Wash interacts with lamin and affects global nuclear organization

Abstract

The cytoplasmic functions of Wiskott-Aldrich syndrome family (WAS) proteins are well established and include roles in cytoskeleton reorganization and membrane-cytoskeletal interactions important for membrane/vesicle trafficking, morphogenesis, immune response, and signal transduction. Misregulation of these proteins is associated with immune deficiency and metastasis [1-4]. Cytoplasmic WAS proteins act as effectors of Rho family GTPases and polymerize branched actin through the Arp2/3 complex [1, 5]. Previously, we identified Drosophila washout (wash) as a new member of the WAS family with essential cytoplasmic roles in early development [6, 7]. Studies in mammalian cells and Dictyostelium suggest that WASH functions primarily in a multiprotein complex that regulates endosome shape and trafficking in an Arp2/3-dependent manner [8-11]. However, roles for classically cytoplasmic proteins in the nucleus are beginning to emerge, in particular, as participants in the regulation of gene expression [12, 13]. Here, we show that Drosophila Wash is present in the nucleus, where it plays a key role in global nuclear organization. wash mutant and knockdown nuclei disrupt subnuclear structures/organelles and exhibit the abnormal wrinkled morphology reminiscent of those observed in diverse laminopathies [14-16]. We find that nuclear Wash interacts with B-type Lamin (Lamin Dm0), and, like Lamin, Wash associates with constitutive heterochromatin. Wash knockdown increases chromatin accessibility of repressive compartments and results in a global redistribution of repressive histone modifications. Thus, our results reveal a novel role for Wash in modulating nucleus morphology and in the organization of both chromatin and non-chromatin nuclear sub-structures.

Figure 1

Wash is in the nucleus and disrupts nuclear morphology. (A) Wash is expressed in both the nucleus and cytoplasm as shown by western blot analysis of nuclear and cytoplasmic Drosophila Kc167 cell extracts. Extract specificity shown by western blot analysis with Lamin (nuclear) and β-tubulin (cytoplasmic). (B-B’) Micrographs of immunostained S2R+ cells (single focal plane) showing Wash is both cytoplasmic and nuclear. (C-D) Wash staining in S2R+ cells treated with dsRNA to GFP (control; C) or Wash (D) showing specificity of the Wash antibody. (E-E’) Confocal micrograph of larval salivary glands (projection) showing Wash is present in both the cytoplasm and the nucleus. (F-G) Nuclear and cytoplasmic Wash staining in wildtype salivary gland cells (F) and its absence in wash^Δ185 mutants (G). (H-I’) S2R+ cells treated with 5mg/ml digitonin (to permeabilize only the plasma membrane; I-I’) or 0.2% triton X-100 (to permeabilize both the plasma and nuclear membranes; H-H’), then stained for Wash and H3K27me3. Nuclear Wash and H3K27me3 staining are not detected when the nuclear membrane is not permeabilized (I-I’). (J-K’) Wash associates with specific regions on third-instar larval polytene chromosomes. (L-O’) wash knockdown mutants exhibit morphological alterations in nuclear shape. Confocal projections of S2R+ cells treated with dsRNA for GFP (L-L’, N-N’) or Wash (M-M’, O-O’) then stained for Lamin (L-M’), microtubules (MT) (L-M), actin (N-O) and DNA (DAPI; L-M, N-O) showing that Wash knockdown disrupts nuclear morphology in addition to cytoplasmic architecture. (P) Quantification of nuclear shape concavity in S2R+ cells treated with dsRNA to GFP (9.8±1.2%, n=176) and Wash (17.0±1.6%, n=211) (P=0.0003). Scatterplot with median ± IQR shown. (Q-V) Confocal projections of wildtype versus wash^Δ185 mutant salivary glands stained for actin and Lamin show that Wash affects nuclear morphology without gross cytoplasmic defects. Whole salivary glands (Q-R), salivary gland cells (S-T), and salivary gland nuclei (U-V). (X) Quantification of nuclear shape concavity in wildtype (2.7±0.68%, n=23) and wash^Δ185 (18.9±3.2%, n=23) salivary gland nuclei (P<0.0001). Scatterplot with median ± IQR shown. See also Figure S1.

Figure 2

Wash disrupts nuclear and genome organization. (A-F) Confocal projections of wildtype and wash^Δ185 mutant salivary gland polytene chromosomes; wash^Δ185 chromosomes show aberrant alignment and banding (D) and are extremely fragile (E-F). (G-H) Fluorescent in situ hybridization of chromosome-specific BAC pools hybridized to the X (yellow), second (green) and third (red) chromosome in wildtype and wash^Δ185 mutant salivary gland nuclei shows less compact chromosome territories in wash mutants. (I-N) Confocal micrographs of wildtype and wash^Δ185 mutant salivary gland nuclei stained for DNA and nuclear markers. HP1 (green) chromocenter localization is lost in wash^Δ185 nuclei (I-J). Coilin (green) cajal body localization is disrupted in wash^Δ185 nuclei while Mtor (red) remains localized to the periphery of salivary gland nuclei (K-L). Fibrillarin (green) localization at the nucleolus is disrupted in wash^Δ185 nuclei (M-N). While MOF localizes properly to the X chromosome in both wildtype and wash^Δ185 nuclei, it highlights the disruption to chromosome territory compaction observed in wash mutants (M-N). (O-T) Confocal projections of histone modifications in wildtype and wash^Δ185 salivary gland nuclei. Both repressive histone marks (H3K9me3 (O-P) and H4K20me2 (Q-R)) and the active histone mark H3K4me3 (S-T) are reduced in wash nuclei. (U) Schematic diagram of the Wash constructs used to generate the transgenic lines indicating the position of the added NES, GFP fusion, and the substitution mutations (WKRS>AAAA) in the Wash NLS (not drawn to scale). (V-DD) Specific reduction of Wash in the nucleus results in disrupted nuclear shape and sub-nuclear structure/organelle organization. Confocal projections of salivary gland nuclei from wash^Δ185 P{GFP-Wash^WT} (V-V’”, X, AA) and wash^Δ185 P{GFP-Wash^+NESΔNLS} (W-W’”, Y, BB) stained for Lamin (V-V’, W-W’, AA-BB), GFP (V’-V”, W’-W”), Wash (X-Y), and Coilin (AA-BB). Quantification of nuclear shape concavity in wash^Δ185 P{GFP-Wash^WT} (WT: 4.74±0.67%; n=27) or wash^Δ185 P{GFP-Wash^+NESΔNLS} (ΔNLS: 8.58±1.09%; n=27) transgenic salivary gland nuclei (P=0.0045) (Z). Scatterplot with median ± IQR shown. Quantification of the number of Coilin puncta in wash^Δ185 P{GFP-Wash^WT} (WT: 1.04±0.04; n=26) or wash^Δ185 P{GFP-Wash^+NESΔNLS} (ΔNLS: 2.38±0.40; n=24) transgenic salivary gland nuclei (P=0.0028) (CC). Quantification of average dispersion of Coilin puncta in wash^Δ185 P{GFP-Wash^WT} (WT: 0.04±0.04; n=26) or wash^Δ185 P{GFP-Wash^+NESΔNLS} (ΔNLS: 0.69±0.14; n=24) transgenic salivary gland nuclei (P=0.0001) (DD). Boxplot graphs show the median and 25% and 75% percentile measures. The whiskers indicate variability outside the upper and lower quartiles. See also Figure S1 and Movie S1.

Figure 3

Wash interacts with lamin and associates with LADs. (A-B) Wash interacts with Lamin directly by GST-pulldown assays using IVT (A) or bacterially expressed proteins (B). (C) Wash and Lamin interact in vivo. Western blot of immunoprecipitations from embryo nuclear extracts with Wash, no primary antibody, or with an unrelated antibody (9e10). (D) Duolink Proximity Ligation Assay performed with antibodies recognizing Wash and GFP in GFP-Lamin expressing salivary glands. Duolink signal is only observed if the two antibodies are within 30 nm. (E) Genome-wide chromatin profile of Lamin (red) and Wash (blue) bound regions. (F) Chromosome-3R region aligned to developmental expression showing a significant overlap of LADs with Wash-associated regions at transcriptionally silent regions. Track with squished view of genes is shown in blue. (G) Venn-diagram comparing LADs to Wash associated chromatin regions (P<1×10⁻⁶). (H) LADs were aligned at their ends (+/− 3 Kb) and the normalized DamID probe signals were averaged in 150-bp bins for Wash occupancy. (I) DamID-based 5 color chromatin states and random sequences were aligned at their ends and the normalized DamID probe signals were averaged in 150-bp bins for Wash occupancy. YELLOW and RED chromatin contain proteins and histone modifications characteristic of active chromatin. BLUE chromatin contains H3K27me3 and PcG proteins, GREEN chromatin contains HP1 and Su(var)3-9, and BLACK chromatin contains Lamin and histone H1. (J) Wash chromatin profile on a 60Kb region of chromosome 2L showing an inverse correlation with developmental gene expression (similar results were obtained with the RNAseq data for Kc167 cells). See also Figure S2.

Figure 4

Wash and Lamin affect chromatin accessibility at heterochromatic regions and Position Effect Variegation. (A-F) Distribution of M.SssI-based chromatin accessibility around active promoter, constitutive heterochromatin and heterochromatin-like euchromatin (HLE) regions. modENCODE Consortium generated chromatin states were aligned at their 5’ and 3’ (+/− 1.5 Kb) and the normalized probe signals were averaged in 50-bp bins. The y-axis in each plot represents the relative enrichment of M.SssI-methylated DNA for mock and Wash knockdown (A-C) and for mock and Lamin knockdown (D-F) in S2R+ cells. Statistical significance (determined using the two-sample KS test): TSS: Wash, P=0.823, Lamin P<4×10⁻³; Heterochromatin: Wash, P<2×10⁻³, Lamin P<4×10⁻³; HLE: Wash, P= 0.256, Lamin P<2×10⁻³. (G-J) wash and lamin suppress brown-mediated PEV. Wash and Lamin mediate suppression of the classical PEV allele bw^VDE2, a chromosomal rearrangement that juxtaposes the bw gene near 2R heterochromatin (G-H). wash (male & female) and lamin (female only) mediate suppression of the non-classical PEV allele bw^D, which inserts heterochromatin into the bw gene and silences the homolog (I-J). (K-L) wash and lamin enhance white mediated centromeric PEV (w⁺ gene inserted in proximal X chromosome heterochromatin). Bar plots showing percentage of flies falling into each expression quintile (flies sorted into one of five bins based on the percentage of ommatidia expressing the w⁺ marker or bw gene (bin 1 = 0-20% to bin 5 = 80-100%). The median±SEM and P-values are given in each panel. Eyes shown are representative of the average phenotype for each genotype and each pair is an age-matched, sibling pair (G-L). The number of eyes scored (N) is given beside each eye. See also Figures S3 and S4.