In 1877 Claude Bernard declared (71, ll?),“the indisputable fact is that cane sugar administration considerably increases liver glycogen content, but how does the sugar act-as a nutritive stimulator or a substance directly converted to glycogen?” In 1923 intermediary metabolism of carbohydrates was reviewed in Physiologicul Reviews (98). At that time, the initial step in glucose conversion to glycogen could only be ascribed to the transformation of glucose to a more reactive form that might undergo condensation to glycogen. In 1931 Cori (18) described, again in Physiological Reviews, the cycle, now bearing his name, in which glucose is converted to lactic acid by muscle and the lactic acid reconverted to glucose by liver. However, as Stetten and Stetten (109) wrote in 1960 in Physiological Reviews, measurement of the actual pathway by which glycogen is formed from glucose awaited the availability of isotopes. Hastings and colleagues (14, 112) in 1941-1942 showed that “C from [llC] bicarbonate and [llC] lactate was incorporated into liver glycogen. Bollman (8) in 1943 stated:“this may indicate that glucose is broken down to a three carbon intermediate before resynthesis to glycogen.” Boxer and Stetten (9) in 1944, from the enrichment of deuterium in liver glycogen from fasted rats given glucose and lactate along with ‘H, O, concluded that “glycogenesis was from fragments smaller than hexose rather than glucose directly.” However, Cook and Lorber (17), eight years later, on injecting [l-14C] glucose into rats, found that the major portion of 14C in the glucosyl units of hepatic glycogen was retained in Cl. This finding was confirmed by others, ie, 280% of the label from specifically labeled glucose was unrandomized in its conversion to glycogen. Hers (31) in 1955 concluded:“this provides a strong indication that glucose is built into glycogen molecules essentially as intact 6-carbon units.” Some 20 years later, Shikama and Ui (99) administered glucose to fasted rats injected with [14C] bicarbonate and used the incorporation of 14C into blood glucose and liver glycogen as the measure of gluconeogenesis and glycogenesis. Increased incorporation of 14C into glycogen compared with glucose consequent to a glucose load led them to conclude that a glucose load directs the final product of hepatic gluconeogenesis from blood glucose to hepatic glycogen. Although glycogen was readily formed in those in vivo glucose administrations, glucose proved a poor precursor of glycogen in rat liver preparations in vitro (eg, see Refs. 10, 26, 30, 42). That finding led to the suggestion that much of hepatic glycogen initially accumulating in rat liver following refeeding after a period of starvation is derived from nonglucose precursors. That glucose administration readily replenished glycogen stores in vivo and yet glucose conversion to glycogen is very limited in vitro was called by Katz and McGarry and co-workers (47, 88) the “glucose paradox.” They reexamined the pathways by which hepatic glycogen stores, depleted on fasting, are replenished on glucose administration in vivo. From their quantitations they concluded that the major pathway of glucose conversion to glycogen in vivo, after a fast, is by an indirect pathway, ie, glucose-glucose 6-phosphate-three-carbon compound (s)-glucose 6-phosphate-glycogen, and not directly, ie, glucose-glucose 6-phosphate-glycogen (43, 47, 71).

Their experimental findings, which led to and were used to further support that conclusion, are contained in five papers (46, 48, 51, 77, 78). These are summarized after first examining the definition of the pathways. Then that conclusion is evaluated in consideration of the findings by other …