The goal of this study was to examine the transmural distribution of ryanodine receptors in left ventricular (LV) hypertrophy (LVH) and its in vivo consequences. Dogs were chronically instrumented with an LV pressure gauge, ultrasonic crystals for measurement of LV internal diameter and wall thickness, and a left circumflex coronary blood flow velocity transducer. Severe LVH was induced by chronic banding of the aorta (12+/-1 months), which resulted in a 78% increase in LV/body weight. When ryanodine was infused directly into the circumflex coronary artery, it did not affect LV global function or systemic hemodynamics; however, it reduced LV wall thickening and delayed relaxation in the posterior wall in control dogs but was relatively ineffective in dogs with LVH. In LV sarcolemmal preparations, [3H]ryanodine ligand binding revealed a subendocardial/subepicardial gradient in normal dogs. In LVH there was a 45% decrease in ryanodine receptor binding and a loss in the natural subendocardial/subepicardial gradient, which roughly correlated inversely with the extent of LVH and directly with regional wall motion. Both mRNA and Western analyses revealed similar findings, with a reduction of the transmural mRNA levels and a loss in the natural gradient between subendocardial and subepicardial layers in LVH. Thus, ryanodine receptor message and binding in LVH is reduced preferentially in the subendocardium with consequent attenuation of the action of ryanodine in vivo. The selectively altered ryanodine regulation subendocardially in LVH could reconcile some of the controversy in this field and may play a role in mediating decompensation from stable LVH.