LYSOSOMES, A NliW GROUP OF CYTOPLASMIC PARTICLES 1 33 



found to show a distribution similar to that ol acid phosphatase. This fact, to- 

 gether with the close similarities in structure-linked latency and activation men- 

 tioned above, led to the conclusion that all five enzymes belong to the same 

 special class of cytoplasmic particles, which were termed lysosomes as a reminder 

 of their manifold hydrolytic activities (29). 



The inclusion of aryl-sulphatases A and B amongst the lysosomal enzymes 

 (see fig. i) rests on the fractionation experiments of Roy (84, 85) and of Dodg- 

 son, Spencer and Thomas (30) and on the comparative investigation by Viala 

 and Gianetto ( 101 ) of the activation of these enzymes and of acid phosphatase. 

 According to recent observations by Paigen and (iriffiths ( personal communica- 

 tion), the group, as defined by the above criteria, may also include a phospho- 

 protein phosphatase. 



In recent years, our efforts have been directed towards a better dissociation 

 between lysosomes and mitochondria, taking advantage of the new possibilities 

 provided by density-gradient centrifugation. Figures 3-6 illustrate some results 

 which have been obtained by Beaufay, Baudhuin and Bendall (6). These graphs 

 inay serve to delineate the precise experimental basis as well as the actual limita- 

 tions of our concept of lysosomes. 



The frequency curves of figure 4 indicate a mean sedimentation constant ( in 

 0.25 M sucrose and at 0°) of about 5,000 Svedberg units for acid phosphatase 

 and of 10,000 S for cytochrome oxidase. The respective mean densities (in 0.25 m 

 sucrose) of the particles bearing these enzymes have recently been estimated by 

 density-equilibration experiments in gradients of colloidal thorium oxide (Thoro- 

 trast) at 1.13-1.15 for acid phosphatase and i.io-i.ii for cytochrome oxidase 

 (Beaufay, unpublished). From these values, average diameters of 0.4 and 0.8 

 micron can be calculated. The shape of the two curves suggests strongly the 

 presence of two distinct populations, which cannot be separated quantitatively 

 from each other by centrifugation owing to the fairly wide overlapping of their 

 sedimentation constants. All that can be done is to cut in somewhere between 

 the two curves and to separate two fractions differing in their relative content 

 of the two enzymes. That is actually what we believe has been done to some 

 extent with the classical fractionation scheme (fig. 2./), and in a more clearcut 

 fashion with the new scheme (fig. 2B). 



Our thesis, therefore, which has been set forth more elaborately in previous 

 reviews (24, 21), is that results such as those of figure 2 indicate that the iso- 

 lated fractions are heterogeneous and contain mixtures of at least two distinct 

 populations of particles in different proportion. Implicit in this interpretation are 

 the rejection of the alternative hypothesis that the results may be due to the 

 heterogeneity of the particles themselves within a single group, and therefore 

 the acceptance of the postulate that "granules of a given population are en- 

 zymically homogeneous or at least cannot be separated by centrifuging into sub- 

 groups differing significantly in relative enzymic content" (29). 



