standard deviation. The underlying data for this figure can be found in S1 Data. kD, kilodalton; MW, molecular weight; N, number of biological replicates; qPCR, quantitative polymerase chain reaction.
Having generated a nonlocalizing slam[ACU] RNA, we could test whether Slam protein accumulation at the FC would depend on RNA localization. We stained embryos injected with the nonlocalizing slam[ACU] mRNA for Slam protein (Fig 5B). slam embryos were recognized by the absence of overall slam RNA or protein signal. Staining restricted to the injected site is due to the injected construct. We clearly detected Slam protein at the FC in slam embryos (Fig 5B). We conclude that Slam protein has an intrinsic RNA-independent affinity for the FC. These data also indicate that we preserved the coding capacity of slam[ACU].
Translational control of slam at the FC
We got the impression that protein levels with slam[ACU] were lower than with wild-type slam. The difference may be due to inefficient translation of slam[ACU], to inefficient localization of Slam protein, or to the injection procedure resulting in lower mRNA levels than endogenous slam expression.
In order to test translation efficiency in embryos, we generated a genomic transgene with the slam[ACU] sequence, preserving introns, 5′ untranslated regions, and 3′ untranslated regions (Fig 5C, S1 Fig). A corresponding genomic transgene with the endogenous coding sequence fully complements a slam deficiency . slam[ACU] RNA expressed from the genomic transgene was comparably abundant as the endogenous wild-type allele (Fig 5D), confirming the integrity of the transgene and transcript. slam[ACU] RNA showed a low degree of FC localization in wild-type background and slightly more so in slam deficient embryos (Fig 5C). This low degree of localization is likely due to the 5′ untranslated region, which is sufficient to localize at the FC in the injection assay (S4B Fig).
Next, we assayed protein levels by embryo staining and western blot with total extracts. Staining for Slam protein in slam[ACU] embryos in comparison to rescued siblings revealed strongly reduced protein levels at the FC (Fig 5C and 5E). A similarly clear difference in total protein levels was detected by western blot (Fig 5F). For this, we manually sorted embryos according to stage (mid-cellularization) and genotype. No Slam was detected in slam deficient embryos (Fig 5F). slam[ACU] embryos contained much less Slam protein than the rescued siblings, which zygotically expressed Slam from the endogenous gene (Fig 5F). These data indicate Slam protein derived from slam[ACU] is much less abundant and that slam[ACU] was thus much less translated than wild-type slam.
The reduced translation may be due to secondary RNA structures or codon usage affecting translation efficiency. We distinguished these options by expression of slam[ACU] in comparison to slam[WT] in cultured Drosophila melanogaster Schneider 2 cells (S2 cells). S2 cells do not express Slam protein in detectable levels (S6A Fig). Following transient transfection, we detected comparable slam RNA and protein levels for slam[ACU] and wild-type slam (S6B and S6C Fig). These data indicate that generic translation in S2 cells is comparably efficient for slam[ACU] and wild-type slam. These data do not rule out the conceivable option that slam[ACU] contains secondary RNA structures or peculiar codon usage affecting translation efficiency that are specifically present in the embryo but not in S2 cells.
Consistent with the reduced protein expression from slam[ACU] in embryos, the slam[ACU] genomic transgene only partially complemented the cellularization phenotype and did not rescue the lethality of a slam deficiency. Embryos maternally and zygotically deficient for slam but with slam[ACU] only formed a short cellularization furrow and did not complete cellularization (S7 Fig). This functional test showed that efficient expression of slam is physiologically important. Taken together, these data suggest that localization of slam RNA at the basal domain or its colocalization with Slam protein are important for full translation.
Slam protein attracts slam mRNA to the FC
The transient FC accumulation of not yet translated slam RNA suggests a local translation of slam, at least partially (Model 1, Fig 6A). In addition to local translation, slam RNA may recruit Slam protein to the basal region, which was synthesized by translation within the cytoplasm (Model 2, Fig 6A). To functionally assess the significance of Model 2, we injected a translation-incompetent but localization-competent slam RNA (GFP-stop-slam) into the slam[ACU] embryos (Fig 6B). We would expect localization of the injected RNA and in the case of Model 2, a corresponding local increase of Slam protein levels.
(A) Two models for accumulation of slam RNA and protein. RNA (red) with ribosomes, protein (green dots). Protein is synthetized by local translation after the RNA was attracted by protein (Model 1). Protein is synthesized in the cytoplasm and attracted by localized RNA (Model 2). (B) Test of these models by a nontranslatable, localizing RNA (red) and a translatable, localization-impaired RNA (orange). Embryos from slam germ line clones with genomic slam[ACU] injected with GFP-stop-slam RNA and stained for the injected RNA (grey/red), Slam protein (grey/green), and DNA (blue). Insets are at 2.25× magnification. Yellow arrows point to localized RNA and protein. * marks posterior pole cells. Scale bars = 10 μm.
We detected the injected GFP-stop-slam RNA at the FC in levels comparable to slam in control embryos (Fig 6B). Slam protein staining was uniform in low levels, similar to slam[ACU] embryos (Fig 6B). We did not detect any increased signal at the injection site. These data suggest that localized slam RNA does not attract cytoplasmic Slam protein to the FC.
Biochemical association of slam mRNA and protein
A key feature of slam is, on the one side, the functional interdependence of RNA and encoded protein and, on the other side, the temporally and spatially restricted translation. We hypothesized that the functional interaction of slam RNA and protein is based on a biochemical interaction. By Slam immunoprecipitation, we found that slam mRNA was enriched in the bound fraction as compared to immunoprecipitates by Dia antibodies (Fig 7A and 7B). The formin Dia nucleates and elongates F-actin and is enriched at the FC [23–25]. The enrichment of slam RNA was higher than for a series of control RNAs (Fig 7A). We confirmed the specific association of slam RNA with Slam protein by conducting immunoprecipitation with single-chain GFP antibody (GFP-binder) and lysates of embryos expressing GFP-slam from a genomic transgene in comparison to wild-type extracts (Fig 7C and 7D). We found a higher enrichment of slam mRNA than other RNAs in the bound fraction (Fig 7C). This enrichment confirmed the specificity of the RNA-protein interaction.
Immunoprecipitation from (A, B) wild-type extracts with Slam and Dia antibodies or from (C, D) GFP-slam and wild-type extracts with GFP antibody. (A, C) Associated RNA of selected genes were quantified by RT-qPCR. Error bars indicate standard error of mean. The underlying data for this figure can be found in S1 Data. (B, D) Slam and GFP-slam proteins were detected by western blot with Slam antibody. (E) Transcripts associated with immunoprecipitates were identified by next generation sequencing. An enrichment factor was calculated for each transcript as the ratio of transcript numbers between immunoprecipitates (1) with Slam and Dia antibody and wild-type lysates or (2) with GFP binder and lysates from GFP-slam and wild-type embryos. Indicated are the number of transcripts above threshold (4-fold enrichment). The intersection contains 3 transcripts. ΔCq, difference in qPCR cycles; GFP, green fluorescent protein; N, number of biological replicates; RT-qPCR, reverse transcription quantitative polymerase chain reaction.
To assess whether the bound fraction contained other mRNAs beside slam RNA, we identified associated transcripts in an unbiased manner by next generation sequencing (Fig 7E). RNA was isolated and sequenced from the bound fractions of the immunoprecipitation experiments with Slam antibody and GFP-binder, including the controls. In the bound fractions, we detected 102 and 16 transcripts, respectively, which were enriched more than 4-fold (Fig 7E, S1 Table). The intersection of the 2 experimental approaches contained slam with the strongest enrichment and 2 more unrelated transcripts. These data indicate that slam mRNA is present in a specific complex with Slam protein in embryos. Although slam is among the most abundant transcripts during this stage, none of the other abundant transcripts were enriched in our immunoprecipitation experiments (S1 Table). As the primary structure of Slam does not contain any obvious structural domains related to RNA binding, other RNA binding proteins likely mediate the biochemical association of slam RNA and protein.
The primary function of slam RNA is encoding Slam protein. In addition to coding information, slam RNA contains 2 more pieces of noncoding information: (1) information for specific subcellular localization of the RNA at the FC, which is at least partially mediated by an interaction with Slam protein, and (2) information for spatial and temporal control of translation, which is high at the FC during the onset of cellularization. By comparing wild-type RNA with a variant RNA, slam[ACU] with the same coding information, the relevance of the noncoding information was uncovered. slam[ACU] RNA is widely distributed in the cell and gives rise to much less Slam protein. Containing coding and noncoding information distinguishes slam RNA from generic mRNAs, which contain only coding information. slam RNA may be related to the growing class of mRNAs with coding and noncoding functions (cncRNA) .
Using the injection assay with synthetic RNA transcribed from truncation constructs, we identified several regions of slam RNA that are sufficient for localization to the FC in wild-type embryos. This includes the 5′ untranslated region and at least 2 large regions within the coding sequence, which we have so far not further defined. Each of these regions is able to localize to the FC on its own in wild-type embryos, which indicates redundancy in the mechanism of RNA localization. Interpretation of these data is difficult, however, given that endogenous slam RNA and protein were present in our assay, which may lead to localization by oligomerization or other indirect binding mechanisms.
slam RNA is subject to spatiotemporal control of translation. Although the RNA is present in high levels during the first half of cellularization, translation is restricted to the initial minutes of cellularization. The almost constant protein levels throughout cellularization are due to the stability of Slam protein, as inhibition of new synthesis by cycloheximide leads only to a weak decrease of GFP-slam fluorescence. In contrast to these constant levels during cellularization is the sharp increase in protein levels during the onset of cellularization.
This initial rise in protein levels is partly due to the transcriptional up-regulation of slam. The transcriptional regulation appears not to be sufficient, as we observed a striking difference between slam[wild-type] RNA with slam[ACU] RNA. Although both RNAs contain the same coding information and give rise to similar amounts of Slam protein in cultured cells, slam[wild-type] RNA is more efficiently translated than slam[ACU] RNA in blastoderm embryos. Based on the correlation of impaired RNA localization and reduced translational efficiency, we favor the model that efficient translation is linked to RNA localization or interaction with Slam protein at the FC. However, the embryo-specific lower efficiency of slam[ACU] translation may alternatively be due to secondary RNA structures or disadvantageous codons, which were introduced in our mutagenesis.
A particular feature of slam is that the encoded protein is required for the elaboration of noncoding features. slam RNA requires Slam protein for accumulation at the FC. Not only do we observe a functional interaction of the RNA and protein but also colocalization and biochemical association, indicating molecular interactions. These molecular interactions between RNA and protein may be direct or indirect. They are likely to be indirect, as Slam protein does not contain a canonical RNA binding domain or does not share any detectable sequence homology to RNA binding proteins. Analysis of transcripts associated with Slam protein provided distinct lists depending on the experimental procedure. Importantly, slam RNA was identified by both approaches, which is consistent with our quantitative polymerase chain reaction (qPCR) analysis for a few selected genes. The list of associated transcripts contained transcripts with high and low abundance, indicating that the procedure was sufficiently sensitive to also detect weakly expressed genes. Screening through the gene functions, we did not detect a specific set of genes, such as genes involved in cellularization or cytoskeletal organization.
The biological function of the intimate relationship of slam mRNA and its protein has been unclear. Given the observed specific spatiotemporal profile of slam translation and Slam dynamics, we propose the following model (Fig 8A). Initially, Slam protein accumulates in low levels at the FC independently of its mRNA. Starting with the onset in zygotic transcription, slam mRNA exits the nucleus as part of a complex that is not competent for efficient translation. At least a fraction of slam RNA molecules accumulates at the basal domain forming the FC prior to the first round of translation. At the basal domain, slam mRNA becomes competent for efficient translation, which leads to an increase in Slam protein at the FC. The increased amounts of Slam protein further promote accumulation of slam mRNA, leading to even more Slam protein. Such a mechanism constitutes a positive feedback loop, which provides an explanation for the observed switch-like profile of Slam protein staining at the FC (Fig 8B). Some minutes later, when full Slam levels have been reached, translation is turned off. Slam protein then functions in spatially restricted activation of Rho signaling, actomyosin organization, Patj recruitment, cortical compartmentalization, and furrow invagination (Fig 8) [10, 17, 25]. As slam is a key regulator of cellularization, the rapid increase of Slam protein may be important for a coordinated and timely onset of its downstream processes.
(A) Conceptual steps for accumulation of slam RNA and protein at the basal compartment/prospective FC. (B) Scheme for time course of Slam accumulation at the FC with a positive feedback or according to a linear model. Mit13, mitosis of cycle13; Inter 14, interphase of cycle 14.
We identified slam as an mRNA with noncoding information for localization and translational control in addition to its coding information. On the molecular level, slam is special in that mRNA and protein associate in a complex as demonstrated by co-immunoprecipitation. This unconventional relationship of slam RNA and protein may be important for the tightly restricted subcellular localization and strong increase in protein levels at the FC. The functional interaction of slam RNA and protein constitutes a positive feedback loop, which contributes to the fast increase in Slam protein levels at the FC.
Materials and methods
Fly stocks were obtained from the Bloomington stock center, if not otherwise noted. The following fly strains and mutations were used: Df(2L)slam Frt2L slam5′rescue , nuf1 , shi . The following transgenes were used: slam[wild-type], genomic transgene with the slam locus , GFP-slam, genomic transgene with GFP at N-terminus, slam[ACU], genomic transgene with alternative coding sequence, slamPP7, genomic transgene with PP7 sites inserted within the coding sequence, UASp-myc-GFP-slam[wild-type], UASp-myc-GFP-slam[ACU], transgenes with GAL4 driven expression. Genomic transgenes were generated by PhiC31 integrase-mediated site-specific insertions on the third chromosome at cytological position 68A4 . slam deficient embryos were generated by slam germ line clones in progenies of the cross Df(2L)slam Frt2L slam5′rescue/CyO females with hs-Flp122; ovoD2L Frt2L/CyO males and heat shock for 2× 1 h at 37°C in larvae . The Df(2L)slam Frt2L slam 5′rescue is also deficient for some proximal genes in addition to the slam locus. These genes do not function in early embryonic development, as the cellularization phenotype is rescued with the genomic slam transgene . For microinjection experiments, females with slam germ line clones were crossed with males Df(2L)slam Frt2L slam5′rescue/ CyO. In the case of transgenes, females with slam germ line clones (hsFlp122/+; Df(2L)slam Frt2L slam5′rescue/ovo[D2L] Frt2L; transgene/+ were crossed with males Df(2L)slam Frt2L slam5′rescue/CyO; transgene/transgene. Rescue of viability was tested with Df(2L)slam Frt2L slam5′rescue/Df(2L)slam Frt2L slam3′rescue; transgene/+ . All embryos from crosses with slam germ line clones are maternally deficient for slam. 50% of the progeny are also zygotically deficient for slam, whereas the other half contain a zygotic wild-type allele of slam (zygotic rescue). Embryos with zygotic rescue form a furrow during cellularization and have strong and uniform slam RNA and protein expression. For imaging of not yet translated slam RNA, embryos were obtained from females expressing PCP-GFP [21, 22] crossed to males with the slamPP7 transgene. Mapping of the RNA localization regions was performed with wild-type flies (OregonR). The shibire phenotype was induced in embryos from heterozygous females by shifting the embryos to 32°C 10 min prior to the FRAP experiment .
Truncations and gene fusions of slam cDNA as specified in Table 1 were cloned by PCR-based cloning. The vector pCS2 or pCS2-6xmyc (obtained from R. Rupp) contains a short leader sequence derived from the beta globin leader sequence. Genomic constructs were cloned into a pattB plasmid, suitable for generation of transgenic flies . Details of the cloning procedures and PCR are available on request. cDNA constructs were based on cDNA clone LD22808, which served as template for PCR. In comparison to other cDNA clones available now, LD22808 lacks 69 nucleotides at position CDS 1083, which do most likely represent an additional, rarely used intron. The slam gene with alternative codon usage (slam[ACU]) was designed by each 3-nucleotide codon with another suitable codon, according to the codon usage frequency in Drosophila  and synthesized by MWG Eurofins. Noncoding parts of the gene (5′ and 3′ untranslated regions and introns) were not mutated. slam[ACU] was cloned into the pattB-slam8.6 genomic construct  by PasI and SphI restriction sites. Introns were preserved. For slamPP7, a sequence with 12xPP7 [21, 22] was inserted at the unique SacII site of the coding sequence within the 8.6-kb genomic DNA of the slam locus cloned in pBKS and transferred to the transformation vector pattB. This construct also contained a 6x MS2 site inserted at position 50 of the 3′ untranslated region. For the transgenes with GAL4 driven expression pUASp-myc-GFP-slam and pUASp-myc-GFP-slamACU, slam and slam[ACU], cDNAs were cloned in frame into pUASp-myc-GFP.
RNA synthesis in vitro
For microinjection, capped transcripts were synthesized with linearized plasmid templates and the SP6 mMESSAGE mMACHINE high yield capped RNA transcription kit (Applied Biosystems). For live imaging of injected RNA, the reaction mix was complemented with 0.5 μl aminoallyl-UTP (25 mM, Jena bioscience). Isolated RNA (4 μl) was labeled with 5 μl Rhodamine Red-X, (succinimidyl ester in 56 μl DMSO, Invitrogen), 1 μl 0.1 M Na-Borate [pH 9] at RT overnight. The labeling dye and salts were removed using a desalting Sephadex G50 spin column. For RNA in situ hybridization, slam RNA probes were generated with T7 RNA polymerase and antisense templates as previously described .
RNA extraction, reverse transcription, quantitative PCR, RNA sequencing
Total RNA was extracted from staged embryos using Trizol Reagent (Invitrogen). Reverse transcription was performed with 1–2 μg RNA and oligo-dT primers according to the instructions of the manufacturer (Roche). Two μl (12.5% of the sample) of reverse transcripts were analyzed by quantitative PCR with specific primers. qPCR reactions were performed in duplicates. Specificity was controlled by a sample in which reverse transcriptase enzyme was omitted and in reactions with defined amounts of templates. The following primer pairs were used for quantitative PCR: WT1 SY88 (603–622) + SY89 (885–903), WT2 SY94 (−334–−316) + SY95 (−24–−1), ACU SY92 (603–620) + SY93 (886–905). ACU is specific for the slam[ACU] allele. The numbers in parentheses specify the position of the nucleotide within the cDNA, according to the coding sequence. For genome-wide analysis, RNA extracted from immunoprecipitates was subjected to next generation sequencing (Illumina HiSeq2000), according to the manufacturer’s protocols. Analysis was performed with 2–4 biological replicates. Weakly expressed genes (N < 50 or N < 10 for the experiment with Slam/Dia antibodies or GFP binder, respectively) were not considered in the analysis. RNA expression data were obtained from wild-type embryos with 1.5–2.5 h (, GEO Series accession number GSE97557). The data from Next Generation Sequencing have been deposited in NCBI’s Gene Expression Omnibus  and are accessible through GEO Series accession number GSE99761 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE99761.
Embryos were fixed in 8% formaldehyde in phosphate-buffered saline (PBS) and fluorescent RNA in situ hybridization was performed as previously described . Protein staining with a specific antibody was performed after RNA staining. The following antibodies were used: primary antibodies: rabbit/guinea pig-α-Slam , mouse-α-myc-9E10 (Roche); secondary antibodies: Alexa-coupled goat-anti-mouse/rabbit/guinea pig antibodies (Invitrogen), peroxidase coupled α-digoxigenin Fab-fragments (Roche). RNA probes against slam cDNA sequence, against slam[ACU], against GFP, or against PP7 were used. DNA was costained by DAPI. Embryos were mounted in Aquapolymount.
Fluorescent images of fixed and immune-stained embryos and live imaging experiments were recorded with a confocal microscope (Zeiss LSM780). Images were processed with ImageJ/Fiji and Photoshop (Adobe). Images from FRAP experiments were analyzed as previously described . Time-lapse recordings of embryos injected with fluorescently labeled slam RNA were recorded with a Leica epifluorescence microscope. Time-lapse recordings with differential interference contrast optics were recorded with an inverted microscope (Zeiss Observer Z.1, 25× NA0.7/oil) with a computer-controlled stage, allowing simultaneous recordings of multiple embryos.
Embryos were injected as previously described . The mRNA constructs were injected at concentrations of about 2–5 μg/μl. Fixation was performed about 1–1.5 h after injection, when the majority of embryos were in the cellularization stage. Cycloheximide (1 mg/ml) or buffer (0.1 M Na-phosphate, 5 mM KCl) was injected from the posterior end. About 10 minutes later imaging or FRAP was started. The embryos contained a genomic GFP-slam transgene.
Expression of slam and slamACU in Drosophila Schneider cells
slam wild type and slam[ACU] (on pCS2 plasmid, driven by a CMV promoter) were tested for protein production levels in cultured D. melanogaster Schneider line 2 (S2) cells. S2 cells were grown in Schneider’s Drosophila medium (Gibco/Invitrogen) supplemented with 10% fetal bovine serum (Gibco/Invitrogen) at 25°C. Cells were reseeded 1 day before transfection at about 1.0 million cells/ml (counted in Neubauer chamber). Transfection was performed according to the instruction of the manufacturer (Qiagen Transfection Kit containing Effectene Transfection Reagent). A control reaction (using pCS-GFP) indicated an approximately 6% transformation efficiency. Cells were harvested after 48–72 h. Half of the cells were dissolved in SDS loading buffer (Laemmli) to a final concentration of 50,000 embryos/μl. SDS gel electrophoresis was performed with 1 million cells per sample lane. RNA was extracted from the second half of the cells using Trizol Reagent (Invitrogen).
Guinea pig-α-Slam antibody  or guinea pig-α-Dia antibody  was bound to Dynabeads coated with Protein A (Invitrogen) in PBT for 1 h at 4°C. Beads were washed 5 times with PBT. Embryonic extracts were prepared by lyzing about 100 mg of 1.5–3-hour-old embryos in 1 ml YSS buffer  (50 mM Tris/HCl [pH8], 75 mM NaCl2, 1 mM MgCl2, 0.05% NP40, 100 mM sucrose, 1 M DTT, protease inhibitors [Roche], RNAase inhibitors) in a Dounce homogenizer. The lysate was centrifuged for 15 min at 14,000 rpm to remove debris. The cleared supernatant was mixed with the beads coated with antibody and incubated on a rotator for 1 h at 4°C. A 200-μl sample was taken from the supernatant (unbound fraction). Beads were then washed 4 times with cold YSS buffer. 200 μl of YSS buffer was added to the bound samples. Unbound and bound samples were complemented with 20 μl 10% SDS and 0.5 μl glycogen (20 μg/ml). Samples were extracted with phenol/chloroform/isoamylalcohol. Protein was precipitated from the organic phase with 1 ml acetone. RNA was precipitated by the addition of 40 μl Na-acetat (3 M) and 500 μl 100% ethanol to the aqueous phase. Protein and RNA samples were analyzed by western blot, quantitative RT-PCR, and next generation sequencing, respectively. For co-immunoprecipitation with the single-chain GFP antibody (GFP binder or GFP-TRAP, Chromotek), staged embryos from GFP-slam transgenic or wild-type flies were collected and lysed as described above, except for the following: 0.5 μl biotinylated-GFP binder (50 mM) was mixed with cleared embryo lysate for 1 h at 4°C. Then, 20 μl PBS-washed streptavidin-coupled Dynabeads were added and mixed for a further 1 h at 4°C.
Western blots and analysis of manually selected embryos
SDS polyacrylamide gel electrophoresis (PAGE) and western blot were performed as previously described . Primary antibodies were guinea pig or rabbit-α-Slam , mouse-α-tubulin-α (Sigma), and rabbit or guinea pig-α-Dia . Western blots were developed with fluorescently labeled secondary antibodies (800CW Donkey-α-guinea pig/mouse/rabbit IgG) and detected with a LICOR system. Sixteen-bit images were processed by Photoshop and FIJI/ImageJ. For analysis of embryos with defined genotype and stage, embryos were heat fixed as previously described , stained for Slam and DAPI, and mounted in 50% glycerol. Under fluorescent microscope, embryos in cellularization were sorted by their zygotic genotype and stage. Sorted embryos (N = 10–20) were lysed in Laemmli buffer with a Dounce homogenizer and analyzed by western blot.
S1 Fig. slam constructs and transgenes.
Schematic drawing of genomic slam transgenes and slam cDNA constructs in comparison to the slam locus. Boxes indicate transcribed regions. Coding sequence is marked in light orange, sequences with alternative coding in red, untranslated region in grey, GFP in green, 6xmyc tag in blue, and 12xPP7 in dark blue. The stop codon following the GFP tag is marked with a triangle.
S2 Fig. Expression and localization of slam RNA and protein in blastoderm embryos.
(A) Relative abundance of slam RNA determined by reverse transcription and quantitative PCR, with RNA isolated from embryo collections staged by time after egg lay. Corresponding nuclear cycles are indicated. (B) Abundance of Slam protein analyzed by western blot with extracts from heat fixed embryos staged by morphology and nuclear cycle. Detection of α-tubulin serves as a loading control. Corresponding absolute time after fertilization is indicated. (C) Wild-type embryos were fixed and stained for slam RNA by fluorescent RNA in situ hybridization (grey/red), for Slam protein by immunostaining (grey/green), and for DNA (blue). Images were recorded from 1 slide in 1 session, with the unchanged settings of the confocal microscope. Embryos were staged by the length of the furrow and nuclear density and morphology. Arrows in yellow point to punctate RNA staining inside the nucleus, which likely represent sites of primary transcription. (D) Expression of slamPP7 RNA in embryos with zygotic expression of slamPP7 in a wild-type background. RNA was detected with a probe specific for the PP7 sequence by fluorescent RNA in situ hybridization. For staging, embryos were stained for Slam protein (red) and DNA (blue), shown in side view. Scheme of the genomic slamPP7 transgenic construct is shown above the images. Scale bar = 10 μm. C, late cellularization (cycle 14); G, early gastrulation (stage 7); MW, apparent molecular weight; P, preblastoderm (nuclear cycle 1–9).
S3 Fig. Characterization of embryos injected with cycloheximide.
Images from time-lapse recordings of embryos expressing GFP-slam from a genomic transgene prior to and 10 min and 20 min after injection. Embryos were injected with either buffer or cycloheximide (1 mg/ml). Embryos were injected (A) in mitosis 13, (B) at the onset of cellularization, and (C) in early cellularization. Scale bar = 10 μm.
S4 Fig. Mapping of parts of the slam RNA sufficient for localization in wild-type embryos.
(A) Wild-type embryos or slam deficient embryos from germ line clones were injected with indicated mRNA and fixed and stained for Slam protein using Slam or Myc antibodies and RNA by probes for slam or GFP. slam embryos were recognized by the absence of overall slam RNA or protein signal. Scale bar = 10 μm. (B) Schematic representation of constructs. Detected FC localization of the injected RNA is indicated by “+” and red color of the construct and non-FC localization by “–”and faint red color of the construct. slam coding sequence is marked in red, 6xmyc tag in blue, and GFP in green. Dashed lines indicate deletions. Boxes in yellow mark the mapped regions, Loc1, Loc2, Loc3.
S5 Fig. Sequences of slamACU allele.
Comparison of coding sequences of wild-type (capital letters) and ACU (small letters) alleles. Sequences changes in ACU are marked in red. The exon-exon junction of the second intron is marked in yellow. cDNA clone LD22808 contains a third small intron (marked in blue), which is not annotated in the current genome annotation. This piece of sequence was not mutated in the ACU allele.
S6 Fig. Expression of the slam in cultured S2 cells.
(A, B) Western blot probed with Slam and α-tubulin antibodies with extracts from cultured S2 cells transiently transfected for expression for slam, slam[ACU], or GFP. * marks a cross-reacting band that served as loading control. (C) Abundance of slam transcripts and Slam protein expressed from slam[ACU] in relation to slam[wild type] was determined by western blotting (protein) and reverse transcription with qPCR (RNA) in extracts of transiently transfected S2 cells. Error bar indicates standard error of the mean. N = 3, 3 biological replicates. The underlying data for this figure can be found in S1 Data. ACU, alternative codon usage; GFP, green fluorescent protein; kD, kilodalton; MW, molecular weight; qPCR, quantitative polymerase chain reaction; S2 cells, Drosophila melanogaster Schneider 2 cells.
S7 Fig. Phenotype of slamACU embryos.
The length of the cellularization furrow was measured from time-lapse recordings with differential interference contrast of wild-type embryos and embryos from females with slam germ line clones with or without the slam[ACU] genomic transgene. Embryos were grouped into mutant (red) and zygotically rescued (blue) according to the cellularization phenotype. Bars = standard error of the mean. The underlying data for this figure can be found in S1 Data. ACU, alternative codon usage; N, number of embryos; slam, slow as molasses.
S1 Table. Transcripts specifically associated with Slam protein.
S1 Movie. Injected fluorescent slam RNA associates with FC.
Rhodamine-labelled slam mRNA was injected into wild-type embryos. Images were recorded every 20 s. The movie starts in interphase 13. During mitosis, the furrow extends and retracts. During interphase 14, the furrow gradually elongates with the FC moving inwards.
We thank M. Claussen, A. Ephrussi, R. Rupp, F. Schnorrer, B. Sullivan, and the Developmental Studies Hybridoma bank created by NICHD of the NIH/USA and maintained by the University of Iowa, the Bloomington Drosophila stock center (supported by NIH P40OD018537), the Drosophila genomics and genetics resources, Kyoto, the BACPAC resource center at Children’s Hospital Oakland, and the Genomic Resource center at Indiana University (supported by NIH 2P40OD010949-10A1) for discussions, materials, or fly stocks. We thank the Transcriptome analysis laboratory of the Göttingen Medical School for transcriptome analysis.
- 1. Holt CE, Bullock SL. Subcellular mRNA localization in animal cells and why it matters. Science. American Association for the Advancement of Science; 2009;326: 1212–1216. pmid:19965463
- 2. Martin KC, Ephrussi A. mRNA localization: gene expression in the spatial dimension. Cell. 2009;136: 719–730. pmid:19239891
- 3. Lasko P. mRNA localization and translational control in Drosophila oogenesis. Cold Spring Harb Perspect Biol. Cold Spring Harbor Lab; 2012;4: a012294–a012294. pmid:22865893
- 4. Cody NAL, Iampietro C, Lécuyer E. The many functions of mRNA localization during normal development and disease: from pillar to post. Wiley Interdiscip Rev Dev Biol. John Wiley & Sons, Inc; 2013;2: 781–796. pmid:24123937
- 5. Lécuyer E, Yoshida H, Parthasarathy N, Alm C, Babak T, Cerovina T, et al. Global analysis of mRNA localization reveals a prominent role in organizing cellular architecture and function. Cell. 2007;131: 174–187. pmid:17923096
- 6. Gonsalvez GB, Long RM. Spatial regulation of translation through RNA localization. F1000 Biol Rep. 2012;4: 16. pmid:22912650
- 7. Kindler S, Wang H, Richter D, Tiedge H. RNA transport and local control of translation. Annu Rev Cell Dev Biol. Annual Reviews; 2005;21: 223–245. pmid:16212494
- 8. St Johnston D. Moving messages: the intracellular localization of mRNAs. Nat Rev Mol Cell Biol. Nature Publishing Group; 2005;6: 363–375. pmid:15852043
- 9. Kugler J-M, Lasko P. Localization, anchoring and translational control of oskar, gurken, bicoid and nanos mRNA during Drosophila oogenesis. Fly (Austin). 2009;3: 15–28.
- 10. Wenzl C, Yan S, Laupsien P, Grosshans J. Localization of RhoGEF2 during Drosophila cellularization is developmentally controlled by slam. Mech Dev. 2010;127: 371–384. pmid:20060902
- 11. Foe VE, Odell GM, Edgar BA. Mitosis and Morphogenesis in the Drosophila Embryo: Point and Counterpoint. In: Bate M, Arias AM, editors. The Development of Drosophila melanogaster. Cold Spring Harbor Laboratory Press; 1993.
- 12. Schejter ED, Wieschaus E. Functional elements of the cytoskeleton in the early Drosophila embryo. Annu Rev Cell Biol. Annual Reviews 4139 El Camino Way, P.O. Box 10139, Palo Alto, CA 94303–0139, USA; 1993;9: 67–99. pmid:8280474
- 13. Mazumdar A, Mazumdar M. How one becomes many: blastoderm cellularization in Drosophila melanogaster. Bioessays. 2002;24: 1012–1022. pmid:12386932
- 14. Liu B, Grosshans J. Link of Zygotic Genome Activation and Cell Cycle Control. Methods Mol Biol. New York, NY: Springer New York; 2017;1605: 11–30. pmid:28456955
- 15. Lecuit T, Samanta R, Wieschaus E. slam encodes a developmental regulator of polarized membrane growth during cleavage of the Drosophila embryo. Dev Cell. 2002;2: 425–436. pmid:11970893
- 16. Stein JA, Broihier HT, Moore LA, Lehmann R. Slow as molasses is required for polarized membrane growth and germ cell migration in Drosophila. Development. 2002;129: 3925–3934. pmid:12135929
- 17. Acharya S, Laupsien P, Wenzl C, Yan S, Grosshans J. Function and dynamics of slam in furrow formation in early Drosophila embryo. Dev Biol. 2014;386: 371–384. pmid:24368071
- 18. Rothwell WF, Fogarty P, Field CM, Sullivan W. Nuclear-fallout, a Drosophila protein that cycles from the cytoplasm to the centrosomes, regulates cortical microfilament organization. Development. 1998;125: 1295–1303. pmid:9477328
- 19. Sung H-W, Spangenberg S, Vogt N, Grosshans J. Number of nuclear divisions in the Drosophila blastoderm controlled by onset of zygotic transcription. Curr Biol. 2013;23: 133–138. pmid:23290555
- 20. Swanson MM, Poodry CA. The shibire(ts) mutant of Drosophila: a probe for the study of embryonic development. Dev Biol. 1981;84: 465–470. pmid:20737885
- 21. Halstead JM, Lionnet T, Wilbertz JH, Wippich F, Ephrussi A, Singer RH, et al. Translation. An RNA biosensor for imaging the first round of translation from single cells to living animals. Science. American Association for the Advancement of Science; 2015;347: 1367–1671. pmid:25792328
- 22. Halstead JM, Wilbertz JH, Wippich F, Lionnet T, Ephrussi A, Chao JA. TRICK: A Single-Molecule Method for Imaging the First Round of Translation in Living Cells and Animals. Meth Enzymol. Elsevier; 2016;572: 123–157. pmid:27241753
- 23. Afshar K, Stuart B, Wasserman SA. Functional analysis of the Drosophila diaphanous FH protein in early embryonic development. Development. 2000;127: 1887–1897. pmid:10751177
- 24. Grosshans J, Wenzl C, Herz H-M, Bartoszewski S, Schnorrer F, Vogt N, et al. RhoGEF2 and the formin Dia control the formation of the furrow canal by directed actin assembly during Drosophila cellularisation. Development. The Company of Biologists Limited; 2005;132: 1009–1020. pmid:15689371
- 25. Yan S, Lv Z, Winterhoff M, Wenzl C, Zobel T, Faix J, et al. The F-BAR protein Cip4/Toca-1 antagonizes the formin Diaphanous in membrane stabilization and compartmentalization. J Cell Sci. The Company of Biologists Ltd; 2013;126: 1796–1805. pmid:23424199
- 26. Sampath K, Ephrussi A. CncRNAs: RNAs with both coding and non-coding roles in development. Development. Oxford University Press for The Company of Biologists Limited; 2016;143: 1234–1241. pmid:27095489
- 27. Bischof J, Maeda RK, Hediger M, Karch F, Basler K. An optimized transgenesis system for Drosophila using germ-line-specific phiC31 integrases. Proc Natl Acad Sci USA. National Acad Sciences; 2007;104: 3312–3317. pmid:17360644
- 28. Vicario S, Moriyama EN, Powell JR. Codon usage in twelve species of Drosophila. BMC Evol Biol. BioMed Central Ltd; 2007;7: 226. pmid:18005411
- 29. Winkler F, Gummalla M, Künneke L, Lv Z, Zippelius A, Aspelmeier T, et al. Fluctuation Analysis of Centrosomes Reveals a Cortical Function of Kinesin-1. Biophys J. 2015;109: 856–868. pmid:26331244
- 30. Edgar R, Domrachev M, Lash AE. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. Oxford University Press; 2002;30: 207–210.
- 31. Kress TL, Yoon YJ, Mowry KL. Nuclear RNP complex assembly initiates cytoplasmic RNA localization. J Cell Biol. Rockefeller University Press; 2004;165: 203–211. pmid:15096527