Oxidation resistance 1 regulates post-translational modifications of peroxiredoxin 2 in the cerebellum

Abstract

Protein aggregation, oxidative and nitrosative stress are etiological factors common to all major neurodegenerative disorders. Therefore, identifying proteins that function at the crossroads of these essential pathways may provide novel targets for therapy. Oxidation resistance 1 (Oxr1) is a protein proven to be neuroprotective against oxidative stress, although the molecular mechanisms involved remain unclear. Here, we demonstrate that Oxr1 interacts with the multifunctional protein, peroxiredoxin 2 (Prdx2), a potent antioxidant enzyme highly expressed in the brain that can also act as a molecular chaperone. Using a combination of in vitro assays and two animal models, we discovered that expression levels of Oxr1 regulate the degree of oligomerization of Prdx2 and also its post-translational modifications (PTMs), specifically suggesting that Oxr1 acts as a functional switch between the antioxidant and chaperone functions of Prdx2. Furthermore, we showed in the Oxr1 knockout mouse that Prdx2 is aberrantly modified by overoxidation and S-nitrosylation in the cerebellum at the presymptomatic stage; this in-turn affected the oligomerization of Prdx2, potentially impeding its normal functions and contributing to the specific cerebellar neurodegeneration in this mouse model.

Keywords: Antioxidant; Chaperone; Mouse; Neurodegeneration; Oxidative stress; Peroxiredoxin.

Fig. 1

Oxr1 plays a role in the regulation of the oligomerization status of Prdx2 in vitro and in vivo. (A-B) Representative western blot of Prdx2 from protein extracts of P18 wild-type control (Oxr1^+/+) and Oxr1^d/d brain tissue with the cerebellum removed (A) and the cerebellum alone (B). Quantification of the HMW Prdx2 complexes (###) and the Prdx2 dimer (#) is shown (N = 8 animals per group (A) and N = 3 (B)). (C) Western blot of Prdx2 from protein extracts of adult wild-type and Oxr1 Tg mouse brains without the cerebellum. Quantification of the HMW Prdx2 complexes and the dimer is shown (N = 3 animals per group). (D) Western blot of Prdx2 from protein extracts of N2A cells transfected with either an empty vector or Oxr1-FL and treated with 300 μM H₂O₂ for 1 h. Quantification of the HMW Prdx2 complexes, the Prdx2 dimer (#) and the decamer (##) is shown (N = 3 independent repeats). For all panels, extracts were run on 10% BN-PAGE (native) gels and levels of Prdx2 and β-actin (input) were determined for the same samples, but run in reducing conditions on separate gels. β-actin was used as a loading control. Data are represented as mean ± SEM. Panels A-D: t-test; *p < 0.05, ***p < 0.001.

Fig. 2

Prdx2 binds to TLDc proteins, but does not interact functionally with Oxr1 to regulate H₂O₂-induced cell death. (A) N2A cells were co-transfected with Prdx2-MYC and HA-tagged TLDc constructs as indicated and data from co-immunoprecipitations (IP) of the TLDc proteins using an anti-MYC antibody are shown by immunoblotting for HA (left panel). α-tubulin was used as a loading control for the input samples (right panel). The asterisk (*) represents non-specific IgG heavy chains. (B) Co-immunoprecipitation of Oxr1-FL with Prdx2 from adult wild-type mouse brain tissue (excluding the cerebellum). (C) N2A cells transfected with the indicated constructs were treated with 500 μM H₂O₂ for 5 h and cell death was quantified as the number of pyknotic nuclei. 10 fields of view per condition were quantified (N = 100–150 cells per field of view). (D) Changes in H₂O₂ concentration were monitored using a colorimetric ferrous oxidation-xylenol orange (FOX) assay over 25 min with either recombinant purified Prdx2, Oxr1-C, OXR1-FL or Prdx2 with Oxr1-C proteins with 50 μM H₂O₂. (N = 3 independent repeats). Data are represented as mean ± SEM. Panel C: 1-way ANOVA; ***p < 0.001 compared to empty vectors, ###p < 0.001 compared to transfection with Prdx2.

Fig. 3

Prdx2 and Oxr1 possess holdase activity. (A-C) DTT-induced insulin aggregation assay using recombinant Prdx2 and Oxr1-C with lysozyme as a negative control. Absorbance readings at 650 nm were taken every 3 min (A). Quantification of the aggregation rate (B) and absorbance at endpoint (C) are shown. (D-F) Thermally-induced citrate synthase (CS) aggregation assay using recombinant Prdx2 and Oxr1-C. Absorbance readings at 360 nm were taken every 3 min (D). Aggregation rate (E) and absorbance at end point (F) are shown. (G-I) DTT-induced insulin aggregation assay using two concentrations (6 μM and 9 μM) of recombinant Oxr1-C or human OXR1-FL (6 μM), in parallel with human TBC1D24 and the TLDc domain of TBC1D24 (TBC1D24 TLDc). Absorbance readings at 650 nm were taken every 3 min (G). Quantification of the aggregation rate (H) and absorbance at endpoint (I) are presented. (A-I) All values are shown as normalized to the background signal. Data are represented as mean ± SEM. B, C, H, I, E, F: t-test, *p < 0.05, **p < 0.01, ***p < 0.001 as compared to insulin (B, C, H, I) or CS (E, F), ##p < 0.01, ###p < 0.001 as compared to insulin + Prdx2 (B, C, H, I) or CS + Prdx2 (E, F), ^§§§p < 0.001 as compared to insulin + Oxr1-C 6 μM (H, I).

Fig. 4

Oxr1 modulates overoxidation of Prdx2 catalytic residues in vitro and in vivo. (A-B) N2A cells were transfected with either Oxr1-FL (A) or Oxr1-C (B) and treated with 300 μM H₂O₂ for 1 h. Representative image of western blots and quantification of Prdx2 overoxidation is shown as the ratio of Prdx2-SO_2/3 to total Prdx2 (N = 6 independent repeats for Oxr1-FL and N = 3 independent repeats for Oxr1-C). (C-F) Representative westerns blots of protein extracts from an adult Oxr1 overexpressing mouse (Oxr1 Tg) (C-D) and a P18 Oxr1 knockout (Oxr1^d/d) mouse (E-F) from both the brain lacking the cerebellum and cerebellar tissues, with quantification of Prdx2 overoxidation compared to wild-type controls (N = 3–4 animals per group). β-actin was used as a loading control. Data presented as mean ± SEM. Panels A-F: t-test; *p < 0.05.

Fig. 5

Oxr1 regulates S-nitrosylation of Prdx2 in vitro and in vivo in the cerebellum. (A) SH-SY5Y cells were treated for 30 min with 200 μM SNOC or vehicle (veh) and processed for the biotin switch assay to quantify the levels of S-nitrosylated endogenous PRDX2 (SNO-PRDX2) by western blot in the presence of the constructs as indicated. (N = 3 independent repeats). (B-C) Representative western blots of SNO-Prdx2 levels in brain (B) or cerebellum (C) as determined by biotin switch from mice over-expressing Oxr1 (Oxr1 Tg) compared to wild-type controls (Oxr1^+/+) (N = 3 animals per group). (D-E) SNO-Prdx2 levels in brain (D-F) or cerebellum (E-G) determined by biotin switch (D-E) or organomercury resin capture (ORC, F-G) from Oxr1 knockout (Oxr1^d/d) mice compared to wild-type controls (N = 3 animals per group). In all panels, quantification is shown as the ratio of SNO-Prdx2 to total Prdx2 with α-tubulin used as a loading control. Data presented as mean ± SEM. Negative controls (-ve Ctrl) were combined proteins diluted in dimethylformamide without labelling/reducing reagent (for biotin switch) or proteins incubated with 10 mM DTT for 15 min (for ORC). Panel A: 1-way ANOVA (comparison of constructs to empty vector in either vehicle or SNOC-treated condition) and t-test (comparison of each construct between vehicle and SNOC-treated conditions), Panels B-G: t-test; *p < 0.05, **p < 0.01, ***p < 0.001 as compared to empty vector-transfected cells or to control mice, ^#p < 0.05, ^##p < 0.01 as compared to vehicle treated cells.