Visualization of a repeating subunit organization in rat hepatocyte chromatin fixed in situ

0 Laboratoire de Pathologie cellulaire, 15 rue de I'Ecole de Medecine , 75270 Pans Cedex 06 , France CHROMATIN FIXED IN SITU" The chromatin structure of rat hepatocytes fixed in situ was investigated in thin sections selectively stained for DNA with the Feulgen-like osmium-ammine reaction. Visualization of the chromatin structure was facilitated by the dispersal that occurred in the chromatin of regenerating rat hepatocytes at 24 h after partial hepatectomy. Our results show that repeating subunits were present both in the clumps and in the extended fibres of chromatin. These subunit particles appeared to be composed of an electron-translucent core delimited by a strongly electron-opaque DNA ring. Depending on the section plane, they displayed a morphology varying from a flat, oval shape to a rough circle. The mean diameter of the repeating subunits was 11 0 8 nm (S.D.) with a range from 9-5 to 13 nm. The thickness of the DNA ring was about 3nm. Regarding the association of the subunits in the chromatin fibres, two types of fibres were visualized, 11 nm and 20-25 nm thick, respectively. The thinner fibres were composed of a linear array of adjoining subunits. In the thicker fibres the subunit organization did not clearly form discrete superunit structures. VISUALIZATION OF A REPEATING SUBUNIT ORGANIZATION IN RAT HEPATOCYTE CHROMATIN FIXED IN SITU" HERNANDEZ-VERDUN1 Evidence that the bulk of chromatin is organized into repeating subunits called nucleosomes, as well as knowledge about nucleosome structure, have been derived from investigations carried out withm vitro chromatin (Chambon, 1978; Felsenfeld, 1978; Kornberg, 1977; McGhee & Felsenfeld, 1980). Since the first electron microscopic observation of this particulate structural organization of chromatin (Olins & Olins, 1974), artificial dispersal of chromatin has become a necessary step for the visualization of nucleosomes. The morphological study of chromatin structure in situ has been hindered by the relatively low resolution of chromatin structures when they are observed in ultrathin sections stained with the contrasting agents routinely employed. Together with the deoxyribonucleoproteins, many intermingled and varied substances are rendered electron-opaque within the thickness of the specimen, thus masking the morphology of the chromatin structures (Derenzini, Lorenzoni, Marinozzi & Barsotti, 1977). As a consequence of the lack of the morphological #We dedicate this work to the memory of our friend and colleague Dr Gilles Moyne. ("Author for correspondence, at Bologna. demonstration of nucleosome structure in chromatin fixed in situ, proof is still lacking that chromatin exhibits the same particulate organization in situ as that observed in vitro, even though psoralen was demonstrated to crosslink DNA in vivo at nucleosomal length intervals (Cech, Potter & Pardue, 1978). Recently, stereo electron micrographs have indicated the presence of nucleosome particles in conventionally fixed, embedded and sectioned preparations of chicken erythrocyte chromatin that had been previously dispersed in solutions of low ionic strength (Olins, Olins, Zentgraf & Franke, 1980). Both unravelling of the compact, highly ordered chromatin fibres and solubilization of the interchromatin substances that occur in artificially dispersed chromatin have greatly facilitated visualization of the particulate nature of chromatin in thin sections; at the same time, these processes have indicated changes in the chromatin structure and composition. Therefore, it remains open to question whether the nucleosome morphology really exists in chromatin in situ, or whether it is an artifact caused by the procedure used to obtain chromatin dispersal. We therefore studied the structure of chromatin in regenerating rat hepatocytes fixed in situ in thin sections stained with the Feulgen-like osmium-ammine reaction according to Cogliati & Gautier (1973). The osmium-ammine was shown to stain the DNA selectively in this reaction (Cogliati & Gautier, 1973; Gautier, 1976; Moyne, 1980). By rendering only the DNA electron-opaque, this staining procedure thus removed the abovementioned handicap when investigating chromatin structure in situ. Moreover, the physiological dispersal that occurs in regenerating rat hepatocytes at 24 h after partial hepatectomy (Derenzini, 1979) greatly improved the visualization of DNAcontaining structures in thin sections. The data thus obtained are compatible with chromatin organization into repeating subunits with a mean diameter of 11 0-8 nm (S.D.), made up of an electron-translucent inner core delimited by a thin, electronopaque DNA ring. Male Wistar rats, weighing 110130g, were used. Partial hepatectomy was performed according to Higgins & Anderson (1931) under ether anaesthesia. Rats were killed 24h after partial hepatectomy. Samples from regenerating livers were immediately fixed in a solution of 4 % formaldehyde in 0-1 M Sorensen buffer (pH 7-2). Formaldehyde was chosen as fixative to ensure that only DNA molecules in the chromatin were stained by the Feulgen-like procedure and to allow DNA structures with a thickness of about 3 nm to be visualized clearly (Derenzini, Viron & Puvion-Dutilleul, 1982). The samples were dehydrated in ethanol and embedded in Epon. Ultrathin sections were cut using a diamond knife. The Feulgen-like osmium-ammine technique was carried out on thin sections mounted on gold Fig. 1. Rat hepatocyte at 24 h after partial hepatectomy. Only the structures known to contain DNA are contrasted in the nucleus; no cytoplasmic organelles are stained. The DNA-containing structures appear to be dispersed in form. Many extended fibres are visible in the nucleolar body (n). Arrow indicates the structures visualized at higher magnification in Fig. 2. Feulgen-like reaction with osmium-ammine. X28 000. Fig. 2. Detail of area indicated by an arrow in Fig. 1. Arrowheads indicate roundish subunit particles made up of a thin electron-opaque filament encircling an electrontranslucent core in the compact DNA structures. Arrows show the subunits in two thin fibres whose thickness is formed by only one subunit. X175 000. grids without Formvar and hydrolysed for 30min at room temperature with freshly prepared 5 MHC1. The grids were then left for about 1 h in distilled water. During that time, osmium-ammine/ SO2 was prepared by bubbling SO2 for 30min through 10 ml of a suspension of 1 % osmiumammine in distilled water contained in a 25 ml cylinder. During SO2 bubbling the suspension turned black. The suspension was then left to settle until the liquid in the middle zone of the cylinder turned a yellow-brown colour. This yellow-brown liquid was used for the staining reaction. The grids were then thoroughly rinsed with distilled water. Some grids were post-stained with 0-5 % uranyl acetate in water. The sections were examined with a Jeol 100X electron microscope, equipped with a goniometer stage, operating at 100 kV with a (...truncated)