The structure, redox chemistry and motor neuron toxicity of heterodimeric zinc-deficient SOD1-Implications for the toxic gain of function observed in ALS

Abstract

A subset of familial cases of amyotrophic lateral sclerosis (fALS) are caused by mutations to copper, zinc superoxide dismutase (Cu, Zn SOD1). There are over 200 mutations to SOD1 that have been associated with fALS and the majority of these mutations are dominantly inherited. Thus, individuals are heterozygous and express both wild-type SOD1 and the mutant form of the protein. Paradoxically, when rodent models are produced that mimic the co-expression of wild-type SOD1 with mutant fALS SOD1 the motor neuron disease accelerates. Previously, we have shown that the loss of zinc from an SOD1 kills cultured motor neurons due to a gained, redox activity catalyzed by the active-site copper. Furthermore, motor neuron toxicity of zinc-deficient SOD1 is enhanced by wild-type Cu, Zn SOD1. Because SOD1 exists as a non-covalent dimer, the enhanced toxicity might result from stabilization of the heterodimeric interface between zinc-deficient SOD1 and Cu, Zn-SOD1. However, experimentation with the heterodimer is difficult because SOD1 subunits exchange in minutes. To better characterize the role of dimer stabilization on the enhanced toxicity of fALS mutant SOD1 by wild type SOD1, we genetically tethered a zinc-deficient SOD1 subunit with a Cu, Zn SOD1 subunit with a 16-residue linker. The x-ray structure of the tethered heterodimer shows that zinc-deficient subunit adopts a wild-type-like conformation and is not misfolded. The heterodimer intermediate also produced peroxynitrite from nitric oxide, and the tethered SOD1 was strikingly toxic to primary cultures of motor neurons. This work supports the concept that zinc-deficient SOD1 is a likely toxic intermediate in ALS. Furthermore, the wild-type allele in human familial-SOD1 ALS patients may physically contribute to the dominant inheritance of SOD1 mutations through heterodimer formation.

Keywords: Superoxide dismutase; amyotrophic lateral sclerosis; protein engineering.

Figure 1.

Tethered SOD1 sequence. Cloning, expression, and crystallization of C111S-D83S/C111S SOD1 heterodimer (Het-SOD1). (A) The SOD1 heterodimer is a single polypeptide chain of ~32.7 kDa comprising tandem SOD1 protein monomers connected by a 16-residue flexible linker sequence (bold underlined). The darker shade with bold and underline indicate the location of mutations. The N-terminal monomer is wild type-like with a C111S mutation for better expression and solubility and the second constitutively zinc-deficient D83S/C111S C-terminal SOD1. (B) Size exclusion inductively coupled plasma mass spectrometry quantitation of Cu and Zn content. (C) Photograph of the crystals used for structural determination.

Figure 3.

Motor neuron survival assays. Motor neuron survival was determined after administration of vehicle only (black bar vehicle, n=13 each condition), C111S-SOD1 (grey bar WT, n=12 TF, n=8 TFD), zinc deficient D83S/C111S SOD1 (green, n=12 TF, n=8 TFD), a 50/50 mixture of zinc-deficient D83S/C111S-SOD1 and holo SOD1(C111S) (grey diagonal lines, Cu, Zn SOD1, n=13 TF, n=8 TFD) and the Het-SOD1 with the glycine serine linker (red bar D83S/C111S-C111S SOD1 heterodimer, n=12 TF, n=8 TFD). Cultures treated with the trophic factors BDNF, GDNF and cardiotrophin-1 (TF) and separately after trophic factor deprivation (TFD). Values are mean with standard deviation. Statistical analysis with two-way ANOVA with Tukey’s multiple comparison post-hoc test (*p-value 0.02, ***p-value <0.001).

Figure 4.

Peroxynitrite production due to re-oxidation of SOD. Preparations of Zinc-deficient WT SOD including wild type Cu, E SOD (green squares) and the D83S D83S+WT heterodimer (red triangles) are readily reduced via ascorbate to produced superoxide. In the presence of NO, superoxide reacted to form peroxynitrite, which was detected via oxidation of coumarin boronate assay. WT Cu,Zn SOD (blue circles) did not oxidize coumarin boronate any faster than the spontaneous hydrolysis compared to buffer control (black squares).

Figure 5.

Structure of heterodimer D83S-C111S SOD1 (yellow, 6DTK), wild type SOD1 (1PU0) (cyan) and D83S mutant (2R27) (purple) are overlaid using one monomeric subunit A (left) to show perfect superposition of the second subunit B heterodimer and wild type and significant twist of the mutant monomer. A). Front view and B) Top view. The approximate two-fold axis is shown by the arrow, Zn is in blue and Cu^II is in orange. C) Representative coordination environment of Zn and Cu in heterodimer D83S/C111S-C111S SOD1, and D) overlaid with the electron density map shown as a grey mesh contoured at 1.5σ within 1.6 Å of ligands.

Figure 6.

XANES spectra for (A) Zn K-edges of D83S-C111S SOD1 heterodimer (HD) (blue), reduced Het-SOD1 (red), and WT SOD1 reduced (green), and for (B) Cu K-edges of D83S+WT SOD1 heterodimer (HD) (blue), reduced HD SOD1 (red), WT SOD1 (green), and reduced WT SOD1 (purple).

Figure 7.

A). Overlay of the tethered heterodimer SOD1 (orange) with WT SOD1 (grey). Only the metals bound to the tethered SOD1 are shown (Cu in orange and Zn in grey sphere) (Het-SOD1 PDB:6Dtk, WT SOD1 PDB:1PU0). B.) shows the electrostatic and zinc-binding loops disordered in zinc-deficient homodimer (blue, PDB:2R27) compared to the right zinc-deficient subunit in 7A the tethered heterodimer (16). C.) Overlay of tethered Het-SOD1 (orange) with the homodimeric zinc-deficient SOD1 illustrating the greater twist in the dimer interface in the zinc deficient-homodimer (blue).