Inactive PARP1 causes embryonic lethality and genome instability in a dominant-negative manner

Proteogenomic reprogramming to a functional human blastomere-like stem cell state via a PARP-DUX4 regulatory axis

Here, we show that conventional human pluripotent stem cells cultured with non-specific tankyrase-PARP1-inhibited conditions underwent proteogenomic reprogramming to functional blastomere-like tankyrase/PARP inhibitor-regulated naive stem cells (TIRN-SC). TIRN-SCs concurrently expressed hundreds of pioneer factors in hybrid 2C-8C-morula-ICM programs that were augmented by induced expression of DUX4. Injection of TIRN-SCs into 8C-staged murine embryos equipotently differentiated human cells to the extra-embryonic and embryonic compartments of chimeric blastocysts and fetuses. Ectopic expression of murine-E-Cadherin in TIRN-SCs further enhanced interspecific chimeric tissue targeting. TIRN-SC-derived trophoblast stem cells efficiently generated placental chimeras. Proteome-ubiquitinome analyses revealed increased TNKS and reduced PARP1 levels and an ADP-ribosylation-deficient, hyper-ubiquitinated proteome that impacted expression of both tankyrase and PARP1 substrates. ChIP-seq of NANOG-SOX2-OCT4 and PARP1 (NSOP) revealed genome-wide NSOP co-binding at DUX4-accessible enhancers of embryonic lineage factors; suggesting a DUX4-NSOP axis regulated TIRN-SC lineage plasticity. TIRN-SCs may serve as valuable models for studying the proteogenomic regulation of pre-lineage human embryogenesis. VIDEO ABSTRACT.

A PARP2 active site helix melts to permit DNA damage-induced enzymatic activation

Poly(ADP-ribose) polymerase 1 (PARP1) and PARP2 recognize DNA breaks immediately upon their formation, generate a burst of local PARylation to signal their location, and are co-targeted by all current FDA-approved forms of PARP inhibitors (PARPi) used in the cancer clinic. Recent evidence indicates that the same PARPi molecules impact PARP2 differently from PARP1, raising the possibility that allosteric activation may also differ. We find that, unlike for PARP1, destabilization of the autoinhibitory domain of PARP2 is insufficient for DNA damage-induced catalytic activation. Rather, PARP2 activation requires further unfolding of an active site helix. In contrast, the corresponding helix in PARP1 only transiently forms, even prior to engaging DNA. Only one clinical PARPi, Olaparib, stabilizes the PARP2 active site helix, representing a structural feature with the potential to discriminate small molecule inhibitors. Collectively, our findings reveal unanticipated differences in local structure and changes in activation-coupled backbone dynamics between human PARP1 and PARP2.

XRCC1 mediates PARP1- and PAR-dependent recruitment of PARP2 to DNA damage sites

Poly-ADP-ribose polymerases 1 and 2 (PARP1 and 2) are critical sensors of DNA-strand breaks and targets for cancer therapy. Upon DNA damage, PARP1 and 2 synthesize poly-ADP-ribose (PAR) chains on themselves and other substrates, facilitating DNA single-strand break repair by recruiting PAR-binding DNA repair factors, including X-ray repair cross-complementing group 1 (XRCC1) and aprataxin and polynucleotide kinase phosphatase-like factor (APLF). While diverse DNA lesions activate PARP1, PARP2 is selectively activated by 5' phosphorylated nicks. They function independently and compensate for each other. Previous studies suggest that PARP1 and its PAR chains act upstream to recruit PARP2 to DNA damage sites. Here, we report that the scaffold protein XRCC1 mediates PARP1- and PAR-dependent recruitment of PARP2 to damage sites. XRCC1-deficiency causes hyperactivation of PARP1 while attenuating micro-irradiation-induced PARP2 foci. Mechanistically, the BRCT1 domain of XRCC1 binds to PAR, while its BRCT2 domain interacts with the PARP2 catalytic domain independently of the PARP2 enzymatic activity and the LIG3 BRCT domain via residues D575 and Y576. This mode of PARP2 enrichment is important for the recruitment of certain PAR-binding proteins, such as APLF, but dispensable for others, such as the XRCC1-BRCT1 domain. These findings highlight the distinct role of PARP1 and PARP2 in PAR synthesis and uncover unexpected hierarchical roles of PARP1 and XRCC1 upstream of PARP2.

Caspase 3 and caspase 7 promote cytoprotective autophagy and the DNA damage response during non-lethal stress conditions in human breast cancer cells

Cell stress adaptation plays a key role in normal development and in various diseases including cancer. Caspases are activated in response to cell stress, and growing evidence supports their function in non-apoptotic cellular processes. A role for effector caspases in promoting stress-induced cytoprotective autophagy was demonstrated in Drosophila, but has not been explored in the context of human cells. We found a functionally conserved role for effector caspase 3 (CASP3) and caspase 7 (CASP7) in promoting starvation or proteasome inhibition-induced cytoprotective autophagy in human breast cancer cells. The loss of CASP3 and CASP7 resulted in an increase in PARP1 cleavage, reduction in LC3B and ATG7 transcript levels, and a reduction in H2AX phosphorylation, consistent with a block in autophagy and DNA damage-induced stress response pathways. Surprisingly, in non-lethal cell stress conditions, CASP7 underwent non-canonical processing at two calpain cleavage sites flanking a PARP1 exosite, resulting in stable CASP7-p29/p30 fragments. Expression of CASP7-p29/p30 fragment(s) could rescue H2AX phosphorylation in the CASP3 and CASP7 double knockout background. Strikingly, yet consistent with these phenotypes, the loss of CASP3 and CASP7 exhibited synthetic lethality with BRCA1 loss. These findings support a role for human caspases in stress adaptation through PARP1 modulation and reveal new therapeutic avenues for investigation.

DNA damage response inhibitors in cancer therapy: lessons from the past, current status and future implications

The DNA damage response (DDR) is a network of proteins that coordinate DNA repair and cell-cycle checkpoints to prevent damage being transmitted to daughter cells. DDR defects lead to genomic instability, which enables tumour development, but they also create vulnerabilities that can be used for cancer therapy. Historically, this vulnerability has been taken advantage of using DNA-damaging cytotoxic drugs and radiotherapy, which are more toxic to tumour cells than to normal tissues. However, the discovery of the unique sensitivity of tumours defective in the homologous recombination DNA repair pathway to PARP inhibition led to the approval of six PARP inhibitors worldwide and to a focus on making use of DDR defects through the development of other DDR-targeting drugs. Here, we analyse the lessons learnt from PARP inhibitor development and how these may be applied to new targets to maximize success. We explore why, despite so much research, no other DDR inhibitor class has been approved, and only a handful have advanced to later-stage clinical trials. We discuss why more reliable predictive biomarkers are needed, explore study design from past and current trials, and suggest alternative models for monotherapy and combination studies. Targeting multiple DDR pathways simultaneously and potential combinations with anti-angiogenic agents or immune checkpoint inhibitors are also discussed.

The complex universe of inactive PARP1

Poly(ADP-ribose) polymerase 1 (PARP1) is a crucial member of the PARP family, which modifies targets through ADP-ribosylation and plays key roles in a variety of biological processes. PARP inhibitors (PARPis) hinder ADP-ribosylation and lead to the retention of PARP1 at the DNA lesion (also known as trapping), which underlies their toxicity. However, inhibitors and mutations that make PARP1 inactive do not necessarily correlate with trapping potency, challenging the current understanding of inactivation-caused trapping. Recent studies on mouse models indicate that both trapping and non-trapping inactivating mutations of PARP1 lead to embryonic lethality, suggesting the unexpected toxicity of the current inhibition strategy. The allosteric model, complicated automodification, and various biological functions of PARP1 all contribute to the complexity of PARP1 inactivation.

PARG inhibition induces nuclear aggregation of PARylated PARP1

Poly (ADP-ribose) glycohydrolase (PARG) inhibitors are currently under clinical development for the treatment of DNA repair-deficient cancers; however, their precise mechanism of action is still unclear. Here, we report that PARG inhibition leads to excessive PARylated poly (ADP-ribose) polymerase 1 (PARP1) reducing the ability of PARP1 to properly localize to sites of DNA damage. Strikingly, the mis-localized PARP1 accumulates as aggregates throughout the nucleus. Abrogation of the catalytic activity of PARP1 prevents aggregate formation, indicating that PAR chains play a key role in this process. Finally, we find that PARP1 nuclear aggregates were highly persistent and were associated with cleaved cytoplasmic PARP1, ultimately leading to cell death. Overall, our data uncover an unexpected mechanism of PARG inhibitor cytotoxicity, which will shed light on the use of these drugs as anti-cancer therapeutics.

Inactive Parp2 causes Tp53-dependent lethal anemia by blocking replication-associated nick ligation in erythroblasts

Poly (ADP-ribose) polymerase (PARP) 1 and 2 enzymatic inhibitors (PARPi) are promising cancer treatments. But recently, their use has been hindered by unexplained severe anemia and treatment-related leukemia. In addition to enzymatic inhibition, PARPi also trap PARP1 and 2 at DNA lesions. Here we report that, unlike Parp2-/- mice, which develop normally, mice expressing catalytically inactive Parp2 (E534A and Parp2EA/EA) succumb to Tp53- and Chk2-dependent erythropoietic failure in utero, mirroring Lig1-/- mice. While DNA damage mainly activates PARP1, we demonstrate that DNA replication activates PARP2 robustly. PARP2 is selectively recruited and activated by 5'-phosphorylated nicks (5'p-nicks), including those between Okazaki fragments, resolved by ligase 1 (Lig1) and Lig3. Inactive PARP2, but not its active form or absence, impedes Lig1- and Lig3-mediated ligation, causing dose-dependent replication fork collapse, which is detrimental to erythroblasts with ultra-fast forks. This PARylation-dependent structural function of PARP2 at 5'p-nicks explains the detrimental effects of PARP2 inactivation on erythropoiesis, shedding light on PARPi-induced anemia and the selection for TP53/CHK2 loss.

Adaptive genetic mechanisms in mammalian Parp1 locus

Poly(ADP-ribose) polymerase 1 (PARP1) is a highly conserved nuclear protein in multicellular organisms that by modulating chromatin opening facilitates gene expression during development. All reported Parp1 null knockout mouse strains are viable with no developmental anomalies. It was believed that functional redundancy with other PARP family members, mainly PARP2, explains such a controversy. However, while PARP2 has similar catalytic domain to PARP1, it lacks other domains, making the absence of developmental problems in Parp1 mice knockouts unlikely. Contrary to prior assumptions, in our analysis of the best-investigated Parp1 knockout mouse strain, we identified persistent mRNA expression, albeit at reduced levels. Transcript analysis revealed an alternatively spliced Parp1 variant lacking exon 2. Subsequent protein analysis confirmed the existence of a truncated PARP1 protein in knockout mice. The decreased level of poly(ADP-ribose) (pADPr) was detected in Parp1 knockout embryonic stem (ES) cells with western blotting analysis, but immunofluorescence staining did not detect any difference in distribution or level of pADPr in nuclei of knockout ES cells. pADPr level in double Parp1 Parg mutant ES cells greatly exceeded its amount in normal and even in hypomorph Parg mutant ES cells, suggesting the presence of functionally active PARP1. Therefore, our findings challenge the conventional understanding of PARP1 depletion effects.

Regulation of PARP1/2 and the tankyrases: emerging parallels

ADP-ribosylation is a prominent and versatile post-translational modification, which regulates a diverse set of cellular processes. Poly-ADP-ribose (PAR) is synthesised by the poly-ADP-ribosyltransferases PARP1, PARP2, tankyrase (TNKS), and tankyrase 2 (TNKS2), all of which are linked to human disease. PARP1/2 inhibitors have entered the clinic to target cancers with deficiencies in DNA damage repair. Conversely, tankyrase inhibitors have continued to face obstacles on their way to clinical use, largely owing to our limited knowledge of their molecular impacts on tankyrase and effector pathways, and linked concerns around their tolerability. Whilst detailed structure-function studies have revealed a comprehensive picture of PARP1/2 regulation, our mechanistic understanding of the tankyrases lags behind, and thereby our appreciation of the molecular consequences of tankyrase inhibition. Despite large differences in their architecture and cellular contexts, recent structure-function work has revealed striking parallels in the regulatory principles that govern these enzymes. This includes low basal activity, activation by intra- or inter-molecular assembly, negative feedback regulation by auto-PARylation, and allosteric communication. Here we compare these poly-ADP-ribosyltransferases and point towards emerging parallels and open questions, whose pursuit will inform future drug development efforts.

PARticular MARks: Histone ADP-ribosylation and the DNA damage response

Cellular and molecular responses to DNA damage are highly orchestrated and dynamic, acting to preserve the maintenance and integrity of the genome. Histone proteins bind DNA and organize the genome into chromatin. Post-translational modifications of histones have been shown to play an essential role in orchestrating the chromatin response to DNA damage by regulating the DNA damage response pathway. Among the histone modifications that contribute to this intricate network, histone ADP-ribosylation (ADPr) is emerging as a pivotal component of chromatin-based DNA damage response (DDR) pathways. In this review, we survey how histone ADPr is regulated to promote the DDR and how it impacts chromatin and other histone marks. Recent advancements have revealed histone ADPr effects on chromatin structure and the regulation of DNA repair factor recruitment to DNA lesions. Additionally, we highlight advancements in technology that have enabled the identification and functional validation of histone ADPr in cells and in response to DNA damage. Given the involvement of DNA damage and epigenetic regulation in human diseases including cancer, these findings have clinical implications for histone ADPr, which are also discussed. Overall, this review covers the involvement of histone ADPr in the DDR and highlights potential future investigations aimed at identifying mechanisms governed by histone ADPr that participate in the DDR, human diseases, and their treatments.

New Horizons of Synthetic Lethality in Cancer: Current Development and Future Perspectives

In recent years, synthetic lethality has been recognized as a solid paradigm for anticancer therapies. The discovery of a growing number of synthetic lethal targets has led to a significant expansion in the use of synthetic lethality, far beyond poly(ADP-ribose) polymerase inhibitors used to treat BRCA1/2-defective tumors. In particular, molecular targets within DNA damage response have provided a source of inhibitors that have rapidly reached clinical trials. This Perspective focuses on the most recent progress in synthetic lethal targets and their inhibitors, within and beyond the DNA damage response, describing their design and associated therapeutic strategies. We will conclude by discussing the current challenges and new opportunities for this promising field of research, to stimulate discussion in the medicinal chemistry community, allowing the investigation of synthetic lethality to reach its full potential.

A PARP2-specific active site α-helix melts to permit DNA damage-induced enzymatic activation

PARP1 and PARP2 recognize DNA breaks immediately upon their formation, generate a burst of local PARylation to signal their location, and are co-targeted by all current FDA-approved forms of PARP inhibitors (PARPi) used in the cancer clinic. Recent evidence indicates that the same PARPi molecules impact PARP2 differently from PARP1, raising the possibility that allosteric activation may also differ. We find that unlike for PARP1, destabilization of the autoinhibitory domain of PARP2 is insufficient for DNA damage-induced catalytic activation. Rather, PARP2 activation requires further unfolding of an active site α-helix absent in PARP1. Only one clinical PARPi, Olaparib, stabilizes the PARP2 active site α-helix, representing a structural feature with the potential to discriminate small molecule inhibitors. Collectively, our findings reveal unanticipated differences in local structure and changes in activation-coupled backbone dynamics between PARP1 and PARP2.

XRCC1 mediates PARP1- and PAR-dependent recruitment of PARP2 to DNA damage sites

Poly-ADP-ribose polymerases 1 and 2 (PARP1 and PARP2) are crucial sensors of DNA-strand breaks and emerging cancer therapy targets. Once activated by DNA breaks, PARP1 and PARP2 generate poly-ADP-ribose (PAR) chains on themselves and other substrates to promote DNA single-strand break repair (SSBR). PARP1 can be activated by diverse DNA lesions, whereas PARP2 specifically recognizes 5' phosphorylated nicks. They can be activated independently and provide mutual backup in the absence of the other. However, whether PARP1 and PARP2 have synergistic functions in DNA damage response remains elusive. Here, we show that PARP1 and the PAR chains generated by PARP1 recruit PARP2 to the vicinity of DNA damage sites through the scaffold protein XRCC1. Using quantitative live-cell imaging, we found that loss of XRCC1 markedly reduces irradiation-induced PARP2 foci in PARP1-proficient cells. The central BRCT domain (BRCT1) of XRCC1 binds to the PAR chain, while the C-terminal BRCT domain (BRCT2) of XRCC1 interacts with the catalytic domain of PARP2, facilitating its localization near the breaks. Together, these findings unveil a new function of XRCC1 in augmenting PARP2 recruitment in response to PARP1 activation and explain why PARP1, but not PARP2, is aggregated and hyperactivated in XRCC1-deficient cells.

Inactive Parp2 causes Tp53-dependent lethal anemia by blocking replication-associated nick ligation in erythroblasts

PARP1&2 enzymatic inhibitors (PARPi) are promising cancer treatments. But recently, their use has been hindered by unexplained severe anemia and treatment-related leukemia. In addition to enzymatic inhibition, PARPi also trap PARP1&2 at DNA lesions. Here, we report that unlike Parp2 -/- mice, which develop normally, mice expressing catalytically-inactive Parp2 (E534A, Parp2 EA/EA ) succumb to Tp53- and Chk2 -dependent erythropoietic failure in utero , mirroring Lig1 -/- mice. While DNA damage mainly activates PARP1, we demonstrate that DNA replication activates PARP2 robustly. PARP2 is selectively recruited and activated by 5'-phosphorylated nicks (5'p-nicks) between Okazaki fragments, typically resolved by Lig1. Inactive PARP2, but not its active form or absence, impedes Lig1- and Lig3-mediated ligation, causing dose-dependent replication fork collapse, particularly harmful to erythroblasts with ultra-fast forks. This PARylation-dependent structural function of PARP2 at 5'p-nicks explains the detrimental effects of PARP2 inhibition on erythropoiesis, revealing the mechanism behind the PARPi-induced anemia and leukemia, especially those with TP53/CHK2 loss.

Significance

This work shows that the hematological toxicities associated with PARP inhibitors stem not from impaired PARP1 or PARP2 enzymatic activity but rather from the presence of inactive PARP2 protein. Mechanistically, these toxicities reflect a unique role of PARP2 at 5'-phosphorylated DNA nicks during DNA replication in erythroblasts.