What Is The Difference Between Synthesis Dependent Strand Annealing And Double Strand Break Repair?

Abstract

DNA double-strand breaks (DSBs) are one of the most deleterious types of lesions to the genome. Synthesis-dependent strand annealing (SDSA) is thought to be a major pathway of DSB repair, just directly tests of this model accept only been conducted in budding yeast and Drosophila. To better understand this pathway, we developed an SDSA analysis for use in human cells. Our results support the hypothesis that SDSA is an of import DSB repair mechanism in human cells. Nosotros used siRNA knockdown to assess the roles of a number of helicases suggested to promote SDSA. None of the helicase knockdowns reduced SDSA, merely knocking down BLM or RTEL1 increased SDSA. Molecular analysis of repair products suggests that these helicases may prevent long-tract repair synthesis. Since the major alternative to SDSA (repair involving a double-Holliday junction intermediate) tin lead to crossovers, we also adult a fluorescent assay that detects crossovers generated during DSB repair. Together, these assays volition be useful in investigating features and mechanisms of SDSA and crossover pathways in homo cells.

Keywords: double-strand intermission repair, crossing over, synthesis-dependent strand annealing

Double-strand breaks (DSBs) are considered to be one of the most detrimental types of Deoxyribonucleic acid damage. There are numerous mechanisms for repairing DSBs, broadly classified into cease joining and homology-directed recombination (HDR). Among the latter, the double-strand break repair (DSBR) (Figure 1) model has been popular since it was proposed >xxx yr agone (Szostak et al. 1983). A hallmark of this model is the double-Holliday junction (dHJ) intermediate, which has two of the four-stranded junctions originally hypothesized by Holliday (1964). In DSBR, as in Holliday's model, specialized nucleases resolve Holliday junctions (HJs) by introducing symmetric nicks; contained resolution of the two HJs results in 50% of repair events having a reciprocal crossover. It has as well been proposed that dHJs tin can be processed without the activeness of a nuclease if a helicase and topoisomerase migrate the two HJs toward 1 another and then decatenate the remaining link (Figure i) (Thaler et al. 1987); this process has been called dissolution to distinguish it from endonucleolytic resolution (Wu and Hickson 2003).

Models of DSB repair past homologous recombination. (A) Blue lines stand for 2 strands of a DNA duplex that has experienced a DSB. HDR begins with resection to expose unmarried-stranded DNA with 3′ ends (arrows). Ane of these can undergo strand invasion into a homologous duplex (red) to generate a D-loop; the 3′ invading stop is then extended past synthesis. (B) In SDSA, the nascent strand is dissociated and anneals to the other resected finish of the DSB. Completion of SDSA may result in noncrossover gene conversion (cherry-red patch, shown after repair of any mismatches). (C) An alternative to SDSA is annealing of the strand displaced past synthesis to the other resected end of the DSB. Additional synthesis tin can lead a dHJ intermediate. (D) In DSBR, the dHJ is resolved by cutting to generate either crossover or noncrossover products (one of two possible outcomes for each instance is shown). (E) The dHJ can too be dissolved by a helicase-topoisomerase complex to generate noncrossover products.

In studies of Dna DSB repair resulting from transposable element excision in Drosophila, Nassif et al. (1994) noted that crossovers were exceptional and the 2 ends of a single DSB could use unlike repair templates. To explicate these results, they proposed the synthesis-dependent strand annealing (SDSA) model (Figure 1). In addition to connected use of Drosophila gap-repair assays (e.g., Kurkulos et al. 1994; Adams et al. 2003), other types of evidence have been interpreted equally support for the SDSA model. In Saccharomyces cerevisiae meiotic recombination, gel-based separation and quantification of intermediates and products showed that noncrossovers are made before dHJs appear, suggesting that these noncrossovers are generated by SDSA (Allers and Lichten 2001). In vegetatively growing Due south. cerevisiae, Mitchel et al. (2010) studied repair of a small gap DSB in cells defective in mismatch repair. Based on the loftier frequency with which heteroduplex Deoxyribonucleic acid tracts (regions that contain one template strand and i recipient strand) in noncrossover products were restricted to one side of the DSB, they ended that almost noncrossover repair occurred through SDSA. Miura et al. (2012) used an Due south. cerevisiae assay designed specifically to detect SDSA. A plasmid with a DSB was introduced into cells in which templates homologous to the two sides of the DSB were on different chromosomes, eliminating the possibility of a dHJ intermediate. Based on results of these diverse assays, many researchers now believe SDSA to be the most mutual mechanism of mitotic DSB repair past HDR (reviewed in Andersen and Sekelsky 2010; Verma and Greenberg 2016).

In mammalian cells, the directly-repeat GFP (DR-GFP) assay (Pierce et al. 1999) has been an instrumental tool for studying DSB repair past HDR. In this analysis, an upstream GFP gene (SceGFP) is disrupted past insertion of an I-SceI site, and a downstream GFP fragment (iGFP) serves as a template for repair. Gene conversion replaces the region surrounding the I-SceI site in SceGFP, generating an intact GFP gene. This gene conversion has been suggested to ascend through SDSA, but it is not possible to distinguish between SDSA and other noncrossover DSB repair in this assay (run into Figure 1). Xu et al. (2012) adult a novel human cell assay in which gene conversion could be detected simultaneously at the DSB site and at another site ∼1 kbp abroad. They establish that the two were frequently independent and ended that SDSA is a major mechanism for DSB repair in human cells, just they too could non exclude DSBR as a possible source.

Development of the CRISPR/Cas9 system for genome engineering science (Cong et al. 2013; Republic of mali et al. 2013) provides additional accent on the importance of understanding SDSA mechanisms in human cells, as it has been suggested that replacement of multi-kilobase pair fragments after Cas9 cleavage, and probably other HDR events, occurs through SDSA (Byrne et al. 2015). We therefore designed an assay to detect DSB repair by SDSA in human cells. Here, we draw this assay and evidence that, as hypothesized, SDSA appears to be an important pathway for HDR in human cells. Nosotros report the effects of knocking down diverse proteins proposed to part during SDSA. We also describe a fluorescence-based assay for detecting crossovers generated during DSB repair. Use of these assays should help to further our understanding of DSB repair pathways used in human cells.

Materials and Methods

Structure of analysis plasmids

The SDSA assay construct, pGZ-DSB-SDSA, was based on pEF1α-mCherry-C1 vector (itemize no. 631972; Clonetech). A fragment of mCherry was removed by cutting with NheI and HindIII and inserting annealed oligonucleotides containing an I-SceI recognition sequence and a office of the mCherry sequence. The product, pEF1a-mCherry-I, had 350 bp of mCherry deleted and replaced with an I-SceI recognition sequence. In parallel, 5′ and iii′ mCherry fragments, overlapping by 350 bp, were PCR-amplified and cloned into a vector containing a fragment of the copia retrotransposon from D. melanogaster. A fragment of HPRT was cloned out of the DR-GFP construct. This entire module (5′ mCherry–copia–3′ mCherry–HPRT) was PCR-amplified and cloned into the pEF1a-mCherry-I to produce pGZ-DSB-SDSA. The full sequence was deposited in GenBank under accretion {"type":"entrez-nucleotide","attrs":{"text":"KY447299","term_id":"1148299218","term_text":"KY447299"}}KY447299.

The crossover assay construct, pGZ-DSB-CO, was based on pHPRT-DRGFP (Pierce et al. 2001) and the intron-containing mCherry gene from pDN-D2irC6kwh (Nevozhay et al. 2013). The iGFP fragment was expanded to include the entire 3′ end of the transcribed region, and this was cloned into the mCherry intron of pDN-D2irC6kwh. This module (mCherry with an intron containing three′ GFP) was cloned into the pHPRT-DRGFP vector cutting with HindIII, so that it replaced the iGFP fragment and was in reverse orientation relative to the SceGFP gene. The total sequence of pGZ-DSB-CO was deposited in GenBank nether accession {"type":"entrez-nucleotide","attrs":{"text":"KY447298","term_id":"1148299215","term_text":"KY447298"}}KY447298.

Generation of stably-transfected jail cell lines

U2OS and HeLa cells were cultured under normal conditions (DMEM + ten% FBS + pen/strep) for 24 hr until they reached eighty% confluency before transfection with either SDSA or crossover assay constructs using a Nucleofector 2b Device (catalog no. AAB-1001; Lonza) and Cell Line Nucleofector Kit V (itemize no. VCA-1003; Lonza). At ane wk postal service-transfection, appropriate antibiotics were added to select for the cells with a stably-integrated construct. pGZ-DSB-SDSA analysis has a factor for neomycin resistance; cells receiving this construct were treated with 700 μg/ml G418 (catalog no. A1720; Sigma) for 1 wk then single-cell clones were derived. pGZ-DSB-CO contains a PGK1 gene that confers resistance to puromycin; cells receiving this construct were treated with 10 μg/ml puromycin (itemize no. P8833; Sigma) for 1 wk then single-prison cell clones were derived. Initial attempts to determine re-create number by Southern absorb were unsuccessful; still, the analyses described below strongly suggested that the lines we characterized each carried a single insertion or possibly a single tandem assortment.

Dna repair assays and flow cytometry

U2OS cells with pGZ-DSB-SDSA integrated were cultured in ten-cm dishes containing 10 ml of DMEM medium with high glucose (Corning) until divide onto half-dozen-well plates at a concentration of five × 10⁴ cells/ml using 0.05% trypsin 0.53 mM EDTA solution (Corning). Upon reaching ∼60% confluency, the cells were treated with an siRNA reaction mixture (xc nmol siRNA and 8 μl lipofectamine 2000 reagent per well; Invitrogen). At 24 hr later transfection, the siRNA reaction mixture was replaced with the fresh culture medium. After 12 hr the cells were split so that knockdown could assessed in one half (see qPCR evaluation of the siRNA knockdown efficiency). The other half was treated with 100 μl I-SceI–expressing adenovirus (Anglana and Bacchetti 1999) (previously titrated to a nonlethal concentration). Subsequently some other 24 hr, the medium was replaced and thus the adenovirus removed. After some other 72 hr, the cells were harvested and resuspended in one× PBS (Corning) supplemented with 2% fetal bovine serum (FBS) and five mM EDTA, for flow cytometry acquisition on a BD LSRFortessa, using 488 and 561 nm lasers to observe the mCherry fluorescence.

U2OS cells with pGZ-DSB-CO integrated were cultured and treated under the aforementioned conditions. Flow cytometry acquisition was conducted on a BD FACSAriaII, using 388 and 532 nm lasers to detect GFP and mCherry fluorescence.

U2OS genomic Dna isolation

Cells were cultured in a 15-cm dish until they reached 100% confluency, then rinsed with 1× PBS and harvested in 0.05% trypsin, 0.53 mM EDTA, by centrifuging for 3 min at 2000 rpm. Cells were washed with PBS and transferred to one.5-ml microfuge tubes and spun for 10 sec to repellet. PBS was removed and cells were resuspended in TSM (10 mM Tris-HCl, pH 7.four; 140 mM NaCl; i.v mM MgCl_ii) with 0.5% NP-40 and incubated on ice for 2–3 min. Later on pelleting, cells were resuspended in 1 ml nuclei dropping buffer (0.075 M NaCl; 0.024 K EDTA, pH eight.0). The intermission was transferred to a 15-ml tube containing 4 ml nuclei dropping buffer with ane mg Proteinase K (last Proteinase Thou concentration = 0.2 mg/ml), and 0.v% SDS. The cells were lysed overnight at 37°. The side by side day, an equal volume of phenol was added and mixed on an orbital shaker for 2 hr followed by a 5-min spin at 2000 rpm. The aqueous phase was transferred to a clean tube, an equal volume of chloroform was added, and the mix was incubated for 30 min on an orbital shaker. After spinning, the aqueous stage was transferred to a new tube and 0.1 vol three M NaOAc was added, followed by 1 vol isopropanol. The Deoxyribonucleic acid was spooled out using a glass Pasteur pipette and resuspended overnight in 1 ml TE buffer (x mM Tris-HCl, pH viii.0; 1 mM EDTA). The side by side mean solar day, the DNA was precipitated using 0.v vol 7.v M NH_fourOAc and 2 vol ethanol. DNA was spooled out and resuspended in 0.5 ml TE-4 buffer (ten mM Tris-HCl, pH 8.0; 0.1 mM EDTA). Samples were stored in iv° until analyzed.

PCR analysis of the repair events

DNA from the BRCA2 knockdown repair events was isolated according to the protocol described in U2OS genomic DNA isolation, and used in a PCR reaction to amplify a desired DNA fragment for sequencing or fragment length characterization. A total of 1.v μl Dna was added to each PCR mixture containing primer sets according to Supplemental Textile, Table S1 in File S1, iProof polymerase (catalog no. 424264; BioRad) and buffer. PCR amplification reaction program was 33 cycles of the post-obit: twenty sec at 98°, 20 sec at 64°, and 20–150 sec at 72°. Products were run on 1–1.five% agarose gels with ethidium bromide earlier being imaged.

Western absorb of BLM protein in siRNA-treated cells

Cells treated with siRNA as described in DNA repair assays and menses cytometry were harvested on the 3rd solar day post-transfection using 0.05% Trypsin, 0.53% mM EDTA solution (Corning). Later washing with 1× PBS, the cells were resuspended in a protein sample buffer (Tris-HCl; SDS; glycerol; bromophenol blue; 150 mM DTT) and boiled. A total of 20 μl of the protein sample was loaded on a 7.5% SDS-PAGE gel and the gel was run for 1–2 hr at 100 Five. Poly peptide was transferred to a PVDF membrane using a moisture transfer method (1.five hr at xc 5 in iv°). The membrane was blocked in PBS with 5% powdered milk and incubated in PBS plus 0.i% Triton-X plus main antibodies [rabbit anti-BLM (itemize no. 2179; Abcam) at 1:2000 and mouse anti-αTubulin (catalog no. T9026; Sigma) at 1:8000] overnight at 4° on a rocker. The membrane was and so washed iii times in PBS-T solution. HDRP-conjugated secondary antibodies were added (goat anti-rabbit at 1:5000 and caprine animal anti-mouse at 1:100,000) and the blot was incubated for 1 hr at room temperature. The membrane was done three times in PBS-T solution and then incubated in an ECL solution (Thermo Fisher Scientific) for chemiluminescence for ii min. The Western absorb epitome was taken using a BioRad Molecular Imager (ChemiDoc XRS+) or the X-ray motion-picture show was developed using a developer.

qPCR evaluation of the siRNA knockdown efficiency

Cells treated with siRNA every bit described in DNA repair assays and menstruum cytometry were harvested on the third solar day mail-transfection using 0.05% Trypsin 0.53% mM EDTA solution (Corning). RNA was extracted using the manufacturer's protocol for ReliaPrep RNA Cell Miniprep Organization (Promega). Purified RNA was used as a template to generate the cDNA library with QuantiTect Reverse Transcription Kit (itemize no. 205310; Qiagen). The qPCR mix independent cistron-specific DNA primers, cDNA, and the QuantiTect SYBR Dark-green PCR kit (catalog no. A 204141; Qiagen). Amplification and quantification was conducted on a RealTime PCR machine (QuantStudio 6 flex Real Time PCR System).

Statistical analysis

Statistical comparisons were performed on the raw data (Tables S2 and S3 in File S1) using GraphPad Prism version 6.07 for Windows (GraphPad Software Inc., La Jolla, CA). In the case of BLM knockdown in the SDSA assay, one value (271% of control) was found to be a significant outlier based on the Grubb's test using GraphPad QuickCalcs online (https://graphpad.com/quickcalcs/grubbs2/), and was excluded from farther analysis.

Results and Discussion

An SDSA analysis for human cells

To study SDSA in human cells, we used an approach conceptually similar to the P{w^a } assay used in Drosophila (Adams et al. 2003; McVey et al. 2004). In this analysis, if both ends of the DSB generated past P chemical element excision are extended past synthesis from the sis chromatid, the nascent strands can anneal at repeats inside the P{w^a } chemical element, generating a product that is unique to SDSA and hands distinguishable by phenotype. To mimic this situation in human being cells, we built a construct (Figure 2) that has an mCherry gene in which a 350-bp segment was replaced with the 18 bp I-SceI recognition sequence, rendering the factor nonfunctional. When a DSB is generated past I-SceI (Effigy 2A), HDR tin exist completed using a downstream repair template. The repair template is split up: each half has 800 bp of homology adjacent to the intermission site plus the 350 bp of deleted mCherry sequence. The two halves are separated by a 3-kbp spacer of unique sequence. Since the 350-bp sequence is on both sides of the spacer, information technology constitutes a direct repeat. We hypothesized that both ends of the I-SceI–induced break will invade the side of the template to which they are homologous, either simultaneously or sequentially (Figure 2B). If synthesis on both sides extends through or far enough into the 350-bp repeat earlier the nascent strands are dissociated from the template, the overlapping regions can anneal to one another (Figure 2B). Completion of SDSA restores a functional mCherry gene at the upstream location.

An SDSA assay for human cells. (A) Schematic of the analysis construct. The mCherry coding sequence is represented with a large arrow, with the xanthous box indicating the site at which a 350-bp fragment was removed and replaced with an I-SceI site. The thick red lines designate the promoter and 5′ and 3′ untranslated regions. Downstream of mCherry is the neo selectable marker and vector sequences (black arrow and line, respectively). The repair template consists of 800 bp of homology to the left of the DSB site, then the 350-bp deleted fragment (magenta), a three-kbp spacer (bluish), another copy of the 350-bp fragment, and 800 bp of homology to the right end of the DSB site. The lines below are a representation of the same sequences as duplex Deoxyribonucleic acid, for use in subsequent panels. Arrowheads point 3′ ends at the outer edges of the construct in all panels. (B) Two-ended SDSA repair. The DSB is shown as resected and paired with a template. This diagram shows interchromatid repair, where the template on the sister chromatid is used (whether or non it was cut by I-SceI). Intrachromatid repair may also be possible; in this case, the black lines to the right of the DSB would exist continuous with those to the left of the template. After strand substitution, synthesis into the 350-bp regions, and dissociation, complementary sequences tin can anneal. In the example shown here, the left end has been extended through the 350-bp fragment into the spacer; the right end synthesized but midway into the 350-bp fragment. After trimming of the spacer sequence, filling of gaps, and ligation, a restored mCherry is produced, resulting in crimson fluorescence. (C) One-ended sequential SDSA. This is similar to (B), except only the left terminate of the DSB has paired with the template. After synthesis and dissociation, the nascent strand can pair with the 2nd half of the template. Additional synthesis can extend the nascent strand to provide complementarity to the other end of the DSB, assuasive annealing and completion of SDSA. (D) If one or both ends invade the template and synthesis traverses the entire template, a dHJ tin form (height). If the dHJ is dissolved or resolved in the noncrossover orientation, the product shown in the middle is generated. The bottom function of this panel shows the chromosomal product of crossover resolution (in that location is also an acentric extrachromosomal circle that will be lost). (E) A product produced by initiation of repair by SDSA merely completion past end joining. In this case, role of the spacer has been copied into the upstream mCherry (sequences from the template are indicated with green lines).

The scenario higher up requires two-ended SDSA, but sequential one-ended SDSA is also possible (Figure 2C). If but one end of the pause invades the downstream template, is extended past repair synthesis, and is then dissociated from the template, the nascent strand will not be complementary to the other resected end; however, this nascent strand will take homology to the repeat on the other side of the template. A 2nd wheel of strand exchange and repair synthesis using the other echo could lead to add-on of sequences complementary to the other resected DSB stop. This would also restore a functional mCherry gene by SDSA.

A functional mCherry gene might also be generated through a combination of SDSA and DSBR. In the sequential SDSA scenario, the 2nd strand commutation upshot could be processed into a dHJ. The product of dissolution or noncrossover resolution of such a dHJ will be identical to that of SDSA (Figure 2B), but if the dHJ is resolved as a crossover, generation of a functional mCherry gene will be accompanied by a deletion of all sequences between the upstream mCherry and the downstream template. Dissolution or noncrossover dHJ resolution in this scenario cannot be distinguished from SDSA, just it should be noted that germination of such a dHJ intermediate still requires at least one cycle of D-loop disassembly—a cardinal step that separates SDSA from DSBR (Effigy one).

Other types of repair that practice not generate a functional mCherry are possible. A dHJ can be generated if synthesis extends through one mCherry 350-bp repeat, the entire spacer, and the other 350-bp repeat (Figure second). Processing of this dHJ would give a product in which the entire template, including the duplicated 350-bp sequences and the spacer, was copied into the upstream mCherry gene. Dissolution or noncrossover resolution would result in two copies of the template (Effigy 2d, middle), whereas crossover resolution would delete intervening sequences (Figure 2D, bottom). Nonhomologous end joining (NHEJ) tin can restore or disrupt the I-SceI recognition sequence, depending on whether information technology is precise or imprecise (not depicted). Hybrid repair, in which repair is initiated by HDR but completed past stop joining instead of annealing, can give rise to an mCherry in which the 350-bp gap is not completely filled or, if synthesis extends into the spacer, in which role of the spacer is copied into the upstream mCherry gene (Figure 2E).

To generate cell lines with the SDSA repair construct, we transfected both U2OS and HeLa cells with linearized SDSA construct and used G418 to select stably-transfected lines. To induce DSBs, we infected cells with an adenovirus expressing I-SceI (Anglana and Bacchetti 1999). We detected mCherry activity by fluorescence microscopy (Figure 3A and Effigy S2A in File S1) as early on every bit 2 d after viral infection. Stable expression persisted through months of cell culturing. Molecular assay of genomic Deoxyribonucleic acid from clones derived from single mCherry-positive cells confirmed the absenteeism of the I-SceI, restoration of a complete mCherry cistron, and the presence of an intact repair template (Figure S1 in File S1). We quantified SDSA repair through flow cytometry (Figure 3B and Tabular array S2 in File S1).

Results from the SDSA assay. (A) Cells fluorescing blood-red due to mCherry expression subsequently I-SceI infection. (B) Flow cytometry of cells later I-SceI expression. In this instance, ten,000 singlet cells were assayed and 200 were gated as exhibiting red fluorescence. (C) Furnishings of siRNA knockdown on acquisition of mCherry expression (see Materials and Methods). Fluorescence frequencies from period cytometry were normalized to NT; raw data are given in Table S2 in File S1. Mistake bars indicate SD. The ratio paired t-test was used to compare raw values for each siRNA target to its NT control and each double-knockdown to the single knockdown of the corresponding helicase. For helicase single knockdowns, P values were Bonferroni-corrected for multiple (half dozen) comparisons. * P < 0.05, ** P < 0.01, **** P < 0.0001.

We obtained mCherry-positive cells in multiple isolates of each cell type, and nosotros selected one U2OS prison cell isolate for farther label. Since nosotros intended to apply siRNA to knock down proteins that might be required for SDSA, we get-go knocked downwardly BRCA2 as a positive command. BRCA2 is essential for RAD51-mediated strand exchange and for initiation of HDR (Sharan et al. 1997). Consequent with this function, knocking downwardly BRCA2 resulted in a significant decrease in red-fluorescing cells relative to the nontargeting (NT) siRNA-negative command (Figure 3C). Although there was substantial residuum HDR, probably due to incomplete loss of BRCA2 (Figure S3 in File S1), we conclude that SDSA does occur in human cells and that our assay can be used to study this process.

An assay for detecting crossovers generated during DSB repair in homo cells

D-loop dissociation is a critical footstep in the SDSA pathway. If the D-loop is non dissociated, 2d-terminate capture may lead to formation of a dHJ intermediate (Effigy 1C). Information technology is thought that most dHJs formed in proliferating cells are candy past dissolution to give noncrossover products, but resolution by nicking can generate crossovers (Effigy 1D). To complement our analysis for SDSA, we adult an assay that detects crossovers generated later DSB formation (Figure 4A). Our analysis is based on the DR-GFP gene conversion assay (Pierce et al. 2001), which has a GFP gene interrupted by an I-SceI site (SceGFP), and a downstream GFP fragment (iGFP) that serves as a repair template. We added an mCherry gene with an intron and placed the iGFP fragment (modified to contain the entire iii′ end of GFP) inside the intron. This is in reverse orientation relative to SceGFP, to forbid the possibility of the single-strand annealing pathway. In cells transfected with this construct, mCherry is expressed but GFP is non. Afterwards DSB induction with I-SceI, factor conversion of the I-SceI site restores GFP expression. Even so, if cistron conversion is accompanied past a crossover, the region between the GFP fragments becomes inverted, resulting in loss of mCherry expression.

An assay to detect crossovers generated during DSB repair. (A) Schematic of the assay. The diagram at the top (i) has the SceGFP factor in solid green, with the yellowish box indicating insertion of an I-SceI site. The modified iGFP fragment (with the entire iii′ finish of the cistron) is indicated in hatched green. The mCherry gene is shown in magenta, with the intron represented equally a dotted line. (ii) Representation of the construct every bit double-stranded Dna. (3) After DSB introduction and resection, resected ends tin pair with the homologous iGFP template. The example here shows intrachromatid repair. (four) Gene conversion without a crossover results in replacement of the I-SceI site with GFP sequences, generating a functional GFP cistron. mCherry expression is unaffected. (five) If a crossover is generated, the cardinal region is inverted and mCherry expression is lost. Repair with the sister chromatid (interchromatid repair) may also exist possible. In that case, a crossover results in a dicentric chromatid and an acentric chromatid; these products will non be recovered, but noncrossover gene conversion will exist recovered. (B) A field of cells afterward I-SceI expression, showing the green (left) and crimson (middle, magenta) channels, and a merged image (right). Greenish arrowhead marks a cell that gained GFP expression but lost mCherry expression (a crossover); white arrow indicates a jail cell that gained GFP expression and retained mCherry expression (noncrossover factor conversion); magenta arrow indicates a cell in which GFP was non converted. (C) Menses cytometry showing gating for GFP and mCherry fluorescence. (D) Increased noncrossover and crossover gene conversion later on knocking down BLM. Bars bespeak percentage of cells expressing GFP and mCherry (noncrossover gene conversion; upper right quadrant of flow cytometry output) or GFP only (crossovers; lower right quadrant). Error confined are SD based on three different wells per handling. Raw data are in Table S3 in File S1. *** P = 0.0010, **** P < 0.0001 based on unpaired t-test.

Nosotros introduced this construct into both U2OS and HeLa cells and selected for stably-transfected lines that expressed mCherry. Nosotros tested several U2OS lines for response to I-SceI introduction. We selected a line that consistently yielded cells that were positive for both GFP and mCherry (noncrossover gene conversions) and cells that were positive for GFP simply (crossovers). In our standard protocol (come across Materials and Methods), treatment of NT siRNA control cells with I-SceI resulted in 15.6% of cells gaining GFP expression (Effigy 4, B and C and Table S3 in File S1); 3.eight% of these (0.six% of all cells) had lost mCherry expression. This is likely to exist an underestimate of the crossover frequency because only intrachromatid crossovers can be recovered, as interchromatid crossovers result in dicentric and acentric products (come across Figure 4 legend).

Knocking down BLM or RTEL1 elevates SDSA frequency

In the model shown in Figure one, SDSA diverges from dHJ pathways when a helicase dissociates the nascent strand from the template. Several helicases have been suggested to perform this stride. Drosophila gap repair assays institute roles for Blm and Fancm helicases in SDSA (Adams et al. 2003; Kuo et al. 2014). The Arabidopsis orthologs of these enzymes promote noncrossover recombination in meiosis, possibly by SDSA (Crismani et al. 2012; Séguéla-Arnaud et al. 2015). Sgs1, the yeast ortholog of Blm, is required for noncrossover recombination in budding yeast meiosis (De Muyt et al. 2012), and Sgs1 and Mph1 (the orthologs of Fancm) take been implicated in SDSA in vegetative cells (Mitchel et al. 2013). In Schizosaccharomyces pombe, Fml1, the ortholog of Fancm/Mph1, was suggested to promote SDSA in DSB repair during replication and in meiosis (Sun et al. 2008; Lorenz et al. 2012). Yet another helicase, RTEL-1, was hypothesized to disrupt D-loops in Caenorhabditis elegans meiosis (Barber et al. 2008; Youds et al. 2010). In human being cells, RECQ5 is proposed to promote SDSA (Paliwal et al. 2014). Aside from the experiments in Drosophila and budding yeast, none of the assays performed could distinguish between SDSA and other pathways. Thus, we used our assay to ask whether the orthologs of any of these or related helicases affect SDSA in human being cells.

We did not find any modify in the frequency of ruddy-fluorescing cells after knocking downwards FANCM, RECQ5, WRN, or FBXO18 (Figure 3C and Tabular array S2 in File S1). Knockdown of BLM or RTEL1 significantly altered the frequency of red-fluorescing cells; however, instead of decreasing SDSA as expected, both knockdowns resulted in an increase in red-fluorescing cells (Effigy 3C). BLM has been shown to accept several functions in HDR pathways, including in DSB end resection in a pathway redundant with ExoI (Zhu et al. 2008) and dHJ dissolution (Wu and Hickson 2003; Wu et al. 2006; Singh et al. 2008). Also, RTEL1, which was initially identified equally a telomere length regulator and is responsible for T-loop disruption (Ding et al. 2004; Sarek et al. 2015), can besides unwind Dna secondary structures and promote replication fork progression (Barber et al. 2008). To brainstorm to assess how knocking downward BLM and RTEL1 might increment SDSA, we knocked down each in combination with BRCA2. In simultaneous knockdowns, red-fluorescing cells were decreased from the frequency observed of BLM and RTEL1 single knockdowns (Figure 2C). The magnitude of the decrease was similar to that of the BRCA2 unmarried knockdown relative to the negative command (44% decrease for BRCA2 relative to NT, 47% for BRCA2 + BLM relative to BLM solitary, and 54% for BRCA2 + RTEL1 relative to RTEL1 lone), indicating that BLM and RTEL1 touch on SDSA through functions after strand exchange into a homologous template.

We also tested the effects of knocking down BLM in the crossover assay. Transfection with siRNA to knockdown BLM resulted in 25% of cells gaining GFP expression (Effigy 3C). This is a sixty% increase compared to the NT control, similar to the average increment of 59% in the SDSA assay. Crossovers were elevated, with 6.4% of the GFP-positive cells (1.half dozen% of all cells) having as well lost mCherry expression. This is consequent with studies showing elevated spontaneous crossing over in cells from Blossom syndrome patients (German et al. 1977).

Knocking down BLM or RTEL1 alters repair outcomes

To further develop our SDSA assay and proceeds additional insights into the furnishings of knocking down BLM and RTEL, we determined the structures of repair events produced in knockdown cells. Nosotros analyzed 55 clones derived from unmarried red-fluorescing cells, including 23 from the NT control, 21 from BLM knockdown, 10 from FANCM knockdown, and i from RTEL1 knockdown. All but one had the structure expected of SDSA (Figure 2B). The remaining clone, which came from NT siRNA treatment, had lost neo and the template spacer, and therefore may have arisen from SDSA followed by DSBR with crossover resolution (Figure 1D). These results support our decision that cells with restored mCherry utilized SDSA to repair the DSB, perhaps occasionally coupled with use of DSBR.

Nosotros also analyzed cells that failed to produce mCherry. In the NT control, all 45 lines examined appeared to be identical to the initial construct (Effigy 5A). We did not measure cleavage efficiency in our analysis, but we titrated adenovirus infection to the highest dose that did not cause detectable cell lethality. Delivery of I-SceI by adenovirus infection of HEK293 cells resulted in 85% of sites beingness cut (Anglana and Bacchetti 1999), so it is likely that most or all of the cells with intact I-SceI sites are likely to event from cleavage followed by precise NHEJ using the 4 nt complementary overhangs left by I-SceI.

Structures of repair events in cells not expressing mCherry. Single cells non exhibiting red fluorescence were grown and and so analyzed. (A) The starting construct, and the structure found in the bulk of clones in each siRNA treatment group. (B) Copying of the unabridged template into the upstream mCherry I-SceI site. This might occur through long-tract SDSA (one synthesis bicycle or several) or a dHJ intermediate (as in Effigy 2D, center). (C) Copying of the entire template with loss of intervening sequences. This is proposed to occur when a dHJ intermediate is resolved in the crossover orientation, every bit in Figure second, bottom. (D) Copying of role of the template into the upstream mCherry site, as predicted from initiation of SDSA but completion of repair by cease joining, as in Effigy 2E. (Due east) Retention of the I-SceI site with loss of the neo and ori sequences.

The bulk of clones from BLM or RTEL1 knockdown cells that did not produce mCherry also had an intact I-SceI site; however, structures indicating other repair processes were observed in 11 out of 34 of these clones from BLM knockdown (P < 0.0001 compared to NT) and 24 out of 133 of these clones from RTEL1 knockdown cells (P = 0.0007). In four of the BLM knockdown clones and 14 of the RTEL1 knockdown clones, the entire 3-kbp spacer sequence was copied from the repair template into the upstream mCherry (Figure five, B and C). This extensive repair synthesis might occur through multiple cycles of strand exchange, as is believed to occur in Drosophila gap repair by SDSA (McVey et al. 2004), or through a single, continuous synthesis event. Among the 17 examples in which the entire spacer was copied, one from each knockdown sample had lost neo, a construction that is most consistent with a dHJ existence resolved to give a crossover (Figure 2D and Figure 5C). The other 15 may have arisen by long-tract SDSA or by dissolution or noncrossover resolution of a dHJ (Effigy 2D and Effigy 5B).

There were additional repair events from knockdown cells that also had prove of long-tract synthesis. Three events from BLM knockdown and vii from RTEL1 knockdown had a subset of the spacer copied into the upstream mCherry (Figure 2E and Effigy 5D). These are most likely hybrid repair that involved long-tract synthesis followed by terminate joining. There was i upshot from the BLM knockdown that had an intact I-SceI site but had lost neo (Effigy 5E). The source of this event and whether it occurred following I-SceI cleavage is unknown.

Roles of BLM and RTEL1 in SDSA

Information technology was surprising that none of the helicase knockdowns led to decreased SDSA, since orthologs of all of these have been suggested to promote SDSA. Ane possibility is that at that place is redundancy among ii or more of these proteins. This possibility can be addressed through simultaneous knockdowns or use of doubly mutant cells. In the case of BLM and RTEL1, knockdown led to dramatic increases in HDR, both in the frequency of red-fluorescing cells (Effigy 3C) and in the fraction of nonfluorescing cells that had evidence for HDR (nix out of 45 for NT compared to 11 out of 34 for BLM knockdown and 24 out of 133 for RTEL1 knockdown). This is consequent with a reported increase in gene conversion in the DR-GFP assay after siRNA knockdown of BLM (Paliwal et al. 2014).

Molecular assay of repair events provides some insights into sources of increased HDR. Several repair products from both BLM and RTEL1 knockdowns had extensive synthesis (Effigy 5). Extensive synthesis has too been reported in another assay when the HDR proteins BRCA1 or CtIP were knocked down, and it was suggested that these proteins preclude long-tract HDR (Chandramouly et al. 2013). BLM and RTEL1 might prevent long-tract HDR through their D-loop disruption activities (Van Brabant et al. 2000; Barber et al. 2008). Our assay requires enough synthesis from both ends of the DSB to allow annealing inside the 350-bp repeat (Figure 2B), or sequential strand invasion and synthesis (Figure 2C). If BLM and RTEL1 unremarkably disrupt D-loops after less than a few hundred nucleotides of synthesis, the nascent strands would not be able to amalgamate at the 350-bp repeat and we would see reduced SDSA. In this scenario, we might await the NT sample to have some hybrid repair events with short synthesis tracts (as in Figure 2E merely with only part of the 350-bp echo on one or both ends and none of the spacer sequence). We did not notice any such products, but it is possible our sample size was non large enough to find these exceptional events.

Some repair products in knockdown cells had synthesis that spanned the unabridged 3-kbp spacer (e.1000., Figure 5B). These could generate dHJ intermediates that would then exist resolved or dissolved. In the case of RTEL1 knockdown, merely one out of thirteen events in which the unabridged spacer was copied had the structure expected of dHJ crossover resolution (Effigy 5C). If crossover and noncrossover resolution occur at equal frequencies, then either SDSA or dHJ dissolution were probable responsible for almost of these events. The BTR circuitous, which contains the BLM helicase, is believed to be the major or sole dissolvase (Wu and Hickson 2003; Seki et al. 2006; Dayani et al. 2011). Although there was also only a single result in the BLM knockdown suggestive of dHJ crossover resolution (Figure 5C), at that place were only three other events in which the entire spacer was copied; it is possible that all three of these came from dHJ resolution that had a noncrossover consequence.

In our assay, SDSA may occur if both ends of the DSB appoint with the template. This suggests the possibility that a office of BLM and RTEL1 is to ensure that merely one end engages with the repair template. Among the x repair events in which just part of the spacer was copied into the upstream mCherry (Figure 5D), all of them appeared to accept synthesis from the left end but. This is in contrast to the Drosophila P{due west^a } excision assay, where most repair events have several kilobase pairs of synthesis from both ends of the intermission (Adams et al. 2003). The disparity could arise from the divergence between organisms or tissues, dissimilar structures of the DSB ends (4 nt 3′ overhangs for I-SceI; 17 nt three′ overhangs for P element excision), or distance between the sequence homologous to the left side of the DSB and the sequence homologous to the right side (∼3.five kbp for this assay but >14 kbp for P{w^a }).

The cases of fractional copying of the spacer almost likely derive from hybrid repair in which the initial steps of SDSA are executed merely dissociation of the nascent strand does not reveal complementary sequences for annealing, so repair is instead completed by DNA polymerase Θ-mediated end joining (also chosen microhomology-mediated end joining) (Chan et al. 2010; Wyatt et al. 2016). As in SDSA, these D-loops must have been disassembled by another helicase than the 1 knocked down, or by residual helicase nowadays after the knockdown.

Knocking down BLM did pb to elevated crossovers in the crossover analysis (Figure 4D). This issue was expected, based on phenotypes like elevated sister chromatid substitution (German et al. 1977). However, there was also an overall increment in HDR, as noncrossovers were likewise elevated. Thus, knocking down BLM resulted in elevated HDR in the DR-GFP assay (Paliwal et al. 2014), our SDSA assay, and our crossover assay. This might exist expected if knockdown affects the jail cell cycle profile, such that more cells are in Due south or G2 phases and therefore more likely to repair by HDR instead of NHEJ. We conducted cell bicycle profiling of cells in which BLM was knocked down, but did not detect any significant differences in the cell bicycle profile compared to the NT command (Figure S4 in File S1).

The causes of increased HDR when BLM or RTEL1 is knocked down remain unknown. This is an interesting area for time to come investigation, as understanding this unexpected effect will no uncertainty reveal important functions of these proteins.

Final remarks

We have demonstrated that our assays efficiently detect DSB repair past SDSA or that lead to crossovers, and that these assays tin can be used to study the effects of knocking downwards or knocking out different repair genes. Strengths of the assays include the ease of identifying the SDSA or crossover outcomes and the power to investigate other types of repair based on structures of repair products. Nosotros did non make up one's mind whether the lines we used had only a single re-create of the assay construct integrated (run across Materials and Methods), only analyses of cells exposed to I-SceI strongly contend that at that place was only one insertion location in both cases. In the SDSA analysis, an average of 2% of cells acquired ruby fluorescence in any given experiment. If there were insertions at ii unlike sites repairing independently, we would expect that in the vast majority of cases SDSA would occur in simply ane of the ii insertions. PCR across the I-SceI site would give two bands, one respective to the original construct and a larger band resulting from replacement of the 350-bp fragment. In opposition to this expectation, 55 out of 55 red-fluorescing clones examined had only the larger band. Similarly, crossovers in the crossover assay result in loss of mCherry. If in that location were several integrations every site would have to lose mCherry simultaneously to be scored as a crossover. It remains possible that one or both constructs integrated in a tandem assortment in the lines we used. If this happened, so it is likely that all I-SceI sites were cut, leaving some extrachromosomal fragments but simply a unmarried chromosomal repair template. It is unknown what issue the extrachromosomal fragments would accept on repair of the chromosomal DSB.

These assays can be modified to tailor their utilize in addressing specific questions. With the development of CRISPR/Cas9 genome editing (Cong et al. 2013; Mali et al. 2013), factor knockouts could be done instead of knockdowns, at least for genes that are not essential in the timeframe of these assays. It may be advantageous to use other cell lines for this approach, as U2OS cells accept more than two copies of many genes (Forbes et al. 2015). For some questions, it would be informative to incorporate SNP markers into the template then that gene conversion tract properties could exist measured at a college resolution than reported here. Differences betwixt the ii 350-bp repeats in the SDSA assay could be used to identify cases of template switching or two-ended invasions. Nosotros did non attempt to develop loftier-throughput sequencing of repair products, but amplification of the unabridged SDSA module with single-molecule tagging would make it possible to sequence a big number of contained repair events simultaneously. Finally, diverse distance parameters, such as changes to the amount of synthesis required to attain the repeats (they are immediately adjacent to the DSB in our assay, simply >5 kbp into the gap in the P{w^a } assay) or length of the repeats (350 bp in this assay, 275 bp in P{due west^a }) could provide insight into the frequency of template switching, the length of a typical synthesis tract, and the ability to repair larger gaps.

Elucidating details of SDSA and crossover repair is important for understanding DSB repair in general, but will also evidence vital in future optimization of CRISPR/Cas9 gene replacement or integration strategies that have been hypothesized to occur through SDSA (Byrne et al. 2015). We believe both assays we describe tin be useful in achieving these goals.

Acknowledgments

We thank Jan LaRocque for cells and teaching us the DR-GFP analysis, Mira Pronobis for U2OS and HeLa cells, and Yangze Gao and Cyrus Vaziri for the I-SceI adenovirus. Nosotros give thanks Diana Chong for assistance in acquiring confocal images of mCherry-positive cells. We give thanks Lydia Morris, Talia Hatkevich, and Juan Carvajal Garcia for comments on the manuscript. This work was supported by grants from the National Institutes of General Medical Sciences to J.S. under laurels numbers 5R01GM099890 and 1R35GM118127. The University of North Carolina (UNC) Menstruum Cytometry Core Facility is supported in part by P30 CA016086 Cancer Center Cadre Back up Grant to the UNC Lineberger Comprehensive Cancer Center.

Footnotes

Communicating editor: K. S. McKim

Source: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5386867/

Posted by: metzgersprevelink.blogspot.com

Share this post