RID is required for both repeat-induced point mutation and nucleation of a novel transitional heterochromatic state for euchromatic repeats

Abstract

To maintain genome integrity, repeat sequences are subject to heterochromatin inactivation and, in Neurospora, repeat-induced point mutation (RIP). The initiating factors behind both are poorly understood. We resolve the paradoxical observation that newly introduced Repeat-Linker-Repeat (R-L-R) constructs require RID alone for RIP, while genomic repeats are RIPed in the absence of RID, showing that eu- and hetero- chromatic repeats are handled differently, the latter additionally requiring DIM-2. The differences between mechanisms associated with older and newer duplicates caution against extrapolation from mechanisms inferred from model experimental systems. Additionally, while chromatin status affects RIP, we also show that RID, when tethered with LexA, acts as a nucleation center for the transition from euchromatin to heterochromatin in an HDA-1 dependent fashion. Constitutive heterochromatin by contrast is largely HDA1 independent and depends on HDA-1 paralogs. RID is thus a dual function initiator of both RIP and the transition to heterochromatin.

Graphical Abstract

Figure 1.

RIP patterns in R-L-R constructs and native RIPed duplicates colocalizing with heterochromatic regions. (A) RIP on R-L-R segments with different divergence. Besides the original R-L-R segment, two additional tester constructs with 5% and 10% divergence (C and D) were made in WT, rid^Δ and dim-2^Δ background. (B) The RIP on genomic duplicates from parents in WT, rid^Δ, dim-2^Δ and rid^Δ; dim-2^Δ background to progeny ascospores. Significance levels: ***P ≤ 0.001, **0.001 < P ≤ 0.01, *0.01 < P ≤ 0.05; ns, P > 0.05. (E) The colocalization of native duplicates, heterochromatic H3K9me3 regions, and 5mC methylation. 1.5–1.9 Mb on Chr7 is shown here.

Figure 2.

RID is tightly involved in RIP and de novo heterochromatin formation on unmethylated Sly1-1. (A) The hybridization diagram to investigate RID’s function on different types of duplicates. (B) RID controls de novo methylation on copy A (Sly1-1). The level of DNA methylation on Sly1-1 was analyzed in WT, rid^Δ and dim-2^Δ backgrounds. Methylation level was calculated as 50bp windows for the 11 kb Sly1-1 sequence and 500 bp windows for 1.75–1.9 Mb on Chr7 as the 5mC constitutive control. (C) and (D) RIP on Sly1-1 and genomic duplicates from crosses of WT, rid^Δ and dim-2^Δ backgrounds to progeny ascospores. Significance levels: ***P ≤ 0.001, **0.001 < P ≤ 0.01, *0.01 < P ≤ 0.05; ns, P > 0.05. (E) H3K9me3 modification on Sly1-1 requires RID. The relative enrichment of H3K9me3 at the site of Sly1-1 was assessed by ChIP-qPCR in WT and rid^Δ background. Data were normalized to ratios obtained without immunoprecipitation (total input), and the H3K9me3 hotspot and coldspot as the positive control and negative control, respectively. Error bars represent the standard error of the mean.

Figure 3.

RID-LexADBD under native rid promoter could induce RIP and de novo heterochromatin formation. (A) Quantification of rid expression levels in asexual hypha and sexual ascocarp using RT-qPCR, including both the WT and RID::LexADBD strains. (B) Western Blot analysis of 6xHis::RID in hypha and ascocarp shows much higher expression in the sexual stage, while expression in asexual stage is barely detectable. (C) Tethered RID induces DNA methylation at a euchromatic locus. The LexADBD-fused protein was targeted to LexAO sequence integrated downstream of the his-3 locus. Before any protein is tethered, LexAO as well as surrounding regions are confirmed to be unmethylated (methylation level in 50 bp windows). When HP1 or RID was tethered in vegetative hyphae, DNA methylation was induced only when dim-2 is present. (D) Tethered RID can establish H3K9me3 near the region of LexAO. Enrichment of H3K9me3 is displayed using the Integrative Genomics Viewer, and the y-axis represents RPKM. (E) The rate of mutation occurring on Sly1-1 (left) and constitutive heterochromatin (right) were compared between WT and RID-LexADBD backgrounds. The mutation rate in RID-LexADBD is significantly higher for Sly1-1 (Brunner–Munzel test, P = 0.022 for Sly1-1 and P= 0.48 for constitutive heterochromatin).

Figure 4.

HDA-1 is required for RID-mediated de novo heterochromatin formation. (A) Distinct influence of HDA-1 and HDA-2/3 on maintenance of 5mC methylation. The genome was divided into 100 bp windows, and the methylation level for each window was quantified as 5mCs/(Cs + 5mCs). The histogram shows regions with altered methylation levels compared to WT, calculated as (“methylation level of mutant”—“methylation level of WT”)/ “methylation level of WT,” where 0 indicates no change and −100% indicates complete loss. Different bin sizes were tested, and only bins with a methylation level > 0.1 in WT were included to highlight the influences of HDACs on methylation maintenance. (B) HDA-1 is necessary for H3K9me3 induction on Sly1-1. The relative enrichment of H3K9me3 at the site of Sly1-1 was assessed by ChIP-qPCR in WT, hda-1 knocked out and re-introducing strains. Data were normalized to ratios obtained without immunoprecipitation (total input), and the H3K9me3 hotspot and coldspot as the positive control and negative control, respectively. Error bars represent the standard error of the mean. (C) 5mC Methylation changes on Sly1-1 and constitutive heterochromatin after HDA-1 KO and its re-introduction. Methylation levels are quantified in 50 bp windows for the 11kb Sly1-1 sequence (left) and in 500 bp windows for 1.75–1.9 Mb on Chr7, the 5mC constitutive control (right). The 5mC methylation level of WT parent (W₀), the sexual progenies of WT (W₁), sexual progenies of hda1^Δ (h₁) and h₁ with reintroduced HDA-1 (h₁^HDA-1) were shown. KO of HDA-1 hindered the increase of methylation level on Sly1-1(A) but do not affect the methylation level at many constitutive heterochromatic regions. (D) HDA-1 is essential for maintenance of methylation in transitional heterochromatin but is dispensable for stable constitutive heterochromatin. Strains containing AB copies were repeatedly crossed in both hda-1⁺ and hda-1⁻ background and DNA methylation levels were assessed in both WT and knock out strains. The percentages represent the level of DNA methylation of Sly1-1(A) before and after HDA-1 KO.

Figure 5.

Transition from euchromatin to constitutive heterochromatin in N. crassa. (A) Schematic of the RID and HDA-1 KO in different generations. (B) RID functions differently in RIP on Sly1-1(A) from W₁ to W₂. KO of RID results in complete loss of RIP in first-generation progeny ascospores (Fig. 2C) but a few RIP mutations remain in second-generation ascospores (r₂). The remaining few mutations disappear after additional KO of HDA-1. Error bars represent the standard error of the mean (s.e.m.), and significance levels are indicated as follows: ***P ≤ 0.001, **0.001 < P ≤ 0.01, *0.01 <P ≤ 0.05; ns, P > 0.05. (C) Model for de novo heterochromatin formation of unRIPed duplicates. DNA is shown wrapped around nucleosomes with a region of repeats, and the dashed lines represent the repeated DNA interactions in the sexual phase. In the upper section, RIP (horizontal bars) can be mediated by RID. In the lower section, RID and HDA-1 are crucial for the establishment of DNA methylation and H3K9me3, serving as signals for the formation of de novo heterochromatin. Subsequently, the transition of de novo heterochromatin into constitutive heterochromatin is dependent on HDA-1. After multiple rounds of RIP, constitutive heterochromatin becomes insensitive to HDA-1.