BIOLOGICAL OVERVIEW

The Drosophila Bithorax Complex encodes three well-characterized homeodomain proteins that direct segment identity, as well as several noncoding RNAs of unknown function. This study analyzes the iab-4 locus, which produces the microRNAs iab-4-5p and iab-4-3p. iab-4 is analogous to miR-196 in vertebrate Hox clusters. Previous studies demonstrated that miR-196 interacts with the Hoxb8 3' untranslated region. Evidence is presented that miR-iab-4-5p directly inhibits Ubx activity in vivo. Ectopic expression of mir-iab-4-5p attenuates endogenous Ubx protein accumulation and induces a classical homeotic mutant phenotype: the transformation of halteres into wings. These findings provide the first evidence for a noncoding homeotic gene and raise the possibility that other such genes occur within the Bithorax complex (Ronshaugen, 2005).

Classical genetic studies suggest that the Bithorax Complex (BX-C) contains as many as nine homeotic genes (Lewis 1978). However, only three encode Hox proteins: Ultrabithorax (Ubx), abdominal-A (abd-A), and Abdominal-B (Abd-B). The bulk of genomic DNA comprising the BX-C is thought to function as cis-regulatory DNA which controls the timing and site of Hox expression. Nevertheless, it has been known for more than 15 years that intergenic regions of the BX-C are extensively transcribed (Cumberledge, 1990; Bae, 2002; Drewell, 2002). The possible functional activities of the noncoding RNAs have received little attention, despite the fact that these transcripts, including iab-4 (Cumberledge, 1990), are expressed in restricted domains along the anterior-posterior axis, like conventional Hox genes (Ronshaugen, 2005).

Hox gene clusters contain conserved miRNAs. For example, miR-10 is located within the Drosophila Antennapedia gene complex (ANT-C) between the Hox genes Deformed and Sex combs reduced (Lagos-Quintana, 2001). The sequence and location of miR-10 are conserved in vertebrate Hox complexes (Tanzer, 2005). Sex combs reduced has been proposed as a direct miR-10 target in insects (Brennecke, 2005). A second group of Drosophila miRNAs map to a hairpin located at the distal end of the iab-4 locus (Aravin, 2003), which resides between abd-A and Abd-B (). miRNAs were cloned from both arms of this hairpin and are termed iab-4-5p and iab-4-3p (Aravin, 2003). miR-iab-4-5p was recently predicted to regulate Ubx activity (Stark, 2003; Grun, 2005). Although vertebrates lack an iab-4 ortholog, as defined by sequence identity, a different miRNA, miR-196, resides at an analogous position adjacent to the posterior-most HOX 9-13 paralogs. Tissue culture assays, in vivo cleavage products, and transgenic lacZ 'sensors' indicate that miR-196 inhibits Hoxb8 activity (Mansfield, 2004; Yekta, 2004). Despite these provocative target relationships, no phenotypes have been associated with any Hox miRNA (Ronshaugen, 2005).

miRNAs are short, 21-24-nt RNAs that attenuate protein synthesis by binding complementary sites in target mRNAs. An unexpectedly modest amount of base-pairing appears to underlie target recognition. Experimental and computational studies have converged on the principle of 'seed-pairing,' whereby ~7 continuous Watson-Crick base pairs at the 5'-end of the miRNA mediate target recognition. The limited sequence requirement for miRNA-mRNA interactions has fueled current proposals that a third or more of all mRNAs may be regulated by miRNAs. As tallies of miRNA loci continue to grow (with current estimates for humans ranging from 800 to 1000), the network of possible miRNA:target interactions will expand (Ronshaugen, 2005).

Only a small number of miRNA:target interactions have been studied in vivo. This study presents evidence that iab-4 microRNAs selectively attenuate Ubx activity in vivo. The Ubx 3' untranslated region (3' UTR) contains predicted target sites for miR-iab-4-5p and expression of a GFP-Ubx-3' UTR 'sensor' transgene is repressed by ectopic expression of a mir-iab-4 minigene. This minigene also reduces Ubx protein levels in haltere imaginal discs, thereby inducing a classical homeotic transformation of halteres into wings. Taken together, these results suggest that the iab-4 transcription unit encodes an authentic homeotic regulatory gene. It is suggested that additional noncoding RNAs correspond to 'missing' homeotic genes in the Bithorax complex, and novel mechanisms of iab-4 regulation during development are discussed (Ronshaugen, 2005).

The detailed analysis of intergenic transcripts in the abd-A/Abd-B interval suggests a 'strand exclusion' model for iab-4 regulation. The iab-4 locus is unusual in that both strands are transcribed (Bae, 2002). However, each strand displays a distinct pattern of expression. The strand producing the iab-4 pri-miRNA is broadly expressed in the A2-A7 region of the germband, while the other strand is expressed in A8 and A9. Double RNA FISH assays suggest that the miR strand is initially expressed in A2-A8, but expression is lost in A8 as transcription of the other strand progresses from the iab-8 domain. Perhaps transcription from one strand diminishes transcription from the other. Although the detailed mechanism may be different, these observations are evocative of the mutually exclusive expression of Xist and Tsix RNAs on mammalian X chromosomes. Additional target predictions for miR-iab-4 imply that exclusion of iab-4 expression from A8 might be important for stable accumulation of other potential iab-4 target mRNAs, such as Abd-B (Enright, 2003; Stark, 2003; Ronshaugen, 2005 and references therein).

Traditionally, recessive loss-of-function mutations are used to assess the in vivo activities of patterning genes. The principal argument for iab-4:Ubx interactions in development rests with the analysis of dominant gain-of-function phenotypes arising from the misexpression of miR-iab-4 in the haltere imaginal discs. The specificity of the resulting haltere-to-wing homeotic transformation correlates with reduced levels of Ubx protein accumulation specifically in the regions where miR-iab-4 products are misexpressed. It is likely that there is redundancy in the transcriptional repression of Ubx by Abd-A and Abd-B products, and the inhibition of Ubx protein synthesis by iab-4 miRNAs. Indeed, noncoding genes in the BX-C were mainly identified by dominant mutations such as chromosomal rearrangements. Therefore, misexpression assays may prove effective in analyzing the function of other Hox noncoding genes (Ronshaugen, 2005).

This study provides evidence that iab-4 encodes a novel homeotic regulatory activity, which functions, at least in part, by producing miRNAs inhibiting Ubx. iab-4 miRNAs may regulate additional target mRNAs. For example, computational analyses identify homothorax as another potential target of interest (Grun, 2005). Homothorax works in parallel with the Hox cofactor Extradenticle and various Hox proteins to control the patterning of legs and antennae. It is also possible that downstream transcriptional targets of Ubx ('realizators') might be modulated by iab-4 miRNAs. Additional noncoding RNAs in the BX-C, such as cbx, pbx and bxd, might also possess homeotic regulatory activities and account for the remaining genes identified within the sorrounding region (Ronshaugen, 2005).

A single Hox locus in Drosophila produces functional microRNAs from opposite DNA strands

MicroRNAs (miRNAs) are ~22-nucleotide RNAs that are processed from characteristic precursor hairpins and pair to sites in messages of protein-coding genes to direct post-transcriptional repression. The miRNA iab-4 locus in the Drosophila Hox cluster is transcribed convergently from both DNA strands, giving rise to two distinct functional miRNAs. Both sense and antisense miRNA products target neighboring Hox genes via highly conserved sites, leading to homeotic transformations when ectopically expressed. Sense/antisense miRNAs are also present in the mouse and antisense transcripts are found close to many miRNAs in both flies and mammals, suggesting that additional sense/antisense pairs exist (Stark, 2008).

Hox genes are highly conserved homeobox-containing transcription factors crucial for development in animals. Genetic analyses have identified them as determinants of segmental identity that specify morphological diversity along the anteroposterior body axis. A striking conserved feature of Hox complexes is the spatial colinearity between Hox gene transcription in the embryo and the order of the genes along the chromosome. Hox clusters also give rise to a variety of noncoding transcripts, including microRNAs (miRNAs) mir-10 and mir-iab-4/mir-196, which derive from analogous positions in Hox clusters in flies and vertebrates (Yekta, 2004). miRNAs are ~22-nucleotide (nt) RNAs that regulate gene expression post-transcriptionally. They are transcribed as longer precursors and processed from characteristic pre-miRNA hairpins. In particular, Hox miRNAs have been shown to regulate Hox protein-coding genes by mRNA cleavage and inhibition of translation, thereby contributing to the extensive regulatory connections within Hox clusters (Mansfield, 2004; Yekta, 2004; Hornstein, 2005; Ronshaugen, 2005). Several Hox transcripts overlap on opposite strands, providing evidence of extensive antisense transcription, including antisense transcripts for mir-iab-4 in flies (Bae, 2002) and its mammalian equivalent mir-196 (Mainguy, 2007). However, the function of these transcripts has been elusive. This study shows that the iab4 locus in Drosophila produces miRNAs from opposite DNA strands that can regulate neighboring Hox genes via highly conserved sites. Evidence is provided that such sense/antisense miRNA pairs are likely employed in other contexts and a wide range of species (Stark, 2008).

Examination of the antisense transcript that overlaps Drosophila mir-iab-4 revealed that the reverse complement of the mir-iab-4 hairpin folds into a hairpin reminiscent of miRNA precursors. Moreover, 17 sequencing reads from small RNA libraries of Drosophila testes and ovaries mapped uniquely to one arm of the iab-4 antisense hairpin. All reads were aligned at their 5' end, suggesting that the mir-iab-4 antisense hairpin is processed into a single mature miRNA in vivo, which is referred to as miR-iab-4AS. For comparison, six reads were found consistent with the known miR-iab-4-5p (or miR-iab-4 for short) and one read was foudn for its star sequence (miR-iab-4-3p). Interestingly, the relative abundance of mature miRNAs and star sequences for mir-iab-4AS (17:0) and mir-iab-4 (6:1) reflects the thermodynamic asymmetry of the predicted miRNA/miRNA* duplexes (Khvorova, 2003; Schwarz, 2003). Because they derived from complementary near palindromes, miR-iab-4 and miR-iab-4AS had high sequence similarity, only differing in four positions at the 3' region. However, they differed in their 5' ends, which largely determine miRNA target spectra (Brennecke, 2005; Lewis, 2005): miR-iab-4AS was shifted by 2 nt, suggesting targeting properties distinct from those of miR-iab-4 and other known Drosophila miRNAs (Stark, 2008).

Robust transcription of mir-iab-4 sense and antisense precursors was confirmed by in situ hybridization to Drosophila embryos. Both transcripts were detected in abdominal segments in the posterior part of the embryo, but intriguingly in nonoverlapping domains. As described previously (Bae, 2002; Ronshaugen, 2005), mir-iab-4 sense was expressed highly in abdominal segments A5-A7, showing modulation in levels within the segments: abdominal-A (abd-A)-expressing cells appeared to have more mir-iab-4, whereas Ultrabithorax (Ubx)-positive cells appeared to have little or none (this study; Ronshaugen, 2005). In contrast, mir-iab-4AS transcription was detected in the segments A8 and A9, where Abdominal-B (Abd-B) is known to be expressed (this study; Yoder, 2006). Primary transcripts for mir-iab-4 and mir-iab-4AS were also detected by strand-specific RT-PCR in larvae, pupae, and male and female adult flies, suggesting that both miRNAs are expressed throughout fly development (Stark, 2008).

To assess the possible biological roles of the two iab-4 miRNAs, fly genes were examined for potential target sites by searching for conserved matches to the seed region of the miRNAs (Lewis, 2005). Highly conserved target sites were found for miR-iab-4AS in the 3' untranslated regions (UTRs) of several Hox genes that are proximal to the iab-4 locus and are expressed in the neighboring more anterior embryonic segments: abd-A, Ubx, and Antennapedia (Antp) have four, five, and two seed sites, respectively, most of which are conserved across 12 Drosophila species that diverged 40 million years ago. More than two highly conserved sites for one miRNA is exceptional for fly 3' UTRs, placing these messages among the most confidently predicted miRNA targets and suggesting that they might be particularly responsive to the presence of the miRNA. The strong predicted targeting of proximal Hox genes was reminiscent of previously characterized miR-iab-4 targeting of Ubx in flies and miR-196 targeting of HoxB8 in vertebrates (Mansfield, 2004; Yekta, 2004; Hornstein, 2005; Ronshaugen, 2005; Stark, 2008 and references therein).

To test whether miR-iab4AS is functional and can directly target abd-A and Ubx, Luciferase reporters were constructed carrying the corresponding wild-type 3' UTRs and control 3' UTRs in which each seed site was disrupted by point substitutions. mir-iab-4AS potently repressed reporter activity for abd-A and Ubx. This repression was specific to the miR-iab-4AS seed sites; expression of the control reporters with mutated sites was not affected. Tested were perform to see whether mir-iab-4AS reduced expression of a Luciferase reporter with the Abd-B 3' UTR, which has no seed sites. As expected, mir-iab-4AS expression did not affect reporter activity, consistent with a model where miRNAs do not target genes that are coexpressed at high levels. In addition to demonstrating specific repression dependent on the predicted target sites, these assays confirmed the processing of the mir-iab-4AS hairpin into a functional mature miRNA (Stark, 2008).

If miR-iab-4AS were able to potently down-regulate Ubx in the fly, its misexpression should result in a Ubx loss-of-function phenotype, a line of reasoning that has often been used to study the functions and regulatory relationships of Hox genes. Ubx is expressed throughout the haltere imaginal disc, where it represses wing-specific genes and specifies haltere identity. When mir-iab-4AS was expressed in the haltere imaginal disc under bx-Gal4 control, a clear homeotic transformation of halteres to wings was observed. The halteres developed sense organs characteristic of the wing margin and their size increased severalfold, features typical of transformation to wing. Consistent with the increased number of miR-iab4AS target sites, the transformation was stronger than that reported for expression of iab-4 (Ronshaugen, 2005), for which changes in morphology were confirmed wing-like growth was not found (Stark, 2008).

It is concluded that both strands of the iab-4 locus are expressed in nonoverlapping embryonic domains and that each transcript produces a functional miRNA in vivo. In particular, the novel mir-iab-4AS is able to strongly down-regulate neighboring Hox genes. Interestingly, vertebrate mir-196, which lies at an analogous position in the vertebrate Hox clusters, is transcribed in the same direction as mir-iab-4AS and most other Hox genes, and targets homologs of both abd-A and Ubx (Mansfield, 2004; Yekta, 2004; Hornstein, 2005). With its shared transcriptional orientation and homologous targets, mir-iab-4AS appears to be the functional equivalent of mir-196 (Stark, 2008).

The expression patterns and regulatory connections between Hox genes and the two iab-4 miRNAs show an intriguing pattern in which the miRNAs appear to reinforce Hox gene-mediated transcriptional regulation. In particular, miR-iab-4AS would reinforce the posterior expression boundary of abd-A, Ubx, and Antp, supporting their transcriptional repression by Abd-B. mir-iab-4 appears to support abd-A- and Abd-B-mediated repression of Ubx, reinforcing the abd-A/Ubx expression domains and the posterior boundary of Ubx expression. Furthermore, both iab-4 miRNAs have conserved target sites in Antp, which is also repressed by Abd-B, abd-A, and Ubx. The iab-4 miRNAs thus appear to support the established regulatory hierarchy among Hox transcription factors, which exhibits 'posterior prevalence,' in that more posterior Hox genes repress more anterior ones and are dominant in specifying segment identity. Interestingly, Abd-B and mir-iab-4AS are expressed in the same segments, and the majority of cis-regulatory elements controlling Abd-B expression are located 3' of Abd-B. This places them near the inferred transcription start of mir-iab-4AS, where they potentially direct the coexpression of these genes. Similarly, abd-A and mir-iab-4 may be coregulated as both are transcribed divergently, potentially under the control of shared upstream elements (Stark, 2008).

These data demonstrate the transcription and processing of sense and antisense mir-iab-4 into functional miRNAs with highly conserved functional target sites in neighboring Hox genes. In an accompanying study (Bender 2008), genetic and molecular analyses in mir-iab-4 mutant Drosophila revealed that the proposed regulation of Ubx by both sense and antisense miRNAs occurs under physiological conditions and, in particular, the regulation by miR-iab-4AS is required for normal development. These lines of evidence establish miR-iab-4AS as a novel Hox gene, being expressed from within the Hox cluster and regulating Hox genes during development (Stark, 2008).

The genomic arrangement of two miRNAs that are expressed from the same locus but on different strands might provide a simple and efficient means to create nonoverlapping miRNA expression domains. Such sense/antisense miRNAs could restrict each other's transcription, either by direct transcriptional interference, as shown for overlapping convergently transcribed genes (Shearwin, 2005; Hongay, 2006), or post-transcriptionally, possibly via RNA-RNA duplexes formed by the complementary transcripts. Sense/antisense miRNAs would usually differ at their 5' ends and thereby target distinct sets of genes, which might help define and establish sharp boundaries between expression domains. Coupled with feedback loops or coregulation of miRNAs and genes in cis or trans, this arrangement could provide a powerful regulatory switch. The iab-4 miRNAs might be a special case of tight regulatory integration in which miRNAs and proximal genes appear coregulated transcriptionally in cis and repress each other both transcriptionally and post-transcriptionally (Stark, 2008).

It is perhaps surprising that no antisense miRNA had been found previously, even though, for example, the intriguing expression pattern of the iab-4 transcripts had been reported nearly two decades ago (Cumberledge, 1990; Bae, 2002), and iab-4 lies in one of the most extensively studied regions of the Drosophila genome. The frequent occurrence of antisense transcripts (Yelin, 2003; Katayama, 2005) suggests that more antisense miRNAs might exist. Indeed, up to 13% of known Drosophila, 20% of mouse, and 31% of human miRNAs are located in introns of host genes transcribed on the opposite strand or are within 50 nt of antisense ESTs or cDNAs. These include an antisense transcript overlapping human mir-196 (see also Mainguy, 2007). However, because of the contribution of noncanonical base pairs, particularly G:U pairs that become less favorable A:C in the antisense strand, many miRNA antisense transcripts will not fold into hairpin structures suitable for miRNA biogenesis, which explains the propensity of miRNA gene predictions to identify the correct strand. Nonetheless, in a recent prediction effort, 22 sequences reverse-complementary to known Drosophila miRNAs showed scores seemingly compatible with miRNA processing. Deep sequencing of small RNA libraries from Drosophila confirmed the processing of small RNAs from four of these high-scoring antisense candidates (Ruby, 2007), and the ovary/testes libraries used here showed antisense reads for an additional Drosophila miRNA (mir-312). In addition, using high-throughput sequencing of small RNA libraries from mice, sequencing reads were found that uniquely matched the mouse genome in loci antisense to 10 annotated mouse miRNAs. Eight of the inferred antisense miRNAs were supported by multiple independent reads, and two of them had reads from both the mature miRNA and the star sequence. These results suggest that sense/antisense miRNAs could be more generally employed in diverse contexts and in species as divergent as flies and mammals (Stark, 2008).

MicroRNAs in the Drosophila bithorax complex

The iab-4 noncoding RNA from the Drosophila bithorax complex is the substrate for a microRNA (miRNA). Gene conversion was used to delete the hairpin precursor of this miRNA; flies homozygous for this deletion are sterile. Surprisingly, this mutation complements with rearrangement breakpoint mutations that disrupt the iab-4 RNA but fails to complement with breaks mapping in the iab-5 through iab-7 regulatory regions. These breaks disrupt the iab-8 RNA, transcribed from the opposite strand. This iab-8 RNA also encodes a miRNA, detected on Northern blots, derived from the hairpin complementary to the iab-4 precursor hairpin. Ultrabithorax is a target of both miRNAs, although its repression is subtle in both cases (Bender, 2008).

A large number of microRNA (miRNA) clones prepared from Drosophila RNA have been characterized at a variety of developmental stages. Two of these clones matched sequences from the BX-C, mapping to the 3' end of a ncRNA discovered by Cumberledge (1990). This ncRNA was called the 'iab-4 RNA,' because it was thought to come from the iab-4 segmental domain of the BX-C, and the miRNAs were named miR-iab-4 5p (five prime) and miR-iab-4 3p (three prime). More recent mapping of segmental domains (Bender, 2000) has shown that the RNA actually lies in the iab-3 domain (regulating parasegment 8), and indeed, the ncRNA is expressed from parasegment 8 through parasegment 12 (Cumberledge, 1990). However, the iab-4 nomenclature is maintained here to avoid confusion with the designations in other studies. Two cDNA clones for the iab-4 RNA were described by Cumberledge (1990), with alternate 3' poly(A) sites separated by 304 base pairs (bp). The two miRNAs come from this region between these two poly(A) sites; both are presumably derived from a 70-base hairpin RNA precursor predicted from the sequence (Bender, 2008).

A recent study (Ronshaugen, 2005) suggested that the miR-iab-4 5p miRNA might be responsible for repression of Ubx in the abdominal segments where the miRNA is expressed. The conclusion was based on experiments in which miR-iab-4 5p was expressed at high levels in tissues, including the wing and haltere discs, where miR-iab-4 5p is not normally found. However, the pattern of UBX expression in PS8, where the miRNA is expressed, is very similar to the UBX pattern in PS7, which lacks the miRNA. The obvious repression of Ubx in both of these parasegments is clearly dependent on abd-A; any effect of miR-iab-4 5p must be subtle or redundant. Moreover, misexpression of miRNAs in other systems have been shown to give misleading effects. The function of miR-iab-4 5p can best be examined by mutating or deleting the miRNA (Bender, 2008).

Prior studies characterized a large number of P element insertions in the BX-C, including one called HCJ200, which maps only ~200 bp proximal to the miRNAs. This provided the opportunity to mutate the miRNAs by P-element-mediated gene conversion. Loss of the miRNAs derived from the iab-4 ncRNA causes no apparent morphological or behavioral phenotype, but the analysis revealed a functional miRNA derived from the opposite strand (Bender, 2008).

A 3.7-kb conversion donor fragment was constructed with a mutated version of the miRNA precursor sequence. The precursor sequence is symmetrically cut by the BstZ17I restriction endonuclease; this permitted the replacement of most of the precursor sequence with a double-stranded oligonucleotide-containing sites for the HindIII and I-SceI endonucleases. A plasmid with the donor sequence and a plasmid to supply P-element transposase were both injected into embryos with the HCJ200 (rosy⁺) P insertion. Offspring from the injected individuals were screened for loss of the HCJ200 P element (i.e., rosy^-), and progeny from these exceptional flies were screened by PCR for a change in the size of the genomic DNA at the site of the insertion. One of 86 fertile injected animals gave the expected convertants. The conversion events were verified by sequencing the PCR product, and by whole-genome Southern blots. Genomic DNA was cut with HindIII, which cuts the oligonucleotide introduced in the conversion but does not cut the wild-type sequence within the 3.7-kb donor fragment. The sizes of the HindIII fragments on the Southern blot confirmed that the donor sequence was at the expected position in the BX-C and not at another genomic location (Bender, 2008).

Flies homozygous for the conversion chromosome (henceforth called 'δ') appeared normal. In particular, no evidence of segmental transformation was seen in mounted adult abdominal cuticles of either sex. However, both sexes were sterile. Females had ovaries with eggs of normal size, but only very rare individuals ever laid an egg, even after mating with wild-type males, and these rare eggs never hatched. Males had morphologically normal testes containing motile sperm. In single-fly tests for mating behavior, δ homozygous females mated with wild-type males as readily as their heterozygous siblings. The δ homozygous males showed normal courtship behavior toward wild-type females, except that they never completed copulation. The mutant males mounted the females, but they did not bend their abdomens quite far enough to mate. Thus, the sterility in both sexes appeared to be behavioral, due to a defect in the nerves or muscles required to lay eggs or to curl the abdomen (Bender, 2008).

The δ mutation was tested for complementation with rearrangement mutations in the BX-C, including several that should disrupt the iab-4 RNA transcript upstream of the position of the miRNA precursor. Surprisingly, breaks disrupting the iab-4 transcription unit complemented with δ -- i.e., trans-heterozygotes were fertile. Thus, the iab-4 RNA is not the precursor for any miRNA that is responsible for fertility. In contrast, rearrangements distal to the position of the miRNAs failed to complement, even with breaks >50 kb distal. Assuming that noncomplementing rearrangements are upstream in the precursor, one can deduce that the precursor is transcribed distal to proximal on the chromosome, and that the miRNA responsible for fertility comes from the opposite strand to those detected by Aravin (2003). Similarly, one would predict that the precursor transcript for the fertility miRNA spans at least the iab-7 through the iab-3 segmental domains (Bender, 2008).

Several studies have detected such a transcript (Sánchez-Herrero, 1989; Zhou, 1999; Bae, 2002; Bender, 2002), which is usually called the iab-8 ncRNA. It has been detected and mapped solely by in situ hybridization to RNA in embryos, although the complementation analysis now corroborates the in situ mapping, at least for the iab-4 through iab-7 region. Its start site is near Abd-B; it was detected by probes 4 kb proximal to the 3' end of the Abd-B transcripts. Zhou (1999) defined a potential promoter for the iab-8 transcript, which lies ~5 kb downstream from Abd-B, although the evidence did not preclude a start site further upstream. Bae (2002) detected the iab-8 RNA upstream of the Abd-B class A RNA start site (the 'BPP' probe, suggesting a more distal start site. However, hybridization to the Abd-B class B RNA in the ninth abdominal segment (PS14) could have been mistaken for the iab-8 RNA pattern. Moreover, the iab-8^S10 breakpoint, just proximal to Abd-B, does complement the sterility phenotype of δ, and so the promoter for the iab-8 fertility function should be to the left of that break (Bender, 2008).

At the 3' end, the iab-8 RNA extends through abd-A. The iab-8 RNA in situ pattern was detected by a probe 5.5 kb proximal to the 3' end of the abd-A poly(A) site. Thus the transcription unit spans at least 120 kb. The iab-8 RNA has not been detected by probes in the bxd regulatory region, further proximal to abd-A (Bender, 2008).

The iab-8 transcript initiates at the cellular blastoderm stage (Zhou, 1999), as do most of the other embryonic ncRNAs (Akam, 1985; Cumberledge, 1990; Bae, 2002). However, it should take ~45 min to transcribe to the position of the miRNA precursor hairpin, assuming a transcription speed of ~1.3 kb/min. This would account for the developmental delay in the appearance of the RNA signal. The iab-8 RNA is located in the eighth abdominal segment and in more posterior segmental rudiments. In late embryos, the transcript persists in the posterior end of the ventral nerve chord (Bender, 2008).

Embryos homozygous for the δ mutation showed no apparent changes in the patterns of ABD-A and ABD-B, but there were subtle differences in the UBX pattern. UBX is expressed strongly and comprehensively in the cells of parasegment 6 (PS6, primarily the first abdominal segment). In the second abdominal segment (PS7), ABD-A appears and turns off UBX, especially in the more anterior cells of the parasegment. In the more posterior segments, the UBX staining pattern weakens progressively, and the ABD-A pattern becomes somewhat stronger. However, in δ embryos, the UBX pattern is nearly constant from PS7 through PS12. Thus, the progressive posterior decline in wild-type embryos appears not to be caused by ABD-A or ABD-B but rather by miR-iab-4 5p, whose expression shows a progressive posterior increase in PS8-12 (Bender, 2008).

In the eighth abdominal segment (PS13) of a wild-type embryo, UBX is almost completely absent in both the epidermis and the CNS. Homozygotes for δ show a partial derepression of UBX in the CNS in PS13. The derepression is similar in pattern and intensity to that seen in AbdB^-/+ heterozygotes (data not shown). The repression of Ubx could be indirect; miR-iab-8 could be a positive regulator of Abd-B (and miR-iab-4 a positive regulator of abd-A). But all known targets of miRNAs are negatively regulated, and so it seems more likely that both miRNAs directly regulate Ubx (Bender, 2008).

It is possible that the effects of these two miRNAs are masked by functional redundancy with abd-A (for miR-iab-4) and Abd-B (for miR-iab-8). Embryos lacking ABD-A (but retaining miR-iab-4) show a dramatic derepression of UBX in the second through seventh abdominal segments. There does appear to be a slight decline in UBX levels in the more posterior segments, which could be attributed to miR-iab-4 repression. A complete analysis would include the UBX expression in embryos lacking both abd-A and miR-iab-4, but that will require construction of an abd-A, δ double mutant chromosome. In any case, miR-iab-4 repression of Ubx is subtle, even in the absence of ABD-A. Similarly, in an Abd-B homozygous mutant embryo (retaining miR-iab-8), UBX expression in the eighth abdominal segment closely resembles that in the seventh. Again, a δ, Abd-B double mutant chromosome would be useful for comparison, but again it is clear that the repression of Ubx by miR-iab-8 is still subtle in the absence of ABD-B. There is no reason to expect that Ubx is the only target of these miRNAs; perhaps other target genes will be discovered which the miRNAs repress more dramatically (Bender, 2008).

MiR-iab-8 is the first example of a functional product of a ncRNA in the BX-C. There are no other predicted miRNA precursor sequences in the iab-8 RNA or elsewhere in the BX-C (the Antennapedia complex includes miR-10), but there are many other ncRNAs. The possibility that they also include functional products now seems more likely (Bender, 2008).

Functionally distinct regulatory RNAs generated by bidirectional transcription and processing of microRNA loci

Many microRNA loci exhibit compelling hairpin structures on both sense and antisense strands; however, the possibility that a miRNA gene might produce functional species from its antisense strand has not been examined. Antisense transcription of the Hox miRNA locus mir-iab-4 generates the novel pre-miRNA hairpin mir-iab-8, which is then processed into endogenous mature miRNAs. Sense and antisense iab-4/iab-8 miRNAs are functionally distinguished by their distinct domains of expression and targeting capabilities. miR-iab-8-5p, like miR-iab-4-5p, is also relevant to Hox gene regulation. Ectopic mir-iab-8 can strongly repress the Hox genes Ultrabithorax and abdominal-A via extensive arrays of conserved target sites, and can induce a dramatic homeotic transformation of halteres into wings. The antisense miRNA principle is generalizable: it has been shown that several other loci in both invertebrates and vertebrates are endogenously processed on their antisense strands into mature miRNAs with distinct seeds. These findings demonstrate that antisense transcription and processing contributes to the functional diversification of miRNA genes (Tyler, 2008).

These studies of BX-C miRNAs reveal two principal insights into miRNA regulatory biology. First, a new Hox cluster miRNA, mir-iab-8, was identified. Using gain-of-function assays it was shown that it can strongly inhibit Ubx and abd-A and generate homeotic phenotypic transformations. Indeed, the Hox-regulatory properties of mir-iab-8 are far more potent than those of mir-iab-4 (Ronshaugen, 2005), and correlate directly with the properties of its target sites in their 3' UTRs. Curiously, both BX-C miRNAs obey organizational and functional rules previously defined for the protein-encoding members of the BX-C. These regulatory RNAs exhibit colinearity, in that transcription of pri-mir-iab-8 initiates more distally on the chromosome and is expressed more posteriorly in the embryo relative to pri-mir-iab-4. They also exhibit posterior prevalence, in that both sense and antisense iab-4 miRNAs directly repress multiple homeotic genes located more anteriorly in the Hox cluster. In fact, the next most-anterior Hox gene Antp is a third likely endogenous iab-miRNA target that contains highly conserved target sites with t1A features for miR-iab-8-5p. In contrast, Abd-B contains no conserved sites for either iab-4 or iab-8 miRNAs in its long (>2 kb) 3' UTR. Therefore, BX-C miRNAs and homeobox genes are governed by the same regulatory logic (Tyler, 2008).

It is worth recalling that saturation mutagenesis screens of the BX-C revealed only three loci that are required for viability and exhibit homeotic defects, corresponding to the homeobox genes Ubx, abd-A, and Abd-B. In contrast, pioneering studies by Lewis (1978) considered rearrangements and dominant alleles suggested the existence of at least eight homeotic 'factors' in this region of the genome. Although many of these are now recognized as cis-regulatory elements that regulate Hox gene transcription, this work with BX-C miRNAs reveals two bona fide Hox regulators that are capable of inducing severe dominant homeotic transformations. The endogenous requirement for iab-4/iab-8 miRNAs appears to be subtle, possibly due to compensatory transcriptionally based regulatory mechanisms. Nevertheless, loss of function analysis corroborates that these miRNAs are required for normal expression of Hox targets in the nervous system and for normal development. These data emphasize that loss-of-function and gain-of-function genetics are complementary approaches to uncover important regulatory molecules (Tyler, 2008).

Antisense transcription and processing were uncovered in this study as a mechanism to generate new functional miRNAs. Bioinformatic analysis suggests that a large fraction of miRNA loci are theoretically competent to produce antisense miRNAs. Extant cloning efforts suggest that few miRNA loci actively produce large quantities of antisense miRNAs. Nevertheless, the sequencing effort reported in this study has revealed additional instances of putative antisense miRNAs. Although none of these was cloned more than twice, genetics demonstrates that rare miRNAs (e.g., lsy-6, expressed by a handful of neurons in a whole animal) can be critical components of regulatory networks and can have potent biological activities. Therefore, assessment of the biological relevance of the other antisense miRNA candidates awaits further study (Tyler, 2008).

In the case of the iab-4 locus, the regulatory diversity afforded by sense and antisense transcription of a single miRNA hairpin is manifested by altering the seed regions of their respective miRNA products and by deploying the sense and antisense pri-miRNA transcripts in distinct spatial domains. It might in fact be deleterious for a given locus to be simultaneously transcribed on both strands -- either because of transcriptional interference from colliding polymerase complexes, or because of the possibility to inadvertently generate dsRNA. Further analysis is needed to test the notion that it is favorable for sense/antisense miRNA pairs not to be expressed in the same cells. Overall, though, as animal genomes are quite extensively transcribed, and many miRNA genes adopt extensive hairpins on both strands, the potential for endogenous antisense processing of miRNA hairpins is theoretically quite broad. It is proposed that antisense transcription of other miRNA loci might generate novel small RNAs whose potentially beneficial regulatory activities are available for selection and stabilization by natural selection. This identification of several confident examples of antisense miRNAs, whose processing and/or targets have been conserved among diverse species, provides compelling support for this hypothesis (Tyler, 2008).

Developmental inhibition of miR-iab8-3p disrupts mushroom body neuron structure and adult learning ability

MicroRNAs are small non-coding RNAs that inhibit protein expression post-transcriptionally. They have been implicated in many different physiological processes, but little is known about their individual involvement in learning and memory. Several miRNAs have been identified that either increased or decreased intermediate-term memory when inhibited in the central nervous system, including miR-iab8-3p. This paper reports a new developmental role for this miRNA. Blocking the expression of miR-iab8-3p during the development of the organism leads to hypertrophy of individual mushroom body neuron soma, a reduction in the field size occupied by axonal projections, and adult intellectual disability. Four potential mRNA targets of miR-iab8-3p were identified whose inhibition modulates intermediate-term memory including Ceramide phosphoethanolamine synthase, which may account for the behavioral effects produced by miR-iab8-3p inhibition. These results offer important new information on a microRNA required for normal neurodevelopment and the capacity to learn and remember normally (Busto, 2016)

A double-negative gene regulatory circuit underlies the virgin behavioral state

Virgin females of many species conduct distinctive behaviors, compared with post-mated and/or pregnant individuals. In Drosophila, this post-mating switch is initiated by seminal factors, implying that the default female state is virgin. However, it was recently shown that loss of miR-iab-4/miR-iab-8-mediated repression of the transcription factor Homothorax (Hth) within the abdominal ventral nerve cord (VNC) causes virgins to execute mated behaviors. This study used genomic analysis of mir-iab-4/8 deletion and hth-microRNA (miRNA) binding site mutants (hth[BSmut]) to elucidate doublesex (dsx) as a critical downstream factor. Dsx and Hth proteins are highly complementary in CNS, and Dsx is downregulated in miRNA/hth[BSmut] mutants. Moreover, virgin behavior is highly dose sensitive to developmental dsx function. Strikingly, depletion of Dsx from very restricted abdominal neurons (SAG-1 cells) abrogates female virgin conducts, in favor of mated behaviors. Thus, a double-negative regulatory pathway in the VNC (miR-iab-4/8 -| Hth -| Dsx) specifies the virgin behavioral state (Garaulet, 2021).

Females of diverse invertebrate and vertebrate species coordinate multiple behavioral programs with their reproductive state. Mature female virgins are receptive to male courtship and copulation, but following mating and/or pregnancy, they decrease sexual activity and modulate behaviors to generate and foster their children. Behavioral remodeling associated with the female reproductive state includes increased aggression and nest building in avians and mammals and decreased male acceptance, increased egg-laying, and appetitive/metabolic changes in insects. The genetic and neurological control of this process has been intensively studied in fruit flies, where sexual activity induces the post-mating switch, a host of behavioral changes collectively known as post-mating responses (PMRs) (Garaulet, 2021).

In Drosophila, as in other species, 'virgin' is typically considered the default behavioral state, because factors that induce PMRs are transferred in seminal fluids during copulation. Among these, Sex Peptide (SP) is necessary and sufficient to drive most female post-mated behaviors. SP signals via uterine SP sensory neurons (SPSNs). Some SPSN+ neurons contact abdominal interneurons in the ventral nerve cord (VNC) that express myoinhibitory peptide, which input into a restricted population of ascending neurons (SP abdominal ganglion [SAG] neurons) that project to the posterior brain, including pC1 neurons. This outlines an ascending flow of information for how a seminal fluid peptide can alter female brain activity. The brain integrates this with auditory and visual cues to coordinate diverse behaviors mediated by distinct lineages of descending neurons and VNC populations that modulate specific behaviors according to internal state and external stimuli (Garaulet, 2021).

Recently, it was found that post-transcriptional suppression of the homeobox gene homothorax (hth) within the VNC is critical to implement the virgin behavioral state (Garaulet, 2020). Of note, deletion of the Bithorax Complex (BX-C) locus mir-iab-4/8, point mutations of their binding sites in hth, or deletion of the hth neural-specific 3' UTR extension bearing many of these microRNA (miRNA) sites all cause mutant female virgins to perform mated behaviors. Thus, the failure to integrate two post-transcriptional regulatory inputs at a single target gene prevents females from appropriately integrating their sexual internal state with external behaviors (Garaulet, 2021).

Recognition of the transcription factor Hth as a target of regulatory circuits for virgin behavior implies that downstream loci may serve as a functional output for this process. This study used molecular genetic profiling to identify a critical requirement for Doublesex (Dsx) to implement the female virgin behavioral state. Dsx has been well studied with respect to differentiation of sexually dimorphic traits, but its roles in post-mitotic neurons are little known. This study found that expression of Dsx in the VNC mediates virgin behavior, and that modulation of Dsx in only a few abdominal VNC neurons is sufficient to convert the suite of female virgin behaviors into mated conducts (Garaulet, 2021).

Recent work established how miRNA mediated suppression of the transcription factor Hth to safeguard the virgin female behavioral state. Using engineered alleles and spatio-temporal hth manipulations, this study demonstrated a developmental requirement for post-transcriptional regulation of Hth within the abdominal ganglion of the CNS for female behavior. However, Hth was not required in otherwise wild-type VT-switch neurons for execution of virgin behaviors, implying that expression of Hth in the abdominal VNC must normally be prevented. This involves integration of two mechanisms: a high density of BX-C miRNA binding sites (miR-iab-4/8) within the hth-HD 3' UTR, as well as a neural-specific 3' UTR elongation, which unveils many of these sites only on neural hth isoforms (Garaulet, 2021).

This study has extended this regulatory axis by showing that loss of BX-C miRNAs, acting through derepressed Hth, leads to downregulation of the Dsx in the abdominal VNC. Dsx is well-known as a master sex determination transcription factor, and it shows localized expression in specific CNS domains. However, although the activity of Dsx-expressing neurons per se has been implicated in the switch in females, the functions of Dsx in post-mitotic neurons are less well defined. This work reveals that Dsx itself is a central component in specifying virgin behavior, because its restricted suppression in as few as four (SAG-1+) neurons is sufficient to induce post-mated behaviors. It remains to be better defined how SAG-1 neurons are affected by depletion of Dsx. No overt differentiation defects were observed, but an effect of masculinization cannot be ruled out. Otherwise, the recent work suggests an activity defect in a general population of switch neurons in the miRNA mutant, but more direct analysis of dsx-depleted SAG-1 neurons awaits (Garaulet, 2021).

Altogether, in contrast with highly branched regulatory networks that are bioinformatically inferred to lie downstream of individual miRNAs, this study revealed a linear, double-negative regulatory cascade comprising miRNAs and two transcription factors (see SAG-1 neurons specifically require Dsx for a suite of female virgin behaviors). These findings provide impetus to assess possible direct regulation of Dsx by Hth, as well as to elucidate Dsx targets that are relevant to female behavioral control. Overall, this study expands a genetic hierarchy that is essential for females to couple the virgin internal state with appropriate behaviors (Garaulet, 2021).

date revised: 25 August 2022

Home page: The Interactive Fly © 2006 Thomas Brody, Ph.D.

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