CuII (atsm) inhibits ferroptosis: Implications for treatment of neurodegenerative disease

Abstract

Background and purpose: Diacetyl-bis(4-methyl-3-thiosemicarbazonato)copper^II (Cu^II (atsm)) ameliorates neurodegeneration and delays disease progression in mouse models of amyotrophic lateral sclerosis (ALS) and Parkinson's disease (PD), yet the mechanism of action remains uncertain. Promising results were recently reported for separate Phase 1 studies in ALS patients and PD patients. Affected tissue in these disorders shares features of elevated Fe, low glutathione and increased lipid peroxidation consistent with ferroptosis, a novel form of regulated cell death. We therefore evaluated the ability of Cu^II (atsm) to inhibit ferroptosis.

Experimental approach: Ferroptosis was induced in neuronal cell models by inhibition of glutathione peroxidase-4 activity with RSL3 or by blocking cystine uptake with erastin. Cell viability and lipid peroxidation were assessed and the efficacy of Cu^II (atsm) was compared to the known antiferroptotic compound liproxstatin-1.

Key results: Cu^II (atsm) protected against lipid peroxidation and ferroptotic lethality in primary and immortalised neuronal cell models (EC₅₀ : ≈130 nM, within an order of magnitude of liproxstatin-1). Ni^II (atsm) also prevented ferroptosis with similar potency, whereas ionic Cu^II did not. In cell-free systems, Cu^II (atsm) and Ni^II (atsm) inhibited Fe^II -induced lipid peroxidation, consistent with these compounds quenching lipid radicals.

Conclusions and implications: The antiferroptotic activity of Cu^II (atsm) could therefore be the disease-modifying mechanism being tested in ALS and PD trials. With potency in vitro approaching that of liproxstatin-1, Cu^II (atsm) possesses favourable properties such as oral bioavailability and entry into the brain that make it an attractive investigational product for clinical trials of ferroptosis-related diseases.

Figure 1

The chemical structures of M^II(atsm) and M^II(atsm₂) (where M = either Cu^II or Ni^II), Cu^II(gtsm), ferrostatin‐1, and liproxstatin‐1

Figure 2

Cu^II(atsm) rescues ferroptosis and lipid peroxidation in cultured neurons. (a) Viability of cells treated with RSL3 (1 μM) ± liproxstatin‐1 or Cu^II(atsm) for 24 hr. Data are means ± SEM, n = 6 independent experiments. EC₅₀ was determined by non‐linear regression, with 95% CI shown in parenthesis. (b) Lipid peroxidation measured by C11‐BODIPY(581/591) in cells treated with RSL3 (0.2 μM) ± liproxstatin‐1 or Cu^II(atsm) (1 μM) for 24 hr. Data are means ± SEM, n = 6 independent experiments. P values were calculated using the Kruskal–Wallis test corrected for multiple comparisons by controlling the false discovery rate with the Benjamini, Krieger, and Yekutieli test. *P < .05 compared to vehicle treated cells. (c–e) Viability of cells treated with RSL3 (0.1 μM, c), erastin (1 μM, d) or Fe^II (1 mM, e) ± liproxstatin‐1 or Cu^II(atsm) for 24 hr. Data are means ± SEM, n = 5 independent experiments. EC₅₀ was determined by non‐linear regression, with 95% CI shown in parenthesis. (f) Lipid peroxidation measured by C11‐BODIPY(581/591) in cells treated with RSL3 (50 nM), erastin (0.5 μM), or Fe^II (0.5 mM) ± liproxstatin‐1 or Cu^II(atsm) (1 μM) for 24 hr. Data are means ± SEM, n = 5 independent experiments. P values were calculated using the Kruskal–Wallis test corrected for multiple comparisons by controlling the false discovery rate with the Benjamini, Krieger, and Yekutieli test. *P < .05 compared to vehicle treated cells

Figure 3

Cu does not prevent ferroptosis in N27 cells. (a–c) Viability of cells treated with RSL3 (0.1 μM) or erastin (1 μM) ± CuSO₄ (a), Cu^II (gtsm) (b), or clioquinol (c) for 24 hr. Data are means ± SEM, n = 5 independent experiments. (a, b) P values were calculated using the Kruskal–Wallis test corrected for multiple comparisons by controlling the false discovery rate with the Benjamini, Krieger, and Yekutieli test. (c) P values were calculated using the Mann–Whitney test. *P < .05 compared to vehicle

Figure 4

Ni^II(atsm) prevents ferroptosis in N27 cells and lipid peroxidation induced by Fe^II. (a, b) Viability in cells treated with RSL3 (0.1 μM) for 24 hr. (a) Cells treated with RSL3 ± Cu^II(atsm), Ni^II(atsm) or H₂(atsm) (1 μM) for 24 hr in the presence or absence of the Cu chelator BCS (100 μM). Data are means ± SEM, n = 6 independent experiments. P values were calculated using the Kruskal–Wallis test corrected for multiple comparisons by controlling the false discovery rate with the Benjamini, Krieger, and Yekutieli test. *P < .05 compared to vehicle with RSL3, ^P < .05 compared to vehicle with RSL3 and BCS. (b) Cells treated with RSL3 ± Cu^II(atsm) or Ni^II(atsm) for 24 hr. Data are means ± SEM, n = 5 independent experiments. EC₅₀ was determined by non‐linear regression, with 95% CI shown in parenthesis. (c, d) Lipid peroxidation measured by C11‐BODIPY(581/591) in Locke's buffer with arachidonic acid (50 μM) treated with Fe^II (2 μM) for 30 min. (c) Arachidonic acid treated with Fe^II ± liproxstatin‐1, Cu^II(atsm), Ni^II(atsm) or H₂(atsm) (1 μM) in the presence or absence of the Cu chelator BCS (12.5 μM). Data are means ± SEM, n = 6 independent experiments. P values were calculated using the Kruskal–Wallis test corrected for multiple comparisons by controlling the false discovery rate with the Benjamini, Krieger and Yekutieli test. *P < .05 compared to vehicle with Fe, ^P < .05 compared to vehicle with Fe^II and BCS. (d) Arachidonic acid treated with Fe^II ± liproxstatin‐1, Cu^II(atsm), Ni^II(atsm), or H₂(atsm). Data are means ± SEM, n = 5 independent experiments

Figure 5

Cu^II(atsm₂) is impaired in preventing ferroptosis in N27 cells or lipid peroxidation induced by Fe^II. (a, b) Viability in cells treated with RSL3 (a, 0.1 μM) or erastin (b, 1 μM) ± Cu^II(atsm) or Cu^II(atsm₂) for 24 hr. Data are means ± SEM, n = 5 independent experiments. EC₅₀ was determined by non‐linear regression, with 95% CI shown in parenthesis. Cu^II(atsm₂) EC₅₀ could not be determined as viability did not exceed 50%. (c) Lipid peroxidation measured by C11‐BODIPY(581/591) in Locke's buffer using arachidonic acid (50 μM) treated with Fe^II (2 μM) ± Cu^II(atsm) or Cu^II(atsm₂) for 30 min. Data are means ± SEM, n = 5 independent experiments

Figure 6

Cu^II(atsm) does not react with Fe^II or Fe^III to suppress lipid peroxidation. Arachidonic acid (10 μM) was treated with either Fe^II (10 μM, a, c) or Fe^III (10 μM, b, d) ± ceruloplasmin (200 μg·ml⁻¹), ascorbate (1 mM), liproxstatin‐1 (10 μM) or Cu^II(atsm) (10 μM) in Locke's buffer for 30 min. (a, b) Lipid peroxidation was measured by C11‐BODIPY(581/591) every 90 sec. Data are means ± SEM, n = 5 independent experiments. Non‐linear regression was conducted. (c, d) Fe^II measured by Ferene‐S after 1 min and 30 min. P values were calculated using the Kruskal–Wallis test corrected for multiple comparisons by controlling the false discovery rate with the Benjamini, Krieger,and Yekutieli test. *P < .05, compared to vehicle treated cells. (e) Redox cycling reactions of Fe species that can lead to the propagation of lipid peroxides and radicals