PALADE 43 



tions. We found that pancreatic microsomes [53] are closed vesicles (figs. 4 

 and 5) almost exclusively derived from the rough-surfaced part of the cells' 

 voluminous reticula. They are, however, more labile structures than their 

 hepatic counterparts: even when isolated in 0.88 M sucrose they enspherulate 

 and do not retain the flattened cisternal form that the elements of the endo- 

 plasmic reticulum have in situ. They are also more susceptible to various treat- 

 ments, especially to deoxycholate, which, at 0.3 per cent final concentration, 

 solubilizes ~ 85 per cent of their protein and ~ 40 per cent of their RNA. As 

 in the case of the liver, the DOC-insoluble material consists of RNA particles 

 (fig. 6) with a relatively high RNA/ protein ratio. The main difference be- 

 tween hepatic and pancreatic microsomes concerns their phospholipide content: 

 there is ~ 8 times less phospholipide in pancreatic microsomes than in their 

 hepatic counterparts, and the phospholipide concentration in the microsomal 

 fraction is equal to, or only slightly higher than, that in the original pancreatic 

 brei. 



Centrifugation of the microsomal supernatant resulted in the sedimentation of 

 further material which was arbitrarily divided into two postmicrosomal frac- 

 tions °( PMi and PM 2 ). We found that the corresponding pellets consist almost 

 exclusively of small, dense particles ~ 15 m^ in diameter (fig. 7), with a mod- 

 erate admixture of small vesicles in the PMi. Chemically both fractions were 

 made up of ribonucleoproteins with a negligible amount of phospholipides. 

 In the original fractionation scheme [53], the two postmicrosomal fractions 

 contained RNA in comparable concentrations, but in a recent modification 

 PMi has a noticeably higher RNA/protein ratio than PM 2 . 



The results obtained with the pancreas confirmed and extended our previous 

 findings on liver. In both the microsomes were found to be vesicles derived 

 either to a large extent [51], or almost exclusively [53], from the rough-surfaced 

 part of the endoplasmic reticulum. We used a characteristic structural detail, 

 namely the attached particles, to establish this derivation, and our conclusion 

 appears reasonably valid; there is good general agreement between the basic 

 morphological features of the microsomes and the fine structural details of the 

 endoplasmic reticulum. There is, however, no agreement as far as general dimen- 

 sions are concerned: the network is a continuous structure which may spread 

 throughout the entire cytoplasm of the intact cell (~ 20 X ~ 30 X ~ 40 fi), 

 whereas the microsomes are considerably smaller (0.05 to 0.3 n), corresponding 

 in dimensions to the vesicles, tubules, and cisternae which by their interconnec- 

 tions form the reticulum. Accordingly we must assume that the network is 

 broken during tissue grinding, the resulting fragments being the microsomes. 

 Since these fragments are closed vesicles, we are further obliged to postulate that 

 the broken segments heal readily into closed structures or that the fragmenta- 

 tion of the reticulum occurs by a generalized pinching-ofT process rather than 

 by mechanical disruption [51, 53]. 



