Hexose monophosphate shunt and labelling of amino acids in rat brain after injection of [U-14C]glucose.

unlabelled glucose 1-phosphate was added and changes in the concentration of glucose 1-phosphate and glucose 1-[32P]phosphate activity were followed as the system relaxed back towards equilibrium. The results from a number of such experiments at low concentrations of glucose monophosphates are shown in Fig. 1. With 32P-labelled substrates, co-transport of glucose I-[32P1phosphate into glucose 6-phosphate was observed. The activity of glucose l-[32P]phosphate fell to 37% of its original value and then returned to that value as chemical equilibrium was reached. Any fall in 32P radioactivity in the I-position of glucose 1,6-biphosphate would not affect this value significantly. The fall in glucose l-[32Plphosphate radioactivity is very similar to that expected for a phosphoenzyme mechanism (see line 3, Fig. 1). In similar experiments with 14C-labelled substrates no induced transport was observed (line 1, Fig. 1). Taken together these results support mechanism 1 and exclude mechanisms 2 and 3. However, mechanism 3 could give rise to the observed induced-transport pattern if the isomerization of the enzymeglucose 1,6-biphosphate complex is rate-limiting as the degree of co-transport for 14Cand 32P-labelled substrates is decreased towards that of mechanism 1. This possibility has been excluded, for to decrease the '*C-co-transport to less than that indicated by line 4 in Fig. 1, the K , for glucose 1-phosphate would need to be five times less than the experimental value (see Britton & Clarke, 1968). Similarly any isomerization of the phosphoenzyme (mechanism 1) should decrease the extent of 3zP-co-transport and lead to 14C-counter-transport especially at high substrate concentrations. No 14C-counter-transport was observed, and under these conditions, experiments indicated that the rate constants for any isomerization must be in excess of 2.7 x Thus these results are only compatible with mechanism 1 or 1'. To determine the nature of the phosphoenzyme, 10 units of enzyme were incubated for 1 min at 3OoC with 0.25 pCi each of 32Pand "C-labelled glucose 1,6-biphosphate in 1 6 0 m ~ histidine/40m~-tris/5 mM-MgCI,, pH 7.4. In the same buffer at 4OC, the mixture was passed down a Sephadex G-25 column. No radioactivity was associated with the enzyme fraction, which suggests that the phosphoenzyme is unstable (probable half-life <6min at 4OC). In contrast the phosphoenzyme from muscle is stable. Gel electrophoresis by the method of Laemmli & Favre (1973) showed that the isoenzyme patterns of the rabbit muscle and liver enzymes are different. It is therefore possible that the apparent bisphosphatase activity associated with the liver enzyme is of metabolic importance.