Loss of SLX4IP leads to common fragile site instability and compromises DNA interstrand crosslink repair in vivo

Abstract

Common fragile sites (CFSs) are chromosomal loci with inherent characteristics that make them difficult to fully replicate thus rendering them vulnerable to replication stress (RS). Under-replicated CFSs manifest as cytogenetic gaps and breaks on metaphase chromosomes. Moreover, CFSs are hotspots for tumorigenic chromosomal rearrangements. The Fanconi anemia (FA) pathway is at the core of a network of proteins that work to safeguard CFSs during replication and RS. Here, we uncover a novel role of SLX4IP in maintaining CFS stability. We show that SLX4IP localizes stressed CFSs and that its loss exacerbates genome instability, including CFS expression. Furthermore, direct SLX4IP depletion leads to impaired replication and growth deficiencies. SLX4IP and FANCP/SLX4 are epistatic, suggesting that SLX4IP acts with SLX4 to maintain CFS stability. Finally, zebrafish larvae with homozygous knockout of the slx4ip gene showed higher frequency of embryonic anomalies and sensitivity to DNA crosslinking agents, a typical cellular characteristic of patients with FA. Our results establish a causal link between SLX4IP deficiency and chromosomal instability, which may explain how SLX4IP dysregulation contributes to cancer development.

Faithful DNA replication is essential for error-free genome duplication and complications that perturb the replication machinery can lead to genomic instability and cancer predisposition (1, 2). Difficult-to-replicate chromosomal regions, such as telomeres, centromeres, and common fragile sites (CFSs), halt replication locally. Consequently, these regions exhibit high propensity to end up under-replicated following partial inhibition of DNA synthesis (3, 4, 5), resulting in mis-segregation and rearrangement events (6). Difficult-to-replicate regions share characteristics that contribute to their instability. They often contain repetitive sequences that are prone to form DNA secondary structures, such as hairpins (7, 8), and they are replicated late in S-phase (9, 10, 11). CFSs are unique in that they contain long genes that are undergoing active transcription throughout S-phase, which leads to an increased incidence of replication and transcription machinery collisions, R-loop formation, and displacement of replication origins (12, 13). While CFSs pose a serious threat to the integrity of DNA replication, endogenous RS is rarely sufficient to cause their expression. However, the induction of RS, such as that seen with oncogenic transformation or the addition of low doses (0.2–0.4 μM) of the polymerase inhibitor Aphidicolin (APH), exacerbates CFS expression that can be visualized as chromosomal breaks and gaps in metaphase chromosomes (14, 15).

The Fanconi anemia (FA) pathway is responsible for repairing interstrand crosslinks (ICLs), but it is also central to maintaining unperturbed replication of CFSs. Mutations in at least 22 known proteins causes FA (16), eight of which make up the FA core complex that monoubiquitinates the FANCD2-FANCI heterodimer in response to RS. This in turn orchestrates subsequent nucleolytic cleavage, translesion synthesis and recombination steps, partly through the remaining 12 FA proteins (16). FANCP/SLX4 and FANCD2 have been well-characterized in the context of CFSs. FANCD2 directly facilitates replication through CFSs in the absence of exogenous RS, safeguarding their stability (17). If FANCD2-mediated protection fails, CFSs are not properly replicated during S-phase and thus enter prophase under-replicated, which triggers mitotic DNA synthesis (MiDAS) (18). Analogous to break-induced replication (BIR) in yeast, MiDAS is homologous recombination (HR)-based and requires initial endonucleolytic cleavage of stalled replication forks flanking an under-replicated region, followed by template strand invasion. Unlike canonical HR, this process is RAD52-dependent and requires FANCD2, SLX4, the endonuclease MUS81-EME1, and the polymerase δ subunit POLD3 (19, 20). Without SLX4, CFSs fail to resolve via MiDAS (18). Notably, SLX4 localizes to difficult-to-replicate regions independently of replication, suggesting additional roles at CFSs (21). Unresolved under-replicated CFSs form ultrafine anaphase bridges which repair is mediated by mechanisms involving TOP2A, PICH and the BTRR complex (22, 23, 24). Finally, pervasive under-replication of CFSs throughout mitosis can lead to sequestration of CFSs into micronuclei (25).

Recently, a comprehensive analysis of proteins recruited to CFSs generated through chromatin immunoprecipitation of FANCD2 coupled to mass spectrometry in G2/M synchronized HeLa cells (26) revealed new CFS-associated proteins, including SLX4-interacting protein (SLX4IP). SLX4IP, previously named C20orf94, is a direct interactor of SLX4 (27), and its disruption has been linked to cancer, as demonstrated by its dysregulation in Merkel cell carcinoma (28), frequent deletion in acute lymphoblastic leukemia (29), and downregulation in myeloid leukemia cells (30). Notably, SLX4IP plays an important role at telomeres. Loss of SLX4IP dramatically changes telomeric proteomes (31), and molecular studies have revealed a close relationship between SLX4 and SLX4IP in alternative lengthening of telomeres (ALT)-recombination. SLX4IP modulates ALT HR intermediates by interacting with SLX4, BLM, and XPF, and the loss of SLX4IP induces phenotypes associated with telomere instability (32). While SLX4IP is lost in some ALT-positive cancers, others have reported that SLX4IP-deficient cells switch from ALT to a telomerase-positive state, suggesting a crucial role for SLX4IP in ALT induction (33). SLX4IP furthermore plays a critical role in the repair of ICLs by ensuring the stable tethering of XPF to the SLX4 trinuclease complex (34). As CFSs are phenotypically similar to telomeres in repair and replication stress (35), we sought to establish whether SLX4IP is as critical for CFS stability as it is for telomere stability. Our results show that SLX4IP localizes to stressed CFSs and that depletion of SLX4IP is sufficient to cause increased chromosomal instability with exacerbated expression of CFSs. SLX4IP-depleted U2OS cells also exhibit growth perturbations and impaired S-phase nucleotide incorporation. SLX4IP appears to act in coordination with SLX4 to maintain chromosomal stability. In vivo studies using zebrafish show that SLX4IP is necessary for proper embryogenesis, and its loss confers sensitivity to a DNA crosslinking agent, DEB, similar to the loss of known FA disease-causing genes.

Results

SLX4IP localizes to stressed CFSs

To determine whether SLX4IP plays a role in maintaining CFS stability, we first evaluated SLX4IP foci under conditions that perturb CFS integrity. APH-induced RS causes CFS under-replication and activates the FA pathway (18). Therefore, we used a low dose of APH previously reported to induce RS (26), and foci formation was evaluated by high-content microscopy. SLX4IP forms distinct nuclear foci whose number and mean intensity are induced upon treatment with APH (Fig. 1A; Fig. S1A). By labeling the cells with EdU and DAPI to allow for cell cycle profiling, we found that SLX4IP foci formation is most prominent in U2OS cells during mid-S phase to G2 and that APH-induced RS increases SLX4IP foci mostly during later stages of S-phase (Fig. 1B). Next, we sought to establish whether SLX4IP localizes to stressed CFSs. FANCD2 associates with CFSs in late G2 and early mitosis and can thus be used as CFS marker (15, 28). SLX4IP strongly colocalized with FANCD2 on prometaphase chromatin (Fig. 1C). SLX4IP was also found to often colocalize with FANCD2 and constitutively colocalize with SLX4 in G2 cells (Fig. S1, B–D), suggestive of CFS association. Importantly, a low dose of APH increased the number of colocalization events in all cases as well as the fraction of FANCD2 foci colocalizing with SLX4IP (Fig. 1D). SLX4IP foci formation was completely abrogated upon siRNA-mediated SLX4IP depletion, confirming the specificity of the signal (Fig. S1E). In U2OS cells, FANCD2 and SLX4 are found at both telomeres and CFSs. Staining for SLX4IP/FANCD2/TRF2 confirmed that SLX4IP colocalizes to telomeres in G2 cells as previously reported but also revealed SLX4IP/FANCD2 colocalization events without TRF2 (Fig. S1F). To further confirm that SLX4IP associates with CFSs upon RS, we used a SLX4IP-KO U2OS cell-line reconstituted with GFP-SLX4IP. ChIP-qPCR revealed an enrichment of GFP-SLX4IP at the fragile site FRA7H, but not at the non-fragile region GAPDH, upon treatment with APH (Fig. 1E). Together, these data indicate that SLX4IP associates with CFSs during APH-induced RS.

Figure 1.

SLX4IP localizes to stressed CFSs. A, representative images of SLX4IP foci. SLX4IP forms distinct nuclear foci induced by APH (0,2 μM). B, for cell cycle analysis, U2OS cells are sorted by mean EdU and total DAPI content. This allows for quantification of SLX4IP foci in each cell cycle stage (mean ± SD; n = 3 biological replicates; >260 cells analyzed per condition). C, representative images of SLX4IP colocalization with FANCD2 in U2OS prometaphase cells. D, quantification of foci in (C) (mean ± SD; n = 3 biological replicates; 50 cells analyzed per condition). E, ChIP-qPCR analysis of GFP-SLX4IP or GFP localization to GAPDH or FRA7H in cells treated with DMSO or APH (0,2 μM) (mean ± SD; n = 2 biological replicates). Significance was assessed by Student’s unpaired t-tests. Scale bars: 10 μM.

SLX4IP deficiency impairs replication and promotes genome instability

To investigate the cellular consequences of SLX4IP deficiency, protein depletion was achieved using siRNA in U2OS cells (Fig. 2A). We observed that transient depletion of SLX4IP impairs immediate cellular proliferation as measured by high-content microscopy of transfected H2B-GFP U2OS cells (Fig. 2B). SLX4IP kd cells exhibit impaired S-phase EdU incorporation both in untreated and in APH-treated conditions (Fig. 2, C and D) but induced no considerable differences in cell cycle distribution (Fig. S2A). Staining with a thymidine analog such as EdU and DAPI allows for assessment of the degree of polyploidy in cell populations (36, 37) (Fig. S2B). Quantification of this fraction revealed that transient depletion of SLX4IP induced an increase in polyploid U2OS cells (Fig. S2C), which was not further exacerbated by APH.

Figure 2.

Transient SLX4IP-depletion perturbs replication. A, Western blot showing effect of siRNA SLX4IP depletion. B, representative images and corresponding proliferation curves for H2B-GFP U2OS cells relative to day 0 transfected with indicated siRNA (mean ± SD; n = 3 biological replicates). C, representative cell cycle profiles for U2OS cells transfected with indicated siRNA. 5000 cells are displayed per condition. D, graph of S-phase EdU incorporation from conditions in (C) (mean ± SD; n = 3 biological replicates). Scale bar: 100 μM.

To determine whether SLX4IP promotes stable replication of CFSs, we evaluated parameters associated with CFS instability. 53BP1 nuclear bodies (53BP1 N.Bs) protect under-replicated CFSs during the subsequent G1 phase (38). SLX4IP deficiency increased the number of G1 53BP1 N.Bs, suggestive of increased genomic instability (Fig. 3A). A similar increase in the number of micronuclei, structures associated with CFS disruption, was also observed (Fig. 3B). Consistent with previous reports, we were able to generate viable SLX4IP-KO U2OS cells displaying complete abrogation of SLX4IP (Fig. S3B). These cells did not exhibit consistent growth defects but showed an increase of G1 53BP1 N.Bs similar to transiently depleted cells (Figs. 3C; Fig. S3C). Strikingly, the increase in 53BP1 N.Bs and micronuclei observed upon loss of SLX4IP as compared to WT was not recapitulated by APH treatment, which induces genome instability to a similar extent in both conditions (Fig. 3, A–C), raising the possibility that SLX4IP safeguards chromosome stability under basal conditions, while its role may become saturated under exogenous replication stress. Together, these findings indicate that SLX4IP plays a role in maintaining normal replication progression and genome stability.

Figure 3.

SLX4IP-deficient U2OS cells exhibit global chromosomal instability markers. A, G1 53BP1 N.Bs in siCtrl- and siSLX4IP-transfected U2OS cells (mean ± SD; n = 3 biological replicates; >2000 cells analyzed per condition). Cyclin A-positive cell marked by an asterisk for reference. B, micronuclei (MN) in siCtrl- and siSLX4IP-transfected U2OS cells (mean ± SD; n = 3 biological replicates; >1000 cells analyzed per condition). C, G1 53BP1 N.Bs in U2OS SLX4IP KO cells and EV cells (mean ± SD; n = 3 biological replicates; >1500 cells analyzed per condition). Significance was assessed by Student’s unpaired t-tests. Scale bars: 10 μM.

SLX4IP depletion causes chromosomal breakage

We observed that transient SLX4IP depletion exacerbated chromosomal breaks and gaps (Fig. 4, A and B). Treatment with a low dose of APH increased the number of breaks and gaps, which was further exacerbated with the loss of SLX4IP. This suggests that SLX4IP is integral for the protection of CFS stability, and to confirm that these broken loci indeed represent CFSs, we assessed whether they actively undergo mitotic DNA synthesis (MiDAS), a hallmark of CFSs (18), by scoring the number of breaks with prometaphase EdU incorporation. SLX4IP depletion increased the fraction of EdU-negative breaks slightly but more consistently increased the number of EdU-positive breaks corresponding to under-replicated CFSs (Fig. 4, C and D). Treatment with low-dose APH increases the number of broken CFSs, but SLX4IP depletion combined with APH does not consistently exacerbate CFS breakage. Instead, the increase in chromosomal fragility upon APH appears driven by EdU-negative breaks (Fig. 4D), perhaps reflecting broader failure of replication completion. Together, this indicates that SLX4IP depletion alone is sufficient to drive CFS instability. Depletion of SLX4 is sufficient to abrogate MiDAS (18). To assess whether MiDAS is impaired by SLX4IP kd, we counted EdU foci on metaphase chromosomes. MiDAS is intact and increased slightly upon SLX4IP depletion, further suggesting impaired CFS integrity (Fig. 4E).

Figure 4.

SLX4IP-deficient U2OS cells exhibit chromosomal breakage and CFS underreplication. A, metaphase chromosome spreads of U2OS cells transfected with indicated siRNA. Gaps and breaks are indicated with arrows. B, quantification of chromosomal gaps and breaks as in (A) (mean ± SD; n = 2 biological replicates; >40 metaphase spreads analyzed per condition). C, representative images of EdU-positive chromosomes marking CFSs. D, quantification of (C) (mean ± SD; n = 2 biological replicates; >1800 chromosomes analyzed per condition). E, quantification of metaphase EdU foci (mean ± SD; n = 2 biological replicates; >600 chromosomes analyzed per condition). Cells were treated with 0,4 μM APH for 24h in all experiments. Significance was assessed by Student’s unpaired t-tests. Scale bar: 10 μM.

SLX4IP functions with SLX4 to maintain genome stability but also displays SLX4-independent functions

Our data indicate that SLX4IP associates with CFSs and that its loss hinders normal proliferation and causes genomic instability, including CFS breakage. Together with FANCP/SLX4, SLX4IP plays a role in interstrand crosslink-repair and homologous recombination (30, 32). We therefore examined whether the observed functions of SLX4IP require SLX4. To this end, we assayed for epistasis between SLX4 and SLX4IP. SLX4 depletion was achieved using siRNA, and two independent siRNA abrogate SLX4 foci formation and lower SLX4 protein levels substantially (Fig. 5A; Fig. S4, A and B). SLX4 and SLX4IP colocalize, and the number of colocalization events increased upon APH-induced RS (Fig. 5A; Fig. S1B). SLX4IP foci were fully abrogated upon SLX4 knockdown; however, a SLX4IP pan-nuclear signal remained. Conversely, SLX4 foci were not significantly affected by SLX4IP depletion (Fig. S4A). Similarly, SLX4IP protein levels were decreased upon SLX4 depletion, while SLX4 was unaffected by SLX4IP depletion (Fig. S4B). Consistent with reported data, (39) SLX4 depletion increased G1 53BP1 N.B formation, but to a lesser extent than SLX4IP depletion (Fig. 5B). Importantly, co-depletion did not further increase the amount of 53BP1 N.Bs. Depletion of SLX4 did not impair cell proliferation, while both SLX4IP depletion alone and co-depletion significantly impaired proliferation (Fig. 5C). However, S-phase progression was directly impaired upon SLX4 depletion in a similar fashion to SLX4IP depletion (Fig. 5, D and E). This impairment was not further exacerbated upon co-depletion. Although this analysis is limited by the combined use of siRNA, these data suggest that SLX4IP has SLX4-dependent roles in maintaining genome stability and S-phase progression, and roles independent of SLX4 in maintaining cell proliferation.

Figure 5.

SLX4IP functions with SLX4 to maintain chromosome stability but also display SLX4-independent functions. A, Representative pictures of SLX4IP-, SLX4-and co-depleted U2OS cells. B, G1 53BP1 N.Bs in cells treated with indicated siRNA (mean ± SD; n = 3 biological replicates; >2000 cells analyzed per condition). C, proliferation of cells treated with indicated siRNA. Co-depleted conditions are pooled (mean ± SD; n = 2 biological replicates). D, S-phase EdU intensity and (E) representative EdU profiles from indicated transfections (mean; n = 2 biological replicates; 10′000 cells per condition are shown). Significance was assessed by Student's unpaired t-tests. Scale bar: 10 μM.

SLX4IP-deficient zebrafish display embryonic defects and are hypersensitive to crosslinking-agent DEB

Given the important role of SLX4IP in the maintenance of genome stability, we sought to establish the physiological consequences of SLX4IP loss in vivo. Zebrafish larvae function as an effective vertebrate model to assess in vivo consequences of altered DNA damage response and have proven valuable in unraveling FA gene functions (40, 41, 42). SLX4IP is conserved in zebrafish (Fig. S5A). A zebrafish mutant line with a 40 bp insertion in exon three was therefore generated using CRISPR/Cas9 mutagenesis, and this mutation results in encoding an early frameshift/truncated product (p.L18Rfs59∗), predicted to be a null allele. We performed transcript analysis to confirm the presence of the predicted mutation at the RNA level and to check for the presence of any insertion-induced in-frame, aberrant, splice products that may provide residual functional protein. We confirmed the expression of the 40 bp insertion mutant RNA, and sequencing of the cDNA transcript did not reveal any aberrant splice products (Fig. 6A). To determine whether slx4ip homozygous knockouts survive to adulthood, we grew progenies from heterozygous in cross to adulthood and determined their genotypes. The homozygous knockout fish were observed at the expected Mendelian ratio among the surviving adults (Fig. 6B), indicating no lethality at earlier developmental stages. Previously, female-to-male sex reversal has been observed in all 17 FA gene knockouts tested (40), suggesting FA pathway plays an essential role in gametogenesis. However, among the slx4ip homozygous knockout fish, we observed nearly an equal number of male and female fish (Fig. 6C), suggesting no sex reversal phenotype for slx4ip mutants. Length of the adult male fish was also not affected by slx4ip ablation (Fig. S5C).

Figure 6.

Evaluation of slx4ip knockout zebrafish for survival, growth, and response to a DNA crosslinking agent. A, image of wildtype (slx4ip+/+) and knockout (slx4ip−/−) allele RT-PCR products resolved on 2% agarose gel. The mutant allele product has expected 40 bp insertion and no other aberrant splice product. Bands 400 bp and 500 bp marked by ∗ and ∗∗, respectively. B, data from genotyped adults indicates normal Mendelian birth ratios for wildtype, heterozygous (slx4ip+/−) and knockout zebrafish. C, data from genotyped adults by gender, indicating no female-to-male sex reversal bias in knockouts. D, phenotypical analysis of zebrafish larvae of indicated genotypes reveal higher rates of yolk sac edema and eye size reduction in larvae devoid of slx4ip. E, representative images of zebrafish larvae of indicated genotype. The red arrow demonstrates yolk edema. F, results of heterozygous incross to assess hypersensitivity of embryos to the administration of 1.0 μg/ml DEB, with phenotype-based separation at 72 hpf. G, results of heterozygous fish crossed with knockout fish to assess hypersensitivity to administration of 0.8 μg/ml DEB, with phenotype-based separation at 120 hpf. H, representative images of progenies from wildtype incross and knockout incross imaged at 120 hpf, after administration of varying concentrations of DEB.

To assess the developmental morphology of the zebrafish larvae, an array of phenotypes was considered, including yolk edema, eye size, heart edema, fin- and body-related parameters at four dpf. Morphological aberrations are considered significant when the percentage of affected larvae is larger than 20%. Although with only partial penetrance, we found that eye size was affected by slx4ip ablation, and that yolk edema was significantly more frequent in slx4ip mutant larvae (Fig. 6, D and E). While other morphological abnormalities appear to occur with elevated frequency in slx4ip mutant larvae (Fig. S5B), these phenotypes do not meet the threshold of 20% penetrance and are thus not considered significant.

In cells, SLX4IP and SLX4 cooperate to maintain genome stability. Zebrafish larvae deficient in FA genes, including slx4, display strong sensitivity to the crosslinking agent DEB (40). To determine whether slx4ip null mutant exhibit hypersensitivity to DEB, we treated the embryos produced from various crosses starting at six hpf to varying time points, and at the end of treatment, separated them based on gross morphological changes. The embryos from inbred heterozygous were subjected to 1.0 μg/ml DEB treatment until 72 hpf. An increased number of homozygous knockouts exhibited moderate and severe morphological changes compared to wild-type and heterozygous embryos (Fig. 6F), suggesting hypersensitivity of knockout embryos to DEB treatment. In addition, the embryos produced from crossing of heterozygous fish with knockout fish were treated with 0.8 μg/ml DEB until 120 hpf. Strikingly, nearly all knockout embryos exhibited either moderate or severe morphological changes, whereas heterozygous embryos appeared normal (Fig. 6G). The longer duration of DEB treatment at lower dose appears to produce more distinctive genotype-based morphological changes in this cross compared to that from the heterozygous incross. We also tested the hypersensitivity of null incross and wild-type incross embryos to increasing concentrations of DEB (0.2 μg/ml −1.2 μg/ml). Nearly all knockout embryos exhibited severe phenotype starting from 0.6 μg/ml DEB, whereas the wildtype embryos did not exhibit severe phenotype until DEB concentration reached 1.2 μg/ml (Fig. 6H). All these experiments consistently show increased sensitivity of slx4ip knockout embryos to DEB treatment. Taken together, this suggests that functional slx4ip expression is needed for correct embryogenesis, and that its depletion confers a similar phenotype as FA genes upon treatment with DEB.

Discussion

Chromosomes carry inherently difficult-to-replicate regions such as centromeres, telomeres, and CFSs, which need specialized pathways to aid full chromosome duplication. Our findings reveal that the SLX4-associated protein SLX4IP is recruited to CFSs during interphase into prometaphase, and that its loss compromises CFS stability in a SLX4-dependent manner. We further note perturbed S-phase progression and proliferation of acutely SLX4IP-depleted U2OS cells and find that slx4ip null mutant zebrafish display elevated developmental aberrations. Finally, slx4ip null zebrafish embryos exhibit increased sensitivity to crosslinking agents, similar to effects observed upon ablation of FA genes.

Important work by Panier et al. (32) has demonstrated that SLX4IP maintains telomere hyper-recombination in ALT cell lines via direct interaction with both BLM and SLX4. Accordingly, loss of SLX4IP in U2OS is sufficient to increase ALT phenotypes such as ABPs and C-circles. Based on our data in U2OS, we conclude that the loss of SLX4IP is also sufficient to increase phenotypes associated with CFS under-replication, including EdU-positive metaphase chromosome breaks and G1 53BP1 N.Bs. This raises the possibility that processes of hyper-recombination in ALT-positive cells might not only be sequestered to telomeres but may also affect other repeat-rich DNA stretches with similar characteristics as telomeres, such as CFSs. Our co-depletion studies indicate that loss of SLX4 and SLX4IP function epistatically with respect to CFS instability, in contrast to previous observations that find an additive rather than epistatic effect of co-depletion on telomere stability and cell proliferation (32). These differences likely reflect methodological variation. Our method of co-depletion may not fully eliminate SLX4- and SLX4IP gene expression for additive effects to appear, and siRNA-mediated knockdown does not exclude compensatory expression. Additionally, short-term proliferation experiments may not capture longer-term synthetic phenotypes. Nevertheless, our results suggest that SLX4IP has broader roles in genome stability beyond telomere maintenance.

Generating slx4ip null mutants in zebrafish allowed us to demonstrate that loss of functional slx4ip alleles results in genomic instability as observed by the hypersensitivity of the embryo/larvae to treatment with DEB. The sensitivity to DEB is consistent with what was observed for the other FA gene mutants in zebrafish. Like null mutants for most FA genes, slx4ip nulls showed no defects in survival to adulthood, fertility, and growth differences. However, unlike most of the FA gene knockouts that showed partial or complete female-to-male sex reversal phenotype, slx4ip nulls did not appear to display the gender bias, perhaps indicating that slx4ip may not play a critical role in gonadogenesis. In general, zebrafish lacking functional slx4ip resemble those lacking a known disease-causing FA gene, prompting to speculate the candidacy of slx4ip as an FA gene.

SLX4IP is a conserved protein implicated in tumorigenesis, yet its molecular functions remain poorly understood. We and others (32) report that SLX4IP does not simply phenocopy SLX4 but has functions that extend from its scaffold-tethered state. Contrary to SLX4, SLX4IP loss increases MiDAS, indicating it is dispensable for mitotic repair and thus suggesting that SLX4IP-depleted cells enter mitosis with a higher burden of under-replicated CFSs. Although SLX4IP has not previously been directly linked to CFS stability, it has been associated with several DNA repair processes, including ICL repair and HR, both pathways that play vital roles in the maintenance and repair of CFSs (28). Notably, SLX4IP is reported to tether endonuclease XPF-ERCC1 to SLX4 for ICL resolution, and XPF is known to promote restart of stalled replication forks at secondary structures (43) which are frequent at CFSs. Loss of SLX4IP could therefore compromise fork restart and lead to an increase in under-replicated CFSs and replication difficulties. In addition, SLX4IP reportedly promotes promiscuous dissolution of HR intermediates at telomeres in ALT-cells (30). Perhaps HR intermediate integrity is lost at CFSs upon SLX4IP loss, leading to persistent under-replicated CFSs and genomic instability. Together, these observations support a model where SLX4IP maintains CFS stability by coordinating resolution of stalled forks and recombination intermediates through interaction with the SLX1-SLX4, MUS81-EME1, and XPF-ERCC1 (SMX) trinuclease complex. Further mechanistic dissection will be necessary to test this hypothesis directly.

Our study has several limitations. Although we observe that SLX4IP foci formation is induced by APH, its loss does not exacerbate chromosome instability under RS, suggesting unclear compensatory mechanisms. Knockdown and knockout models of SLX4IP yielded partially divergent phenotypes, possibly reflecting adaptation to chronic loss. Our findings are currently limited to U2OS, an ALT-positive cell line, and would benefit from additional work in other cellular backgrounds. Further mechanistic works and studies in diverse models will be important to clarify the full extent of SLX4IP in genome maintenance.

Experimental procedures

Cell culture

U2OS and HeLa cells were grown and maintained in Dulbecco’s Modified Eagle Medium (DMEM) (Gibco, 31966) supplemented with 6% fetal bovine serum (FBS) (Gibco, 6270) and 1% penicillin/streptomycin (PS) (Gibco, 15140). For the proliferation assays, H2B-GFP U2OS cells were maintained at similar conditions. GFP-SLX4IP Flp-In TREx U2OS cells were a kind gift from Ivan Muñoz and John Rouse. GFP-SLX4IP plasmid expression was maintained by selection with 10 μg/ml Hygromycin (Thermofisher, 10687010) and 2 μg/ml Blasticidin (Invivogen, ant-bl-5b). All cell lines were maintained at 37 °C in a 5% CO₂ incubator.

RNA interference

U2OS cells were reverse-transfected in DMEM with 6% FBS and no PS using 20 nM of indicated siRNAs in OptiMEM (Gibco, 31985) with transfection reagent DHarmaFECT (Dharmacon, T-2001–03) following the manufacturer’s instructions into a 6-well plate (Costar, 3516). Transfected cells were split into new medium after 24h at 70% confluency. All experiments were performed 72h post-transfection. All siRNA was diluted in dH₂O (Invitrogen, 10977). The following sequences of the siRNA sense strand were purchased from OriGene as a set (OriGene, SR315041) and used: UAACGACGCGACGACGUAA (siCtrl#1); universal scrambled negative control (OriGene, SR30004) (siCtrl#2); GAAAGAUUAUCUACAAUUCAGAACA (siSLX4IP#1); CUUUUAGCAAGUCAGAAUGAAGATT (siSLX4IP#2); CAUCACCAUGUCCAAAACAAAGUCC (siSLX4IP#3); GAGAAGAACCCUAAUGAAATT (siSLX4#1); GCACAAGGGCCCAGAACAATT (siSLX4#2). For epistatic analyses, 40 nM of total siRNA per condition was used.

CRISPR-generated SLX4IP KO clones

SLX4IP-KO U2OS cells were generated via the CRISPR-Cas9 system in accordance with Ran et al., 2013 (44). gRNA targeting SLX4IP exon 2 (forward: 5′-CACCGAGCAAATTTCTTAGATGCCA-3′, reverse: 5′-AAACTGGCATCTAAGAAATTTGCTC-3′) were used for SLX4IP KO#1, and gRNA targeting SLX4IP exon 8 (forward: 5′-CACCGTGTTTCCACAGTGTCTCTTC-3′, reverse: 5′-AAACGAAGAGACACTGTGGAAACAC-3′) were used for SLX4IP KO#2. The gRNA was then cloned into pSpCas9(BB)-2A-GFP vectors for co-transfection of gRNA and Cas9. U2OS cells were seeded into 6-well plates at 70% confluency and on the subsequent day, transfected with 1,25 μg of the pSpCas9(BB)-2A-GFP plasmids using FuGENE transfection reagent (Promega, E2311) following the manufacturer’s instructions. For controls EV#1 and EV#2, 1,25 μg of empty vector was used. After 24 h, single GFP-positive cells were sorted into clear 96-well plates with a BD FACSAria II cell sorter. After 2 weeks of expansion, the cells were screened by immunofluorescence microscopy and subsequent western blotting. Sequences and targets were verified by PCR.

Western blotting

For protein extract preparation, cells were washed twice in ice-cold PBS (gibco, 14190) and lysed using RIPA buffer (Sigma, R0278) with protease inhibitor (Roche), β-glycerophosphate, NaF and sodium orthovanadate. Sample protein amounts were quantified with a DC Protein Assay kit (Bio-Rad, 500–001). Adjusted samples were mixed with LDS 4x sample buffer (Thermofisher, NP0007) and DTT, and then boiled at 70 °C for 15 min. The samples were separated on a NuPAGE 4 to 12% Bis-Tris gel (Invitrogen, NP0321) in NuPAGE MOPS running buffer (Life technologies, NP0001) at 180 V for 90 min before being transferred to a PDVF membrane (Amersham, 10600001) at 350 mV for 2h. The membrane was then washed in PBST (PBS with 0.1% Tween-20 (Sigma, P2287)), blocked with 5% skimmed milk in PBST for 1h and incubated with primary antibodies overnight at 4 °C. On the subsequent day, the membrane was washed in PBST and incubated with appropriate HRP-conjugated secondary antibody for 1h at room temperature (RT). The membrane was developed using an ECL Prime Western Blot Detection Reagent kit (Amersham, RPN2232) for 1 min, and chemiluminescence was detected with an Amersham Imager 600. SeeBlue Plus2 prestained standard protein ladder (Invitrogen, LC5925) has been used for all western blots. For displayed western blots, the following primary antibodies have been used: 1:500 mouse monoclonal SLX4IP (SantaCruz, sc-377066); 1:500 sheep polyclonal SLX4 (University of Dundee, S586D); 1:2000 mouse vinculin (Sigma, V9131); 1:5000 mouse β-Actin (Sigma, A2228). HRP-conjugated secondary antibodies used include anti-Mouse IgG-peroxidase antibody (Sigma, A3682) and anti-Sheep IgG peroxidase conjugate (Sigma, A3415), both at dilution 1:10,000.

High-content microscopy immunofluorescence assays

For all high-content immunofluorescence (IF) experiments, 5000 cells were seeded into μClear Cellstar 96-well microscopy plates (Greiner Bio-one, 655090). After 24 h, the cells were treated with 0,2 μM aphidicolin (APH) (Sigma, A0781) or DMSO (Sigma, D2438) and fixed after 20h unless otherwise indicated. The plates were pre-extracted using 0,2% Triton X-100 (merck, X100) in PBS for 1 min on ice and then fixed by adding 4% formaldehyde (VWR Chemicals, 9713) for 15 min RT. The wells were blocked with blocking buffer (3% BSA (Sigma, A7030) in PBST) for 1h at RT and incubated with primary antibody in blocking buffer overnight at 4 °C. On the subsequent day, the wells were washed with PBST and incubated with appropriate Alexa Flour fluorescence-tagged secondary antibody for 2h at RT in dark. Afterwards, the wells were washed with PBST and incubated with 0,5 μg/ml DAPI in PBS for 10 min at RT in dark. 53BP1/Cyclin A staining was done through direct fixation with 4% formaldehyde for 15 min RT and subsequent permeabilization with 0,5% Triton X-100 in PBS. For EdU incorporation assays, 20 μM EdU was added 30 min prior to pre-extraction and the click-it reaction prior to the blocking step. Click-it reagent was prepared by adding freshly prepared ascorbic acid (100 nM), CuSO₄ (2 mM), azide 647 (Life technologies, 10277) to PBS, which was then added onto EdU-treated wells for 1h at RT in dark. All images were acquired through an Olympus IX-81 wide-field microscope. High-content image analysis and foci counts were performed with software Olympus ScanR Analysis 3.2 and exported to Tibco Spotfire v.10.3 for appropriate gating. For displayed IF, following primary antibodies have been used: 1:250 mouse SLX4IP (SantaCruz, sc-377066); 1:250 sheep SLX4 (University of Dundee, S587D); 1:500 rabbit TRF2 (novus, IMG-124A); 1:1000 rabbit FANCD2 (novus, NB100–182); 1:200 mouse Cyclin A (SantaCruz, sc-271682); 1:1000 rabbit 53BP1 (novus, NB100–304).

ChIP-qPCR

70% confluent 15 cm dishes with doxycycline-induced GFP-SLX4IP Flp-In TREx U2OS cells were used as starting material per condition. For preparation of sheared chromatin, the SimpleChIP Plus Enzymatic Chromatin IP Kit was employed following manufacturer’s instructions. Briefly, cells were treated with 0,2 μM APH or DMSO for 20h then crosslinked in 1% formaldehyde (Sigma, 252549) for 10 min and quenched with glycine. Cells were pelleted into tubes, washed with PBS, and lysed. For optimal DNA digestion, micrococcal nuclease was added and the tubes incubated for 20 min at 37 °C. The reaction was then quenched with EDTA, nuclei were pelleted, resuspended in ChIP buffer, and sonicated on a Bioruptor Plus (Hologic, BD1020001). Lysates were cleared and transferred to new tubes. For analysis of fragment length and total DNA content, a fraction of the lysate is cleared with Proteinase K and eluted using a Qiagen DNA Extraction Kit (Qiagen, 69504). Fragments corresponded to a length between 1 to 5 nucleosomes. For IP, the chromotek GFP-Trap Agarose Kit (chromotek, gta-10) were used following manufacturer’s instructions. 10 μg chromatin were added to the IPs, and the lysate was precleared with binding control agarose beads (chromotek, bab-20) before overnight pulldown. Beads were washed according to protocol and reverse crosslinked overnight with proteinase K. For DNA purification, kit NucleoSpin Gel and PCR Clean-up was used (Macherey-Nagel, 740609.250), and DNA is finally eluted in supplied elution buffer. qPCR is set up with SYBRGreen and following primers: FRA7H-F: TAATGCGTCCCCTTGTGACT, FRA7H-R: GGCAGATTTTAGTCCCTCAGC, GAPDH-F: CCCTCTGGTGGTGGCCCCTT, GAPDH-R: GGCGCCCAGACACCCAATCC.

Preparation and analysis of chromosome spreads

To obtain images of metaphase chromosome spreads, reverse transfected U2OS cells were seeded into 10 cm dishes, and after 24h, 0.4 μM APH or DMSO was added to the plates. To synchronize the cells in G2/M-phase, CDK1 inhibitor RO-3306 (Sigma, SML0569) was added 13h after APH. After an additional 6h, cells were washed twice in prewarmed PBS and once in DMEM before the cells were released in DMEM containing 0, 1 μg/ml of the spindle formation inhibitor Karyomax Colcemid (Gibco, 15212012) and EdU (20 μM) for 1h. After in total 20h, the DMEM was collected, and the cells were harvested with trypsin (Gibco). The cells were then pelleted, most of the medium was removed, and the cells were resuspended in the remaining medium. 5 ml prewarmed 75 mM KCl was added dropwise to the cells while vortexing, and the resuspension was incubated at 37 °C for 20 min. The cells were then pelleted again, and the medium was removed. 5 ml ice-cold, freshly prepared methanol:acetic acid (3:1) fix solution was used to fix the cells by resuspending dropwise before the solution was incubated at RT for 30 min and −20 °C overnight. On the same day, glass slides were washed in 96% ethanol (VWR 26 Chemicals, 20821.365), then dH₂O and finally incubated in dH₂O overnight. The next day, cells were pelleted, the solution was removed, and the pellet was resuspended in 200 μl fresh fix solution. The glass slides were prepared on a pre-chilled tray, and a drop of resuspended cells was dropped from a distance on the glass slides. The slides were then left to evaporate at RT overnight. The remaining solution was stored at −20 °C. The following day, the dry slides were immersed in PBS and shaken for 90 min. The slides were then fixed in 4% PFA in PBS for 4 min RT and washed in a 3% BSA-0,5% Triton-X 100 in PBS solution for 2 times 5 min. To visualize EdU, the slides were incubated with 100 μl click-it reaction mix (Molecular Probes, C10339) for 1h at RT in dark and then washed thrice with 3% BSA-0,5% Triton-X 100 in PBS. The spreads were then rinsed with PBS for 30 min RT, dried, and mounted using DAPI VectaShield mounting medium (Vector Laboratories, LS-J1033–10). The slides were finally sealed by a Menzel-Gläser coverslip with nail polish (MaxFactor, B009S2GHE4). For acquisition, an Olympus BX63 microscope with objective 60X/OIL was used. The images were acquired as Z-stack, and EdU foci, gaps and breaks were scored in ImageJ v.1.54f.

Zebrafish husbandry and ethics statement

All zebrafish experiments were performed in compliance with the National Institutes of Health guidelines for animal handling and research under NHGRI Animal Care and Use Committee (ACUC) approved protocol G-17-3 assigned to SCC. Zebrafish husbandry was performed as described previously (45).

Generation of slx4ip knockout mutants

Adult wildtype zebrafish (Danio rerio) were purchased from the European Zebrafish Resource Center (EZRC) and maintained at 26 to 28 C on a light cycle of 12h light: 12 dark. crRNAs with sequences TGAGCCGCGATGCTTCTGCT (#1) and CCACAGTTTTATTATCACAT (#2) were bought from IDT Biotechnology (Alt-R CRISPR-Cas9 crRNA, 2 nmol), as were Cas9 endonuclease and tracrRNA. Cas9/sgRNA ribonucleoprotein complexes were assembled according to the manufacturer’s instructions. One-cell-stage embryos were injected with a mix containing the Cas9/sgRNA complexes as described by Essner, J. on the integrated DNA Technologies platform.

DNA and RNA analysis of zebrafish slx4ip knockout mutants

Isolation of DNA from zebrafish and identification of their genotype using fluorescent PCR (fPCR) is as described earlier (40). The forward (TGTAAAACGACGGCCAGTaaggggatgatgtgacttgc) and reverse (GTGTCTTtgtgggccaatgaaaacata) primers were used for determining the wildtype, heterozygous, and homozygous knockout slx4ip genotypes. The caudal fin tissue of adult wildtype and homozygous knockout fish obtained by the ACUC-approved fin clip method were used for RNA extraction using standard TRI Reagent (Ambion) protocol following tissue homogenization with a Ribolyzer (MP Biomedicals). The RNA purification and cDNA synthesis were performed using RNeasy mini kit (Qiagen) and the Superscript IV First-Strand Synthesis system kit (Invitrogen), respectively, as per the manufacturer’s instructions. The slx4ip RT-PCR primers, e x .2_F (TGGGGGAAATCTAACGACTG) and e x .6 (AACACGCTCGGGAAATACAC) were designed to amplify the exon three containing the indel mutation along with its flanking exons. The fPCR and RT-PCR products were analyzed on an agarose gel, and their sequence was determined by Sanger’s dideoxy sequencing method.

Evaluation of mutant adult fish for viability, female-to-male sex reversal, and body length.

The progenies generated from inbreeding of heterozygote mutant fish were grown to adulthood and their DNA was genotyped using fPCR method using forward (TGTAAAACGACGGCCAGTaaggggatgatgtgacttgc) and reverse (GTGTCTTtgtgggccaatgaaaacata) primers. Deviations from expected number of nulls were tested with goodness-of-fit Chi-square statistical analysis. To determine the presence of both sexes among surviving adults, all genotyped fish were categorized as males and females and counted. The standard body length data from an adult fish were collected as described (40). The body length data was collected from wildtype and null progenies from inbred heterozygote mutant fish and analyzed using unpaired t test.

Evaluation of developmental defects of zebrafish larvae

Nine phenotypes (yolk edema, eye size defects, heart edema, body length, fin absence, inner ear abnormalities, tail bending, scoliosis, body axis abnormalities) were analyzed at the stage of 96 h post-fertilization (hpf). At four dpf, larvae derived from a cross of two slx4ip heterozygous larvae were anesthetized and placed in a robotic microfluidic system (VAST: Vertebrate Automated Screening Technology, Union Biometrica), which allows the automatic aspiration, placement, and rotation of the larvae. Four images, two lateral, one dorsal, and one ventral, were taken per larva. The nine phenotypes were quantified as Yes/No according to the presence or absence of the phenotype. All teratogenic phenotypes were scored as 0 = normal larva or 1 = affected larva. Phenotypes based on quantitative measurements (heart edema, eye size defects, embryo size defects) were measured using FIJI and compared to respective controls. In that case, larvae with parameters 2 standard deviations from the population mean were considered to be affected. Teratogenic phenotypes are considered statistically significant when the percentage of affected larvae is >20%.

Evaluating the sensitivity of zebrafish slx4ip null larvae to DEB treatment

The larvae obtained from indicated breeding crosses were treated with diepoxybutane (DEB) (Sigma Aldrich) at indicated concentrations in egg water with methylene blue starting six hpf. The embryos were separated at the end of treatment into three groups based on the severity of the observed morphological changes (normal, moderate and severe) and genotyped using fPCR primers as described earlier (40). Images of DEB-treated embryos representing homozygous mutant incross were taken using LAS X Imaging software on a Leica M205 microscope with a DFC7000 color camera.

Sequence alignment and conservation

All multiple sequence alignment and conservation score output was done through Jalview 2.11.2.0.

Statistical analyses

Statistical analyses have been performed using RStudio and GraphPad Prism 8. For significance testing of differences in means, two-tailed unpaired Student’s t-tests were used without correction for multiple comparisons. Applied statistical test is annotated in figure legends, and p-values of significant or near-significant comparisons are indicated in figures.

Data availability

Raw, uncropped western blots and statistical source data for all panels in main Figures 1-5 have been deposited to a public Mendeley Data repository and can be accessed via the following DOI: https://doi.org/10.17632/yhg5dkx2xx.1. All other data that support the findings of this study are available upon reasonable request to the corresponding author (andres.lopez@cabimer.es).

Supporting information

This article contains supporting information.

Conflict of interest

The authors declare that they have no conflicts of interest with the contents of this article.

Reviewed by members of the JBC Editorial Board. Edited by Patrick J. O'Brien