J. Cell Sci. 51, 85-94 (1981) 85Printed in Great Britain © Company of Biologists Limited 1981

ELECTRON MICROSCOPIC STUDIES OF

SILVER-STAINED PROTEINS IN NUCLEOLAR

ORGANIZER REGIONS: LOCATION IN

NUCLEOLI OF RAT SYMPATHETIC NEURONS

DURING LIGHT AND DARK PERIODS

MARIE-JOSEPHE PfiBUSQUE AND RAYMOND SEITE,with the technical assistance of Mireille Vio (C.N.R.S.)Groupe de Neurocytobiologie, Laboratoirt de Biologie Cellulaire et Histologie,Faculty de Midecine, 27 Boulevard Jean Moulin, 13385 Marseille Cedex 5 -France

SUMMARY

Ag-AS staining of nucleolar organizer regions was carried out on interphasic superiorcervical ganglia neurons of rats sacrificed during light and dark periods. While the Ag-AStechnique has mostly been used on monolayer cell lines or cell suspensions, the present studyshowed that in electron microscopy this technique is also applicable to small pieces of tissues.The finest pictures are obtained when (1) all solutions used for the staining procedure are atpH 4-5-4-7 and (2) the second step of the reaction involving ammoniacal silver and formalindeveloping solutions does not exceed 3 min. The results indicate that in the 2 time periodsstudied, a positive reaction took place exclusively in nucleolar fibrillar centres and in thefibrillar ribonucleoprotein (RNP) component (dense fibrillar component). The other nucleolarcomponents, i.e. granular and vacuolar, were devoid of silver deposits. As previously shown insympathetic neurons, the fibrillar centres of the nucleoli show a 10-fold increase in volumeduring the dark period. In this period, silver grains were located on both 'giant' and small-sized fibrillar centres. The fibrillar RNP component seen either at the periphery of fibrillarcentres or in the form of a well-delimited network showed the strongest reaction. The samedistribution of silver grains was observed in the sympathetic neurons of rats sacrificed duringthe light period. Here again, silver accumulation occurred exclusively in the fibrillar centresand the fibrillar RNP component. The same difference in reactivity was observed as for thedark period, the fibrillar RNP component being the main site of the reaction.

INTRODUCTION

The Ag-AS method of Goodpasture & Bloom (1975) is known to specifically stainthe nucleolar organizer regions (NORs) in chromosome preparations under lightmicroscopy. The silver staining has been shown to be associated with non-histonenucleolar protein (Goodpasture & Bloom, 1975; Howell, Denton & Diamond, 1975;Schwarzacher, Mikelsaar & Schnedl, 1978; Olert, Sawatzki, Kling & Gebauer, 1979)at definite reactive sites (Lischwe, Smetana, Olson & Busch, 1979; Mamrack, Olson& Busch, 1979) involving reducing groups (Buys & Osinga, 1980). But the nature ofthe stained nucleolar material has not yet been ascertained, although it has beenshown that proteins C23 and B23 are the major nucleolar silver-stained proteins(Busch, Daskal, Gyorkey & Smetana, 1979; Lischwe et al. 1979; Mamrack et al. 1979;

86 M.-J. Pdbusque and R. Seite

Daskal, Smetana & Busch, 1980). In several kinds of cells at different stages ofactivity, the degree of silver staining of the nucleoli and the rate of RNA synthesiswere found to be correlated strongly (Miller, Dev, Tantravahi & Miller, 1976; Engel,Zenzes & Schmid, 1977; Hansmann, Gebauer, Bihl & Grimm, 1978; Schwarzacheret al. 1978; Busch et al. 1979; Hofgartner, Krone & Jain, 1979).

Recently, the Ag-AS method has been adapted to the ultrastructural level in cellcultures in interphase, mitosis and meiosis. The silver staining was located in thefibrillar components (Schwarzacher et al. 1978; Daskal et al. 1980; Hernandez-Verdun, Hubert, Bourgeois & Bouteille, 1980), especially in the fibrillar centresregarded as the interphasic counterpart of the NOR (see reviews by Jordan &Loening, 1977; Bouteille & Hernandez-Verdun, 1979; Goessens & Lepoint, 1979) andin the fibrillar ribonucleoprotein (RNP) component, also called the dense fibrillar com-ponent, which is the site of the ribosomal RNA synthesis (Goessens, 1976; Knibiehler.Navarro, Mirre & Stahl, 1977; Mirre & Stahl, 19786; Fakan & Puvion, 1980).

However, it is generally claimed that the fibrillar centres are small-sized and difficultto identify in interphase nuclei. By using a stereological and ultrastructural analysis,we found previously that, in the rat sympathetic neuron (a permanent interphasiccell), the development of fibrillar centres exhibits circadian changes with significantlylarger nucleolar centres in neurons of animals sacrificed at night (Pebusque & Selte,1980). Recently, in rats subjected to an artificial synchronization, we demonstratedthat the variations in the mean volumes of both nucleoli and fibrillar centres insuperior cervical ganglia (SCG) neurons follow a diurnal rhythm with a 10-foldincrease of the latter (Seite, Pebusque, Robaglia & Cau, 1980; Pebusque, Robaglia &Seite, 1981; Pebusque & Seite, unpublished). The nucleoli of rat SCG neurons canthus be used as a model for morphological, cytochemical and functional investigationsof the nucleolus. The aims of the present study were as follows: (1) to adapt theAg-AS method at the ultrastructural level to our material, i.e. rat SCG, in view ofthe fact that we used block tissues and not cell cultures as described by severalauthors (Schwarzacher et al. 1978; Daskal et al. 1980; Hernandez-Verdun et al.1980); (2) to determine the optimal staining conditions for increasing resolution byreducing the size of silver grains; (3) to identify the interphasic silver-stainedcomponents in nucleoli of SCG neurons of rats sacrificed during both light and darkperiods and to assess a possible variability in stained structures, according to the 24-hcycle. A preliminary note dealing with Ag-stained NOR proteins in nucleoli of ratSCG neurons during the dark period only has been published elsewhere (Pebusque,Vio & Seite, 1981).

MATERIALS AND METHODS

Eight male adult Sprague—Dawley rats (250-300 g) were used in these experiments. ArtificialLighting was applied and the light/dark cycle was 12 h of light (07-19 h) and 12 h of darkness(19—07 h). Access to food and water was allowed ad libitum. The rats were randomly assignedto one of two groups.

Group 1 (4 rats) sacrificed between 00 and 01 h (local time)Group 2 (4 rats) sacrificed between 14 and 15 h (local time)

Ag-stained nucleolar proteins in neurons 87

At the time of sacrifice, all the animals were anaesthetized with 40 mg/kg intraperitonealsodium barbiturate. The superior cervical ganglia (SCG) were removed and immediatelyimmersed in 1 -6 % glutaraldehyde in Sorensen buffer (pH 72 at 4 °C). The SCG were eitherleft intact after removal or desheathed and cut into 8-10 small pieces in the fixative solution.The time taken to perform the latter operations did not exceed 10 min. The pieces were fixedfor s min at 4 °C in Carnoy's solution (acetic acid/ethanol, 1:3), gradually rehydrated inethanol and rinsed in distilled water for 10 min. The Ag-AS procedure of Goodpasture &Bloom (1975) was then applied as described by Hernandez-Verdun et al. (1980), but theammoniacal silver solution and formalin-developing solution step was performed in successivetrials for 215, 2-30, 3, 3-30 and 4 min. Moreover, the entire procedure was performed at2 different pH values, 4-5-4-7 and 5-3-5-5, respectively.

The fragments were then rinsed in distilled water at the appropriate pH, dehydrated inethanol and embedded in Epon. Thin sections were made on a Reichert Ultramicrotome,contrasted or not by uranyl acetate and lead citrate and examined in a Siemens Elmiskop 101electron microscope at 80 kV.

RESULTS

Under our experimental conditions, a satisfactory preservation of the nuclear andcytoplasmic ultrastructure was observed with a glutaraldehyde prefixation prior tofixation with Carnoy's solution, despite the extractive effect of the latter. The nucleoliwere well preserved and it was possible to identify all the nucleolar components.

It must be pointed out that when the silver reaction was performed in superiorcervical ganglia that had been removed and preserved intact, no silver-stainedstructure was found. On the contrary, in SCG that had been desheathed, dilaceratedand cut into small pieces prior to the staining procedure, the silver staining waspresent on well-defined nucleolar components.

In order to define optimum conditions for locating the stained nucleolar structures,several improvements involving monitoring of the pH of each solution and anammoniacal silver-formalin step for different times were tested.

It was found that the optimal pH of each solution during the entire procedure was4-5-4-7. In this case, indeed, no significant background was found in the nucleoplasmor in the cytoplasm, whereas it was very noticeable when solutions were used at pH

5-3-5-5-Moreover, the finest pictures were obtained when the ammoniacal silver solution-

formalin developing solution step did not exceed 3 min and all solutions used for thestaining procedure were at pH 4-5-4-7. It was then possible to compare the differentstained nucleolar components in the same nucleolus and also in nucleoli of rat SCGneurons during the dark and light periods.

Silver-stained structures in nucleoli of SCG neurons of rats sacrificed during the darkperiod

In all sections, whether post-stained or not with uranyl acetate and lead citrate, itwas found that the staining reaction was restricted to the nucleoli in specific delimitedareas corresponding to fibrillar components. So, the highly voluminous fibrillarcentres clearly appeared as specific sites of the silver reaction (Fig. 1). In addition, thefibrillar RNP component located at the periphery of the large fibrillar centre was also

88 M.-J. Pibusque and R. Seite

•a

Ag-stained nucleolar proteins in neurons

t ..,

tig. 3. Section ot Ag-AS stained reticulated nucleolus in sympathetic neurons ofsuperior cervical ganglia of rats sacrificed during the dark period. In this case, thelarge fibrillar centre (/c), the smaller one (arrow) and the well-delimited network ofthe fibrillar RNP component are exclusive sites of silver accumulation. Silver grainsare always denser in the fibrillar RNP component than in the fibrillar centres them-selves. Post-staining with uranyl acetate and lead citrate, x 28000.

covered with silver grains. The most striking fact was that silver grains were denserin the latter than in the large fibrillar centres themselves.

As shown by comparison with nucleoli obtained under standard fixing and stainingconditions (Fig. 2), the other small, more or less round-shaped, nucleolar areas alsocovered with silver deposits appeared to be small fibrillar centres with their associated

Figs. 1-2. Sections of reticulated nucleoli in sympathetic neurons of superiorcervical ganglia of rats sacrificed during the dark period.

Fig. 1. Ag-AS stained section. Small-sized silver grains are located in the largefibrillar centre (Jc) and in the smaller ones (arrows). Note the high density of silvergrains on the fibrillar RNP component located either at the periphery of the fibrillarcentres or in form of a discontinuous layer. No grains are visible in the granular RNPand vacuolar components. The thin section was post-stained with uranyl acetate andlead citrate, x 25 000.

Fig. 2. Nucleolus ultrastructure after conventional fixation and staining for electronmicroscopy. A highly voluminous fibrillar centre (fc) and some small ones (arrows)can be seen. A network of the fibrillar RNP component (electron-opaque fibrillarcomponent) is present in the nucleolus. x 25 000.4 CEL 51

M.-J. Pibusque and R. Seite

m.

Ag-stained micleolar proteins in neurons 91

fibrillar RNP component (Fig. 1). It should be noted also, that the density of silvergrains was stronger at the periphery than in the fibrillar centres themselves. Withoutpost-staining with uranyl acetate and lead citrate, the same difference was observed.

As can be clearly seen in Fig. 3, the Ag-positive nucleolar structures may exhibita well-delimited network according to the section plane. Comparison of silver-stainedstructures (Fig. 3) with the nucleoli obtained under standard fixing and stainingconditions (Fig. 2) clearly identifies the Ag-positive network as the fibrillar RNPcomponent.

It should be noted that the other nucleolar components, i.e. the granular RNPcomponent and the vacuolar component, are devoid of silver deposits.

Silver-stained structures in nucleoli of SCG neurons of rats sacrificed during the lightperiod

The same distribution of silver grains was observed in these nucleoli as in those ofSCG neurons sacrificed at night. Silver deposits were located either in the small-sizedfibrillar centres with their surrounding RNP component, giving a pattern of more orless round-shaped structures (Fig. 4), or in a network corresponding to the fibrillarRNP component (Fig. 5). It should be noted that we still observed a difference insilver-grain density between the 2 fibrillar components. The silver grains located inthe fibrillar RNP component were indeed denser than in the fibrillar centres them-selves (Fig. 4). This difference was not due to superimposition of standard post-staining and Ag-staining.

DISCUSSION

It appears from the foregoing observations that the distribution of the grains wasrestricted to particular nucleolar areas that could be identified as fibrillar centres andto the fibrillar RNP component, since grains were not detected in granular elementsof the nucleoli of SCG neurons. The fibrillar centres are indeed selectively stained bythe Ag-AS staining of nucleolar organizer regions, both in the case of the small onespresent during the light and dark periods and the highly voluminous fibrillar centreencountered only during the dark period as described in previous studies (Pebusque& Selte, 1980; Seite et al. 1980). Earlier studies have been carried out at the ultra-structural level by several authors (Hernandez-Verdun et al. 1978,1980; Schwarzacheret al. 1978; Daskal et al. 1980) and the silver-stained proteins located on the fibrillar

Figs. 4-5. Sections of Ag-AS stained reticulated nucloli of superior cervical ganglionneurons of rats sacrificed during the light period. Silver grains are accumulatedexclusively in the small-sized fibrillar centres (arrows) and in the fibrillar RNPcomponent.

Fig. 4. Silver grains are denser in the fibrillar RNP component than in thefibrillar centres themselves. The thin section was post-stained with uranyl acetate andlead citrate, x 25000.

Fig. 5. A silver-stained network corresponding to the fibrillar RNP componentcan be seen; without post-staining with uranyl acetate and lead citrate, x 25000.

•4-2

92 M.-J. Pibusque and R. Seite

component have been described. Our findings are in agreement with the latter, but itmust be pointed out that this is the first time fibrillar centres have been shown to bewholly silver-stained. On the other hand, under our experimental conditions, themost striking observation was that the accumulation of silver grains was more pro-nounced in the fibrillar RNP component than in the fibrillar centres themselves. Theproblem is to determine whether this difference is due to a difference in group re-activity or to the presence of a greater quantity of silver-staining proteins, or merelyto the superimposition of Ag-AS staining, and uranyl acetate and lead citrate used forpost-staining. It should be noted that, on sections without post-staining, the samedifference was observed. This fact eliminates the third possibility, but on the basis ofour results we cannot define an exact hypothesis.

The present experimental staining conditions, using small fragments of tissue,involving solutions at pH 4-5-4-7 and the second step of the Ag-AS procedure notexceeding 3 min, present a considerable advantage, in our opinion, in that the grainsize was 10-15 n m o r ^ess- Under these conditions, the small-sized silver deposits donot mask the underlying structures.

It has now been well established that the fibrillar RNP component is the main siteof rRNA synthesis (Lepoint & Goessens, 1978; Mirre & Stahl, 1978 a, b) and it isgenerally accepted that the transcriptional units are located in this component. Onthe other hand, in all the systems studied with light microscopy, the amount of silver-positive nucleolar material has been correlated to the functional activity of the nucleolarorganizer regions (Miller et al. 1976; Howell, 1977; Hansmann et al. 1978; Mikelsaar& Schwarzacher, 1978; Busch et al. 1979; Buys & Osinga, 1980; Hofgartner et al.1979; Schmiady, Miinke & Sperling, 1979). The fibrillar RNP component is thereforedirectly related to the transcriptional activity of nucleoli and we have found that,under our experimental conditions, this fibriller RNP component is the site of a greataccumulation of silver grains. However, as stated by Hernandez-Verdun et al. (1980),'neither the reason why NOR silver proteins are localized in the fibrillar centres, northe relationship between these proteins and the transcriptional activity of the nucleolihas yet been clarified'.

In SCG neurons, the pattern observed in the Ag-stained structures was that ofa well-delimited network corresponding to fibrillar RNP component linking thefibrillar centres together. Estable & Sotelo (1951), studying the nucleolus with Ag-staining methods, described an Ag-positive network of fibrils, which they called'nucleonema' (Estable, 1966). On the other hand, if multiple fibrillar centres represent(as suggested by Chouinard, 1971) sections of a long meandering stretch of chromatinsurrounded by nucleolar fibrillar component belonging to a chromosome nucleolarorganizer, the large fibrillar centre encountered during the dark period in nucleoli ofrat SCG neurons would only involve some and not all of the nucleolar organizingregions.

Specifically stained material is obviously present in nucleoli of sympathetic neuronsduring both light and dark periods. No significant difference was observed in thelocation and amount of silver grains in the small-sized fibrillar centres present duringthe 2 periods of time studied. But the most striking fact concerns the large fibrillar

Ag-stained nucleolar proteins in neurons 93

centre, always seen during the dark period, in which a great accumulation of silver-stained material is observed. This fact raises the question as to whether or not theincreased size of fibrillar centres during darkness might be correlated to an accumula-tion of silver-stained proteins.

At the ultrastructural level, it is not possible to make a quantitative estimate of thenumber of silver grains that accumulated in nucleolar sections. But our presentfindings suggest that there is an overall increase in the amount of argyrophylicproteins since silver deposits are located in the highly voluminous nocturnal fibrillarcentres as well as in the small-sized daytime ones.

We are grateful to H. Debroas/Soler for her expert assistance in microphotography andE. Canale for her skilful assistance in preparing the manuscript.

REFERENCES

BOUTEILLE, M. & HERNANDEZ-VERDUN, D. (1979). Localization of a gene: the nucleolarorganizer. Biomedtcine 30, 282-287.

BUSCH, H., DASKAL, Y., GYORKEY, F. & SMETANA, K. (1979). Silver staining of nucleolargranules in tumor cells. Cancer Res. 39, 857-863.

BUYS, C. H. C. M. & OSINGA, J. (1980). Abundance of protein-bound sulfhydryl and disulfidegroups at chromosomal nucleolus organizing regions. A cytochemical study on the selectivesilver staining of NORs. Chromosoma 77, 1-11.

CHOUINARD, L. A. (1971). A light- and electron-microscope study of the nucleolus duringgrowth of the oocyte in the prepubertal mouse, J. Cell Set. 9, 637-663.

DASKAL, Y., SMETANA, K. & BUSCH, H. (1980). Evidence from studies on segregated nucleolithat nucleolar silver staining proteins C23 and B23 are in the fibrillar component. Expl CellRes. 127, 285-291.

ENGEL, W., ZENZES, M. T. & SCHMID, M. (1977). Activation of mouse ribosomal RNA genesat the 2-cell stage. Hum. Genet. 38, 57-63.

ESTABLE, C. (1966). Morphology, structure and dynamics of the nucleonema. Natn. CancerInst. Monogr. 23, 91-105.

ESTABLE, C. & SOTELO, J. R. (1951). Una nueva estructura celular : el nucleonema. Publs Inst.Invest. Cien. Biol., Montivideo 1, 105-126.

FAKAN, S. & PUVION, E. (1980). The ultrastructural visualization of nucleolar and extra-nucleolar RNA synthesis and distribution. Int. Rev. Cytol. 65, 255-300.

GOESSENS, G. (1976). High resolution autoradiographic studies of Ehrlich tumour cell nucleoli.Expl Cell Res. 100, 88-94.

GOESSENS, G. & LEPOINT, A. (1979). The nucleolus-organizing regions (NOR'S): recent dataand hypotheses. Biologie cell. 35, 211-220.

GOODPASTURE, C. & BLOOM, S. E. (1975). Visualization of nucleolar organizer regions inmammalian chromosomes using silver staining. Chromosoma 53, 37-50.

HANSMANN, I., GEBAUER, J., BIHL, L. & GRIMM, T. (1978). Onset of nucleolus organizeractivity in early mouse embryogenesis and evidence for its regulation. Expl Cell Res. 114,263-268.

HERNANDEZ-VERDUN, D., HUBERT, J., BOURGEOIS, C. & BOUTEILLE, M. (1978). Identificationultrastructurale de l'organisateur nucle'olaire par la technique a l'argent. C.R. hebd. Sianc.Acad. Set., Paris 287, 1421-1423.

HERNANDEZ-VERDUN, D., HUBERT, J., BOURGEOIS, C. A. & BOUTEILLE, M. (1980). Ultrastruc-tural localization of Ag-NOR stained proteins in the nucleolus during the cell cycle and inother nucleolar structures. Chromosoma 79, 349-362.

HOFGARTNER, F. J., KRONE, W. & JAIN, K. (1979). Correlated inhibition of ribosomal RNAsynthesis and silver staining by actinomycin D. Hum. Genet. 47, 329-333.

HOWELL, W. M. (1977). Visualization of ribosomal gene activity: silver stained proteinsassociated with rRNA transcribed from oocyte chromosomes. Chromosoma 6a, 361-367.

94 M.-J. Pibusque and R. Seite

HOWELL, W. M., DENTON, T. E. & DIAMOND, J. R. (1975). Differential staining of the satelliteregions of human acrocentric chromosomes. Experentia 31, 260-262.

JORDAN, E. G. & LOENING, U. E. (1977). The nucleolus at Salmanca. Nature, Land. 269,376-378.

KNIBIEHLER, B., NAVARRO, A., MIRRE, C. & STAHL, A. (1977). Localization of ribosomalcistrons in the quail oocyte during meiotic prophase I. Expl Cell Res. n o , 153-157.

LEPOINT, A. & GOESSENS, G. (1978). Nucleologenesis in Ehrlich tumour cells. Expl Cell Res." 7 , 89-94.

LISCHWE, M. A., SMETANA, K., OLSON, M. O. J. & BUSCH, H. (1979). Proteins C23 and B23are the major nucleolar silver staining proteins. Life Set. 25, 701-708.

MAMRACK, M. D., OLSON, M. O. J. & BUSCH, H. (1979). Amino acid sequence and sites ofphosphorylation in a highly acidic region of nucleolar non histone protein C23. Biochemistry18, 338i-3386.

MIKELSAAR, A. V. & SCHWARZACHER, H. G. (1978). Comparison of silver staining of nucleolusorganizer regions in human lymphocytes and fibroblasts. Hum. Genet. 42, 291-299.

MILLER, D. A., DEV, V. G., TANTRAVAHI, R. & MILLER, O. J. (1976). Suppression of humannucleolus organizer activity in mouse-human somatic hybrid cells. Expl Cell Res. 101,235-243-

MIRRE, C. & STAHL, A. (1978a). Infrastructure and activity of the nucleolar organizer in themouse oocyte during meiotic prophase. J. Cell Set. 31, 79-100.

MIRRE, C. & STAHL, A. (19786). Peripheral RNA synthesis of fibrillar center in nucleoli ofJapanese quail oocytes and somatic cells. J. Ultrastruct. Res. 64, 377-387.

OLERT, J., SAWATZKI, G., KLING, H. & GEBAUER, J. (1979). Cytological and histochemicalstudies on the mechanism of the selective silver staining of nucleolus organizer regions(NORs). Histochem. J. 60, 91-99.

PEBUSQUE, M.-J., ROBAGLIA, A. & SEITE, R. (1981). Diurnal rhythm of nucleolar volume insympathetic neurons of the rat superior cervical ganglion. Eur. J. Cell Biol. 24, 128—130.

PEBUSQUE, M.-J. & SEITE, R. (1980). Circadian change of fibrillar centers in nucleolus ofsympathetic neurons: an ultrastructural and stereological analysis. Biologie cell. 37, 219-222.

PEBUSQUE, M.-J., Vio, M. & SETTE, R. (1981). Ultrastructural location of Ag-NOR stainedproteins in nucleoli of rat sympathetic neurons during the dark period. Biologie cell. 40,I5I-I54-

SCHMIADY, H., MUNKE, M. & SPERLING, K. (1979). Ag-staining of nucleolus organizer regionson human prematurely condensed chromosomes from cells with different ribosomal RNAgene activity. Expl Cell Res. 121, 425-428.

SCHWARZACHER, H. G., MIKELSAAR, A. V. & SCHNEDL, W. (1978). The nature of the Ag-staining of nucleolus organizer regions: electron- and light-microscopic studies on humancells in interphase, mitosis and meiosis. Cytogenet. Cell Genet. 20, 24-39.

SEYTE, R., PEBUSQUE, M.-J., ROBAGLIA, A. & CAU, P. (1980). Stereological analysis of circadianrhythm of nucleolar fibrillar centers in neurons of the rat superior cervical ganglion. Eur.J. Cell Biol. 22, 499.

{Received 1 April 1981)