MacroD1 Is a Promiscuous ADP-Ribosyl Hydrolase Localized to Mitochondria

Abstract

MacroD1 is a macrodomain containing protein that has mono-ADP-ribose hydrolase enzymatic activity toward several ADP-ribose adducts. Dysregulation of MacroD1 expression has been shown to be associated with the pathogenesis of several forms of cancer. To date, the physiological functions and sub-cellular localization of MacroD1 are unclear. Previous studies have described nuclear and cytosolic functions of MacroD1. However, in this study we show that endogenous MacroD1 protein is highly enriched within mitochondria. We also show that MacroD1 is highly expressed in human and mouse skeletal muscle. Furthermore, we show that MacroD1 can efficiently remove ADP-ribose from 5' and 3'-phosphorylated double stranded DNA adducts in vitro. Overall, we have shown that MacroD1 is a mitochondrial protein with promiscuous enzymatic activity that can target the ester bonds of ADP-ribosylated phosphorylated double-stranded DNA ends. These findings have exciting implications for MacroD1 and ADP-ribosylation within the regulation of mitochondrial function and DNA-damage in vivo.

Keywords: ADP-ribose; ADP-ribosylation; DNA damage repair; PARP; hydrolase; macrodomain; mitochondria.

FIGURE 1

MacroD1 is differentially expressed in human and mouse. (A) Schematic diagram of human MacroD1 macrodomain. MTS^∗ – predicted mitochondrial targeting sequence using MitoProt II – v1.101 (Claros and Vincens, 1996). (B) RNA-seq analysis of MacroD1 (red) mRNA transcript levels in human tissues. Data from GTEx (http://www.gtexportal.org). Data shown as median mRNA transcript levels. RPMK – reads per kilobase per million mapped reads. (C) Immunoblot analysis of MacroD1 steady state protein levels in HeLa cells following 72 or 96 h MacroD1 gene silencing by control siRNA, MacroD1 siRNA 1 or MacroD1 siRNA 2. (D) Immunoblot analysis of MacroD1 steady state protein levels in human cell lines. (E) Immunoblot analysis of Macrod1 steady state protein levels in tissues from Macrod1^+/+ and Macrod1^-/- mice. BAT, brown adipose tissue; WAT, white adipose tissue; short, short exposure time; medium, medium exposure time.

FIGURE 2

Endogenous MacroD1 localizes to mitochondria. (A) Immunoblot analysis of endogenous MacroD1 protein in whole cell extracts (WCE) and cytosolic [glyceraldehyde 3-phosphate dehydrogenase (GAPDH)], nuclear [DNA ligase 3 (LIG3)], microsomal and mitochondrial [ATP synthase, H+ transporting, mitochondrial F1 complex, alpha subunit 1, cardiac muscle (ATP5A)] sub-cellular fractions from HeLa cells. (B) Indirect immunofluorescence imaging of RD cells following 96 h MacroD1 gene silencing by control siRNA, MacroD1 siRNA 1 or MacroD1 siRNA 2 co-stained with MacroD1 (green), ATP5A (red), and DAPI (blue). Scale bar = 10 μm.

FIGURE 3

The N-terminus of MacroD1 is required for mitochondrial localization and import. (A) Schematic diagram of full-length MacroD1 (MacroD1-GFP) and N-terminal truncated MacroD1 (Δ77MacroD1-GFP) constructs. MTS^∗ – Predicted mitochondrial targeting sequence was ascribed using the MTS cleavage site as predicted by MitoFates, located between amino acid 77 and 78 (Fukasawa et al., 2015). (B) MTS probability prediction scores for MacroD1-GFP and Δ77MacroD1-GFP cDNA sequences using MitoFates (Fukasawa et al., 2015), TargetP 1.1 Server (Emanuelsson et al., 2000) and MitoProt II – v1.101 (Claros and Vincens, 1996). (C) Live cell imaging of U2OS cells transfected with MacroD1-GFP (green) or Δ77MacroD1-GFP (green) for 48 h; mitochondria and nuclei were labeled with MitoTracker Deep Red FM (red) and Hoechst 33258 (blue), respectively. (D) Isolated mitochondrial fractions from HeLa cells were resuspended in isolation buffer or hypotonic buffer (2 mM HEPES) before treatment with proteinase K (PK) and/or Triton X-100 (as indicated). Immunoblot analysis was performed to determine PK accessibility to MacroD1 and sub-mitochondrial protein markers: AIF [apoptosis-inducing factor 1, mitochondrial; intermembrane space (IMS) marker, HSP60 (60 kDa heat shock protein, mitochondrial; mitochondrial matrix (MM) marker], UQCRC2 [cytochrome b-c1 complex subunit 2, mitochondrial; inner mitochondrial membrane (IMM) marker] and TOM20 [mitochondrial import receptor subunit; outer mitochondrial membrane (OMM)] marker. W, whole cell extract; D, cell debris; P, post mitochondrial supernatant; M, isolated mitochondrial fraction.

FIGURE 4

MacroD1 and Streptomyces coelicolor protein SCO6450 efficiently remove mono-ADP-ribose adducts from ADP-ribosylated phosphorylated double stranded DNA ends. (A) De-ADP-ribosylation of ADP-ribosylated thymidine base on single stranded DNA by DarG (lane 4), MacroD1 WT (lane 5), SCO6450 (lane 6), and MacroD1 mutant (MacroD1 G270E – lane 7) macrodomain-containing proteins. Thymidine base on single stranded DNA was ADP-ribosylated by DarT (lanes 2–8). In both (B,C) phosphorylated (lanes 3–9) double stranded DNA was mono-ADP-ribosylated by PARP1 E998Q. Removal of mono-ADP-ribosylated 5′ (B) and 3′ (C) phosphorylated double stranded DNA ends by MacroD1 WT (lane 6), SCO6450 (lane 7), and MacroD1 mutant (MacroD1 G270E – lane 8) macrodomain containing protein; ddH₂O (lane 5) and BSA (lane 9) were used as negative controls. PARP1 E998Q was unable to ADP-ribosylate non-phosphorylated double stranded DNA (noP) (lane 2). Mono-ADP-ribosylated phosphorylated double-stranded DNA samples were treated with (lane 3) or without PK (lane 4) to confirm DNA and not protein was the substrate of PARP1 E998Q modification. ³²P radiolabelled 21mer DNA was used as a size marker (M, lane 1 in all experiments). PARPi, PARP inhibitor; BSA, bovine serum albumin.