In eukaryotes, two partially overlapping enzymes mediate the degradation of R-loops; RNase H1 and RNase H2 (Figure 1A; Lazzaro et al., 2012; Lockhart et al., 2019). To determine whether the Smc5/6 complex contributes to R-loop repair in proliferating cells, we introduced the smc6-9 and nse4-4 alleles of the Smc5/6 complex (Figure 1B) into budding yeast strains defective for RNase H1 (rnh1Δ) and RNase H2 (rnh201Δ/rnh202Δ/rnh203Δ) activity. We used the smc6-9 and nse4-4 alleles for our analysis because they correspond to moderate and a strong temperature-sensitive mutants of the Smc5/6 complex, respectively, and inactivate its DNA repair activity in a general/non-specific manner (Ben-Aroya et al., 2008; Hwang et al., 2008). For simplicity, we will refer to strains defective in both Smc5/6 complex and RNase H activities as double mutants and their parent strains as single mutants (i.e., even if the corresponding strains carry more than one mutant allele).

The proliferation rate of the Smc5/6-RNase H double mutants was compared to that of parental single mutants at various temperatures (23 °C, 30 °C, and 32 °C) and in the presence of DNA-damaging agents (hydroxyurea [HU], 4-nitroquinoline 1-oxide [4NQO], and methyl methanesulfonate [MMS]) (Figure 1C and D and Figure 1—figure supplement 1A). Loss of both Smc5/6 and RNase H activities resulted in impaired proliferation even under optimal growth conditions, as illustrated by the growth patterns of haploid spores after sporulation and dissection of heterozygous diploid strains (Figure 1C and D; YPD 23 °C). Competitive proliferation assays revealed that double mutant strains exhibit heightened temperature sensitivity at 30 °C and 32 °C compared to their corresponding parental strains (Figure 1D). Yeast cells without RNase H activity can replicate their chromosomes normally under unchallenged growth conditions, but their replication is hampered when exposed to DNA-damaging agents or replication inhibitors (Lockhart et al., 2019; Heuzé et al., 2023). Consistent with this, most of the Smc5/6-RNase H double mutants were more sensitive than single mutant strains when exposed to genotoxic stress (e.g., see HU at 23 °C in Figure 1D; MMS and 4NQO in Figure 1—figure supplement 1A). A similar growth exacerbation phenotype was observed when Smc5/6 complex mutations were combined with RNase H2 mutations in a RNase H1 proficient background (Figure 1—figure supplement 1B). Loss of Smc5/6 complex E3-ligase activity in the mms21-H202Y mutant also gave rise to a synthetic growth defect when combined with RNase H mutations (Figure 1—figure supplement 1C), thereby implicating sumoylation in RNA-DNA hybrid processing. Taken together, these experiments reveal that simultaneous loss of Smc5/6 complex and RNase H activities results in severe growth defects in yeast.

Next, we wanted to evaluate whether the growth defect observed in Smc5/6-RNase H double mutants is due to defective RNA-DNA hybrid removal. We quantified RNA-DNA hybrid foci by indirect immunofluorescence on chromosome spreads stained with the S9.6 antibody (Bou-Nader et al., 2022). Remarkably, we detected a large increase in the numbers of S9.6 foci on chromatin spreads prepared from double mutants compared to those from single mutant strains (Figure 2 and Figure 2—figure supplement 1). Whereas RNase H mutants alone typically show a few RNA-DNA hybrid/S9.6 foci per nucleus, inactivating Smc5/6 components in this genetic background resulted in the accumulation of more than 10 foci per nucleus. We observed similar results when measuring total S9.6 RNA-DNA hybrid fluorescence intensity in multiple fields of view (Figure 2—figure supplement 1). The presence of S9.6 foci was highest in the nse4-4 rnh1Δ rnh201Δ and smc6-9 rnh1Δ rnh201Δ strains, consistent with the fact that rnh201Δ represents the deletion of the catalytic subunit of RNase H2 (Figure 2A). RNA-DNA hybrids accumulated to similar levels in RNase H-only mutants (see rnh1Δ rnh201Δ, rnh1Δ rnh202Δ, and rnh1Δ rnh203Δ strains carrying 3–10 S9.6 foci per nucleus; Figure 2—figure supplement 1). nse4-4 and smc6-9 mutations alone did not result in a statistically significant increase in RNA-DNA foci formation, indicating that cells possess excess R-loop processing capacity when RNase H and other alternate R-loop metabolism pathways are active. Importantly, overexpression of an ectopic copy of RNase H1 gene (P_GAL1-RNH1) by galactose induction in the single and double mutant strains largely suppressed the S9.6 signal on chromatin spreads (compared to control cells without RNase H1 overexpression; Figure 2B). These results indicate that the foci described in Figure 2A are reflective of RNA-DNA hybrid formation in double mutant strains. To further confirm the accumulation of RNA-DNA hybrids in Smc5/6-RNase H double mutants, we performed a S9.6 antibody-mediated immunoprecipitation assay followed by qPCR analysis, as described in El Hage and Tollervey, 2018. We focused this analysis on loci that are known to be highly enriched in R-loop formation (Figure 3A). Consistent with the chromosome spread analysis, we observed increased RNA-DNA hybrid immunoprecipitation specifically at telomeres and ribosomal genes in the Smc5/6-RNase H double mutant strains compared to the parental strains deficient in only RNase H activity (Figure 3A).

In addition to the results presented above that are based on the binding between RNA-DNA hybrids and the S9.6 antibody; we wanted to test if the accumulation of RNA-DNA hybrid structures in the Smc5/6-RNase H double mutant strain could be detected using an alternative approach. Previous studies have established that the single-stranded DNA region of R-loop structures can be directly targeted by various mutagenic enzymes, one of them being the activation-induced cytosine deaminase (AID). This enzyme is highly active on single-stranded DNA during active transcription and creates mutations in DNA by deamination of cytosine and converting cytosine into uracil. This event leads to increased Rad52 foci formation in yeast, a homologous recombination-related phenotype that can be exploited to quantify R-loop abundance upon overexpression of the AID enzyme (Cañas et al., 2022). In line with our previous results, we observed a substantially higher level of AID-induced Rad52-GFP foci formation in the nse4-4 rnh1Δ rnh201Δ mutant strain compared to a wild-type control or the parental single mutants (rnh1Δ rnh201Δ; Figure 3B, wild-type and nse4-4; Figure 3—figure supplement 1). These results suggest that the severe growth defects of yeast defective in both RNase H and the Smc5/6 complex can be linked to an accumulation of R-loops in the chromosomes of these cells.

R-loops are most frequently observed at highly transcribed genes but can also be formed at telomeres and near DNA replication forks (Figure 4A). To identify the source of R-loops that are substrates for the Smc5/6 complex, we introduced mutant alleles of the Smc5/6 complex in yeast backgrounds that accumulate R-loops (or RNA-DNA hybrid structures) at specific genomic locations.

We first asked if unscheduled R-loops formed during active transcription are natural substrates/targets of the Smc5/6 complex. To test this notion, we used yeast strains defective for the Sen1 helicase (sen1-1 carrying a point mutation in the helicase domain of Sen1; Mischo et al., 2011) and THO complex (hpr1Δ; Luna et al., 2019), two conditions that lead to very high levels of RNA-DNA hybrids in actively transcribed genes (Figure 4A). Synthetic/aggravating interactions with these two mutant conditions are frequently used as a genetic assay to test the contribution of putative effectors of R-loop metabolism (e.g., Appanah et al., 2020). Interestingly, deletion of HPR1 was synthetic lethal when combined with the nse4-4 mutation, as evidenced by the growth pattern of haploid spores following sporulation and dissection of a heterozygous diploid strain (Figure 4B). Although hpr1Δ smc6-9 double mutants were viable, the growth defect associated with the hpr1Δ mutation was strongly exacerbated in the presence of smc6-9 at both permissive and restrictive temperatures (Figure 4B). smc6-9, nse4-4 and mms21-H202Y alleles also experienced synthetic growth defects when combined with sen1-1 (see 30 °C /32 °C and HU/MMS conditions in Figure 4B and Figure 1—figure supplement 1C). The nse4-4 sen1-1 mutant showed defective proliferation even under normal/unchallenged growth conditions, consistent with the more severe temperature sensitivity of this allele compared to that of smc6-9 (Figure 4B). Moreover, we detected a significant increase in the numbers of S9.6 foci on chromatin spreads prepared from nse4-4 sen1-1 and hpr1Δ smc6-9 mutants compared to those from single mutant strains (Figure 4—figure supplement 1). Taken together, these genetic interactions suggest that R-loop formed at highly transcribed genes are physiological substrates for the Smc5/6 complex.

Next, we combined smc6-9 and nse4-4 mutations with alleles of DNA polymerase ε and Sen1 helicase that increase the formation of RNA-DNA hybrids during DNA replication. Specifically, the pol2-M644G mutant exhibits a 10-fold increased ribonucleotide incorporation during DNA replication (McElhinny et al., 2010) whereas the sen1-3 allele (carrying point mutations in the N-terminal domain of Sen1) impairs the interaction of Sen1 with the replication machinery, thereby increasing R-loop formation in the vicinity of replication forks (Figure 4A; Appanah et al., 2020). We did not observe synthetic growth defects in double mutants of these genes with smc6-9 and nse4-4 mutations (Figure 4B), suggesting that DNA replication-associated RNA-DNA hybrid structures are not substrates for the Smc5/6 complex in vivo.

Finally, we investigated whether the Smc5/6 complex interacts with a natural R-loop formed at telomeres; the telomeric repeat-containing RNA [TERRA]-DNA hybrid. To this end, we combined smc6-9 and nse4-4 mutations with the rat1-1 allele defective in the 5' to 3' exonuclease activity responsible for TERRA removal (Figure 4A; Luke et al., 2008). The resulting double mutant strains exhibited stronger growth defects than the corresponding single mutants (Figure 4B), indicating a role for the Smc5/6 complex in R-loop metabolism at telomeres. This result is consistent with the significant accumulation of increased RNA-DNA hybrids specifically at telomeres in Smc5/6-RNase H double mutants (Figure 3A). Combining rat1-1 and smc6-9 mutations to sen1-1 phenocopied the sen1-1 smc6-9 double mutant (Figure 4C), consistent with an involvement of Sen1 in telomeric R-loops metabolism. Overall, these genetic analyses suggest that R-loops formed at highly transcribed genes and telomeres are likely physiological targets for the Smc5/6 complex.

The role we uncovered above for the Smc5/6 complex prompted us to investigate whether this enzyme was capable of recognizing R-loop structures directly. Extensive research over recent years has revealed that yeast and human Smc5/6 complexes are highly conserved and show strong binding affinities towards DNA substrates that mimic single-stranded (ss)–double-stranded (ds) DNA junctions, supercoiled or catenated DNA, and even branched DNA structures (Serrano et al., 2020; Gutierrez-Escribano et al., 2020; Tanasie et al., 2022). We have also established that the human SMC5/6 complex can bind short RNA-DNA duplexes in vitro (Serrano et al., 2020), suggesting the enzyme is capable of recognizing the more complex R-loop structure. To test this notion, we purified the human SMC5/6 complex and prepared R-loop and D-loop substrates (Figure 5A–B and Figure 5—figure supplement 1A) to conduct binding experiments by electrophoretic mobility shift assays (EMSAs) (Figure 5B–D and Figure 5—figure supplement 1B). The size of R-loop and D-loop substrates used in binding experiments are the same, enabling direct comparison of SMC5/6 complex affinity for these substrates (Figure 5—figure supplement 1A). We observed that the SMC5/6 complex can bind both R-loop and D-loop structures in EMSA experiments, but its specificity for the R-loop structure appeared slightly greater (Figure 5C and D). For instance, the SMC5/6 complex can bind the R-loop substrate even at a lower concentration of 6.25 nM and 12.5 nM, as evident from the protein/DNA band shift formed at the top of the gel at that concentration of enzyme (lane 2 and 3; Figure 5C). In contrast, we observed little to no binding to D-loops at 6.25 nM and 12.5 nM of SMC5/6 complex (lane 2 and 3; Figure 5D) and visibly moderate binding at the higher concentrations of protein. Consistent with this, the calculated equilibrium dissociation constant (K_D) of the Smc5/6 complex was slightly lower for R-loop substrates than for D-loops, but the significance of this difference, if any, remains to be established (Figure 5—figure supplement 1C). The similar range of binding affinities of the SMC5/6 complex for the R-loop and D-loop suggests that the single-stranded DNA part of the R-loop may be important for SMC5/6 complex binding on these structures—a mechanism previously observed for DICER-mediated R-loop resolution (Camino et al., 2023). Importantly, we observed that the SMC5/6 complex maintained its binding to R-loops even when challenged with 100-fold excess double-stranded DNA analogue (i.e., Poly[d(I-C)]), a behavior not observed with D-loops (lane 11–13 Figure 5C vs lane 10–12 Figure 5D). These data suggest that the binding of the Smc5/6 complex to R-loops is more resilient to challenge by a competitor substrate than the interaction of the enzyme to D-loops. Together, our EMSA experiments indicate that the SMC5/6 complex can efficiently associate with R-loops.

The results described above prompted us to test if R-loops are universal binding substrates for SMC complexes. To test this notion, we purified yeast condensin in monomeric and multimeric forms (Figure 5—figure supplement 2A–B; St Pierre et al., 2009; Keenholtz et al., 2017) and conducted EMSA experiments. While multimeric condensin bound R-loop substrates with high efficiency, similar concentrations of the enzyme in monomeric form failed to bind R-loop substrates in EMSA experiments (Figure 5—figure supplement 2C–F). This result suggests that the ability to bind R-loops is not an intrinsic property of monomeric condensin.

Next, we investigated whether binding of the SMC5/6 complex to an R-loop substrate can affect its degradation and/or stability in vitro. RNase H1 and RNase H2 are the primary enzymes responsible for the removal of R-loops in chromosomes, but recent studies suggest that RNase H2 is the only one that acts throughout the cell cycle (Lockhart et al., 2019; Zimmer and Koshland, 2016). Consistent with this, we also observed a remarkably strong genetic interaction when combining Smc5/6 complex mutations with RNase H2 enzyme inactivation (Figure 1—figure supplement 1B). We therefore tested the impact of the human SMC5/6 complex on the catalytic activity of purified human RNase H2 enzyme in a reconstituted R-loop degradation assay (Figure 6A). Similar to the EMSA, we introduced a ³²P radiolabel on the RNA moiety of our R-loop substrate to allow direct visualization of RNA degradation (Figure 6B and Figure 6—figure supplement 1). We first confirmed that human RNase H2 was able to cleave the radiolabeled RNA in a time and concentration-dependent manner (Figure 6—figure supplement 1A–B). We also observed that the purified RNase H2 enzyme was not active on a D-loop structure, thereby demonstrating the specificity of the enzyme in our reaction conditions (Figure 6—figure supplement 1C). Next, we incubated increasing concentrations of the SMC5/6 complex with the R-loop substrate in the absence of RNase H2. Under these conditions, the complex did not induce degradation of the RNA moiety or otherwise affect the stability of the R-loop structure (lanes 2–6; Figure 6C). However, in the presence of low levels of RNase H2 enzyme, the same concentrations of SMC5/6 complex induced a major stimulation of R-loop degradation. This led to a rapid accumulation of radiolabeled product at the bottom of the gel (lanes 8–12; Figure 6C), reflecting the nucleolytic processing of the RNA moiety within the R-loop by RNase H2. The stimulation of RNase H2 activity by the SMC5/6 complex was concentration-dependent and evident even at the lowest concentration of the SMC5/6 complex tested in this experiment (1.25 nM, lane 8; Figure 6C). To test the possibility that the SMC5/6 complex modulates RNase H2 activity by direct binding, we performed an in vitro pull-down assay and found no detectable physical interaction connecting these proteins, a result further confirmed by co-immunoprecipitation assays from yeast extracts (Figure 6—figure supplement 2A–B). Taken together, these results indicate that the promotion of R-loop degradation by the SMC5/6 complex does not depend on this enzyme establishing a strong physical interaction with RNase H2.

All yeast strains used in this study are derivatives of strain K699/K700. The complete list of the relevant genotypes of the strains used in the study are provided in Table 1. Yeast growth conditions, procedures for genetic analysis, and creation of strains carrying relevant mutations was performed as previously described in Serrano et al., 2020. To create double mutant strains, haploid mutants carrying the specified alleles were mated to produce heterozygous diploid yeasts. Sporulation and dissection of diploid strains was subsequently performed at 23 °C. For the cell viability assay, performed under conditions of DNA damage or replication stress, yeast strains were grown on a solid medium containing MMS, 4-NQO, and HU at different temperatures. Specifically, a fivefold dilution series of wild-type and mutant yeast cultures (the first spot on the left side of the plate corresponds to a culture at OD₆₀₀ of 0.2) were spotted on solid YPD (yeast extract, peptone, 2% glucose) with or without the presence of DNA-damaging agent and grown in temperature-controlled incubators for 28–72 hr before scanning the plates in a scanner (Serrano et al., 2020). Strains expressing RNase H1 were generated by integrating YIplac204::Pgal1::RNH1 at the TRP1 locus and integration was confirmed by PCR screening. To create strains expressing the AID enzyme, we transformed relevant strains tagged with GFP at the C-terminus of Rad52 with pESC-LEU-HsAIDSc plasmid (Addgene plasmid #60810) (Mayorov et al., 2005). A complete list of plasmids used in this study is also provided in Table 2.

Chromatin spreads were performed as described previously (Grubb et al., 2015) with some minor modifications. Briefly, cells from the mid-log phase grown in YPD at 23 °C were collected and washed in 1 mL ZK buffer (25 mM Tris, pH 7.5, 0.8 M KCl) and were resuspended in ZK buffer supplemented with 1 M DTT. Spheroplasting of cells was achieved by the addition of 5 μl of Zymolyase (20 mg/ml) and incubation at 30 °C with gentle rotation. Subsequently, spheroplast cells were centrifuged (2000 rpm/5 min) and resuspended in MES/Sorbitol buffer (0.1 M MES pH 6.5, 0.5 mM MgCl₂, 1 mM EDTA, 1 M sorbitol). Next, cells were added to the glass slide (Corning) and immediately were fixed and lysed by the addition of a fixative solution (3% paraformaldehyde in 4% sucrose) and 1% NP-40 substitute solution. Cells were spread using a plastic pipette rolled from one end of the slide to the other end. Slides with chromatin spreads were dried overnight. The next day, slides were washed with 1 x TBS (Tris-buffered saline) for 10 min and blocked for 15 min with 5% BSA (Bovine serum albumin) in 1 x TBS. Chromatin spreads were incubated with mouse monoclonal antibody S9.6 (MABE1095) (1:250 dilution) for overnight followed by Cy3-conjugated goat anti-mouse antibody (Jackson Laboratories, #115-165-003) (1:700 dilution) for 2 hr. Nuclei were counterstained with 50 μl of VectaShield (Vector Laboratories, CA) plus 1 x DAPI (4’,6- diamidino-2-phenylindole) and sealed with nail polish. Images were acquired using Nikon Eclipse Ti2 inverted microscopy with an oil immersion 100 x objective. For each replicate (n>3), about 150 nuclei were visualized and manually counted to obtain the fraction with detectable RNA-DNA hybrid foci using the 3D measurement module of the NIS-Elements software (Nikon Instruments Inc). Fluorescence intensity of RNA-DNA hybrid structures (arbitrary units; A.U.) was quantified using ImageJ (N.I.H, USA).

Mid-log cultures grown in YPD at 23 °C were collected. RNA-DNA hybrids were processed and analyzed as described in El Hage and Tollervey, 2018. Real-time quantitative PCR was performed at the indicated regions using the SsoAdvanced SYBR Green PCR Master Mix (Bio-Rad) with a CFX384 Real-Time PCR System (C-1000 Touch Thermal Cycler). Data was analyzed using the CFX Maestro Bio-Rad software and the relative abundance of RNA-DNA hybrid immunoprecipitated in each region was normalized to the signal obtained in the inputs. Average and standard error of at least three independent experiments are shown.

The human SMC5/6 core complex was purified using a triple affinity purification approach followed by size exclusion chromatography, as described by Serrano et al., 2020 with minor modifications. 35 L of an S. cerevisiae strain overexpressing the core complex was cultured under optimal growth conditions in a bioreactor (Techfors-S-42L) to an OD₆₀₀ of 0.7–1.0. Protein expression was induced by the addition of 2% galactose and cells were grown further for 16 hr at 18 °C. Briefly, yeast pellets were resuspended in 200 ml buffer N (50 mM K₂HPO₄ / KH₂PO₄ pH 8, 50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10% glycerol, 0.5% Triton X-100, 2 mM β-Mercaptoethanol) supplemented with 20 mM imidazole and protease inhibitors (E64, Pepstatin A, 4-(2-aminoethyl) benzene sulfonyl fluoride hydrochloride [AESBF]). Yeast popcorn is made by the dropwise freezing of the cell suspension in liquid nitrogen. The popcorns were further lysed two cycles in a freezer mill. The lysates were resuspended in 1 L of buffer N and passed through a high-pressure homogenizer (Avestin EmulsiFlex-C3) at an operating pressure of 25,000 psi. The final lysate was centrifuged at 24,000 rpm for 45 min at 4 °C. The soluble lysates were passed through a column packed with Nickel-NTA resin at a flow rate of 5 ml/min for slower binding. The unbound fractions from the first purification column were loaded to a second Ni-NTA column for a second-round binding to maximise yield of purified protein. Both the columns were washed with 10 column volumes (CV) of buffer N supplemented with 60 mM imidazole. Complex was eluted with buffer SB (50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10% glycerol, 0.5% Tween 20, 2 mM βME) supplemented with 500 mM imidazole. The combined fractions from both the columns were loaded into a StrepTractin XT 5 mL column using an AKTA prime FPLC purification system. The column was programed to wash with 10 CV of buffer SB supplemented with 0.5% Triton X-100 and eluted with 5 CV of buffer GB (25 mM K₂HPO₄/KH₂PO₄ pH 8, 500 mM NaCl, 10% glycerol, and 2 mM βME) supplemented with 50 mM biotin. The elution was mixed and incubated with 5 ml of pre-equilibrated Glutathione S-transferase (GST)-Sepharose resin, in GST binding buffer GB (20 mM K₂HPO₄/KH₂PO₄ pH 8, 200 mM NaCl, 10% glycerol and 2 mM DTT) for 2 hr at 4 °C. The resin was washed with 10 CV of buffer GEB (50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10% glycerol, and 2 mM βME). The SMC5/6 complex was eluted with 5 CV of buffer GEB supplemented with 25 mM of reduced Glutathione. Linker, poly-histidine, Strep-tag II, and GST tags were cleaved by an overnight digestion with 1 mg of TEV protease per 4 mg/mL of fusion protein. Digestion was carried out in GEB buffer supplemented with 1 mM DTT. Digestion product was loaded into a Superose 6 10/300 size exclusion chromatography column in GF buffer (50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10% glycerol, and 2 mM βME) in order to remove the cleaved tags, digested linker, and TEV protease. Elution fractions containing highly purified and stoichiometric complexes were concentrated, quantified, snap-frozen, and stored at −80 °C.

A polycistronic vector allowing the co-expression of all the subunits of human RNase H2 (pET-hH2ABC); was obtained from Robert J. Crouch (NIH). All the subunits of the holoenzyme–namely, RNase H2A, H2B, and H2C–are expressed from this vector as N-terminal hexahistidine fusion proteins. For purification, Escherichia coli BL21 was transformed with pET-hH2ABC, and 6 L of culture was grown at 37 °C to an OD₆₀₀ of 0.4–0.6 before being induced with 0.3 mM of IPTG. The culture was grown further at 18 °C for 16 hr after induction. Bacterial pellets were resuspended in 100 ml buffer N (50 mM K₂HPO₄ /KH₂PO₄ pH 8, 50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10% glycerol, 0.5% Triton X-100, 2 mM βME) supplemented with 20 mM imidazole and protease inhibitors (E64, Pepstatin A, AESBF). The lysate was passed twice through a high-pressure homogenizer (Avestin EmulsiFlex-C3) at an operating pressure of 15,000 psi and centrifuged at 24,000 rpm for 45 min at 4 °C. The soluble lysates were passed through a column packed with Ni-NTA resin at a flow rate of 5 ml/min for slower binding. The column was washed with 10 column volumes (CV) of buffer N supplemented with 60 mM imidazole. Complex was eluted with buffer SB (50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10% glycerol, 0.5% tween 20, 2 mM βME) supplemented with 500 mM imidazole. The eluted fractions were dialyzed against 20 mM HEPES pH 7.6, 250 mM NaCl, 10% glycerol, and 1 mM DTT, quantified, snap frozen, and stored at −80 °C.

The DNA or RNA oligos are 5’-labeled with ATP-[γ³²P] (PerkinElmer Life Sciences) using T4 polynucleotide kinase (New England BioLabs). Radiolabeled oligos were then annealed to a complementary strand by heating to 95 °C and slow cooling over a long period of time in PNK buffer (70 mM Tris-HCl pH 7.6, 10 mM MgCl₂, 5 mM DTT). Annealed substrates were separated from free ATP-[γ³²P] on an 8% native PAGE in Tris Borate/EDTA buffer at room temperature. The gel band corresponding to the annealed substrate was excised, purified, and finally eluted. The eluted substrates were quantified (nM) using a scintillation counter.

Two types of R-loop substrates were synthesized. First, the R-loop substrate was constructed by annealing [³²P]-labeled RNA strand (DD 4265) with DNA strand 1 (DD 4263) and DNA strand 2 (DD 4264). Another R-loop substrate was constructed by annealing [³²P]-labeled DNA strand 2 (DD 4264) with DNA strand 1 (DD 4263) and RNA strand (DD 4265). Second, a D-loop substrate was also generated by annealing [³²P]-labelled DNA strand 2 with DNA strand 1 and a DNA strand 3 (DD 4266). A single [³²P]-labeled ssDNA strand (DD4264) was used as a control for electrophoretic migration in a gel.

The sequence of oligonucleotides used as in vitro substrates are: DNA strand 1 (DD 4263; 1.5 µg/µl or 100 µM): 5’GGGTGAACCTGCAGGTGGGCGGCTGCTCATCGTAGGTTAGTTGGTAGAATTCGGCAGCGTC-3’ (61 mer); DNA strand 2 (DD 4264; 1.8 µg/µl or 100 µM): 5’GACGCTGCCGAATTCTACCAGTGCCTTGC TAGGACATCTTTGCCCACCTGCAGGTTCACCC-3’ (61 mer); RNA strand 1 (DD 4265; 0.5 µg/µl or 100 µM): 5’-AAAGArUGrUCCrUAGCAAGGCAC-3’ (21 mer); DNA strand 3 (DD 4266; 0.6 µg/µl or 100 µM): 5’-AAAGATGTCCTAGCAAGGCAC-3’ (21 mer).

The DNA binding activity of human SMC5/6 complex and yeast condensin was determined by electrophoretic mobility shift assay (EMSA). The condensin holoenzyme was purified according to a published procedure (St Pierre et al., 2009). Reactions containing 40 nM [³²P]-labeled oligonucleotides and the indicated concentrations of SMC5/6 complex and condensin (i.e. 0, 6.25, 12.5, 25, 50, and 100 nM) were incubated in binding buffer A (25 mM MOPS [morpholinepropanesulfonic acid] pH 7.6, 60 mM KCl, 0.2% NP40, 2 mM DTT, 5 mM MgCl₂) in a total volume of 15 μl. Reactions were incubated at 24 °C for 10 min and loaded on a 6% acrylamide gel, electrophoresed at 150 volts for 240 min in 1 X TBE buffer for EMSA. Gels were then dried onto DE81 filter paper and visualized by autoradiography. To evaluate the specificity of the SMC5/6 complex binding to R-loop as compared to D-loop, increasing concentrations of human SMC5/6 complex (i.e. 0, 6.25, 12.5, 25, 50, and 100 nM) were incubated with 100 x molar excess concentration of unlabeled poly-deoxy-inosinic-deoxy-cytidylic acid (poly[d(I-C)], Roche, Cat. No. 0108812001) along with 40 nM of [³²P]-labelled R-loop and D-loop, respectively, followed by EMSA. Quantification of unbound and protein-bound DNA was performed with ImageJ using the histogram function. Data was fitted to a Hill equation with dissociation constants (K_D) using GraphPad Prism 9.0 (GraphPad Software Inc).

RNase assays were performed in Buffer C (20 mM HEPES [4-(2-hydroxyethyl)–1-piperazineethanesulfonic acid] pH 7.5, 150 mM NaCl, 10 mM MgCl₂, 0.5 mM DTT). The R-loop (40 nM) substrates were pre-incubated with SMC5/6 complex at the indicated concentration in buffer C for 10 min at 24 °C followed by the addition of RNase H2 complex for 7 min at the same reaction conditions. Reactions were deproteinized in a one-fifth volume of stop buffer (Buffer A, 1% SDS, 5 mM EDTA, and 0.2 mg/ml proteinase K) for 15 min at 24 °C. Reactions were loaded on an 8% acrylamide gel, electrophoresed at 150 volts for 150 min, dried onto DE81 filter paper, and visualized by autoradiography.

Detection of R-loops through AID-induced DNA damage and subsequent rise of Rad52 foci was performed as described in Cañas et al., 2022. Strains were grown at 23 °C in minimal media followed by galactose induction for 2 hr to overexpress AID enzymes and collected. Cells were fixed in formaldehyde (10% in 0.1 M KPO₄ pH 6.4) for 30 min at room temperature and washed twice in 0.1 M KPO₄ pH 7.0 buffer. Staining of the nuclei was performed with DAPI at a final concentration of 2 µg/ml in cells suspended in 0.1 M KPO₄ pH 7.0 buffer. For DAPI staining and Rad52-GFP visualization, images were acquired using Nikon Eclipse Ti2 inverted microscopy with an oil immersion 100 x objective. For each replicate (n=3), about 100 cells were visualized and manually counted to obtain the fraction with detectable Rad52 foci using the 3D measurement module of the NIS-Elements software (Nikon Instruments Inc).

Active RNase H2 enzyme and SMC5/6 complex were purified as described above. The resulting complexes carry tandem Strep-tag II (3xSTII) tags on SMC6 and a 10xHis tag on NSMCE4–SMC5. RNase H2 subunits were tagged with a 6xHis tag. For the pull-down procedure, Streptactin XT resin (10 µL per reaction) was washed with ddH₂O and with binding buffer (50 mM HEPES-NaOH [4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid, N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid)–sodium hydroxide] pH 7.5, 150 mM NaCl, 5% glycerol, 1% BSA, 5 mM MgCl₂) before being incubated in binding buffer in a total volume of 300 µL for 2 hr at 4 °C with or without 0.8 µM human SMC5/6 complex. The resins were then washed with wash buffer (50 mM HEPES-NaOH pH 7.5, 150 mM NaCl, 5% glycerol, 0.2% NP40, 5 mM MgCl₂). For each condition, 10 µL of Streptactin XT resin with or without 1.95 µM of human RNase H2 were incubated in a total volume of 300 µL of binding buffer at 4 °C overnight (~12 hr). The reactions were then washed with wash buffer. The pellet was resuspended in wash buffer and 4 x sample buffer (90% 4 x Laemmli sample buffer, 10% β-mercaptoethanol) at a ratio of 1:1. The reactions were then loaded on an 4–12% Bis-Tris SDS-PAGE gel and ran in 1 x MOPS buffer for 2 hr at 120 Volts, before being analyzed by immunoblotting with 1:2500 dilution of anti-HIS antibody (Qiagen 34660).

Co-immunoprecipitation assay was performed as described previously (Pastic et al., 2024) with minor modifications. Yeast cultures (50 mL) were grown in YPD at 23 °C to an OD₆₀₀ of 0.7–0.8 and washed with ddH₂O before being snap-frozen in liquid nitrogen and stored at –80 °C. For the co-IP method, magnetic protein G beads (Dynabeads Protein G; Thermo Fisher Scientific, Cat #1003D) were washed with the PBS-BSA buffer (1 x PBS [Phosphate buffered saline], 1% BSA) and incubated for 2 hr at 4 °C with 4 µg of mouse anti-Strep antibody (Qiagen 34850) antibody or 4 µg of mouse IgG Isotype control (Invitrogen 02–6502) per strain to confirm the specificity of the anti-Strep antibody. Cells were lysed by vigorous shaking with glass beads in 800 µL of co-IP lysis buffer (50 mM HEPES–KOH pH 7.5, 140 mM NaCl, 1% Trition X-100, 0.1% Sodium deoxycholate, 1 mM EDTA pH 8.0, 1 mM AEBSF, 10 µM Pep-A, 10 µM E64). The lysates (~1 mL) were then cleared by centrifugation. Once the beads were washed with the PBS-BSA buffer, the lysate were equally split between the two beads conditions and incubated at 4 °C overnight (~12 hr). The beads were then washed with the co-IP lysis buffer, with high salt buffer (50 mM HEPES – KOH pH 7.5, 360 mM NaCl, 1% Trition X-100, 0.1% Sodium deoxycholate, 1 mM EDTA pH 8.0) and with TE buffer (10 mM Tris-HCl pH 8.0, 0.1 mM EDTA) and resuspended in 2 X sample buffer (4 X Sample buffer, 0.2% of BPB [bromophenol blue],~5% ddH₂O, 5% of β-mercaptoethanol). Proteins were separated on an SDS-PAGE gel [4–12% Bis-Tris] in 1 x MOPS buffer and ran for 2 hr at 120 Volts. Proteins were subsequently transferred on a membrane before being analyzed by immunoblotting with 1:2000 dilution of mouse anti-Strep antibody and 1:5000 dilution of mouse anti-Myc (9E10) antibody (GeneTex 369661).

Results presented in this study are representative examples of at least three independent experiments. All statistical analyses were performed using GraphPad Prism 7 (GraphPad Software Inc). Sample size (n) and statistical tests performed in each experiment are described in the relevant figure legends. In all cases, p-values expressed as *p<0.05, **p<0.005, and ***p<0.0005 are considered significant.