Intermolecular transmission of superoxide dismutase 1 misfolding in living cells

Abstract

Human wild-type superoxide dismutase-1 (wtSOD1) is known to coaggregate with mutant SOD1 in familial amyotrophic lateral sclerosis (FALS), in double transgenic models of FALS, and in cell culture systems, but the structural determinants of this process are unclear. Here we molecularly dissect the effects of intracellular and cell-free obligately misfolded SOD1 mutant proteins on natively structured wild-type SOD1. Expression of the enzymatically inactive, natural familial ALS SOD1 mutations G127X and G85R in human mesenchymal and neural cell lines induces misfolding of wild-type natively structured SOD1, as indicated by: acquisition of immunoreactivity with SOD1 misfolding-specific monoclonal antibodies; markedly enhanced protease sensitivity suggestive of structural loosening; and nonnative disulfide-linked oligomer and multimer formation. Expression of G127X and G85R in mouse cell lines did not induce misfolding of murine wtSOD1, and a species restriction element for human wtSOD1 conversion was mapped to a region of sequence divergence in loop II and β-strand 3 of the SOD1 β-barrel (residues 24-36), then further refined surprisingly to a single tryptophan residue at codon 32 (W32) in human SOD1. Time course experiments enabled by W32 restriction revealed that G127X and misfolded wtSOD1 can induce misfolding of cell-endogenous wtSOD1. Finally, aggregated recombinant G127X is capable of inducing misfolding and protease sensitivity of recombinant human wtSOD1 in a cell-free system containing reducing and chelating agents; cell-free wtSOD1 conversion was also restricted by W32. These observations demonstrate that misfolded SOD1 can induce misfolding of natively structured wtSOD1 in a physiological intracellular milieu, consistent with a direct protein-protein interaction.

Fig. 1.

Sequence overview of (A) wtSOD1 and (B) G127X SOD1. Residue G85 (mutated in some FALS cases) and the relative locations of the DSE, GX-CT, and WT-CT epitopes are noted, with the corresponding antibodies used in this study. The pan-SOD1 antibody used in this study was a rabbit polyconal antibody prepared by immunization with the whole wtSOD1 polypeptide. (C–F) Expression of G127X mutant misfolded SOD1 induces misfolding in wtSOD1. (C and D) IF images of G127X-transfected HEK 293FT (HEK) cells probed with (C) GX-CT polyclonal IgG, specific for the nonnative C terminus of G127X SOD1, and (D) the DSE2-specific mAb 3H1, which recognizes only misfolded full-length SOD1. (E) Merge of C and D with the nuclear-specific stain DAPI; Inset: untransfected cells stained with GX-CT, 3H1, and DAPI. Cells positive for 3H1 immunoreactivity also coexpress G127X. (F) Empty vector (EV) control stained with GX-CT, 3H1, and DAPI; Inset: EV control stained with DAPI and pan-SOD1 antibody.

Fig. 2.

Association with and conformational conversion of misfolded mutant SOD1 and misfolded wtSOD1. (A) IP of lysates from transiently transfected HEK cells. DSE mAbs 3H1 and 10C12 precipitate wtSOD1 in lysates from cells transfected with G127X or G85R but not with EV control. The quantitation summary is shown below; error bars represent SE. Values are the average of a minimum of five independent IP experiments. *Statistically significant difference compared with EV control. (B) IP of PK-treated lysates from G127X- and G85R-transfected HEK cells, demonstrating marked protease sensitivity exhibited by the mutant SOD1 variants and misfolded wtSOD1 immunoprecipitated by the DSE mAbs. A minimum of five replicates were performed for all immunoprecipitation experiments. (C) Protease sensitivity comparison of wtSOD1 and G127X mutant SOD1. Note that G127X is only detectable in the absence of PK. All immunoblots were probed with pan-SOD1 pAb.

Fig. 3.

Mouse SOD1 is not a substrate for conformational conversion by human misfolded SOD1, and a single missense mutation W32S in FALS-linked SOD1 mutants prevents conversion of wtSOD1 to a misfolded form. (A and B) IF images of G127X transiently transfected HEK (A) and mouse N2a (B) cells probed with GX-CT and 3H1 (the DSE2-specific mAb). HEK cells expressing G127X display 3H1 immunoreactvity for misfolded wtSOD1, but no 3H1 immunoreactivity is observed in N2a cells. (C and D) IP of lysates from HEK cells transfected with G127X/W32S (C) or G85R/W32S (D) double mutants. SOD1-DSE mAbs immunoprecipitate SOD1 from lysates endogenously expressing original FALS-linked SOD1 single mutants but not from those expressing the W32S mutation.

Fig. 4.

Misfolded wtSOD1 is sustained after induction by misfolding-competent SOD1 molecular species. Lysates from cells collected at time points after transfection were examined by immunoprecipitation with DSE mAbs 3H1 and 10C12, and the relative amount of immunoprecipitated misfolded SOD1 calculated and represented in graphical form over time. (A) G127X vs. (B) G127X/W32S; G127X/W32S is incompetent to induce wtSOD1 misfolding, but G127X-induced wtSOD1 misfolding persists at least 4 d after G127X is no longer detectable, consistent with wtSOD1 misfold propagation. The relative abundance of the expressed transgene (in G127X in A and G127X/W32S in B) is shown by GX-CT IP. (C) Overexpressed hWT SOD1 vs. overexpressed W32S SOD1. No antibody probes exist to distinguish whether misfolded SOD1 derives from transfection or endogenous SOD1. There is a significant difference at day 10 of DSE immunoreactivity for wtSOD1 vs. SOD1/W32S transfectants (P < 0.0001, Mann-Whitney test). (D) Representative immunoblots (probed with pan-SOD1 pAb) from HEK cell lysates collected 2 d and 10 d after transfection with EV, G127X, G127X/W32S, hWT SOD1, and W32S SOD1.

Fig. 5.

Recombinant G127X can induce wtSOD1 misfolding in a cell-free system under reducing, metal cation-buffered conditions. (A) Representative Biacore sensorgram showing misfolded SOD1 binding to 3H1 mAb after 24 h incubation at 37 °C in the presence and absence of recombinant misfolded G127X protein in Hepes-buffered saline (HBSN) containing 50 mM DTT and 5 mM EDTA. Analyte is applied to the sensor surface during the association phase and allowed to elute during the dissociation phase. (B) Same experiment as in A performed in Hepes buffered saline without DTT or EDTA. (C) Box plot showing difference in 3H1 binding between 24-h incubated samples in the presence and absence of G127X for both buffer conditions. Binding levels are normalized to account for variation in immobilization levels of 3H1 between replicates. In the buffer containing DTT and EDTA there is a significant increase in 3H1 binding when G127X is added (99% confidence interval 22–103 response units, RU), whereas in the buffer without DTT or EDTA there is no significant difference due to the addition of G127X (95% confidence interval −26 to 17 RU). P values reported are from a paired t test of independent replicates. (D) Treatment of misfolded wtSOD1 generated by coincubation with G127X with PK (1 μg/mL for 30 min) causes its complete degradation. (E) Coincubation with G127X/W32S causes significantly less induction of wtSOD1 misfolding compared with coincubation with G127X (P = 0.02). Error bars show mean and SD of binding at the end of the association phase for three independent experimental replicates.