NAD+ Precursors Reverse Experimental Diabetic Neuropathy in Mice

Abstract

Abnormal NAD⁺ signaling has been implicated in axonal degeneration in diabetic peripheral neuropathy (DPN). We hypothesized that supplementing NAD⁺ precursors could alleviate DPN symptoms through increasing the NAD⁺ levels and activating the sirtuin-1 (SIRT1) protein. To test this, we exposed cultured Dorsal Root Ganglion neurons (DRGs) to Nicotinamide Riboside (NR) or Nicotinamide Mononucleotide (NMN), which increased the levels of NAD⁺, the SIRT1 protein, and the deacetylation activity that is associated with increased neurite growth. A SIRT1 inhibitor blocked the neurite growth induced via NR or NMN. We then induced neuropathy in C57BL6 mice with streptozotocin (STZ) or a high fat diet (HFD) and administered NR or NMN for two months. Both the STZ and HFD mice developed neuropathy, which was reversed through the NR or NMN administration: sensory function improved, nerve conduction velocities normalized, and intraepidermal nerve fibers were restored. The NAD⁺ levels and SIRT1 activity were reduced in the DRGs from diabetic mice but were preserved with the NR or NMN treatment. We also tested the effect of NR or NMN administration in mice that overexpress the SIRT1 protein in neurons (nSIRT1 OE) and found no additional benefit from the addition of the drug. These findings suggest that supplementing with NAD+ precursors or activating SIRT1 may be a promising treatment for DPN.

Keywords: NAD+; NEDD4-1; SIRT1; diabetic neuropathy; high fat diet; mitochondria; streptozotocin.

Figure 1

Addition of NMN or NR or SIRT1OE promoted neurite growth in cultured DRG neurons in a SIRT1-dependent manner. (A) Primary DRGs were cultured from 3-month-old nSIRT1OE mice and wild-type (WT) mice for 4 days and the length of neurite outgrowth was measured. Neurite outgrowth was defined as the longest neurite extension in μm per neuron, and the average value was calculated for each culture. Protein extracts were prepared, and the levels of SIRT1 protein were measured with a Western blot analysis and the SIRT1 activity was measured via fluorometric assay [3]. (B) Cultured immortalized rat DRGs (50B11) were differentiated with forskolin (75 µM) in the absence or in the presence of added NR (150 μg/mL) or NMN (100 μg/mL). 50B11 cells were transfected with CMV-SIRT1 over expression (OE) vector. Duplicate cultures were also treated with SIRT1 inhibitor EX527 (5 μM). Neurite length was measured 72 h after the addition of NR/NMN/EX527 or after transfection with SIRT1OE vector. The addition of NMN or NR increased neurite length compared to vehicle control (p < 0.01), and transfection with SIRT1OE vector also increased neurite length (p < 0.001). Addition of EX527 inhibited forskolin-induced neurite growth. (p < 0.001). *** = p < 0.001 in subfigure A; Circle represent individual experimental value in subfigure B.

Figure 2

NR reverses STZ-induced neuropathy in C57BL6 mice (n = 6/group). Three-month-old C57BL6 WT and Streptozotocin- (STZ-) induced diabetic mice were purchased from Jackson Labs. Baseline nerve conduction studies (NCSs) were completed at the beginning of the study. Sciatic motor nerve conduction velocity (SMNCV; (A)), tail sensory nerve conduction velocity (TSNCV; (B)), tail motor latency (TML; (C)), and mechanical allodynia (MA; (D)) were measured in non-diabetic (control) and in STZ mice at 1 and 2 months. After confirming that the STZ mice had developed neuropathy as observed by the changes in the NCSs and MA, the STZ mice were fed with NR in the diet at a dose of 300 mg/kg for an additional 2 months. NCSs were performed at 3 and 4 months, namely after 1 and 2 months of the administration of NR. The results are shown as follows: sciatic SMNCV (A), TSNCV (B), TML (C), and MA (D) via the Von Frey filament paw withdrawal threshold method. Statistical comparisons were made among the three groups with the ANOVA and post hoc Tukey test. *** p < 0.001; STZ at 2 months compared to 0-month-old STZ and non-diabetic mice in all parameters. ⁺⁺⁺ p < 0.001, STZ + NR at 4 months compared to STZ at 2 months in all parameters. Dietary administration of NR reversed all the deficits of STZ-induced neuropathy. Administration of NR to non-diabetic mice had no significant effect [1].

Figure 3

NR reverses HFD-induced neuropathy in C57BL6 mice (n = 6/group). Three-month-old, male, C57BL6 WT mice were fed with either a control diet (CD) or a high fat diet (HFD) for 2 months. Baseline NCSs were completed at the beginning of the study. SMNCV, TML, TSNCV, and MA were measured in the mice fed a CD and the mice fed an HFD at 1 and 2 months. After confirming that consumption of the HFD for 2 months induced development of peripheral neuropathy as observed by the changes in the NCSs and MA, NR was added to the HFD mice at a dose of 300 mg/kg for an additional 2 months. Nerve conduction studies were performed after 1 and 2 months of the administration of the NR. SMNCV (A), TML (B), TSNCV (C), and mechanical allodynia via the Von Frey filament paw withdrawal threshold method (D). Statistical comparisons were made among the three groups with the ANOVA and post hoc Tukey test. *** p < 0.001; HFD mice at 2 months compared to 0-month-old HFD and CD mice in all parameters. ⁺⁺⁺ p < 0.001, STZ + NR mice at 4 months compared to STZ at 2 months in all parameters. The dietary administration of NR reversed all the deficits of STZ-induced neuropathy. The administration of NR to non-diabetic mice had no significant effect [1].

Figure 4

The NAD⁺ and SIRT1 activity levels were decreased in the DRGs from the STZ diabetic animals. NMN treatment repairs the NAD⁺ deficit, increasing SIRT1 activity. (A) Perchloric acid extraction of the frozen DRG tissue with added internal standard of NAD⁺ was completed as described [1]. NAD⁺ analysis in tissue samples was carried out with QTRAP^® 5500 mass spectrometer (AB Sciex, Framingham, MA, USA) equipped with a turbo–electrospray interface in positive ionization mode. The results showed a significant decrease in NAD⁺ levels in diabetic DRG neuronal tissue, and the treatment of diabetic mice with NMN or NR increased the NAD⁺ levels to tissue levels in non-diabetic DRGs. (B) SIRT1 activity was measured in the DRG tissue extracts following the SIRT1 Activity Assay Kit (Fluorometric; ab 156065). The results show a significant decrease in SIRT1 activity in diabetic DRG neuronal tissue, and the treatment of diabetic mice with NMN or NR increased the SIRT1 activity to tissue levels in non-diabetic DRGs. *** = p < 0.001; Triangles, circles and squares represent individual sample values.

Figure 5

NMN administration does not enhance reversal of HFD neuropathy in nSIRT1OE mice (n = 6/group). The nSIRT1OE mice were generated as previously described [3]. The expression of nSIRT1 was shut off by feeding the bigenic mouse with doxycycline (DOX, 200 mg/kg diet). Three-month-old, DOX-fed, male, nSIRT1OE-OFF mice were fed with either a control diet (CD; Group #1) or a high fat diet (HFD; Group #2) for 4 months. Baseline NCSs were performed at the beginning of the study. The SMNCV, TML, TSNCV, and MA were measured in the CD mice and the HFD mice at 2 and 4 months. After confirming the induction of peripheral neuropathy in the HFD, nSIRTOE-OFF mice at 2 months, the nSIRT1OE was tuned on by removing the DOX from the diet. The nSIRT1OE mice were divided into two groups: Group #3 was fed with a DOX-free HFD for an additional 2 months. Group #4 was fed with a DOX-free HFD, and NMN was administered on alternate days at a dose of 100 mg/kg ip for an additional 2 months. NCSs were performed after 2 weeks and after 2 months of the administration of NMN. The model for the experiment (A), IENFD (B), SMNCV (C), mechanical allodynia through the Von Frey filament paw withdrawal threshold method (D). There was no significant difference with NMN administration in the nSIRT1OE mice compared to nSIRT1OE alone. *** = p < 0.001; the # represents the groups that were compared.