nisms with respect to modulation of basal hepatic gly-We perfused livers from fed rats with a balanced salt cogenolysis. In the presence of Ca 2/ , activation of solution containing 1 mmol/L glucose. Under these conphosphorylase with stimulation of glycogenolysis ditions a low steady rate of glycogenolysis was observed dominates. Cyclooxygenase inhibitors abolish this ef-(approximately 1.7 mmol glucose equivalents/g/min; 20% fect. Activation by NO of the cyclooxygenase in nonpaof the maximal glycogenolytic activity). Nitric oxide renchymal cells is a distinct possibility. In the absence (NO) transiently stimulated hepatic glucose production. of Ca 2/ , inhibition of basal glycogenolysis becomes ob-A maximal response (on average doubling basal glucose servable. It remains to be established whether this reoutput) was observed with 34 mmol/L NO. The same consults from cGMP-mediated stimulation of hydrolysis centration of nitrite (NO 2 0 ) was ineffective. Half-maxiof cAMP. (HEPATOLOGY 1996;23:1564-1571.) mal effects were seen at 8 to 10 mmol/L NO, irrespective of the flow direction (portocaval or retrograde). This glycogenolytic response to NO corresponded to a partial
Glycogenolysis is catalysed by glycogen phosphorylactivation of phosphorylase. The NO effect was not addiase, an enzyme that exists in two interconvertible tive to maximal stimulation of glycogenolysis (7.7 { 0.2 forms, the active phosphorylated a-form and the inac-
hexose equivalents/g/min; n Γ
4) by 100 mmol/L ditive dephosphorylated b-form. Hepatic glycogenolysis butyryl cyclic adenosine monophosphate (Bt 2 cAMP).
strictly depends on covalent activation of phosphoryl-
The requirement for activation of phosphorylase was also evidenced by the ineffectiveness of NO in phosphor-ase, 1 by a specific phosphorylase kinase. The kinase is ylase-kinase-deficient livers of gsd/gsd rats. The NO activated by cyclic adenosine monophosphate (cAMP)effect was blocked by co-administration of cyclooxydependent protein kinase, or by direct interaction of genase inhibitors (50 mmol/L ibuprofen, 50 mmol/L in-Ca 2/ with a calmodulin subunit. Ultimately, the catadomethacin, or 2 mmol/L aspirin), suggesting a medialytic expression of phosphorylase a is under substrate tory role of prostanoids from nonparenchymal cells. control, 2 because cytosolic levels of inorganic phosphate This conclusion was confirmed by the fact that NO did (P i ) fail to saturate substrate demands of activated not activate phosphorylase in isolated hepatocytes. phosphorylase a.
Moreover, NO was no longer glycogenolytic in livers
Control of hepatic glycogenolysis involves intricate perfused with Ca 2/ -free medium, in agreement with paracrine communication between nonparenchymal the known mediatory role of Ca 2/ in prostanoid-mediated responses. Surprisingly, in Ca 2/ -free medium NO cells and hepatocytes. Evidence has accumulated inhibited the basal glucose production. This coincided that prostanoids are engaged in the control of glycogenwith an increased elution of cyclic guanosine monoolysis in liver. Although endotoxin, 8 platelet-activating phosphate (cGMP). Inhibition of glycogenolysis by NO factor, 9 and phorbol esters 3,8 stimulate glycogenolysis under these conditions was blocked by 1 mmol/L thein intact liver, these compounds are ineffective when ophylline, suggestive for involvement of cGMP-stimutested directly on isolated hepatocytes. In the intact lated cAMP phosphodiesterase. However, we could not liver, cyclooxygenase inhibitors suppress stimulation confirm that an increase in cGMP resulted in a drop of prostanoid synthesis and activation of phosphorylase in cAMP. In conclusion, NO recruits opposing mechaby the former agents. Prostanoids may be directly involved in covalent activation of hepatocellular phosphorylase. Indirectly, eicosanoids may trigger vaso-Abbreviations: cAMP, cyclic adenosine monophosphate; Pi, inorganic phosconstriction, reduce O 2 extraction, increase P i availabilphate; cGMP, cyclic guanosine monophosphate; Bt2cAMP, dibutyryl cyclic ity secondary to hypoxia, and thus stimulate the adenosine monophosphate.