Effect of a cyclooxygenase-2 inhibitor on postexercise muscle protein synthesis in humans

Abstract

Nonselective blockade of the cyclooxygenase (COX) enzymes in skeletal muscle eliminates the normal increase in muscle protein synthesis following resistance exercise. The current study tested the hypothesis that this COX-mediated increase in postexercise muscle protein synthesis is regulated specifically by the COX-2 isoform. Sixteen males (23 +/- 1 yr) were randomly assigned to one of two groups that received three doses of either a selective COX-2 inhibitor (celecoxib; 200 mg/dose, 600 mg total) or a placebo in double-blind fashion during the 24 h following a single bout of knee extensor resistance exercise. At rest and 24 h postexercise, skeletal muscle protein fractional synthesis rate (FSR) was measured using a primed constant infusion of [(2)H(5)]phenylalanine coupled with muscle biopsies of the vastus lateralis, and measurements were made of mRNA and protein expression of COX-1 and COX-2. Mixed muscle protein FSR in response to exercise (P < 0.05) was not suppressed by the COX-2 inhibitor (0.056 +/- 0.004 to 0.108 +/- 0.014%/h) compared with placebo (0.074 +/- 0.004 to 0.091 +/- 0.005%/h), nor was there any difference (P > 0.05) between the placebo and COX-2 inhibitor postexercise when controlling for resting FSR. The COX-2 inhibitor did not influence COX-1 mRNA, COX-1 protein, or COX-2 protein levels, whereas it did increase (P < 0.05) COX-2 mRNA (3.0 +/- 0.9-fold) compared with placebo (1.3 +/- 0.3-fold). It appears that the elimination of the postexercise muscle protein synthesis response by nonselective COX inhibitors is not solely due to COX-2 isoform blockade. Furthermore, the current data suggest that the COX-1 enzyme is likely the main isoform responsible for the COX-mediated increase in muscle protein synthesis following resistance exercise in humans.

Fig. 1.

Timeline of the experimental protocol. Meal, standardized meal; MB, muscle biopsy; pill, placebo or cyclooxygenase (COX)-2 inhibitor (celecoxib) consumption; RE, resistance exercise bout. *Venous blood sample. There was ∼1 wk between infusion protocols.

Fig. 2.

Mixed muscle protein fractional synthesis rate (FSR) in the placebo and COX-2 inhibitor groups at rest and 24 h following the resistance exercise bout. *P < 0.05, main effect for exercise. Analysis of covariance revealed no significant difference (P > 0.05) between the placebo and COX-2 inhibitor groups postexercise when controlling for resting FSR.

Fig. 3.

Fold changes in mRNA levels of COX variant 1 (COX-1v1) and variant 2 (COX-1v2) and COX-2 in the placebo and COX-2 inhibitor groups from rest to 24 h following the resistance exercise bout. mRNA levels were normalized to GAPDH using the 2^−ΔΔC_T relative quantification method. *P < 0.05 from placebo.

Fig. 4.

COX-1 and COX-2 protein levels in the placebo and COX-2 inhibitor groups at rest (open bars) and 24 h following the resistance exercise bout (shaded bars). COX-2 protein levels were undetectable at rest or 24 h postexercise in either group (see materials and methods for further description). Inserts in COX-1 and COX-2 graphs are representative Western blots with positive controls. The additional insert at left in the COX-2 graph is the human recombinant positive control run separately and loaded per the manufacturer‘s directions, with the lower 2 bands likely representing at least 2 versions of the glycosylated COX-2 protein (39). The amount of human recombinant positive control loaded in the representative COX-2 image at right is 0.3% of the manufacturer’s directions.