A One and a Two…Expanding Roles for Poly(ADP-ribose) Polymerases in Metabolism

Summary

In two related papers in this issue, Bai et al. (2011a; 2011b) observe that PARP-1- or PARP-2-deficient mice show increased energy expenditure and protection against diet-induced obesity. This metabolic phenotype is achieved, in part, through SIRT1 activation, either by increasing NAD⁺ levels or by promoting SIRT1 expression.

Poly(ADP-ribose) polymerases (PARPs) comprise a family of enzymes sharing a conserved catalytic domain that supports mono- or poly(ADP-ribosyl)transferase activity using NAD⁺ as a donor of ADP-ribose units. Members of the PARP family, such as PARP-1 and PARP-2, are involved in a wide range of molecular and cellular processes, including maintenance of genome stability, regulation of chromatin structure and transcription, cell proliferation, and apoptosis (Krishnakumar and Kraus, 2010). The focus of the field has been on the molecular mechanisms underlying the role of PARPs in these processes. Recently, however, physiological roles of PARPs have been moving to the forefront, as evidenced by two articles in this issue from Bai et al. (2011a; 2011b). In these papers, the authors demonstrate an expanded metabolic role of PARP-1 and PARP-2 upon individual genetic depletion in mice, both of which lead to increased energy expenditure and mitochondrial oxidation.

PARP-1 is the founding member of the PARP family and contributes up to 90 percent of the total cellular PARP activity. PARP-1 controls transcriptional outcomes by modulating chromatin structure altering the activity of chromatin- and transcription-related factors through poly(ADP-ribosyl)ation (PARylation) (Krishnakumar and Kraus, 2010). To investigate the role of PARP-1 in metabolic homeostasis, Bai et al. (2011b) characterized the metabolic phenotypes of PARP-1^−/− mice. Despite increased food intake, the mice were leaner with reduced fat accumulation, higher energy expenditure, enhanced glucose oxidation, and protection from diet-induced obesity than their wild-type littermates. PARP-1^−/− mice also showed increased mitochondrial content, as well as elevated oxidation- and respiration-related gene expression in key metabolic tissues, such as muscle and brown adipose tissue (BAT) (Fig. 1A).

Figure 1. Studies with PARP-1^−/− and PARP-2^−/− mice illustrate essential roles for PARP-1 and PARP-2 in metabolic homeostasis and reveal several potential modes of functional interplay with SIRT1.

(A) The schematic summarizes the effects of PARP-1 or PARP-2 knockout on a variety of metabolism-related phenotypes in five different tissues. The effects (increase or decrease upon knockout) are illustrated by the use of color-coded arrows (PARP-1, red; PARP-2, blue). Many of the effects illustrated here are based on the two papers by Bai et al., but information from other recent papers is included as well. The phenotypes shown for PARP-1 are based on PARP-1^−/− mice with a predominantly C57BL/6J background. Opposite effects for some parameters may be noted for PARP-1^−/− mice made in a predominantly SV129 background (see the text). (B) Potential mechanisms for functional interplay and potential antagonism between PARP-1 or PARP-2 and SIRT1. The schematic illustrates four different modes of functional interplay and potential antagonism between PARP-1 or PARP-2 (collectively labeled “PARP”) and SIRT1, labeled a through d.

Interestingly, the metabolic phenotypes in PARP-1^−/− mice resemble those in which SIRT1, an NAD⁺-dependent protein deacetylase, is overexpressed or chemically activated (Herranz and Serrano, 2010). These results suggest functional interplay between PARP-1 and SIRT1, perhaps since both are tightly controlled by cellular NAD⁺ levels (Fig. 1B). Indeed, elevated NAD⁺ levels were observed in both muscle and BAT in PARP-1^−/− mice, as well as increased SIRT1 activity as reflected by reduced acetylation of downstream SIRT1 targets, including the metabolic regulators PGC-1α and FOXO1 (Bai et al., 2011b). The authors also recapitulated the mitochondrial phenotypes in cell-based assays, where PARP-1 was depleted by RNAi or chemically inhibited by the drug PJ34. These effects were blunted by SIRT1 depletion, suggesting a direct link between PARP-1 and SIRT1. The authors posit that increased NAD⁺ availability might be a mechanism by which SIRT1 is activated upon PARP-1 knockout, depletion, or inhibition, although other mechanisms are also possible (Fig. 1B).

The role of PARP-1 in metabolic homeostasis, however, may not be so simple, since other genetic determinants could modify the phenotypic outcomes. For example, PARP-1^−/− mice on a predominantly SV129 (obesity-resistant) background are more susceptible to age-related weight gain and diet-induced obesity than wild-type littermates (Devalaraja-Narashimha and Padanilam, 2010; Wang et al., 1995), while PARP-1^−/− mice on a predominantly C57BL/6J (obesity-prone) background have lower weights, exhibit improved metabolic profiles, and are protected against diet-induced obesity and insulin resistance (Bai et al., 2011b; de Murcia et al., 1997). C57BL/6J mice harbor a loss-of-function deletion in the gene encoding nicotinamide nucleotide transhydrogenase, an enzyme that catalyzes the production of NAD⁺ through the reversible reduction of NADP⁺ by NADH (Freeman et al., 2006). This mutation, or perhaps a similar mutation elsewhere in the genome, might impact the phenotypes described herein. Such possibilities should be considered in future studies.

In a second study, Bai et al. (2011a) used both cell- and mouse-based studies to determine if PARP-2 plays a role in metabolic homeostasis. PARP-2, the closest paralog of PARP-1, shares ~70 percent similarity with the catalytic domain of PARP-1 and is responsible for most of the residual PARP activity in PARP-1^−/− mice. In C2C12 myocytes, the authors observed that RNAi-mediated depletion of PARP-2 did not alter cellular NAD⁺ levels, but increased the levels of SIRT1 mRNA, protein, and activity. This was reflected by reduced acetylation of downstream SIRT1 targets, as well as increased mitochondrial biogenesis and oxidation. Interestingly, the authors found that PARP-2 localizes to the SIRT1 promoter where it functions as a negative transcription regulator (Fig. 1B(b)). Thus, PARP-2 depletion is expected to cause similar metabolic effects as SIRT1 activation. Indeed, the authors found that PARP-2^−/−mice exhibit elevated energy expenditure, oxygen consumption, and lipid oxidation associated with a leaner phenotype and protection against diet-induced obesity (Fig. 1A).

The authors also observed that PARP-2^−/− mice displayed a tissue-specific mitochondrial metabolic phenotype, which was more evident in skeletal muscle and liver, and less so in BAT (Bai et al., 2011a). Surprisingly, despite an improved metabolic profile and insulin sensitivity, PARP-2^−/− mice were glucose intolerant due to pancreatic failure when fed a high fat diet. Several key genes involved in pancreatic β-cell proliferation and function were not induced in the PARP-2^−/− mice, which can be partially explained by the activation of FOXO1, a negative regulator of pancreatic growth (Fig. 1A). The tissue-specific metabolic effects, however, suggest that SIRT1 activation is not the only explanation for the observed phenotypes. Further studies are needed to resolve this issue.

These two studies fit well with other recent studies using knockout mouse models to highlight the metabolic role of PARP family members. For example, a previous study showed that PARP-2^−/− mice have defects in adipogenesis with reduced expression of adipogenic genes (Bai et al., 2007). Tankyrase (PARP-5) knockout mice exhibit increased energy expenditure, increased fatty acid and glucose utilization, and reduced adiposity (Yeh et al., 2009). Furthermore, PARP-1^−/− mice exhibit impaired food entrainment of peripheral circadian clocks, supporting a role for PARP-1 in linking circadian rhythms to metabolic homeostasis (Asher et al., 2010) (Fig. 1A).

Taken together, the studies summarized herein raise a number of challenging, but exciting, questions. What are the molecular mechanisms underlying the functional interplay of SIRT1 with PARP-1 and PARP-2? Although potential mechanisms have been suggested (Fig. 1B), these models need to be verified more thoroughly. In addition, what is the molecular basis of PARP-1- and PARP-2-dependent, SIRT1-independent outcomes? How might cellular compartmentalization of NAD⁺ pools and consumers (e.g., PARPs and SIRT1) contribute? Finally, might PARP inhibitors be useful for the treatment of metabolic diseases? Although PARP inhibitors have demonstrated safety and efficacy as potential treatments for various types of cancers (Krishnakumar and Kraus, 2010), their current lack of specificity and potential deleterious effects on DNA repair pathways make their long-term use as a treatment of metabolic diseases a questionable, but intriguing, possibility.