108 



I. R. GIBBONS AND J. R. G. BRADFIELD 



BRAIN-TISSUE 

 EXTRACTABILITY 



D Ffesh motenol 



■ Os(X - filed moleriol 



obviously makes an exception in this series, because 

 extractability is considerably increased. An explana- 

 tion would be that here the fixation has opened up 

 the tissue and rendered the lipids accessible for 

 extraction. 



The 2 "^'o lipids bound by the osmium fixation must 

 consist of unsaturated compounds. In rats fed with 

 a diet containing 5 °o fat, about 70 °o of the tissue 

 fatty acids are unsaturated (3). Mammalian fat has 

 as a rule an iodine number of 65 corresponding to 

 a total capacity to bind 0.47 g Os per gram fat. 

 With regards to the fact that only 70 % of such fat is 

 actually unsaturated, we calculate the osmium up- 

 take by the unsaturated fat fraction to 0.63 g osmium 

 per gram. 



Ten per cent fat can ordinarily be extracted from 

 tissues, which means that 100 g tissue dry weight 

 has a maximal capacity to bind 4.7 g osmium by 

 its lipids. Our analysis showed that maximal uptake 

 values for liver tissue were about 15 g osmium per 



100 g tissue dry weight. Thus the uptake exceeds 

 the theoretical capacity of the tissue lipids by |. 

 In other words, a considerable part of the osmium 

 is bound by substances other than lipids. 



Direct analysis shows that but little of the binding 

 capacity of the lipids is actually used. Instead of 

 theoretically possible 4.7 g only 1.3 g are taken up 

 by the lipids in 100 g tissue dry weight. That is 8.4% 

 of the total osmium uptake, (fig. 2.) 



In consequence neither proteins nor lipids are 

 significantly preferred by osmium. 



Upon a question by Dr. Sjostrand if there might 

 possibly be a preference of certain compounds in the 

 lipid fraction itself, an analysis of fixed and unfixed 

 brain lipids was carried out. (fig. 3.) Technical de- 

 tails of the rather complicated analytical procedure 

 cannot be discussed here and will be published else- 

 where. 



To demonstrate preferences of the osmium tetroxide 

 fixation towards certain lipids an amount of osmium 

 tetroxide was added corresponding to about 30 % 

 of the theoretical capacity of uptake. Six per cent 

 lipids have been rendered unextractable but again 

 no real significant preference for one type of lipid 

 could be observed. 



It may thus be said that osmium tetroxide fixation 

 stains reactive compounds in tissues in even pro- 

 portions, without giving significant preference to 

 certain substances. Unfortunately more cannot be 

 said from this study but that the well-known pattern 

 of repeating dense and light bands in thin sections 

 corresponds to accumulations of reacting and un- 

 reacting groups. 



References 



1. Bahr, G. F., Exptl. Cell Research 7, 457 (1954). 



2. — ibid. 9, 277 (1955). 



3. Deuel, H. J., The Lipids. N.Y. Interscience Publ. (1951). 



The Fixation of Nuclei in Locust Testis 



I. R. Gibbons and J. R. G. Bradfield 



Cavendish Laboratory, Cambridge 



Previous studies on the problem of fixation, both 

 with light and electron microscopes, have demon- 

 strated the excellence of buffered osmium tetroxide in 

 the preservation of cytoplasmic ultrastructure (4, 5). 

 This work has been confirmed by the highly organ- 

 ized and highly reproducible ultrastructures observed 

 with the electron microscope in the cytoplasm of 

 cells after osmium fixation. However, with regard to 

 the nucleus the situation is quite different. Little 

 apparently organized ultrastructure has been observed 

 in osmium-fixed nuclei either interkinetic or dividing. 

 In view of this fact we have considered it desirable 

 to re-examine the question of the fixation of nuclei. 



The chromatin distribution within live nuclei of 

 locust primary spermatocytes has been investigated 

 by observing squashes of testis follicles in Belar 

 solution with an ultra-violet microscope. Nuclei in 

 meiotic prophase cells appear to have their chroma- 

 tin arranged in intertwined strands, the thickness of 

 these strands is variable and increases as the cell 

 approaches diakinesis. When, however, osmium 

 fixed nuclei of this material are examined in electron 

 micrographs of thin sections they usually appear to 

 have a rather homogeneous fine-grained appearance, 

 and show little, if any, trace of the inhomogeneity 

 of chromatin existing in live nuclei. In order to re- 



