Structural basis for interaction between CLAMP and MSL2 proteins involved in the specific recruitment of the dosage compensation complex in Drosophila

Abstract

Transcriptional regulators select their targets from a large pool of similar genomic sites. The binding of the Drosophila dosage compensation complex (DCC) exclusively to the male X chromosome provides insight into binding site selectivity rules. Previous studies showed that the male-specific organizer of the complex, MSL2, and ubiquitous DNA-binding protein CLAMP directly interact and play an important role in the specificity of X chromosome binding. Here, we studied the highly specific interaction between the intrinsically disordered region of MSL2 and the N-terminal zinc-finger C2H2-type (C2H2) domain of CLAMP. We obtained the NMR structure of the CLAMP N-terminal C2H2 zinc finger, which has a classic C2H2 zinc-finger fold with a rather unusual distribution of residues typically used in DNA recognition. Substitutions of residues in this C2H2 domain had the same effect on the viability of males and females, suggesting that it plays a general role in CLAMP activity. The N-terminal C2H2 domain of CLAMP is highly conserved in insects. However, the MSL2 region involved in the interaction is conserved only within the Drosophila genus, suggesting that this interaction emerged during the evolution of a mechanism for the specific recruitment of the DCC on the male X chromosome in Drosophilidae.

Figure 1.

Drosophila protein CLAMP interacts with MSL2 through its N-terminal zinc-finger domain. (A) NMR structure of CLAMP^87–153 (PDB: 7NF9). (B) Relative perturbation of CLAMP^87–153 residues’ chemical shifts upon interaction with MSL2. Full spectra are shown in the Supplementary Figure S6. (C) Residues with the strongest chemical shifts are shown in red at the CLAMP N-terminal zinc finger structure. (D) The MSL2-contact surface of the CLAMP N-terminal zinc finger. (E) Multiple sequence alignment of CLAMP N-terminal zinc-fingers from various insects. Typical DNA-binding residues are shown, and red asterisks mark the residues displaying the largest chemical shift perturbations. Triangles depict residues subjected to mutagenesis, red represents a negative effect on binding, and green represents no detectable effect. (F) Schematic representation of zinc-finger structure showing the secondary structure, DNA- and MSL2-binding residue positions are depicted. Residues subjected to mutagenesis are shown in red circles. (G) GST-pulldown (left) and yeast two-hybrid assays (right) of the interaction between GST-tagged MSL2^618–655 and 6xHis-thioredoxin-tagged CLAMP^1–153 bearing point mutations within the zinc-finger domain. AD, activation domain; BD, DNA-binding domain of GAL4 protein. + or – denotes the ability of yeasts to grow on the media without histidine; assay plates are shown in the Supplementary Figure S9.

Figure 2.

The small unfolded sequence of MSL2 interacts with CLAMP. (A) Sections of ¹H–¹⁵N MSL2^618–655 spectra showing perturbations of chemical shifts of MSL2 residues upon CLAMP binding. Full spectra are shown in the Supplementary Figure S13. (B) Relative perturbation of MSL2 residues’ chemical shifts upon interaction with CLAMP. (C) Multiple sequence alignment of CLAMP-interacting MSL2 region from Drosophila species. Red asterisks show the residues with the largest chemical shift perturbations upon binding to CLAMP. Triangles mark residues subjected for mutagenesis, red represents a negative effect on binding, and green represents no detectable effect. (D) GST-pulldown assay of interaction between GST-tagged MSL2^618–655 (represented by *) and 6xHis-thioredoxin-tagged CLAMP^1–153.

Figure 3.

The CLAMP–MSL2 interaction is required for the correct recruitment of the dosage compensation complex. (A) Schematic representation of rescue constructs expressing 3xHA-tagged CLAMP 3xFLAG-tagged MSL2 proteins under the control of ubiquitin-p63E promoter; SV40 poly(A)–SV40 polyadenylation signal; attB is the site for φC31-mediated recombination used for site-specific insertion of the construct; yellow represents the intronless yellow gene used as a reporter. (B) Comparison of the viability (relative to clamp²/+ mutant background) of males and females upon the rescue of the clamp²/clamp² mutant background with CLAMP proteins expressed in heterozygous (k/TM6) and homozygous (k/k) transgenic constructs. Complete data are shown in the Supplementary Table S2. (C) Relative viability of females with ectopic expression of the MSL2 variants. (D) Effect of single amino-acid substitutions and replacement of the CLAMP binding domain (CBD) in the FLAG-tagged MSL2 variants on DCC recruitment shown by immunostaining of polytene chromosomes with anti-FLAG and anti-MSL1 antibodies in females. All staining's are shown in Supplementary Figures S18A and S19. Polytene chromosomes stained with MSL2 antibodies are shown in Supplementary Figure S18B. Scale bar is 20 μm.

Figure 4.

Comparison of the zinc-finger residues used for protein binding. Cartoons were drawn according to the following structures (PDB ID): 1F2I (Zif268-artificial peptide (40)), 1Y0J (FOG F1–GATA (38)), 1SRK (FOG F3–TACC3 coiled-coil (37)), 3W5K (Snail F2–importin beta (39)). The cartoons for intramolecular interactions mediated by zinc fingers are shown in the Supplementary Figure S21.