FUNCTIONAL CIIANGUS IN STIU'CTUKK ^3 



partially freed from the surrounding membranes. To he sure, this heavy sub- 

 fraction is a heterogeneous preparation which contains, in addition to intracisternal 

 granules, various microsomal detritus, mainly damaged membranes. We also ob- 

 tained an intermediary subfraction made up primarily ot detached RNP particles 

 and a light subfraction representing the deoxycholate-soluble components ol the 

 microsomes, primarily their licjuid content antl their solubilized membranes. In 

 fed animals, proteolytic and ribonuclease activities were more concentrated in the 

 heavy microsomal subfraction than in the parent microsomes. In starved animals, 

 the corresponding subfraction showed no increase in specific activity over the 

 microsomes (37). These findings clearly indicate that the intracisternal granules 

 are masses of segregated digestive enzymes. To my knowledge, this is the first 

 instance in which a product of the endoplasmic reticulum has been demonstrated 

 in the form of well defined granules within the cavities of the system, and has 

 been identified biochemically. 



The study of the microsomal subfractions has revealed another feature worth 

 mentioning. The intermediary subfractions, which primarily consist of detached 

 particles, contain proteases and ribonuclease in remarkably high concentrations. 

 If we assume that the finding reflects the situation in situ, then we may have 

 reached here, at the level of the attached RNP particles, the site, or at least one of 

 the sites, of protein synthesis in the cell. To verify this assumption we must 

 demonstrate that the enzymes in question are new proteins still attached to their 

 site of synthesis, not old enzymes relocated by adsorption during tissue homog- 

 enization. This important distinction remains to be established by future work. 

 For the moment we can strengthen the first alternative by some evidence de- 

 rived from experiments in which the incorporation of leucine-i-C'^ into the mixed 

 proteins of various cell fractions was followed with time. The curves (fig. 5) show 

 that the highest early incorporation takes place in the attached particles and the 

 crossings on the graph are compatible with a transfer of labeled proteins from 

 the particles to the microsomal content, that is, the intracisternal granules, and 

 from the latter to the zymogen granules. It should be stressed that the results of 

 these experiments do not validate our hypothetical pathway; they are only com- 

 patible with it. A more convincing test can be obtained by following with time 

 the radioactivity of a single protein, chymotrypsinogen for instance, produced by 

 the cell. Information of this kind is not yet available. 



Membrane Exchange During Zymogen Discharge. The third example con- 

 cerns the membrane of the zymogen granules. It is finally known, as a result 

 of the work carried out by Marshall with fluorescein-labeled antibodies (2^), 

 and by Hokin (13) and us (36) with cell fractions, that Heidenhain's hypothesis 

 is correct and that the zymogen granules deserve their name. Indeed, they contain 

 trypsinogen and chymotrypsinogen — the classical zymogens — and in addition pro- 

 carboxypeptidase, ribonuclease, lipase and amylase. All these hydrolytic enzymes. 



