Limitation of myocardial injury and infarction has been demonstrated by interventions such as ischemic preconditioning or the use of pyruvate as a substrate, which reduces glycogen content before, and acidosis during, ischemia. An isolated perfused rat heart model of global ischemia was employed to test the hypothesis that glycogen depletion reduces ischemic injury as measured by creatine kinase release. 31P-nuclear magnetic resonance spectroscopy was used to measure high-energy phosphates (ATP and phosphocreatine), phosphomonoesters (PME), and intracellular pH. Compared with control glucose-perfused hearts with normal glycogen content (1.49 +/- 0.13 mg Glc/g wet wt), glycogen-depleted pyruvate, ischemic preconditioned, and glycogen-depleted glucose hearts all had reduced glycogen content before ischemia (0.62 +/- 0.16, 0.81 +/- 0.10, and 0.67 +/- 0.12 mg Glc/g wet wt, respectively; P = 0.003) and significantly higher pH at the end of ischemia (5.85 +/- 0.02, 6.33 +/- 0.06, 6.24 +/- 0.04, and 6.12 +/- 0.02 in control, glycogen-depleted pyruvate, preconditioned, and glycogen-depleted glucose-perfused hearts, respectively; P < 0.01), although acidification during the initial phase of ischemia was differentially affected by the three interventions. Glycogen-depleted pyruvate and preconditioned hearts had reduced PME accumulation, greater recovery of function and phosphocreatine, and lower creatine kinase release on reperfusion, whereas glycogen-depleted glucose-perfused hearts were similar to control hearts. In summary, glycogen depletion by these three methods limits the fall in pH during global ischemia, although glycogen depletion in the absence of preconditioning does not limit ischemic injury.(ABSTRACT TRUNCATED AT 250 WORDS)