Coincidence of L1 retrotransposition and DNA repair deficiency elicits synthetic lethality in S-phase mature neurons Authors: Fumio Takahashi, Takashi Yamamura, Shinji Oki Published: 2026 ABSTRACT SummaryNeurodegeneration involves the entangled processes of cell-autonomous and non-cell-autonomous neuronal cell death, which leads to a collapse in the integrity of the neural network and causes behavioral symptoms. Here, we demonstrate that aberrant cell cycle re-entry (CCR) is prominent in mature neurons and that the replication fork acts as a target site for long interspersed nuclear element-1 (L1) retrotransposition during neurodegeneration. The fatal susceptibility of S-phase neurons in 5xFAD mice is attributed to a DNA repair deficiency in response to L1-mediated replication stress. Impaired sirtuin 6 expression seems to allow stochastic L1 activation and enhanced retrotransposition. Reduced estrogen/prolactin signaling correlates well with reduced Brca1 expression that is thought to protect neurons from replication stress. Importantly, L1-mediated pathogenesis correlates well with conventional Alzheimer’s disease pathology. In summary, the combination of enhanced L1 retrotransposition and DNA repair deficiency elicited synthetic lethality in S-phase neurons, highlighting a previously unknown pathogenic mechanism of neurodegeneration. Introduction Neurodegeneration comprises the entangled processes of cell-autonomous and non-cell-autonomous neuronal cell death (NCD), which leads to the collapse of neural networks and distinctive behavioral symptoms. Despite intensive causative research, a mechanistic overview of neurodegeneration remains elusive. Neurodegenerative diseases (NDs) are conventionally defined by the type of protein aggregates, and each ND is considered to stem from a discrete pathogenesis. Recently, several pathologies, including neuroinflammation, mitochondrial dysfunction, and DNA damage responses (DDRs), have been recognized as common pathophysiologies in diverse NDs.1,2,3,4 However, their pathogenic significance is not necessarily clear, and integrated insights into the potent interplay between several common ND-associated pathophysiologies are still lacking. Most research on NDs has focused on how protein aggregates are formed as a direct and irreversible cause of NCD, and detailed analysis of the molecular mechanism of NCDs after protein aggregation toward NCD has long been overlooked. We previously showed that the L1-encoded ORF1 protein that accumulates after aberrant neuronal L1 reactivation acts as a prototypic antigen for Eomes-positive helper T (Eomes+ Th) cells (immune-mediated neurodegeneration) and exerts neurotoxicity in patients and animal models of amyotrophic lateral sclerosis (ALS) and Alzheimer’s disease (AD).5,6 Interestingly, stochastic L1 reactivation and L1 retrotransposition are considered potent causes of DDRs, genomic instability, and NCDs in diverse NDs.7,8,9 Half a million copies of long interspersed nuclear element-1 (LINE-1 or L1) loci account for 17% of the human genome, and approximately 100 L1 retrotransposons are predicted to be retrotransposition competent.10,11 A typical L1 element is 6 kb long and encodes two proteins, ORF1 and ORF2. ORF1 stabilizes L1 transcripts as an RNA-binding protein, and ORF2 is an endonuclease/reverse transcriptase that enables the insertion of L1 complementary DNA into the host genome via target-primed reverse transcription. Mechanistically, ORF2 preferentially recognizes two typical DNA structures: hairpin templates that fold back on themselves, and the initial cleavage of a single-stranded DNA substrate close to the junctions between single-and double-stranded DNA,12,13 the latter of which is commonly formed during DNA replication. Accordingly, replication forks have been reported to be preferential target sites for L1 retrotransposition.14,15,16 In line with the L1 tropism of retrotransposition toward the replication fork, accumulating evidence suggests that a significant proportion of mature neurons in diverse NDs show an increased frequency of partial cell cycle re-entry (CCR). Although post-mitotic mature neurons are considered not to revert to the cell cycle once terminally differentiated and after cell division has ceased, approximately 10% of mature neurons in the normal brain have been demonstrated to be hyperploid.17 In addition, neuronal CCR is frequent in AD neurons; 20%–30% of them are hyperploid and are distributed in the middle of the G1/S phase.18,19,20,21 As mature neurons never undergo cell division to restore aneuploidy, the persistent formation of replication forks is associated with recurrent L1 attacks in such hyperploid neurons. Accordingly, a clear cell cycle bias has been observed for L1 reactivation and L1 retrotransposition, which peaks during the S phase.22 Therefore, the persistent interaction of the L1 retrotransposition intermediate gives rise to a stalled replication fork and induces DDRs, unless properly resolved by the DNA repair machinery. However, the molecular mechanism underlying this aberrant cell cycle progression and L1 retrotransposition targeting the replication fork in hyperploid neurons has not been well elucidated as a possible pathogenic mechanism for NCD and neurodegeneration. In this study, we integrated the aforementioned common defects across NDs, including L1 reactivation/retrotransposition and CCR, using 5xFAD mice as a model of neurodegeneration. Here, we provide evidence that the coincidence of L1 retrotransposition and impaired operation of the DNA repair machinery is the key to inducing synthetic lethality in S-phase neurons, which provides a novel molecular mechanism for L1-mediated autonomous NCD in NDs. Results Aberrant CCR is associated with L1 retrotransposition and acquisition of damaged phenotype in mature neurons of 5xFAD mice Aberrant CCR in mature neurons has been described in the context of aging and NDs,22,23,24 and a clear cell cycle bias has been observed for L1 retrotransposition, which peaks during the S phase.22 Furthermore, L1 reactivation is commonly observed in NDs.7,8,9 Therefore, we examined whether CCR and L1 retrotransposition are involved in neurodegeneration. We compared the cell cycle status of mature neurons derived from 5xFAD mice and their littermate controls using flow cytometry. As shown in Figure 1A, the number of neurons in the G0/G1 phase was comparable; however, a significant fraction of 5xFAD neurons were distributed in the S phase. To confirm the enhanced CCR in 5xFAD neurons, we crossed 5xFAD mice with FUCCIS/G2-M (mAG-hGem) mice (a derived strain of FUCCI mice) containing an expression cassette of the gene for geminin fused to the gene for the monomeric version of Azami Green fluorescent protein (mAG).25,26 As shown in Figure 1B, we observed a higher frequency of Neu N+ cells coexpressing geminin. These data further indicated an increased frequency of CCR in 5xFAD neurons.Figure 1Cycle re-entry and L1 retrotransposition in association with selective NCD of 5xFAD mouse brains(A) Neurons isolated from 3-month-old mice were stained with Hoechst 33342 and cell cycle phase distributions were evaluated by flow cytometry for live neurons. Data are represented as mean ± SD. Individual p values were shown in the image. Statistical analysis was performed by two-way ANOVA with Tukey’s multiple comparisons test. ∗∗∗p < 0.001.(B) Brain sections of 5xFAD/mAG-hGem mice or mAG-hGem mice were stained with anti-NeuN Ab (Red) and DAPI (Blue). mAG-hGem+/NeuN+ cells were visualized and quantified. Scale bars: 200 μm.(C) Frequency of L1 retrotransposition was analyzed by flow cytometry for neurons in each cell cycle phase. Representative flow cytometry data for GFP expression and live/dead status are shown in a pseudocolor plot (left). GFP+ cells in live or dead neurons were quantified in each cell cycle fraction (right). Data are represented as mean ± SD. Statistical analysis was performed by unpaired Student’s t test. ∗p < 0.05.(D) Immunofluorescence images of L1-GFP/5xFAD or L1-GFP brain stained with anti-geminin (red) and DAPI (blue). The relative frequencies of GFP+/geminin+ cells were quantified by automated cell counting. Scale bars: 200 μm. Representative summary data (n = 3) of three independent experiments with similar results are shown as mean ± SD. Statistical analysis was performed by unpaired Student’s t test. ∗∗p < 0.01.(E) Annexin V+ or γH2AX + S-phase neurons within GFP+ fractions were quantified by flow cytometry. Representative data (n = 3) from 2 to 3 independent experiments with similar results are shown as mean ± SD. Statistical analysis was performed by unpaired Student’s t test. ∗∗p < 0.01, ∗p < 0.05. Cycle re-entry and L1 retrotransposition in association with selective NCD of 5xFAD mouse brains (A) Neurons isolated from 3-month-old mice were stained with Hoechst 33342 and cell cycle phase distributions were evaluated by flow cytometry for live neurons. Data are represented as mean ± SD. Individual p values were shown in the image. Statistical analysis was performed by two-way ANOVA with Tukey’s multiple comparisons test. ∗∗∗p < 0.001. (B) Brain sections of 5xFAD/mAG-hGem mice or mAG-hGem mice were stained with anti-NeuN Ab (Red) and DAPI (Blue). mAG-hGem+/NeuN+ cells were visualized and quantified. Scale bars: 200 μm. (C) Frequency of L1 retrotransposition was analyzed by flow cytometry for neurons in each cell cycle phase. Representative flow cytometry data for GFP expression and live/dead status are shown in a pseudocolor plot (left). GFP+ cells in live or dead neurons were quantified in each cell cycle fraction (right). Data are represented as mean ± SD. Statistical analysis was performed by unpaired Student’s t test. ∗p < 0.05. (D) Immunofluorescence images of L1-GFP/5xFAD or L1-GFP brain stained with anti-geminin (red) and DAPI (blue). The relative frequencies of GFP+/geminin+ cells were quantified by automated cell counting. Scale bars: 200 μm. Representative summary data (n = 3) of three independent experiments with similar results are shown as mean ± SD. Statistical analysis was performed by unpaired Student’s t test. ∗∗p < 0.01. (E) Annexin V+ or γH2AX + S-phase neurons within GFP+ fractions were quantified by flow cytometry. Representative data (n = 3) from 2 to 3 independent experiments with similar results are shown as mean ± SD. Statistical analysis was performed by unpaired Student’s t test. ∗∗p < 0.01, ∗p < 0.05. We previously reported ectopic L1 reactivation and CCR in 5xFAD and mSOD1 mice.5,6 We examined the correlation between L1 reactivation and CCR in 5xFAD neurons. L1-GFP27/5xFAD mice, in which cells undergoing L1 retrotransposition were visualized using GFP expression, were analyzed for neuronal GFP expression during individual cell cycles using flow cytometry. The GFP+ signal was significantly increased in neurons with aberrant CCR in L1-GFP/5xFAD mice compared to that in L1-GFP mice (Figure 1C). Remarkably, dead S-phase neurons were more frequent in L1-GFP/5xFAD mice than in the littermate controls, which contained more GFP+ cells (Figure 1C), suggesting that L1 retrotransposition may induce selective cell death in S-phase neurons. Next, we examined the distribution of geminin-expressing neurons in the cortical region of 3-month-old L1-GFP/5xFAD mice. We identified few L1-GFP+ cells with no geminin expression in the cortex of L1-GFP mice. In contrast, there were many L1-GFP+ neurons distributed in the cortices of L1-GFP/5xFAD mice (Figure 1D, upper). In the cortical layer V, GFP+ and geminin-expressing cells were detected only in L1-GFP/5xFAD mice (Figure 1D, lower). Similarly, we detected L1-GFP+ geminin-expressing neurons in the subiculum, but not in the hippocampus, of 5xFAD mice (Figure S1). These data suggested that neurons with aberrant CCR and L1 retrotransposition reside in the cortical region and subiculum of the L1-GFP/5xFAD brain. Notably, during disease progression, the geminin signal extended to the cytosol of neurons in 6-month-old 5xFAD/mAG_h-Gem mice, and these geminin+ cells co-expressed caspase-3 (Figure S2). Intriguingly, a significant proportion of neurons from 6-month-old L1-GFP/5xFAD mice showed reduced reactivity to Neu O, a lipophilic probe incorporated into the membranes of live neurons, and those with reduced Neu O staining lost reactivity to anti-NeuN Abs (Figure S3), suggesting that neurons with CCR may lose NeuN antigen expression upon disease progression. Because hyperploidy enhances the susceptibility of mammalian cells to DDRs28 and a number of hyperploid, post-mitotic, mature neurons were stained with Hoechst 33342 or expressed GMNN. (Figures 1A and 1D), we evaluated DDRs in those GFP+ neurons with aberrant CCR. As shown in Figure 1E, GFP+ S-phase neurons in L1-GFP/5xFAD mice contained a higher frequency of neurons stained with annexin V, a probe for the detection of phosphatidylserine in the outer plasma membrane of pro-apoptotic cells. Furthermore, S-phase neurons of 5xFAD mice showed significant accumulation of γH2AX, a marker of DNA double-strand breaks (DSBs) (Figure 1E), and significant numbers of GFP+ 5xFAD neurons, but not of GFP+ littermate neurons coexpressing caspase-3 (Figure S4). Together, these data indicate that aberrant CCR is tightly associated with L1 retrotransposition and acquisition of the DDR phenotype, which confers pro-apoptotic/apoptotic features to mature neurons in 5xFAD mice. RNA-seq analysis of differential gene expression in S-phase neurons Our data suggest that L1 retrotransposition and the subsequent accumulation of DNA damage may induce the selective vulnerability of S-phase neurons to cell death, as significant numbers of GFP+ 5xFAD neurons coexpress caspase-3 (Figure S4). To elucidate the molecular basis of L1 retrotransposition-induced neuronal cell status, we performed RNA sequencing (RNA-seq) analysis of S-phase neurons derived from 3-month-old L1-GFP/5xFAD or L1-GFP mice with or without GFP expression. We confirmed L1 reactivation status in the individual fraction and confirmed exclusive L1 reactivation in GFP+ S-phase neurons of 5xFAD mice (Figure S7). A heatmap illustrating differentially expressed genes (DEGs) in S-phase neurons highlighted the distinctive clusters of unique gene upregulation in each fraction (Figure 2A). Relatively few DEGs were found in neurons from L1-GFP mice, and a remarkable difference was observed in the genes related to cell cycle checkpoints, which were selectively upregulated in GFP+ cells (Figures S5 and S6). Meanwhile, S-phase neurons derived from 5xFAD mice contained a large cluster of DEGs in both the GFP+ and GFP− fractions. Interpretive phenomenological analysis (IPA) revealed that GFP+ neurons from L1-GFP/5xFAD mice highlighted a canonical pathway of active nucleic acid metabolism compared to those from L1-GFP mice (Figure 2B), suggesting active DNA synthesis after the transition to the S phase. Neddylation, a post-translational modification with the ubiquitin-like protein Nedd8 (neural precursor cell expressed, developmentally downregulated protein 8) to substrate proteins and a well-recognized event downstream of amyloid pathology of AD, was one of the top canonical pathways when comparing GFP+ 5xFAD neurons and GFP+ control neurons (Figure S8). In contrast, a comparison between L1-GFP/5xFAD and L1-GFP mice-derived GFP− neurons highlighted the term “cellular effect of sildenafil,” a cellular status after the inhibition of phosphodiesterase 5 (PDE5) (Figure S9). Together, neurons from L1-GFP/5xFAD mice showed distinct gene expression, distinguished by increased gene expression related to nucleic acid metabolism and neddylation in GFP+ neurons and a cellular phenotype conferred by PDE5 inhibition in GFP− neurons.Figure 2RNA-seq analysis followed by pathway analysis for S-phase neurons(A) Heatmap of differentially expressed genes (DEGs) based on RNA-seq data in sorted GFP+ and GFP− S-phase neurons derived from L1-GFP/5xFAD or L1-GFP mice (comparison of pooled samples from three 3-month-old mice). p < 0.05 was considered significant by two-tailed t tests. GEO: GSE300383.(B) Interpretive phenomenological analysis (canonical pathway) between GFP+ and GFP− neurons derived from L1-GFP mice. RNA-seq analysis followed by pathway analysis for S-phase neurons (A) Heatmap of differentially expressed genes (DEGs) based on RNA-seq data in sorted GFP+ and GFP− S-phase neurons derived from L1-GFP/5xFAD or L1-GFP mice (comparison of pooled samples from three 3-month-old mice). p < 0.05 was considered significant by two-tailed t tests. GEO: GSE300383. (B) Interpretive phenomenological analysis (canonical pathway) between GFP+ and GFP− neurons derived from L1-GFP mice. Impaired expression of DNA repair factors is closely correlated with vulnerability of L1-retrotransposed S-phase neurons To clarify the molecular basis of the increased susceptibility to L1 retrotransposition in S-phase neurons of 5xFAD mice, we performed IPA and identified the possible involvement of Bard1 (Brca1-associated RING domain protein), an early-onset breast cancer susceptibility gene, as one of the genes exclusively downregulated in the GFP+ S-phase neurons of 5xFAD mice (Figure 3A). Bard1 is an obligatory partner of Brca1 (breast cancer susceptibility gene 1), which controls the DNA repair system and resolves stalled replication forks after DDR, nuclease attack/DSB, and subsequent attrition.29 Thus, we examined genes related to the DNA repair machinery in S-phase neurons of 5xFAD mice. Surprisingly, the expression of 21 DNA repair-related genes, including Brca1 and Bard1, was significantly suppressed in the GFP+ neurons of 5xFAD mice compared to that in GFP+ control neurons (Figure 3B). These included DSB repair factors (Atm, Rad51, Palb2, Hus1, Brca1, Rad51ap1, Bard1, Mapk8, Rmi2, and Rfc3), Fanconi anemia-related factors (Rmi1, Ube2t, Poln, Fancd2, Fancf, Top3a, and Cemps), nucleotide excision repair factors (Uvssa and Ercd8), and base excision repair factors (Neil3 and Ung). We confirmed the reduced expression of the representative genes, Brca1, Palb2, Bard1, and Rad51, by quantitative reverse transcription PCR (RT-qPCR) (Figure 3C), and a significant reduction in the protein expression of Brca1 in neurons from 5xFAD mice, especially in the S and G2/M phases, compared with those of littermates (Figure 3D). While playing a pivotal role in DNA repair and cell cycle checkpoints, Brca1 is known to protect stalled replication forks during replication stress. As the mode of Brca1-mediated stalled fork protection is distinct from that of general homologous recombination,30 we focused on Brca1/Bard1 rather than other DDR genes. Therefore, we hypothesized that the malfunction of Brca1/Bard1 may cf. the preferential fragility of 5xFAD neurons against L1 retrotransposition-induced NCDs. To further investigate the fate of L1-GFP+ 5xFAD neurons with reduced Brca1/Bard1 expression, we examined the live/dead status of S-phase neurons with or without Brca1 expression. GFP+ 5xFAD neurons that lacked Brca1 expression had a significantly higher proportion of dead cells, suggesting that Brca1 plays an important role in neuroprotection against L1 retrotransposition-induced DDRs, and that a defect in Brca1 expression confers vulnerability to NCDs in S-phase neurons in 5xFAD mice (Figure 3E).Figure 3Neuroprotective properties of Brca1 against L1 retrotransposition-induced DNA damage and subsequent cell death in S-phase neurons(A) Interpretive phenomenological analysis (Brca1 in protein ubiquitination) between GFP+ neurons derived from L1-GFP/5xFAD or L1-GFP mice.(B) Heatmap of DNA repair genes downregulated in GFP+ S-phase neurons from L1-GFP/5xFAD mice (left). Scattered plot of comparative gene expression between GFP+ or GFP− S-phase neurons (right). The 22 genes described in the main text are highlighted as red dots.(C) Relative expression of representative DNA repair-related genes was quantified by RT-qPCR.(D) Neurons were analyzed for their Brca1 protein expression in individual cell cycle phases by flow cytometry.(E) S-phase neurons with or without Brca1 expression were analyzed for their live/dead status by flow cytometry.Representative data (n = 3) from 2 to 3 independent experiments with similar results are shown, except for (A) and (B). Data are represented as mean ± SD. Statistical analysis was performed by one-way ANOVA with Tukey’s multiple comparisons test for (C) and (E) and two-way ANOVA with Tukey’s multiple comparisons test for (D). ns, not significant; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. Neuroprotective properties of Brca1 against L1 retrotransposition-induced DNA damage and subsequent cell death in S-phase neurons (A) Interpretive phenomenological analysis (Brca1 in protein ubiquitination) between GFP+ neurons derived from L1-GFP/5xFAD or L1-GFP mice. (B) Heatmap of DNA repair genes downregulated in GFP+ S-phase neurons from L1-GFP/5xFAD mice (left). Scattered plot of comparative gene expression between GFP+ or GFP− S-phase neurons (right). The 22 genes described in the main text are highlighted as red dots. (C) Relative expression of representative DNA repair-related genes was quantified by RT-qPCR. (D) Neurons were analyzed for their Brca1 protein expression in individual cell cycle phases by flow cytometry. (E) S-phase neurons with or without Brca1 expression were analyzed for their live/dead status by flow cytometry. Representative data (n = 3) from 2 to 3 independent experiments with similar results are shown, except for (A) and (B). Data are represented as mean ± SD. Statistical analysis was performed by one-way ANOVA with Tukey’s multiple comparisons test for (C) and (E) and two-way ANOVA with Tukey’s multiple comparisons test for (D). ns, not significant; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. Brca1 is critically involved in DNA repair and survival of S-phase neurons after replication fork-targeted L1 attack As concurrent downregulation of DDR-related genes and enhanced frequency of L1 retrotransposition were extensively observed in S-phase neurons of 5xFAD mice, we focused on the functional involvement of Brca1/Bard1 in replication fork protection and investigated the functional relevance of reduced Brca1 expression and increased L1 retrotransposition in S-phase cells in terms of promoting the cell death cascade. Brca1 expression was significantly upregulated during the transition to S-phase in SH-SY5Y neuroblastoma cells after treatment with aphidicolin for 24 h (Figure 4A). To evaluate L1 retrotransposition in vitro, we employed SH-SY5Y cells transduced with an L1-GFP reporter construct (L1-GFP/SH-SY5Y cells).31 As shown in Figure 4B, aphidicolin treatment increased GFP+ L1-GFP/SH-SY5Y cells, which were co-stained with annexin V. Annexin V binds to phosphatidylserine exposed to the outer leaflet of the plasma membrane that occurs at the early stage of apoptosis. These data suggest that L1-GFP/SH-SY5Y cells arrested at S phase were more vulnerable to spontaneous L1 attack and accumulation of DDRs. To confirm the protective effect of Brca1 against accumulation of annexin V+ cells, we transfected L1-GFP/SH-SY5Y cells with Brca1-specific small interfering RNA (siRNA) prior to aphidicolin treatment. Brca1 protein expression was significantly decreased after treatment with Brca1-specific siRNA (Figure S10). As shown in Figure 4C, Brca1-specific siRNA treatment significantly increased the accumulation of annexin V+ cells compared to that in control siRNA-treated cells.Figure 4Neuroprotective properties of Brca1 against L1 retrotransposition-induced DDR(A) Brca1 expression in SH-SY5Y cells with or without aphidicolin treatment was quantified by RT-qPCR (n = 3).(B) L1-GFP/SH-SY5Y cells with or without aphidicolin treatment were stained with annexin V and analyzed by flow cytometry. Typical plot (left) and representative summary data (right) are shown.(C) Workflow of the experiment (top). Relative frequency of annexin V+ cells by immunohistochemistry (bottom). Respective data of 3 independent experiments with similar results are shown.(D) (Top) Workflow of the experiment. Proportion of annexin V+ cells within whole live cells (bottom left). Proportion of dead cells within whole cells (bottom right).(E) DNA:RNA immunoprecipitation data of genomic DNA purified from aphidicolin-treated SH-SY5Y cells. L1 transcripts with or without RNase A treatment before cDNA synthesis were quantified by RT-qPCR with ORF1- or ORF2-specific primers.(F) Chromatin immunoprecipitation employing anti-Polδ Ab of genomic DNA from SH-SY5Y cells with or without aphidicolin treatment. After digestion with restriction enzymes (HindIII or HaeIII), L1 transcripts were amplified by using ORF1- and ORF2-specific primers.Respective data of 2–3 independent experiments with similar results are shown as mean ± SD. Statistical analysis was performed by unpaired Student’s t test for (A), (B), and (C), and one-way ANOVA with Tukey’s multiple comparisons test for (D). ns, not significant; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Neuroprotective properties of Brca1 against L1 retrotransposition-induced DDR (A) Brca1 expression in SH-SY5Y cells with or without aphidicolin treatment was quantified by RT-qPCR (n = 3). (B) L1-GFP/SH-SY5Y cells with or without aphidicolin treatment were stained with annexin V and analyzed by flow cytometry. Typical plot (left) and representative summary data (right) are shown. (C) Workflow of the experiment (top). Relative frequency of annexin V+ cells by immunohistochemistry (bottom). Respective data of 3 independent experiments with similar results are shown. (D) (Top) Workflow of the experiment. Proportion of annexin V+ cells within whole live cells (bottom left). Proportion of dead cells within whole cells (bottom right). (E) DNA:RNA immunoprecipitation data of genomic DNA purified from aphidicolin-treated SH-SY5Y cells. L1 transcripts with or without RNase A treatment before cDNA synthesis were quantified by RT-qPCR with ORF1- or ORF2-specific primers. (F) Chromatin immunoprecipitation employing anti-Polδ Ab of genomic DNA from SH-SY5Y cells with or without aphidicolin treatment. After digestion with restriction enzymes (HindIII or HaeIII), L1 transcripts were amplified by using ORF1- and ORF2-specific primers. Respective data of 2–3 independent experiments with similar results are shown as mean ± SD. Statistical analysis was performed by unpaired Student’s t test for (A), (B), and (C), and one-way ANOVA with Tukey’s multiple comparisons test for (D). ns, not significant; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Although our data indicate that Brca1 plays a crucial role in protection against L1 retrotransposition-induced DDRs, we detected minimal authentic aspects of death in SH-SY5Y cells, even after treatment with Brca1 siRNA, which is in sharp contrast to the apparent cell death observed in GFP+ S-phase neurons of 5xFAD mice. Considering that the successful retrotransposition of a single copy of the L1-GFP cassette is sufficient to induce GFP expression, the frequency of stochastic L1 retrotransposition in normal cultured cells may not be sufficient to induce cell death via DDRs. We then treated the cells with bleomycin, which induces DNA strand scission protected by Brca132 to mimic nuclease-induced DSB responses. The combination of Brca1 siRNA and bleomycin increased the annexin V+ population in live cells and significantly increased the number of dead cells (Figure 4D). These results suggest that the combination of reduced Brca1 expression and DSBs, which may be caused by L1 retrotransposition in S-phase neurons, is sufficient to induce synthetic lethality, although we do not exclude the possibility that bleomycin may have additional effects. Interaction of the L1 transcript with the replication fork promotes the configuration of putative DNA:RNA triplet structures recognized by the S9.6 Ab.33 As 5xFAD-derived CCR mature neurons are in the very early stages of S phase with a limited number of replication forks, we explored whether the L1 transcript preferentially targets the replication fork in S-phase cells. As shown in Figure 4E, the L1 transcript was detected after DNA-RNA immunoprecipitation (DRIP) with S9.6 Ab. However, RNase A treatment before immunoprecipitation moderately decreased the level of the L1 transcript, possibly because of the dense association of ORF1 with the L1 transcript, which may interfere with effective RNase activity. Next, we applied chromatin immunoprecipitation (ChIP) analysis using anti-DNA PolδAb, a replication fork-specific DNA polymerase involved in the synthesis of lagging strand DNA, because L1 endonuclease shows a preference for cleaving the lagging strand DNA of replication templates.34 Genomic DNA derived from aphidicolin-treated or untreated SH-SY5Y cells was immunoprecipitated with anti-DNA PolδAb after enzymatic digestion with either HindIII (6-base cutter) or HaeIII (4-base cutter). As shown in Figure 4F, we detected L1 transcripts corresponding to the ORF1 and ORF2 regions in aphidicolin-treated cells but not in untreated cells. These data suggest that cells stalled in the early to middle phases of cell cycle progression are susceptible to L1 attacks at replication forks, which may cause replication conflicts and promote cell death. Estrogen/Prl axis as a potent upstream regulator of neuronal Brca1 expression Since we revealed that Brca1 plays a critical role in the cellular stabilization of aneuploid neurons and that the inhibition of Brca1 provoked them into DDRs via L1 retrotransposition-induced replication fork stress, we next explored the upstream regulators critical for sufficient Brca1 expression. IPA showed that hormone and hormone receptor genes were possible candidates (Figure 5A). As Brca1 mutation carriers pose a higher risk for estrogen-responsive tissues, such as the breast and ovary, we focused on estrogen and prolactin (Prl), which have been reported to induce BRCA1 expression in human breast cancer cell lines.35 As shown in Figure 5C, Prl treatment of isolated neurons in vitro significantly upregulated Brca1 expression in a dose-dependent manner. Furthermore, CCR-associated upregulation of Brca1 and FancD2 was compromised in the neurons of 5xFAD/PrlR heterozygous mice (Figure 5D). Interestingly, aphidicolin treatment alone was sufficient to induce Prl expression in SH-SY5Y cells (Figure 5E). Additionally, SH-SY5Y cells pretreated with PrlR-specific siRNA showed increased reactivity to γH2AX after aphidicolin treatment (Figure 5F). Interestingly, fluoro-Jade C-positive (FJC+) areas were significantly enlarged in broad brain regions, including the primary motor cortex, frontal and somatosensory cortex, visual cortex, and subiculum, but not in the hippocampus, of 5xFAD/PrlR heterozygous mice compared to 5xFAD mice (Figure 5G). Since estrogen has been reported to induce Prl gene expression,35 we evaluated the in vivo effects of tamoxifen treatment on the neuronal status of 5xFAD mice. One week of oral administration of tamoxifen significantly upregulated Prl and Brca1 expression in live S-phase neurons (Figure 5H). As expected, the cortical area stained with FJC was significantly reduced after tamoxifen treatment in the 5xFAD mice (Figure 5I). These data suggest that estrogen-Prl signaling plays a crucial role in neuroprotection against L1 retrotransposition-mediated DDRs that lead to NCDs, possibly via Prl-mediated Brca1 induction in S-phase neurons.Figure 5Functional significance of estrogen/Prl signals for induction of Brca1 expression in S-phase neurons(A) Interpretive phenomenological analysis (secretion from pituitary) between GFP+ or GFP− neurons derived from L1-GFP/5xFAD mice.(B) Heatmap of hormone-related differentially expressed genes (left). Relative expression of Esr2 and Prl (right).(C) Induction of Brca1 expression in Prl-treated primary neurons isolated from 3-month-old wild-type mice.(D) Relative expression of Brca1 and FancD2 in neurons derived from wild-type or PrlRhet mice. Data are shown in individual cell cycle.(E) Expression of Prl in SH-SY5Y cells with or without aphidicolin treatment (n = 3).(F) SH-SH5Y cells treated with PrlR siRNA followed by aphidicolin treatment were stained with anti-γH2AX Ab (green) and DAPI (Blue). Scale bars: 50 μm.(G) Fluorescence image of wild-type, 5xFAD, or 5xFAD/PrlRhet brain after FJC staining (left) and summary data (right). Scale bars: 1 mm.(H) 5xFAD mice were injected intraperitoneally with tamoxifen for 1 week. After 1 month from the last treatment, Prl+ or Brca1+ neurons were quantified by flow cytometry.(I) FJC staining was performed for 5xFAD mice 1 month after the same treatment as described in (H).Representative data (n = 3) from 2 to 3 independent experiments with similar results are shown as mean ± SD., except for (A), (B), (F), and (G). Statistical analysis was performed by unpaired Student’s t test for (E), (G), (H), and (I), one-way ANOVA with Tukey’s multiple comparisons test for (C), and two-way ANOVA with Tukey’s multiple comparisons test for (D). ns, not significant; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001. Functional significance of estrogen/Prl signals for induction of Brca1 expression in S-phase neurons (A) Interpretive phenomenological analysis (secretion from pituitary) between GFP+ or GFP− neurons derived from L1-GFP/5xFAD mice. (B) Heatmap of hormone-related differentially expressed genes (left). Relative expression of Esr2 and Prl (right). (C) Induction of Brca1 expression in Prl-treated primary neurons isolated from 3-month-old wild-type mice. (D) Relative expression of Brca1 and FancD2 in neurons derived from wild-type or PrlRhet mice. Data are shown in individual cell cycle. (E) Expression of Prl in SH-SY5Y cells with or without aphidicolin treatment (n = 3). (F) SH-SH5Y cells treated with PrlR siRNA followed by aphidicolin treatment were stained with anti-γH2AX Ab (green) and DAPI (Blue). Scale bars: 50 μm. (G) Fluorescence image of wild-type, 5xFAD, or 5xFAD/PrlRhet brain after FJC staining (left) and summary data (right). Scale bars: 1 mm. (H) 5xFAD mice were injected intraperitoneally with tamoxifen for 1 week. After 1 month from the last treatment, Prl+ or Brca1+ neurons were quantified by flow cytometry. (I) FJC staining was performed for 5xFAD mice 1 month after the same treatment as described in (H). Representative data (n = 3) from 2 to 3 independent experiments with similar results are shown as mean ± SD., except for (A), (B), (F), and (G). Statistical analysis was performed by unpaired Student’s t test for (E), (G), (H), and (I), one-way ANOVA with Tukey’s multiple comparisons test for (C), and two-way ANOVA with Tukey’s multiple comparisons test for (D). ns, not significant; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001. Selective downregulation of Sirt6 is a possible trigger for enhanced L1 reactivation in 5xFAD mice Thus far, our data indicated that L1 retrotransposition was significantly increased in aphidicolin-treated SH-SY5Y cells; however, the extent of this increase seemed insufficient to induce explicit cell death in normal cultured cells. Therefore, the S-phase neurons of 5xFAD mice must have additional defects that allow for excessive L1 reactivation and stochastic retrotransposition. Interestingly, IPA revealed that sirtuin family genes, pleiotropic factors with either mono-ADP-ribosyltransferase or deacylase activity, showed dynamic changes in GFP+ S-phase neurons in 5xFAD mice (Figure 6A). Among the seven sirtuin family genes, Sirt6 was selectively downregulated in the GFP+ S-phase neurons of 5xFAD mice (Figure 6B), which was confirmed by RT-qPCR (Figure 6C). Sirt6 is critically involved in the regulation of L1 reactivation and DDR accumulation, and inhibition of L1 reactivation by reverse transcriptase inhibitors prevents DDR accumulation.36,37,38 To investigate the functional relevance of impaired Sirt6 expression on L1 retrotransposition, L1-GFP/SH-SY5Y reporter cells were transfected with Sirt6-specific siRNA before aphidicolin treatment. As shown in Figure 6D, suppression of Sirt6 expression after siRNA treatment significantly augmented GFP expression in L1-GFP/SH-SY5Y cells, suggesting that reduced expression of Sirt6 may be attributed to enhanced L1 retrotransposition in 5xFAD neurons.Figure 6Functional significance of Sirt6 expression for regulation of L1 reactivation in S-phase neurons(A) Interpretive phenomenological analysis (sirtuin signaling pathway) between GFP+ or GFP− neurons derived from L1-GFP/5xFAD mice.(B) Heatmap of differential expression of sirtuin family genes.(C) Relative expression of Sirt6 quantified by RT-qPCR in GFP+ or GFP− S-phase neurons from L1-GFP/5xFAD or L1-GFP mice.(D) Numerical analysis for L1 retrotransposition in SH-SY5Y cells with or without Sirt6 siRNA treatment. Scale bars: 50 μm.(E) 5xFAD mice received efavirenz in drinking water (1.6 μg/mL) ad libitum for 1 month. Numbers of NeuN cells and FJC+ area were quantified in the mouse cortex.Representative data from 2 to 3 independent experiments with similar results are shown as mean ± SD., except for (A), (B), and (D). Statistical analysis was performed by unpaired Student’s t test for (D) and (E), and one-way ANOVA with Tukey’s multiple comparisons test for (C). ns, not significant; ∗p < 0.05, ∗∗p < 0.01. Functional significance of Sirt6 expression for regulation of L1 reactivation in S-phase neurons (A) Interpretive phenomenological analysis (sirtuin signaling pathway) between GFP+ or GFP− neurons derived from L1-GFP/5xFAD mice. (B) Heatmap of differential expression of sirtuin family genes. (C) Relative expression of Sirt6 quantified by RT-qPCR in GFP+ or GFP− S-phase neurons from L1-GFP/5xFAD or L1-GFP mice. (D) Numerical analysis for L1 retrotransposition in SH-SY5Y cells with or without Sirt6 siRNA treatment. Scale bars: 50 μm. (E) 5xFAD mice received efavirenz in drinking water (1.6 μg/mL) ad libitum for 1 month. Numbers of NeuN cells and FJC+ area were quantified in the mouse cortex. Representative data from 2 to 3 independent experiments with similar results are shown as mean ± SD., except for (A), (B), and (D). Statistical analysis was performed by unpaired Student’s t test for (D) and (E), and one-way ANOVA with Tukey’s multiple comparisons test for (C). ns, not significant; ∗p < 0.05, ∗∗p < 0.01. To reveal the pathogenic significance of enhanced L1 retrotransposition, we administered efavirenz, a non-nucleoside reverse transcriptase inhibitor class of antiretroviral drugs that also inhibits L1 retrotransposition, via drinking water to 5xFAD mice for four weeks, and the neuronal status and neurodegenerative index were evaluated. The oral administration of efavirenz significantly preserved the number of NeuN+ cells and reduced the expansion of the FJC+ area in the cortex of 5xFAD mice (Figure 6E). These results suggest that excessive L1 reactivation, concomitant with reduced Sirt6 expression, may be the primary cause of NCD, leading to neurodegeneration in 5xFAD mice. Additional research is needed to determine the precise contribution of Sirt6 to L1 reactivation in those neurons. Pathogenic relevance of L1 retrotransposition-mediated synthetic lethality with conventional AD pathology Finally, we confirmed pathogenic causality between L1 retrotransposition-mediated synthetic lethality and conventional AD pathology. Since 5xFAD mice exhibit amyloid-associated neurodegeneration that resembles AD pathology, we examined whether ectopic CCR and subsequent L1 retrotransposition observed in mature S-phase neurons have significant relevance to amyloid-associated AD pathology. Interestingly, aberrant CCR and L1 retrotransposition were preferentially observed in the cortical neurons of 5xFAD mice that were co-stained with Abs reactive to both the Ab peptide and amyloid precursor proteins (APPs) (Figure S11), suggesting that L1 retrotransposition-mediated synthetic lethality correlates well with conventional amyloid-associated pathology. The preceding loss of cholinergic transmission has been highlighted as a key pathogenic feature of AD. Interestingly, we detected a higher frequency of L1 retrotransposition in neurons expressing nAChR α4, α5, or α7, regardless of their background (L1-GFP vs. L1-GFP/5xFAD) (Figure 7A). RT-qPCR analysis confirmed that skewed L1 GFP+ expression was predominantly observed in nAChRα4+ neurons compared with AChRα7+ neurons (Figure 7B). Similar results were obtained by flow cytometry analysis using anti-nAChRα4 or anti-nAChRα7 Abs, in which the skewed upregulation of nAChRα4 was much more apparent in the GFP+ S-phase neurons of L1-GFP/5xFAD mice (Figure 7C). These results suggest that neurons expressing nAChRα4 are selectively vulnerable to L1 retrotransposition. To generalize the selective vulnerability of nAChR-expressing neurons to CCR and L1 retrotransposition, we compared the relative distribution of nAChRα4- and nAChRα7-expressing neurons in each phase of the cell cycle and estimated the frequency of L1-GFP+ cells. Although nAChRα7+ neurons from either L1-GFP/5xFAD or L1-GFP were distributed from the G0/G1 to the S phase, most nAChRa4+ neurons from either L1-GFP/5xFAD or L1-GFP were distributed in the S phase, whereas the rest were in the G2/M phase. Interestingly, GFP+ nAChRα4 + neurons of 5xFAD mice were exclusively accumulated in the S phase compared with GFP+ nAChRα4 + neurons of L1-GFP mice (Figure 7D). Meanwhile, no distribution bias was observed in nAChRα4+ neurons in either the G0/G1 or G2/M phases, or in nAChRα7+ neurons in all phases of the cell cycle (Figure 7D). Using conventional cell cycle analysis, we confirmed that a significant proportion of nAChRα4+ neurons were distributed in S phase (Figure 7E). Accordingly, immunohistochemical analysis demonstrated that most geminin+ cortical neurons coexpressed nAChRα4 in 5xFAD/mAG_h-Gem mice (Figure 7F). Together, these data suggest that nAChRα4-expressing neurons are prone to aberrant CCR and subsequent L1 retrotransposition in 5xFAD mice, which may explain why acetylcholine signaling is preferentially compromised in early AD pathology.Figure 7Preferential L1 retrotransposition in neurons expressing specific AChRa subunit in 5xFAD mice(A) RNA-seq data for differential expression of AChRα4, α5, and α7 of L1-GFP neuron (blue) or L1-GFP/5xFAD neuron (red).(B) Relative expression of AChRα4 and AChRα7 was quantified by RT-qPCR.(C) After Hoechst 33342 staining, S-phase neurons from 3-month-old L1-GFP/5xFAD mice were quantified for their expression of AChRα4 or AChRα7 by flow cytometry.(D) Neurons isolated from 3-month-old L1-GFP/5xFAD or L1-GFP mice were stained with Hoechst 33342 and anti-AChRα4 or AChRα7 Ab, then analyzed by flow cytometry.(E) Cell cycle distribution of neurons from 5xFAD mice expressing AChRα4 or AChRα7.(F) Immunofluorescence images of cortical neurons from 5xFAD/mAG_h-Gem mice after staining with anti-AChRα4 Ab. Scale bars: 50 μm.Representative data from 2 to 3 independent experiments with similar results are shown as mean ± SD, except for (A), (E), and (F). Statistical analysis was performed by unpaired Student’s t test for (C) and (D), and one-way ANOVA with Tukey’s multiple comparisons test for (B). ns, not significant; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗, p < 0.0001. Preferential L1 retrotransposition in neurons expressing specific AChRa subunit in 5xFAD mice (A) RNA-seq data for differential expression of AChRα4, α5, and α7 of L1-GFP neuron (blue) or L1-GFP/5xFAD neuron (red). (B) Relative expression of AChRα4 and AChRα7 was quantified by RT-qPCR. (C) After Hoechst 33342 staining, S-phase neurons from 3-month-old L1-GFP/5xFAD mice were quantified for their expression of AChRα4 or AChRα7 by flow cytometry. (D) Neurons isolated from 3-month-old L1-GFP/5xFAD or L1-GFP mice were stained with Hoechst 33342 and anti-AChRα4 or AChRα7 Ab, then analyzed by flow cytometry. (E) Cell cycle distribution of neurons from 5xFAD mice expressing AChRα4 or AChRα7. (F) Immunofluorescence images of cortical neurons from 5xFAD/mAG_h-Gem mice after staining with anti-AChRα4 Ab. Scale bars: 50 μm. Representative data from 2 to 3 independent experiments with similar results are shown as mean ± SD, except for (A), (E), and (F). Statistical analysis was performed by unpaired Student’s t test for (C) and (D), and one-way ANOVA with Tukey’s multiple comparisons test for (B). ns, not significant; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗, p < 0.0001. Discussion L1 reactivation and subsequent retrotransposition are the major causes of DDR leading to NCDs. We have previously reported ectopic L1 reactivation in 5xFAD and mSOD1 mice and in neurotoxic Eomes+ Th cells that recognize the ORF1 protein derived from dying neurons for the propagation of neurodegeneration.5 In this study, we demonstrated that a previously unappreciated synthetic lethality consisting of enhanced L1 retrotransposition and DNA repair deficiency under aberrant CCR is critically involved in authentic NCD and subsequent neurodegeneration in 5xFAD mice. Once terminally differentiated, the neurons remain quiescent. However, neuronal upregulation of cell cycle-related genes and cell cycle dysregulation have long been suggested to be involved in the pathogenesis of diverse NDs.18,39 For instance, patients with AD contain more than 20% of hyperploid neurons with hyperploid DNA content, which further increases with disease progression.19 Importantly, these hyperploid neurons seem to be selectively eliminated during the progression to severe AD.20 We observed a similar tendency for neuronal hyperploidy in the 5xFAD mice. Intriguingly, the affected neurons of 6-month-old 5xFAD mice showed significantly reduced Neu O staining, and most of those with reduced Neu O staining lost reactivity to the anti-NeuN Ab. The NeuN antigen (hexaribonucleotide binding protein-3 or Rbfox3) is exclusively expressed in postmitotic neurons but not in proliferating progenitor cells. Mature neurons that regain CCR may lose NeuN expression,40 which may disturb the exact identification of live neurons upon disease progression. Aberrant reactivation of transposable elements (TEs) is another common event widely observed in diverse NDs.7,41 This neuronal reawakening of TEs is a potent pathogenic driver of neurodegeneration that promotes genomic instability, ectopic gene expression, and the induction of neuroinflammatory responses. However, the comprehensive insight into how such a disturbance in cellular homeostasis via TE reactivation leads to NCDs is not fully understood. The fact that the emergence of L1 retrotransposition shows a strong bias toward the S phase in cultured cells22 and that L1 preferentially targets replication forks14 prompted us to hypothesize a possible link between CCR and L1 reactivation during neurodegeneration. RNA-seq and IPA analyses revealed that increased nucleotide metabolism and neddylation were the most upregulated pathways in GFP+ S-phase neurons of 5xFAD mice. Neddylation is commonly observed in the neuronal and glial ubiquitinated inclusions of NDs in AD, ALS, Parkinson’s disease, dementia with Lewy bodies, and multiple system atrophy.42 Although Nedd8 is localized in the nucleus under physiological conditions, the overexpression of APP activates APP-BP1 and promotes the cytoplasmic accumulation of Nedd8, which triggers ectopic CCR, aging, and cell death.43 Importantly, dysregulation of neddylation is a typical characteristic of AD.42,44 As dysfunction of the neddylation pathway is considered a key event in the pathogenesis of AD, the exclusive upregulation of the neddylation pathway in GFP+ S-phase neurons of 5xFAD mice implies that this particular subset of neurons may preferentially be fatal at the very early stage of neurodegeneration. GFP neurons of 5xFAD mice showed upregulation of the canonical cellular effects of the sildenafil pathway. Many PDE5 inhibitors, including sildenafil, exhibit neuroprotective effects, improve learning deficiencies in rats, and significantly attenuate AD risk.45,46 These data suggest that GFP+ S-phase neurons may be a bona fide neuronal subset that is susceptible to neurodegeneration in 5xFAD mice. Accordingly, exclusive L1 reactivation was observed in GFP+ 5xFAD neurons, but not in GFP+ littermate neurons (Figure S7), suggesting that the background condition for GFP expression may differ between neurons from GFP+ 5xFAD mice and littermates, where L1 reactivation leads to following L1 retrotransposition only in GFP+ 5xFAD neurons. RNA-seq and IPA revealed that the expression of 21 DNA repair-related genes, including Brca1, Bard1, and FancD2, was uniformly decreased in the GFP+ S-phase neurons of 5xFAD mice, suggesting that these neurons are more vulnerable to replication stress after L1 retrotransposition. Brca1 is a pleiotropic protein with tumor suppressor and caretaker functions in familial breast and ovarian cancers via the regulation of diverse cellular responses, including DDRs, cell cycle checkpoints, apoptosis, transcription regulation, replication fork protection, and heterochromatin formation. Brca1 functions in association with its obligatory partner Bard129 through its N-terminal RING domain. In addition, Palb2 associates with Brca1 via a coiled-coil domain, and phospho-dependent interactions between ABRAXAS, BRIP1, and CtIP are mediated by the C-terminal BRCT domain of Brca1.47 The Brca1/Bard1 complex plays a critical role in the resolution of replication stress, restarting stalled replication forks, protection from nuclease attacks, and subsequent attrition.29 Earlier histological analyses of postmortem brain samples from patients with AD and a mouse model suggested reduced expression of Brca1,45 although subsequent reports provided conflicting results.48,49 Brca1 expression was significantly upregulated in healthy S-phase neurons and strongly downregulated in dying neurons. Thus, drastic fluctuations in neuronal Brca1 expression may lead to conflicting results in mixed live and dead neurons. Importantly, we do not exclude the possibility that other DNA repair factors downregulated in the GFP+ S-phase neurons of 5xFAD mice may be involved in NCD, which is to be clarified in a future study. There are different types of replication stress hindering the progression of DNA synthesis through lesions or alternative structures, such as single-strand breaks, DNA secondary structures (stem-loop, cruciform structures, or G-quadruplex DNA), inter-strand crosslinks, and stable R-loops.50,51,52 R-loop-related replication stress is usually induced by head-on collisions between DNA replication and transcription and is suggested to provide a pathological mechanism that underpins neurodegeneration.53 Brca1 interacts with senataxin (SETX), a putative RNA-DNA helicase and the causative gene of oculomotor apraxia type 2 and ALS type 4, to promote transcription termination and R-loop resolution.54 In addition, Bcra1 is involved in R-loop removal during DSB repair.55 The transcription-associated R-loop (cis-R-loop) can be detected by DRIP using S9.6 Ab. However, S9.6 Ab also recognizes a DNA:RNA triplet structure consisting of a replication fork associated with the L1 transcript (trans-R-loop).33 We detected L1 transcripts in S-phase cells by both DRIP and ChIP analyses using the S9.6 Ab and an anti-DNA polδ Ab, suggesting that L1 preferentially targets the replication fork. Several key attenuators of the L1 reactivation have been identified.56 Highlighted among them via IPA was Sirt6, a nuclear protein involved in histone deacetylation, defatty-acylation, and mono-ADP-ribosylation that regulates senescence, neurodegeneration, immunity, and cancer.57,58,59 Sirt6 knockout (KO) mice show an age-like degenerative phenotype due to the accumulation of DDR, whereas Sirt6-overexpressing mice show an extended lifespan.38 Intriguingly, deletion of Sirt6 impairs mono-ADP-ribosylation of the corepressor KAP1 and allows the reactivation of L1 and other TEs.36 Furthermore, reverse transcriptase inhibitors prevent the accumulation of DNA damage and extend the lifespan of Sirt6 KO mice by suppressing L1 reactivation.37 Although we have not shown direct evidence that downregulation of Sirt6 expression causes L1 reactivation, the administration of efavirenz was effective in maintaining neuronal survival and reducing the FJC+ area, suggesting that impaired Sirt6 expression in S-phase neurons may allow L1 reactivation in L1-GFP/5xFAD mice. Synthetic lethality is defined as the synergistic outcome of polygenic and complex cellular events, in which the comorbidity of two independent components constitutes a higher-order synergy, compromises cellular robustness, and causes cell death.37,60,61 Mutation of Brca1 and concomitant inhibition of poly ADP-ribose polymerase (PARP) is one such synthetic lethality.62,63,64 The concept of Brcaness,65 a DNA repair defect with no Brca mutation, has now been extended to defects in replication fork protection.66 In this study, we identified synthetic lethality between L1 retrotransposition and DNA repair deficiency, which was attributed to the impaired expression of Sirt6 and Brca1, corresponding to downstream L1 reactivation and replication stress, respectively, in S-phase neurons. Similar synthetic lethality between sirtuins and Brca1/2 has been reported in BRCA1-defective triple-negative breast cancer cell lines.67 BRCA1 mutations are closely associated with hereditary breast and ovarian cancers; however, they cf. an increased risk of other hormone-dependent cancers of the uterus, cervix, and prostate.68 Prl plays a significant role in neuroprotection against excitotoxicity69 and induces BRCA1 expression in human breast cancer cell lines.70 We confirmed that Prl and Brca1 expression was significantly increased in S-phase cells, which prevented the accumulation of γH2AX signals71 and replication fork degradation in the absence of Brca1-deficient tumors.72 5xFAD/PrlRhet mice showed a significant increase in DNA damage and an expanded neurodegenerative area in 5xFAD mice. Furthermore, tamoxifen treatment increased the neuronal expression of Prl and Brca1 and prevented the expansion of the neurodegenerative area in 5xFAD mice. Importantly, menopause has been suggested to be a potent disease-promoting factor for AD in women, and estrogen replacement therapy is a possible option for the treatment of AD after menopause.73 Although we cannot exclude the possibility that other pathways may lead to Brca1 induction, our data indicate that estrogen/Prl signaling plays an important role in Brca1 upregulation in S-phase neurons. Further investigation is required to clarify the therapeutic potential of estrogen/Prl signaling in neurodegeneration. Finally, we evaluated the functional relevance of synthetic lethality in conventional AD pathology and confirmed that aberrant CCR and/or L1-GFP expression was observed in neurons stained intracellularly with Abs that recognize all forms of amyloid proteins, including Ab and APP. APP overexpression leads to the activation of APP-BP1, cytoplasmic translocation of Nedd8, induction of CCR, and neuronal apoptosis,43 which may be a relay point between L1 retrotransposition-mediated synthetic lethality and conventional AD pathology. Cholinergic defects, which underlie the cholinergic hypothesis, have been implicated in the pathogenesis of AD.74,75,76 Three different nAChR subunits (α4, α5, and α7) are predominantly expressed in the cortex in a layer- and cell-type-specific manner77 and play an important role in AD pathogenesis.78 Interestingly, the selective loss of neuronal nAchRα4 and/or nAchRα7 expression has been described as a key feature of early AD pathology.79,80,81,82,83,84 We observed preferential L1 retrotransposition in nAChRα4+ S-phase neurons in 5xFAD mice, suggesting that nAchRα4+ neurons are more vulnerable to CCR and L1 retrotransposition. Although the fate of these neurons and disease progression should be investigated in future studies, these data demonstrate a link between selective nAchR expression and CCR-associated L1 retrotransposition in the cortical neurons of 5xFAD mice. In conclusion, we demonstrated that the coincidence of L1 retrotransposition and DNA repair deficiency causes synthetic lethality in S-phase neurons of 5xFAD mice. Most of the events and phenomena described in this study were individually postulated to be associated with diverse NDs. Therefore, the significance of our study was to consolidate these events and phenomena to establish a consistent and comprehensive pathogenic mechanism of neurodegeneration. Together, autonomous NCDs are not monocausal, and multifaceted synthetic lethality plays a pivotal role in the pathogenesis of neurodegeneration. Limitations of the study Here, we report synthetic lethality elicited by the coincidence of enhanced L1 retrotransposition and DNA repair deficiency for selective NCD of S-phase neurons, providing a novel molecular mechanism for the initial stage of cell-autonomous NCD. However, we previously identified the involvement of immune-mediated neurodegeneration in promoting the propagation of neurodegenerative lesions at the progressive stage of NDs. Therefore, a comprehensive analysis of the neuron-autonomous and immune-mediated pathogenesis of clinical manifestations of NDs should be performed to evaluate the significance of this study. Accordingly, the efficacy of combined therapeutic interventions targeting the biphasic processes of neuron-autonomous and immune-mediated pathogenesis and their potential synergistic practicality should be examined to establish a pragmatic therapeutic strategy for NDs. Resource availability Lead contact Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Shinji Oki (soki@ncnp.go.jp). Materials availability This study did not generate new, unique reagents. Data and code availability •Sequencing data generated in this study have been deposited in the Gene Expression Omnibus (GEO) under accession number GSE300383 and are publicly available as of the date of publication.•This paper does not report original code.•All other data supporting the findings of this study are available from the lead contact upon reasonable request. •Sequencing data generated in this study have been deposited in the Gene Expression Omnibus (GEO) under accession number GSE300383 and are publicly available as of the date of publication. Sequencing data generated in this study have been deposited in the Gene Expression Omnibus (GEO) under accession number GSE300383 and are publicly available as of the date of publication. •This paper does not report original code. This paper does not report original code. •All other data supporting the findings of this study are available from the lead contact upon reasonable request. All other data supporting the findings of this study are available from the lead contact upon reasonable request. Acknowledgments We are grateful to Dr. T. Okamura (National Center for Global Health and Medicine [NCGM], Tokyo, Japan) and Dr. A. Miyawaki (National Research and Development Agency [RIKEN], Saitama, Japan) for providing L1-GFP and FUCCI mice, respectively. We thank Drs. Y. Izhizaka (NCGM), N. Gilbert (Inserm), and N. Luning Prak for providing the L1-GFP expression vector. We thank A. Takeo, C. Koto, and K. Inoue for their excellent technical assistance and assistance with animal care. We thank T. Araki and S. Wakatsuki for critically reviewing this manuscript. This work was supported by JSPS KAKENHI grant nos. 23KJ2199 and 24k18634 (F.T.); 20K07895, 25K10652, and 23K24802 (S.O.); the National Institute of Neuroscience Intramural Research Grant for Neurological and Psychiatric Disorders 30-5/1-4 (S.O.); and the Mitsubishi Foundation Research Grants in the Natural Sciences (S.O.). Author contributions Conceptualization, F.T. and S.O.; methodology, F.T. and S.O.; investigation, F.T. and S.O.; writing – original draft, F.T.; writing – review and editing, T.Y. and S.O.; funding acquisition, F.T. and S.O.; resources, T.Y. and S.O.; supervision, T.Y. and S.O. Declaration of interests The authors declare no competing interests. Declaration of generative AI and AI-assisted technologies in the writing process The authors declare that no generative AI and AI-assisted technologies were applied in the writing process. STAR★Methods Key resources table REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesNeuroFluor NeuOSTEMCELLCat#01801 Cas:1616355-50-0Hoechst 33342DOJINDOH342; Cas:23491-52-3Live/Dead Near-IR Dead Cell Stain KitInvitrogenCat#L34976anti-Brca1 MoAbSanta CruzCat#sc-6954; AB_626761anti-Prl MoAbBIOSSCat#bs-0508R-A647;AB_11088132Alexa Fluor® 647 anti-H2A.X Phospho (Ser139)BioLegendCat#613408; AB_2295046Alexa Fluor® 647 Annexin VBioLegendCat#640911; AB_2561293anti-nAChRα4 MoAbSanta CruzCat#sc-74519 AB_2080535anti-nAChRα7 MoAbSanta CruzCat#sc-58607, AB_784835Cleaved Caspase-3 (Asp175) (D3E9) Rabbit mAb (Alexa Fluor® 647 Conjugate)CSTCat#9602; AB_2687881anti-DNA pol δ MoAb (A-9)Santa CruzCat#sc-17776; AB_675484Alexa Fluor® 647 anti-β-Amyloid, 17-24BioLegendCat# 800717; AB_2721295AF647-conjugated rabbit anti-GemininSanta CruzCat#sc-74456; AB_2943326AF647-conjugated rabbit anti-NeuNAbcamab177487; AB_2532109Chemicals, peptides, and recombinant proteinsTamoxifenNACALAI TESQUECat#19885-34CAS:10540-29-1EfavirenzTCICat#E0997 Cas: 154598-52-4AphidicolinSIGMA ALDRICHCat#A4487 Cas:38966-21-1Bleomycin SulfateTCICat#B3972Super BLOCK Blocking BufferThermo Fisher ScientificCat#37515HaeIIITAKARA BioCat#1051AHindIIITAKARA BioCat#1060ATrue Black Lipofuscin Auto fluorescence plus reagent (BTI, USA)BIOTIUMCat#23014-TDAPI Fluoromount-GSouthen BiotechCat#0100-20DPX Mount for histologySIGMA ALDRICHCat#06522Dulbecco’s Modified Eagle MediumGibcoCat#12430-054Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 supplementedGibcoCat#11320-033Critical commercial assaysMACS® adult brain dissociation kitMiltenyi BiotechCat#130-107-677Neuron isolation kit mouseMiltenyi BiotechCat#130-115-389Magnetofection NeuroMag kitOZ BIOSCIENCESNM50500RNeasy Mini KitQiagenCat#74106First-strand cDNA KitTakara BioCat#6215ACUT&Tag-IT® R-loop Assay KitACTIVE MOTIFCat#53167Immunoprecipitation kit-Dynabeads proteinGInvitrogenCat#10007DFluoro-Jade C Ready-to-Dilute Staining KitbiosensisCa#TR-160-FJCDeposited dataNCBI Gene Expression OmnibusNCBI Gene Expression Omnibushttps://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE300383Experimental models: Cell linesSH-SY5Y cellsATCC®CRL-2266Experimental models: Organisms/strains5xFAD miceThe Jackson Laboratoryhttps://doi.org/10.1016/j.isci.2022.104278FUCCI miceRIKEN BRCRBRC02817L1-GFP miceProvided by Dr. Okamura (National Center for Global Health and Medicine (NCGM), Tokyo, Japan).https://doi.org/10.1111/j.1349-7006.2011.02060.xPrlR het miceThe Jackson LaboratoryB6.129P2-Prlrtm1Cnp/J, #003142OligonucleotideshL1 ORF1 primerhttps://doi.org/10.1038/s41586-018-0784-9N/AhL1 ORF2 primerhttps://doi.org/10.1038/s41586-018-0784-9N/APrl Mouse qPCR Primer Pair (NM_011164)OrigeneMP211723Mm_Palb2_1_SG QuantiTect Primer AssayQIAGENQT00291928Mm_Bard1_1_SG QuantiTect Primer AssayQIAGENQT00112280Mm_Brca1_1_SG QuantiTect Primer AssayQIAGENQT00145810Hs_BRCA1_1_SG QuantiTect Primer AssayQIAGENQT00039305Mm_Rad51_1_SG QuantiTect Primer AssayQIAGENQT00102921Mm_Fancd2_1_SG QuantiTect Primer AssayQIAGENQT00267155Mm_Sirt6_1_SG QuantiTect Primer AssayQIAGENQT00112700MM_CHRNA4_2070690 QuantiNova LNA PCR AssayQIAGENSBM0826587MM_CHRNA7_2205642 QuantiNova LNA PCR AssayQIAGENSBM0961501FlexiTube GeneSolution for SIRT6QIAGENGS51548FlexiTube GeneSolution for BRCA1QIAGENGS672FlexiTube GeneSolution for PRLRQIAGENGS5618Software and algorithmsFACS Diva software and FlowJo softwareBD BioScienceshttps://www.bdbiosciences.com/ja-jp/products/software/instrument-software/bd-facsdiva-softwareIPA pathway analysisQIAGENhttps://digitalinsights.qiagen.com/products-overview/discovery-insights-portfolio/analysis-and-visualization/qiagen-ipa/RNA-seq using SMART-Seq v4 Ultra Low Input KitTAKARA BioN/AOtherGentleMACS Octo DissociatorMiltenyi Biotechhttps://www.miltenyibiotec.com/US-en/products/gentlemacs-octo-dissociator-with-heaters.htmlFACS Canto IIBD BioscienceN/AFACS Aria II cell sorterBD BioscienceN/AReal-time PCR instrumentRoche DiagnosticsLightCycler 96 instrumentCryostatThermo Fisher ScientificCryostar NX70Fluorescence microscopeKeyenceBZ-X710 automated fluorescence microscope REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesNeuroFluor NeuOSTEMCELLCat#01801 Cas:1616355-50-0Hoechst 33342DOJINDOH342; Cas:23491-52-3Live/Dead Near-IR Dead Cell Stain KitInvitrogenCat#L34976anti-Brca1 MoAbSanta CruzCat#sc-6954; AB_626761anti-Prl MoAbBIOSSCat#bs-0508R-A647;AB_11088132Alexa Fluor® 647 anti-H2A.X Phospho (Ser139)BioLegendCat#613408; AB_2295046Alexa Fluor® 647 Annexin VBioLegendCat#640911; AB_2561293anti-nAChRα4 MoAbSanta CruzCat#sc-74519 AB_2080535anti-nAChRα7 MoAbSanta CruzCat#sc-58607, AB_784835Cleaved Caspase-3 (Asp175) (D3E9) Rabbit mAb (Alexa Fluor® 647 Conjugate)CSTCat#9602; AB_2687881anti-DNA pol δ MoAb (A-9)Santa CruzCat#sc-17776; AB_675484Alexa Fluor® 647 anti-β-Amyloid, 17-24BioLegendCat# 800717; AB_2721295AF647-conjugated rabbit anti-GemininSanta CruzCat#sc-74456; AB_2943326AF647-conjugated rabbit anti-NeuNAbcamab177487; AB_2532109Chemicals, peptides, and recombinant proteinsTamoxifenNACALAI TESQUECat#19885-34CAS:10540-29-1EfavirenzTCICat#E0997 Cas: 154598-52-4AphidicolinSIGMA ALDRICHCat#A4487 Cas:38966-21-1Bleomycin SulfateTCICat#B3972Super BLOCK Blocking BufferThermo Fisher ScientificCat#37515HaeIIITAKARA BioCat#1051AHindIIITAKARA BioCat#1060ATrue Black Lipofuscin Auto fluorescence plus reagent (BTI, USA)BIOTIUMCat#23014-TDAPI Fluoromount-GSouthen BiotechCat#0100-20DPX Mount for histologySIGMA ALDRICHCat#06522Dulbecco’s Modified Eagle MediumGibcoCat#12430-054Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 supplementedGibcoCat#11320-033Critical commercial assaysMACS® adult brain dissociation kitMiltenyi BiotechCat#130-107-677Neuron isolation kit mouseMiltenyi BiotechCat#130-115-389Magnetofection NeuroMag kitOZ BIOSCIENCESNM50500RNeasy Mini KitQiagenCat#74106First-strand cDNA KitTakara BioCat#6215ACUT&Tag-IT® R-loop Assay KitACTIVE MOTIFCat#53167Immunoprecipitation kit-Dynabeads proteinGInvitrogenCat#10007DFluoro-Jade C Ready-to-Dilute Staining KitbiosensisCa#TR-160-FJCDeposited dataNCBI Gene Expression OmnibusNCBI Gene Expression Omnibushttps://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE300383Experimental models: Cell linesSH-SY5Y cellsATCC®CRL-2266Experimental models: Organisms/strains5xFAD miceThe Jackson Laboratoryhttps://doi.org/10.1016/j.isci.2022.104278FUCCI miceRIKEN BRCRBRC02817L1-GFP miceProvided by Dr. Okamura (National Center for Global Health and Medicine (NCGM), Tokyo, Japan).https://doi.org/10.1111/j.1349-7006.2011.02060.xPrlR het miceThe Jackson LaboratoryB6.129P2-Prlrtm1Cnp/J, #003142OligonucleotideshL1 ORF1 primerhttps://doi.org/10.1038/s41586-018-0784-9N/AhL1 ORF2 primerhttps://doi.org/10.1038/s41586-018-0784-9N/APrl Mouse qPCR Primer Pair (NM_011164)OrigeneMP211723Mm_Palb2_1_SG QuantiTect Primer AssayQIAGENQT00291928Mm_Bard1_1_SG QuantiTect Primer AssayQIAGENQT00112280Mm_Brca1_1_SG QuantiTect Primer AssayQIAGENQT00145810Hs_BRCA1_1_SG QuantiTect Primer AssayQIAGENQT00039305Mm_Rad51_1_SG QuantiTect Primer AssayQIAGENQT00102921Mm_Fancd2_1_SG QuantiTect Primer AssayQIAGENQT00267155Mm_Sirt6_1_SG QuantiTect Primer AssayQIAGENQT00112700MM_CHRNA4_2070690 QuantiNova LNA PCR AssayQIAGENSBM0826587MM_CHRNA7_2205642 QuantiNova LNA PCR AssayQIAGENSBM0961501FlexiTube GeneSolution for SIRT6QIAGENGS51548FlexiTube GeneSolution for BRCA1QIAGENGS672FlexiTube GeneSolution for PRLRQIAGENGS5618Software and algorithmsFACS Diva software and FlowJo softwareBD BioScienceshttps://www.bdbiosciences.com/ja-jp/products/software/instrument-software/bd-facsdiva-softwareIPA pathway analysisQIAGENhttps://digitalinsights.qiagen.com/products-overview/discovery-insights-portfolio/analysis-and-visualization/qiagen-ipa/RNA-seq using SMART-Seq v4 Ultra Low Input KitTAKARA BioN/AOtherGentleMACS Octo DissociatorMiltenyi Biotechhttps://www.miltenyibiotec.com/US-en/products/gentlemacs-octo-dissociator-with-heaters.htmlFACS Canto IIBD BioscienceN/AFACS Aria II cell sorterBD BioscienceN/AReal-time PCR instrumentRoche DiagnosticsLightCycler 96 instrumentCryostatThermo Fisher ScientificCryostar NX70Fluorescence microscopeKeyenceBZ-X710 automated fluorescence microscope Experimental model and study participant details Mice The 5xFAD and L1-GFP mice have been previously described 5,84. Prlr-het mice (B6.129P2-Prlrtm1Cnp/J, #003142) were obtained from Jackson Laboratory. FUCCI mice (RBRC02817) 25,26 were obtained from RIKEN BRC. Male mice were used throughout the experiments unless otherwise described. All mice were maintained in ventilated racks under a 12-h light/dark cycle with food and water ad libitum in temperature-controlled, specific pathogen-free facilities. Breeding, backcrossing, and maintenance of the mice were performed at the animal facility of the National Center of Neurology and Psychiatry (NCNP). All animal experiments performed in this study were approved by the Animal Care and Use Committee of the National Institute of Neuroscience, Japan (Authorization #R6-53). All efforts were made to minimize suffering and reduce the number of animals. Method details Mouse treatment with reagents Two-month-old mice were treated intraperitoneally with tamoxifen (60 mg/kg, Fujifilm Wako, Osaka, Japan) dissolved in sunflower oil or sunflower oil anole for five consecutive days. Two-month-old mice received efavirenz (1.6 μg/mL, TCI, Tokyo, Japan) or 0.07% dimethyl sulfoxide in drinking water ad labium for 6 weeks. Isolation of adult neurons Brains and spinal cords dissected from 3-month-old mice were digested using a GentleMACS Octo Dissociator (Miltenyi Biotech, Bergisch Gladbach, Germany) according to the manufacturer’s instructions. Neurons were isolated using the MACS® adult brain dissociation kit (Miltenyi Biotech) with non-neuron antibody-conjugated magnetic beads. Isolated neurons were suspended in FACS buffer for flow cytometry analysis or adjusted to maximum of 3.5×105 cells/mL in Dulbecco’s modified Eagle medium (DMEM; Thermo Fisher Scientific, Waltham, MA) supplemented with 10% fetal calf serum (FCS) for either siRNA or Prl treatment. Flow cytometry analysis Flow cytometry analysis and sorting neurons isolated from brain and spinal cords were stained with neuron-staining buffer NeuroFluor NeuO (STEMCELL Technologies, Vancouver, Canada) for 1 h and Hoechst 33342 (DOJINDO, Kumamoto, Japan) for 45 min in medium, then stained with the Live/Dead Near-IR Dead Cell Stain Kit (Invitrogen, Waltham, MA) for 15 min at 37°C in a CO2 incubator. Cells in individual cell cycle phases (G0/G1, S, and G2/M) were sorted using a FACS Aria II cell sorter (BD Biosciences, Franklin Lakes, NJ). Fluorescence dye-conjugated anti-Brca1 MoAb (Santa Cruz Biotechnologies, Dallas, TX), anti-Prl MoAb (BIOSS, Woburn, MA), anti-γH2AX MoAb, anti-Annexin V (BioLegend, San Diego, CA), nAChRα4 and nAChRα7 (Santa Cruz Biotechnologies) were used at 1:100 dilution in phosphate-buffered saline (PBS) supplemented with 0.05% FCS. For cell surface staining, cells were incubated with MoAbs for 30 min on ice. The splenocytes were stained with MoAbs and Hoechst 33342 in the same manner and used for gate setting. Flow cytometry data were processed using FACSDiva software (v8.0.1) and analyzed using FlowJo software (v10.6.2) (BD BioSciences). Cell culture and siRNA treatment For Prl treatment, isolated neurons were adjusted to a maximum of 3.5×105 cells/mL in DMEM supplemented with 10% FCS, 200 μL of suspended cells were seeded to 48-well plates coated with poly-L-lysine and laminin, and cultured at 37°C in 5% CO2. Twenty-four hours later, cells were treated with Prl for another 72 h. For siRNA treatment, 1.0×105 SH-SY5Y cells/well in DMEM/Nutrient Mixture F-12 supplemented (DMEM/F12; Thermo Fisher Scientific) with 10% FCS were transfected with Sirt6 siRNA, Brca1 siRNA, or Prlr siRNA (10 nM each) with the Magnetofection NeuroMag kit (OZ Biosciences, Marseille, France), cultured up to 24 h, and then treated with 0.4 μM aphidicolin for another 72 h. For bleomycin treatment, the cells were treated with bleomycin 8 h after aphidicolin treatment. RNA extraction and qRT-PCR Total RNA was extracted from isolated neurons using the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. RNA was reverse transcribed using a first-strand cDNA kit (Takara Bio, Kusatsu, Japan) with random hexamer primers and subjected to qRT-PCR analysis. Gene expression was quantified by qRT-PCR using a LightCycler 96 instrument (Roche Diagnostics, Basel, Switzerland) and SYBR Green Master Mix (Roche). RNA-seq and IPA pathway analysis Conventionally, RNA samples were sent to Rhelixa Inc (Tokyo Japan) for RNA-seq analysis. Raw read counts were normalized to the trimmed mean of M values, and DEG analysis was performed using edgeR (version 3.26.8). DEGs were detected with thresholds of |log2FC| > 1 and p < 0.05. Data, network, and canonical pathway analyses were conducted using Ingenuity Pathway Analysis (IPA) (QIAGEN). Immunofluorescence imaging Isolated brains were fixed with 4% paraformaldehyde (PFA) for 24 h at 4°C and dissected by using a Cryostar NX70 cryostat (Thermo Fisher Scientific). The sections were dried, rehydrated, defatted, and blocked with Super BLOCK Blocking Buffer (Thermo Fisher Scientific) for 1 h at room temperature. Specimens were incubated with fluorescent dye-conjugated primary antibodies and mounted using a fluorescence mounting medium with DAPI (Southern Biotech, Birmingham, AL). All samples were imaged using a BZ-X710 automated fluorescence microscope (Keyence, Osaka, Japan). The following antibodies were used: AF647-conjugated rabbit anti-geminin (1:50) (Santa Cruz Biotechnology) and AF647-conjugated rabbit anti-NeuN (1:200) (Abcam, Cambridge, UK). SH-SY5Y cells were seeded in 8-well LAB-TEK chambers (Nalge Nunc International, Rochester, NY) after manual collagen coating. Cells were fixed with 1:1 acetone/methanol fixative for 20 min at −20°C, rinsed three times with PBS, blocked with Super BLOCK Blocking Buffer for 60 min at room temperature, and incubated with fluorescent dye-conjugated anti-γH2AX Ab (1:100) overnight at 4°C. After three washes with PBS, the samples were treated with True Black Lipofuscin Auto Fluorescence Plus Reagent (Biotium, Fremont, CA) before mounting with DAPI Fluoromount-G (Southern Biotech). For the evaluation of neurodegenerative status, brain sections were stained with a Fluoro-Jade C stain reagent kit (Biosensis, Thebarton, Australia) and compounded with DPX mount medium (Sigma-Aldrich, St. Louis, MO). DRIP and ChIP For DRIP analysis with S9.6 Ab, DNA was extracted from 1.0×106 SH-SY5Y cells treated with aphidicolin for 72h and were analyzed using the CUT&Tag-IT® R-loop Assay Kit (Active Motif, Carlsbad, CA) with slight modification. Briefly, RNA was purified from immunoprecipitates using the RNeasy Mini Kit (Qiagen), and L1 cDNAs were quantified in the ORF1 and ORF2 regions by qPCR. For ChIP analysis, DNA extracted from aphidicolin-treated SH-SY5Y cells was digested with restriction enzymes (HaeIII or HindIII). Fragmented DNA were immunoprecipitated with anti-DNA PolδAb or control IgG (Santa Cruz Biotechnology). RNA was purified from immunoprecipitates using the RNeasy Mini Kit (Qiagen), and L1 cDNAs were quantified at the ORF1 and ORF2 regions using qPCR. Quantification and statistical analysis Statistical analysis was performed using GraphPad Prism 8 (GraphPad Software, San Diego, CA) by applying the unpaired Student’s t test and one- or two-way analysis of variance (ANOVA) with Tukey’s multiple comparisons test. ns, not significant; ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001; ∗∗∗∗, p < 0.0001 throughout the manuscript. Statistical significance was set at p < 0.05.