HYDROT.YSIS PRODUCTS OF CEREBROSIDES 501 



since the cerebrosides obtained from all the normal and pathological organs 

 examined contained small amounts of the glucose type of cerebroside. 

 This would indicate that it is a normal cell constituent. The problem still 

 unsolved is why the glucose form should be deposited exclusively in the 

 spleen in most cases of Gaucher's disease, while the normal galactose type 

 predominates in the brain.'""' One also wonders whether the galactose and 

 glucose residues may be interchanged without a complete rupture of the 

 molecule. 



8. iMethods for Determination of the Cerebrosides 



Although it is possible to determine the cerebroside content by ascertain- 

 ing the amount which can be isolated from a tissue, this procedure is far 

 from quantitative and is exceedingly time-consuming. Moreover, it can- 

 not be effectively applied where the concentrations of cerebrosides are 

 small. Most of the quantitative methods have been based upon the 

 determination of the sugar present in a sample of cerebroside after hydroly- 

 sis. 



NolP^i first employed an acid hydrolysis of the lipid mixture followed by 

 the determination of the liberated sugar by means of Fehling copper re- 

 agent. A somewhat similar procedure was used by Koch''" and also by 

 Winterstein and Hirschberg,^^^ who used the Bertrand method. In these 

 procedures the total amount of reducing substance found after hydrolysis 

 was considered to be galactose. NolP^^ reported that no reduction occurs 

 with the unhydrolyzed lipid mixture. Kirk^''^ reports, however, that when 

 the Somogyi reagent was used for the determination of the reducing sugar, 

 the alcoholic brain extracts invariably showed a definite reduction before 

 hydrolysis. This author interprets this as indicating that the results ob- 

 tained by the earlier methods are too high because of the failure to correct 

 for the original reducing action of components in the tissue. 



The first method introduced by KimmelstieP^" offered a considerable 

 improvement over the earlier ones. The Hagedorn-Jensen ferricyanide 

 method was employed for the determination of the reducing subtances, 

 and the initial reduction values of the lipid mixture were subtracted from 

 the results obtained after hydrolysis. However, the values used for cor- 

 rection were about 50% of the totals obtained after hydrolysis; it is there- 

 fore evident that they cannot be very accurate. In his second procedure, 

 Kimmelstiel^''^ modified his technique by the introduction of a preliminary 

 precipitation with zinc hydroxide, which considerably lowered the reducing 

 value on the unhydrol3''zed sample. However, Kirk^^^ points out that an 



"3 W. Koch, Am. J. Physiol., 11, 303-329 (1904). 



«* H. Winteretein and E. Hirschberg, Biochem. Z., 159, 351-369 (1925). 



^'6 E. Kirk, J. Biol. Chem., 123, G13-621 (1938). 



"« P. Kimmelstiel, Mikrochemie {Pregl Feslschrifl), 165-177 (1929). 



