CD38 Dictates Age-Related NAD Decline and Mitochondrial Dysfunction through an SIRT3-Dependent Mechanism

Abstract

Nicotinamide adenine dinucleotide (NAD) levels decrease during aging and are involved in age-related metabolic decline. To date, the mechanism responsible for the age-related reduction in NAD has not been elucidated. Here we demonstrate that expression and activity of the NADase CD38 increase with aging and that CD38 is required for the age-related NAD decline and mitochondrial dysfunction via a pathway mediated at least in part by regulation of SIRT3 activity. We also identified CD38 as the main enzyme involved in the degradation of the NAD precursor nicotinamide mononucleotide (NMN) in vivo, indicating that CD38 has a key role in the modulation of NAD-replacement therapy for aging and metabolic diseases.

Keywords: CD38; NAD(+); aging; glucose intolerance; mitochondrial function.

Figure 1. CD38 expression increases with aging

(A) Immunoblots for CD38, PARP1, SIRT1, NAMPT and Tubulin in mice liver (left). On the right, graphs show relative protein expression of SIRT1, PARP1, CD38 and NAMPT. Relative expression of each protein was calculated as a ratio to Tubulin levels in each lane, and then calculated relative to 5 month old mice (n=12 for each age, *p<0.05, ^**p<0.01). In the experiments of Figure 1 A–H animals were obtain from the NIA aging colony. (B–D) Immunoblots for CD38, PARP1, SIRT1, NAMPT and Tubulin in mice adipose tissue (B), spleen (C), and skeletal muscle (D) (left). On the right, graphs show relative protein expression of SIRT1, PARP1, CD38 and NAMPT in the tissues. Relative expression of each protein was calculated as a ratio to Tubulin in each lane, and then calculated relative to 5 month old mice (n=4–6 for each age, *p<0.05, ^**p<0.01). (E) Relative mRNA levels of CD38, PARP1, PARP2, SIRT1, NAMPT, NAPRT1, NMNAT-1, NMNAT-2, NMNAT-3 and NADK in liver of young and old mice (3 and 32 months of age, respectively). (n=4 animals for each age, *p<0.05, versus 3 month old mice). (F–H) Relative mRNA levels of CD38, PARP1, PARP1, SIRT1 NAMPT, and NAPRT1 in adipose tissue (F), spleen (G), and skeletal muscle (H) (3 and 32 months of age, respectively) (n = 4 animals for each age, *p <0.05, versus 3 month old mice). (I) Relative mRNA levels of CD38, PARP1, and NAMPT in omental adipose tissue of human subjects with average ages of 34 (young) and 61 (old) (n = 9 for each age, *p <0.05, versus young group). All values are mean ± SEM.

Figure 2. CD38 regulates the age-related NAD⁺ decline

(A) CD38 activity in liver, adipose tissue, spleen and skeletal muscle of aging mice (n=6 animals for each age, ^**p<0.01 versus 5 month old mice). (B) CD38 activity in liver of 1 year old WT and CD38 KO litter mate mice (*p<0.05). (C) Total liver NAD⁺ and NADH levels in WT and CD38KO mice during aging measured by mass spectroscopy (n=4 mice for each age, *p<0.05, ^**p<0.01 versus 5 month old mice). (D) Total liver NAD⁺ and NADH levels in WT and CD38KO mice during aging measured by cycling assay (n=14 animals for each age, *p<0.05, ^**p<0.01 versus 5 month old mice). (E) NAD⁺ levels in adipose tissue, spleen, and skeletal muscle of WT and CD38KO mice during aging (n=4 animals for each age, *p<0.05, ^**p<0.01 versus 5 month old mice). All values are mean ± SEM.

Figure 3. CD38 deficiency upregulates mitochondrial function and increases NAD⁺/NADH ratio in mitochondria

Liver of WT and CD38KO 12 month old litter mate mice were used for the measurements below in B-H. All values are mean ± SEM. (A) Percentage of oxygen consumption coupled to ATP synthesis in isolated mitochondria during aging in WT and CD38KO mice (n=4 animals for each age, *p<0.05, ^**p<0.01 versus 5 month old mice). (B) Oxygen consumption rates in isolated mitochondria. The following drugs were added for the experimental profile: Succinate 10mM and rotenone 1µM (Succ+Rot), 0.15mM ADP, 1µg/mL oligomycin (Oligo), and 1µM FCCP (inset) (n=4, *p<0.05, ^**p<0.01 versus WT mice). (C) Relative membrane potential in mitochondria (*p<0.05 versus WT mice). (D–F) Total NAD⁺ levels (D), NADH levels (E), and the NAD⁺/NADH ratio (F) in isolated mitochondria (n=3, *p<0.05 versus WT mice). (G) Relative mtDNA quantification of COX I and ND4 as mitochondrial-encoded genes normalized by GAPDH. (H) Relative mRNA expression of glucose metabolism pathway enzymes (n=6, *p<0.05 versus WT mice).

Figure 4. Changes in CD38 expression regulate NAD⁺ levels and mitochondrial function in cells

293T cells expressing vector or Flag. CD38 were used in experiments from (A–L) and A549 cells were used in (M–O). All values are mean ± SEM. *p<0.05, ^**p<0.01, compared to control cells. (A) Immunoblotting for CD38 and Tubulin. (B) CD38 NADase activity (n=5). (C) and (D) Total NAD⁺ and NADH levels in whole cells (n=5). (E) Mitochondrial oxygen consumption in intact cells. Histogram represents the % oxygen consumption under FCCP-induced maximum respiration (n=6). (F) Lactate released in media (n=5). (G) Transmission electron microscopy. Scale bar: 200nm. OM: outer membrane, IM: internal membrane, MT: mitochondrial matrix, MC: mitochondrial cristae, ER: endoplasmic reticulum. (H) Relative mtDNA quantification (n=4). (I) Citrate synthase activity (n=9). (J) CD38 NADase activity (n=5) and immunoblotting for CD38 and Tubulin. (K) Total NAD⁺ levels (n=5). (L) Routine oxygen consumption (n=3).

Figure 5. CD38 controls the metabolic response in aged mice by activating SIRT3-dependent mechanisms

All values are mean ± SEM. (A) Mitochondrial protein acetylation profile (upper panel) and immunoblotting for SIRT3 in one year old WT and CD38KO litter mate mice (lower panel). Each lane represents one independent mouse. (B) SIRT3 activity in liver mitochondria isolated from one year old WT and CD38KO litter mate mice with endogenous contaminant NAD⁺ or addition of a saturating dose of 2mM NAD⁺ (n=4, *p<0.05, ns= non-significant p>0.05 versus WT control). (C) NAD⁺ levels in isolated liver mitochondria from two year old litter mate WT, CD38KO, and CD38/SIRT3KO mice (n=6, *p<0.05 versus WT mice). (D) Glucose concentration in (F) two year old litter mate WT, SIRT3KO, CD38KO, and CD38/SIRT3KO mice after intraperitoneal injection of glucose (left graphs). Area under the curve for glucose concentrations in different mice (right graphs) (n=8, ^**p<0.01, *p<0.05 versus WT mice). (E) Oxygen consumption rates coupled to ATP synthesis in liver mitochondria isolated from two year old litter mate WT, CD38KO, and CD38/SIRT3KO mice (n=6, *p<0.05, versus WT mice).

Figure 6. CD38 regulates the NR-induced metabolic improvement in high fat diet

All values are mean ± SEM. (A) Nicotinamide concentration in vitro after incubation of 1mM NR or 0.5mM NMN with CD38 recombinant protein. The metabolites were measured by HPLC (n=3, ^**p<0.01 versus NMN incubation). (B) Nicotinamide concentration in different tissue homogenates from WT and CD38KO mice after incubation with 1mM NMN. The metabolites were measured by a coupled assay (n=3,*p<0.05, ^**p<0.01 versus WT mice). (C) Nicotinamide concentration in blood from WT and CD38KO mice after incubation with 1mM NMN. The metabolites were measured by HPLC (n=3, *p<0.05, ^**p<0.01 versus WT mice). (D) Levels of NR, NMN, and NAD in blood of WT and CD38KO mice after injection of 500mg/kg NR intraperitoneally (i.p.). The metabolites were measured by HPLC (n=3, *p<0.05 versus WT mice). Due to the technique use it is possible that the NR peak in the figure could contain other metabolites. (E and F) Glucose concentration in WT and CD38KO mice after 16 weeks of high fat diet, treated with 500mg/Kg of Nicotinamide Riboside (NR) (dotted line) i.p. The controls received PBS injection (solid line) (E). Two-way ANOVA and Bonferroni's post-test with repeated measures show significant interaction between the glucose curve for CD38KO and CD38KO + NR (****p<0.0001). (F) Quantification of glucose AUC (**p<0.01, n=10).