RRM2B deficiency causes dATP and dGTP depletion through enhanced degradation and slower synthesis

Abstract

Mitochondrial DNA (mtDNA) replication requires a steady supply of deoxyribonucleotides (dNTPs), synthesized de novo by ribonucleotide reductase (RNR). In nondividing cells, RNR consists of RRM1 and RRM2B subunits. Mutations in RRM2B cause mtDNA depletion syndrome, linked to muscle weakness, neurological decline, and early mortality. The impact of RRM2B deficiency on dNTP pools in nondividing tissues remains unclear. Using a mouse knockout model, we demonstrate that RRM2B deficiency selectively depletes dATP and dGTP, while dCTP and dTTP levels remain stable or increase. This depletion pattern resembles the effects of hydroxyurea, an inhibitor that reduces overall RNR activity. Mechanistically, we propose that the depletion of dATP and dGTP arises from their preferred degradation by the dNTPase SAMHD1 and the lower production rate of dATP by RNR. Identifying dATP and dGTP depletion as a hallmark of RRM2B deficiency provides insights for developing nucleoside bypass therapies to alleviate the effects of RRM2B mutations.

Keywords: dNTP metabolism; genome stability; mtDNA stability; ribonucleotide reductase.

Fig. 1.

Characterization of phenotypic, morphological, and metabolic alterations in RRM2B Hemi and KO mice. (A) Kaplan–Meier survival curves of RRM2B Hemi and WT mice (n = 39 to 41). (B) Longitudinal weight charts of RRM2B Hemi and WT mice (n = 30). (C) Longitudinal weight charts of RRM2B KO mice from week 7 to terminal week 12 (n = 22 to 23). WT groups included six females and 16 males, while KO groups included 13 females and 10 males. (D) Gross images showing gastric distension and food accumulation in forestomach of RRM2B KO mice. (E) Distribution of free water, muscle mass, and fat deposits in WT and RRM2B KO mice measured by echo-MRI (n = 5) WT groups included one female and four males, while KO groups included three females and two males. (F) Histological sections of 12-wk-old WT and RRM2B KO kidneys and thigh muscle stained for morphology. (G) Histological sections of 2-y-old WT and RRM2B Hemi kidneys and thigh muscle stained for morphology. (H) COX, SDH, and NADH-TR staining in thigh muscles of 12-wk-old WT and RRM2B KO mice. (I) COX, SDH, and NADH-TR staining in thigh muscles of 2-y-old WT and RRM2B Hemi mice.

Fig. 2.

Effect of RRM2B deficiency on mtDNA levels and alkali sensitivity. (A) MtDNA copy number measured as CytB/Actin in 12-wk-old WT and RRM2B KO tissues (n = 6 to 10). (B) MtDNA copy number measured as CytB/Actin in 12-wk-old WT and RRM2B Hemi tissues (n = 6 to 10). (C) Representative agarose gel image showing mtDNA fragmentation following alkali treatment in WT and RRM2B KO tissues. Fragments in alkali-treated RRM2B KO brains and hearts are largely shorter than 500 nucleotides and therefore undetectable on the blot. The molecular weights of the ladder bands are indicated in kbp. (D) Quantification of mtDNA fragmentation from four mice per genotype, including the ones shown in C.

Fig. 3.

Analysis of dNTP pools in RRM2B KO and Hemi mice across developmental stages and mouse fibroblasts treated with HU. (A) One-year-old RRM2B Hemi mice compared to WT (n = 6 to 10). (B) RRM2B KO embryos at E13.5 compared to WT embryos (n = 6 to 9). (C) Seven-week-old RRM2B KO mice compared to WT (n = 5 to 7). (D) Twelve-week-old RRM2B KO mice compared to WT (n = 7 to 13). (E) Mouse BALB/c 3T3 fibroblasts (n = 2) treated with increasing concentrations of HU for 1 h. Data shown as mean ± SEM.

Fig. 4.

Analysis of dNTP pools in SAMHD1 KO mice and dNDP synthesis rates by RNR. (A) dNTP pools in postmitotic tissues of 15-wk-old SAMHD1 KO mice compared to WT controls (n = 6 to 10). (B) Enzymatic activity of recombinant RNR (composed of RRM1 and RRM2 subunits) assessed using a competition assay with all four ribonucleoside diphosphate (NDP) substrates (ADP, GDP, CDP, and UDP) at equal concentrations (200 μM). Reactions were performed in the presence of physiological concentrations of dNTPs (22 μM dCTP, 30 μM dTTP, 14 μM dATP, 5.6 μM dGTP) and NTPs (600 μM CTP, 1300 μM UTP, 2400 μM ATP, 500 μM GTP). Data shown as mean ± SEM; n = 3. (C) Schematic illustrating differential dNTP production in proliferating, postmitotic, and RRM2B-deficient cells. The balance of dNTP production and degradation varies across cell states, driven by the composition and activity of ribonucleotide reductase (RNR), SAMHD1, and deoxynucleoside (dN) kinases. dNs can be exported from the cell and subsequently reimported, where they undergo phosphorylation by deoxynucleoside kinases through the salvage pathway. Mitochondrial DNA (mtDNA) is depicted as circular DNA molecules, while nuclear DNA is represented as mitotic chromosomes. Proliferating cells: RNR is composed of RRM1 and RRM2, enabling high levels of de novo dNTP production to support both nuclear and mitochondrial DNA replication. SAMHD1 activity is low, ensuring sufficient dNTP levels for replication, while RRM2B is expressed at low levels and is not essential. Postmitotic cells: RRM2B replaces RRM2 as the primary RNR small subunit, producing lower but adequate dNTP pools required for continuous mtDNA replication and nuclear DNA repair. SAMHD1 activity is elevated, playing a significant role in degrading excess dNTPs to maintain homeostasis. Although RRM2 is downregulated in postmitotic cells, it is unclear whether it is completely absent or if it has some contribution to RNR activity. To reflect this uncertainty, the contribution of RRM2 to dNTP pools in postmitotic cells is depicted with a dashed arrow. RRM2B-deficient postmitotic cells: In the absence of functional RRM2B, residual RRM2 expression is insufficient to sustain adequate dNTP production. SAMHD1-dependent hydrolysis disproportionately depletes purine dNTPs, leading to inadequate dNTP pools for mtDNA replication. This deficiency results in mtDNA depletion and impaired mitochondrial function.