elav: Biological Overview | Evolutionary Homologs | Regulation | Developmental Biology | Effects of Mutation | References

Gene name - embryonic lethal, abnormal vision

Synonyms - elav

Cytological map position - 1B5-9

Function - RNA-binding

Keywords - pan-neural

Symbol - elav

FlyBase ID:FBgn0260400

Genetic map position - 1-[0.0]

Classification - RNP-1 signature

Cellular location - nuclear

NCBI links: Precomputed BLAST | Entrez Gene
Recent literature
Zaharieva, E., Haussmann, I. U., Brauer, U. and Soller, M. (2015). Concentration and localization of co-expressed ELAV/Hu proteins control specificity of mRNA processing. Mol Cell Biol 35(18):3104-15. PubMed ID: 26124284
Neuronally co-expressed ELAV/Hu proteins comprise a family of highly related RNA binding proteins, which bind to very similar cognate sequences. How this redundancy is linked to in vivo function and how gene specific regulation is achieved, has not been clear. Analysis of mutants in Drosophila ELAV/Hu family proteins ELAV, FNE and RBP9, and genetic interactions among them, indicates mostly independent roles in neuronal development and function, but convergence in the regulation of synaptic plasticity. Conversely, ELAV, FNE, RBP9 and human HuR bind ELAV target RNA in vitro with similar affinity. Likewise, all can regulate alternative splicing of ELAV target genes in non-neuronal wing-disc cells and substitute ELAV in eye development with artificially increased expression, but can also substantially restore ELAV's biological functions, when expressed under the control of the elav gene. Furthermore, ELAV related Sex-lethal can regulate ELAV targets and ELAV/Hu proteins can interfere with sexual differentiation. An ancient relationship to Sex-lethal is revealed by gonadal expression of RBP9 providing a maternal failsafe for dosage compensation. These results indicate that highly related ELAV/Hu RNA binding proteins select targets for mRNA processing based on expression levels and sub-cellular localization, but only minimally by altered RNA binding specificity.
Sanfilippo, P., Smibert, P., Duan, H. and Lai, E. C. (2016). Neural specificity of the RNA binding protein Elav is achieved by post-transcriptional repression in non-neural tissues. Development [Epub ahead of print]. PubMed ID: 27802174
Drosophila Elav is the founding member of the conserved family of Hu RNA binding proteins (RBPs), which collectively play critical and diverse roles in post-transcriptional regulation. Surprisingly, although Elav has a well-characterized neural cis-regulatory module, endogenous Elav is also ubiquitously transcribed and post-transcriptionally repressed in non-neural settings. In particular, mutant clones of multiple miRNA pathway components derepress ubiquitous Elav protein. Re-annotation of the elav transcription unit shows that not only does it generate extended 3' UTR isoforms, its universal 3' UTR isoform is much longer than previously believed. This longer common 3' UTR region includes multiple conserved, high-affinity sites for the miR-279/996 family. Notably, out of several miRNA mutants tested, endogenous Elav and a transgenic elav 3' UTR sensor are derepressed in mutant clones of mir-279/996, Cross-repression of Elav by another RBP was observed to be derepressed in non-neural miRNA pathway clones, namely Mei-P26. Finally, it was demonstrated that ubiquitous Elav has regulatory capacity, since derepressed Elav can stabilize an Elav-responsive sensor. It is critical to restrict Elav outside of the nervous system as misexpression of Elav in non-neural territories has profoundly adverse consequences. Altogether, this study defined unexpected post-transcriptional mechanisms that direct appropriate cell-type specific expression of a conserved neural RBP.

The quantity of a gene product may be controlled in a number of ways: regulation of the gene's rate of transcription initiation, the rate of protein synthesis and degradation, and the stability of the RNA are all factors that determine the level of any given protein in a cell. For years these ideas have reverberated like mantras in molecular biology circles, but evidence for the mechanisms of posttranscriptional regulation were sorely lacking. Recently it has been found that whole classes of cytoplasmic mRNAs can be made inactive, that proteins affect mRNA stability and proteins regulate the structure of mRNA. ELAV is one of a class of proteins that binds to mRNA and presumably has a regulatory function.

ELAV is required for correct differentiation and maintenance of the nervous system. The gene encodes an RNA binding protein expressed in all neurons after birth, qualifying elav as a pan-neural gene. The subcellular distribution of ELAV was investigated using ELAV-specific antibodies and scanning confocal laser microscopy. ELAV is predominantly localized within the nucleus where it concentrates within discrete domains described as dots and webs. To characterize these discrete domains an analysis of Drosophila coiled bodies was initiated. Coiled bodies are non-capsular nuclear bodies that appear to be composed of coiled fibrils. Most coiled bodies disassemble prior to or during mitosis. After cell division, the reassembly of coiled bodies occurs during G1 phase and is preceded by the reformation of nucleoli. Coilin is a 80-kD nuclear protein, identified with autoimmune serum, that is found as an integral component of coiled bodies. The polyclonal antibody R288 raised against human coilin was used to identify coiled bodies in cells of the Drosophila larval central nervous system. Double-labeling immunohistochemistry shows that, similar to vertebrate and plant systems, small nuclear ribonucleoproteins are enriched within these structures. The nuclear distribution of ELAV is reminiscent of the distribution of a number of splicing factors, including snRNPs, snRNAs, U1 70K and U2AF, as well as the snRNA-specific 2,2,7 trimethyl guanosine cap within mammalian nuclei (see Sans fille for more information on Drosophila splicing factors). Further analysis of ELAV reveals that subnuclear domains enriched with this molecule localize within and close to coiled bodies and close to subnuclear domains enriched with splicing factors. Deletion of the ELAV alanine/glutamine-rich amino-terminal auxiliary domain has no discernible effect on localization; proteins produced from elav lethal alleles distribute normally. This morphological study provides the first hint of a role for ELAV in the generation of alternatively spliced neural-specific mRNAs (Yannoni, 1997).

Although the Drosophila erect wing (ewg) gene is broadly transcribed in adults, an unusual posttranscriptional regulation involving alternative and inefficient splicing generates a 116-kDa Ewg protein in neurons, while protein expression elsewhere (or of other isoforms) is below detection at this stage. This posttranscriptional control is important, since broad expression of Ewg can be lethal. Elav is necessary to regulate Ewg protein expression in Elav-null eye imaginal disc clones and Elav is sufficient for Ewg expression in wing disc imaginal tissue after ectopic expression. Analysis of Ewg expression elicited from intron-containing genomic transgenes and cDNA minitransgenes in Elav-deficient eye discs shows that this regulation is dependent on the presence of ewg introns. Analyses of the ewg splicing patterns in wild-type and Elav-deficient eye imaginal discs and in wild-type and ectopic Elav-expressing wing imaginal discs, show that certain neuronal splice isoforms correspond to Elav levels. The data presented in this paper are consistent with a mechanism by which Elav increases the splicing efficiency of ewg transcripts in alternatively spliced regions rather than with a mechanism by which stability of specific splice forms is enhanced by Elav (Koushika, 2000).

The primary transcript of ewg, which has 10 exons, A to J, is alternatively spliced in two regions. Neuron-enriched heads and neuron-poor bodies have different EWG RNA splicing profiles. Heads show enrichment for a transcript encoding a 116-kDa protein, whereas bodies have lower amounts of the transcript that encodes the 116-kDa protein and greater amounts of unprocessed RNA. The head-enriched transcript encoding the 116-kDa protein results from inclusion of exon D and exclusion of exons E and I. Additionally, splicing of introns 3a, 3c, and 6 is inefficient, since these introns are retained in polyadenylated EWG RNA (Koushika, 2000).

Additionally, Elav promotes a neuron-enriched splice isoform of Drosophila armadillo transcript. The neuron-specific arm transcript, n-arm, is generated by an alternative splice event that results from the exclusion of exon 6 of ubiquitous-arm (u-arm). The primer pair used amplifies both u-arm and n-arm transcripts; the 147-bp smaller band corresponds to n-arm, while the 244-bp band corresponds to u-arm. To test if Elav has a role in the formation of n-arm transcripts, RNA from wild-type and elav null allele (edr) eye discs, as well as from wild-type eye discs and wing discs ectopically expressing Elav were analyzed by RT-PCR. The amount of n-arm is reduced in Elav-deficient eye discs, and in the ectopically expressing wing discs expression of n-arm is clearly induced. No change was observed in the band representing u-arm splicing. In summary, the presence of n-arm is correlated with the presence of Elav in both neural and nonneural tissues, implying that arm transcripts are regulated by Elav. Similar data were obtained for splicing of exons VIIa and VIIb of Neuroglian transcripts (Koushika, 2000).

Elav ensures that the correct alternatively spliced protein isoforms of certain genes are generated in neurons. Currently three target genes, ewg, Nrg, and arm have been identified. Both Nrg and arm are vital genes and are broadly transcribed and ubiquitous protein isoforms are broadly expressed, but an additional isoform, encoded by an alternatively spliced transcript, is pan-neurally expressed. The significance of the neural Nrg (n-Nrg) isoform is not known, but the distinct cytoplasmic domain could be important in signaling. The n-Arm isoform differs from the ubiquitous Arm (u-Arm) isoform because it lacks the Wingless interacting domain; moreover, it preferentially interacts with DE-cadherins. Even with these differences in properties, the current evidence suggest that the u-Arm is sufficient to provide the n-Arm function. Perhaps a more detailed phenotypic analysis may reveal a specific role for n-Arm (Koushika, 2000).

ewg, also a vital gene, is broadly transcribed, but the protein product, a likely transcriptional regulator, is almost exclusively neural. In the case of ewg, it is clear that the expression of the 116-kDa protein isoform is essential for viability in the nervous system and that, when expressed in nonneural tissues, it can be lethal. These Elav-regulated genes provide insight into the regulatory role of Elav in neurons. Experiments reported here demonstrate for the first time that the prevalence of neuron-specific ewg, nrg, and arm transcripts positively correlates with Elav levels, and these results are achieved through the increased use of specific splice sites (Koushika, 2000).

ELAV links paused Pol II to alternative polyadenylation in the Drosophila nervous system

Alternative polyadenylation (APA) has been implicated in a variety of developmental and disease processes. A particularly dramatic form of APA occurs in the developing nervous system of flies and mammals, whereby various developmental genes undergo coordinate 3' UTR extension. In Drosophila, the RNA-binding protein ELAV inhibits RNA processing at proximal polyadenylation sites, thereby fostering the formation of exceptionally long 3' UTRs. This study presents evidence that paused Pol II promotes recruitment of ELAV to extended genes. Replacing promoters of extended genes with heterologous promoters blocks normal 3' extension in the nervous system, while extension-associated promoters can induce 3' extension in ectopic tissues expressing ELAV. Computational analyses suggest that promoter regions of extended genes tend to contain paused Pol II and associated cis-regulatory elements such as GAGA. ChIP-seq assays identify ELAV in the promoter regions of extended genes. This study provides evidence for a regulatory link between promoter-proximal pausing and APA (Oktaba, 2014).

ELAV is an RNA-binding protein that has been shown to bind to U-rich regions in target mRNAs, including neuroglian and erect wings. Recently, the Hox gene Ultrabithorax (Ubx) was shown to be bound by ELAV through similar elements to regulate alternative splicing, but ELAV was not found to bind to predicted binding sites in the Ubx 3' UTR. Similarly, this study also failed to identify specific ELAV recognition sequences within extended 3' UTRs. The present study investigated how ELAV is selectively recruited to appropriate targets during neurogenesis (Oktaba, 2014).

The activities of synthetic reporter genes were exanubed in transgenic embryos to determine whether extended 3' UTRs are sufficient for the selective recruitment of ELAV in vivo. Transgenes contain the Drosophila synthetic core promoter (DSCP) attached to a GFP coding sequence followed by the entire extended 3' UTR of elav, one of the targets of ELAV. If elav 3' UTR sequences are sufficient to recruit ELAV, then this transgene should produce mRNAs containing extended 3' UTRs (Oktaba, 2014).

Expression of 3' UTR sequences was monitored via double labeling assays with GFP coding sequences to distinguish transgene mRNAs from endogenous elav transcripts. Expression of the transgene was confirmed by colocalization of GFP with a probe directed against the short 3' UTR. However, colocalization of GFP with extended sequences was not observed, indicating that mRNAs produced from the transgene lack 3' extensions. The only signals containing 3' extensions corresponded to endogenous elav mRNAs (Oktaba, 2014).

Additional experiments were done to determine why the transgene fails to produce extended transcripts. The possibility that the GFP coding sequence somehow inhibits expression of extended sequences by creating GFP transgenes lacking proximal poly(A) signals was excluded. Such constructs no longer depended on ELAV for 3' extension and were found to produce mRNAs containing extended 3' UTR sequences when expressed in ectopic tissues lacking ELAV (Oktaba, 2014).

To test whether promoter sequences play a role in ELAV recruitment, the DSCP sequence was swapped with a 333 base pairs (bp) genomic DNA fragment encompassing the native elav promoter region, consisting of 92 bp upstream and 241 bp downstream of the (TSS). Strikingly, colocalization of GFP and extension sequences was observed indicating expression of the elav 3' UTR extension, as seen for the endogenous locus (Oktaba, 2014).

To confirm that 3' extension depends on native promoter regions of extended genes, a construct bearing the fully extended brat 3' UTR downstream of GFP was also tested, using three different promoters: the DSCP, the native promoter producing the short form of brat, and the native promoter producing the extended form of brat. Only the brat promoter associated with endogenous extension mediated expression of transgenic transcripts containing 3' UTR extensions. These observations suggest that the promoter regions of extended genes are essential for the ELAV-mediated expression of 3' UTR extensions (Oktaba, 2014).

The preceding results suggest that promoter sequences are important for the synthesis of 3' extensions in the developing nervous system. Their importance was determined by examining nonneural tissues. Ectopic ELAV can drive 3' UTR extension in ectopic tissues from endogenous loci. Attempts were made to determine whether ectopic ELAV could also induce ectopic 3' extensions from transgenic DNAs (Oktaba, 2014).

Both the GFP-elav transgene and ELAV protein were expressed in muscle cells using a Mef2-Gal4 driver. In this context, mRNA expression from the reporter is easily distinguished from endogenous elav expression, which occurs only in the nervous system. The DSCP fails to generate 3' UTR extensions, and only endogenous elav transcripts in the CNS were detected. In contrast, the GFP-elav transgene containing the native elav promoter produced transcripts with extended 3' UTRs in muscle tissue. Quantification of transgene expression in dissected muscle tissue using quantitative PCR (qPCR) shows that both promoters drive robust transgene expression (GFP signal), but only the native promoter drives expression of extension sequences. Similarly, the second brat promoter, but not the DSCP, was also able to drive expression of an extended brat 3' UTR in muscle cells (Oktaba, 2014).

Whether the promoter sequence from one extended gene could promote extension of the 3' UTR of another such gene was also tested. Indeed, a GFP transgene containing the elav promoter and brat extended 3' UTR exhibited ELAV-mediated APA. These observations suggest a link between transcription initiation and ELAV-mediated APA (Oktaba, 2014).

To determine whether the promoter regions of extended genes share common sequence motifs, 252 neural-specific transcripts produced by 219 different genes exhibiting 3' UTR extensions were examined. The most significantly enriched motif is the GAGA element, which occurs in nearly half of all extended genes. To investigate the functional significance of the GAGA element in promoters of extended genes, whether 3' UTR extension is diminished in animals lacking the GAGA-binding protein, Trithorax-like (Trl) was tested. For all six genes examined, the ratio between extension sequences and coding sequences was reduced between 15% and 75% in Trl mutant flies. These observations suggest that the GAGA motifs in the promoters of extended genes are important for proper 3' UTR extension (Oktaba, 2014).

The GAGA element is a motif commonly found in the promoter regions of genes containing paused Pol II. Paused Pol II is a pervasive feature of gene regulation in metazoan development, and at least 10%-30% of all genes in Drosophila contain paused Pol II. It is thought that paused promoters are poised for rapid activation and thereby exhibit synchronous induction in the different cells of a tissue. Another function of promoter pausing might be to ensure proper recruitment of essential factors for RNA elongation and processing (Oktaba, 2014).

Most extended genes contain paused Pol II, based on whole genome Pol II ChIP-seq assays. Some extended genes express both short and long isoforms from the same promoter (for example elav), while others (e.g., brat) employ different promoters for the different isoforms. In the latter case, only the promoter driving the extended isoform contains paused Pol II (Oktaba, 2014).

To determine whether paused Pol II might be associated with the formation of 3' UTR extensions, the overall Pol II pausing index (PI) of extended genes and various control genes was examined. Extended transcripts were found to be derived from significantly more paused promoters than any of the control groups, including neural-specific (but nonextended) genes. Thus, there is a clear association between Pol II pausing and 3' UTR extension, which transcends the general pausing seen for neural-specific gene expression. Extended transcripts are also strongly paused in muscle cells, where they are not actively transcribed and where ELAV is not expressed. Thus, Pol II pausing at extended genes occurs independently of ELAV (Oktaba, 2014).

The preceding analyses raise the possibility that ELAV is selectively recruited to the promoter regions of extended genes. To test this hypothesis, ChIP-seq assays were performed using anti-ELAV antibodies. ELAV is an RNA-binding protein that directly binds and inhibits proximal poly(A) elements of target transcripts. It was therefore reasoned that it should be possible to identify the genome-wide distribution of ELAV by crosslinking ELAV/RNA complexes to associated DNA templates. ELAV ChIP-seq assays were conducted with nuclei obtained from 6-8 hr and 10-12 hr embryos. These stages were selected based on previous observations regarding the timing of 3' extensions in the nervous system (Oktaba, 2014).

6,879 genomic regions bound by ELAV were identified in 6-8 hr embryos and 8,076 regions in 10-12 hr embryos. There is a striking enrichment of ELAV in the promoter regions of extended genes. For example, argonaute1 (ago1) produces multiple APA isoforms driven from three different promoters. The two promoters that produce extended transcripts display ELAV peaks, whereas the promoter that expresses the short (ubiquitous) isoform does not. High levels of ELAV are also found at 3' poly(A) sites, consistent with previous RNA immunoprecipitation assays (Oktaba, 2014).

The ChIP-seq data were combined into a 'meta-gene' plot that provides simple visualization of key sites of ELAV binding. There is a significant enrichment of ELAV at the promoter regions of extended genes as compared with neural-specific nonextended genes. A distinct ELAV peak is seen near the TSS, although ELAV binding continually increases across the 5' UTR and peaks at ~300 bp downstream of the start site (Oktaba, 2014).

ELAV not only binds to promoter regions, but also to 3' UTRs and introns of extended genes. ELAV is strikingly depleted from coding sequences. As expected, binding markedly increases in the vicinity of proximal poly(A) sites and remains high across extended regions where there are additional poly(A) elements (Oktaba, 2014).

A meta-gene analysis was also performed of previously published Pol II ChIP-seq data. Pol II binding is highly enriched in the promoter regions of extended genes, which is consistent with earlier evidence that such genes tend to contain paused Pol II. The Pol II binding profile did not otherwise differ from nonextended neural-specific genes. It is possible that ELAV binds to both nascent transcripts and associated DNA templates, since ELAV is usually detected at distal poly(A) sites of extended genes prior to full transcriptional extension (Oktaba, 2014).

This study has presented evidence that paused Pol II fosters selective recruitment of ELAV and coordinates expression of extended 3' UTR sequences during neurogenesis. The basis for selective recruitment of ELAV is a bit of a mystery since it has been shown to interact with broadly distributed low-complexity RNA sequences (e.g., U-rich). Increased interaction between paused promoters and termination regions might help promote 3' extension, for example, by bringing ELAV to the promoter via gene looping. The observed association of ELAV with the paused promoter regions of extended genes provides a foundation for selectivity and also strengthens the link between transcription initiation and 3' cleavage. It is improbable that paused Pol II is sufficient for recruitment of ELAV, since not all paused genes exhibit APA. It is therefore likely that additional sequence elements, for example, in extended 3' UTRs, are essential for recruitment. ELAV proteins are highly conserved, and it is easy to imagine that the regulation of 3' extension in the vertebrate CNS depends on selective promoter recruitment as seen in Drosophila (Oktaba, 2014).


cDNA clone length - Five major species of transcripts can be detected including those containing intronic sequences (Yao, 1991).

Bases in 5' UTR - 491

Exons - three


Amino Acids - 519

Structural Domains

The protein contains three repeats of a pair of sequences defined as RNA-binding consensus sequences, an octopeptide (RNP1), and an appropriately spaced hexapeptide, RNP2 (Yao, 1991).

The neuron specific Drosophila Elav protein belongs to the Elav family of RNA binding proteins, characterized by three highly conserved RNA recognition motifs; an N-terminal domain, and a hinge region between the second and third RNA recognition motifs. The role of the Elav hinge in localization and function has been examined in vivo using transgenes encoding Elav hinge deletions. Subcellular localization of the hinge mutant proteins reveals that residues between amino acids 333-374 are necessary for nuclear localization. This delineated sequence has no significant homology to classical nuclear localization sequences, but it is similar to the recently characterized nucleocytoplasmic shuttling sequence, the HNS, from a human Elav family member, HuR. However, this defined sequence is insufficient for nuclear localization as tested using hinge-GFP fusion proteins. Functional assays have revealed that mutant proteins that fail to localize to the nucleus are unable to provide Elav vital function, but their function is significantly restored when translocated into the nucleus by a heterologous nuclear localization sequence tag (Yannoni, 1999).

elav: Evolutionary Homologs | Regulation | Developmental Biology | Effects of Mutation | References

date revised: 20 Nov 97

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

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