Next-generation humanized patient-derived xenograft mouse model for pre-clinical antibody studies in neuroblastoma.

Faithful tumor mouse models are fundamental research tools to advance the field of immuno-oncology (IO). This is particularly relevant in diseases with low incidence, as in the case of pediatric malignancies, that rely on pre-clinical therapeutic development. However, conventional syngeneic and genetically engineered mouse models fail to recapitulate the tumor heterogeneity and microenvironmental complexity of human pathology that are essential determinants of cancer-directed immunity. Here, we characterize a novel mouse model that supports human natural killer (NK) cell development and engraftment of neuroblastoma orthotopic patient-derived xenograft (O-PDX) for pre-clinical antibody and cytokine testing. Using cytotoxicity assays, single-cell RNA-sequencing, and multi-color flow cytometry, we demonstrate that NK cells that develop in the humanized mice are fully licensed to execute NK cell cytotoxicity, permit human tumor engraftment, but can be therapeutically redirected to induce antibody-dependent cell-mediated cytotoxicity (ADCC). Although these cells share phenotypic and molecular features with healthy controls, we noted that they lacked an NK cell subset, termed activated NK cells, that is characterized by differentially expressed genes that are induced by cytokine activation. Because this subset of genes is also downregulated in patients with neuroblastoma compared to healthy controls, we hypothesize that this finding could be due to tumor-mediated suppressive effects. Thus, despite its technical complexity, this humanized patient-derived xenograft mouse model could serve as a faithful system for future testing of IO applications and studies of underlying immunologic processes.

Immunodetection of human topoisomerase I-DNA covalent complexes.

A number of established and investigational anticancer drugs slow the religation step of DNA topoisomerase I (topo I). These agents induce cytotoxicity by stabilizing topo I-DNA covalent complexes, which in turn interact with advancing replication forks or transcription complexes to generate lethal lesions. Despite the importance of topo I-DNA covalent complexes, it has been difficult to detect these lesions within intact cells and tumors. Here, we report development of a monoclonal antibody that specifically recognizes covalent topo I-DNA complexes, but not free topo I or DNA, by immunoblotting, immunofluorescence or flow cytometry. Utilizing this antibody, we demonstrate readily detectable topo I-DNA covalent complexes after treatment with camptothecins, indenoisoquinolines and cisplatin but not nucleoside analogues. Topotecan-induced topo I-DNA complexes peak at 15-30 min after drug addition and then decrease, whereas indotecan-induced complexes persist for at least 4 h. Interestingly, simultaneous staining for covalent topo I-DNA complexes, phospho-H2AX and Rad51 suggests that topotecan-induced DNA double-strand breaks occur at sites distinct from stabilized topo I-DNA covalent complexes. These studies not only provide new insight into the action of topo I-directed agents, but also illustrate a strategy that can be applied to study additional topoisomerases and their inhibitors in vitro and in vivo.

Nonhomologous end joining drives poly(ADP-ribose) polymerase (PARP) inhibitor lethality in homologous recombination-deficient cells.

Poly(ADP-ribose) polymerase (PARP) inhibitors are strikingly toxic to cells with defects in homologous recombination (HR). The mechanistic basis for these findings is incompletely understood. Here, we show that PARP inhibitor treatment induces phosphorylation of DNA-dependent protein kinase substrates and stimulates error-prone nonhomologous end joining (NHEJ) selectively in HR-deficient cells. Notably, inhibiting DNA-dependent protein kinase activity reverses the genomic instability previously reported in these cells after PARP inhibition. Moreover, disabling NHEJ by using genetic or pharmacologic approaches rescues the lethality of PARP inhibition or down-regulation in cell lines lacking BRCA2, BRCA1, or ATM. Collectively, our results not only implicate PARP1 catalytic activity in the regulation of NHEJ in HR-deficient cells, but also indicate that deregulated NHEJ plays a major role in generating the genomic instability and cytotoxicity in HR-deficient cells treated with PARP inhibitors.

Integration of Pan-Cancer Cell Line and Single-Cell Transcriptomic Profiles Enables Inference of Therapeutic Vulnerabilities in Heterogeneous Tumors.

Single-cell RNA sequencing (scRNA-seq) greatly advanced the understanding of intratumoral heterogeneity by identifying distinct cancer cell subpopulations. However, translating biological differences into treatment strategies is challenging due to a lack of tools to facilitate efficient drug discovery that tackles heterogeneous tumors. Developing such approaches requires accurate prediction of drug response at the single-cell level to offer therapeutic options to specific cell subpopulations. Here, we developed a transparent computational framework (nicknamed scIDUC) to predict therapeutic efficacies on an individual cell basis by integrating single-cell transcriptomic profiles with large, data-rich pan-cancer cell line screening data sets. This method achieved high accuracy in separating cells into their correct cellular drug response statuses. In three distinct prospective tests covering different diseases (rhabdomyosarcoma, pancreatic ductal adenocarcinoma, and castration-resistant prostate cancer), the predicted results using scIDUC were accurate and mirrored biological expectations. In the first two tests, the framework identified drugs for cell subpopulations that were resistant to standard-of-care (SOC) therapies due to intrinsic resistance or tumor microenvironmental effects, and the results showed high consistency with experimental findings from the original studies. In the third test using newly generated SOC therapy-resistant cell lines, scIDUC identified efficacious drugs for the resistant line, and the predictions were validated with in vitro experiments. Together, this study demonstrates the potential of scIDUC to quickly translate scRNA-seq data into drug responses for individual cells, displaying the potential as a tool to improve the treatment of heterogenous tumors. SIGNIFICANCE: A versatile method that infers cell-level drug response in scRNA-seq data facilitates the development of therapeutic strategies to target heterogeneous subpopulations within a tumor and address issues such as treatment failure and resistance.

An integrated single-cell RNA-seq map of human neuroblastoma tumors and preclinical models uncovers divergent mesenchymal-like gene expression programs.

BACKGROUND: Neuroblastoma is a common pediatric cancer, where preclinical studies suggest that a mesenchymal-like gene expression program contributes to chemotherapy resistance. However, clinical outcomes remain poor, implying we need a better understanding of the relationship between patient tumor heterogeneity and preclinical models. RESULTS: Here, we generate single-cell RNA-seq maps of neuroblastoma cell lines, patient-derived xenograft models (PDX), and a genetically engineered mouse model (GEMM). We develop an unsupervised machine learning approach ("automatic consensus nonnegative matrix factorization" (acNMF)) to compare the gene expression programs found in preclinical models to a large cohort of patient tumors. We confirm a weakly expressed, mesenchymal-like program in otherwise adrenergic cancer cells in some pre-treated high-risk patient tumors, but this appears distinct from the presumptive drug-resistance mesenchymal programs evident in cell lines. Surprisingly, however, this weak-mesenchymal-like program is maintained in PDX and could be chemotherapy-induced in our GEMM after only 24 h, suggesting an uncharacterized therapy-escape mechanism. CONCLUSIONS: Collectively, our findings improve the understanding of how neuroblastoma patient tumor heterogeneity is reflected in preclinical models, provides a comprehensive integrated resource, and a generalizable set of computational methodologies for the joint analysis of clinical and pre-clinical single-cell RNA-seq datasets.

An integrated single-cell RNA-seq map of human neuroblastoma tumors and preclinical models uncovers divergent mesenchymal-like gene expression programs.

Neuroblastoma is a common pediatric cancer, where preclinical studies suggest that a mesenchymal-like gene expression program contributes to chemotherapy resistance. However, clinical outcomes remain poor, implying we need a better understanding of the relationship between patient tumor heterogeneity and preclinical models. Here, we generated single-cell RNA-seq maps of neuroblastoma cell lines, patient-derived xenograft models (PDX), and a genetically engineered mouse model (GEMM). We developed an unsupervised machine learning approach ('automatic consensus nonnegative matrix factorization' (acNMF)) to compare the gene expression programs found in preclinical models to a large cohort of patient tumors. We confirmed a weakly expressed, mesenchymal-like program in otherwise adrenergic cancer cells in some pre-treated high-risk patient tumors, but this appears distinct from the presumptive drug-resistance mesenchymal programs evident in cell lines. Surprisingly however, this weak-mesenchymal-like program was maintained in PDX and could be chemotherapy-induced in our GEMM after only 24 hours, suggesting an uncharacterized therapy-escape mechanism. Collectively, our findings improve the understanding of how neuroblastoma patient tumor heterogeneity is reflected in preclinical models, provides a comprehensive integrated resource, and a generalizable set of computational methodologies for the joint analysis of clinical and pre-clinical single-cell RNA-seq datasets.

A spatial cell atlas of neuroblastoma reveals developmental, epigenetic and spatial axis of tumor heterogeneity.

Neuroblastoma is a pediatric cancer arising from the developing sympathoadrenal lineage with complex inter- and intra-tumoral heterogeneity. To chart this complexity, we generated a comprehensive cell atlas of 55 neuroblastoma patient tumors, collected from two pediatric cancer institutions, spanning a range of clinical, genetic, and histologic features. Our atlas combines single-cell/nucleus RNA-seq (sc/scRNA-seq), bulk RNA-seq, whole exome sequencing, DNA methylation profiling, spatial transcriptomics, and two spatial proteomic methods. Sc/snRNA-seq revealed three malignant cell states with features of sympathoadrenal lineage development. All of the neuroblastomas had malignant cells that resembled sympathoblasts and the more differentiated adrenergic cells. A subset of tumors had malignant cells in a mesenchymal cell state with molecular features of Schwann cell precursors. DNA methylation profiles defined four groupings of patients, which differ in the degree of malignant cell heterogeneity and clinical outcomes. Using spatial proteomics, we found that neuroblastomas are spatially compartmentalized, with malignant tumor cells sequestered away from immune cells. Finally, we identify spatially restricted signaling patterns in immune cells from spatial transcriptomics. To facilitate the visualization and analysis of our atlas as a resource for further research in neuroblastoma, single cell, and spatial-omics, all data are shared through the Human Tumor Atlas Network Data Commons at www.humantumoratlas.org.

Enhanced killing of cancer cells by poly(ADP-ribose) polymerase inhibitors and topoisomerase I inhibitors reflects poisoning of both enzymes.

Poly(ADP-ribose) polymerase-1 (PARP1) plays critical roles in the regulation of DNA repair. Accordingly, small molecule inhibitors of PARP are being developed as agents that could modulate the activity of genotoxic chemotherapy, such as topoisomerase I poisons. In this study we evaluated the ability of the PARP inhibitor veliparib to enhance the cytotoxicity of the topoisomerase I poisons topotecan and camptothecin (CPT). Veliparib increased the cell cycle and cytotoxic effects of topotecan in multiple cell line models. Importantly, this sensitization occurred at veliparib concentrations far below those required to substantially inhibit poly(ADP-ribose) polymer synthesis and at least an order of magnitude lower than those involved in selective killing of homologous recombination-deficient cells. Further studies demonstrated that veliparib enhanced the effects of CPT in wild-type mouse embryonic fibroblasts (MEFs) but not Parp1(-/-) MEFs, confirming that PARP1 is the critical target for this sensitization. Importantly, parental and Parp1(-/-) MEFs had indistinguishable CPT sensitivities, ruling out models in which PARP1 catalytic activity plays a role in protecting cells from topoisomerase I poisons. To the contrary, cells were sensitized to CPT in a veliparib-independent manner upon transfection with PARP1 E988K, which lacks catalytic activity, or the isolated PARP1 DNA binding domain. These results are consistent with a model in which small molecule inhibitors convert PARP1 into a protein that potentiates the effects of topoisomerase I poisons by binding to damaged DNA and preventing its normal repair.

Inferring therapeutic vulnerability within tumors through integration of pan-cancer cell line and single-cell transcriptomic profiles.

Single-cell RNA sequencing greatly advanced our understanding of intratumoral heterogeneity through identifying tumor subpopulations with distinct biologies. However, translating biological differences into treatment strategies is challenging, as we still lack tools to facilitate efficient drug discovery that tackles heterogeneous tumors. One key component of such approaches tackles accurate prediction of drug response at the single-cell level to offer therapeutic options to specific cell subpopulations. Here, we present a transparent computational framework (nicknamed scIDUC) to predict therapeutic efficacies on an individual-cell basis by integrating single-cell transcriptomic profiles with large, data-rich pan-cancer cell line screening datasets. Our method achieves high accuracy, with predicted sensitivities easily able to separate cells into their true cellular drug resistance status as measured by effect size (Cohen's d > 1.0). More importantly, we examine our method's utility with three distinct prospective tests covering different diseases (rhabdomyosarcoma, pancreatic ductal adenocarcinoma, and castration-resistant prostate cancer), and in each our predicted results are accurate and mirrored biological expectations. In the first two, we identified drugs for cell subpopulations that are resistant to standard-of-care (SOC) therapies due to intrinsic resistance or effects of tumor microenvironments. Our results showed high consistency with experimental findings from the original studies. In the third test, we generated SOC therapy resistant cell lines, used scIDUC to identify efficacious drugs for the resistant line, and validated the predictions with in-vitro experiments. Together, scIDUC quickly translates scRNA-seq data into drug response for individual cells, displaying the potential as a first-line tool for nuanced and heterogeneity-aware drug discovery.

Neuroblastoma: When differentiation goes awry.

Neuroblastoma is a leading cause of cancer-related death in children. Accumulated data suggest that differentiation arrest of the neural-crest-derived sympathoadrenal lineage contributes to neuroblastoma formation. The developmental arrest of these cell types explains many biological features of the disease, including its cellular heterogeneity, mutational spectrum, spontaneous regression, and response to drugs that induce tumor cell differentiation. In this review, we provide evidence that supports the notion that arrested neural-crest-derived progenitor cells give rise to neuroblastoma and discuss how this concept could be exploited for clinical management of the disease.

Identification of a modular super-enhancer in murine retinal development.

Super-enhancers are expansive regions of genomic DNA comprised of multiple putative enhancers that contribute to the dynamic gene expression patterns during development. This is particularly important in neurogenesis because many essential transcription factors have complex developmental stage- and cell-type specific expression patterns across the central nervous system. In the developing retina, Vsx2 is expressed in retinal progenitor cells and is maintained in differentiated bipolar neurons and Müller glia. A single super-enhancer controls this complex and dynamic pattern of expression. Here we show that deletion of one region disrupts retinal progenitor cell proliferation but does not affect cell fate specification. The deletion of another region has no effect on retinal progenitor cell proliferation but instead leads to a complete loss of bipolar neurons. This prototypical super-enhancer may serve as a model for dissecting the complex gene expression patterns for neurogenic transcription factors during development. Moreover, it provides a unique opportunity to alter expression of individual transcription factors in particular cell types at specific stages of development. This provides a deeper understanding of function that cannot be achieved with traditional knockout mouse approaches.

The myogenesis program drives clonal selection and drug resistance in rhabdomyosarcoma.

Rhabdomyosarcoma (RMS) is a pediatric cancer with features of skeletal muscle; patients with unresectable or metastatic RMS fare poorly due to high rates of disease recurrence. Here, we use single-cell and single-nucleus RNA sequencing to show that RMS tumors recapitulate the spectrum of embryonal myogenesis. Using matched patient samples from a clinical trial and orthotopic patient-derived xenografts (O-PDXs), we show that chemotherapy eliminates the most proliferative component with features of myoblasts within embryonal RMS; after treatment, the immature population with features of paraxial mesoderm expands to reconstitute the developmental hierarchy of the original tumor. We discovered that this paraxial mesoderm population is dependent on EGFR signaling and is sensitive to EGFR inhibitors. Taken together, these data serve as a proof of concept that targeting each developmental state in embryonal RMS is an effective strategy for improving outcomes by preventing disease recurrence.

EP300 Selectively Controls the Enhancer Landscape of MYCN-Amplified Neuroblastoma.

Gene expression is regulated by promoters and enhancers marked by histone H3 lysine 27 acetylation (H3K27ac), which is established by the paralogous histone acetyltransferases (HAT) EP300 and CBP. These enzymes display overlapping regulatory roles in untransformed cells, but less characterized roles in cancer cells. We demonstrate that the majority of high-risk pediatric neuroblastoma (NB) depends on EP300, whereas CBP has a limited role. EP300 controls enhancer acetylation by interacting with TFAP2ß, a transcription factor member of the lineage-defining transcriptional core regulatory circuitry (CRC) in NB. To disrupt EP300, we developed a proteolysis-targeting chimera (PROTAC) compound termed "JQAD1" that selectively targets EP300 for degradation. JQAD1 treatment causes loss of H3K27ac at CRC enhancers and rapid NB apoptosis, with limited toxicity to untransformed cells where CBP may compensate. Furthermore, JQAD1 activity is critically determined by cereblon (CRBN) expression across NB cells. SIGNIFICANCE: EP300, but not CBP, controls oncogenic CRC-driven transcription in high-risk NB by binding TFAP2ß. We developed JQAD1, a CRBN-dependent PROTAC degrader with preferential activity against EP300 and demonstrated its activity in NB. JQAD1 has limited toxicity to untransformed cells and is effective in vivo in a CRBN-dependent manner. This article is highlighted in the In This Issue feature, p. 587.

Mesenchymal and adrenergic cell lineage states in neuroblastoma possess distinct immunogenic phenotypes.

Apart from the anti-GD2 antibody, immunotherapy for neuroblastoma has had limited success due to immune evasion mechanisms, coupled with an incomplete understanding of predictors of response. Here, from bulk and single-cell transcriptomic analyses, we identify a subset of neuroblastomas enriched for transcripts associated with immune activation and inhibition and show that these are predominantly characterized by gene expression signatures of the mesenchymal lineage state. By contrast, tumors expressing adrenergic lineage signatures are less immunogenic. The inherent presence or induction of the mesenchymal state through transcriptional reprogramming or therapy resistance is accompanied by innate and adaptive immune gene activation through epigenetic remodeling. Mesenchymal lineage cells promote T cell infiltration by secreting inflammatory cytokines, are efficiently targeted by cytotoxic T and natural killer cells and respond to immune checkpoint blockade. Together, we demonstrate that distinct immunogenic phenotypes define the divergent lineage states of neuroblastoma and highlight the immunogenic potential of the mesenchymal lineage.

A single-cell and single-nucleus RNA-Seq toolbox for fresh and frozen human tumors.

Single-cell genomics is essential to chart tumor ecosystems. Although single-cell RNA-Seq (scRNA-Seq) profiles RNA from cells dissociated from fresh tumors, single-nucleus RNA-Seq (snRNA-Seq) is needed to profile frozen or hard-to-dissociate tumors. Each requires customization to different tissue and tumor types, posing a barrier to adoption. Here, we have developed a systematic toolbox for profiling fresh and frozen clinical tumor samples using scRNA-Seq and snRNA-Seq, respectively. We analyzed 216,490 cells and nuclei from 40 samples across 23 specimens spanning eight tumor types of varying tissue and sample characteristics. We evaluated protocols by cell and nucleus quality, recovery rate and cellular composition. scRNA-Seq and snRNA-Seq from matched samples recovered the same cell types, but at different proportions. Our work provides guidance for studies in a broad range of tumors, including criteria for testing and selecting methods from the toolbox for other tumors, thus paving the way for charting tumor atlases.

Author Correction: A single-cell and single-nucleus RNA-Seq toolbox for fresh and frozen human tumors.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Gene expression profiles of the rat brain both immediately and 3 months following acute sarin exposure.

We have studied sarin-induced global gene expression patterns at an early time point (15 min; 0.5xLD50) and a later time point (3 months; 1xLD50) using Affymetrix: Rat Neurobiology U34 chips in male, Sprague-Dawley rats and have identified a total of 65 (early) and 38 (late) genes showing statistically significant alterations from control levels at 15 min and 3 months, respectively. At the early time point, those that are classified as ion channel, cytoskeletal and cell adhesion molecules, in addition to neuropeptides and their receptors predominated over all other groups. The other groups included: cholinergic signaling, calcium channel and binding proteins, transporters, chemokines, GABAnergic, glutamatergic, aspartate, catecholaminergic, nitric oxide synthase, purinergic, and serotonergic signaling molecules. At the late time point, genes that are classified as calcium channel and binding proteins, cytoskeletal and cell adhesion molecules and GABAnergic signaling molecules were most prominent. Seven molecules (Ania-9, Arrb-1, CX-3C, Gabab-1d, Nos-2a, Nrxn-1b, PDE2) were identified that showed altered persistent expression in both time points. Selected genes from each of these time points were further validated using semi quantitative RT-PCR approaches. Some of the genes that were identified in the present study have been shown to be involved in organophosphate-induced neurotoxicity by both other groups as well as ours. Principal component analysis (PCA) of the expression data from both time points was used for comparative analysis of the gene expression, which indicated that the changes in gene expression were a function of dose and time of euthanasia after the treatment. Our model also predicts that besides dose and duration of post-treatment period, age and possibly other factors may be playing important roles in the regulation of pathways, leading to the neurotoxicity.

Poly (ADP-Ribose) Polymerase Inhibitor Hypersensitivity in Aggressive Myeloproliferative Neoplasms.

PURPOSE: DNA repair defects have been previously reported in myeloproliferative neoplasms (MPN). Inhibitors of PARP have shown activity in solid tumors with defects in homologous recombination (HR). This study was performed to assess MPN sensitivity to PARP inhibitors ex vivo EXPERIMENTAL DESIGN: HR pathway integrity in circulating myeloid cells was evaluated by assessing the formation of RAD51 foci after treatment with ionizing radiation or PARP inhibitors. Sensitivity of MPN erythroid and myeloid progenitors to PARP inhibitors was evaluated using colony formation assays. RESULTS: Six of 14 MPN primary samples had reduced formation of RAD51 foci after exposure to ionizing radiation, suggesting impaired HR. This phenotype was not associated with a specific MPN subtype, JAK2 mutation status, or karyotype. MPN samples showed increased sensitivity to the PARP inhibitors veliparib and olaparib compared with normal myeloid progenitors. This hypersensitivity, which was most pronounced in samples deficient in DNA damage-induced RAD51 foci, was observed predominantly in samples from patients with diagnoses of chronic myelogenous leukemia, chronic myelomonocytic leukemia, or unspecified myelodysplastic/MPN overlap syndromes. CONCLUSIONS: Like other neoplasms with HR defects, MPNs exhibit PARP inhibitor hypersensitivity compared with normal marrow. These results suggest that further preclinical and possibly clinical study of PARP inhibitors in MPNs is warranted. Clin Cancer Res; 22(15); 3894-902. ©2016 AACR.

Sarin (nerve agent GB)-induced differential expression of mRNA coding for the acetylcholinesterase gene in the rat central nervous system.

We carried out a time-course study on the effects of a single intramuscular (i.m.) dose (0.5x LD(50)) of sarin (O-isopropyl methylphosphonofluoridate), also known as nerve agent GB, on the mRNA expression of acetylcholinesterase (AChE) in the brain of male Sprague-Dawley rats. Sarin inactivates the enzyme AChE which is responsible for the breakdown of the neurotransmitter acetylcholine (ACh), leading to its accumulation at ACh receptors and overstimulation of the cholinergic system. Rats were treated with 50 microg/kg of sarin (0.5x LD(50)) in 1 mL saline/kg and terminated at the following time points: 1 and 2 hr and 1, 3, and 7 days post-treatment. Control rats were treated with normal saline. Total RNA was extracted, and northern blots were hybridized with cDNA probes for AChE and 28S RNA (control). Poly-A RNA from both treated and control cortex was used for reverse transcription-polymerase chain reaction (RT-PCR)-based verification of the data from the northern blots. The results obtained indicate that a single (i.m.) dose of sarin (0.5x LD(50)) produced differential induction and persistence of AChE mRNA levels in different regions of the brain. Immediate induction of AChE transcripts was noted in the brainstem (126+/-6%), cortex (149+/-4%), midbrain (153+/-5%), and cerebellum (234+/-2%) at 1 hr. The AChE expression level, however, increased over time and remained elevated after a decline at 1 day in the previously shown more susceptible brainstem. The transcript levels remained elevated at a later time point (3 days) in the midbrain, after a dramatic decline at day 1 (110+/-2%). In the cortex, transcript levels came down to control values by day 1. The cerebellum also showed a decline of the elevated levels observed at 2 hr (275+/-2%) to control values by day 1. RT-PCR analysis of the AChE transcript at 30 min in the cortex showed an induction to 213+/-3% of the control level, confirming the expression pattern obtained by the northern blot data. The immediate induction followed by the complex pattern of the AChE mRNA time-course in the CNS may indicate that the activation of both cholinergic-related and unrelated functions of the gene plays an important role in the pathological manifestations of sarin-induced neurotoxicity.

The Elephant and the Blind Men: Making Sense of PARP Inhibitors in Homologous Recombination Deficient Tumor Cells.

Poly(ADP-ribose) polymerase 1 (PARP1) is an important component of the base excision repair (BER) pathway as well as a regulator of homologous recombination (HR) and non-homologous end-joining (NHEJ). Previous studies have demonstrated that treatment of HR-deficient cells with PARP inhibitors results in stalled and collapsed replication forks. Consequently, HR-deficient cells are extremely sensitive to PARP inhibitors. Several explanations have been advanced to explain this so-called synthetic lethality between HR deficiency and PARP inhibition: (i) reduction of BER activity leading to enhanced DNA double-strand breaks, which accumulate in the absence of HR; (ii) trapping of inhibited PARP1 at sites of DNA damage, which prevents access of other repair proteins; (iii) failure to initiate HR by poly(ADP-ribose) polymer-dependent BRCA1 recruitment; and (iv) activation of the NHEJ pathway, which selectively induces error-prone repair in HR-deficient cells. Here we review evidence regarding these various explanations for the ability of PARP inhibitors to selectively kill HR-deficient cancer cells and discuss their potential implications.