AMP-activated protein kinase regulates nicotinamide phosphoribosyl transferase expression in skeletal muscle

Abstract

Deacetylases such as sirtuins (SIRTs) convert NAD to nicotinamide (NAM). Nicotinamide phosphoribosyl transferase (Nampt) is the rate-limiting enzyme in the NAD salvage pathway responsible for converting NAM to NAD to maintain cellular redox state. Activation of AMP-activated protein kinase (AMPK) increases SIRT activity by elevating NAD levels. As NAM directly inhibits SIRTs, increased Nampt activation or expression could be a metabolic stress response. Evidence suggests that AMPK regulates Nampt mRNA content, but whether repeated AMPK activation is necessary for increasing Nampt protein levels is unknown. To this end, we assessed whether exercise training- or 5-amino-1-β-D-ribofuranosyl-imidazole-4-carboxamide (AICAR)-mediated increases in skeletal muscle Nampt abundance are AMPK dependent. One-legged knee-extensor exercise training in humans increased Nampt protein by 16% (P < 0.05) in the trained, but not the untrained leg. Moreover, increases in Nampt mRNA following acute exercise or AICAR treatment (P < 0.05 for both) were maintained in mouse skeletal muscle lacking a functional AMPK α2 subunit. Nampt protein was reduced in skeletal muscle of sedentary AMPK α2 kinase dead (KD), but 6.5 weeks of endurance exercise training increased skeletal muscle Nampt protein to a similar extent in both wild-type (WT) (24%) and AMPK α2 KD (18%) mice. In contrast, 4 weeks of daily AICAR treatment increased Nampt protein in skeletal muscle in WT mice (27%), but this effect did not occur in AMPK α2 KD mice. In conclusion, functional α2-containing AMPK heterotrimers are required for elevation of skeletal muscle Nampt protein, but not mRNA induction. These findings suggest AMPK plays a post-translational role in the regulation of skeletal muscle Nampt protein abundance, and further indicate that the regulation of cellular energy charge and nutrient sensing is mechanistically related.

Figure 2. Three weeks of one-legged knee extensor exercise training in humans increases Nampt protein in trained, but not untrained, skeletal muscle

Human volunteers performed 15 sessions of one-legged knee extensor endurance training over the course of 3 weeks. Skeletal muscle biopsies were obtained from both vastus lateralis muscles before and after training (n= 8). * Indicates treatment × time interaction effect (P < 0.05).

Figure 3. Nampt protein levels are related to AMPK activity in mouse skeletal muscle

Mouse skeletal muscle Nampt protein was measured in tibialis anterior muscle of mouse models with reduced AMPK activity, such as (A) LKB1 KO (n= 9–11), (B) AMPK α2i (n= 6) or (C) AMPK γ1 transgenic mice, which show chronically elevated AMPK activity in skeletal muscle (n= 5–6). # Indicates vs. WT (P < 0.05).

Figure 4. Acute exercise increases mouse skeletal muscle Nampt mRNA in an AMPK α2-independent manner

Nampt mRNA was measured in AMPK α2 WT and KO mouse muscle 3 h after completion of a 90 min treadmill exercise bout (n= 6–13). * Indicates vs. Pre (P < 0.05); †† indicates vs. 0, 1 h (P < 0.01).

Figure 5. Combined wheel-cage and treadmill training increases Nampt protein in mouse skeletal muscle in an AMPK α2- and PGC-1α-independent manner

Quadriceps muscles of sedentary or trained (6.5 weeks of combined voluntary wheel-cage and forced exercise training) WT and AMPK α2 KD mice (n= 12–16) were removed the morning following the final exercise bout, and (A) Nampt protein, (B) hexokinase II protein and (C) Nampt mRNA levels were measured. D, Nampt protein abundance was measured in WT and PGC-1α KO mice that underwent 5 weeks of combined voluntary wheel-cage and forced endurance training, or served as sedentary controls (n= 16). * Indicates vs. control (P < 0.05); ** indicates vs. control (P < 0.01); # indicates vs. WT (P < 0.05).

Figure 6. Acute AICAR treatment increases Nampt mRNA independent of AMPK α2

A, Nampt mRNA was measured in C57BL/6J mouse quadriceps muscle 2, 4 and 8 h after AICAR injection (500 mg kg⁻¹ body weight; n= 6). B, Nampt mRNA concentrations and C) Nampt protein abundance were assessed 8 h after AICAR treatment (500 mg kg⁻¹ body weight; n= 10–13). * Indicates vs. saline (P < 0.05); † indicates vs. 2 and 4 h (P < 0.05); # indicates vs. WT (P < 0.05).

Figure 7. Repeated AICAR administration increases skeletal muscle Nampt in an AMPK-dependent but PGC1α-independent manner

A, Nampt protein; B, hexokinase II protein and C, Nampt mRNA levels were measured in quadriceps of WT or AMPK α2 KD animals following 4 weeks of daily treatment with AICAR (500 mg kg⁻¹ body weight) or saline (n= 7–8). D, Nampt protein levels were measured in both WT and PGC-1α KO mice following 4 weeks of daily treatment with AICAR or saline (n= 8). * Indicates vs. saline (P < 0.05); # indicates vs. WT (P < 0.05); ** indicates vs. saline (P < 0.01).

Figure 8. Effect of repeated metformin treatment on skeletal muscle Nampt concentrations

Nampt concentrations were measured in white and red gastrocnemius muscle of WT and AMPK α2 KD that were treated with 2 weeks of oral metformin treatment (300 mg kg⁻¹ body weight) or saline. # Indicates vs. WT (P < 0.05); ** indicates vs. red gastrocnemius (P < 0.01); n= 10–12. Metformin treatment increased Nampt protein nearly significantly in white gastrocnemius (two-way ANOVA; main metformin treatment effect, P= 0.06; observed power = 0.39).

Figure 1. Nampt antibody specificity and validity

Nampt knockdown reduces (A) Nampt mRNA levels and (B) Nampt protein expression in C2C12 myoblasts. C, three identical C2C12 lysate aliquots from cells overexpressing FLAG-tagged Nampt were resolved alongside a mouse liver sample (right-hand lane). After protein transfer to polyvinylidene difluoride, the membrane was cut in two pieces as indicated by the dashed line, incubated with antibodies specific to FLAG (left side) and Nampt (right side), and re-assembled before visualisation. D, antibody specificity in human and mouse tissues. In separate gels, mouse muscle, human muscle and mouse liver lysates were resolved alongside a mouse liver sample (right-hand lane). After protein transfer to polyvinylidene difluoride, the membrane was then cut in two pieces as indicated by the dashed line, incubated with human Nampt (left side) or mouse Nampt antibody (right side), and re-assembled before visualisation. sh1, sh2, short hairpin clone 1, 2.