H3K9 Promotes Under-Replication of Pericentromeric Heterochromatin in Drosophila Salivary Gland Polytene Chromosomes

Abstract

Chromatin structure and its organization contributes to the proper regulation and timing of DNA replication. Yet, the precise mechanism by which chromatin contributes to DNA replication remains incompletely understood. This is particularly true for cell types that rely on polyploidization as a developmental strategy for growth and high biosynthetic capacity. During Drosophila larval development, cells of the salivary gland undergo endoreplication, repetitive rounds of DNA synthesis without intervening cell division, resulting in ploidy values of ~1350C. S phase of these endocycles displays a reproducible pattern of early and late replicating regions of the genome resulting from the activity of the same replication initiation factors that are used in diploid cells. However, unlike diploid cells, the latest replicating regions of polyploid salivary gland genomes, composed primarily of pericentric heterochromatic enriched in H3K9 methylation, are not replicated each endocycle, resulting in under-replicated domains with reduced ploidy. Here, we employ a histone gene replacement strategy in Drosophila to demonstrate that mutation of a histone residue important for heterochromatin organization and function (H3K9) but not mutation of a histone residue important for euchromatin function (H4K16), disrupts proper endoreplication in Drosophila salivary gland polyploid genomes thereby leading to DNA copy gain in pericentric heterochromatin. These findings reveal that H3K9 is necessary for normal levels of under-replication of pericentric heterochromatin and suggest that under-replication at pericentric heterochromatin is mediated through H3K9 methylation.

Keywords: Drosophila; Endoreplication; H3K9; H4K16; heterochromatin; under-replication.

Figure 1

H3K9 promotes endoreplication of the Drosophila salivary gland. (A) Polytene chromosome spreads from HWT pre-wandering third-instar larvae stained for PCNA (green) and DAPI to detect DNA (blue). Representative early-replicating (ER), early/mid-replicating (E-MR), mid/late-replicating (M-LR), late-replicating (LR), very late-replicating (VLR) and non-replicating (nR) PCNA patterns are shown. White boxes designate the chromocenter as identified by HP1a staining (HWT and H4K16R) or cytologically (H3K9R) (not shown). (B) Percentage of PCNA-positive (S phase; black) and PCNA-negative (G phase; grey) polytene chromosome spreads for HWT (n = 211), H3K9R (n = 286; p < 0.001) and H4K16R (n = 284; p > 0.05) genotypes. Significance was determined using the Chi-squared test. All S phase measurements were taken at the pre-wandering developmental stage. (C) Salivary gland area of HWT and H3K9R at 67–72 h after egg deposition (p = 0.0229), 91–96 h after egg deposition (p < 0.0001) and at the pre-wandering third-instar larval stage (p = 0.1838) (Student’s T test). (D) Percentage of PCNA-positive polytene chromosome spreads in each of the five replication timing pattern categories shown in A for HWT (n = 96), H3K9R (n = 65; p < 0.00001) and H4K16R (n = 101; p = 0.004956) genotypes. Significance determined using the Chi-squared test and HWT data as the expected categories. (E) Representative chromosome arms with E-MR and LR PCNA-patterns for HWT, H3K9R and H4K16R. In E-MR patterns, PCNA colocalizes with DAPI-dim inter-bands (green arrowheads) but not DAPI-bright bands (yellow arrowheads). In LR patterns, PCNA colocalizes with DAPI-bright bands (green/yellow double arrowheads).

Figure 2

DNA copy number in pericentric heterochromatin is elevated in H3K9R mutants. (A,B) Heatscatter plot of (A) H3K9R/HWT log₂ ratio and (B) H4K16R/HWT log₂ ratio of normalized copy number at 10kb windows along Chromosomes 2 and 3. LOESS regression line of modENCODE H3K9me2 ChIP signal is shown in red (GSE47260). (C) Quantification of mutant/HWT ratio of normalized copy number at 10kb windows for all major chromosome scaffolds (Chromosomes 2L, 2R, 3L, 3R, 4 and X) separated into pericentromeres (Peri) and chromosome arms (Arms) (* = p < 0.001; Student’s T test). Coordinates for pericentromeres and chromosome arms were defined in References [52,53] (see also Figure S1). (D) Representative polytene chromosome chromocenter from HWT, H3K9R and H4K16R wandering third-instar salivary glands stained with DAPI. Scale bar = 10 µM. (E) Quantification of cytological categories for HWT and H4K16R chromocenters as performed in Reference [48]. HWT and H4K16R chromocenters shown in panel D represent the organized category whereas the H3K9R chromocenter shown represents the severely disorganized category, which we previously reported comprises 72% of H3K9R chromocenters [48]. The distribution of chromocenters among the three categories between HWT and H4K16R is not statistically different (p > 0.0001; Chi squared test).

Figure 3

Under-replication of pericentric heterochromatin is H3K9-dependent. (A) Histogram of under-replicated domains identified by CNVnator [51] for HWT, H3K9R and H4K16R genotypes. Bin size is set to 10 kb. (B) Boxplot of the SuUR/OregonR, H3K9R/HWT and H4K16R/HWT log₂ ratios of DNA copy number at 10 kb windows at fully replicated (Full) and under-replicated (UR) domains as defined by CNVnator [51] in HWT (* = p < 0.001; Student’s T test). (C) Boxplot of the SuUR/OregonR, H3K9R/HWT and H4K16R/HWT log₂ ratios of normalized signal at 10kb windows at under-replicated domains (top panel) and fully replicated domains (bottom panel) in pericentromeres (Peri) or chromosome arms (Arms) (* = p < 0.001; Student’s T test). (D) Heatscatter plot of normalized signal at 10 kb windows at under-replicated domains in HWT versus H3K9R (top panels) or versus H4K16R (bottom panels) at pericentromeres (left panels) and chromosome arms (right panels).