PALADE 41 



and more or less extensive association was considered a secondary phenomenon 

 that occurs at a relatively late stage in the evolution of cellular organization. 

 The view, originally based on findings on undifferentiated or embryonic cells, 

 subsequently received full support from electron-microscope studies of various 

 bacteria [48, 49, 50] which revealed that bacterial protoplasm contains a large 

 population of small, dense particles (usually smaller than those found in animal 

 cytoplasm), but apparently no internal membranous system comparable to the 

 endoplasmic reticulum [50]. 



A Correlated Morphological and Biochemical Analysis of Hepatic Micro- 

 somes. The next task was to put the hypothesis to a test by trying to find out in 

 what cell fractions the particles segregate during the differential centrifugation 

 of tissue breis. Fraction chemistry and particle distribution were correlated by 

 using duplicate pellets of the fractions under study : one pellet for biochemical 

 analysis, and the other for electron microscopy after appropriate fixation, em- 

 bedding, and sectioning. It was found both necessary and expeditious to fix the 

 pellets in toto, and to cut them in such a way as to be able to survey them from 

 top to bottom. With such precautions, the existence and the extent of inter- 

 contamination among cell fractions could be easily detected and the presence of 

 distinct layers in some pellets clearly demonstrated. 



The work, carried out in collaboration with Dr. Philip Siekevitz, started 

 with an analysis of the microsomal fraction isolated from rat-liver breis [51]. 

 We found that this fraction consists almost exclusively of closed vesicles limited 

 by a dense continuous membrane, ~ 7 mp thick, and filled with a material of 

 relatively low density. Most of these vesicles are derived from the rough-surfaced 

 part of the endoplasmic reticulum as indicated by the small (~ 15 mp), dense 

 particles attached to the outer surface of their limiting membrane. Smooth- 

 surfaced vesicles are also present in the microsomal fraction, but their origin is 

 more difficult to ascertain; they may represent fragments of the smooth-surfaced 

 part of the reticulum or they may be derived from other sources (Golgi com- 

 plex? 8 cell membrane?). In 0.88 M sucrose, the medium used in our experi- 

 ments, the microsomal vesicles retained the flattened appearance of intracellular 

 cisternae and reacted like osmometers to changes in the concentration of the 

 medium: they swelled in hypotonic media. Treatment with versene (2 per cent 

 in 0.88 M sucrose, for 60 minutes at 0° C) removed ~ 60 per cent of the micro- 

 somal RNA and resulted in extensive loss of attached particles. Incubation in 

 ribonuclease (0.5 mg/ml 0.88 M sucrose; 60 minutes at 37° C) caused RNA 

 losses of ~ 85 per cent and produced a heavy agglutination of microsomal 

 vesicles. During the incubation, the attached particles apparently were lost. 

 Finally treatment with sodium deoxycholate (DOC) (0.5 per cent in 0.88 M 

 sucrose, at pH 7.5) "solubilized" most of the protein and phospholipides of the 



8 According to our interpretation [38], the "Golgi complex" is a differentiated part of 

 the endoplasmic reticulum. Other cytologists [52] consider this structure a distinct and 

 independent cell organelle. 



