- Email: [email protected]
Three-dimensional organization of the endoplasmic reticulum in supraoptic neurons of the rat. A structural functional correlation
Brain Research, 170 (1979) 247-258 '(h Elsevier/North-Holland Biomedical Press
247
T H R E E - D I M E N S I O N A L O R G A N I Z A T I O N OF THE ENDOPLASMIC RETIC U L U M 1N SUPRAOPTIC NEURONS OF T H E RAT. A S T R U C T U R A L FUNCTIONAL CORRELATION
GI~RARD ALONSO and IVAN ASSENMACHER Laboratoire de Neuroendocrinologie, ERA 85 du CNRS, Universit~~ de Montpellier 11, Montpellier (France)
(Accepted November 9th, 1978)
SUMMARY Double impregnation staining of tissue was used to study on thick sections the three-dimensional configuration of the peripheral endoplasmic reticulum in neurosecretory neurons of the supraoptic nucleus in control and water-deprived rats. According to the spatial organization of the endoplasmic reticulum, two types of neurons are described in this report: Type I neurons which predominated in control rats (70~) exhibited characteristically large lamellar structures connected to loosely anastomosed tubules. In type I1 neurons, which prevailed in water-deprived rats (85 o/) the endoplasmic reticulum had the appearance of a highly-developed network of interconnected tubules, with no lamellar structures. Double impregnation staining combined with high resolution radioautography after [~H]leucine administration showed that the tubular configuration of the endoplasmic reticulum was the main active site of protein synthesis by contrast with the lamellar components, whose activity seemed poor. In terms of protein synthesis, the three-dimensional configuration of the peripheral endoplasmic reticulum of the supraoptic neurons appeared therefore to be closely connected with their functional state.
INTRODUCTION The ultrastructural aspects of the magnocellular neurosecretory neurons of the supraoptic nucleus(SON)have been extensively studied in the ratS, u,12,14,17,24,'~6,30,36. Experimental studies in osmotically stimulated animals have shown that in these neurons the more drastic changes in intracellular organization were related to the highly developed system of rough endoplasmic reticulum that forms the Ni,~sl substance at the periphery of the perikarya3,1~,~5,a6,19,27,28,32,33,3~. The present work was designed to provide insight into the spatial organization of the endoplasmic reticulum (ER) in relation to its functional significance in the neuro-
248 secretory neurons of the SON. In a first series of investigations using selective staining of the intracellular membranous system 3° thick sections were studied by electron microscopy in order to visualize the three-dimensional configuration of the ER in the supraoptic neurons of normal and water deprived rats. In a second set o f experiments, this particular staining technique was combined with high resolution radioautography on thick sections after administration of tritiated amino acids in order to explore the possible functional significance of the ER morphological configuration. MATERIAL AND METHODS
Animals The animals used were adult male Wistar Rats (200-250 g) housed in temperature (21 i 1 °C) and lighting (I2L-12D) controlled rooms. Control rats had free access to dry pellets of standard food (Provimi, Paris) and to drinking water. For the osmotically stimulated rats, the water supply was removed 5 days before the sacrifice.
Morphological study Five normal and 5 water-deprived rats were perfused with a 2 ~o glutaraldehyde solution in 0.1 M cacodylate buffer whose pH was adjusted to 7.5. Both SON of each animal were rapidly dissected and immersed in fixative for 1 h. The tissue blocks were then transferred to a 5 ~ uranyl acetate solution in distilled water for 1 h at room temperature. After rinsing in distilled water they were poststained in double lead and copper citrate for 1 h at room temperature (for details see ref. 31). Following several rinses in distilled water, the tissue blocks were postfixed overnight in t ~ osmium tetroxide at 4 °C. Finally they were dehydrated and embedded in Epon. Ultra-thin or 0.5-1 /~m thick sections were examined at 100 kV in a JEOL 100 B or Philips 30l electron microscope. Stereoscopic pairs were obtained by tilting the section from ~ 5° to ~5 ° in the goniometric stage of the Philips 301 electron microscope. For the morphological analysis of the supraoptic neurons, 1 ~m thick sections of the SON were randomly photographed at a constant magnification of 2000. These photographs were taken in order to centre individual neuronal perikarya. To minimize the effect of regional variations within the nuclei, thick sections were successively cut from the anterior, medial and posterior parts of the SON. Lastly, 620 photographs were analyzed on 3 ~ prints at a final magnification of 6000.
Radioautographic study Six normal and 6 water-deprived rats were stereotaxically injected, unilaterally in the supraoptic region with 0,5 mCi [aH]leucine in 2/zl saline (concentration of the injected leucine: 25 × 10-4 M). Thirty minutes, 3 h and 8 h after the tracer injection two rats from each group were sacrificed by intracardiac perfusion of 2 ~o glutaraldehyde in 0.1 M cacodylate buffer (pH 7.5) containing 0.5 M 'cold' leucine to avoid artefactual retention of free [3H]leucine by the glutaraldehyde fixative4,10,24. Fixation lasted for 1 h and SON tissue blocks were stained as described above. Sections (0.5/~mthick) were prepared for electron microscope radioautography by dipping into ttford L4 emulsion.
249
Fig. 1. Type 1 n e u r o n s - - 0.5 i t m thick sections. A : The peripheral e n d o p l a s m i c reticulum a p p e a r s as poorly a n a s t o m o s e d tubules (upper part of picture) or as large flattened sheets (lower part) perforated by pores on their periphery. In the perinuclear zone the tubular elements of the e n d o p l a s m i c reticu[um are connected to the nuclear m e m b r a n e (arrows). N, nucleus ; M i, mitochondria ; G, Golgi complex; L, lysosomes. , 9000, B a n d C: Stereoscopic pair. Tiltage angle I 5 to 5 . The unfenestrated flattened sheets o f e n d o p l a s m i c reticulum are parallel to the plasma m e m b r a n e delimiting the n e u r o n (pro) a n d are connected to each other by elongated e n d s . . 9000.
250
Fig. 2. Type 1I neurons 0.5/~m thick sections. A : The peripheral endoplasmic reticulum appears as a highly developed network of interconnected tubular elements with numerous polygonal meshes. Mi, mitochondria; G, Golgi complex..." 9500. B and C: Stereoscopic pair. Tiltage angle * 5 to - - 5 ~. Small sheets of endoplasmic reticulum are perforated by many pores giving rise to lace-like structures (arrows) which constitute extensions to the tubular network. G, Golgi complex. - J 0,000.
251 TABLE 1
Percentage o f Type 1 and Type H neurons in the S O N of control and water-deprived rats R controls vs water-deprived. Type I neurons
Type l I neurons R -
Controls Water-deprived
68 ~ 15~
32 ~o 85~
Type I Type 1I
2.12 0.18"*
** P ~ 0.01 vs controls. After 15-25 days exposure they were developed with microdol X and examined in a JEOL 100 B electron microscope. R E S U LTS
Morphological stud), As a result of the technique used in this work, the intracytoplasmic membranous system (nuclear envelope, Golgi apparatus, endoplasmic reticulum and mitochondria) was intensely stained whereas the ribosomes, neurosecretory granules, microtubules and microfilaments remained unstained. Staining of lysosomes varied and they frequently appeared as homogenous dense structures. Two morphological zones were distinguishable in the perikaryon of most neurons: (a) a perinuclear zone, including the Golgi complexes, generally somewhat heterogenous in appearance, and (b) a peripheral zone, essentially containing a well-developed endoplasmic reticulum system. In thick sections of the SON of both normal and water deprived rats, the peripheral ER appeared to be constituted of tubular elements and lamel[ar structures. Stereoscopic study of these thick sections showed very characteristic spatial arrangement of the peripheral ER within a given perikaryon, so that two types of neurons could generally be distinguished. They were tentatively described as Type I and Type |I neurons.
Type I neurons In these neurons the peripheral ER system comprised large lamellar structures connected by their elongated ends to loosely anastomosed tubular elements (Fig. I A). These structures were mainly located in the outermost regions of the perikarya and were frequently arranged as large parallel sheets extending into the neuroplasma (Fig. 1B, C). These ER sheets were sometimes perforated by pores, 100-150 nm in diameter, occurring mainly at the periphery of the membranous structure. Type 11 neurons In these neurons, the ER essentially appeared as a highly developed tubular
252
Fig. 3. Radioautography of Type I neurons. A : 30 min after [3H]leucine injection, numerous silver grains are superimposed on the elongated tubules whereas a small unfenestrated sheet is only slightly labelled (arrow). G, Golgi complex, z 12,500. B: 3 h after [ZH]leucine injection, no silver grains are observed on the lamellar structures of the peripheral ER (*) whereas tubular elements are intensely labelled. ~ 10,000.
253
Fig. 4. Radioautography of a Type 1 neuron 3 h after [~H]leucine injection. Intense labelling is observed over the tubular elements of the ER. A few silver grains are located at the periphery of the large sheets of ER which appear to be perforated by many pores, whereas no labelling is visible on the surface of unfenestrated sheets (arrow). , 8500. system. These tubules, which in some neurons looked slightly dilated, always formed a continuous three-dimensional network with polygonal meshes (Fig. 2A). The large lamellar structures prevailing in Type I neurons were not visible in Type II, In connection with the tubular network only small perforated sheets were observed, some of them resembling lace-like membranous structures (Fig. 2B, C). Both types of neurons were generally visible in all SON of control and waterdeprived rats. However, very few neurons were found exhibiting the characteristics of both types in different regions of their perikaryon. As Table I clearly shows, Type I neurons predominated in the SON of control rats, and Type 11 in that of waterdeprived animals.
Racfioautographic study In light microscopic radioautography intense labelling over all SON magnocellular neurons was observed in both control and water-deprived rats at all time intervals after injection of [3H]leucine. As early as 30 rain after the injection, the silver grains were very abundant over the cytoplasm and nucleus of the neurons. After longer intervals, labelling gradually extended over the axonal processes located in the ventral part of the SON. High resolution radioautography showed that the general pattern of labelling did not differ significantly in control and water-deprived rats, since in both groups the
254
Fig. 5. Radioautography of a Type 11 neuron. 30 min after [3H]leucineinjection, silver grains are homogeneously distributed over the tubular network of the peripheral endoplasmic reticulunl. ~: 9000. entire intracellular membrane system appeared to be labelled. However, at all times after [ZH]leucine administration, the most intense labelling was localized over the peripheral ER system. Careful examination of the radioautographs revealed that in Type I and Type I1 neurons the silver grains were essentially superimposed on the tubular elements of the ER (Figs. 3-5). By contrast, the large sheets of ER observed in Type I neurons were usually unlabelled (Fig. 3). However, a few scattered silver grains were sometimes visible at the periphery of some sheets, especially those perforated by many pores (Fig. 4). In Type I1 neurons, the small perforated sheets were, as a rule, also labelled. DISCUSSION The chemical basis of the double-impregnation staining technique has not yet been elucidated. As a result of the pH of the gtutaraldehyde fixative and uranyt stain used in this work selective impregnation of the intracytoplasmic membranes could be obtained. However, as noted by Thierry and Rambourg 31 the low pH (3.5) of the uranyl acetate solution might lead to shrinkage of membranous celt structures, and it is true that the diameter of both the smooth endoplasmic reticulum and microvesicles of axonal endings was generally smaller after double impregnation of the tissue than: after the use of more conventional techniques 1'18. It is also true that double-impregnation staining did not reveal neurotubules, microfilaments, ribosomes or neuroplasm,
255 which means caution must be exercised when comparing the present morphological descriptions with ultrastructural data obtained by classical electron microscopic techniques. Nevertheless, the general organization of the intracellular membranous system did not appear markedly altered in the present study, and on the whole, the classical pattern observed in neurosecretory neurons was reproduced. Thus, the importance of the double-impregnation staining technique lies in the fact that the ER stood out in sharp contrast in the cellular matrix and therefore is clearly visible at 100 kV even in 0.5-1 /tin thick sections. The present study of the supraoptic neurons of control and water-deprived rats made it possible to distinguish two main three-dimensional configurations of the peripheral endoplasmic reticulum. Type I neurons in which a loosely arranged tubular system was connected to large lamellar structures that appeared to be the characteristic feature of this neuronal type (Fig. I) and Type il neurons, whose ER consisted essentially of a highly developed network of tubular element (Fig. 2). Statistical comparison of the distribution of each cell type within the SON of control and waterdeprived rats clearly shows a drastic increase in Type II neurons in the latter group. This result appears in good agreement with previous data 3,12,1~.1!~,a~,'33,3'~, indicating that in supraoptic neurons, osmotic stimulation induces changes in the peripheral ER leading to a more developed and interconnected system of so called 'cisternae'. However, these earlier studies using classical staining techniques on ultra-thin sections generally concluded from the examination of several cytological parameters in a marked heterogeneity of SON neurons supposedly resulting from a phasic activity of these neurons2,~, 3~. By contrast, the present observations that almost all supraoptic neurons displayed a similar structural configuration of their ER after water-deprivation appear in better agreement with results of electrophysiological studies showing that in osmotically stimulated animals, a majority of supraoptic neurons exhibited electrical patterns of activated secretion2, 34. This leads to the tentative conclusion that Type Ii neurons correspond to highly stimulated neurons. The absence of lamellar ER within stimulated neurons where protein synthesis is known to be highly enhanced ')1,22 led us to investigate by high resolution radioautography, the possible functional significance of the different ER structures obselved in thick sections. Although very thin sections are known to provide the best resolution in radioautography ~9, it was recently shown that self-absorption of/J-particles in tissue after metallic impregnation may to some extent compensate for the lack of resolution resulting from the thickness of the section 13.1s. Since no diffuse radioautographic reaction was detectable in our material (Figs. 3 and 5), it may be assumed, in accordance with previous studies 6,20 that most of the radioactive molecules remaining in the neurons belonged to proteins. It is generally accepted that radioautographic reactions detected over the peripheral ER of neurons shortly after [all]amino acid administration visualize newly synthetized proteins at their production sites 7,8.9. Consequently, the most salient feature of the present radioautographic study was that in all SON neurons from both control and water-deprived rats, the silver grains were essentially superimposed on the tubular components of the peripheral ER at all times after
256 tritiated precursor administration, whereas the lamellar structure of the Type ! neurons showed little or no labelling. As the large sheets of ER were not found to be labelled, it could be argued that radioactive proteins may have migrated from these structures at the time of sacrifice. However, under the present conditions, in which high doses oftritiated precursor were injected locally, persistent and very intense labelling of the supraoptlc neurons was observed over 8 h after the injection, even in the water-deprived rats where the proximodistal migration of synthetized proteins is known to be highly enhanced 23. This seems to indicate that the exogenous precursor remains available for continued uptake and incorporation throughout this period. It therefore seems very likely that the unlabelled lamellar ER structures can be considered as sites of reduced protein synthesis as opposed to the active tubular structures. The results of the radioautographic study indicating that the lametlar ER structures may be poorly active as protein synthesis sites, as well as the previous morphological observations concerning the predominance of the tubular vs lamellar structures during osmotic stress, enable us to conclude that in stimulated supraoptic neurons, enhanced protein synthesis might be associated with expansion of the tubular components of the peripheral ER at the expense of the lamellar structures. Some arguments do favour the idea of a gradual transformation of the inactive sheets of Type I neurons, into the active tubules of Type I!. In our experiments, the perforated periphery of some ER sheets of Type 1 neurons, as well as the small lace-like membranous structures of the Type !I neurons were found to be slightly labelled after [ZH]leucine administration t Fig. 4j indicating a moderate level of protein synthesis. On the other hand, stereoscopic observation of the perforated structures peculiar to Type I1 neurons (Fig. 2B. C) strongly suggests that the structural transformation may be brought about by an increasing number of pores perforating the ER sheets, thus producing the lace-like ER structures which very often closely resemble the surrounding polygonal network of tubular elements. In conclusion, the three-dimensional organization of the peripheral ER of neurosecretory SON neurons appears closely related to their functional state. In this connection, the relative abundance within the peripheral ER system of the tubular (active) and lamellar linactive) configuration might provide some indication of the neuron's functional activity. Further investigations will be needed to determine whether similar morphofunctional correlations can be extended to other types of neurons or even to other cell types. ACKNOWLEDGEM ENTS We are grateful to Professor B. Droz and D r A. Rambourg of the Department of Biology, CEA. Saclay, France. for their valuable comments and advice in connection with the present work.
257 REFERENCES I Alonso, G. and Assenmacher, I., The smooth endoplasmic reticulum in neurosecretory axons of the rat neurohypophysis, J. Cell Biol., 32 (1978) 205 208. 2 Arnauld, E., Dufy, B. and Vincent, J. D., Hypothalamic supraoptic neurones : rates and patterns of action potential firing during water deprivation in the unanesthetized monkey, Brain Re,search, 100 (1975) 315 325. 3 Bara, D., Bartok, I. und Csapo, Z. S., Beitriige zur Ultrastruktur der hypothalamischen Neurosekretion, Endokrinologie, 53 (1968) 385 396. 4 Bergeron, M. et Droz, B., Analyse critique des conditions de fixation et de preparation des tissus pour la detection radioautographique des protdines neoformdes en microscopie 61ectronique, J. Microscopie, 7 (1968) 51 62. 5 Cotte, G. et Picard, D., Etude ultrastructurale des neurones du noyau supraoptique du rat, Bull. Ass. Anat., 141 (1968) 738 747. 6 Droz, M. and Warshawsky, Reliability of the radioautographic technique for the detection of newly synthetized protein, J. Histoehem. Cytochem., 11 (1963) 426-435. 7 Drox, B., Fate of newly synthetized proteins in neurons. In The Use ~/" Radioautography ./br Im,est(eatioH ~1 ProteiH Synthesis, Academic Press, New York, 1965, pp. 159 175. 8 Droz, B., Synthese et transfert des protdines cellulaires dans les neurones ganglionnaires. Etude radioautographique en microscopie 61ectronique, J. Mieroscopie, 6 (1967) 201 208. 9 Droz, B., Kinetic analysis of macromolecular movements in cells by means of radieautography, J. Microsc. Biol. Cell., 27 (1976) 191 200. 10 Droz, B. and Simon, C., Radieautographic visualization of newly synthetized proteins: a critical appraisal, J. Microsc. Biol. Cell., 27 (1976) 203 206. II Eichner, D. and Theman, H., Elektronenmikroskopische Untersuchungen am Nucleus supraopticus hypothalami der Albino-Ratte, Z. Zel[/brsch., 74 (1966) 242 251. 12 Enestr6m, S., Nucleus supraopticus, a morphological and experimental study in the rat, Acta path. microhiol, stand., Suppl. 186 (1967) I 99. 13 Gupta, B. L., Moreton, R. B. and Cooper, N. C., Reconsideration of resolution in the ME autoradiography using biological line source, J. Microse., 99 (1973) I 25. 14 Kalimo, H., UItrastructural studies on the hypothalamic neurosecretory neurons of the rat. 11. The hypothalamo-neurohypophysial system in rats vdth hereditary hypothalamic diabetes insipidus, Z. Zell/orsch., 134 (1972) 205-225. 15 Kalimo, H., UItrastructural studies on the hypothalamic neurosecretory neurons of the rat. 111. Paraventricular and supraoptic neurons during lactation and dehydration, Cell Tiss. Reds., 163 (1975) 151 168. 16 Klein, M. J., Porte, A. et Stutinsky, E., Comparaison ultrasiructurale des noyaux ncurosdcretoires hypothalamiques chez le Rat normal ou en 6tat de surcharge osmotique, Bull. Ass. Anat., 142 (1969) 1066 1072. 17 Lafarga, M., Palacios, G. and Perez, R., Morphological aspects of the functional synchronization of supraoptic nucleus neurons, Experientia (Basel/, 31 (1975) 348-349. 18 Markov, D., Rambourg, A. and Droz, B., Smooth endoplasmic reticulum and fast axonal transport of glycoproteins, an electron microscope radioautographic study of thick sections alter heavy metals impregnation, J. Microsc. Biol. Cell., 25 (1976) 57 60. 19 Morris, J. F., Dyball, R. E. J., A quantitative study of the ultrastructural changes in the hypotbalamo-neurohypophysial system during and after exI3erimentally induced hypersecretion, Cell. Tiss. Res., 149 fl974) 525 535. 20 Monncron, A. and Moule, Y., Critical evaluation of specificity in electron microscopical radioautography in animal tissues, Exp. Cell. Res., 56 (1969) t79 193. 21 Norstrg, m, A., A functional study of the hypothalamo-neurohypophysial system of the rat, w,ith the use of a ne,a.ly developed method for localized administration of labelled precursor, Brain Resem'eh, 28 (1971) 131 142. 22 NorstrOm, A., Enerstr6m, S. and Hamberger, A., Amino acid incorporation into proteins of the supraoptic nucleus of the rat after osmotic stress, Brain Research, 26 (1971) 95 103. 23 Norstr/3m, A. and Sj6strand, J., Effect of salt loading, thirst and w,ater-loading on transport and turnover of" neurohypophysial proteins of the rat, J. Endocr., 52 (1972) 87 105. 24 Peters, T. Jr., Asley, C. A., An artefact in radioautography due to binding of free amino acids to tissue by fixation, J. Cell. Biol., 33 (1967) 53 60. 25 Picard, D. La synchronisation fonctionnelle des cellules neuros6cr6trices du noyau supraoptique
258 du rat deshydrat6 et rehydrat6; 6tude ultrastructurale, C.R. Acad. Sci. {Paris), 271 (1970} 847-850. 26 Pilgrim, C., Morphologische und funktionnelle Untersuchungen zur Neurosekretbildung. Enzymhistochemische, autoradiographische und elektronenmikroskopische Beobachtungen an Ratten unter osmotiseher Belastung, Ergebn. Anat. EntwickL Gesch., 41 (t969) 1--79. 27 Rechardt, L,, Electron microscopic and histochemical observations on the supraoptic nucleus of normal and dehydrated rats, Acta physiol, scand, Suppl., 329 (1969) 1-79. 28 Reinhardt, H. F., Henning, L. C. und Rohr, H. P., Morphometrisch-ultrastruktureUe Untersuchungen am Nucleus supraopticus der Ratte nach Dehydration, Z. Zellfors~'h., 102 (1969) 172-181. 29 Rogers, A. W.. Techniques of Autoradiography, 2nd Ed., Elsevier, Amsterdam, 1973. 30 Sloper, J. C. and Bateson, R. G.. Ultrastructure of neurosecretory cells in the supraoptic nucleus of the dog and rat, J. Endocr., 31 (1966) 139-150. 31 Thiery, G. and Rambourg, A., A new staining technique for studying thick sections in electron microscope, J. Micros. Biol. Cell., 26 (1976) 103-106. 32 Tweedle, C. D. and Hatton, G. I.. Ultrastructural comparisons of neurons of supraoptic and circularis nuclei in normal and dehydrated rats, Brain Res. Bull., 1 (1976) 103-121 33 Tweedle, C. D., Hatton, G. I., Ultrastructural changes in rat hypothalamic neurosecretory cells and their associated glia during minimal dehydration and rehydration, Cell Tiss. Res., 181 (19777 50-72. 34 Waiters, J. K. and Hatton, G. I., Supraoptic neuronal activity in rats during five days of water deprivation, Physiol. Behav.. 13 (I974) 661-667. 35 Zambrano, D. and De Robertis, E., The secretory cycle of supraoptic neurons tn the rat. A structural functional correlation, Z. Zellforsch., 73 (1966) 414-431. 36 Zambrano, D., Mordoh, J.. Neurosecretory activity in supraoptic nucleus of normal rats, Z. Zellforsch., 73 (1966) 405~113.
247
T H R E E - D I M E N S I O N A L O R G A N I Z A T I O N OF THE ENDOPLASMIC RETIC U L U M 1N SUPRAOPTIC NEURONS OF T H E RAT. A S T R U C T U R A L FUNCTIONAL CORRELATION
GI~RARD ALONSO and IVAN ASSENMACHER Laboratoire de Neuroendocrinologie, ERA 85 du CNRS, Universit~~ de Montpellier 11, Montpellier (France)
(Accepted November 9th, 1978)
SUMMARY Double impregnation staining of tissue was used to study on thick sections the three-dimensional configuration of the peripheral endoplasmic reticulum in neurosecretory neurons of the supraoptic nucleus in control and water-deprived rats. According to the spatial organization of the endoplasmic reticulum, two types of neurons are described in this report: Type I neurons which predominated in control rats (70~) exhibited characteristically large lamellar structures connected to loosely anastomosed tubules. In type I1 neurons, which prevailed in water-deprived rats (85 o/) the endoplasmic reticulum had the appearance of a highly-developed network of interconnected tubules, with no lamellar structures. Double impregnation staining combined with high resolution radioautography after [~H]leucine administration showed that the tubular configuration of the endoplasmic reticulum was the main active site of protein synthesis by contrast with the lamellar components, whose activity seemed poor. In terms of protein synthesis, the three-dimensional configuration of the peripheral endoplasmic reticulum of the supraoptic neurons appeared therefore to be closely connected with their functional state.
INTRODUCTION The ultrastructural aspects of the magnocellular neurosecretory neurons of the supraoptic nucleus(SON)have been extensively studied in the ratS, u,12,14,17,24,'~6,30,36. Experimental studies in osmotically stimulated animals have shown that in these neurons the more drastic changes in intracellular organization were related to the highly developed system of rough endoplasmic reticulum that forms the Ni,~sl substance at the periphery of the perikarya3,1~,~5,a6,19,27,28,32,33,3~. The present work was designed to provide insight into the spatial organization of the endoplasmic reticulum (ER) in relation to its functional significance in the neuro-
248 secretory neurons of the SON. In a first series of investigations using selective staining of the intracellular membranous system 3° thick sections were studied by electron microscopy in order to visualize the three-dimensional configuration of the ER in the supraoptic neurons of normal and water deprived rats. In a second set o f experiments, this particular staining technique was combined with high resolution radioautography on thick sections after administration of tritiated amino acids in order to explore the possible functional significance of the ER morphological configuration. MATERIAL AND METHODS
Animals The animals used were adult male Wistar Rats (200-250 g) housed in temperature (21 i 1 °C) and lighting (I2L-12D) controlled rooms. Control rats had free access to dry pellets of standard food (Provimi, Paris) and to drinking water. For the osmotically stimulated rats, the water supply was removed 5 days before the sacrifice.
Morphological study Five normal and 5 water-deprived rats were perfused with a 2 ~o glutaraldehyde solution in 0.1 M cacodylate buffer whose pH was adjusted to 7.5. Both SON of each animal were rapidly dissected and immersed in fixative for 1 h. The tissue blocks were then transferred to a 5 ~ uranyl acetate solution in distilled water for 1 h at room temperature. After rinsing in distilled water they were poststained in double lead and copper citrate for 1 h at room temperature (for details see ref. 31). Following several rinses in distilled water, the tissue blocks were postfixed overnight in t ~ osmium tetroxide at 4 °C. Finally they were dehydrated and embedded in Epon. Ultra-thin or 0.5-1 /~m thick sections were examined at 100 kV in a JEOL 100 B or Philips 30l electron microscope. Stereoscopic pairs were obtained by tilting the section from ~ 5° to ~5 ° in the goniometric stage of the Philips 301 electron microscope. For the morphological analysis of the supraoptic neurons, 1 ~m thick sections of the SON were randomly photographed at a constant magnification of 2000. These photographs were taken in order to centre individual neuronal perikarya. To minimize the effect of regional variations within the nuclei, thick sections were successively cut from the anterior, medial and posterior parts of the SON. Lastly, 620 photographs were analyzed on 3 ~ prints at a final magnification of 6000.
Radioautographic study Six normal and 6 water-deprived rats were stereotaxically injected, unilaterally in the supraoptic region with 0,5 mCi [aH]leucine in 2/zl saline (concentration of the injected leucine: 25 × 10-4 M). Thirty minutes, 3 h and 8 h after the tracer injection two rats from each group were sacrificed by intracardiac perfusion of 2 ~o glutaraldehyde in 0.1 M cacodylate buffer (pH 7.5) containing 0.5 M 'cold' leucine to avoid artefactual retention of free [3H]leucine by the glutaraldehyde fixative4,10,24. Fixation lasted for 1 h and SON tissue blocks were stained as described above. Sections (0.5/~mthick) were prepared for electron microscope radioautography by dipping into ttford L4 emulsion.
249
Fig. 1. Type 1 n e u r o n s - - 0.5 i t m thick sections. A : The peripheral e n d o p l a s m i c reticulum a p p e a r s as poorly a n a s t o m o s e d tubules (upper part of picture) or as large flattened sheets (lower part) perforated by pores on their periphery. In the perinuclear zone the tubular elements of the e n d o p l a s m i c reticu[um are connected to the nuclear m e m b r a n e (arrows). N, nucleus ; M i, mitochondria ; G, Golgi complex; L, lysosomes. , 9000, B a n d C: Stereoscopic pair. Tiltage angle I 5 to 5 . The unfenestrated flattened sheets o f e n d o p l a s m i c reticulum are parallel to the plasma m e m b r a n e delimiting the n e u r o n (pro) a n d are connected to each other by elongated e n d s . . 9000.
250
Fig. 2. Type 1I neurons 0.5/~m thick sections. A : The peripheral endoplasmic reticulum appears as a highly developed network of interconnected tubular elements with numerous polygonal meshes. Mi, mitochondria; G, Golgi complex..." 9500. B and C: Stereoscopic pair. Tiltage angle * 5 to - - 5 ~. Small sheets of endoplasmic reticulum are perforated by many pores giving rise to lace-like structures (arrows) which constitute extensions to the tubular network. G, Golgi complex. - J 0,000.
251 TABLE 1
Percentage o f Type 1 and Type H neurons in the S O N of control and water-deprived rats R controls vs water-deprived. Type I neurons
Type l I neurons R -
Controls Water-deprived
68 ~ 15~
32 ~o 85~
Type I Type 1I
2.12 0.18"*
** P ~ 0.01 vs controls. After 15-25 days exposure they were developed with microdol X and examined in a JEOL 100 B electron microscope. R E S U LTS
Morphological stud), As a result of the technique used in this work, the intracytoplasmic membranous system (nuclear envelope, Golgi apparatus, endoplasmic reticulum and mitochondria) was intensely stained whereas the ribosomes, neurosecretory granules, microtubules and microfilaments remained unstained. Staining of lysosomes varied and they frequently appeared as homogenous dense structures. Two morphological zones were distinguishable in the perikaryon of most neurons: (a) a perinuclear zone, including the Golgi complexes, generally somewhat heterogenous in appearance, and (b) a peripheral zone, essentially containing a well-developed endoplasmic reticulum system. In thick sections of the SON of both normal and water deprived rats, the peripheral ER appeared to be constituted of tubular elements and lamel[ar structures. Stereoscopic study of these thick sections showed very characteristic spatial arrangement of the peripheral ER within a given perikaryon, so that two types of neurons could generally be distinguished. They were tentatively described as Type I and Type |I neurons.
Type I neurons In these neurons the peripheral ER system comprised large lamellar structures connected by their elongated ends to loosely anastomosed tubular elements (Fig. I A). These structures were mainly located in the outermost regions of the perikarya and were frequently arranged as large parallel sheets extending into the neuroplasma (Fig. 1B, C). These ER sheets were sometimes perforated by pores, 100-150 nm in diameter, occurring mainly at the periphery of the membranous structure. Type 11 neurons In these neurons, the ER essentially appeared as a highly developed tubular
252
Fig. 3. Radioautography of Type I neurons. A : 30 min after [3H]leucine injection, numerous silver grains are superimposed on the elongated tubules whereas a small unfenestrated sheet is only slightly labelled (arrow). G, Golgi complex, z 12,500. B: 3 h after [ZH]leucine injection, no silver grains are observed on the lamellar structures of the peripheral ER (*) whereas tubular elements are intensely labelled. ~ 10,000.
253
Fig. 4. Radioautography of a Type 1 neuron 3 h after [~H]leucine injection. Intense labelling is observed over the tubular elements of the ER. A few silver grains are located at the periphery of the large sheets of ER which appear to be perforated by many pores, whereas no labelling is visible on the surface of unfenestrated sheets (arrow). , 8500. system. These tubules, which in some neurons looked slightly dilated, always formed a continuous three-dimensional network with polygonal meshes (Fig. 2A). The large lamellar structures prevailing in Type I neurons were not visible in Type II, In connection with the tubular network only small perforated sheets were observed, some of them resembling lace-like membranous structures (Fig. 2B, C). Both types of neurons were generally visible in all SON of control and waterdeprived rats. However, very few neurons were found exhibiting the characteristics of both types in different regions of their perikaryon. As Table I clearly shows, Type I neurons predominated in the SON of control rats, and Type 11 in that of waterdeprived animals.
Racfioautographic study In light microscopic radioautography intense labelling over all SON magnocellular neurons was observed in both control and water-deprived rats at all time intervals after injection of [3H]leucine. As early as 30 rain after the injection, the silver grains were very abundant over the cytoplasm and nucleus of the neurons. After longer intervals, labelling gradually extended over the axonal processes located in the ventral part of the SON. High resolution radioautography showed that the general pattern of labelling did not differ significantly in control and water-deprived rats, since in both groups the
254
Fig. 5. Radioautography of a Type 11 neuron. 30 min after [3H]leucineinjection, silver grains are homogeneously distributed over the tubular network of the peripheral endoplasmic reticulunl. ~: 9000. entire intracellular membrane system appeared to be labelled. However, at all times after [ZH]leucine administration, the most intense labelling was localized over the peripheral ER system. Careful examination of the radioautographs revealed that in Type I and Type I1 neurons the silver grains were essentially superimposed on the tubular elements of the ER (Figs. 3-5). By contrast, the large sheets of ER observed in Type I neurons were usually unlabelled (Fig. 3). However, a few scattered silver grains were sometimes visible at the periphery of some sheets, especially those perforated by many pores (Fig. 4). In Type I1 neurons, the small perforated sheets were, as a rule, also labelled. DISCUSSION The chemical basis of the double-impregnation staining technique has not yet been elucidated. As a result of the pH of the gtutaraldehyde fixative and uranyt stain used in this work selective impregnation of the intracytoplasmic membranes could be obtained. However, as noted by Thierry and Rambourg 31 the low pH (3.5) of the uranyl acetate solution might lead to shrinkage of membranous celt structures, and it is true that the diameter of both the smooth endoplasmic reticulum and microvesicles of axonal endings was generally smaller after double impregnation of the tissue than: after the use of more conventional techniques 1'18. It is also true that double-impregnation staining did not reveal neurotubules, microfilaments, ribosomes or neuroplasm,
255 which means caution must be exercised when comparing the present morphological descriptions with ultrastructural data obtained by classical electron microscopic techniques. Nevertheless, the general organization of the intracellular membranous system did not appear markedly altered in the present study, and on the whole, the classical pattern observed in neurosecretory neurons was reproduced. Thus, the importance of the double-impregnation staining technique lies in the fact that the ER stood out in sharp contrast in the cellular matrix and therefore is clearly visible at 100 kV even in 0.5-1 /tin thick sections. The present study of the supraoptic neurons of control and water-deprived rats made it possible to distinguish two main three-dimensional configurations of the peripheral endoplasmic reticulum. Type I neurons in which a loosely arranged tubular system was connected to large lamellar structures that appeared to be the characteristic feature of this neuronal type (Fig. I) and Type il neurons, whose ER consisted essentially of a highly developed network of tubular element (Fig. 2). Statistical comparison of the distribution of each cell type within the SON of control and waterdeprived rats clearly shows a drastic increase in Type II neurons in the latter group. This result appears in good agreement with previous data 3,12,1~.1!~,a~,'33,3'~, indicating that in supraoptic neurons, osmotic stimulation induces changes in the peripheral ER leading to a more developed and interconnected system of so called 'cisternae'. However, these earlier studies using classical staining techniques on ultra-thin sections generally concluded from the examination of several cytological parameters in a marked heterogeneity of SON neurons supposedly resulting from a phasic activity of these neurons2,~, 3~. By contrast, the present observations that almost all supraoptic neurons displayed a similar structural configuration of their ER after water-deprivation appear in better agreement with results of electrophysiological studies showing that in osmotically stimulated animals, a majority of supraoptic neurons exhibited electrical patterns of activated secretion2, 34. This leads to the tentative conclusion that Type Ii neurons correspond to highly stimulated neurons. The absence of lamellar ER within stimulated neurons where protein synthesis is known to be highly enhanced ')1,22 led us to investigate by high resolution radioautography, the possible functional significance of the different ER structures obselved in thick sections. Although very thin sections are known to provide the best resolution in radioautography ~9, it was recently shown that self-absorption of/J-particles in tissue after metallic impregnation may to some extent compensate for the lack of resolution resulting from the thickness of the section 13.1s. Since no diffuse radioautographic reaction was detectable in our material (Figs. 3 and 5), it may be assumed, in accordance with previous studies 6,20 that most of the radioactive molecules remaining in the neurons belonged to proteins. It is generally accepted that radioautographic reactions detected over the peripheral ER of neurons shortly after [all]amino acid administration visualize newly synthetized proteins at their production sites 7,8.9. Consequently, the most salient feature of the present radioautographic study was that in all SON neurons from both control and water-deprived rats, the silver grains were essentially superimposed on the tubular components of the peripheral ER at all times after
256 tritiated precursor administration, whereas the lamellar structure of the Type ! neurons showed little or no labelling. As the large sheets of ER were not found to be labelled, it could be argued that radioactive proteins may have migrated from these structures at the time of sacrifice. However, under the present conditions, in which high doses oftritiated precursor were injected locally, persistent and very intense labelling of the supraoptlc neurons was observed over 8 h after the injection, even in the water-deprived rats where the proximodistal migration of synthetized proteins is known to be highly enhanced 23. This seems to indicate that the exogenous precursor remains available for continued uptake and incorporation throughout this period. It therefore seems very likely that the unlabelled lamellar ER structures can be considered as sites of reduced protein synthesis as opposed to the active tubular structures. The results of the radioautographic study indicating that the lametlar ER structures may be poorly active as protein synthesis sites, as well as the previous morphological observations concerning the predominance of the tubular vs lamellar structures during osmotic stress, enable us to conclude that in stimulated supraoptic neurons, enhanced protein synthesis might be associated with expansion of the tubular components of the peripheral ER at the expense of the lamellar structures. Some arguments do favour the idea of a gradual transformation of the inactive sheets of Type I neurons, into the active tubules of Type I!. In our experiments, the perforated periphery of some ER sheets of Type 1 neurons, as well as the small lace-like membranous structures of the Type !I neurons were found to be slightly labelled after [ZH]leucine administration t Fig. 4j indicating a moderate level of protein synthesis. On the other hand, stereoscopic observation of the perforated structures peculiar to Type I1 neurons (Fig. 2B. C) strongly suggests that the structural transformation may be brought about by an increasing number of pores perforating the ER sheets, thus producing the lace-like ER structures which very often closely resemble the surrounding polygonal network of tubular elements. In conclusion, the three-dimensional organization of the peripheral ER of neurosecretory SON neurons appears closely related to their functional state. In this connection, the relative abundance within the peripheral ER system of the tubular (active) and lamellar linactive) configuration might provide some indication of the neuron's functional activity. Further investigations will be needed to determine whether similar morphofunctional correlations can be extended to other types of neurons or even to other cell types. ACKNOWLEDGEM ENTS We are grateful to Professor B. Droz and D r A. Rambourg of the Department of Biology, CEA. Saclay, France. for their valuable comments and advice in connection with the present work.
257 REFERENCES I Alonso, G. and Assenmacher, I., The smooth endoplasmic reticulum in neurosecretory axons of the rat neurohypophysis, J. Cell Biol., 32 (1978) 205 208. 2 Arnauld, E., Dufy, B. and Vincent, J. D., Hypothalamic supraoptic neurones : rates and patterns of action potential firing during water deprivation in the unanesthetized monkey, Brain Re,search, 100 (1975) 315 325. 3 Bara, D., Bartok, I. und Csapo, Z. S., Beitriige zur Ultrastruktur der hypothalamischen Neurosekretion, Endokrinologie, 53 (1968) 385 396. 4 Bergeron, M. et Droz, B., Analyse critique des conditions de fixation et de preparation des tissus pour la detection radioautographique des protdines neoformdes en microscopie 61ectronique, J. Microscopie, 7 (1968) 51 62. 5 Cotte, G. et Picard, D., Etude ultrastructurale des neurones du noyau supraoptique du rat, Bull. Ass. Anat., 141 (1968) 738 747. 6 Droz, M. and Warshawsky, Reliability of the radioautographic technique for the detection of newly synthetized protein, J. Histoehem. Cytochem., 11 (1963) 426-435. 7 Drox, B., Fate of newly synthetized proteins in neurons. In The Use ~/" Radioautography ./br Im,est(eatioH ~1 ProteiH Synthesis, Academic Press, New York, 1965, pp. 159 175. 8 Droz, B., Synthese et transfert des protdines cellulaires dans les neurones ganglionnaires. Etude radioautographique en microscopie 61ectronique, J. Mieroscopie, 6 (1967) 201 208. 9 Droz, B., Kinetic analysis of macromolecular movements in cells by means of radieautography, J. Microsc. Biol. Cell., 27 (1976) 191 200. 10 Droz, B. and Simon, C., Radieautographic visualization of newly synthetized proteins: a critical appraisal, J. Microsc. Biol. Cell., 27 (1976) 203 206. II Eichner, D. and Theman, H., Elektronenmikroskopische Untersuchungen am Nucleus supraopticus hypothalami der Albino-Ratte, Z. Zel[/brsch., 74 (1966) 242 251. 12 Enestr6m, S., Nucleus supraopticus, a morphological and experimental study in the rat, Acta path. microhiol, stand., Suppl. 186 (1967) I 99. 13 Gupta, B. L., Moreton, R. B. and Cooper, N. C., Reconsideration of resolution in the ME autoradiography using biological line source, J. Microse., 99 (1973) I 25. 14 Kalimo, H., UItrastructural studies on the hypothalamic neurosecretory neurons of the rat. 11. The hypothalamo-neurohypophysial system in rats vdth hereditary hypothalamic diabetes insipidus, Z. Zell/orsch., 134 (1972) 205-225. 15 Kalimo, H., UItrastructural studies on the hypothalamic neurosecretory neurons of the rat. 111. Paraventricular and supraoptic neurons during lactation and dehydration, Cell Tiss. Reds., 163 (1975) 151 168. 16 Klein, M. J., Porte, A. et Stutinsky, E., Comparaison ultrasiructurale des noyaux ncurosdcretoires hypothalamiques chez le Rat normal ou en 6tat de surcharge osmotique, Bull. Ass. Anat., 142 (1969) 1066 1072. 17 Lafarga, M., Palacios, G. and Perez, R., Morphological aspects of the functional synchronization of supraoptic nucleus neurons, Experientia (Basel/, 31 (1975) 348-349. 18 Markov, D., Rambourg, A. and Droz, B., Smooth endoplasmic reticulum and fast axonal transport of glycoproteins, an electron microscope radioautographic study of thick sections alter heavy metals impregnation, J. Microsc. Biol. Cell., 25 (1976) 57 60. 19 Morris, J. F., Dyball, R. E. J., A quantitative study of the ultrastructural changes in the hypotbalamo-neurohypophysial system during and after exI3erimentally induced hypersecretion, Cell. Tiss. Res., 149 fl974) 525 535. 20 Monncron, A. and Moule, Y., Critical evaluation of specificity in electron microscopical radioautography in animal tissues, Exp. Cell. Res., 56 (1969) t79 193. 21 Norstrg, m, A., A functional study of the hypothalamo-neurohypophysial system of the rat, w,ith the use of a ne,a.ly developed method for localized administration of labelled precursor, Brain Resem'eh, 28 (1971) 131 142. 22 NorstrOm, A., Enerstr6m, S. and Hamberger, A., Amino acid incorporation into proteins of the supraoptic nucleus of the rat after osmotic stress, Brain Research, 26 (1971) 95 103. 23 Norstr/3m, A. and Sj6strand, J., Effect of salt loading, thirst and w,ater-loading on transport and turnover of" neurohypophysial proteins of the rat, J. Endocr., 52 (1972) 87 105. 24 Peters, T. Jr., Asley, C. A., An artefact in radioautography due to binding of free amino acids to tissue by fixation, J. Cell. Biol., 33 (1967) 53 60. 25 Picard, D. La synchronisation fonctionnelle des cellules neuros6cr6trices du noyau supraoptique
258 du rat deshydrat6 et rehydrat6; 6tude ultrastructurale, C.R. Acad. Sci. {Paris), 271 (1970} 847-850. 26 Pilgrim, C., Morphologische und funktionnelle Untersuchungen zur Neurosekretbildung. Enzymhistochemische, autoradiographische und elektronenmikroskopische Beobachtungen an Ratten unter osmotiseher Belastung, Ergebn. Anat. EntwickL Gesch., 41 (t969) 1--79. 27 Rechardt, L,, Electron microscopic and histochemical observations on the supraoptic nucleus of normal and dehydrated rats, Acta physiol, scand, Suppl., 329 (1969) 1-79. 28 Reinhardt, H. F., Henning, L. C. und Rohr, H. P., Morphometrisch-ultrastruktureUe Untersuchungen am Nucleus supraopticus der Ratte nach Dehydration, Z. Zellfors~'h., 102 (1969) 172-181. 29 Rogers, A. W.. Techniques of Autoradiography, 2nd Ed., Elsevier, Amsterdam, 1973. 30 Sloper, J. C. and Bateson, R. G.. Ultrastructure of neurosecretory cells in the supraoptic nucleus of the dog and rat, J. Endocr., 31 (1966) 139-150. 31 Thiery, G. and Rambourg, A., A new staining technique for studying thick sections in electron microscope, J. Micros. Biol. Cell., 26 (1976) 103-106. 32 Tweedle, C. D. and Hatton, G. I.. Ultrastructural comparisons of neurons of supraoptic and circularis nuclei in normal and dehydrated rats, Brain Res. Bull., 1 (1976) 103-121 33 Tweedle, C. D., Hatton, G. I., Ultrastructural changes in rat hypothalamic neurosecretory cells and their associated glia during minimal dehydration and rehydration, Cell Tiss. Res., 181 (19777 50-72. 34 Waiters, J. K. and Hatton, G. I., Supraoptic neuronal activity in rats during five days of water deprivation, Physiol. Behav.. 13 (I974) 661-667. 35 Zambrano, D. and De Robertis, E., The secretory cycle of supraoptic neurons tn the rat. A structural functional correlation, Z. Zellforsch., 73 (1966) 414-431. 36 Zambrano, D., Mordoh, J.. Neurosecretory activity in supraoptic nucleus of normal rats, Z. Zellforsch., 73 (1966) 405~113.