Rachel's family, especially her mother Cathy, was also very supportive and encouraged me to go back to graduate school and fulfill this dream. I was lucky to have several people who encouraged me to go back to graduate school and believed in me even before I believed in myself.
INTRODUCTION OVERVIEW
DNA replication in the context of development will be described in detail, with particular emphasis on the work of the fruit fly Drosophila melanogaster, which has been an invaluable tool for the study of DNA for more than a century. These ideas are central to the research I conducted, which focused on the central question of how DNA replication and the chromatin environment are regulated at the limits of development.
The fundamentals of DNA replication
Mcm10 promotes further unwinding of DNA and replication protein A (RPA) binds to the newly available single-stranded DNA (ssDNA) (Douglas et al. 2018). While the essential components of the replication fork have been defined in vitro, the assembly of the replication fork in vivo is much more complex (Baris et al. 2022).
The unique challenges of development
- loops play multifaceted roles in the cell
- loops are a three-stranded nucleic acid structure canonically formed when nascent RNA from transcription reanneals to the template DNA strand, resulting in a
- loop formation also occurs in the context of phage T4 infection and in the replication of mitochondrial DNA. At least two different origins of replication within the T4
- loops are most frequently found at the promoter, the 5’UTR, TSS, TTS and 3’
Components of the Fanconi-Anemia complex have also been linked to R-loop resolution (García-Rubio et al. 2015). For example, the formation of R-loops at the promoter of COOLAIR causes transcriptional repression (Sun et al. 2013).
Rif1 inhibits replication fork progression and controls DNA copy number in Drosophila
Abstract
Introduction
Both heterochromatin and several euchromatic regions of the genome have reduced DNA copy number relative to overall ploidy (Nordman et al., 2011). Recently, however, SUUR has been shown to control copy number by directly reducing replication fork progression (Nordman et al., 2014).
Results
The mechanism by which SUUR is recruited to replication forks and how it inhibits their progression remains poorly understood. Here we investigate how SUUR is recruited to replication forks and how it inhibits fork progression.
The SNF2 domain is essential for SUUR function and replication fork localization As a first step in understanding the mechanism of SUUR function, we wanted to
In contrast, the SuUR DSNF mutant localized to heterochromatin, but its recruitment to active replication forks was severely reduced (Figure 2-1E; Supplemental Figure 2-2). Together, these results demonstrate that the SNF2 domain is important for ACID recruitment to replication forks and essential for ACID-mediated subreplication.
SUUR associates with Rif1
We verified that both full-length SUUR and the SNF2 domain were equally expressed (Supplementary Figures 2–3B). We conclude that SUUR and Rif1 exist in the same protein complex and the interaction between SUUR and Rif1 is independent of chromatin overhang.
Underreplication is dependent on Rif1
Indeed, SUUR associates with late replicating regions of the genome (Filion et al., 2010; Pindyurin et al., 2007). While characterizing the role of Rif1 in subreplication and patterns of DNA replication in endo-cycling cells, we observed differences in the heterochromatic regions of SuUR and Rif1 mutants.
Rif1 affects replication fork progression
This quantitative analysis of origin fork and replication fork progression revealed that origin fork was unaffected in the Rif1 mutant, as no major change in copy number at the origin of replication was detected when comparing wild-type and Rif1 mutant stage 13 follicle cells are not (Supplementary Table). 2-2). To determine whether Rif1 controls replication fork progression by increasing the period of EdU incorporation within the 7.5 h time window of gene amplification, comparable to a SuUR mutant, we quantified the fraction of stage 13 follicle cells that were EdU positive .
Rif1 acts downstream of SUUR
We sought to distinguish between these possibilities by determining whether SUUR could still associate with replication forks in the absence of Rif1 function. SUUR localized to both replication forks and heterochromatin in the absence of Rif1 function (Figure 2-5; Supplementary Figure 2-9).
Rif1 localizes to active replication forks
Localization of replication forks (EdU) and SUUR in the nuclei of wild-type and Rif1 mutant follicular cells. Together, these results indicate that Rif1 is associated with replication forks in amplifying follicular cells and cultured cells.
SUUR is required to retain Rif1 at replication forks
Consistent with Rif1 association with replication forks in amplifying follicle cells, Rif1 was enriched in EdU pulse samples compared to chase samples in cultured cells (Supplemental Figure 2-11A). If true, then Rif1 association with replication forks should be at least partially dependent on SUUR.
The PP1-interacting motif of Rif1 is necessary for underreplication
We isolated salivary glands from wandering 3rd instar larvae of the Rif1 PP1 mutant and Rif1 PP1/+ heterozygous animals as a wild-type control. Similar to the Rif1 mutant, underreplication was largely abolished in the Rif1 PP1 mutant (Figure 2-7A-C; Supplemental Table 2-1).
Discussion
Third, SUUR localizes to replication forks and heterochromatin in a Rif1 mutant, but is unable to inhibit replication fork progression in the absence of Rif1. Based on these observations, we defined a novel function of Rif1 as a regulator of replication fork progression.
SNF2 domain and fork localization
Overexpression of the C-terminal two-thirds of ACID is able to induce ectopic sites of underreplication. We suggest that the affinity of ACID for replication forks is significantly reduced at physiological levels in the absence of the SNF2 domain.
Rif1 controls underreplication
By subsequent analysis, we demonstrated that Rif1 plays a direct role in copy number control and that Rif1 acts downstream of SUUR in the underreplication process. It is possible that Rif1 may promote underreplication through a mechanism independent of SUUR.
Rif1 regulates replication fork progression
Our observations, together with these previous studies, leave open the possibility that Rif1-mediated control of replication fork progression could be an evolutionarily conserved function of Rif1. In addition, underreplication is dependent on the Rif1 binding motif on PP1, raising the possibility that the Rif1/PP1 complex is required to inhibit replication fork progression.
Materials and Methods Key Resources Table
Reagent type (species)
Designation Source or reference Identifiers Additional information
SUUR
DRSC
25437540) Antibody anti-Rif1
11056799 Recombinant
2012) (PMID:22454539) Software,
2016) (PMID:27079975) Software,
Strain List
BAC-mediated recombineering
Next, a gene block (IDT) was used to replace the galK cassette and generate a precise deletion within the SuUR gene. The SuUR ΔSNF BAC was injected into a strain harboring the 86 F8 landing site (Best Gene Inc.).
Generation of heat shock-inducible, FLAG-tagged SuUR transgenic lines
CRISPR mutagenesis
Both Rif1 1 and Rif1 2 mutants had substantial deletions of the Rif1 gene and both had frameshift mutations early in the coding region. Rif1 1 has a frameshift mutation at amino acid 14, while Rif1 2 has a frameshift mutation at amino acid 11.
Cytological analysis and microscopy
To generate a Rif1 allele defective for PP1 binding, the pU6-BbsI vector expressing gRNA targeting the 3' end of Rif1 was co-injected with a recovery vector containing mutagenized SILK and RVSV sites (SAAK and RASA) with 1 kb of homology to upstream and downstream of the mutagenized region. Ten randomly selected regions outside the nucleus were selected, and the mean signal intensity for these regions was averaged to determine the background signal for each image.
Rif1 antibody production
IP-mass spec
An analytical column was packed with 20 cm of C18 reversed phase material (Jupiter, 3 μm beads, 300 Å, Phenomenox) directly into a laser drawn emitter tip. Peptides were loaded onto the capillary reverse phase analytical column (360 μm O.D. x 100 μm I.D.) using a Dionex Ultimate 3000 nanoLC and autosampler.
Genome-wide copy number profiling
Bioinformatics
The average read depth per region was determined by multiplying the number of reads in a region by the read length and dividing by the total region length. For read depth in pericentric heterochromatin regions, the chromatin arm was binned into 10 kb windows and the number of reads for each window was called using bedtools multicov with only uniquely assigned reads.
Copy number analysis by droplet-digital PCR (ddPCR)
Half-maximal analysis of multiplet copy number profiles was performed as described previously (Alexander et al., 2015; Nordman et al., 2014). Briefly, log 2 ratios were generated using bamCompare from deepTools 2.5.0 (Ramírez et al., n.d.) by comparing cell profiles of stage 13 follicles to the 0–2 h embryo sample.
Western blotting
To quantify protein abundance by mass spectrometry, raw mass spectrometry data were imported into Skyline version 4.1 (Schilling et al., 2012).
Proximity Ligation Assay (PLA) with nascent DNA
Cells were washed 2 times in Wash Buffer A, then incubated in amplification buffer at 1:80 for 100 min at 37°C. Slides were washed in Wash Buffer B, then 1:100 dilution of Wash Buffer B before being mounted in Duolink In Situ Mounting Media with DAPI.
Data access
Acknowledgements
Author contributions
Supplementary Table 2-1: Underreplicated regions called by CNVnator
Supplementary Table 2-2: Half-Max called follicle widths Wild
Supplemental Figure 2-9: Quantification of SUUR signal intensity at replication forks in the presence and absence of Rif1. Supplemental Figure 2-10: Quantification of Rif1 signal intensity at replication forks in the presence and absence of SUUR.
Identification of replication fork-associated proteins in Drosophila embryos and cultured cells using iPOND coupled to quantitative mass
In recent years, several techniques have been developed to isolate active replication forks to identify proteins associated with the replication fork (Sirbu et al. Another advantage of iPOND is that it can be coupled with quantitative mass spectrometry to identify proteins of bound to the replication fork in an unbiased manner (Wessel et al.
Establishing iPOND in the developing embryo
To identify proteins at or near active replication forks in Drosophila and to determine whether replication fork composition is affected developmentally, we performed iPOND in combination with quantitative tandem mass spectrometry (TMT) in Drosophila post-MZT embryos and cultured cells. Using an iPOND-TMT approach, together with a stringent statistical analysis, we identified 76 and 278 replication fork-associated proteins in post-MZT embryos and cultured Drosophila S2 cells, respectively.
After purification and validation of EdU via Western blot, peptides derived from the pulse and chase samples were TMT-labeled, separated using multidimensional protein identification technology (MudPIT), and quantified by mass spectrometry (Fig. 3-2A). . A schematic of the labeling and mass spectrometry process for iPOND-TMT in Drosophila post-MZT A.
EdU/Thy Embryo
Difference
Biological ProcessC
A schematic of the labeling and mass spectrometry process for iPOND-TMT in Drosophila post-MZT A. B) Volcano plot visualizing proteins identified as enriched or depleted in the pulse versus the hunting embryo samples. Together, we conclude that iPOND can be used in Drosophila embryos to identify existing and potentially new replication fork-associated proteins.
We also identified networks containing proteosome components, RNA processing factors, protein phosphatase 4 complex (PP4), and a number of proteins with no recognized network connections (Figs. 3–3D).
BRWD3 affects genome stability and replication fork progression
Interestingly, knockdown of polybromine, a component of the Brahma chromatin remodeling complex (Thompson 2009), also caused an increase in DNA damage (Fig. 3- 4a). Depletion of BRWD3 in Drosophila S2 cells causes an increase in ɣ-H2Ax levels in unstressed cells (Fig. 3-4a).
Combing after RNAi
To measure solely the rate of fork progression, we performed DNA comb analysis with CldU as the sole nucleotide analog. To rule out an off-target effect of the RNAi construct, we performed the DNA combing assay with an independent RNAi construct (Supp.
Normalized DNA damage signal intensity (γ-H2Av/DAPI)
GFP DNApol
By coupling iPOND to quantitative mass spectrometry, we identified 76 replication fork-associated proteins in Drosophila post-MZT embryos. In summary, we have developed a protocol for the biochemical isolation of replication fork-associated proteins in Drosophila embryos and cultured cells.
Materials and Methods EdU pulsing of embryos
Importantly, we have provided a resource of replication fork-associated factors in Drosophila for those interested in DNA replication, DNA repair, and chromatin dynamics during replication. A small aliquot of each batch of embryos was taken and biotinylated and incubated with 568-Streptavidin to ensure that at least 50% of the embryos were labeled.
EdU pulsing of S2 cells
After fixation, the bottom layer of PFA was removed and an equal volume of methanol was added. Embryos were washed twice in methanol and transferred to PBS + 0.1% Triton X-100 and permeabilized overnight at 4 o C.
After running the gel, samples were transferred onto 0.2 M PVDF using the Transblot Turbo System (BioRad). Membranes were blocked in 5% milk, and incubated with the appropriate antibody for 1 hour at room temperature.
TMT Labeling
The mixed sample was reduced to 1/6 of the original volume using a SpeedVac, and brought back to original volume with Buffer A (5% acetonitrile, 0.1% formic acid). The samples were centrifuged at 14,000 rpm for 30 minutes and the supernatant was transferred to a fresh tube and stored at -80 o C until mass spectrometry analysis.
Liquid Chromatography – Tandem Mass Spectrometry
Using Q-Exactive HF, data-dependent mass spectrum acquisition was performed by performing a full scan from 300–1800 m/z at a resolution of 60,000. Using Exploris480, acquisition of data-dependent mass spectra was performed by performing a full scan from 400–1600 m/z at a resolution of 120,000.
RNAi and immunofluorescence in S2 cells
For each biological replicate, all samples were taken at the same magnification and exposure time. For quantitative analysis of γ-H2Av levels, regions of interest (ROIs) were defined based on the DAPI signal.
DNA molecular combing
Signal was normalized to DAPI signal intensity to account for differences in total DNA amount. One-way Kruskal-Wallis analysis of variance was performed in GraphPad Prism for statistical significance.
Acknowledgments
T.W was supported by the Vanderbilt Chemistry-Biology Interface Training Program (T32GM065086) and the National Science Foundation Graduate Research Fellowship
Author Contributions
Data Availability Statement
Additional Information
PCNA is a marker of active replication forks and is enriched in the pulse sample purifications. The normalized ratio is target/tubulin in the non-targeting GFP control divided by target/tubulin in the RNAi-treated cells (B) Cell proliferation after five days of RNAi depletion relative to the GFP-non-targeting control.
RNAi efficiency
Supplemental Figure 3-2: Validation of RNAi-based Target Depletion. A) Normalized depletion efficiency for two biological replicates.
Normalized RNA level (target/GFP) Normalized cell number (target/GFP)
Cell proliferation after RNAi treatment
The normalized ratio is target/tubulin in the non-targeting GFP control divided by target/tubulin in the RNAi-treated cells (B). Cell proliferation after five days of RNAi depletion relative to the GFP non-targeting control. hd ELG1 RTEL Cul4 DNApol.
BR WD3-dsRNA-1
BR WD3-dsRNA-2
BRWD3
Histone H3
BR WD3 dsRNA-2
GFP-dsRNA-RP TAATACGACTCACTATAGGGGTGAGTTATAGTTGTATTC GFP-dsRNA-FP TAATACGACTCACTATAGGGGGAGAAACTTTTCACTGG BRWD3-dsRNA1-FP TAATACGACTCACTATAGGGAAACGACTACCCAGGACATT BRWD3-dsRNA1-CACTTACTAGTACGGCTGGCT. BRWD3-dsRNA2-FP TAA TAC GAC TCA CTA TAG GAT GGA AAC TAG ACAACC CAG TTC BRWD3-dsRNA2-RP TAA TAC GAC TCA CTA TAG GAT GTC TGT AAT CTC GGA TGA GG Hd-dsRNA-FP TAATACGACTCACTATAGCAGGCTCGCAG.
R-loop mapping and characterization during Drosophila embryogenesis reveals developmental plasticity in R-loop signatures
- loops are a three-stranded nucleic acid structure canonically formed when nascent RNA from transcription reanneals to the template DNA strand, resulting in a
- loop abundance is developmentally regulated and loop homeostasis is necessary for development
- loop position and properties are influenced during development
- loop enrichment at transcription units changes during development
In addition, we were able to demonstrate changes in the localization of R-loops in gene bodies and the role that AT and GC distortion play in the formation of the Drosophila R-loop. While total GC content does not differ in R-loop positive or negative genes, GC and AT skewing has been shown to be a contributing factor to R-loop formation (Ginno and Chédin et al. 2012).
Common and cell-type specific chromatin features associated with R-loops
We asked which marks are consistently associated with R-loops (positively or negatively) across development and which factors are developmentally specific. We found that the repressive mark H3K27me3 was positively associated with R-loops in all developmental samples, highlighting the association between R-loops and transcriptional repression (Figure 4-4B).
In S2 cells, R-loop-containing genes were slightly overrepresented in the highest expression quartile and, to a lesser extent, in the lowest expression quartile (Figure 4-5B). For this purpose, we selected R-loops in the highest quartile of expression and the lowest quartile of expression from S2 cells (Figure 4-5B).
DNA sequence-specific biases are associated with R-loop formation (Ginno and Chédin et al. 2012; Stolz and Chédin et al. 2019). Furthermore, R-loops can modulate DNA methylation at CpG islands in promoter regions (Ginno and Chédin et al. 2012).
Materials and Methods S9.6 antibody
RNase H1 overexpression
Hatch rate assay
Cell culture
Embryo collection and staging
For this, embryos were fixed in heptane and 2% paraformaldehyde for 20 min with shaking, devitellinized in methanol, washed with methanol and rehydrated in PBS + 0.1%.
Genomic DNA purification and RNase treatment
Slot blot
DRIP-qPCR and ssDRIP-seq
Nucleic acid in the eluate was purified with phenol:chloroform, precipitated and resuspended in 10mM Tris. For DRIP-qPCR, 1μL nucleic acid was diluted 1:10 in water and mixed with 10μL SSoAdvanced Universal Sybr (Bio-Rad).
Alignment and peak calling
MarkDuplicates v2.17.10 and stranded bam files were generated using samtools as described in Xu and Sun et al. Stranded bam files were used to generate ssDRIP peaks with call peaks from MAC2 v2.1.2 (Zhang and Liu et al. 2008).
Stranded measurements were visualized using deep tools bamCoverage with --binSize 50bp, --ignoreForNormalization chrY chrM and -- normalizeUsing RPKM (Ramírez and Manke et al. 2014). GRO-seq FPKM counts were determined with HOMER analyzeRepeats.pl using S2 datasets from Core and Lis et al.
Functional genomic data from modENCODE
Gene Ontology enrichment analysis of R-loop-containing genes was performed using the PANTHER program, with Fisher's exact test and using the Bonferroni correction for multiple testing (Ashburger et al 2000; Consortium et al 2020; Mi et al 2019).
S2 cells
Chromatin associated factor enrichment in R-loops
For the overall analyses, we retained the location of the R-loop peaks and shuffled the locations of the histone modification or transcription factor binding peaks. For a secondary analysis, we examined a subset of R-loops specifically quantified in the TTS and 3' UTR.
Calculation of AT- and GC-skew in R-loops
