Research Article

A nuclease that mediates cell death induced by DNA damage and poly(ADP-ribose) polymerase-1

See allHide authors and affiliations

Science  07 Oct 2016:
Vol. 354, Issue 6308, aad6872
DOI: 10.1126/science.aad6872

You are currently viewing the figures only.

View Full Text

  1. Stressors lead to DNA damage, PARP-1 activation, and PAR formation. PAR facilitates the release of AIF from mitochondria where it binds MIF. This complex translocates to the nucleus to bind DNA; the result is DNA fragmentation and cell death. Interference with this cascade by preventing the formation of the AIF-MIF complex or by a nuclease-deficient MIF prevents DNA fragmentation and promotes cell survival.

  2. Fig. 1 Identification of MIF as a key cell-death effector mediating PARP-1–dependent cell death.

    (A) Strategy for identifying AIF-associated proteins involved in PARP-1–dependent cell death. (B) siRNA-based PARP-1–dependent cell viability high-throughput screening in HeLa cells 24 hours after MNNG treatment (50 μM, 15 min); n = 8. The experiments were repeated in four independent tests ***P < 0.001, one-way ANOVA. (C) Schematic representation of MIF’s PD-D/E(X)K domains. Single-letter abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; Y, Tyr; and X, any amino acid. (D) Alignment of the nuclease domain of human MIF and other nucleases. Arrows above the sequence indicate β-strands and rectangles represent α helices. Amino acid residues that were mutated are indicated with an arrow and number (see Results). Nuclease and CxxCxxHx(n)C domains are highlighted in green and pink, respectively. (E) Crystal structure of MIF trimer (pdb:1GD0) (left) and MIF PD-D/E(x)K motif in trimer (right).

  3. Fig. 2 MIF is a nuclease that cleaves genomic DNA.

    (A) In vitro MIF (2 μM) nuclease assay with pcDNA as substrate. (B) In vitro pulsed-field gel electrophoresis MIF (4 μM) nuclease assay with human genomic DNA (hgDNA) as a substrate in buffer containing Mg2+ (10 mM) with or without EDTA (50 mM) or Ca2+ (2 mM) with or without EDTA (25 mM). (C) Pulsed-field gel electrophoresis assay of MNNG-induced DNA damage in MIF-deficient HeLa cells and wild-type (WT) HeLa cells treated with or without DPQ (30 μM) or ISO-1 (100 μM). NT/MNNG, (nontargeting shRNA/MNNG). (D) Nuclease assay of MIF WT and mutants (4 μM) using human genomic DNA as the substrate.

  4. Fig. 3 MIF binds and cleaves ssDNA.

    (A) MIF DNA binding motif determined by ChIP-seq. (B) Binding of MIF to biotin-labeled small DNA substrates with different structures or different sequences in an EMSA assay (see fig. S8 for illustrations of substrates, and tables S1 and S2 for sequences). Arrow indicates the DNA-MIF complex. Asterisk indicates nonspecific bands. PCS, positive control substrate from the LightShift Chemiluminescent EMSA Kit (Thermo Scientific) containing 60 base pairs (bp) of 5′ biotin-labeled duplex. With or without BSA, bovine serum albumin; PC, Epstein-Barr nuclear antigen extract from the LightShift Chemiluminescent EMSA Kit or MIF. (C) MIF cleavage of unpaired bases at the 3′ end of the stem loop of 5′ or 3′ biotin-labeled small DNA substrates with various structures or sequences in a nuclease assay (see fig. S8 for illustrations of substrates, and tables S1 and S2 for sequences). Experiments were replicated four times with three independent preparations of MIF protein. (D) MIF cleavage of 3′ unpaired bases from nonlabeled PS30 and 3F1 substrates. DNA ladders 1 and 2 were customized with PS30 and its cleavage products by removing its 3′ nucleotides one by one. DNA ladder 1 was prepared using PS30, PS28, PS26, PS24, PS22, and PS20. DNA ladder 2 was prepared using PS29, PS27, PS25, PS23, and PS21. (E) MIF cleavage sites on nonlabeled PS30 and 3F1 substrates.

  5. Fig. 4 Requirement of AIF for the recruitment of MIF to the nucleus in NMDA excitotoxicity.

    (A) Binding of immobilized GST-MIF WT and GST-MIF variants to AIF. (B) Nuclease activity and AIF-binding activity of MIF WT and MIF variants. (C and D) Coimmunoprecipitation (IP) of MIF and AIF in control (CSS) and NMDA-treated cortical neurons. Asterisk indicates IgG. Ab, antibody. (D) Intensity of treated versus untreated cultures. *P < 0.05, Student’s t test. (E) Images of nuclear translocation of AIF and MIF after NMDA treatment in WT, AIF knockdown, and MIF knockdown cortical neurons. AIF shRNA (AIF sh) and MIF shRNA (MIF sh) caused a 71.3 ± 5.2% and 73.3 ± 6.1% protein reduction, respectively. White color indicates the overlay of AIF, MIF, and 4′,6′-diamidino-2-phenylindole (DAPI), showing the nuclear translocation of AIF and MIF. Purple color indicates the overlay of AIF and DAPI, showing the nuclear translocation of AIF. Z stacks illustrating the x,z and y,z axis are provided to demarcate the nucleus. Arrowheads, indicate cells with the high magnification. (F) Quantification of the percentage of cells with nuclear translocation of MIF and AIF after NMDA treatment in WT, AIF knockdown, and MIF knockdown cortical neurons. CSS, control salt solution. ****P < 0.0001, versus its CSS control; ####P < 0.0001, versus its WT treated with NMDA, one-way ANOVA. (G) Immunoblots of nuclear translocation of AIF and MIF after NMDA treatment in WT, AIF knockdown, and MIF knockdown cortical neurons. Compare total protein (T), postnuclear fraction (PN), nuclear fraction (N), and Mito, mitochondrial antibody. (H and I) Relative levels of AIF and MIF in T, PN, and N. Means ± SEM. Experiments were replicated at least three times; ****P < 0.0001, versus its CSS control; ##P < 0.01, ###P < 0.0001, versus its total protein, one-way ANOVA.

  6. Fig. 5 MIF E22A mutant prevents AIF’s recruitment of MIF to the nucleus in NMDA-excitotoxicity.

    (A) Expression of MIF in WT and knockout (KO) neurons. (B) Coimmunoprecipitation of Flag-tagged MIF variants and AIF in cortical neurons after NMDA treatment. (C) Images of nuclear translocation of AIF and exogenous MIF WT and MIF variants after NMDA treatment in MIF KO cortical neurons. Scale bar, 20 μm. White color indicates the overlay of AIF, MIF, and DAPI, showing the nuclear translocation of AIF and MIF. Purple color indicates the overlay of AIF and DAPI, showing the nuclear translocation of AIF. Z stacks illustrating the x,z and y,z axis are provided to demarcate the nucleus. (D) Quantification of the percentage of cells with nuclear translocation of AIF and exogenous MIF WT and MIF variants after NMDA treatment in MIF KO cortical neurons. ****P < 0.0001, versus KO group; ###P < 0.001, versus KO-WT group, one-way ANOVA. (E) Immunoblots of nuclear translocation of AIF and exogenous MIF WT and MIF variants after NMDA treatment in MIF KO cortical neurons. H4, histone H4; mito, mitochondrial antibody. (F and G) Relative levels of AIF and MIF in total protein (T), postnuclear fraction (PN) and nuclear fraction (N). Means ± SEM. Experiments were replicated at least three times. **P <0.01, ****P < 0.001, versus KO control group; ##P < 0.01, ####P < 0.0001, versus its total protein, one-way ANOVA.

  7. Fig. 6 MIF nuclease activity is critical for DNA damage and PARP-1–dependent cell death in cortical neurons.

    (A) Representative images and (B) quantification of NMDA-induced (500 μM for 5 min) excitotoxicity in MIF WT, KO, and KO cortical neurons expressing MIF WT, E22Q, E22A, or P2G. Scale bar, 200 μm. (C) Representative images and (D to F) quantification of NMDA-induced DNA damage 6 hours after treatment determined by the comet assay in MIF WT, KO, and KO neurons expressing MIF WT, E22Q, E22A, or P2G. Dashed lines indicate the center of the head and tail. Scale bar, 20 μm. (G) Pulsed-field gel electrophoresis assay of NMDA-induced DNA damage 6 hours after treatment in MIF WT and KO neurons and KO neurons expressing MIF WT, E22Q, E22A, or P2G. Means ± SEM are shown in (B), (D), (E), and (F). *P < 0.05, **P < 0.01, ***P < 0.001, one-way ANOVA; ns, nonsignificant.

  8. Fig. 7 MIF nuclease activity is critical for DNA damage and ischemic neuronal cell death in vivo.

    (A) Representative images of triphenyl tetrazolium chloride staining of MIF WT, KO, and KO mice that were injected with AAV2-MIF WT, E22Q, or E22A 24 hours after 45 min of middle cerebral artery occlusion (MCAO). (B to D) Quantification of infarction volume in cortex, striatum, and hemisphere 1 day or 7 days after 45-min MCAO. WT MCAO (n = 29); KO MCAO (n = 20); KO-WT MCAO (n = 23). KO-E22Q (n = 22) and KO-E22A MCAO (n = 19). *P < 0.05, ***P < 0.001, versus KO group at the same time point; ##P < 0.01, ###P < 0.001, the same group at 7 days versus at 1 day after 45-min MCAO, one-way ANOVA. (E to G) Neurological deficit was evaluated by [(E) and (F)] open field on a scale of 0 to 5 and (G) corner test evaluated by percentage of right turns at 1 day, 3 days, or 7 days after MCAO surgery. WT MCAO (n = 16); KO MCAO (n = 12); and KO-WT MCAO (n = 16). KO-E22Q MCAO (n = 16) and KO-E22A MCAO (n = 16). Means ± SEM. *P < 0.05, ***P < 0.001, one-way ANOVA in (E) and (G). **P < 0.01, two-way ANOVA in (F), WT and KO-WT versus KO, KO-E22Q, and KO-E22A at different time points. (H) DNA fragmentation determined by pulsed-field gel electrophoresis in the penumbra 1 day, 3 days, or 7 days after 45-min MCAO surgery in MIF WT, KO, and KO mutant mice, which were injected with AAV2-MIF WT, E22Q, or E22A. WT MCAO (n = 15); KO MCAO (n = 15); and KO-WT MCAO (n = 15). KO-E22Q (n = 15) and KO-E22A MCAO (n = 15). (I) Quantification of noncleaved genomic DNA. Means ± SEM. ****P < 0.0001, versus its sham treatment group, one-way ANOVA.