siRNA that participates in Drosophila dosage compensation is produced by many 1.688X and 359 bp repeats

Abstract Organisms with differentiated sex chromosomes must accommodate unequal gene dosage in males and females. Male fruit flies increase X-linked gene expression to compensate for hemizygosity of their single X chromosome. Full compensation requires localization of the Male-Specific Lethal (MSL) complex to active genes on the male X, where it modulates chromatin to elevate expression. The mechanisms that identify X chromatin are poorly understood. The euchromatic X is enriched for AT-rich, ∼359 bp satellites termed the 1.688X repeats. Autosomal insertions of 1.688X DNA enable MSL recruitment to nearby genes. Ectopic expression of dsRNA from one of these repeats produces siRNA and partially restores X-localization of MSLs in males with defective X recognition. Surprisingly, expression of double-stranded RNA from three other 1.688X repeats failed to rescue males. We reconstructed dsRNA-expressing transgenes with sequence from two of these repeats and identified phasing of repeat DNA, rather than sequence or orientation, as the factor that determines rescue of males with defective X recognition. Small RNA sequencing revealed that siRNA was produced in flies with a transgene that rescues, but not in those carrying a transgene with the same repeat but different phasing. We demonstrate that pericentromeric X heterochromatin promotes X recognition through a maternal effect, potentially mediated by small RNA from closely related heterochromatic repeats. This suggests that the sources of siRNAs promoting X recognition are highly redundant. We propose that enrichment of satellite repeats on Drosophilid X chromosomes facilitates the rapid evolution of differentiated sex chromosomes by marking the X for compensation.


Introduction
Many species have evolved highly differentiated X and Y chromosomes that create an imbalance in gene dosage between males and females (Disteche 2012).To ensure equivalent expression of X-linked genes, these species have evolved a process termed dosage compensation.Mammals achieve dosage compensation by inactivating one of the two X chromosomes in females, but Drosophila employ a different strategy (Charlesworth 1996;Lucchesi et al. 2005).Fruit fly males upregulate expression from their single X chromosome about two-fold (Lucchesi 1978).A ribonucleoprotein complex called the Male-Specific Lethal complex or Dosage Compensation Complex (MSL or DCC) is necessary for full dosage compensation in male flies (Gelbart and Kuroda 2009).The MSL complex is made up of five proteins, Male-Specific Lethal 1, -2 and -3 (MSL1; MSL2; and MSL3); Maleless (MLE) and Males absent On the First (MOF).MOF, a histone acetyltransferase, acetylates histone H4 at lysine 16 (H4K16Ac) (Hilfiker et al. 1997;Smith et al. 2000).Acetylation at this position is associated with active genes and decondensed chromatin (Grunstein 1997;Shogren-Knaak et al. 2006).H4K16 is highly enriched over the body of X-linked genes and this mark is thought to be responsible for transcriptional upregulation in males (Bone et al. 1994;Hilfiker et al. 1997;Akhtar and Becker 2000;Smith et al. 2000;Copur et al. 2018).In addition to these five proteins, one of two functionally redundant long noncoding RNAs, RNA on the X 1 and -2 (roX1 and roX2), also participate in MSL complex formation (Amrein and Axel 1997;Meller et al. 1997;Ilik et al. 2013).Simultaneous mutation of both roX genes disrupts X-localization of MSL proteins and leads to male lethality (Meller and Rattner 2002;Deng et al. 2005;Deng and Meller 2006).
Although the components of the dosage compensation complex and their functions are well studied, how the MSL complex selectively recognizes the X chromosome is not clear.One model suggests that the MSL complex forms on roX transcripts and is recruited to 200-300 X-linked Chromatin Entry Sites, also termed High Affinity Sites (CES, HAS) (Kelley et al. 1999;Park et al. 2002;Oh et al. 2003).CES are enriched for a 21 bp GA-rich MRE (MSL Recognition Element) motif (Alekseyenko et al. 2008).A zinc finger protein called Chromatin-Linked Adaptor for MSL Proteins (CLAMP), binds MRE sequences throughout the genome, but selectively recruits the MSL complex to X-linked CES (Soruco et al. 2013).The MSL complex then spreads to nearby genes through binding of MSL3 to the co-transcriptional H3K36me3 mark (Bell et al. 2007;Kind and Akhtar 2007;Larschan et al. 2007;Sural et al. 2008).Although autosomal MRE motifs bind CLAMP, they do not recruit the MSL complex.A CES variant termed PionX (Pioneering sites on the X) is reported to interact with CLAMP and a domain of MSL2 (Villa et al. 2016).These sites are among https://doi.org/10.1093/genetics/iyae074Advance Access Publication Date: 8 May 2024

Investigation
the first CES to be bound during the assembly of MS complex and are enriched on the X chromosome, suggesting a role in distinguishing the X chromosome and autosomes (Albig et al. 2019).
Both roX genes are X-linked and overlap CES.roX transgenes recruit compensation to nearby active genes when inserted on an autosome, suggesting a role in marking the X chromosome (Kelley et al. 1999;Henry et al. 2001;Joshi and Meller 2017).However, roX RNA produced from an autosomal transgene is incorporated into the MSL complex and travels to the X chromosome to rescue compensation in roX1 roX2 males (Meller and Rattner 2002).This indicates that roX genes are not sufficient to confer X-identity.A chromosome conformation study demonstrated that CES are spatially arranged close to each other in the male interphase nucleus (Ramírez et al. 2015).Long-range interactions near active X-linked genes are more pronounced in males compared to females, which suggests that chromosome organization contributes to X recognition and dosage compensation (Grimaud and Becker 2009;Pal et al. 2019).
Previous studies from our lab have established a role for small interfering RNA (siRNA) in X recognition (Menon and Meller 2012;Menon et al. 2014).There are hundreds 1.688 X satellite repeats (CsCl density of 1.688 g/cm 3 ) in short, tandem repeats in euchromatin of the Drosophila X chromosome (Hsieh and Brutlag 1979;Waring and Pollack 1987;DiBartolomeis et al. 1992;Kuhn et al. 2012).These are AT-rich, have a repeat unit of ∼359 bp and are dissimilar to CES.Many 1.688 X repeats are transcribed and small RNA aligning to 1.688 X repeats is present in early embryos and the maternal germ line, but rare or absent in later development (Menon et al. 2014).Ectopic expression of double-stranded RNA (dsRNA) from a repeat at cytological position 3F (1.688 3F ) dramatically rescued the survival of roX1 roX2 males and partially restored X-localization of the MSL proteins (Menon et al. 2014).Sequencing revealed substantial production of siRNAs upon 1.688 3F dsRNA expression.In addition, autosomal transgenes containing 1-2 kb of DNA from 1.688 3F , as well as repeats at 1A and 3C (1.688 1A , 1.688 3C ) recruit MSL proteins in males and increase expression of genes up to 150 kb away (Joshi and Meller 2017;Deshpande and Meller 2018).This suggests a role for 1.688 X DNA in marking the X, and for small RNA from these repeats in the process of X recognition.
Repeats at 1.688 1A , 1.688 3C , and 1.688 4A were selected to create dsRNA-expressing transgenes as these repeats range from very similar to quite divergent from 1.688 3F (89, 67 and 95% identity, respectively).Surprisingly, dsRNA-producing transgenes constructed with 1.688 1A , 1.688 3C , and 1.688 4A sequences failed to rescue roX1 roX2 males (Menon et al. 2014, and unpublished).This was baffling because insertion of 1.688 1A or 1.688 3C DNA on an autosome attracts the compensation machinery to nearby genes, indicating that these repeats do contribute to marking the X for compensation (Joshi and Meller 2017).Intriguingly, 1.688 3F is immediately distal to roX1.The Drosophila genome annotation depicts a roX1 transcript extending through the 1.688 3F repeat, but we have been unable to detect this extended transcript by amplification of cDNA and believe that virtually all roX1 transcripts end at termination signals between roX1 and the 1.688 3F repeats.Nevertheless, the proximity of these elements is provocative and suggested that the 1.688 3F repeat might have a unique biological function.Notably, roX1 is expressed 1.5 h after egg laying (AEL) and supports initial X recognition (Meller 2003).In roX1 mutant males, X-localization of the MSL proteins is delayed until the onset of roX2 expression at 6 h AEL.These observations suggested that two RNA-producing elements at 3F might collaborate to drive initial X recognition.
It is possible that only siRNA from 1.688 3F is capable of promoting X recognition and achieving rescue of roX1 roX2 males, and that sequence differences between the 1.688 X repeats previously tested are critical.Alternatively, details of construction of the transgenes expressing dsRNA from each repeat could determine function.To address these questions, we designed studies to determine if repeat sequence determines biological activity, or if the details of transgene construction play a crucial role.
We show that dsRNA from the 1.688 1A and 1.688 3C repeats rescues roX1 roX2 males when the expressing transgenes are constructed with the same repeat phasing as used to generate the original 1.688 3F dsRNA construct.Selection of these repeats enables direct comparisons between the original 1.688 1A and 1.688 3C dsRNA transgenes that fail to rescue roX1 roX2 males and reconstructed versions containing 1.688 1A and 1.688 3C sequences.Small RNA sequencing revealed that 1.688 1A dsRNA is processed into small RNA when the phasing of the repeat matches that of the original 1.688 3F dsRNA construct.In contrast, virtually no siRNA aligning to 1.688 1A was detected in flies expressing dsRNA 1.688 1A from the original transgene that does not rescue roX1 roX2 males.We conclude that the distinction between the repeats previously tested is due to details of transgene construction, not DNA sequence.This suggests that many of the 1.688 X repeats on the X are capable of producing small RNA promoting X recognition.We provide evidence that the closely related pericentromeric 359 bp repeats, comprising 10 Mb of X heterochromatin, as well as shorter blocks on the second and third chromosomes, act maternally to promote X recognition, possibly through production of siRNAs that are transmitted to the zygote.These findings indicate high redundancy in the sources of satellite siRNAs that promote X recognition.

Fly culture and genetics
Flies were maintained at room temperature on yeast and molasses media.Strains used in these studies are presented in Supplementary Table 1.Transgenes were injected by Rainbow Transgenics (Camarillo, CA).Three independent second chromosome insertions of each construct were selected for further studies.Integrations of roX1 and 1.688 3F recruiting elements in the haf gene were previously described (Joshi and Meller 2017).Integrations were recombined with p[w + Sqh-Gal4]2 and homozygous adults mated to flies carrying reconstructed pWIZ transgenes.Groups of 60 larvae were collected for RNA isolation and RT-qPCR (primers in Supplementary Table 3).roX1 ex33A roX2Δ Zhr 1 recombinants were identified by single fly qPCR.

Generation of reconstructed transgenes
All dsRNA-expressing constructs were created using the pWIZ (White Intron Zipper) vector (Lee and Carthew 2003).Reconstructed repeats with the same size and repeat phasing as the original pWIZ-ds1.6883F transgene are designated by a superscript indicating the repeat followed by R (1.688 1AR , 1.688 3CR ).Due to high AT content, the 1.688 1AR repeat was synthesized as two gene blocks based on sequence downloaded from Flybase (FB2020; Öztürk-Çolak et al. 2024), synthesized by Integrated DNA Technologies and joined at a naturally occurring EcoRI site.Blocks were amplified, digested with EcoRI, ligated, digested with XbaI, gel-purified, ligated into pBlueScript, and confirmed by sequencing.Primers used for the generation and validation of transgenes are presented in Supplementary Table 2. Assembly of 1.688 3CR was accomplished by PCR of overlapping fragments templated with an existing 901 bp 1.688 3C amplicon (Menon et al. 2014).Gel-purified amplicons were mixed and subjected to five cycles of denaturation, annealing, and extension.This templated an amplification with flanking primers 3C_Fo and 3C_Ro.A product of the correct size was gel purified, digested with Xho1 and Not1, inserted into pBlueScript and confirmed by sequencing.1.688 1AR and 1.688 3CR inserts were introduced sequentially into pWIZ and orientation was verified by sequencing after each step.Random insertions were mapped to chromosome and three independent insertions of teach transgene selected for further studies.

Rescue of roX1 roX2 males
Males homozygous for each reconstructed transgene were mated to roX1 roX2; p[w + Sqh-Gal4]2/+ virgins to produce offspring with and without the p[w + Sqh-Gal4]2 driver.Offspring were counted for 10 days following initial eclosion and eye color was used to distinguish flies with the p[w + Sqh-Gal4]2 driver.The driver had no effect on female survival.For this reason, the total number of daughters was used to calculate survival of each class of sons (Supplementary Tables 4 and 5).

RNA preparation
Groups of 60 third instar male larvae were homogenized in Trizol (Invitrogen).RNA was cleaned using the QIAGEN RNeasy kit and measured by nanodrop as previously described (Koya and Meller 2015).One μg of RNA was reverse transcribed using random hexamers and Invitrogen SuperScript IV reverse transcriptase.

Quantitative RT-PCR
Amplification of cDNA was performed using Power SYBR Green PCR master mix on a Quantstudio3 qPCR system (ThermoFisher Scientific).Primers are presented in Supplementary Table 3.Values were normalized to dmn (DCTN2-p50).Fold change in expression was calculated relative to the yw lab reference strain using the efficiency corrected comparative quantification method (Pfaffl 2001).The high copy number of pericentric 359 bp repeats required serial dilution of qPCR template.To determine the influence of dsRNA expression on an autosomal gene with integrated recruiting elements, transgene insertions that achieved the strongest rescue of roX1 roX2 males were selected from the three independent insertions tested.

siRNA sequencing and alignment
Total RNA prepared as described above was shipped to Azenta Life Sciences for siRNA isolation and sequencing.Libraries were generated using the NEB Small RNA library Prep Kit (Ipswich, MA, USA) and Illumina 3′ and 5′ adapters.Index sequences were added by amplification and products of 145-160 bp were gel purified and sequenced on an Illumina NovaSeq (2 × 150 bp).Raw sequencing reads underwent quality analyses and adapter trimming using Trimmomatic (v0.30).Trimmed reads of 18 to 30 bp were aligned to the Drosophila genome (dm6) with bowtie2 with and default parameters (Hoskins et al. 2015).Reads were also aligned to the transcribed region of pWIZ-1.6881AR-inv using bowtie 1 (-l 10 -m 100) with zero mismatches.

Details of transgene construction determine activity
The idea that the 1.688 3F repeat, situated immediately distal to roX1, had special properties was intriguing.However, details of construction differed for the original pWIZ-ds1.6881A , pWIZ-ds1.6883C , and pWIZ-ds1.6883F transgenes (Menon et al. 2014).The amplified repeat segments were of different sizes (Fig. 1a and Supplementary Fig. 3).The ends of each amplicon also fell at different places in the 359 bp repeat unit, and thus each amplicon differed in repeat phasing.Many of the 1.688 X repeats contain an EcoRI site that we arbitrarily designated as the boundary of individual repeats in tandem clusters.A two base pair change in 1.688 3C converts EcoRI to SacI (see Supplementary Fig. 3).Finally, the orientation of repeat DNA introduced into pWIZ differed between pWIZ-ds1.6883F and the other two original constructs (Fig. 1a).We first performed PCR and sequencing on flies with the original pWIZ-ds1.6881A and pWIZ-ds1.6883C transgenes and confirmed that these were constructed as intended and had not been rearranged in transgenic flies.Reverse transcription and qPCR (RT-qPCR) of RNA from flies carrying pWIZ-ds1.6881A and the strong, near-ubiquitous p[w + Sqh-GAL4]2 driver confirmed expression, although at a lower level than the reference pWIZ-ds1.6883F insertion (Supplementary Fig. 1).We also validated substantial rescue of males with the severe roX1 SMC17A roX2Δ mutations upon expression of pWIZ-ds1.6883F , but minimal rescue by pWIZ-ds1.6881A (Fig. 2b).
This raises the question of whether the disparity in biological activity is due to differences in DNA sequence, expression levels, or to differences in the size, orientation, or phasing of repeat DNA in the pWIZ transgene.To address this, we constructed transgenes with 1.688 1A and 1.688 3C sequence but with the same size, orientation, and repeat phasing as pWIZ-ds1.6883F (see Materials and methods).This produced pWIZ-ds1.6881AR (89% sequence identity to 1.688 3F ) and pWIZ-ds1.6883CR (68% identity to 1.688 3F ) (see Supplementary Fig. 3 for sequence alignments).This strategy enables direct comparison to the original transgenes with the same repeat sequence in different phasing.To explore the role of orientation, we constructed pWIZ-ds1.6881AR-inv and pWIZ-ds1.6883CR-inv with the same inserts as pWIZ-ds1.6881AR and pWIZ-ds1.6883CR but inverted in orientation (Fig. 1b).Three independent second chromosome insertions of each new transgene were selected for further studies.
To confirm expression, total RNA was extracted from third instar male larvae carrying reconstructed transgenes and the p[w + Sqh-Gal4]2 driver (Supplementary Table 1).Primers to the w + intron, present in the unprocessed transcript, were used to assess relative expression from each insertion (Supplementary Fig. 1).All transgenes were expressed but with variability due to insertion site.Nevertheless, the range of expression levels overlaps that of the original pWIZ-ds1.6881A and pWIZ-ds1.6883F transgenes.
To assess the biological activity of reconstructed transgenes, expression was driven in roX1 SMC17A roX2Δ males (∼2-3% escapers) and roX1 ex33A roX2Δ males (∼25% escapers).Survival of roX1 SMC17A roX2Δ and roX1 ex33 roX2Δ males expressing RNA from pWIZ-ds1.6881AR and pWIZ-ds1.6881AR-inv matched that achieved by the original pWIZ-ds1.6883F , reaching 30 and 65%, respectively (Fig. 2 and Supplementary Fig. 2; see Supplementary Tables 4  and 5 for raw data).Expression from pWIZ-ds1.6883CR and pWIZ-ds1.6883CR-inv also rescued males but to a somewhat lesser extent.A comparison of transgene expression and male survival reveals that the level of male rescue is largely independent of the level of RNA expression.For example, levels of ds1.688 3CR vary more than 3-fold among strains used in this study, but rescue of roX1 ex33 roX2Δ males is indistinguishable (Supplementary Figs. 1  and 2).This suggests that expression exceeds that necessary for the maximum achievable rescue of roX1 roX2 males by the dsRNAs produced.Taken together, these observations indicate that expression levels and differences in DNA sequence cannot account for the lack of male rescue by previously made pWIZ-ds1.6881A and pWIZ-ds1.6883C transgenes.Instead, the phasing of the repeat fragment introduced into pWIZ appears to determine the biological activity of transgenes that express ds1.688X RNA.
We then asked if differences in rescue of roX1 roX2 males reflect processing into small RNA.To assess this, we sequenced small RNA from male larvae expressing dsRNA from the original pWIZ-ds1.6881A , which does not rescue roX1 roX2 males, and pWIZ-ds1.6881AR-inv , which achieves strong rescue.Both transgenes contain sequence from 1.688 1A introduced into pWIZ in the same orientation, but inserts differ in size and phasing.Small RNA from our lab reference strain (yw) was also sequenced.Consistent with previous studies, we detected minimal siRNA from the 1.688 1A region in yw larvae (Menon et al. 2014).In contrast, abundant siRNAs from pWIZ-ds1.6881AR-inv aligned to 1.688 1A as well as 1.688 4A and other closely related 1.688 X repeats on the X chromosome (Fig. 3a).No siRNA aligning to 1.688 1A was detected in pWIZ-ds1.6881A males.However, endogenous small RNAs from other genomic regions, such as those near the hairpin RNA-producing esi-1 locus, were identified in all samples, demonstrating sample integrity (Fig. 3a).We then aligned all reads to the pWIZ-ds1.6881AR-inv transcript, permitting no mismatches.pWIZ-ds1.6881AR-inv produced abundant small RNAs aligning to both arms (Fig. 3b).However, few alignments to the insert are observed in control flies or those carrying pWIZ-ds1.6881A .In contrast, the SV40 terminator produces small RNAs in both transgene-carrying strains.We conclude that pWIZ-ds1.6881A produces transcript but the double-stranded regions created by the 1.688 1A inserts are not effectively processed into siRNA.The reason for this striking disparity remains speculative.

Expression of reconstructed transgenes enhances compensation near an autosomal 1.688 3F insertion
We postulate that the ability of reconstructed transgenes to rescue roX1 roX2 males involves small RNA-directed chromatin modification at hundreds of related sequences on the X chromosome.We previously demonstrated functional compensation of hemizygous autosomal genes near integrations of DNA from the 1.688 1A , 1.688 3C , and 1.688 3F repeats (Joshi and Meller 2017).Less than 2 kb of 1.688 3F DNA was capable of recruitment of MSL proteins detectable on polytene preparations and increased expression of active genes over 100 kb from an autosomal integration site (Deshpande and Meller 2018).The expression of nearby genes was enhanced upon production of ds1.688 3F , revealing that small RNA from 1.688 3F is able to promote recruitment at related genomic sequences (Deshpande and Meller 2018).We took advantage of autosomal integrations of 1.688 3F and roX1 recruiting elements in the hattifattener (haf) gene (cytological position 22A3) to determine if expression of reconstructed dsRNA-expressing transgenes also enhances recruitment (Fig. 4a).
We generated larvae that express reconstructed transgenes and have integrations of 1.688 3F , roX1, or both elements in a haf intron ([1.688 3F ] 22A3 , [roX1] 22A3 , or [roX1&1.6883F ] 22A3 , Supplementary Table 1).Accumulation of haf mRNA in male larvae was determined by qRT-PCR.Expression was normalized to dmn (DCTN2-p50) and expression in the reference yw strain set to one.Integration of 1.688 3F , roX1, or both recruiting elements increased haf expression in males from 2-to 3-fold (gray bars, Fig. 4b).When only a roX1 recruiting element was present, no further increase was achieved upon expression of dsRNA from reconstructed transgenes.This agrees with the idea that recruitment of dosage compensation by the CES (present in roX1) and the 1.688 X repeats occurs by different mechanisms (Joshi and Meller 2017).In contrast, haf expression increased 5-fold over that in the yw control when the 1.688 3F recruiting element was integrated and dsRNAs were expressed.When both 1.688 3F and roX1 were present, haf expression increased 6-to 10-fold over that of the control in response to dsRNA production (Fig. 4b and Supplementary Table 7).In this context, we see essentially no difference in response to transgenes expressing 1.688 1A and 1.688 3C dsRNA, even though 1.688 3C shares only 68% identify with the 1.688 3F recruiting element but 1.688 1A shares 89%.We conclude that siRNAs from multiple 1.688 X repeats are capable of promoting recruitment by related genomic sequences.This suggests that sources of siRNA that participate in X recognition are likely to be highly redundant.The levels of haf activation achieved are considerably higher than the 2-fold increase anticipated for full dosage compensation.This has been previously noted for autosomal roX1 insertions and is attributed to disruption of chromatin-based silencing upon recruitment of the MSL complex (Kelley and Kuroda 2003).

Pericentromeric 359 bp satellites contribute to X recognition
Over 10 Mb of proximal X heterochromatin is composed of closely related 359 bp satellite repeats (Hsieh and Brutlag 1979;Lohe et al. 1993).The pericentromeric repeats are a potential source of siRNAs that participate in dosage compensation.To test this, we recombined the Zygotic hybrid Rescue (Zhr 1 ) mutation, deleted for the pericentromeric 359 bp repeats, with roX1 ex33 roX2Δ (Sawamura and Yamamoto 1993).Recombinants were validated by qPCR of the 359 repeats.roX1 ex33 roX2Δ Zhr 1 and roX1 ex33 roX2Δ Zhr + virgins were mated to males from our yw reference strain and the eclosion of males from each mating was calculated based on female survival.Control roX1 ex33 roX2Δ Zhr + sons eclosed at 27% (Fig. 5a and Supplementary Table 6).However, roX1 ex33 roX2Δ Zhr 1 sons were recovered at 10-15%, suggesting that the pericentric 359 bp repeats also contribute to compensation.
Zhr + produces a maternally transmitted factor, presumably small RNA, that is necessary for assembly of the 359 bp repeats into heterochromatin in the zygote (Yuan and O'Farrell 2016; Ferree and Barbash 2009).This raised the possibility that maternal deposition of small RNA from the 359 bp repeats might also influence dosage compensation.To determine if this effect was maternal or zygotic, we mated roX1 ex33 roX2Δ females heterozygous for Zhr 1 (roX1 ex33 roX2Δ Zhr 1/+ ) to yw males.We reasoned that if the benefit of wild-type Zhr + was zygotic, most eclosing male offspring would carry this allele.However, if Zhr + confers a maternal effect, equal numbers of Zhr 1 and Zhr + sons are expected.To estimate the fraction of sons with each allele, male offspring were pooled for DNA extraction and the abundance of 359 bp repeats determined by qPCR (Fig. 5b).Amplification reveals approximately equal recovery of each class of sons (Fig. 5c).Lack of enrichment for Zhr + in male offspring suggests that the 359 bp repeats act maternally to support the survival of roX1 roX2 sons.

Discussion
Recognition of X-linked genes is essential for dosage compensation, but how this is accomplished is not fully understood in any system.At least two classes of recruiting elements, the CES and 1.688 X satellites, act to attract the compensation machinery the Drosophila X chromosome but do so by distinct mechanisms.For example, CLAMP binds to and is necessary for recruitment by the CES, but loss of CLAMP does not reduce recruitment by 1.688 X repeats (Makki and Meller 2024).Conversely, genes in the siRNA pathway support recruitment by 1.688 X repeats but are dispensable for the CES.We postulate that cooperation between multiple classes of recruiting elements achieves faithful recognition of the X chromosome by the MSL complex.Both CES and Fig. 3. Biologically active and inactive dsRNAs differ in small RNA production.a) Alignment of small RNA from larvae carrying pWIZ-ds1.6881AR-inv (top; insertion 35) pWIZ-ds1.6881A (middle; insertion 2A) or no transgene (control, bottom).Two independent RNA preparations are shown for each genotype.The orientation of 1.688 1A sequence within pWIZ is illustrated to the left.Tandem 1.688 1A repeats proximal to tyn (left) are depicted by arrows.Related 1.688 4A repeats are 4.3 kb distal to CG43689 (middle).All preparations contained abundant small RNA aligning to the autosomal, hairpin RNA-producing esi-1 locus (right).Alignments are depicted on a linear scale.b) Small RNAs were aligned to the transcript produced by pWIZ-ds1.6881AR-inv , schematically represented with 1.688 1A sequence highlighted.The SV40 terminator is upstream of two polyadenylation sites marked with red arrow heads.Alignments to the pWIZ-ds1.6881AR-inv transcript are presented on a logarithmic scale.All alignments are viewed on igv genome browser (Robinson et al. 2011).Expression of all transgenes is driven by p[w + Sqh-Gal4]2.
1.688 X repeats engage in long-range interactions, suggesting that organization of the X chromosome may contribute to selective identification (Ramírez et al. 2015;Sproul et al. 2020).
The involvement of the siRNA pathway in X recognition, abundance of 1.688 X repeats on the X chromosome, and rescue of roX1 roX2 males by expression from the pWIZ-ds1.6883F transgene, suggested the involvement of an siRNA-directed chromatin modification system (Menon and Meller 2012;Menon et al. 2014).Mutation of a single copy of genes involved in production of siRNA and chromatin modification, including Dicer2, Ago2, and Su(var)3-9, enhance the male lethality of roX1 roX2 mutations supporting this idea (Menon and Meller 2012;Deshpande and Meller 2018).In accord with this, histone 3 di-methylated on lysine 9 (H3K9me2) is enriched on 1.688 X repeats, and this mark is enhanced around an autosomal 1.688 3F insertion upon expression of ds1.688 3F RNA (Deshpande and Meller 2018).Recruitment of the MSL complex and upregulation of autosomal genes near insertions of 1.688 1A or 1.688 3C DNA is also enhanced by ds1.688 3F RNA, suggesting that 1.688 X repeats function redundantly to mark the X (Joshi and Meller 2017).Although these observations implicated 1.688 X siRNA in X recognition, the fact that expression from the original pWIZ-ds1.6881A and pWIZ-ds1.6883C transgenes failed to rescue roX1 roX2 males raised questions regarding the origin of the siRNAs (Menon et al. 2014).The proximity of 1.688 3F and roX1 is intriguing and suggests that 1.688 3F might have a unique function.However, we have been unable to find evidence to support this.roX1 roX2 mutants carrying roX1 VM18A , a deletion of    the essential 3′ end of roX1, the entire 1.688 3F repeat and part of echinus, do not exhibit more complete male lethality than other severe roX1 mutants that retain the 1.688 3F repeats (Menon et al. 2014).Transcripts spanning roX1 and 1.688 3F are depicted in the Drosophila genome annotation, but we have been unable to detect evidence of these longer transcripts in embryos or adults and suspect that annotation is based on rare read through transcripts.Nor have we discovered that the proximity of the 1.688 3F repeats and roX1 is important to their function as each is capable of acting independently to recruit compensation to nearby genes (Joshi and Meller 2017).Demonstration of roX1 roX2 male rescue by reconstructed transgenes that express double-stranded 1.688 1A and 1.688 3C RNAs, and evidence that the pericentromeric 359 bp repeats also contribute to dosage compensation, suggests high redundancy in the sources of small RNAs that support X recognition.
The striking difference in biological activity between the original ds1.688 3F , ds1.688 1A , and ds1.688 3C transgenes now appears due to an inability of cells to process dsRNA from the original transgenes.As phasing of the repeats in the dsRNA was identified as most likely responsible for this, it is possible that dsRNAs with some phasings adopt a conformation that impedes Dicer processing.Dicer preferentially processes substrates based on kinetic and thermodynamic characteristics and sequence composition may also influence Dicer cleavage (Vermeulen et al. 2005;Lee et al. 2023).Modeling of dsRNA secondary structures failed to reveal large differences between the most energetically favorable conformations of ds1.688 1A and ds1.688 1AR-inv .Ineffective RNAi has also been linked to degradation of dsRNA by endogenous nucleases (Singh et al. 2017).Understanding why siRNA production from the original pWIZ-1.6881A and pWIZ-1.6883C transgenes fails will require further studies to explore these ideas.We also acknowledge that the phasing in the original pWIZ-1.6883F transgene may not be optimal for small RNA production, leaving open the possibility that even higher activity might be obtained by trial and error.But the current studies do reveal that the 1.688 3F satellite is not uniquely able to produce siRNA that supports X recognition as this property is shared by other 1.688 X repeats and likely also the pericentric 359 bp satellites.
Female hybrids generated by mating D. melanogaster males to D. simulans females, a species lacking the large block of 359 bp repeats on the X chromosome, fail to assemble heterochromatin over the 359 bp repeats and experience mitotic catastrophe when the D. melanogaster X chromosome fails to segregate (Ferree and Barbash 2009;Yuan and O'Farrell 2016).The maternal factor present in D. melanogaster eggs is suspected to be small RNA from the 359 bp repeats.Zhr 1 lacks 359 bp repeats, thus rescuing mitosis in the daughters of D. melanogaster Zhr 1 males mated to D. simulans females (Sawamura and Yamamoto 1993).Our studies suggest that the pericentric 359 bp repeats also play a maternal role in identification of the X chromosome for dosage compensation.Interestingly, a strong modifier of roX1 roX2 male lethality was previously mapped to the proximal X (Stuckenholz et al. 2003).Our findings suggest that this modifier could be an allelic variant of Zhr.The observation that small RNAs aligning to 1.688 X satellites are abundant in early embryos but absent from wild-type larvae raises the question of when these small RNAs are produced (Menon et al. 2014).Many euchromatic 1.688 X repeats are transcribed in male larvae but small RNAs aligning to these repeats are rare in this stage (Menon et al. 2014).Heterochromatic satellites are transcribed in the female germ line and processed into small RNA (Wei et al. 2021).It is possible that the primary source of the small RNAs active in dosage compensation are maternal.This is consistent with a role for satellite repeats and siRNA in identification of X chromatin during the establishment of dosage compensation in the zygote.
The X chromosomes of many Drosophilids are enriched with chromosome-specific repetitive sequences.These are up to 50 times more abundant than autosome-specific repeats in some species (Gallach 2014).Strikingly, the neo-X chromosome in D. pseudoobscura (arm XR) shares this enrichment, suggesting that acquisition of satellite repeats is an early step in the evolution of differentiated sex chromosomes.Repetitive DNA, marked by turnover in sequence and copy number, is implicated in the evolution of sex chromosomes (Smith 1976;Charlesworth et al. 1994;Bayes and Malik 2009;Gallach 2014).Theoretical models of sex chromosome evolution usually focus on accumulation of mutations on the Y chromosome, erosion of coding potential and the subsequent need to increase expression of hemizygous X-linked genes in males (Charlesworth 1996).Both the CES and X chromosome-specific satellite repeats rapidly populate neo-X chromosomes created by sex chromosome and autosomal fusions (Alekseyenko et al. 2013;Ellison and Bachtrog 2013;Gallach 2014).Given the power of CES and satellite repeats to attract compensation to nearby genes, it is possible that the acquisition of compensation precedes the loss of some Y-linked homologs.This could produce over expression that would exert a selective pressure promoting loss of the Y-linked homolog, accelerating sex chromosome differentiation.

Fig. 2 .
Fig. 2. Expression of reconstructed ds1.688X transgenes rescues roX1 SMC17A roX2Δ males.a) roX1 SMC17A roX2Δ females heterozygous for the p[w + Sqh-GAL4] 2 driver were mated to males carrying pWIZ transgenes.The recovery of male offspring was normalized by multiplying males of each category by two and dividing by the total number of daughters.b) The original pWIZ-ds1.6883F and pWIZ-ds1.6881A (hatched bars) and three independent insertions of each reconstructed transgenic (dark bars) were tested.The survival of males with each pWIZ transgene but without driver is shown by light gray bars.Error bars represent the SEM of four biological replicates.*P < 0.05; **P < 0.01, ***P < 0.001, as determined by Student's t test.

Fig. 4 .
Fig. 4. Expression of reconstructed transgenes modulates expression near an autosomal 1.688 3F integration.a) roX1 and 1.688 3F recruiting elements are integrated in a 40 kb intron of the haf gene (arrow head).Primers to measure transcript accumulation amplify a haf exon and lie within a Rab3GP1 intron (verticle bar).b) Accumulation of transcripts in male larvae with integrated recruiting elements roX1, 1.688 3F , or both is shown by gray bars (insertions [1.688 3F ] 22A3 , [roX1] 22A3 , and [roX1+1.6883F ] 22A3 ; Joshi and Meller 2017).Additional expression of ds1.688 1AR or ds1.688 1AR-inv (blue), and ds1.688 3CR or ds1.688 3CR-inv (green) increased haf expression only when 1.688 3F was present in haf.Expression is normalized to dmn and that in lab reference yw males is set to 1 (line).The significance of expression with recruiting elements alone is with respect to yw males as indicated by asterisks.All other statistical comparisons are between males with recruiting elements alone and those with each recruiting element and dsRNA expression.Error bars represent SEM from three biological replicates.*P < 0.05; **P < 0.01, ***P < 0.001, as determined by Student's t test.