Iron-catalyzed oxidative stress compromises cancer promotional effect of BRCA2 haploinsufficiency through mitochondria-targeted ferroptosis

Abstract

Pathogenic variants in BRCA2 are hereditary risks for various cancers, including breast, ovary, pancreas and prostate. Genomic instability due to insufficient homologous recombination is thought as responsible for carcinogenesis. Reportedly, endogenous or exogenous aldehydes, including formaldehyde and acetaldehyde, suppress BRCA2 function. However, molecular sequences how BRCA2 insufficiency leads to carcinogenesis remains unelucidated. To assess whether Fenton reaction-based oxidative stress is a promotional risk factor of carcinogenesis in BRCA2 haploinsufficiency, we here applied iron-induced renal carcinogenesis to a newly established rat heterozygous mutation model of Brca2 (mutant, T1942Kfs/+; MUT). Rat MUT model, despite significant increase in spontaneous malignant tumors, showed no promotional effect on renal carcinogenesis induced by ferric nitrilotriacetate (Fe-NTA) in contrast to our previous study using Brca1 mutant rats. Array-based comparative genome hybridization of renal cell carcinoma in MUT revealed significant increase in the frequency of homozygous Cdkn2A deletion. Whereas acute-phase analysis of the kidney after single or 1-week Fe-NTA administration to MUT showed suppressed lipid peroxidation, consistent with ferroptosis-resistance, ferroptosis and regeneration of tubular cells were coexistent with higher cytoplasmic catalytic Fe(II) levels in the subacute phase of MUT after 3-week Fe-NTA administration. Mechanistically, mitochondrial dysfunction with excess iron, promoted by insufficient BRCA2 presumably for maintaining DNA integrity, eventually initiated ferroptotic process. In conclusion, iron-dependent oxidative stress plays double-edged roles either for cell death or proliferation in carcinogenesis and its biological consequences are distinct between BRCA2 and BRCA1 haploinsufficiency. Our results suggest that iron-catalyzed oxidative stress is not a major driving force of carcinogenesis in BRCA2 pathogenic variants.

Keywords: BRCA2; Fe-NTA; Ferroptosis; Genome instability; Mitochondria.

Graphical abstract

Fig. 1

Despite high malignant tumor incidence, BRCA2-deficiency does not promote Fenton reaction-based carcinogenesis but favors Cdkn2a deletion (A) Immunoblot analysis of BRCA2 expression of kidney samples after Fe-NTA single injection (W/WT, wild-type; M/MUT, mutant; N/NT, non-treated; n = 3 or 4; means ± SEM; ∗, P < 0.05). (B) Representative immunohistochemical images of γH2AX protein in kidney 3 h after Fe-NTA single injection (bar = 100 μm; n = 4; means ± SEM; ∗, P < 0.05). (C) Probability of survival curve under Fe-NTA-induced renal carcinogenesis experiment in the wild-type (WT; n = 20) and Brca2 heterozygous knockout (Mutant [MUT]; n = 17) groups. (D) A case of Fe-NTA-induced RCC in Brca2 MUT rat (left panel, macroscopic appearance; right panel, histology of the RCC; bar = 100 μm). Li, liver; Ca, carcinoma; arrow heads, peritoneal invasion. (E) Summary of aCGH analysis of early-onset RCC in WT and Brca2 MUT (n = 4, respectively). Individual data on all the 16 RCCs is included in Fig. S1C and Table S3. (F) Representative immunohistochemical images of p16^INK4A protein in RCCs (bar = 50 μm). See text for details.

Fig. 2

BRCA2-deficiency grants ferroptosis-resistance in the acute phase of Fe-NTA-induced renal carcinogenesis (A) Top enriched gene sets in WT compared to Brca2 MUT by GSEA on kidney at 1 week in the Fe-NTA-induced renal carcinogenesis protocol using gene set of “WikiPathways” (n = 3). (B) Enrichment plots of the gene sets associated with ferroptosis. (C) Representative hematoxylin & eosin staining of kidney 3 h after Fe-NTA single injection (bar = 100 μm). Tubular injury area (area enclosed by dotted line) was evaluated (n = 17; means ± SEM; ∗, P < 0.05). (D) Representative immunohistochemical images of HNEJ-1 monoclonal antibody in kidney 3 h after Fe-NTA single injection (bar = 100 μm; n = 4; means ± SEM; ∗∗∗, P < 0.001). (E) Immunoblot analysis of ferroptosis-related proteins in kidney after Fe-NTA single injection (n = 3 or 4; means ± SEM; ∗, P < 0.05; ∗∗, P < 0.01; ns, not significant). Note that the same immunoblot membrane was used as in Fig. S2B. (F) Cell viability by WST assay on WT or Brca2 MUT rat embryonic fibroblast treated for 24 h with FAC or RSL3 in the presence of Ferrostatin-1 (1 μM). Unpaired two-tailed t-test or two-way ANOVA analysis; ∗∗∗, P < 0.001; ∗∗∗∗, P < 0.0001; ns, not significant.

Fig. 3

BRCA2-deficiency promotes lipid peroxidation leading to ferroptosis via mitochondrial damage in the subacute phase of Fe-NTA-induced renal carcinogenesis (A) Immunoblot analysis of phospho-BRCA2(Ser2095) and BRCA2 in kidney at 3 weeks of Fe-NTA injection (n = 3; means ± SEM; ∗, P < 0.05). (B) Representative immunohistochemical images of γH2AX protein in kidney 3 weeks of Fe-NTA injection (bar = 100 μm; n = 3; means ± SEM; ∗, P < 0.05). (C) Representative images of TUNEL staining in kidney at 3 weeks of Fe-NTA injection (bar = 50 μm; n = 3 or 4; means ± SEM; ∗, P < 0.05). (D) Representative immunohistochemical images of HNEJ-1 monoclonal antibody in kidney at 3 weeks of Fe-NTA injection (bar = 100 μm; n = 3; means ± SEM; ∗, P < 0.05). (E) Representative electron microscopic images of mitochondria in kidney (yellow arrows, small-sized deformed mitochondria in the MUT in the untreated control; blue arrows, shrunken mitochondria with immature cristae in the MUT at 3 weeks in the Fe-NTA-induced renal carcinogenesis protocol; ∗, lysosome; ∗∗, autophagosome; bar = 2.0 μm in the left panels; 1.0 μm in the right panels). (F) Analysis of the number of lysosome and autophagosome per 20 μm². (n = 18–21; means ± SEM; ∗, P < 0.05; ∗∗∗, P < 0.001). (G) Analysis of mitochondrial area, number of cristae and mitochondrial roundness (n = 24–31; means ± SEM; ∗, P < 0.05; ∗∗∗∗, P < 0.0001; ns., not significant). (H) Representative immunohistochemical images of GPX4 protein in kidney at 3 weeks of Fe-NTA injection (bar = 100 μm; n = 3 or 4; means ± SEM; ∗, P < 0.05).

Fig. 4

BRCA2-deficiency increases catalytic ferrous iron and transport iron in mitochondria (A) FerroOrange staining of the frozen sections of kidney (bar = 50 μm; n = 3; means ± SEM; ∗, P < 0.05). (B) Immunoblot analysis of iron-related proteins in kidney at 3 weeks of Fe-NTA injection (n = 3; means ± SEM; ∗, P < 0.05; ∗∗, P < 0.01; ns, not significant). (C) Immunoblot analysis of MFN1 and SFXN3 in kidney at 3 weeks of Fe-NTA injection (n = 3; means ± SEM; ∗, P < 0.05; ∗∗, P < 0.01; ∗∗∗, P < 0.001; ns, not significant). (D) Mito-FerroGreen staining of the frozen sections of kidney (bar = 50 μm; n = 3; means ± SEM; ∗, P < 0.05; ns, not significant).

Fig. 5

BRCA2-deficiency causes cell-cycle acceleration in the subacute phase of Fe-NTA-induced renal carcinogenesis (A) Enrichment plots of the gene sets associated with cell-cycle. (B) Representative images of Ki-67 protein in kidney at 3 weeks of Fe-NTA injection (bar = 100 μm; n = 3 or 4; means ± SEM; ∗, P < 0.05). (C) Immunoblot analysis of cell-cycle-related proteins in kidney at 3 weeks of Fe-NTA injection (n = 3; means ± SEM; ∗, P < 0.05; ∗∗, P < 0.01, ∗∗∗, P < 0.001). (D) Immunoblot analysis of p21 in the cytoplasmic or nuclear fraction in kidney at 3 weeks of Fe-NTA injection. Cytoplasmic p21 is normalized by β-actin and nuclear p21 is normalized by LaminB1 (n = 3 or 4; means ± SEM; ∗, P < 0.05). (E) Immunoblot analysis of phospho-AKT (Ser473) and AKT in kidney at 3 weeks of Fe-NTA injection (n = 3; means ± SEM; ∗, P < 0.05).

Fig. 6

Summary of the present study Graphical abstract of how BRCA2-deficiency acts on Fe-NTA induced renal carcinogenesis.