NAD⁺-mediated regulation of mammalian base excision repair

Abstract

The enzymes of the base excision repair (BER) pathway form DNA lesion-dependent, transient complexes that vary in composition based on the type of DNA damage [1]. These protein sub-complexes facilitate substrate/product handoff to ensure reaction completion so as to avoid accumulation of potentially toxic DNA repair intermediates [2]. However, in the mammalian cell, additional signaling molecules are required to fine-tune the activity of the BER pathway enzymes and to facilitate chromatin/histone reorganization for access to the DNA lesion for repair. These signaling enzymes include nicotinamide adenine dinucleotide (NAD⁺) dependent poly(ADP-ribose) polymerases (PARP1, PARP2) and class III deacetylases (SIRT1, SIRT6) that comprise a key PARP-NAD-SIRT axis to facilitate the regulation and coordination of BER in the mammalian cell. Here, we briefly describe the key nodes in the BER pathway that are regulated by this axis and highlight the cellular and organismal variation in NAD⁺ bioavailability that can impact BER signaling potential. We discuss how cellular NAD⁺ is required for BER to maintain genome stability and to mount a robust cellular response to DNA damage. Finally, we consider the dependence of BER on the PARP-NAD-SIRT axis for BER protein complex assembly.

1. Introduction

Base excision repair (BER) is a critical DNA repair pathway found in all species and is essential for the resolution of base lesions in both the nuclear and mitochondrial genomes [1]. As described by Wilson and Kunkel, BER enzymes handoff their enzymatic products to the next enzyme in the pathway [2], thereby avoiding the accumulation of genome destabilizing BER intermediates [3]. This is particularly significant since failure of BER contributes to genome instability, an increase in genetic mutations (base alterations, strand breaks) and can impact cellular metabolism by blocking glycolysis [4]. However, lesion access and formation of BER complexes in chromatinized DNA requires an additional layer of regulation and coordination. This may include protein post-translational modifications (e.g., acetylation/de-acetylation; ADP-ribosylation) to regulate protein function, affect protein sub-cellular location or initiate chromatin unwinding and histone modification to promote DNA access to facilitate repair, followed by chromatin/histone re-compaction. Many of these BER regulatory and/or signaling steps are controlled by nicotinamide adenine dinucleotide (NAD⁺) dependent poly(ADP-ribose) polymerases (PARP1, PARP2) and class III deacetylases (SIRT1, SIRT6) that comprise a key PARP-NAD-SIRT axis to facilitate the regulation and coordination of BER in the mammalian cell.

2. The PARP-NAD-SIRT signaling axis in base excision repair

Base excision repair in the mammalian cell depends on key protein-protein interactions [1]. However, the complexity of DNA lesion repair in the mammalian genome coupled with the impediment of chromatin compacted DNA requires an added layer of signaling and regulation (Figure 1). Broadly speaking, BER is primarily initiated by one of eleven lesion-selective DNA glycosylases [1, 5]. Further, BER can also include several sub-pathways that diverge based on the lesion including single-strand break repair (SSBR), where the DNA backbone is already cleaved and nucleotide incision repair (NIR), a minor sub-pathway in which APE1 initiates DNA glycosylase-independent repair of oxidized cytosines [1]. By encompassing all the sub-pathways, the BER mechanism can be broken into three main functional steps [1]. These steps converge with PARP1 activation (and a small contribution by PARP2) that results in the covalent modification of acceptor proteins with poly-ADP-ribose (PAR) and the recruitment of essential DNA repair proteins. Once PARP1 is activated, over 160 proteins are covalently modified by PAR or form complexes with PAR during base excision repair, highlighting the complexity of the response [4].

Fig. 1 – The PARP-NAD-SIRT signaling axis in base excision repair.

A graphical depiction of base excision repair in a mammalian cell. The critical protein-protein interactions and enzyme activities needed to conduct BER are additionally regulated by factors of the PARP-NAD-SIRT signaling axis to facilitate and regulate enzyme activities (glycosylases, APE1). Further, PARP-activation promotes chromatin and histone reorganization to allow critical BER protein complex assembly and disassembly. Cellular consequences of BER failure are also highlighted. Overall, NAD⁺ plays an essential role as a substrate for both PARP-family and SIRT-family enzymes.

A second layer of regulation and signaling to facilitate DNA repair in the mammalian cell is mediated by the sirtuins, a conserved family of proteins that are NAD⁺-dependent protein deacetylases and in some cases function as ADP-ribosyl transferases [6]. In mammals, there are seven sirtuins, namely SIRT1–SIRT7. Both SIRT1 and SIRT6 have been shown to play critical roles in BER at many levels [6]. Initiation of BER can be modulated by SIRT1- and SIRT6-mediated deacetylation of DNA glycosylases and AP endonuclease 1, APE1 (Figure 1). The role of SIRT6 in BER was first brought to light in a 2006 report, wherein SIRT6 was found to be important for DNA integrity and loss of SIRT6 promoted genomic instability and hypersensitivity to ionizing radiation, hydrogen peroxide and the DNA alkylating agent methyl methanesulfonate (MMS) [7]. SIRT6 also plays a role in the recruitment of chromatin reorganization factors such as CHD4 and SNF2H [6]. There is also crosstalk between PARP1 and SIRT1. PARP1 activity can be directly modulated by SIRT1’s deacetylase activity (Figure 1) or PARP1 can be bound and regulated by the SIRT1 binding protein DBC1. Interestingly, the SIRT1/DBC1 ⇔ PARP1/DBC1 dynamic is regulated by the cellular level of nicotinamide adenine dinucleotide (NAD⁺) [8].

The common thread running through the regulation and signaling events in BER mediated by PARP and sirtuin family proteins is the metabolite NAD⁺. As shown (Figure 1), NAD⁺ is an essential small molecule that can modulate repair initiation via SIRT1 and SIRT6 activation and can affect the primary signaling role for PARP1, impacting chromatin relaxation [9] and BER protein complex assembly (Figure 1). We therefore consider the PARP-NAD-SIRT axis as an essential regulatory and signaling component for BER.

One of the key protein sub-complexes in BER is the DNA polymerase β (Polβ)/XRCC1 heterodimer. Binding of Polβ to XRCC1 protects Polβ from ubiquitin modification and proteasome-mediated degradation [10]. Further, the recruitment of Polβ to sites of DNA damage is significantly attenuated when unable to bind to XRCC1 [10] and the absence of XRCC1 reduces the level of chromatin-bound and nuclear Polβ [11]. It is therefore important to maintain such protein interactions for the completion of repair, as there are significant cellular consequences of repair failure (Figure 1). An essential dictate of the BER handoff model is the sequestration of toxic BER intermediates. PARP1 activation leads to chromatin relaxation and the recruitment of the key BER factors XRCC1 and Polβ to complete repair. However, the absence of Polβ triggers BER failure and the hyper-accumulation of PAR, a phenotype suppressed by the expression of Polβ but not the expression of a mutant of Polβ that is unable to conduct the critical 5’dRp lyase function [3, 12]. It appears that the major activating lesion for PARP1 activation in BER is the major APE1 enzymatic product, a nick containing a 5’dRP lesion that is the substrate for the 5’dRP lyase activity of Polβ [13].

Similarly, since PARP1 activation and the synthesis of PAR is essential for chromatin relaxation and repair protein recruitment, the removal of PAR is equally critical to promote chromatin re-compaction and the disassembly of DNA repair complexes. PARG is the major factor responsible for the degradation of PAR, with ARH3 playing a role to further degrade PAR polymers and TARG1 functioning to remove the final ADP-ribose unit [14, 15]. The absence of PARG also induces repair failure, leading to elevated PAR accumulation and enhanced cell death in response to DNA damaging agents [15]. The cellular consequences of PARP1-hyperactivation that results from repair failure are complex and multifactorial. The initial outcome from PARP1 hyperactivation is the rapid loss of global cellular ATP and NAD⁺ levels with the initial loss of NAD⁺ found in the mitochondria, followed rapidly by the loss of NAD⁺ in both the cytosol and the nucleus [4]. However, the major metabolic phenotype of PARP1-hyperactivation is the inhibition of glycolysis [4]. Whereas NAD⁺ depletion enhances DNA damage-induced cell death [16], the loss of NAD⁺ does not immediately impact glycolysis. As shown (Figure 1), it is the cellular accumulation of PAR that has a direct impact on glycolysis via PAR-mediated inhibition of hexokinase 1 (HK1) [4].

As a critical factor in the PARP-NAD-SIRT axis, we find that cellular NAD⁺ is essential for the repair of genomic DNA damage, particularly with regard to BER. It is also likely that suppressed levels of NAD⁺ will impact other DNA repair pathways. In our early studies, we found that BER failure-induced cell death depends on NAD⁺ bioavailability [3]. Subsequently, we found that combined BER inhibition and NAD⁺ biosynthesis inhibition increases alkylating agent induced cell death [16]. Together, these studies highlight a critical role for NAD⁺ in the regulation and completion of BER. In-line with this concept, we probed a set of DNA damaging agents of different classes for DNA damage and repair by the CometChip Assay [17]. Cancer cells were depleted of cellular NAD⁺ by blocking the salvage pathway enzyme NAMPT and then probed for the repair of the induced DNA damage following genotoxin exposure [18]. Those agents that induce DNA damage repaired by the BER pathway, or requiring PARP1, showed suppressed repair. Further, we found that the PARP1-induced assembly of the BER complex, as measured by the formation of XRCC1-dsRED foci following laser-induced DNA damage, was blocked when NAD⁺ levels were depleted [18]. In all, we find that NAD⁺ plays a critical role in BER. While a major component of its impact on BER is linked to PARP1 activation, the requirement of NAD⁺ for sirtuin function expands the role that NAD⁺ plays in the PARP-NADSIRT axis with regard to the regulation of BER (Figure 1). Together, these studies point to a complicated and not yet resolved mechanism of BER in vivo with regard to the factors required for BER regulation, chromatin unwinding and protein complex assembly to facilitate BER and maintain genome integrity.

3. An essential role for NAD⁺ in base excision repair

As we consider the significance and impact of the PARP-NAD-SIRT axis on DNA repair, it becomes clear that there are numerous co-factors or enzyme substrates essential for the required post-translational modifications (ADP-ribosylation, acetylation/deacetylation) to drive the BER reaction forward, promote chromatin unwinding and complete repair. Substrate (NAD⁺) availability, therefore, becomes a critical factor that may regulate DNA repair capacity of a given cell type, tumor or organ (Figure 2). In this regard, NAD⁺ is an essential factor for the regulation of BER protein complex assembly and contributes to overall cellular BER capacity [3, 16, 18]. This would not be a concern if cellular NAD⁺ levels were constant throughout different cell types. However, fluctuations in cellular levels of NAD⁺ affecting the activity of NAD⁺-consuming enzymes, such as PARP1 and SIRT1, have been linked to the aging process and cancer [19]. As such, low NAD⁺ has been linked with suppressed DNA repair [18] while elevated NAD⁺ provides for robust DNA repair capacity (Figure 2).

Fig. 2 – An essential role for NAD⁺ in base excision repair.

NAD⁺ metabolism is tightly regulated within the cell. Controlled activation of NAD⁺ consumers PARP1 and SIRT1-SIRT7 is essential to maintain high NAD⁺ levels associated with a healthy outcome. However, PARP1 hyperactivation and SIRT1-SIRT7 inhibition leads to a decrease in cellular NAD⁺ levels associated with a poor pathologic outcome for many disease states. Treatment with PARP1 inhibitors or NAD⁺ precursors can reverse the pathologic outcome by increasing cellular NAD⁺ levels.

A recent study further explains the interaction between PARP1, SIRT1 and NAD⁺ in the context of aging. Both SIRT1 and PARP1 are binding partners of the protein DBC1 (deleted in breast cancer 1). DBC1, independent of NAD⁺ content, binds to SIRT1 and inhibits its activity. In contrast, the interaction between PARP1 and DBC1 depends on NAD⁺ availability (Figure 1). As a result of decreased NAD⁺ levels, aging mice demonstrated an increase in the DBC1-PARP1 complex, leading to PARP1 inhibition, reduced DNA repair and increased DNA damage. Interestingly, this outcome was reversed by supplementation with the NAD⁺-biosynthesis precursor nicotinamide mononucleotide (NMN), which increased total NAD⁺ levels, decreased the interaction between DBC1 and PARP1 and restored PARP1 activity in older mice [8]. This is just one of many examples that link changes in NAD⁺ levels due to aging and the level of DNA repair capacity. A cellular or organismal reduction in NAD⁺ is not only associated with aging and age-related diseases but is a prevalent phenotype in cancer as well, suggesting that cancer-related DNA repair defects may be, in part, the result of this NAD⁺ biosynthesis deficit [19]. Some have suggested that the aging/cancer connection may be explained by a mechanism related to this NAD⁺-biosynthesis suppression [19]. Overall, it is now clear, using both cellular and animal models, that increasing NAD⁺ bioavailability with the supplementation of NAD⁺ precursors may provide an opportunity to increase cellular DNA repair capacity (Figure 2). However, it remains to be determined if too much of a good thing (NAD⁺) can also be detrimental.

4. Laser micro-irradiation in the study of base excision repair

Laser micro-irradiation is a technique that has yielded powerful insights into the mechanism of DNA repair in which a microscope is used to focus a visible or near-visible laser on a very small region of the nucleus, leading to tight spatial and temporal control of DNA damage induction [18]. This approach has the advantage of studying processes of DNA repair in the context of intact chromatin in single cells. To assess the assembly and disassembly of repair complexes at sites of induced DNA damage, cells can be fixed and processed for immunofluorescence, or live-cell imaging can be performed with fluorescently labeled DNA repair proteins for real-time analysis (Figure 3).

Fig. 3 – Laser micro-irradiation in the study of base excision repair.

Top panels show confocal fluorescent micrographs revealing the recruitment of XRCC1-dsRED (left) or EGFP-Polβ (right) following DNA damage via laser-induced micro-irradiation. The peak time of 45 seconds shows the fluorescent protein accumulating at the site of micro-irradiation (center image) and the lack of recruitment when PARP1 is inhibited by BMN673 (right image). Plots below show the kinetics of BER protein recruitment to the site of damage in the presence of DMSO (black lines) or the PARP1 inhibitor BMN673 (blue lines).

Laser micro-irradiation has therefore led to significant mechanistic insights into proteins of the canonical BER pathway. Consistent with the critical impact of the PARPNAD-SIRT axis on BER, we have shown that suppression of cellular NAD⁺ levels, by inhibition with the NAMPT inhibitor, FK866, reduces PARP1 activation [4] and suppresses XRCC1-dsRED recruitment to sites of micro-irradiation induced DNA damage [18]. XRCC1 is the primary PARylation target of PARP1 in response to alkylating agent exposure [4]. This PARP1-mediated PARylation of XRCC1 and the binding of XRCC1 to PAR [4] provides rapid recruitment of XRCC1 to sites of DNA damage. Similarly, treatment of cells with the PARP1-inhibitor BMN673 blocks DNA damage-induced PARP1 activation [4] and blocks recruitment of fluorescently-labeled XRCC1 (XRCC1-dsRED) to sites of laser micro-irradiation induced DNA damage (Figure 3). The assembly of the BER complex is rapid, with XRCC1-dsRED peak recruitment seen from 30–45 seconds. Demonstrating the effectiveness of the PARPNAD-SIRT axis in the regulation of BER, both the loss of NAD⁺ [18] and the inhibition of PARP1 (Figure 3) blocks XRCC1-dsRED recruitment to sites of DNA damage. As indicated above and in Figure 1, Polβ forms a tight and functionally important heterodimer complex with XRCC1 [10, 11]. Like XRCC1, EGFP-Polβ is also recruited to sites of laser-induced DNA damage by a PARP1-dependent mechanism (Figure 3). Ongoing studies are needed to fully expand on the kinetics of BER protein complex assembly and disassembly, particularly due to alterations in the PARP-NAD-SIRT BER signaling and regulatory axis.

5. Discussion/Summary

In all, there are over 20 known BER proteins that participate in this handoff process [2], promoting the formation of the essential protein complexes at the site of DNA damage. As we continue to understand the subtle yet critical details of BER regulation in a mammalian cell, the number of proteins and small molecule factors that help to regulate BER are expected to expand significantly. However, these concepts began with work from Dr. Samuel H. Wilson and his laboratory, helping to define the biochemical pathway of BER, exploring concepts and techniques to evaluate these complexes in vitro and in vivo [20]. Together, these studies and those by many others have begun to reveal how these BER factors interact to facilitate repair, leading the way for further studies. Here, we hope to provide some insights in the way forward to identifying new BER regulatory factors, expanding on the significance and specificity of the PARP-NAD-SIRT BER signaling and regulatory axis.