BIOLOGICAL OVERVIEW

The quest for genes expressed in the fly's brain is currently one of the major preoccupations of Drosophila geneticists. The gene roX1, previously known as yang, was identified by two different laboratories using different approaches, both designed to identify genes expressed in the brain. In one approach, an enhancer detector screen was carried out for mushroom body expression of a reporter gene. This study found a reporter expressed in females and not males, but an adjacent gene (roX1) was expressed in males but not females (Meller, 1997). A second study sought male specific genes expressed in mushroom bodies. A line of flies expressing green fluorescent protein in the mushroom body cells was used to obtain purified mushroom body cells using a fluorescence-activated cell sorter. A male cDNA library from GFP-positive cells was constructed and differentially screened with male- and female-specific mushroom body cDNA probes. Two genes were identifed: roX1 and roX2 (Meller, 1997). As will be described later, the RNA transcripts of both genes decorate the X chromosome of male flies, and thus the origin of their name: RNA on the X (roX) (Meller, 1997).

roX1 is expressed during embryonic development in a non-sex delimited fashion. Both males and females show expression of roX1 in neuroblasts, but this expression fades in females. Expression of roX1 is controlled by the Male specific lethal-2 (MSL-2) directed dosage compensation system. MSL-2 is responsible for directing the relatively high level of transcription from the single X chromosome of males, compared with a lower level of transcrition from the two X chromosomes of females (Kelly, 1995). Absence of any one of the four genes responsible for dosage compensation, completely eliminates roX1 expression. Interestingly, fading of the roX1 RNA in females takes place long after the association of MSL proteins (Meller, 1997 and Amrein, 1997).

Presence of functional Sex lethal is responsible for the absence of roX1 RNA in females. As presence of functional Sex lethal protein in females is responsible for a blocking of functional MSL-2 splicing in females, mutation of Sxl results in functional MSL-2 splicing in females and production of roX1 levels comparable to those observed in wild-type males (Meller, 1997).

What is the function(s) of roX1 and the distantly linked roX2 gene, both of which are expressed in adult males and are absent from adult females? Neither RNA codes for a protein, and roX1 RNA is found associated with the single X chromosome of pupae. Mutation of roX1 results in no apparent phenotype, a rather distressing result for geneticists who look to mutation to define gene function. Two facts are known: (1) Both roX1 and roX2 are closely linked to genes that are expressed only in females. Perhaps roX1 and roX2 regulate the expression of these female specific genes. Unfortunately, no evidence is provided for or against this possibility. (2) A similar sex delimited non-transcribed RNA species is involved in X chromosome inactivation in mammals (Penny, 1996 and Lee, 1996). XIST RNA, produced by the inactive X chromosome concentrates near the inactive X in an association with chromatin. XIST RNA remains with the nuclear matrix fraction after removal of chromosomal DNA but is released from its association with the inactive X during mitosis (Clemson, 1996). Incredably, a multicopy Xist transgene located on an autosome appears to produce RNA that binds to and inactivates autosomal chromatin in cis (Lee, 1996).

Might roX1 and roX2 function similarly in Drosophila? It is possible that the RNA of chromatin has a structural role that engenders gene silencing of proximal female specific genes on the male X chromosome of Drosophila. Such a function would have to be carried out by several genes with redundant function, since there is no phenotypic effect associated with roX1 mutation. It is proposed that roX1 and its family members associate along the entire X chromosome to help change chromatin conformation and achieve hypertranscription in the male, or, by their absence, change specific gene activation in the female by changing chromatin coformation, perhaps by associating with the MSLs, histone acetyltransferase, or other constituents of chromatin. This model is analogous to that for Xist, although Xist RNA achieves the opposite goal, that of condensing an X chromsome (Meller, 1997 and Amrein, 1997).

Perhaps no phenotypic effect to Xist mutation exists because the function of Xist in neurons (presumably the activation or repression of genes involved in determining male specific behaviors through modification of the phenotype of neurons) is in fact regulated by multiple backup pathways, ensuring a tight regulation of this important function. There is precedence for this type of redundancy. For instance, misexpression of Fasciclin 3 results in misdirection of certain neurons but does not alter targeting of neurons that normally express Fas3 (Chiba, 1995).

While the hypothesis that the rox RNAs are involved in dosage compensation is attractive, some of the data with regard to the rox RNAs are either lacking or appear to be in conflict with previous knowledge about dosage compensation. (1) Only rox1 has been shown to be associated with the male X chromosome; whether the rox1 RNA is bound to the same sites as the MSLs or is bound in some other pattern is unkown. (2) Some features of the reported temporal and spatial expression patterns of the rox RNAs seem to be incompatible with what is known about the msl-based dosage compensation system. rox1 RNA was reported to be present in both sexes up to germ band retraction with predominant staining in the CNS; a reduction of expression in female embryos occurred late in embryogenesis (Meller, 1997). However, the MSL proteins, known to be involved in dosage compensation, are associated with the X chromosome in all somatic cells of males from early gastrula stages on and not associated with the Xs in female embryos, as has been reported for rox1 (Meller, 1997). Data has been obtained in the current study that contradicts some of the observations of Meller, especially the finding of rox1 RNA with female chromosomes (Franke, 1999).

The strongest evidence that rox1 might be involved in dosage compensation is the fact that it is associated with the X chromosome in male larval salivary gland nuclei (Meller, 1997). Using a similar technique, the same has been found to be true for rox2 RNA. However, this technique does not resolve the distribution of these RNAs along the X chromosome because the extant methods for immunostaining of polytene chromosomes and RNA in situ hybridization are not easily compatible. A protocol that gives an MSL banding pattern without removing the rox RNA molecules was developed by using less concentrated acid solutions for relatively short fixation times. rox1 RNA and MSL-1 protein are colocalized to exactly the same chromosomal bands. No signals were detected that show one of the two molecules alone. The same double staining was performed with rox2 RNA and MSL-1 protein and gave an identical result. These results show that MSL proteins and rox RNAs are distributed on the male X chromosome in exactly the same pattern, providing strong evidence for a direct involvement of the rox RNAs in dosage compensation (Franke, 1999).

The rox1 and rox2 RNAs are essential components of the compensasome, which mediates dosage compensation in Drosophila

The rox RNA distributions were reexamined in embryos: both rox1 and rox2 are expressed in male embryos in patterns completely overlapping with that of the MSL proteins. Both RNAs are detected in a punctate pattern in every cell of the developing male embryo. This nuclear staining pattern is observed from early gastrula stages on. No staining is seen in female embryos, which were stained with anti-Sxl antibodies afterward to verify their sex. To prove that the nuclear pattern represents association with the X chromosome in these diploid cells, double labeling experiments with rox RNAs and anti-MSL-1 antibodies were performed. Both signals completely overlap in all of the embryonic cells, as shown for rox1 and MSL-1. These findings establish that rox RNAs are associated with the male X chromosome from early embryonic stages on and that their expression patterns and subnuclear localizations in embryonic cells correspond exactly to those of the MSL proteins (Franke, 1999).

The MSL protein distribution was examined in sections of adult flies and the pattern was compared to the rox RNA in situ hybridization pattern obtained from similar sections. On sections from male flies, MSL proteins and both rox1 and rox2 can be detected in nuclei in all body parts, most likely in every cell. In every case, the MSL proteins and the rox RNAs are restricted to an area covering about 20% of the nucleus. The distribution of signals obtained with the rox probes is virtually indistinguishable from the signal obtained with anti-MSL antibodies. That the subnuclear localization of the MSL1 protein in adult cells is due to its association with the X chromosome was verified by simultaneous in situ hybridization with a probe that painted the whole X chromosome and anti-MSL-1 antibody staining. To establish that the rox signals overlap with the MSL signal and therefore are associated with the X chromosome, double-labeling with an antibody to the MSL-1 protein and in situ hybridization to the rox RNAs was performed. The rox2 and MSL-1 signals completely overlap. The same result is obtained with rox1 probe, indicating that both rox RNAs are localized on the X chromosome in all somatic cells from adult males. Together with the preceeding results, these findings strongly support the idea that the rox RNAs are involved in msl-mediated dosage compensation beginning at early gastrulation. They continue to be used in all somatic tissues throughout all stages of development (Franke, 1999).

Simultaneous removal of rox1 and rox2 function leads to a loss of X Chromosome binding of the MSL proteins in male embryos. If the rox RNAs are integral members of a nucleic acid-protein complex that regulates dosage compensation, then one might expect that mutations in the rox genes could disrupt the assembly and/or functioning of such a complex. Experiments addressing this possibility have been limited to examining the effects of mutations at the rox1 locus that were shown not to affect MLE binding to the male X chromosome (Meller, 1997 ). The effects of rox2 deletions and rox1/rox2 double mutants were examined on the binding of the MSL proteins to the X chromosome in male embryos. Deletions in the 10B,C region of the X chromosome were used, where the rox2 gene is located (Amrein, 1997). Two deficiencies in the region that remove the rox2 gene [Df(1)DA622 and Df(1)HM456] and two deficiencies that do not remove the rox2 gene [Df(1)KA6 and Df(1)M259-4] were used. These deletions are embryonic lethals, but developed to late embryonic stages. Thus, MSL binding could be studied at earlier stages in these embryos. The rox1MB710 mutation has no effect on the distribution of the MSL-1, -2, and -3 proteins in male embryos, compared to wild-type embryos. The deletions Df(1)DA622 and Df(1)HM456 that remove the rox2 gene, as well as the deletions Df(1)KA6 and Df(1)M259-4 that do not, also have no effects on the MSL distribution. In contrast, double mutants of rox1 and rox2 show a clear lack of MSL association with the X chromosome in male embryos. Sibling control embryos, that are wild type with respect to rox1 and rox2 and are distinguished from the mutant embryos by the ftz-lacZ staining show normal MSL staining. The deletions in the 10B,C region that do not remove the rox2 gene do not show any abnormal MSL distribution when combined with the rox1 mutation. These data thus show that only simultaneous mutations in both rox1 and rox2 abolish MSL association with the X chromosome. These results strongly argue that the rox RNAs are redundant in their function and that they are essential components of the dosage compensation machinery (Franke, 1999).

Thus it has been established that the rox1 and rox2 RNAs function as integral components in dosage compensation. The findings that the rox1 and rox2 genes have an essential but redundant function in dosage compensation and that these RNAs colocalize with the MSL proteins suggest that they are components of an RNA-protein complex that brings about dosage compensation. This RNA-protein complex is here termed "compensasome" (Franke, 1999).

Several results suggest that the rox RNAs are redundant in their function. The absence of any phenotype in flies that are mutant for rox1 led to the initial suggestion that these RNAs might be members of a redundant gene family (Meller, 1997). The finding that embryos exhibit a normal MSL binding pattern with either one or the other of the rox genes mutated shows that neither of these RNAs alone is required for dosage compensation. The result that the MSL binding pattern is disrupted in rox1/rox2 double mutant male embryos shows that these RNAs are functioning as components of the dosage compensation machinery and that they are functionally redundant. However, the results do not address whether the rox1 and rox2 RNAs are present in a fixed stoichiometry in a compensasome or, alternatively, are present in an either/or manner. If as discussed, the rox RNAs are redundant in their function, one might expect similarities in their primary and/or secondary structures. A comparison of the two sequences reveals a region of homology within a 30 bp stretch. Screening the GenBank sequence database with these two 30 nt sequences did not detect other sequences with significant homology. The biological significance of this sequence is not obvious at this point (Franke, 1999).

Colocalization of the rox RNAs and MSL proteins on the X chromosome and the lack of MSL binding to the X chromosome in a rox1/rox2 double mutant background indicates that there is likely a physical interaction between the rox RNAs and the MSL proteins. Such a physical interaction offers a potential basis for the finding that the MLE protein (but not the other MSLs) can be removed from X chromosomal binding sites by treatment with RNAse (Richter, 1996). In contrast, in rox1/rox2 double mutants none of the MSL proteins bind to the male X chromosome. These results are not contradictory, but rather indicate that the rox RNAs are important for the association of all the MSL proteins with the X chromosome, perhaps by being required to form a stable compensasome, and also potentially required for the subsequent RNA dependent stable association of MLE with the X chromosome. Besides being important for the association of the compensasome with the X chromosome, it is possible that the rox RNAs have an additional role(s) in eliciting hypertranscription of the male X chromosome. For several other RNA-protein complexes (i.e., ribosomes, spliceosomes, and telomerases), it has been found that the RNA components of these complexes are not just structural components, but rather perform major functions within these complexes. There are several ways that the rox RNAs could potentially have direct roles in dosage compensation. (1) The rox RNAs may be important for contact between the X chromosome and the compensasome dosage compensation complex. Base pairing between the RNAs and the chromatin to physically connect the compensasome with its target sequences is attractive because the MSL proteins do not appear to have DNA binding activity. Thus, although the MLE protein has been shown to bind nucleic acids, removing this activity did not interfere with the association of MLE with the X chromosome (Lee, 1997). (2) Since an alteration in the X chromosome chromatin structure is certainly correlated with, and may be a prerequisite for, dosage compensation, the rox RNAs might be involved in this alteration of chromatin structure. (3) The rox RNAs may interact with RNA polymerase or other components of the chromatin remodeling and/or transcription machinery to more directly increase transcription rates. The recently discovered 'chromatin remodeling factors' like CHRAC, NURF, and ACF, among others, contain helicase-like components and are involved in the regulation of histone acetylation. More generally, the involvement of RNA molecules in dosage compensation raises the interesting possibility that these factors, like the dosage compensation machinery, might contain important functional RNA components, as well (Franke, 1999).

While the process of dosage compensation in both mammals and Drosophila appears to be quite different, two findings suggest that these two dosage compensation processes might have in common the basic biochemical mechanism through which they modulate X chromosome transcription. In both cases, dosage compensation involves opposite changes in the pattern of acetylation of histone H4. In addition, noncoding RNA molecules are integral components to the dosage compensation machinery in both systems. This suggests that dosage compensation in the two systems is likely mediated by similar biochemical machineries (Franke, 1999).

RNA-on-X 1 and 2 in Drosophila melanogaster fulfill separate functions in dosage compensation

In Drosophila melanogaster, the male-specific lethal (MSL) complex plays a key role in dosage compensation by stimulating expression of male X-chromosome genes. It consists of MSL proteins and two long noncoding RNAs, roX1 and roX2, that are required for spreading of the complex on the chromosome and are redundant in the sense that loss of either does not affect male viability. However, despite rapid evolution, both roX species are present in diverse Drosophilidae species, raising doubts about their full functional redundancy. Thus, this study investigated consequences of deleting roX1 and/or roX2 to probe their specific roles and redundancies in D. melanogaster. Anew mutant allele of roX2 was created, and roX1 and roX2 were shown to have partly separable functions in dosage compensation. In larvae, roX1 is the most abundant variant and the only variant present in the MSL complex when the complex is transmitted (physically associated with the X-chromosome) in mitosis. Loss of roX1 results in reduced expression of the genes on the X-chromosome, while loss of roX2 leads to MSL-independent upregulation of genes with male-biased testis-specific transcription. In roX1 roX2 mutant, gene expression is strongly reduced in a manner that is not related to proximity to high-affinity sites. These results suggest that high tolerance of mis-expression of the X-chromosome has evolved. It is proposed that this may be a common property of sex-chromosomes, that dosage compensation is a stochastic process and its precision for each individual gene is regulated by the density of high-affinity sites in the locus (Kim, 2018).

The dosage compensation machinery involving rox1 and rox2 RNAs provides a valuable model system for studying the evolution of lncRNA-genome interactions, chromosome-specific targeting and gene redundancy. LncRNAs differ from protein coding genes and are often less conserved at the level of primary sequence, as expected due to their lack of protein-coding restrictions. Like those encoding other lncRNAs, rapid evolution, i.e., low conservation of the primary sequences of roX genes has complicated comparative studies. Despite their differences in length and primary sequences, rox1 and rox2 have also been considered functionally redundant in Drosophila melanogaster. However, remarkably considering their rapid evolution and apparent redundancy, orthologs for both rox1 and rox2 have been found in all of 26 species within the Drosophila genus with available whole genome assemblies. Models that explain evolutionarily stable redundancy have been proposed suggesting that the presence of both rox1 and rox2 in these diverged species may be attributable to differences in targets, affinities and/or efficiency or additional functions (Kim, 2018).

On polytene chromosomes, binding patterns of rox1 and rox2 are more or less indistinguishable, except in region 10C where rox2 is almost exclusively present. In the rox2 mutant, genes located in the 10C bin are on average downregulated, but similar downregulation of genes in many other bins is observed, so the effect cannot be directly attributed to loss of rox2. In wildtype 1st instar larvae, levels of rox1 RNA are much higher than levels of rox2 RNA. Interestingly, in rox1 mutant larvae the absolute amount of rox2 RNA increases, but only to ~10% of wildtype levels of total roX RNA. This appears sufficient to avoid lethality, but still causes a significant decrease in X-chromosome expression. However, despite the huge difference in amounts, not only in number but even more considering the size of the two roX RNAs, the staining intensities of roX RNA on rox1 mutant and wildtype polytene chromosomes seem to be roughly equal. On mitotic chromosomes rox1 RNA is only observed in the MSL complexes bound to the distal X-chromosome and this binding is not redundant. This indicates that just after cell division rox1 RNA will be the dominating variant in assembled MSL complexes. Taken together, the results suggest that rox2 RNA has higher affinity than rox1 RNA for inclusion in MSL complexes. Moreover, varying amounts of the two species with different affinities at given cell cycle stages may support proper transmission, spreading of assembled MSL complexes and maintenance of appropriate levels of the complexes (Kim, 2018).

It should be noted that some male rox1 rox2 mutants escaped, so loss of roX is not completely male-lethal, unlike loss of mle, msl1, msl2, msl3 or mof. The complete male lethality in these mutants is attributed to reductions in dosage compensation that have been measured in several studies and observed not only in msl mutants but also following RNAi-mediated depletion of MSL proteins. Notably, the average reduction of X-chromosome expression, relative to wildtype levels, calculated in these cases has varied from ca. 20 to 30%; substantially less than the 35% reduction observed in the rox1 rox2 mutant. Some of the reported differences may be due to use of different techniques and bioinformatics procedures (including use of different cut-offs for expression and developmental stages). However, the reasons why some males can survive the very dramatic imbalance observed in expression of a large portion of the genome are unclear. Furthermore, the reduction in expression of X-chromosome genes observed in the rox1 mutant is not accompanied by any reported phenotypic changes, indicating that D. melanogaster has high intrinsic ability to cope with significant imbalances in X-chromosome expression. It is speculated that in parallel with a compensation mechanism that addresses dosage imbalances the fly has evolved a high degree of tolerance to mis-expression of the X-chromosome (Kim, 2018).

Evolutionary studies have shown that sex chromosomes do not always represent terminal stages in evolution-in fact, the 4th chromosome was ancestrally an X-chromosome that reverted to an autosome. Moreover, the fly shows high and unusual tolerance to dosage differences and mis-expression of the 4th chromosome (although much smaller than the tolerance to those of the X-chromosome). These observations suggest that tolerance of mis-expression is a common outcome in the evolution of sex-chromosomes and this property has been retained with respect to the 4th chromosome, even after its reversion to an autosome. It is proposed that high tolerance of mis-expression in the absence of full functional dosage compensation may be selected for during evolution of sex-chromosomes. This is because gradual degeneration of the proto-Y chromosome will be accompanied by an increasing requirement to equalize gene expression between a single X- (in males) and two X-chromosomes (in females), but changes in genomic location of highly sensitive genes will be favored during periods of incomplete (or shifting) dosage compensation. On transcript level, responses to reductions in dosages of X-chromosome genes have been found to be similar to those of autosomal genes. Thus, potential mechanisms for the higher tolerance are post-transcriptional compensatory mechanisms or selective alterations in gene composition (changes in genomic locations), similar to those proposed for the observed demasculinization of the Drosophila X-chromosome (Kim, 2018).

Prompted by the strong relationship between orchestration of the X- and 4th chromosomes by the MSL complex and POF system, respectively, this study also measured effects of roX suppression on chromosome 4 expression in roX mutants. Weak but significant reduction of expression were observed in the rox2 mutant, but the cause of this reduction remains elusive. In rox2 mutant transcriptional upregulation of X-chromosome genes classified as having low expression levels was also oberved, late replication and weak MSL complex-binding. The loss of rox2 resulting in MSL complexes only including rox1 RNA might alter the spreading properties. It is therefore hypothesized that the observed upregulation might be caused by mis-targeting of the MSL complex in the absence of rox2. However, the ChIP experiment revealed no enrichment of MSL complexes on these genes, and results rather suggest that rox2 directly or indirectly restricts expression of these male-biased genes independently of its role in the MSL complex (Kim, 2018).

It is well known that roX RNAs are important for spreading of the MSL complex in regions between high affinity sites (HAS). It is therefore surprising that loss of roX causes a relatively even reduction in expression of X-chromosomal genes and the decrease is not more dramatic with larger distances, as would be expected for reductions in spreading capacity. Indeed, observed reductions in expression were smaller for genes located far from HAS than for closer genes. A possible explanation is that expression of these genes is compensated by an MSL-independent mechanism. It has been previously shown that most genes on the X-chromosome are dosage-compensated, but a subset are not bound by the MSL complex and do not respond to its depletion. The results corroborate these findings since loss of roX RNA in the rox1 rox2 mutant had little effect on the expression of genes classified as having weak MSL complex binding, clearly indicating that at least one other mechanism is involved. The results further show that high-affinity sites, as defined by MSL-targets in the absence of rox1 and rox2, are highly correlated to genes with the highest MSL binding levels. Therefore, sites targeted in the absence of roX provide a more stringent definition of HAS, with stronger correlation to genes bound by high levels of MSL complex, than targets in the absence of mle, mof or msl3 (Kim, 2018).

The increase in expression mediated by the MSL complex is considered a feed-forward mode of regulation, and appears to be more or less equal (ca. 35%) for all MSL-bound genes. Evidently, highly expressed genes need a stronger increase in transcription than weakly-expressed genes. These results suggest that dosage compensation is a stochastic process that depends on HAS distribution and is correlated with expression levels. Evolutionary analysis has shown that newly formed X-chromosomes acquire HAS, putatively via rewiring of the MSL complex by transposable elements and fine-tuning of its regulatory potential. Such a dynamic process may be required for constant adaptation of the system. Highly expressed genes tend to accumulate HAS in their introns and 3'UTRs, and thus bind relatively high amounts of MSL complex, thereby stimulating the required increase in expression. This also implies that the gene organization on X-chromosomes is under more constraints than autosomes (Kim, 2018).

This study presents the first high-throughput sequencing data and analysis of transcriptomes of rox1, rox2 and rox1 rox2 mutant flies. The results reveal that rox1 and rox2 fulfill separable functions in dosage compensation in D. melanogaster. The two RNA species differ in both transcription level and cell-cycle regulation (Kim, 2018).

In third instar larvae, rox1 is the more abundant variant and the variant that is included in MSL complexes transmitted physically associated with the X-chromosome in mitosis. Loss of rox1, but not loss of rox2, results in decreased expression of genes on the X-chromosome, albeit without apparent phenotypic consequences. Loss of both roX species leads to a dramatic reduction of X-chromosome expression, but not complete male lethality. Taken together, these findings suggest that high tolerance for mis-expression of X-chromosome genes has evolved. It is speculated that it evolved in parallel with dosage compensation mechanisms and that it may be a common property of current and ancient sex-chromosomes (Kim, 2018).

The roX RNAs are important for spreading of the MSL-complex from HAS, but the reduction of X-chromosome expression in rox1 rox2 mutant is not affected by the need for spreading, i.e., distance from HAS. In addition, the genes targeted by the MSL complex in the rox1 rox2 mutant also show strongly reduced expression. The results suggest that the function of the MSL complex which is still present at HAS is compromised in the rox1 rox2 mutant and that the dosage of distant genes is compensated by an alternative, unknown, mechanism. It is proposed that dosage compensation is a stochastic process that depends on HAS distribution. Creation and fine-tuning of binding sites is a dynamic process that is required for constant adaptation of the system. Highly expressed genes will accumulate and be selected for strong HAS (and thus bind more MSL complex) since they require high levels of bound MSL complex for the required increases in expression (Kim, 2018).

Transcription rate of noncoding roX1 RNA controls local spreading of the Drosophila MSL chromatin remodeling complex

The dosage compensation complex in Drosophila is composed of at least five MSL proteins and two noncoding roX RNAs that bind hundreds of sites along the single male X chromosome. The roX RNAs are transcribed from X-linked genes and their RNA products 'paint' the male X. The roX RNAs and bound MSL proteins can spread in cis from sites of roX transcription, but the mechanism controlling spreading is unknown. This study found that cis spreading from autosomal roX1 transgenes is coupled to the level of roX transcription. Low to moderate transcription favors, and vigorous transcription abolishes local spreading. A roX1 minigene one third the size of wild type was constructed as a starting point for mutagenesis. This allowed tests of which evolutionarily conserved motifs were required for activity. One short repeat element shared between roX1 and roX2 was found to be particularly important. When all copies were deleted, the RNA was inactive and unstable, while extra copies seem to promote local spreading of the MSL complex from sites of roX1 synthesis. It is proposed that assembly of the MSL proteins onto the extreme 3' region of elongating roX1 transcripts determines whether the MSL complex spreads in cis (Kelley. 2008).

The genomes of twelve Drosophila species covering approximately 40 Myr of evolution have been sequenced. Each contains a homolog of roX1 and roX2. A simple BLAST search failed in more distantly related species forcing a search with adjacent protein coding gene sequences in the hope that the roX genes had remained in the same syntenic arrangement over time. A combination of computational and manual methods produced a tentative alignment. Being free of the constraints of a translational reading frame, the roX genes have accumulated a huge number of indels (insertions/deletions) in addition to simple nucleotide substitutions. This, along with simple repeats and extensive runs of polyA partially defeated computational alignment methods necessitating manual inspection to identify meaningful islands of homology separated by variable spans of seemingly unrelated sequence (Kelley. 2008).

Focus was placed on the 600 nucleotides at the 3' end of roX1 that previous analysis indicated was important (Stuckenholz, 2003). Several features were apparent from the species alignment. First, the large stem-loop (SL1) analyzed earlier was well conserved in most species except virilis, mojavensis, and grimshawi. A second predicted stem-loop (SL2) followed shortly afterwards, which shows better base pairing potential in non-melanogaster species. It was noticed that rather than the single copy of the 'roXbox' element (RB) previously described (Franke, 1999; Stuckenholz, 2003]) three imperfect copies and a fourth were found in the inverse orientation in all species. Any of the three RBs could potentially base pair with the upstream inverted copy to form three alternative dsRNA stems with loops of variable length, although RB1 is closest in both proximity and sequence complementarity. The genomic sequence suggests that pseudoobscura and persimilis both have four additional copies of the RB. After these sequences are accounted for, the remainder of the region is remarkably rich in runs of (A)_n although the exact length and locations of these varies widely from one species to another. Next the much smaller roX2 sequences were examined. Again, it was noticed that rather than one copy of the RB noted earlier (Franke, 1999), three copies were present in all species with additional copies in a few species. Inverted copies were also found lying near some RBs suggesting that these might fold into dsRNA stem-loops. Again, once the RBs and inverse RB sequences were set aside, the remainder of the sequence was poorly conserved and full of short runs of (A)_n (Kelley. 2008).

Each of these elements as well as one upstream region (Region 1) were deleted from the H83roX1Δ39 minigene and their ability to support dosage compensation was assayed in transgenic male flies. Initial analysis using standard P element transformation provided suggestive results, but the wide variation from one random insertion site to another made direct comparisons between mutants difficult. Therefore all of the deletion mutants were inserted into the same genomic location using the φC31 site-specific integration system. After sampling several attB integration sites, VK11 located at 40E at the base of chromosome 2L was chosen because a nearly full-length 3.4 kb roX1 cDNA at this site gave almost complete male rescue and the 1.2 kb H83roX1Δ39 minigene rescued 47% male viability when the transgenes provided the only source of roX RNA. When all the other deletion mutants were integrated at the same site most continued to rescue male viability near 50%. However, removing SL1 nearly destroyed function, and loss of SL2 reduced activity to a lesser degree. Loss of the very A rich region also reduced activity. Surprisingly, removing any one copy of the RB element or the complementary IRB had no effect on activity. However, removal of all three RB copies or the entire interval containing them completely destroyed activity. Northern analysis showed that roX1 RNAs lacking any RB element were unstable (Kelley, 2008).

A hybrid roX1 minigene, H83roX1Δ39pseudo, was constructed in which the first 600 nt were derived from the 5' end of the D. melanogaster roX1 gene and the last 600 nt were derived from the 3' end of the D. pseudoobscura roX1 gene. Despite the overall poor conservation of primary sequence, many indels between the two distantly related species, and a weak SL2-like element, the hybrid minigene rescued roX1 roX2 male viability to the same degree as the all melanogaster construct. The MSL complex containing only the hybrid H83roX1Δ39pseudo RNA painted the male X in a normal pattern demonstrating that the many sequence differences were inconsequential. However, a striking difference was noted between the melanogaster Δ39 minigene and the pseudoobscura hybrid. The latter supported MSL spreading around the autosomal insertion sites tested despite being driven by the strong hsp83 promoter while none of the melanogaster constructs did. It is not certain which of the many sequence differences are responsible for this enhanced spreading behavior, but the most obvious difference is that the pseudoobscura sequence used here carries six copies of RB at its 3' end rather than three (Kelley, 2008).

This study has shown that only sequences within the roX1 RNA are needed for rox1 spreading. No special spreading initiation sites flank the roX1 gene, and the local chromatin environment surrounding the roX1 transgene is of little importance. Whether or not MSL complex spreads from sites of roX1 transcription depends upon how vigorously the roX1 gene is transcribed. Weak to moderate roX1 transcription favors local spreading of MSL complex surrounding the roX1 gene, but strong transcription favors release of soluble RNA that eventually locates the X in trans. These results are most consistent with a model postulating that MSL subunits begin assembling onto nascent roX transcripts as they emerge from RNA polymerase. The ratio of free MSL protein subunits relative to nascent transcripts strongly influences the final destination of mature complex. This 'Race to Assemble' model was formulated to explain the unexpected MSL distribution pattern seen when the number of roX genes was manipulated in males or when the abundance of MSL subunits was altered. This study varied the amount of transcription from a roX1 transgene and came to the same conclusion. Spreading might require efficient assembly of a mature, functional complex while the nascent roX1 transcript is still tethered to the RNA polymerase. It is proposed that the failure to spread under conditions of abundant roX1 RNA synthesis is because earlier RNA molecules consumed the pool of free MSL subunits available for assembling later roX transcripts. Furthermore, the dense cloud of elongating transcripts may sterically interfere with efficient recruitment of MSL subunits. Consequently, most transcripts are released with an incomplete set of subunits and only finish assembly while diffusing away from the site of transcription (Kelley, 2008).

What biological purpose is served by linking MSL spreading to roX1 transcription is uncertain. Transcriptional control of roX1 varies during development. Perhaps this linkage is useful during early embryogenesis as dosage compensation is being established. Additionally, it can be imagined that production of MSL complex might experience a burst during DNA replication much like histone synthesis. The roX1 gene may need to adjust either its own output or the distribution of mature complexes in order to reestablish the proper level of dosage compensation (Kelley. 2008).

An earlier report concluded that local RNA synthesis was not necessary for MSL complex to spread from a promoterless roX1 cDNA. However, in that study the two transgene insertion sites that supported MSL spreading were found to make low levels of roX1 RNA presumably due to flanking promoter read-through. This is compatible with the current view that low roX transcription favors local spreading (Kelley. 2008).

A second key finding is that when MSL complex spreads into flanking autosomal chromatin, it remains dependent upon constant replenishment from the nearby roX1 locus. Shifting from low to high roX1 transcription caused loss of bound MSL complex surrounding an autosomal transgene after a few hours. This seems counterintuitive because one might have expected that abundant production of roX RNA would have increased the opportunity for newly assembled MSL complex to bind nearby chromatin. Although indirect, this observation is the strongest evidence to date in favor of the idea that MSL complex assembly begins cotranscriptionally, and mature complexes are completed near the roX1 locus only under conditions of moderate to low roX1 RNA synthesis (Kelley. 2008).

Dosage compensation is normally established around the onset of gastrulation. One might imagine that the pattern of target genes bound by the MSL complex is set up once and remains stable over the life of the male, or dosage compensation might respond to changes as gene are utilized in different tissues and times of development. The observation that a nearly 'naked' X can acquire a normal MSL pattern within a few hours of inducing roX RNA synthesis argues against any strict requirement for establishing MSL binding during embryonic development. However, the possibility that the very low levels of MSL complex made early may have somehow prepared the X for later robust binding cannot be excluded (Kelley, 2008).

A sequence of approximately 220 bp forms a male-specific DNase I hypersensitive site (DHS) near the middle of the roX1 gene which is the primary site bound by the MSL complex. This site appears to act as a complex transcriptional control element causing silencing in females and activation in males. MSL complex clearly binds the DHS sequence in vivo when the gene is not transcribed, during very low transcription, and even when transcribed from the hsp83 promoter. However, the MSL complex bound to the UASroX1 transgene could not be detected when GAL4 drives transcription much higher than the wild type rate. This might reveal a negative feedback mechanism that places an upper limit on roX1 expression. In wild type males, as MSL proteins successfully stimulate roX transcription, they might reach a point where they are ejected from the roX gene by a high density of RNA polymerase molecules transiting the DHS. This would ensure that the MSL proteins could drive roX transcription to the correct level, but no higher. This might explain why the MSL-dependent enhancer in located inside the roX1 transcription unit. Although this arrangement results in the DHS sequence being carried along in the mature roX1 RNA, it seems to have no function at the RNA level. In the artificial situation created in this study, roX transcription no longer depends upon MSL proteins, so extraordinarily high roX1 synthesis continues unabated (Kelley, 2008).

Although an initial comparison between roX1 and roX2 failed to detect any obvious similarities at the primary sequence level, later inspection revealed a single 30 nt element (here called the roXbox, RB) at the 3'ends of both roX1 and roX2. Deletion of this element did not result in any significant phenotype. The availability of 12 sequenced Drosophila species genomes allowed determination of which segments of roX1 were best conserved and potentially functionally important. The resulting sequence alignment revealed that the RB was actually present in at least three to seven copies in the roX1 gene of different species, and at least three copies were also present at the 3' end of roX2. Bolstering the notion that these elements are significant, additional conserved copies were found in the inverse orientation that might form dsRNA secondary structures (Kelley. 2008).

Removing the central 70% of the 3.7 kb roX1 sequence only reduces in vivo activity by half, and is fully consistent with prior work. Deletion analysis of the mammalian Xist gene came to a similar conclusion. Small deletions in the 1.2 kb H83roX1Δ39 minigene display more drastic phenotypes than reported for larger deletions in the full-length gene. This is most likely due to the removal of partially redundant domains in the middle of roX1 and the use of a single chromosomal integration site to assay all mutants. Using this system showed that removing any one copy of RB had little effect. However, removing the interval containing all three copies produced an unstable RNA that could no longer support male viability. A similar loss of activity was found when just the three RB elements were deleted precisely. The predicted base pairing between IRB and any copy of RB is not essential for dosage compensation. However, its conservation over time suggests IRB plays some role not measured in this assay or another unidentified sequence can substitute for it (Kelley. 2008).

One might have imagined that the extensive divergence of roX1 primary sequence has been accompanied by equally dramatic coevolution of the MSL subunits. Indeed, while MOF, MSL3, and MLE are very similar across fly species, MSL1, and MSL2 have diverged extensively outside small domains known to contact other protein subunits. This might result in incompatible protein:RNA combinations if components from different species were mixed. A hybrid roX1 minigene was constructed whose first 600 nt came from the 5' end of the D. melanogaster homolog and the last 600 nt from highly diverged D. pseudoobscura gene. This supported dosage compensation in transgenic D. melanogaster males whose endogenous roX1 and roX2 genes had been deleted. This argues that roX1 RNAs from pseudoobscura and melanogaster share a core structure despite the divergence in sequence. However, the much more striking result was that the hybrid RNA promoted local MSL spreading around the sites of the autosomal transgenes. In contrast, the fully melanogaster minigene did not support any MSL binding at or around the transgene. The latter result is easily understood because loss of the DHS sequence from the Δ39 minigene removes the primary MSL binding site within the roX1 gene, and transcription from the strong hsp83 promoter inhibits local spreading of even a full-length roX1 cDNA for reasons discussed above. It is speculated that increasing the number of RBs from three to six somehow tips the balance towards local spreading in spite of vigorous transcription from the hsp83 promoter. The fact that the much smaller H83roX1 Δ39pseudo minigene is still capable of supporting local spreading argues that the 2.4 kb of internal sequence removed was not essential for this aspect of roX1 function, but might somehow enhance it (Kelley. 2008).

If complex assembly does begin cotranscriptionally, then the functionally important sequences studied in this study at the very end of the transcript might link MSL subunit binding to 3' processing and release the transcript from RNA polymerase. This might then influence how the newly formed complex is distributed locally or to more distant targets. Removing all RBs may abort assembly leading to destruction of the defective RNA. Conversely, expanding the number of RB copies may enhance some aspect of assembly required for local spreading (Kelley. 2008).

Autoregulation of the Drosophila Noncoding roX1 RNA Gene

Most genes along the male single X chromosome in Drosophila are hypertranscribed about two-fold relative to each of the two female X chromosomes. This is accomplished by the MSL (male-specific lethal) complex that acetylates histone H4 at lysine 16. The MSL complex contains two large noncoding RNAs, roX1 (RNA on X) and roX2, that help target chromatin modifying enzymes to the X. The roX RNAs are functionally redundant but differ in size, sequence, and transcriptional control. This study examined how roX1 production is regulated. Ectopic DC can be induced in wild-type (roX1⁺) roX2⁺) females if a heterologous source of MSL2 is provided. However, in the absence of roX2, it was found that roX1 expression failed to come on reliably. Using an in situ hybridization probe that is specific only to endogenous roX1, it was found that expression was restored if either roX2 or a truncated but functional version of roX1 were introduced. This shows that pre-existing roX RNA is required to positively autoregulate roX1 expression. Massive cis spreading of the MSL complex was also observed from the site of roX1 transcription at its endogenous location on the X chromosome. It is proposed that retention of newly assembled MSL complex around the roX gene is needed to drive sustained transcription and that spreading into flanking chromatin contributes to the X chromosome targeting specificity. Finally, it was found that the gene encoding the key male-limited protein subunit, msl2, is transcribed predominantly during DNA replication. This suggests that new MSL complex is made as the chromatin template doubles. A model is offered describing how the production of roX1 and msl2, two key components of the MSL complex, are coordinated to meet the dosage compensation demands of the male cell (Lim, 2012).

Previous studies of roX1 transcriptional control argued that either MSL2 alone or with a full set of MSL proteins was sufficient to drive male-specific expression. This study presents evidence that the expression of roX1 gene is instead controlled through an autoregulatory loop. Pre-existing roX RNA, presumably in mature MSL complex, is required to drive new transcription. The reason this study reached a different conclusion is largely attributed to removing the functionally redundant roX2 in most of the experiments and assaying transcription only from the wild type roX1 locus at its normal location on the X (Lim, 2012).

A model is presented for how such an autoregulatory loop might operate (see Autoregulation model). Because roX RNA is not maternally deposited into embryos, one problem is how male embryos could build their first MSL complex needed to initiate the cycle. Earlier studies have shown that roX1 transcription switches on in both sexes around blastoderm, just as general zygotic transcription begins. This suggests that an embryonic roX1 promoter is active without MSL complex and could supply the first roX1 RNA molecules to males, but these RNAs are eventually degraded in females. Early RNA assembles with MSL proteins and then drive roX1 transcription from the known male-specific MSL-dependent promoters, setting up a positive autoregulatory loop necessary for the future maintenance of roX1 expression in males. In the current experiments, it was shown that roX2 RNA or truncated roX1 RNA can also initiate endogenous roX1 expression late in development after the early roX1 transcripts are gone (Lim, 2012).

This model requires that male embryos preferentially sequester newly assembled MSL complex at the roX1 gene to drive sustained transcription instead of allowing it to diffuse away to the vastly larger pool of ordinary X linked genes that must be dosage compensated. Only after roX1 transcription is successfully upregulated can MSL complexes be released to the target genes along the X chromosome. Such behavior has been previously documented in cells containing abundant free MSL subunits and low levels of roX1 transcription, exactly the conditions believed to occur as young male embryos initiate dosage compensation. Examining cells shortly after MSL2 first turned on roX1 transcription showed the earliest roX1 transcripts remained near the site of synthesis consistent with the idea that newly formed MSL complexes preferentially act on the roX1 gene. At later times, such as seen in H83roX1-Δ39 animals that have had five days to drive endogenous roX1 expression, every cell painted the entire X with roX1 RNA. While massive local cis spreading of the MSL complex has been reported for roX1 and roX2 transgenes inserted into autosomes, the physiological relevance of this is not widely accepted. This study instead reports striking local cis spreading of newly made MSL complex from the wild type roX1 gene in its normal X chromosome environment. The data agree well with previous reports of local MSL spreading along the X and support a role for cis spreading in the normal process of dosage compensation (Lim, 2012).

It has not been directly determined what region of the roX1 gene is necessary for autoregulation. However, a strong candidate is the ~200 bp male-specific DNase I hypersensitive site (DHS) found about 1.5 kb downstream of the adult roX1 promoters. The DHS is sufficient to recruit the MSL complex to ectopic sites when moved to the autosomes in a sequence specific manner. While partial complexes lacking MOF or MSL3 remain bound to the DHS, incomplete complexes lacking either roX RNA or the MLE RNA helicase postulated to fold roX RNA bind the DHS poorly. This element stimulates roX1 transcription when MSL complex is bound and represses basal transcription when MSL complex is absent. Also, deleting the DHS greatly reduces transcription of roX1 transgenes. Together, these findings support a model where transcription from the roX1 gene requires pre-existing roX RNA within MSL complexes bound to the internal DHS enhancer. However, this view is likely an oversimplification because very large internal deletions such as roX1^ex7B, comparable to the H83roX1-Δ39 transgene, remain transcriptionally active despite the loss of the DHS enhancer (Lim, 2012).

Translation of msl2 mRNA is normally subject to elaborate controls acting through the 5'and 3' UTRs. Little attention has been given to its transcription control, although recently anti-MSL2 antibodies were found to precipitate msl2 mRNA (Johansson, 2011). msl2 transcription is associated with replication, and this is likely important for its normal control. Without MSL2 protein, naked roX1 RNA is rapidly destroyed (Meller, 2000). This study has found the converse; cells rapidly clear any MSL2 protein not bound to roX RNA. When free MSL2 subunits are artificially stabilized with proteosome inhibitors, they coat all the chromosomes indiscriminately. This implies that the synthesis of MSL2 and roX RNAs are closely coordinated so each component stabilizes the other ensuring that only correctly targeted molecules survive. Although MSL2 lacks known RNA binding motifs, previous work of others is consistent with an intimate interaction between MSL2 and roX RNA (Li, 2008). It is suspected the loss of replication-coupled transcription may contribute to the failure of dosage compensation in some cells relying exclusively on H83M2 for MSL2 protein and roX1 for roX RNA. This defect is corrected when both roX1 and MSL2 are coordinately made from the same hsp83 promoter. Cells in mosaic animals lose dosage compensation sometime between the end of embryogenesis and third instar larvae. Many tissues undergo significant changes in cell cycle near the end of embryogenesis, particularly the introduction of G1, and it is suspected that this shift contributes to loss of dosage compensation in mosaic animals. If MSL complex should ever drop below the level needed to sustain the autoregulatory roX1 loop, it could never recover regardless of later MSL2 production. The details remain unclear because H83M2, the construct used to drive MSL2, is vigorously transcribed at the developmental times examined, including S phase. It is not known whether the regulatory msl2 UTR sequences removed from H83M2 disrupt additional posttranscriptional controls that might promote efficient translation during replication (Lim, 2012).

A second issue is that the X is painted with MSL complex throughout the cell cycle. This might drive continuous rather than cyclic roX1 synthesis. While it was not possible to directly determine if roX1 transcription is cell cycle regulated, it is noted that the replication machinery components ORC2 and MCM are bound specifically to the roX1 DHS enhancer only in male cells. The significance of this is not known, but it is tempting to speculate that components of the pre-initiation complex bound to DHS compete with MSL complex thus inhibiting roX1 transcription in G1. Firing of the replication origins removes ORC2 and MCM possibly allowing MSL complex access to the DHS and so switch on transcription shortly after the onset of S phase. The replication machinery is commonly found near the promoters of many genes, so further experiments will be needed to determine if such binding actually plays any regulatory role here (Lim, 2012).

While this study did not specifically address transcriptional control of roX2, it must differ from roX1 in several important ways. Meller (2003) has previously shown that roX2 transcription lags roX1 by a few hours during early embryonic development and is always limited to males. This study found that roX2 transcription differs by not requiring pre-existing roX RNA and can be switched on several days later during larval development simply by ectopically expressing MSL2. MSL2 protein made by the constitutive H83M2 transgene only sporadically activates roX1, but robustly drives roX2. The roX2 gene also carries a DHS enhancer similar to that found in roX1 (Park, 2003), but if it plays a comparable role in roX2 regulation, it presumably would not require complete MSL complex. The region near the proline rich domain towards the C-terminus of MSL2 is essential for this regulation (Lim, 2012).

The roX1 autoregulation loop described above shares parallels to the SXL autoregulatory loop controlling all aspects of sex determination and dosage compensation in Drosophila. X:A counting elements act upon an early establishment Sxl promoter to make SXL in early female embryos. These first SXL proteins stimulate productive splicing of Sxl mRNAs transcribed from a distinct maintenance promoter ensuring further SXL production. MSL1, MSL3, MOF, and MLE are unable to package early roX1 RNA made from the embryonic promoter in female embryos to form a fully functional mature MSL complex. To do that, MSL2 protein made only in males is required. These earliest MSL complexes are thought to sustain roX1 transcription as embryos switch to the MSL-dependent promoters. Recently a new role for MSL2 in females has been described during the brief window when Sxl autoregulation is established (Gladstein, 2010). Perhaps females also fleetingly utilize the early burst of roX1 before abundant SXL represses msl2 translation (Lim, 2012).

Thousands of large noncoding RNAs have recently been discovered in vertebrates, many of which are associated with chromatin remodeling enzymes. It is likely that some of these will face similar regulatory and functional demands as the roX RNAs and may have evolved comparable strategies to control their production. For instance, the short RepA sequence at the 5′ end of mammalian Xist RNA may influence production of full length transcripts (Lim, 2012).

GENE STRUCTURE

Two transcripts of 3.6 and 3.8 are expressed in adult flies. Two cDNAs are similar in size and sequence, while a third contains a 68 base pair intron (Meller, 1997).

roX1 and roX2 are both linked to female-specific genes. roX2 is closely linked (within 1kb) to a female-specific gene, no distributive disjunction (nod). nod has been mapped to 10B on the X chromosome, the same chromosomal region in which roX2 resides. nod is a member of the kinesin family of genes shown to be essential for proper segregation of nonexchange chromosomes during female meiosis. roX1 is closely linked to oligo peptide transporter 1 (opt1), a gene that shows high homology to members of the class of 12 transmembrane domain oligopeptide transporters. Expression of OPT1 mRNA is largely restricted to nurse cells in the ovary (Amrein, 1997).

Exons - 2

date revised: 15 October 99  

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.