CONVERSION OF FAT TO CARBOHYDRATE 223 



4. General Experimental Evidence on the Question of the Conver- 

 sion of Fat to Carbohydrate 



(/) D\N Ratio 



The D:N ratio was one of the earUest constants which was employed to 

 determine whether or not carbohydrate can originate from fat. It is be- 

 heved that, when a dog is rendered diabetic by phlorhizin and is then sub- 

 jected to fasting, the major part of the reserve glycogen and glucose is 

 flushed out within the first forty-eight to seventy-two hours. After this is 

 accomplished, any glucose subsequently excreted in the urine could have its 

 origin only in the protein catabolized, or in the fat broken down. Accord- 

 ing to Stiles and Lusk,^^ the D:N ratio under ideal conditions, after the 

 extra glucose from stored carbohydrates has been removed, is 3.65:1; 

 most workers have regarded ratios of 3.9 to 3.3 as acceptable in experiments 

 in which this technic was employed. If one accepts the hypothesis that 

 none of the glucose originated from fat, it can be calculated that approxi- 

 mately 58% of the protein molecule is being converted to glucose. This is a 

 reasonable figure, since there is adequate evidence that some of the amino 

 acids are not glucose formers. However, if one assumes that all the carbon 

 in protein is converted to glucose (other than that excreted in urea), one 

 gram of urinary nitrogen would correspond to 8.2 g. of glucose. However, 

 since the protein nitrogen is largely excreted as urea, one must correct the 

 carbon output from protein available for glucose formation by the quantity 

 lost in the urea molecule. This would bring about a revision of the figure 

 for the total potential glucose per g. nitrogen to 6.5 g. On this basis any 

 D : N ratio during fasting which exceeded 6.5 : 1 could be interpreted only as 

 indicative of the transformation of fat to carbohydrate. ^- 



Although most w^orkers have agreed with Stiles and Lusk^^ that the D : N 

 ratio in fasting phlorhizinized dogs approximates 3.6: 1, and that it is unin- 

 fluenced by the administration of fat, contrary results have been reported. 

 Thus, Hartogh and Schumm^* reported D:N ratios as high as 13.0:1 in 

 fasting phlorhizinized dogs between the third and seventh days after the 

 first administration of the glucoside. However, Lusk- strenuously denied 

 the accuracy of these data, and suggested that they might possibly be as- 

 cribed to an inaccuracy in the determination of urinary nitrogen. Thus, ac- 

 cording to Lusk, 2 Hartogh and Schumm reported urine nitrogen values of 

 4.89 and 5.17 g. on the third and seventh days after phlorhizin, respectively, 



8^ P. G. Stiles and G. Lusk, Am. J. Physiol., 10, 67-79 (1903-1904). 

 «* Hartogh and O. Schumm, Naunyn-Schmiedeberg's Arch, exptl. Pathol. Pharmakol., 

 45, 11-45(1901). 



