Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

RNA nucleation by MSL2 induces selective X chromosome compartmentalization

Nature volume 589, pages 137–142 (2021)Cite this article

Subjects

Abstract

Confinement of the X chromosome to a territory for dosage compensation is a prime example of how subnuclear compartmentalization is used to regulate transcription at the megabase scale. In Drosophila melanogaster, two sex-specific non-coding RNAs (roX1 and roX2) are transcribed from the X chromosome. They associate with the male-specific lethal (MSL) complex1, which acetylates histone H4 lysine 16 and thereby induces an approximately twofold increase in expression of male X-linked genes2,3. Current models suggest that X-over-autosome specificity is achieved by the recognition of cis-regulatory DNA high-affinity sites (HAS) by the MSL2 subunit4,5. However, HAS motifs are also found on autosomes, indicating that additional factors must stabilize the association of the MSL complex with the X chromosome. Here we show that the low-complexity C-terminal domain (CTD) of MSL2 renders its recruitment to the X chromosome sensitive to roX non-coding RNAs. roX non-coding RNAs and the MSL2 CTD form a stably condensed state, and functional analyses in Drosophila and mammalian cells show that their interactions are crucial for dosage compensation in vivo. Replacing the CTD of mammalian MSL2 with that from Drosophila and expressing roX in cis is sufficient to nucleate ectopic dosage compensation in mammalian cells. Thus, the condensing nature of roX–MSL2CTD is the primary determinant for specific compartmentalization of the X chromosome in Drosophila.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Drosophila MSL2-CTD and roX2 RNA confer a condensed state.
Fig. 2: Selective compartmentalization by the MSL2-CTD is required for dosage compensation in vivo.
Fig. 3: roX2 RNA nucleation by a drosophilized MSL2 in mammalian cells.
Fig. 4: roX2-MSL2 nucleation elicits dosage compensation in mammalian cells.

Similar content being viewed by others

Data availability

RNA-seq and ChIP–seq data have been deposited to the Gene Expression Omnibus under the accession number GSE129834. All other relevant data supporting the key findings of this study are available within the article and its Supplementary Information or from the corresponding author upon request.

References

  1. Meller, V. H. et al. Ordered assembly of roX RNAs into MSL complexes on the dosage-compensated X chromosome in Drosophila. Curr. Biol. 10, 136–143 (2000).

    Article  CAS  PubMed  Google Scholar 

  2. Samata, M. & Akhtar, A. Dosage compensation of the X chromosome: a complex epigenetic assignment involving chromatin regulators and long noncoding RNAs. Annu. Rev. Biochem. 87, 323–350 (2018).

    Article  CAS  PubMed  Google Scholar 

  3. Lucchesi, J. C. & Kuroda, M. I. Dosage compensation in Drosophila. Cold Spring Harb. Perspect. Biol. 7, a019398 (2015).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  4. Villa, R., Schauer, T., Smialowski, P., Straub, T. & Becker, P. B. PionX sites mark the X chromosome for dosage compensation. Nature 537, 244–248 (2016).

    Article  ADS  CAS  PubMed  Google Scholar 

  5. Alekseyenko, A. A. et al. A sequence motif within chromatin entry sites directs MSL establishment on the Drosophila X chromosome. Cell 134, 599–609 (2008).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Valsecchi, C. I. K. et al. Facultative dosage compensation of developmental genes on autosomes in Drosophila and mouse embryonic stem cells. Nat. Commun. 9, 3626 (2018).

    Article  ADS  PubMed  PubMed Central  CAS  Google Scholar 

  7. Bai, X., Alekseyenko, A. A. & Kuroda, M. I. Sequence-specific targeting of MSL complex regulates transcription of the roX RNA genes. EMBO J. 23, 2853–2861 (2004).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Ilik, I. A. et al. Tandem stem-loops in roX RNAs act together to mediate X chromosome dosage compensation in Drosophila. Mol. Cell 51, 156–173 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Kelley, R. L. et al. Expression of msl-2 causes assembly of dosage compensation regulators on the X chromosomes and female lethality in Drosophila. Cell 81, 867–877 (1995).

    Article  CAS  PubMed  Google Scholar 

  10. Figueiredo, M. L. et al. Non-coding roX RNAs prevent the binding of the MSL-complex to heterochromatic regions. PLoS Genet. 10, e1004865 (2014).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  11. Hallacli, E. et al. Msl1-mediated dimerization of the dosage compensation complex is essential for male X-chromosome regulation in Drosophila. Mol. Cell 48, 587–600 (2012).

    Article  CAS  PubMed  Google Scholar 

  12. Zheng, S. et al. Structural basis of X chromosome DNA recognition by the MSL2 CXC domain during Drosophila dosage compensation. Genes Dev. 28, 2652–2662 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  13. Fauth, T., Müller-Planitz, F., König, C., Straub, T. & Becker, P. B. The DNA binding CXC domain of MSL2 is required for faithful targeting the Dosage Compensation Complex to the X chromosome. Nucleic Acids Res. 38, 3209–3221 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Larson, A. G. & Narlikar, G. J. The role of phase separation in heterochromatin formation, function, and regulation. Biochemistry 57, 2540–2548 (2018).

    Article  CAS  PubMed  Google Scholar 

  15. Roden, C. & Gladfelter, A. S. RNA contributions to the form and function of biomolecular condensates. Nat. Rev. Mol. Cell Biol. https://doi.org/10.1038/s41580-020-0264-6 (2020).

  16. Kato, M. et al. Cell-free formation of RNA granules: low complexity sequence domains form dynamic fibers within hydrogels. Cell 149, 753–767 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Meller, V. H. & Rattner, B. P. The roX genes encode redundant male-specific lethal transcripts required for targeting of the MSL complex. EMBO J. 21, 1084–1091 (2002).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Kim, M., Faucillion, M.-L. & Larsson, J. RNA-on-X 1 and 2 in Drosophila melanogaster fulfill separate functions in dosage compensation. PLoS Genet. 14, e1007842 (2018).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  19. Apte, M. S. et al. Generation of a useful roX1 allele by targeted gene conversion. G3 (Bethesda) 4, 155–162 (2014).

    Article  CAS  Google Scholar 

  20. Straub, T. et al. Stable chromosomal association of MSL2 defines a dosage-compensated nuclear compartment. Chromosoma 114, 352–364 (2005).

    Article  PubMed  Google Scholar 

  21. Lin, Y. et al. Toxic PR poly-dipeptides encoded by the C9orf72 repeat expansion target LC domain polymers. Cell 167, 789–802.e12 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Cheutin, T. et al. Maintenance of stable heterochromatin domains by dynamic HP1 binding. Science 299, 721–725 (2003).

    Article  ADS  CAS  PubMed  Google Scholar 

  23. Strom, A. R. et al. Phase separation drives heterochromatin domain formation. Nature 547, 241–245 (2017).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  24. Schauer, T. et al. Chromosome topology guides the Drosophila dosage compensation complex for target gene activation. EMBO Rep. 18, 1854–1868 (2017).

    Article  CAS  PubMed Central  Google Scholar 

  25. Toll-Riera, M., Radó-Trilla, N., Martys, F. & Albà, M. M. Role of low-complexity sequences in the formation of novel protein coding sequences. Mol. Biol. Evol. 29, 883–886 (2012).

    Article  CAS  PubMed  Google Scholar 

  26. Bertrand, E. et al. Localization of ASH1 mRNA particles in living yeast. Mol. Cell 2, 437–445 (1998).

    Article  CAS  PubMed  Google Scholar 

  27. Baron-Benhamou, J., Gehring, N. H., Kulozik, A. E. & Hentze, M. W. Using the lambdaN peptide to tether proteins to RNAs. Methods Mol. Biol. 257, 135–154 (2004).

    CAS  PubMed  Google Scholar 

  28. Quinn, J. J. et al. Rapid evolutionary turnover underlies conserved lncRNA-genome interactions. Genes Dev. 30, 191–207 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. McFarland, T. J. et al. Evaluation of a novel short polyadenylation signal as an alternative to the SV40 polyadenylation signal. Plasmid 56, 62–67 (2006).

    Article  CAS  PubMed  Google Scholar 

  30. Flemr, M. & Bühler, M. Single-step generation of conditional knockout mouse embryonic stem cells. Cell Rep. 12, 709–716 (2015).

    Article  CAS  PubMed  Google Scholar 

  31. Raja, S. J. et al. The nonspecific lethal complex is a transcriptional regulator in Drosophila. Mol. Cell 38, 827–841 (2010).

    Article  CAS  PubMed  Google Scholar 

  32. Chelmicki, T. et al. MOF-associated complexes ensure stem cell identity and Xist repression. eLife 3, e02024 (2014).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  33. Groth, A. C., Fish, M., Nusse, R. & Calos, M. P. Construction of transgenic Drosophila by using the site-specific integrase from phage phiC31. Genetics 166, 1775–1782 (2004).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Ramírez, F. et al. High-affinity sites form an interaction network to facilitate spreading of the MSL complex across the X chromosome in Drosophila. Mol. Cell 60, 146–162 (2015).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  35. Venken, K. J. T., He, Y., Hoskins, R. A. & Bellen, H. J. P[acman]: a BAC transgenic platform for targeted insertion of large DNA fragments in D. melanogaster. Science 314, 1747–1751 (2006).

    Article  ADS  CAS  PubMed  Google Scholar 

  36. Bischof, J., Maeda, R. K., Hediger, M., Karch, F. & Basler, K. An optimized transgenesis system for Drosophila using germ-line-specific phiC31 integrases. Proc. Natl Acad. Sci. USA 104, 3312–3317 (2007).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  37. Chao, S. H. & Price, D. H. Flavopiridol inactivates P-TEFb and blocks most RNA polymerase II transcription in vivo. J. Biol. Chem. 276, 31793–31799 (2001).

    Article  CAS  PubMed  Google Scholar 

  38. Kroschwald, S., Maharana, S. & Simon, A. Hexanediol: a chemical probe to investigate the material properties of membrane-less compartments. Matters https://doi.org/10.19185/matters.201702000010 (2017).

  39. Johansen, K. M. et al. Polytene chromosome squash methods for studying transcription and epigenetic chromatin modification in Drosophila using antibodies. Methods 48, 387–397 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Chlamydas, S. et al. Functional interplay between MSL1 and CDK7 controls RNA polymerase II Ser5 phosphorylation. Nat. Struct. Mol. Biol. 23, 580–589 (2016).

    Article  CAS  PubMed  Google Scholar 

  41. Mendjan, S. et al. Nuclear pore components are involved in the transcriptional regulation of dosage compensation in Drosophila. Mol. Cell 21, 811–823 (2006).

    Article  CAS  PubMed  Google Scholar 

  42. Yang, Z. & Nielsen, R. Codon-substitution models for detecting molecular adaptation at individual sites along specific lineages. Mol. Biol. Evol. 19, 908–917 (2002).

    Article  CAS  PubMed  Google Scholar 

  43. Neyman, J. & Pearson, E. S. in Breakthroughs in Statistics (eds Kotz, S. & Johnson, N. L.) 73–108 (Springer, 1992).

  44. Yang, Z. & dos Reis, M. Statistical properties of the branch-site test of positive selection. Mol. Biol. Evol. 28, 1217–1228 (2011).

    Article  CAS  PubMed  Google Scholar 

  45. Arrigoni, L. et al. RELACS nuclei barcoding enables high-throughput ChIP-seq. Commun. Biol. 1, 214 (2018).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  46. Skene, P. J. & Henikoff, S. A simple method for generating high-resolution maps of genome-wide protein binding. eLife 4, e09225 (2015).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  47. Boehning, M. et al. RNA polymerase II clustering through carboxy-terminal domain phase separation. Nat. Struct. Mol. Biol. 25, 833–840 (2018).

    Article  CAS  PubMed  Google Scholar 

  48. Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357–359 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  49. Zhang, Y. et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 9, R137 (2008).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  50. Ramírez, F. et al. deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic Acids Res. 44, W160–W165 (2016).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  51. Quinlan, A. R. & Hall, I. M. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841–842 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  52. Li, H. et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).

    PubMed  PubMed Central  Google Scholar 

  53. Grüning, B. A. et al. The RNA workbench: best practices for RNA and high-throughput sequencing bioinformatics in Galaxy. Nucleic Acids Res. 45 (W1), W560–W566 (2017).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  54. Robinson, J. T. et al. Integrative genomics viewer. Nat. Biotechnol. 29, 24–26 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  55. Bailey, T. L. & Elkan, C. Fitting a mixture model by expectation maximization to discover motifs in biopolymers. Proc. Int. Conf. Intell. Syst. Mol. Biol. 2, 28–36 (1994).

    CAS  PubMed  Google Scholar 

  56. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  57. The Gene Ontology Consortium. The Gene Ontology Resource: 20 years and still GOing strong. Nucleic Acids Res. 47, D330–D338 (2019).

    Article  CAS  Google Scholar 

  58. Piovesan, D. et al. MobiDB 3.0: more annotations for intrinsic disorder, conformational diversity and interactions in proteins. Nucleic Acids Res. 46, D471–D476 (2018).

    Article  CAS  PubMed  Google Scholar 

  59. Mier, P. & Andrade-Navarro, M. A. dAPE: a web server to detect homorepeats and follow their evolution. Bioinformatics 33, 1221–1223 (2017).

    CAS  PubMed  Google Scholar 

  60. Hulsen, T., de Vlieg, J. & Alkema, W. BioVenn—a web application for the comparison and visualization of biological lists using area-proportional Venn diagrams. BMC Genomics 9, 488 (2008).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  61. Koulouras, G. et al. EasyFRAP-web: a web-based tool for the analysis of fluorescence recovery after photobleaching data. Nucleic Acids Res. 46 (W1), W467–W472 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  62. Joshi, S. S. & Meller, V. H. Satellite repeats identify X chromatin for dosage compensation in Drosophila melanogaster males. Curr. Biol. 27, 1393–1402.e2 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  63. Patel, S. S., Belmont, B. J., Sante, J. M. & Rexach, M. F. Natively unfolded nucleoporins gate protein diffusion across the nuclear pore complex. Cell 129, 83–96 (2007).

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

We thank the MPI-IE facilities for support: in particular, U. Bönisch for deep sequencing; D. Ryan and G. Renschler for bioinformatics; R. Pohlmeyer and the imaging facility during COVID19 curfew; J. Herman for help in creating the Msl2 knock-in mES cells; W. Szymanski for mass spectrometry; C. Pessoa, U. Gündisch, A. Alexiadis and T. Stehle for technical support; U. Erdogdu and T. Rumpf for sharing reagents and visualizing the CXC domain structure, respectively; S. Hiller and T. Sharpe for advice on biophysical data and conducting SEC–MALS, respectively; G. Pyrowolakis, M. Carmo-Fonseca, V. Meller, A. Brand and M. Bühler for sharing antibodies, plasmids, fly stocks and the parental mES cells, respectively; M. Boehning and P. Cramer for the control protein for the in vitro FRAP experiment; and R. Sawarkar, N. Iovino, G. Renschler, M. Samata, M. Shvedunova and K. Woolcock for input that helped to improve the manuscript. The mouse symbol was modified from an icon made by Freepik from www.flaticon.com. C.I.K.V. was supported by a Human Frontier Science Program long-term fellowship (000233/2014-L). This study was supported by the German Research Foundation (DFG) under Germany’s Excellence Strategy (CIBSS – EXC-2189 – Project ID 390939984). This work was supported by the DFG under the CRC 992 (project A02) and CRC 1381 (project B3) awarded to A.A.

Author information

Author notes

  1. These authors contributed equally: Claudia Isabelle Keller Valsecchi, M. Felicia Basilicata

Authors and Affiliations

  1. Max Planck Institute of Immunobiology and Epigenetics, Freiburg im Breisgau, Germany

    Claudia Isabelle Keller Valsecchi, M. Felicia Basilicata, Plamen Georgiev, Aline Gaub, Janine Seyfferth, Tanvi Kulkarni, Amol Panhale, Giuseppe Semplicio, Vinitha Manjunath, Herbert Holz & Asifa Akhtar

  2. Faculty of Biology, University of Freiburg, Freiburg im Breisgau, Germany

    Tanvi Kulkarni

  3. Institute for Evolutionary Biology and Environmental Studies, University of Zurich, Zurich, Switzerland

    Pouria Dasmeh

Authors
  1. Claudia Isabelle Keller Valsecchi
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. M. Felicia Basilicata
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Plamen Georgiev
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Aline Gaub
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Janine Seyfferth
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Tanvi Kulkarni
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Amol Panhale
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Giuseppe Semplicio
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Vinitha Manjunath
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Herbert Holz
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Pouria Dasmeh
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Asifa Akhtar
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

C.I.K.V. conceptualized the study with M.F.B., P.G., P.D. and A.A. P.G. performed all Drosophila genetic experiments with the help of V.M. A.G. performed ChIP experiments and prepared polytene squashes. J.S. and H.H. assisted with the generation and analysis of mES cells. G.S. processed NGS data. T.K. and A.P. performed immunostainings and microscopy of the S2 cell hexanediol experiment. P.D. performed computational analyses on molecular evolution and fitted rate constants. C.I.K.V. and M.F.B. performed all other experiments and data analyses, and wrote the manuscript with input from all authors. C.I.K.V. and M.F.B. designed, coordinated and supervised the experiments. A.A. provided guidance and secured funding.

Corresponding author

Correspondence to Asifa Akhtar.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature thanks Justin Crocker, Richard Kriwacki and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 Genome-wide analyses of dosage compensation upon loss of msl and roX genes.

a, H4K16ac ChIP–seq profiles were generated from wild-type male, wild-type female or homozygous mutant male L3 larvae. The roX1SMC17A roX2∆ (abbreviated as roX1 roX2), mle9 (abbreviated as mle) and msl-1L60 (abbreviated as msl-1) alleles were compared with controls. The chromatin was fragmented with MNase. Boxplots display the normalized coverage in the input (2-kb bins, n = 11,347 (chr2L), n = 12,083 (chr2R), n = 13,459 (chr3L), n = 15,624 (chr3R), n = 11,377 (chrX); replicates were merged for plotting). The relative coverage of the X chromosome shows that males and females were accurately separated. b, H4K16ac ChIP–seq as in a. The plots show the mean input-normalized log2[fold-change] enrichment at scaled X-linked gene bodies (first 0.6 kb unscaled, left), HAS (middle), and random regions (shuffled HAS positions, right). A Welch Two Sample t-test (two-tailed) for the enrichment of H4K16ac on X-linked gene bodies (n = 2,647) in mutant males compared with females revealed P-values of 3.86 × 10−87 (roX1 roX2), 2.57 × 10−54 (mle) and 0.025 (msl-1). A subset of the same data are also shown in Fig. 1b. Also see Supplementary Data 4 for details on statistics. c, H4K16ac ChIP–seq as in a. Genome-browser snapshot showing the 1x sequencing depth normalized ChIP coverage on an X-linked region containing a HAS, whose position is highlighted with a diamond. d, H4K16ac ChIP–qPCR analyses of genotypes abbreviated as in a. The barplot shows the mean of n = 3 independent biological replicates/experiments ± s.e.m. on the indicated genes. Enrichment values were corrected over the input and expressed relative to the Klg gene (not shown for clarity of the panel). e, Cropped immunoblots of protein levels in nuclear extracts of male and female L3 larvae. Genotypes are wild-type (lanes 1 and 2), msl-2::HA (lanes 3 and 4) and roX1SMC17A roX2∆/Y ; msl-2::HA (lane 5). For source data, see Supplementary Fig. 1. f, Schematic representation of the experimental setup for RNA-seq from dissected wing imaginal discs. The msl-2227 (abbreviated as msl-2), roX1SMC17A roX2∆ (abbreviated as roX1 roX2), mle9 (abbreviated as mle) and msl-1L60 (abbreviated as msl-1) homozygous null mutant males were compared with control males. Data processing is described in Methods. g, RNA-seq as in f, MA-Plots showing normalized counts versus the log2[fold change] obtained with DESeq2. Differentially expressed (DE) genes (FDR < 0.05) are colored in red (upregulation) and blue (downregulation) and the number of DE genes is reported in each panel. h, RNA-seq as in f, Boxplots show the log2[fold-change] of all genes (n = 265) directly overlapping with HAS. The P-values were obtained by a two-sided Wilcoxon rank-sum test. i, RNA-seq as in f, Boxplots show the linear distance in kb to the closest HAS of all DE downregulated genes on the X chromosome scored in each mutant (msl-2: n = 465, roX1 roX2: n = 430, mle: n = 632, msl-1: n = 777 genes). This was compared with the distance to HAS obtained from the same number n of random X-linked genes (grey boxplots). The P-values were obtained by a two-sided Wilcoxon rank sum test. j, RNA-seq as in f, Barplot for the number of DE genes with respect to their location on the chromosomal arms. The P-values for overrepresentation of X-linked genes obtained from a one-sided Fisher’s test are reported. k, RNA-seq as in f, Density plots of the log2[fold change] obtained with DESeq2 for genes on each of the indicated chromosomal arms. All genes were taken into account for the analyses, irrespective of whether they are scored as DE or not. l, RNA-seq as in f, Venn diagram showing the overlaps of the DE genes on the X chromosome and autosomes, where the area is proportional to the number of genes in each group. For clarity of the panel, mle is not shown. m, RNA-seq as in f, Barplot showing the DE gene overlap. The plot illustrates whether DE genes were found exclusively in one particular mutant data set, or also in others. The results for the overlaps of DE genes are reported for up- versus downregulated and X chromosome and autosomes separately. n, MLE ChIP–qPCR analyses from separated male or female msl-2227/CyO-GFP or msl-2227/msl-2227 L3 larvae. The barplot shows the mean ± s.e.m. of n = 3 independent (n = 2 for male homozygous larvae due to limited sample availability) biological replicates/experiments on the indicated HAS or control regions (odsh, Klg). Enrichment values were corrected over the input and expressed relative to the mean of the non-target control regions (odsh, Klg). o, Cropped immunoblots of same genotypes as in e. MSL2 protein is not expressed in females9. For source data, see Supplementary Fig. 1. p, Immunofluorescence of polytene chromosome squashes from endogenously tagged msl-2::HA male (panel 1) and female (panel 2), as well as roX1SMC17A, roX2∆; msl-2::HA male salivary glands. MSL2-HA is shown in red, MOF in green and DAPI in blue. Panels 4 and 5 show the more frequent case of lost X chromosome enrichment for MSL2 with gain at the chromocentres accompanied by a homogenous, female-like staining of MOF. Panel 3 illustrates a rare example, where some X chromosome staining can be scored in roX1 roX2 double nulls, however, the staining pattern is clearly distinct from wild-type. The latter staining is similar to what has been reported in10,17 and we interpret it to be a consequence of an alternative mechanism that can drive MSL complex recruitment to the X62. Scale bars = 20 μm. q, MSL2-HA ChIP–qPCR analyses from wild-type msl-2::HA male, msl-2::HA female or homozygous mutant male L3 larvae abbreviated as in f. The barplot shows the mean of n = 3 independent biological replicates/experiments ± s.e.m. on the indicated HAS or control regions (odsh, Ent2). Enrichment values were corrected over the input and expressed relative to the mean of the non-target control regions (odsh, Ent2).

Extended Data Fig. 2 Computational and in vitro characterization of MSL2 CXC and CTD domains.

a, MSL2 ChIP–seq analyses (data published by10). Boxplots display the change in input-normalized ChIP enrichment difference in roX1ex6 Df(1)roX252 (abbreviated as roX1 roX2) versus wild-type on MSL2-enriched 1 kb bins, HAS (n = 267/26 on X/autosomes) or random regions. The random regions were generated by shuffling the enriched 1 kb bins. A negative value indicates a loss of enrichment in the mutant, while a gain is indicated by a positive value. The enrichment score was calculated using deepTools multibigwig summary. 1 kb bins with a score >0.7 in either genotype were classified as enriched (n = 1,893 enriched/144,916 total bins, of which n = 1,448 reside on X and n = 445 on autosomes). A Welch Two Sample t-test for the difference in enrichment of MSL2 at HAS in wild-type versus roX1 roX2 revealed a P-value of 4.387 × 10−73. b, as in a, Genome-browser snapshot of MSL2 ChIP–seq. Data normalization is described in Methods. c, Analysis of adult male viability in flies expressing transgenic roX2 (tg1, tg2, tg3) in a roX1SMC17A roX2∆ background. The data was expressed relative to females obtained from the same cross. The barplot represents the mean ± s.e.m. with overlaid datapoints reflecting the results from n = 6 vials/crosses per genotype. Details on genotypes and nature of transgenes are provided in Methods and Supplementary Data 4. d, RT–qPCR analysis of the indicated genes in roX1SMC17A roX2+ (bar 1) or roX1SMC17A roX2∆ ; roX2tg1/tg2/tg3 (bars 2–4, see Methods) or wild-type female (bar 5) L3 larvae. All genotypes additionally contained the msl-2::HA allele. The RNA level of each gene was normalized to RpL32 and expressed relative to roX1SMC17A roX2+ male flies. The barplot represents the mean ± s.e.m. of n = 5 larvae with overlaid datapoints reflecting the result from one individual larva. e, Representative coomassie-stained SDS–PAGE of the purified GST-CXC domains of D. melanogaster (Dmel), D. virilis (Dvir) and H. sapiens (Human). GST was used as negative control (−ctrl). For source data, see Supplementary Fig. 1. f, Biolayer interferometry (BLItz) experiments were conducted to quantify binding of the recombinant GST-tagged CXC proteins of different species as in e to the biotinylated ATGAGCGAGATG dsDNA (referred to as S12 in12). The barplots report the mean fitted K(on1) and K(off1) ± s.e.m. rate constants (left), the boxplots display the equilibrium KDs (right). Parameters were determined from two independent protein purifications. In each independent experiment (Dmel CXC n = 12, Dvir CXC n = 7, Human CXC n = 6), the proteins were measured over a concentration range of at least 5 different dilutions. Only measurements at concentrations above 1 μM were taken into account for fitting using a 2:1 binding model (also see Methods). The P-values were obtained by a two-sided Wilcoxon rank sum test. g, as in f, Representative BLItz experiments with observed signals shown in grey and fitted in red-blue colour. h, Sequence alignment of the CTD of MSL2 in different mammalian and Drosophila species abbreviated as Dmel = D. melanogaster, Dsim = D. simulans, Dsec = D. sechellia, Dyak = D. yakuba, Dere = D. erecta, Dana = D. ananassae, Dpse = D. pseudoobscura, Dwil = D. willistoni, Dmoj = D. mojavenis, Dvir = D. virilis, Dgri = D. grimshawi. The numbering at the top refers to the relative amino-acid position, starting from the CTD. i, Evaluation of the poly-proline bias in the MSL2 CTD of Drosophila species abbreviated as in h compared to mammalian species. Prolines are indicated with a grey bar. The amino acid sequence of the D. melanogaster CTD is shown below, with significantly enriched low complexity regions by dAPE software shown in colour. j, Top, illustration of the published crystal structure of the MSL2 CXC domain in complex with DNA (PDB: 4RKH). Residues with signs of positive selection are colored in red, conserved residues in blue. Bottom, sequence alignment of the MSL2 CXC domains of different species abbreviated as in e. Residues with signs of positive selection (also see Supplementary Data 2) are colored in red, conserved residues in blue. k, Representative coomassie-stained SDS–PAGE of the purified Drosophila dCXC-CTD, as well as the Drosophila dCTD proteins. Purified His-Smt3 was used as negative control in several in vitro assays. For source data, see Supplementary Fig. 1. l, Biolayer interferometry (BLItz) experiments were conducted to quantify binding of Drosophila dCXC-CTD or dCTD proteins to the biotinylated dsDNA CGAATATGAGCGAGATGGATG (CES11D1). The BLItz experiments for the dCXC-CTD (20 and 10 μM) are shown in black/grey, while the dCTD (40 and 20 μM) is shown in blue. The two lines at each concentration indicate two independent measurements. m, Absorbance at 280 nm recorded during anion exchange chromatography conducted after the His-Talon purification of the dCXC-CTD construct. Two peaks were monitored, of which the second one corresponds to co-purifying nucleic acids. n, Absorbance at 280 nm recorded during gel filtration of the Drosophila dCXC-CTD and mammalian mCXC-CTD proteins. Final pooled fractions used for in vitro assays are indicated. o, Representative example BLItz experiments of the recombinant D. melanogaster GST-dCXC and untagged dCXC-CTD protein binding to the CGAATATGAGCGAGATGGATG dsDNA (referred to as CES11D1 in12). The experimental signal is colored in grey and the fitted data overlaid in colour. The concentrations from top to bottom in the panels of the GST-dCXC proteins were (40, 16, 4, 2, 0.66, 0.33 μM) and dCXC-CTD (100, 75, 50, 37.5, 25, 10, 4, 1.5, 0.75 μM). Fitted rate constants from n = 5 (dCXC-CTD) or n = 4 (GST-dCXC) independent protein purifications are reported in Fig. 1i. Significance was evaluated with a two-sided Wilcoxon rank sum test, where the difference in equilibrium constant (KD1) revealed a P-value of 1.26 × 10−7, n = 29 (dCXC-CTD) versus n = 21 (GST-dCXC) measurements.

Extended Data Fig. 3 In vitro characterization of dCXC-CTD oligomerization and dCXC-CTD-roX2 droplets.

a, Size exclusion multiangle light scattering (SEC-MALS) experiment performed for the recombinant dCXC-CTD protein. Left, schematic representation of the experimental approach to determine the molar mass of a particle by SEC-MALS. Middle, normalized excess Rayleigh ratio (light scattering, red), normalized differential refractive index (blue) and molar mass (green dots) from SEC-MALS for the dCXC-CTD at three different concentrations with the elution volume displayed on the x axis. The 27.8 kDa dCXC-CTD exhibited a similar elution volume to the 66 kDa BSA monomer, suggesting that it has an unusually large hydrodynamic radius, consistent with parts of it being intrinsically disordered. Right, legend and table with weight-averaged molar mass determined from the SEC-MALS experiment. The loading concentrations of the samples were: 1) high concentration (50 μM, left) and 2) medium (15 μM, right) 3) low concentration (5 μM, right). The observed particle masses at high concentrations are higher than the calculated molar mass of the monomeric dCXC-CTD protein. b, Fluorescence confocal microscopy of 10 μM roX2 RNA spiked at 1:100 molar ratio with fluorescein-labelled roX2, which was mixed with recombinant dCXC-CTD protein at the indicated concentrations and conditions. The buffer was 25 mM MES, pH 6.5, 75 mM NaCl, 0.5 mM TCEP, 2.5% glycerol, 2.5% PEG-8000. In the lower panel, the buffer was 25 mM MES, pH 6.5, 75 mM NaCl, 0.5 mM TCEP, 10% PEG-8000. Scale bars = 20 μm. c, as in b, upon addition of 5% or 10% 1,6-hexanediol, which did not dissolve the roX2-dCXC-CTD particles over the observed time period of around 1 h. Scale bars = 20 μm. d, Average fluorescence intensity over time of bleached area of 10 μM roX2 RNA spiked at 1:100 molar ratio with fluorescein-labelled roX2 mixed with 10 μM recombinant dCXC-CTD protein. The buffer was 25 mM MES, pH 6.5, 75 mM NaCl, 0.5 mM TCEP, 2.5% Glycerol, 2.5% PEG-8000. Entire droplets were bleached and the recovery was recorded over two different time courses (long, 600 s and short, 85 s). For each time course, the black line represents the average of n = 7 FRAP experiments that were performed with 2 different protein purifications with the standard deviation shaded in grey. Data fitting and immobile fraction analysis was obtained with EasyFRAP (see Methods). Representative FRAP images (scale bars = 2 μm) of in vitro roX2-MSL2 (dCXC-CTD) droplets are shown below the data curve panels. e, Fast recovering FRAP control, MBP-Pol2(CTD), as described by47. 75 μM unlabelled protein was spiked at 1:100 molar ratio with Alexa488-labelled protein in 20 mM HEPES, pH 7.4, 220 mM NaCl, 0.2 mM TCEP, 2% Glycerol, 16% Dextran. The black line represents the average of n = 10 FRAP experiments with the standard deviation shaded in grey. The fitted t1/2 recovery time was 5.5 ± 0.8 s.

Extended Data Fig. 4 In vivo dynamics of the MSL complex territory.

a, Schematic representation of the experimental setup to analyse in vivo dynamics of the MSL complex. Drosophila S2 cells or L3 larval imaginal discs were exposed to 5% of 1,6-hexanediol at the indicated time points. For the recovery experiment, 1,6-hexanediol is removed after 5 min and replaced with fresh growth media. The recovery times were 3.5 h in S2 cells and only 10 min in imaginal discs to avoid prolonged incubation of the inverted larval tissues ex vivo (see Methods). The treatments were stopped by fixation. b, The X chromosomal territory enrichment ratio by immunofluorescence is calculated by dividing the average intensity of the territory in one nucleus by the average intensity of an equally sized region elsewhere in the same nucleoplasm. Immobile fractions by 1,6-hexanediol were calculated as percent territory enrichment remaining after treatment. c, Quantification of hexanediol experiment in S2 cells as outlined in a with a recovery phase of 3.5 h see f for representative pictures. The dot plots represent mean ± s.e.m. percentage of nuclei with a rim localization (NUP153, left) or X chromosome territory (MSL1, right) of the total number of nuclei. Each dot represents the n fields of view that were evaluated (n = 4 for 0 min, 2 min and 5 min; n = 3 for recovery), the total number of analysed nuclei is reported for each experimental condition. The P-values were obtained by a Kruskal–Wallis test followed by a Dunn’s statistical test for multiple comparisons. d, Quantification of hexanediol experiment in S2 cells as outlined in a with a recovery phase of 3.5 h see f for representative pictures. The territory enrichment ratio by intensity was determined as delineated in b. The bar plots show the mean ± s.e.m. with overlaid data points reflecting individual nuclei, n = 117 (0 min), n = 116 (2 min), n = 99 (5 min), n = 103 (recovery) for MSL1 and n = 117 (0 min), n = 108 (2 min), n = 99 (5 min), n = 103 (recovery) for H4K16ac. The P-values were obtained by a Kruskal–Wallis test followed by a Dunn’s statistical test for multiple comparisons. e, Quantification of hexanediol experiment in Drosophila L3 larval imaginal discs as delineated in a with recovery phase of 10 min, see g for representative pictures. The territory enrichment ratio by intensity was determined as delineated in b, evaluated areas in territory and nucleoplasm were 0.384 μm2 as of the smaller nuclear size compared to S2 cells. The bar plots show the mean ± s.e.m. with overlaid data points reflecting individual nuclei. n = 120 (0 min), n = 115 (5 min), n = 123 (10 min) n = 109 (recovery) nuclei were counted from three (n = 3) imaginal discs/biological replicates for each condition. The P-values were obtained by a Kruskal–Wallis test followed by a Dunn’s statistical test for multiple comparisons. f, Hexanediol experiment as outlined in a. Immunofluorescence was performed in S2 cells and H4K16ac (red), MSL1 (green), NUP153 (grey) were detected with antibodies, DAPI in blue. NUP153 serves as a control, as FG-nucleoporins are sensitive to 1,6-hexanediol63. The representative pictures are orthogonal projections of an entire z-stack with scale bars = 10 μm. The recovery experiment highlights that cell viability was not compromised by the treatment and cellular function was sustained during the recovery phase. g, Hexanediol experiment as outlined in a. Immunofluorescence was performed in imaginal discs from msl-2::HA flies (endogenously tagged). HA (MSL2, red), MSL1 (green), NUP153 (grey) were detected with antibodies, DAPI in blue. The representative pictures are single z-planes with scale bars = 10 μm. h, RT–qPCR analysis of the indicated genes from dissected wing imaginal discs with msl-2227 null alleles in homozygous or heterozygous (CyO-GFP) condition. The RNA level of each gene was normalized to RpL32 and expressed relative to heterozygous males. The barplot represents the mean ± s.e.m. of n = 5 replicates with overlaid data points representing individual wing disc samples. i, RT–qPCR analysis of the indicated genes from dissected wing imaginal discs in females ectopically expressing UAS-msl-2 or UAS-ctrl with hh-Gal4. The RNA level of each gene was normalized to RpL32 and expressed relative to UAS-ctrl. The barplot represents the mean ± s.e.m. of n = 5 replicates with overlaid data points representing individual wing disc samples. j, RT–qPCR analysis of the indicated genes from male L3 larvae. Heterozygous and homozygous mutants of the msl-1L60 (abbreviated as msl1-1), msl-2227 (abbreviated as msl-2), msl-3083 (abbreviated as msl-3) alleles were analysed. The RNA level of each gene was normalized to RpL32 and expressed relative to wild-type male flies. The barplot represents the mean ± s.e.m. of n = 5 larvae. k, as in j, but for the msl-2227 (abbreviated as msl-2), mle9 (abbreviated as mle) and roX1SMC17A, roX2Δ (abbreviated as roX1, roX2) and recombinants of these alleles.

Extended Data Fig. 5 Characterization of MSL2 hybrids in Drosophila.

a, Cropped immunoblots of total protein extracts of adult male fly heads expressing msl-2 hybrid transgenes (UAS-tg). The genotype was msl-2227/msl-2kmA ; tub-Gal4/UAS-tg. The MSL2* transgene refers to point mutations at I529T, N534P and T535G, which gave largely identical results to the D. virilis MSL2vCXC hybrid construct (also see Supplementary Data 3). Because the CTD and RING hybrids displayed some defects in protein stability, we generated additional transgenic flies with two integrations for the CTD and CXC-CTD hybrids (indicated as 2× and 1×, also see Methods). This served to exclude the possibility that a particular assay is confounded by lower protein levels. For source data, see Supplementary Fig. 1. b, as in a, but for the heterozygous msl-2227/kmA/CyO-GFP; tub-Gal4/UAS-tg-expressing male flies. The no transgene refers to msl-2227/kmA/CyO-GFP ; tub-Gal4/+. c, as in a, but for the msl-2227/msl-2kmA ; tub-Gal4/UAS-tg-expressing female flies. d, as in a, but for the msl-2227/msl-2kmA ; tub-Gal4/UAS-tg-expressing male flies and with extracts prepared from L3 larvae. The lines with two integrations of the MSL2mCXC-CTD construct were created by recombining integrations at the VK33 and 68FB landing sites. e, Co-immunoprecipitations of Flag-tagged full-length MSL2WT, MSL2 hybrids and an untagged control with cropped immunoblots for Flag and MSL1. Flag expression vectors driven by an inducible MtnA promoter were transiently transfected into S2 cells. Following selection with blasticidin, the protein was induced overnight by addition of copper sulfate to the culture media. Flag immunoprecipitations were conducted from nuclear extracts. For source data, see Supplementary Fig. 1. f, ChIP–qPCR analyses of His-Bio-His-tagged full-length MSL2 (MSL2WT), MSL2 domain hybrids and an untagged control. Expression vectors driven by an inducible MtnA promoter were transfected into S2 cells. Following selection with Blasticidin, the protein was induced overnight by addition of copper sulfate to the culture media before formaldehyde fixation and ChIP (Streptavidin IP). The barplot shows the mean of n=3 independent biological replicates/experiments ± s.e.m. on the indicated HAS with overlaid data points. Enrichment values were corrected over the input and expressed relative to the control region (odsh). g, H4K16ac ChIP–qPCR analyses of male L3 larvae expressing msl-2 hybrid transgenes in msl-2227/msl-2kmA backgrounds. The barplot shows the mean of n = 3 independent biological replicates/experiments ± s.e.m. on the indicated genes. Enrichment values were corrected over the input and expressed relative to Klg, which does not display any H4K16ac enrichment in ChIP–seq. CG42732, U6, U12, Klg and Intergenic represent negative controls. h, Immunofluorescence of polytene chromosome squashes from male L3 larvae expressing Flag-tagged msl-2 hybrid transgenes (MSL2WT or MSL2mCXC) in msl-2227/msl-2kmA background. MSL1 (green), Flag (red) and DAPI (blue) are shown, scale bars = 20 μm (left panel) and 10 μm (right panel). i, RT–qPCR analysis of the indicated genes from male L3 larvae of the following genotypes: msl-2227/kmA/CyO-GFP; tub-Gal4/+ (bar 1, abbreviated as het), msl-2227/msl-2kmA; tub-Gal4/+ (bar 2), msl-2227/msl-2kmA; tub-Gal4/UAS-tg (MSL2WT, bar 3, blue) or msl-2227/msl-2kmA; tub-Gal4/UAS-tg (MSL2mCXC, bar 4, orange). The RNA level of each gene was normalized to the geometric mean of RpL32, hh and ssp4 and expressed relative to msl-2227/kmA/CyO-GFP; tub-Gal4/+ flies. The barplot represents the mean ± s.e.m. of at least n = 4 larvae with overlaid datapoints representing individual larva. j, MEME motif analyses of Drosophila MSL2 ChIP peaks overlapping with HAS6 or mouse MSL2 ChIP peaks32 in mouse embryonic stem cells (mES cells) at transcription start sites (TSS ± 1 kb), within gene bodies (intragenic, 1 kb to transcription end site) or outside annotated genes (distal). For mammalian MSL2, only MACS2 peaks associated with differentially expressed genes in Msl2∆ mES cells6 were analysed. The top scoring motif with respective E-value is reported in the panel. k, b-isox precipitation experiments were conducted from wild-type males (msl-2::HA, endogenously tagged) and msl-2227/msl-2kmA ; tub-Gal4/UAS-tg (Flag-tagged MSL2WT or MSL2mCTD hybrid)-expressing female transgenic L3 larvae. Left, schematic representation of the experimental setup as in Fig. 2g. Right, cropped immunoblots of b-isox precipitation fractions. The HA blot was cropped to remove marker positions, the HA and Flag blots were obtained from 2 different gels loaded with the same extracts. For source data, see Supplementary Fig. 1.

Extended Data Fig. 6 ChIP–seq, expression and immunofluorescence analyses of MSL2 hybrids in Drosophila.

a, Cropped immunoblots of a fraction of input and immunoprecipitate of the Flag ChIP–seq samples generated from male L3 larvae expressing msl-2 wild-type or hybrid transgenes with tub-Gal4 in the msl-2227/msl-2kmA background. The MSL2mCTD (2x) was used, all other transgenes are present in a single copy and have a C-terminal 3xFlag tag. RPB3 and histone H3 serve as loading and negative IP control. Blots were cropped at the vertical dashed line to remove extra lanes. Displayed bands were derived from the same gels. No bands were detectable in the ChIP samples for H3. For source data, see Supplementary Fig. 1. b, MSL2–Flag ChIP–seq as in a. Genome-browser snapshot on a selected X-chromosomal region showing the Flag ChIP–seq data (Input subtracted) obtained from MSL2 hybrids together with roX2 ChIRP28, H3K36me3 and H4K16ac from male L3 larvae6. c, MSL2–Flag ChIP–seq analyses from transgenic L3 larvae as in a. Heatmaps show the normalized ChIP enrichment (Input subtracted) on HAS (top) and same number of randomly shuffled regions (bottom). The first four represent male samples, the last females. Biological replicates were merged for visualization. For the MSL2mCXC-CTD male only a single replicate of sufficient quality could be generated. d, MSL2–Flag ChIP–seq as in a. Heatmaps show the log2[fold-change] of the normalized MSL2mCTD or MSL2mRING ChIP enrichment versus the normalized wild-type MSL2WT ChIP enrichment on all MACS2 peaks collectively called in each replicate and ChIP experiment (wild-type and all hybrids combined). Two unsupervised K-means clusters were generated, which separates the regions into gain of enrichment (top, red) and loss of enrichment (bottom, blue) in the hybrids. The signal is sorted by enrichment intensity within each cluster. Right, frequency of peaks per chromosome in the gain of enrichment group in comparison with the size of the Drosophila chromosomal arms. The gained peaks are almost exclusively found on autosomes. e, MSL2–Flag ChIP–seq as in a. MEME-ChIP was used to analyse motifs in peaks within the gain of enrichment group determined in d. No significant motifs were found by MEME, the top scoring motif from DREME with E-value is shown. f, as in e, showing the DREME motif and E-value that displayed the highest central distribution by Centrimo (E = 5.3 × 10−1). g, MSL2–Flag ChIP–seq as in a. Schematic representation of the motif search approach. FIMO was used to find the mMSL2 TSS, distal and intragenic motifs (Extended Data Fig. 5j) in the Drosophila genome. The identified motif positions were intersected with the collective peaks list and analysed for overrepresentation in the gain or loss-group by a one-sided Fisher’s exact test. The peak frequencies in each group are displayed on the right and revealed no statistically significant P-values. h, Immunofluorescence of Armadillo (green) and phosphorylated Histone H3 Ser28 (purple) in male wing imaginal discs from wild-type or msl-2227/msl-2kmA L3 larvae expressing the MSL2WT, MSL2mCTD (2x) or MSL2mCXC-CTD (2x) transgenes with tub-Gal4. DAPI in blue. The pictures are single z-planes with scale bars = 20 μm. i, Immunofluorescence of Armadillo (green) in male salivary glands from L3 larvae as in h. The pictures are single z-planes with scale bars = 100 μm, DAPI is shown in blue. j, RT–qPCR analysis of RNA levels of the indicated genes in homozygous msl-2227/msl-2kmA L3 larvae expressing MSL2 hybrids as in h. The RNA level of each gene was normalized to RpL32 and expressed relative to msl-2227/km/CyO-GFP heterozygous flies expressing MSL2WT. For clarity of the panel, the results of the heterozygous flies are not shown. The barplot represents the mean ± s.e.m. with overlaid data points reflecting one individual larva (n analysed larvae reported for each bar). The selected genes are involved in cellular functions such as cell division (Yippee), cell adhesion and signalling (arm) or cell migration (pico). k, RT–qPCR analysis of roX2 in heterozygous msl-2227/km/CyO-GFP or homozygous msl-2227/msl-2kmA male L3 larvae expressing the MSL2 hybrid transgenes (tub-Gal4/UAS-tg). The “no transgene” (no tg) control refers to tub-Gal4/+ on the 3rd chromosome. All transgenes were present in a single copy. The RNA level of each gene was normalized to RpL32 and expressed relative to the no transgene, CyO-GFP flies. The barplot represents the mean ± s.e.m. of at least 3 replicates with overlaid data points reflecting one individual larva.

Extended Data Fig. 7 Validation of mammalian reconstitution models.

a, Co-immunoprecipitations of Flag-tagged MSL2WT, MSL2dCTD or MSL2dCXC-CTD hybrid knock-in (also see Fig. 3a) and untagged parental control mES cell lines with cropped immunoblots for the MSL complex members MSL1 and MOF. RNA Pol2 serves as an input loading and negative immunoprecipitation control. Immunoprecipitations were conducted from nuclear extracts. Blots were cropped at the vertical dashed line to remove extra lanes. Displayed bands were derived from the same gels. For source data, see Supplementary Fig. 1. b, RT–qPCR analysis of the mouse MSL2 target genes Bex2, Phf8, Zfp185 and Tsix in parental (clone MF, n = 4), Msl2∆ (clone D12, n = 4), MSL2WT (clone D11, n = 4), MSL2dCTD (clone F4, n = 4) and MSL2dCXC-CTD (clone G2, n = 3) mES cells (also see Fig. 3a). The RNA level of each gene was normalized to Hprt and expressed relative to the parental cell line (MF). The barplot represents the mean ± s.e.m. with overlaid data points representing biological replicates. c, RNA-seq was conducted in MSL2WT (clone D11), knockout (Msl2∆, clone D12) or hybrid (MSL2dCTD, clone F4; MSL2dCXC-CTD, clone G2)-expressing mES cells. In the scatterplot, each dot shows the log2[fold-change] of the RNA expression of a mouse gene in a given MSL-hybrid cell line versus the MSL2WT mES cells. Differentially expressed (DE) genes were colored in red (upregulation) or blue (downregulation), if the FDR from DESeq2 was <0.05 in any of the MSL2 hybrid cell lines. The Pearson correlation coefficient of the gene expression changes was calculated with the panel.cor function in R. d, RNA-seq as in c. Heatmap and dendrogram displaying the log2[fold-change] of the significantly (FDR < 0.05) DE genes in any of MSL2 hybrid versus the control (MSL2WT) mES cells by RNA-seq. The log2[fold-change] of these genes in Msl2∆ knockout versus control (MSL2WT) was plotted to compare whether DE genes in MSL2 hybrid expressing cells are also misregulated in the knockout. e, Cropped agarose gel of PCR products across the Zfp185 locus from gDNA of parental and HAS knock-in clones derived from G2 (MSL2dCXC-CTD hybrid) as parental mES cell line (also see Supplementary Data 4) with primers binding outside both homology arms. This product was then re-amplified in a nested PCR for resolving the size shift introduced by replacing the 2 kb Zfp185 region with 1.5 kb of HAS sequences. The original PCR product was analysed by Sanger sequencing to verify insertion of the HAS and deletion (data not shown). For source data, see Supplementary Fig. 1. f, RT–qPCR analysis of Zfp185 and its two neighbouring genes, Nsdhl and Pnma5, the latter of which is also downregulated in Msl2∆ mES cells. Bex2 and Firre are target genes at a different locus, Rplp0 and Tbp serve as non-target controls. The Zfp185 knock-in cell lines displayed with green bars are all derived from the MSL2dCXC-CTD hybrid cell line (G2, blue). The clone IDs from top to bottom are D11 (MSL2WT, black), G2 (MSL2dCXC-CTD, blue), 444-A11, 444-C2, G2-F4, 376-28, 377-A4 and 377-D6. The RNA level of each gene is normalized to Hprt and expressed relative to the MSL2WT cell line. The barplot represents the mean ± s.e.m. of at least n = 4 biological replicates. g, as in Fig. 3f, RT–qPCR analysis of the indicated genes in roX2 (exon 3)- or EGFP control-expressing cell lines that were derived from mES cell clones D11 (MSL2WT), G2 (MSL2dCXC-CTD), G2-F4 (MSL2dCXC-CTD, roX1 HAS at Zfp185), 444-C2 (MSL2dCXC-CTD, socs16D HAS at Zfp185). The RNA level of each gene was normalized to Hprt. The barplot represents the mean with overlaid datapoints of biological replicates (n = 2 independently derived cell lines). h, RT–qPCR analysis of the indicated genes in mES cell clones D11 (MSL2WT), G2 (MSL2dCXC-CTD hybrid) or G2-A5 that expresses roX2 (exon 3) from the HBB-y locus (see Fig. 4c and Supplementary Data 4). The RNA level of each gene was normalized to Hprt. The barplot represents the mean ± s.e.m. of at least n = 2 biological replicates. The P-values were obtained by a two-sided Wilcoxon rank-sum test.

Extended Data Fig. 8 Characterization of roX2 in hybrid and tethered MSL2-expressing mES cells.

a, Schematic representation of the experimental setup showing the RNA levels after transcription inhibition with flavopiridol at the indicated time points in cell lines D11 (MSL2WT) and G2 (MSL2dCXC-CTD hybrid) expressing roX2 (random, single-copy integration). The RNA levels were determined by RT–qPCR, normalized to a spike that was added before RNA isolation and expressed relative to the 0 min control time point in the D11 (MSL2WT) cell line. The line plot connects the mean at each time point ± s.e.m. of n = 7 (MSL2WT) or n = 5 (MSL2dCXC-CTD hybrid) biological replicates. Nascent RNA levels decrease at timeframes correlating with the length of the analysed gene (for example, Pgk1: 16.6 kb, intron levels drop by 50% after around 15 min). mRNAs (for example, Rplp0, Pgk1), rRNA (RNA Polymerase I) or U6 snRNA (RNA Polymerase III) remain rather unaffected within the timeframe of 15 min. Total roX2 levels (0.5 kb) decrease by 50% within 15 min of flavopiridol inhibition, pointing towards a rapid turnover (also see Fig. 3h). b, Cropped immunoblots comparing the Flag-tagged MSL2WT (D11) with the MSL2dCXC-∆CTD-3xFlag-His-λN cell line (clone 513-A6, abbreviated as MSL2∆λN). The coomassie-stained gel, DHX9 and OCT3/4 serve as loading controls. For source data, see Supplementary Fig. 1. c, Representative RNA FISH pictures of roX2 RNA (red) and DAPI (light blue) in the MSL2dCXC-CTD hybrid mES cells (clone G2) with stably integrated roX2 (top) or the MSL2dCXC-∆CTD-3xFlag-His-λN mES cells (clone 513-A6, abbreviated as MSL2∆λN) with stably integrated 2×boxB-roX2 (bottom). The pictures are single z-planes with scale bars = 10 μm. d, Quantification of RNA FISH in c obtained from 3 independent hybridizations in 2 stable clones. Induction of the roX2 foci in mouse ESCs by either the drosophilized hybrid (MSL2dCXC-CTD) or by ectopic tethering of the mouse MSL2 to roX RNA (MSL2∆λN) is compared. Left, dot plots with centre line represent the mean ± s.e.m. foci frequency, where each overlaid data point represents the quantification result of a single tiled image. Each image was quantified as the number of cells that display a confined signal in the nucleus as % of all nuclei identified based on DAPI staining (% of nuclei with foci/total nuclei). The total number of quantified nuclei examined over all experiments was n = 447 or n = 284 cells and is displayed on top of the dot plot. Statistical significance was evaluated by a two-sided Wilcoxon rank-sum test. Right, boxplots representing the area of the roX2 foci obtained from orthogonal projection images. Overlaid data points represent individual foci (n = 140 and n = 85). The P-values were obtained with a two-sided Wilcoxon rank-sum test.

Extended Data Fig. 9 Validation of mammalian reconstitution models expressing roX2 from Zfp185.

a, Schematic representation of experimental outline to create mES cells with targeted integrations of roX2 at mammalian loci to study induction of a local dosage compensation-like phenomenon. A scarless CRISPR strategy without integration of a selection cassette was used30. roX2-expressing cells were derived from the MSL2dCXC-CTD hybrid cell line (clone G2). roX2 or shuffled roX2 expression was driven by the Pgk1 promoter. The roX2 expression cassette is followed by 1.5 kb of Drosophila HAS (roX1, socs16D) or control sequences: odsh (Drosophila non-target) and Bscl2 (Mouse target). All cell lines are reported in Supplementary Data 4. b, Cropped agarose gel of PCR products across the Zfp185 locus from gDNA of parental and Pgk1 promoter-driven roX2 knock-in clones derived from G2 (MSL2dCXC-CTD) as parental mES cell line (also see Supplementary Data 4). The locus was amplified with primers in the Pgk1 promoter and 3′ outside of the homology arm. The PCR product was analysed by Sanger sequencing to verify the insertions. The control PCR shown below each gel amplifies the HBB-y locus. For source data, see Supplementary Fig. 1. c, RNA FISH of roX2 (red), Huwe1 (white) and DAPI (blue) of the mES cells MSL2dCXC-CTD (G2), Pgk1-roX2-mCTRL (422-C9 (Bscl2)), Pgk1-roX2-socs16D HAS (424-B12). The pictures are single z-planes with scale bars = 5 μm, the hybridization was performed three times with similar results (also see Supplementary Note for technical information on RNA FISH). The co-localization with the X-linked Huwe1 RNA implies that these foci, which occur independently of the presence of HAS, correspond to transcripts at the targeted Zfp185 locus. d, RNA FISH for roX2 upon exposure of mES cells to 3% 1,6-hexanediol for 3 or 10 min. mES cells express roX2 from Zfp185 (cell line 419-D2), the MSL2-CTD is drosophilized (MSL2dCXC-CTD hybrid). Scale bars = 5 μm, the hybridization was performed once. Right, quantification of the roX2 foci areas (n = 27 (0 min), n = 20 (3 min), n = 59 (10 min)). Barplots show mean ± s.e.m. with overlaid data points representing individual foci. The P-values were obtained by a Kruskal–Wallis test followed by a Dunn’s statistical test for multiple comparisons (see Extended Data Fig. 4a-g for response of the Drosophila MSL territory to 1,6-hexanediol treatment). e, RT–qPCR analysis of the indicated genes. The roX2 Zfp185 knock-in cell lines are displayed with red shaded bars, shuffled roX2 with grey shaded bars, the wild-type mouse MSL2-expressing cell line (D11) in black and the MSL2dCXC-CTD hybrid cell line (G2) in blue. The RNA level of each gene was normalized to Hprt and expressed relative to the MSL2WT cell line. The barplot represents the mean ± s.e.m. of at least 3 biological replicates. The roX2 and shuffled roX2 RNA levels are expressed as a geometric mean over all cell lines, as these transcripts are not expressed in the MSL2WT cell line. All cell lines and RT–qPCR significance testing for Pnma5 are listed in Supplementary Data 4. A subset of the same data (result for MSL2WT, D11; MSL2dCXC-CTD, G2; 424-B12 (roX2) and 443-C3 (shuffled roX2) is shown in Fig. 4a. f, H4K16ac ChIP–qPCR analyses from MSL2WT (D11), MSL2dCXC-CTD (G2) and Pgk1-roX2-HAS (424-B12, 421-B5, 419-C3)-expressing mES cells. The barplot with overlaid data points shows the mean of n = 2 independent experiments (n = 1 for MSL2WT) at the indicated positions along the Zfp185 locus. Enrichment values were calculated relative to input and serial dilutions performed to account for primer efficiency. The data are expressed as fold change enrichment over the Intergenic 3 region (see Supplementary Data 4), which resides near the transcriptionally inactive HBB-y gene (for clarity not shown in panel).

Extended Data Fig. 10 Validation of mES cells expressing roX2 from Bex2, HBB-y and Drosophila tethering experiment.

a, Cropped agarose gel of PCR products across the Bex2 locus from gDNA of parental and Pgk1 promoter-driven roX2 knock-in clones derived from G2 (MSL2dCXC-CTD) as parental mES cell line (also see Supplementary Data 4). For source data, see Supplementary Fig. 1. b, RT–qPCR analysis of the indicated genes as in Fig. 4b. The RNA level of each gene was normalized to Hprt and expressed relative to the MSL2WT cell line (arbitrary units for roX2, since it is not expressed in MSL2WT cells). The barplot represents the mean ± s.e.m. of at least 3 biological replicates. c, as in a, but for cells with roX2 integrated into the HBB-y locus. For source data, see Supplementary Fig. 1. d, RT–qPCR analysis of the indicated genes as in Fig. 4c. The barplot represents the mean ± s.e.m. RNA levels (normalized to Hprt) of at least 3 biological replicates. RNA levels represent arbitrary units. HBB-bs gene could only be detected in RT–qPCR from a mouse control cDNA, but not from the experimental mES cells. roX2 is not expressed in MSL2WT cells. e, as in b but in Msl3Δ knockout cells derived from MSL2WT (D11 ± Msl3Δ, black), MSL2dCXC-CTD (G2 ± Msl3Δ, blue) and MSL2dCXC-CTD + roX2 (498-B6 ± Msl3Δ red) cells. roX2 in clone 498-B6 (red bars) is expressed from the Bex2 locus (Fig. 4b). f, Cropped immunoblots of total protein extracts of adult male fly heads expressing MS2CP-tagged MSL2 transgenes (also see Fig. 4d) in comparison with wild-type control (msl-2::HA). All transgenes also contain a 2xHA-mCherry tag and were expressed with tub-Gal4 in a heterozygous msl-2227/kmA/CyO-GFP background. To account for differences in protein stability, the MSL2ΔMS2CP (86FB) and MSL2MS2CP (VK33) transgenes are expressed from different landing sites. The star indicates the expected migration of the full-length fusion proteins, whereas the lower products correspond to degradation products. The experiment has been performed once. For source data, see Supplementary Fig. 1. g, Analysis of adult male viability in flies expressing transgenic MS2CP-tagged or untagged transgenes. The genotype was roX1+, roX2+; msl-2kmA/msl-2227; tub-Gal4/UAS-msl-2 or respective heterozygous CyO-GFP control. The data was expressed relative to females obtained from the same cross. The barplot represents the mean ± s.e.m. with overlaid datapoints reflecting the results from each cross/vial (n = 5 for MSL2MS2CP, n = 7 for MSL2ΔMS2CP, n = 6 for MSL2WT). Details on genotypes and nature of transgenes are provided in Methods. h, Transgenic flies expressing MS2CP-tagged MSL2 transgenes are ectopically tethered to MS2-loops tagged roX RNA (also see Fig. 4d–h, MSL2WT = wild-type D. melanogaster transgene, MSL2MS2CP = wild-type MS2CP-tagged and MSL2ΔMS2CP = MSL2(1-520) MS2CP-tagged). The transgenes were expressed as UAS-msl-2 versions in act-Gal4, msl-2Δ/msl-2227 transheterozygous null mutant backgrounds, where the X chromosome is roX1+ roX2+ (no loops) or roX16xMS2, roX2Δ (with tethering, indicated with loops symbol). For cross setup and details see Methods. Immunofluorescence was performed in salivary glands and the msl-2 transgenes were detected with HA antibodies (red), MSL1 is shown in green, DAPI in blue. The pictures are orthogonal projections of an entire stack with scale bars = 50 μm. i, as in h but in imaginal discs. msl-2 transgenes were detected with HA antibodies (yellow), DAPI in blue. The representative pictures are single z-planes with scale bars = 10 μm. j, Top, schematic representation of the strategy to determine the territory enrichment ratio, which is calculated by dividing average intensity of the territory in one nucleus by the average intensity of an equally sized region elsewhere in the same nucleus. Bottom, quantification of h, density plots of the X-chromosome territory enrichment ratio. The number of quantified nuclei is MSL2WT (no loops n = 22, with loops n = 24), MSL2MS2CP (no loops n = 45, with loops n = 42) and MSL2ΔMS2CP (no loops n = 37, with loops n = 37). The P-values were obtained with a two-sided Wilcoxon rank-sum test comparing the MSL2 enrichment in comparison with DAPI. DAPI is equally dispersed within the nucleus and shows a territory enrichment ratio of 1 (= not enriched).

Supplementary information

Supplementary Note

Further explanation and discussion of ChIP-seq and RNA-seq data, Drosophila MSL complex members response to 1,6-hexanediol treatment in vivo, Analysis of MSL2 CXC-domain hybrids, Sequence evolution and in vitro properties of MSL2 domains, Hierarchy between roX2 and roX1, RNA FISH technical notes and Supplementary discussion.

Reporting Summary

Supplementary Figure 1

Uncropped gel pictures and immunoblots.

Supplementary Data 1

List of differentially expressed (DE) genes and GO-terms obtained in RNA-seq from dissected Drosophila wing imaginal discs.

Supplementary Data 2

Branch-site test results for MSL2 RING, CXC and CTD.

Supplementary Data 3

Individual results per cross regarding viability of male msl-2 null mutants expressing MSL2 hybrids.

Supplementary Data 4

Related to Methods, table with primer sequences, plasmids, cell lines, HAS sequences, RNA FISH probes, Drosophila lines, Main Figure Drosophila genotypes and statistical testing.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Valsecchi, C.I.K., Basilicata, M.F., Georgiev, P. et al. RNA nucleation by MSL2 induces selective X chromosome compartmentalization. Nature 589, 137–142 (2021). https://doi.org/10.1038/s41586-020-2935-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41586-020-2935-z

This article is cited by

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing