Neuroscience Vol. 28,No. 2,pp. 42341, 1989
0306-4522/89 $3.00f0.00
Printed in Great Britain
Pergamon Press plc 0 1989IBRO
IMMUNOHISTOCHEMICAL LOCALIZATION OF SECRETOGRANIN II IN THE RAT CEREBELLUM M. G. COZZI,* P. Ros~,*t A. GRECO,* A. HILLE,~ W. B. A. ZANINI*$ and P. DE CAMILLI*§
HUTTNER,t
*CNR Center of Cytopharmacology, Department of Pharmacology, University of Milan, Milan, Italy TEuropean Molecular Biology Laboratory, Postfach 102209, D-6900, Heidelberg, F.R.G. Ah&act-Secretogranin II (chromogranin C) is a peptide related to chromogranin A and secretogranin I (chromogranin B) which is secreted by a regulated pathway from both neurons and endocrine cells. In the present study we have determined by light microscopic immunocytochemistry its distribution in the cerebellum and in adjacent brain stem regions. Secretogranin II was found to be widely distributed throughout the gray matter of these regions. Highly immunoreactive structures in the cerebellar cortex included the majority of climbing fibers, a large number of mossy fibers, sparse varicose fibers in the molecular layer and a subpopulation of neuronal perikarya in the granule cell layer. The location and shape of these neurons are very similar to those of a novel type of cerebellar neurons which has been recently described. A moderate level of immunoreactivity was observed on fibers travelling among Purkinje cells and parallel to the pial surface in the Purkinje cell layer. A variable, but in general low, degree of immunoreactivity was also detectable in the perikarya of Purkinje cells. In the deep cerebellar nuclei a loose network of secretogranin II-positive fibers was visible. Neurons of the nuclei, however, were non-immunoreactive. A dense network of highly immunoreactive fibers was found throughout the brain stem regions adjacent to the cerebellum. Our results indicate that secretogranin II has in the cerebellum and adjacent regions a distribution more widespread than that of known regulatory peptides and suggest that the peptide-mediated signaling in the cerebellum plays a role more important than has been acknowledged so far.
Secretogranin II (SgII) (chromogranin C) is a secretory protein which was first described in the bovine anterior pituitary as a major sulphated component of pituitary gland secretory products.35~36~37 It was independently identified as one of the major tyrosinesulphated secretory proteins of rat pheochromocytoma cells (PC12 41~)~~ and later recognized to be the same as the pituitary protein.‘* SgII is an acidic protein, sulphated on tyrosine and carbohydrate residues and phosphorylated on serine and threonine residues.37v3*It is secreted by a wide variety of endocrine and nerve cells via a regulated pathway.37s38Its tissue distribution and its biochemical and physicochemical properties are similar to those of two other proteins, chromogranin A43 and secretogranin I (chromogranin B). 1,9,38For this reason, these three proteins are considered members of the same class of proteins. 9,38*43Although these three proteins are present in a large number of endocrine and neuronal cells, their tissue and cellular distribution is not always overlapping.“,25,34*38Their relative ratios vary in different tissues and cells, and at certain sites their cellular distribution is almost complementary.38 Several possible physiological functions of these pro-
$To whom correspondence should be addressed. §Present address: Department of Cell Biology, Yale University Medical School, 333 Cedar Street, New Haven, CT 06510, U.S.A. Abbreviations: LDCVs, large dense-core vesicles; SgII, secretogranin II.
teins have been hypothesized. Before secretion, they might play a role in the sorting, processing or condensation of material destined to exocytosis via the regulated pathway.‘6”8 Following their secretion by exocytosis, these proteins, or peptides derived from them by proteolytic processing, might act as peptide hormones or neurotransmitters.8*‘o~‘s~‘6*37~38 Their possible role as peptide precursors, first suggested by the homology between chromogranin A and the regulatory peptide pancreastatin,8J5 has recently gained support from data showing that pancreastatin is indeed derived from chromogranin A.22 Like well established peptides, hormones and neurotransmitters, proteins of the secretogranim chromogranin family are stored in secretory granules of endocrine cells’2,37*43 and in large dense-core vesicles (LDCVs) of neurons. 7*29,30*41 However, both in the endocrine and in the nervous system they appear to have a much more widespread distribution than that of any known regulatory peptide.37.38v4’Thus, mapping the distribution of these proteins in the nervous system not only may be of help to clarify their physiological function, but also may be a way to visualize peptidergic innervation in areas where the identity of peptide neurotransmitters is still unknown. In this study we have endeavored to determine by immunohistochemistry the distribution of SgII in the rat cerebellum and in adjacent brain stem regions. Until recently little information was available on the presence of well-established peptides
423
in
the
have
cerebel]um,2.‘“.“.‘~,~i.411.42
reports
appeared
Only very recently on the presence of a few
peptides,','8.",2'.24,2X,~~
EXPERIMENTAL PROCEDURES Anti-secretogranin
II antibodies
The antiserum directed against rat Sgll, raised in rabbits, was the same as that described in Ref. 38. Antibodies were affinity-purified” by use of SgII partially purified from rat anterior pituitaries’.‘” and coupled to Sepharose 4B. The specificity of these antibodies has been thoroughly characterized in a previous study.?x
Preparation of’ tissue sections. SpragueeDawley male albino rats, 175--250 g, from Charles River (Italy), were anesthetized with an intraperitoneal injection of chloral hydrate (500mg/kg) and fixed by transcardiai perfusion with 4% formaldehyde (freshly prepared from paraformaldehyde) as described.5 At the end of the perfusion blocks of the cerebellum with the attached brain stem were immersed in the same fixative for an additional 3 h and subsequently, after a wash in 120 mM Na-phosphate buffer, pH 7.4 (several hours), infiltrated with 0.5 M sucrose in the
same buffer. Blocks were then frozen in isopentane chilled with liquid nitrogen, Sagittal and coronal sections (IO pm thick) were prepared, in a ReicherttYung Frigocut cryostat. collected on glass slides previously processed as described elsewhere’ and air dried. lmmunostaining and photography. Immunostaining was performed by an indirect technique either by immunorhodamine or by immunoperoxidase as described.h Rhodamine-conjugated goat anti-rabbit IgGs were from Cooper Biochemicals, Inc. (Malvern, PA). Peroxidase-conjugated goat F(ab), anti-rabbit IgGs were from the Pasteur Institute (Paris). Affinity-purified SgII IgGs were used at the final concentrations of 0.1 mgjml for immunofluorescence and 0.01 mg/ml for immunoperoxidase. Controls included substitution of the antiserum directed against SglI with non-immune rabbit IgGs, with preimmune rabbit serum or with rabbit antibodies directed against other antigens. At the end of immunostaining glass slides were mounted with glvcerol 195% (v/v) in 12mM Na-phosphate buffer, pH 7.41: _ Micrographs were taken on a Zeiss photomicroscope III equipped with epifluorescence. Pictures were taken with technical pan 2415 film which was developed with either D19 (fluorescence) or dilution B of HCI 10 (peroxidase). Films and chemicals were from Kodak.
Fig. 1. Low-power view of coronal (a) and sagittal (b) sections of the rat cerebellum and brain stem stained for SgII by immunoperoxidase. The photograph has been obtained by using the immunostained slides as negatives in the photographic enlarger. As a result the dark peroxidase reaction product appears white. lmmunoreactivity is particularly abundant in various regions of the brain stem [see for example in (a) the spina (1) and media (2) vestibular nuclei, the nuclei of the solitary tract (3). the lateral reticular nuclei (4) the nuclei of the inferior olive (5), and in (b) the dorsal (6) and ventral parabranchial (7) nuclei, the pontine (8) and reticular (9) nuclei and the nucleus of the spinal tract of the trigemini (IO)]. At this level of resolution particularly low immunoreactivity is visible in the ventral part of the lateral vestibular nucleus (11) and in the lateral superior olive (12). In the cerebellum the highest level of immunoreactivity is visible in the granule cell layer (gl). The relatively light appearance of the white matter in the cerebellar folia (wm) is due to a refraction effect during the print and not to the presence of detectable immunoreactivity. Bar = 500 pm.
,nrse
parL_nea
:ourse from the Purkinje oz.11layer to the of the molecular layer (ml). The same fibers appear as rows of puncta in the coronal view. &en their mo~hology these fibers probably represent climbing fibers which are known to travel and branch in planes perpendicular to the lon~tudinai axis of the cerebellar folia. Note that in the coronal sections regions rich in SgII-positive fibers (**) alternate with regions which contain only few of them (*). These regions are in register in different folia. Immunoreactivity is practically undectable in the white matter of the folia (wm). A very low level of immunoreactivity is visible in the white matter of the cerebellar medulla (CM). 425
M. G. COZZIet ul
426
Not all climbing fibers were stained (or equally stained). In general, variations were more apparent in the coronal plane, as is visible in Figs 2b and 4. Furthermore, variations in the coronal plane were in register in different folia (Fig. 2b). Occasionally, large variations were also seen in the sagittal plane (Fig. 6a). A few heavily stained fibers which did not have the typical appearance of climbing fibers were also seen (Fig. 6b and c). They were highly fluorescent beaded fibers with a rather rectilinear course travelling both in the sagittal (Fig. 6b) and in the coronal (Fig. 6c) plane. In all regions of the molecular layer no relevant immunoreactivity was detected on perikarya.
RESULTS
A low-magnification view of the cerebellum and of adjacent brain stem regions stained for SgII by an indirect immunoperoxidase technique is shown in Fig. 1. As can be seen in these low-power micrographs, SgII immunoreactivity has a very widespread distribution. It is highly concentrated in a variety of regions of the brain stem and it is also present in all regions of the cerebellum. In the cerebellum it is present at particularly high levels in the granule cell layer. Similar results were obtained by immunofluorescence (not shown). At high-power observation, immunoreactivity in all regions had a very fine punctate appearance, which was evident in sections stained by immunofluorescence (see for example Figs 5, 7b and 10). Puncta probably represent individual LDCVS.‘.~~A description of the pattern of immunoreactivity in the major cerebellar compartments and in adjacent brain stem regions follows.
Purkinje cell layer
Purkinje cells were embedded in a network of immunoreactive varicose fibers which were visible in tangential sections of the layer (Fig. 7a). A moderate, variable degree of immunoreactivity was also visible on the perikarya of some, but not all, Purkinje cells. Such immunoreactivity occurred in the form of small clusters made up of extremely fine puncta (Fig. 7b and c). The cytoplasmic distribution of these patches was reminiscent of the distribution of elements of the Golgi complex in Purkinje cells (our unpublished observations).
Cerebellar cortex
Clearly distinct patterns of immunoreactivity were visible in the molecular and in the granule cell layer (Fig. 2). The overall patterns of staining in the two layers were similar in all cerebellar folia and at all mediolateral levels (vermis, emispheres, flocculus and paraflocculus). However, some regional variations in the density of immunoreactive elements, both in the sagittal and in the coronal plane, were observed (see below).
Granule cell layer
Molecular layer
Sagittal and coronal views of the molecular layer are shown in Figs 2-6. Immunoreactivity occurs primarily in the form of branched varicose fibers which originate at the level of the Purkinje cell layer. From this region they course towards the pial surface giving rise to several collateral branches in the sagittal plane. In coronal sections they appear primarily as almost rectilinear fibers. Their course and their branching follows the course and the branching of major dendrites of Purkinje cells. In some cases they take a spiral course around these dendrites (Fig. 5). As a result, they outline the proximal portions of dendritic trees of Purkinje cells. Given this particular morphology, these fibers are very likely to be climbing fibers. ‘* In peroxidase-stained sections they have a coarse varicose appearance. High-power viewing of sections stained by immunofluorescence reveals the punctate pattern of the staining (Fig. 5). Varicosities appeared to result from the close apposition of a cluster of such puncta (Fig. 5~).
At low-power observation, immunoreactivity in this layer had a very similar patchy appearance in sagittal and coronal views of all cerebellar regions (Figs 2 and 8-10). The patches had irregular highly indented profiles and appeared to be interconnected by moderately-stained fibers which seemed to radiate from the underlying white matter (Fig. 8). This is the typical morphology expected for mossy fiber terminals and for their preterminal axons.32 Immunoreactivity on individual mossy fibers was variable and a comparison with adjacent sections stained for all nerve terminals [using antibodies to synapsin I as markers (not shown)‘] suggested that not all mossy fibers are immunoreactive for SgII. A subpopulation of neuronal somata was also heavily stained (Figs 7b, 9b, d-f and 10~). These neurons were smaller than Golgi cells but slightly larger than granule cells and had a scanty cytoplasm. A subpopulation of neuronal somata was also heavily stained (Figs 7b, 9b, d-f and 10~). These In these regions their numerical density was variable (compare for example Fig. 9a with Fig. 9b and Fig. 1Oawith Fig. IOc), but it was always much lower than that of granule cells.
Fig. 3. Distribution of Se11 in saaittal sections of the cerebellar cortex of three different folia (igmunoperoxidase). Immu~oreactivebranched varicose fibers are visible in the molecular layer (ml). The large majority of these fibers have the typical course of climbing fibers, i.e. they follow and course around the major dendrites of Purkinje cells (PC). A weak non-uniform immunoreactivity is also visible in the cytoplasm of Purkinje cells [best visible in (b) and (c)l. Islands of very intense immunoreactivity. corresponding to the cerebellar glomeruli, are visible in the granule cell layer (gl). Bar = 100 pm.
Fig. 5. High-power view of SgII-immunoreactive climbing fibers stained by immunofluorescence. The course along and around Purkinje cell dendrites of these fibers is clearly visible in the various micrographs. The spiral course of one of such fibers is visible in (d). Note that immunoreactivity of these fibers is accounted for by fine puncta. Varicosities are larger aggregates of such puncta [see arrowheads in (c)l. A moderate level of immunoreactivity is visible in the somata of Purkinje cells (PC). This immunoreactivity is higher in the Purkinje cells shown in (a), Note also the heterogen~us immunoreactivity of the glomeruli in the granule cell layer (gl). An arrow points to a glomerulus which is highiy immunoreactive. ml = molecular layer. Bars = 50 grn.
Fig. 4. Distribution of SgII in coronal sections of the cerebellar cortex (immunoperoxidase). Immunoreactive rectilinear varicose fibers and smah puncta are visible in the molecular layer (ml). These fibers which appear to outline the proximal dendrites of Purkinje cells and in a few cases f(b) and (c)] also their sonata, have the expected morphology of climbing fibers. Note in (a) a region rich in immunoreactive fibers (**) adjacent to a region almost devoid of such fibers (*). Intense immunoreactivity is visible in the glomeruli of the granule cell layer (gl), a moderate immunoreactivity in Purkinje cells (PC). wm = white matter. Bars = 100 pm. 429
430
M. G. Cozzt et al.
White matter
At low-power observation little immunostaining was visible in the cerebellar white matter (Fig. 2a, see also Fig. 11). At high-power observation, a very low density of immunoreactive puncta (probably LDCVs on their way to the terminals) was observed in all white matter regions (not shown). Deep cerebellar nuclei
Immunoreactivity in the deep cerebellar nuclei appeared primarily in the form of a loase plexus of highly stained varicose fibers (Figs 11 and 12). No detectable immunoreactivity was visible in the perikarya of neurons of the nuclei. The same pattern was
observed on all of the three major nuclei and in the dorsal part of the lateral vestibular nucleus (Fig. I), which shares im~rtant simiia~ties with the deep cerebellar nuclei in its synaptic connections.6 Very little SgII immunoreactivity was observed in the ventral part of the lateral vestibular nucleus (Fig. 1). The number of highly immunoreactive fibers was far lower than the number of Purkinje cell fibers innervating the nuclei (see Ref. 6). Brain stem
In most regions of the brain stem adjacent to the cerebellum SgII immunor~acti~ty appeared to be present at high, although variable, concentrations (Figs 1, 13). SgII immunoreactivity was primarily
Fig. 6. Patterns of Sgll immunorea~tivit~ in the cerebellar cortex as visualized by immuno~uor~s~nce. (a) Sagittal section of a region in which only few climbing fibers are highly immunoreactive (arrow). (b) Sagittal section of a region where climbing fibers are not immunoreactive. An intensely stained varicose rectilinear fiber oriented perpendicular to the pial surface is visible. (c) Coronal view of the molecular layer of two adjacent folia. The pial surface of the folia is indicated by arrowheads. Note the presence of both fibers running perpendicular to the pial surface (small arrows), as well as of more intensely immunoreactive varicose fibersinning in the coronal plane parallel to the pial surface (large arrows). ml = molecular layer, gl = granule cell layer, PC = Purkinje cells. 3ars: (a) 100 nm; (b) and (c) 50 pm. Fig. 7. Tangential view of the Purkinje cell layer stained by immunoperoxidase (a) and by immunofluorescence [(b) and (c)l. A moderate, variable, immunoreactivity is visible in Purkinje cells (PC). As visible in the sections stained by immunofluorescence, such immunoreactivity is accounted for by fine puncta sparse in the cytoplasm and often grouped in small clusters. In (b) a Purkinje cell is adjacent to a smaller cell (arrow) which also contains SgII-positive puncta in its thin rim of cytoplasm surrounding the nucleus. A network of varicose immunoreactive fibers is visible around Purkinje cells in (a) (arrows). gl = granule cell layer, ml = molecular layer. Bars: (a) 100 pm; (b) and (c) 50 pm. Fig. 8. Distribution of SgII immunoreactivity in a sagittal section of the granule cell layer (immunoperoxidase). Heavily stained patches of immunoreactivity are visible in this layer. Sometimes moderately immunoreactive fibers which seem to terminate in these patches are visible (arrows). These structures have the typical morphology of mossy fibers and their terminals. Almost no staining is visible in the white matter (wm) from which immunoreactive fibers appear to radiate, indicating that immunoreactivity is concentrated in their terminal regions. ml = molecular layer; PC = Purkinje cell. Bar = 100 pm. Fig. 9. High-power view of selected fields of the granule cell layer (sagittal sections) showing the variability of SgII-positive structures in different regions of this layer (immunoperoxidase). (a) A region where immunoreactivity seems to be primarily concentrate in the glomeruli (g). A glomerulus is shown at higher magnification in (c), (b) A region where immunoreactivity is accounted for not only by the glomeruli but also by cell perikarya (p). Some of these perikarya are also shown in fields (dHf). Immunoreactivity in the cytoplasm of perikarya occurs either in the form of puncta or of larger clumps. In some cases a thick short process which appears to terminate in a glomerulus originates from these cells [arrows in (b), (d) and (f)]. Bar = 50 nm. Fig. 10. High-power view of selected fields of the granule cell layer (sagittal sections) showing SgII-positive structures in different regions of this layer as revealed by immunotluorescence. (a) A region where immunoreactivity seems to be primarily concentrated in the glomeruli (g). Immunoreactive fibers which apparently interconnect the glomeruli are also visible. (b) A higher magnification which shows a fiber (small arrows) which seems to connect two glomeruli (gl and g2). In this field it can be seen that immunoreactivity in both the fiber and the glomerulus is accounted for by very fine puncta. g3 is a glomerulus which is much more immunoreactive than the surrounding glomeruli. (c) A region which contains a high density of immunoreactive perikarya (p). The thin cytoplasm of these neurons is filled with fluorescent puncta. Bar = 50 pm. Fig. 11, Distribution of SgII in a sagittal section of the cerebellar medulla (immunoperoxidase). Several folds of the cerebellar cortex are visible in the field. The field also includes a portion of the nucleus interpositus (Int). A loose network of immunoreactive fibers is visible in this nucleus. Immunoreactivity in most regions of the surrounding white matter (wm) is at the limit of detectability. Low-level immunoreactivity, however, is visible in cross-sectioned bundles of axons (*). Given their course they might represent decussating olivocerebellar fibers, i.e. the axons which generate climbing fibers. ps = pial surface. Bar = 50 pm. Fig. 12. Distribution of SgII in the deep cerebellar nuclei (immunoperoxidase). (a) and (b) Sagittal sections of the nucleus fastigius and dentatus, respectively. (c) A coronal section of the nucleus interpositus. In all fields immunor~ctivity occurs in the form of heavily immunoreactive varicose fibers and isolated puncta, many of which are probably cross-sections of tibers. Small bundles of such fibers are visible in (a) (arrows). Virtually no immunoreactivity is visible in cell somata. wm = white matter. Bar = 100 pm.
Fig. 7 432
Secretogranin II in the rat cerebellum
Fig. 8
433
438
M. G. COZZIet al.
Fig. 13. Distribution of Sgil in the ventral portions of the brain stem (immunoperoxidase). Low-power views of a sag&al (a) and a coronal (b) section. A very widespread distribution of immunoreactivity is visible in all the regions of the neuropile. No immunoreactivity is visible in white matter tracts which are primarily longitudinally-sectioned in (a) and cross-sectioned in (b). Note, for example, the lack of immunoreactivity in the pyramidal tracts (py) visible in (b). 10 = inferior olive, RF = reticular formation. Bar =i 5OOpm.
Secretogranin II in the rat cerebellum confined to neuropile-containing regions and occurred in the form of varicose fibers or isolated boutons. Only a weak immunostaining was observed in perikarya of neurons of the olivary nuclei (Fig. 13b), which are the source of climbing fibers. Little immunoreactivity was observed in white matter bundles with the exception of scattered fine puncta visible only at high magnification.
DISCUSSION
We report in the present study that SgII immunoreactivity is widely distributed in the gray matter of the rat cerebellum and of most adjacent brain stem regions. In all these brain regions SgII is primarily concentrated in varicose processes and in boutons. Some intensely immunoreactive perikarya are also present, but clearly positive perikarya represent only a minority of all the neurons that express SgII. At all sites immunoreactivity has a very fine punctate appearance consistent with a localization of SgII in LDCVS.‘s*~This overall pattern of immunoreactivity is similar to that observed after immunostaining for well-established secretory peptides (see for example Refs 17,23,40). However, SgII is present in a larger number of different cerebellar fiber and cell populations (see for comparison Refs 2, 3, 17, 20, 21, 23, 31, 40). In the molecular layer SgII is primarily concentrated in fibers with the typical course and morphology of climbing fibers3* The intensity of immunoreactivity was found to be variable between individual climbing fibers but the majority of these fibers appeared to contain a high or relevant concentration of immunoreactivity. Recently, two reguand corticotropinlatory peptides, enkephalin detected by factor, have been releasing climbing immunocytochemistry in mammalian fibers3,4.2’.33,39.” (Mugnaini, personal communication). However, enkephalin immunoreactivity so far has been detected only in climbing fibers of the opossum but not of the rat. *’ Transient appearance of calcitonin gene-related peptide in climbing fibers of developing rat cerebellum has also been reported.24 In addition to climbing fibers a few other varicose processes with a rectilinear course were heavily labeled in the molecular layer. Some of these fibers might represent catecholaminergic axons.*’ SgII is present at high concentrations in the granule cell layer of all cerebral regions, where it is primarily localized in mossy fibers’ terminals. Immunostaining intensity on individual mossy fibers was clearly quite variable, but no obvious differences were noted in the numerical density of heavily-stained mossy fibers in the various regions. Heavily-stained and moderately-stained or unstained mossy fibers were almost randomly interspersed in all regions, in contrast to the non-random distribution of climbing fibers with different degrees of immunostaining
439
in the molecular layer. Enkephalin, substance P, corticotropin-releasing factor and cholecystokinin have been found in mammalian mossy fibers.3~20~2’~23~39 However, a much more prominent regional heterogeneity has been observed for SgII. In addition to mossy fibers, a subpopulation of neuronal perikarya was heavily immunostained for SgII in the granule layer. These perikarya were concentrated in specific regions of the cerebellar cortex. In their size, shape and distribution these neurons appeared remarkably similar to a previously unknown population of cerebellar neurons which has been recently described.13 A moderate degree of SgII immunoreactivity was clearly detectable at least in some Purkinje cell perikarya. However, in the deep cerebellar nuclei, where the axons of Purkinje cells terminate, only a loose network of heavily immunoreactive fibers was visible. Terminal axons of Purkinje cells are known to constitute a much denser network.6 Thus, these immunoreactive fibers either did not represent Purkinje cell axons, or represented axons of a minor subpopulation of Purkinje cells. In most of the brain stem adjacent to the cerebellum the concentration of SgII was impressively high (a noticeable exception was the ventral portion of the lateral vestibular nucleus). An identification of immunoreactive processes and terminals was not attempted in this study. To our knowledge, however, no other secretory neuro-peptide was previously found to be so widely distributed within the brain stem. From the preliminary results we have obtained, it appears that the distribution of SgII in the cerebellum and in the brain stem is very different from the distribution of secretogranin I (chromogranin B) in the same regions. This protein appears to be present at an overall lower concentration in these regions and, in particular, in the cerebellum. Furthermore, while SgII is primarily concentrated in varicose processes and terminals, secretogranin I (chromogranin B) appears to be primarily concentrated in perikarya (for example in the perikarya of the deep cerebellar nuclei). The distribution of chromogranin A in these regions has not been described in detail.4’ From the reported brief description of its localization in these brain regions, it appears that the distribution of chromogranin A is more similar to the distribution of secretogranin I (chromogranin B) than to that of SgII. For example, also chromogranin A is present at high concentration in the perikarya of deep cerebellar nuclei. The comparison of the distribution of SgII with that of secretogranin I (chromogranin B) (our unpublished observations) and chromogranin A4’ indicates that although SgII is a peptide much more widely distributed in the nervous system than known peptide neurotransmitters studied so far,38 it is not a general marker for secretory organelles which contain
430
M. G. COZZI et (11.
regulatory peptides. This is in agreement with previous results obtained by us in other regions of the nervous system” and with results obtained by the present authors and by others in the endocrine system.“‘~“~2’~74.‘7~‘R However, a lack of detectable immunoreactivity in many neurons and neuronal processes does not necessarily rule out that low levels of SgII might be expressed in all neurons. In fact, although
the immunoreactive structures that we have described here were observed in all the cerebella that we examined, some variability was found from one specimen to another in the distribution of the lowest levels of immunoreactivity.
CONCLUSION
This study suggests that peptide neurotransmission plays an important role in cerebellar physiology and shows that antibodies directed against peptides of the secretogranin/chromogranin family might be a tool with which to visualize peptidergic pathways for which specific regulatory peptides have not yet been identified. Acknowledgements-We thank M. Moretti and Dr M. Solimena for kind help during some experiments, and P. Tinelli and U. Weib for excellent technical assistance. This work was partially supported by a grant of the Italian National R&ear&Council, Stratkgic Project 04/03 and by an MDA grant to P.D.C. W.B.H. was the recipient of a grant from the Deutsche Forschungsgemeinshaft (Hu 27513-3).
REFERENCES
1. Benedum U. M., Lamouroux A., Konecki D. S., Rosa P., Hille A., Baeuerle P. A., Frank R., Lottspeich F., Mallet
2
3. 4. 5.
6. I.
8. 9. 10.
Il. 12. 13.
J. and Huttner W. B. (1987) The primary structure of human secretogranin I (chromogranin B): comparison with chromogranin A reveals homologous terminal domains and a large intervening variable region. Eur. molec. Biol. Org. J. 6, 1203~1211. Chan-Palay V., Nilaver G., Palay S. L., Beinfeld M. C., Zimmerman E. A., Wu J. and O’Donohue T. L. (1981) Chemical heterogeneity in cerebellar Purkinje cells: existence and coexistence of glutamic acid decarboxylase-like and motilin-like immunoreactivities. Proc. nafn. Acud. Sci. U.S.A. 78, 7787-7791. Cummings S., Bishop G. A. and King J. S. (1987) Corticotropin releasing factor distribution in the cerebellum and precerebellar nuclei of the cat. Society for Neuroscience 17thAnn. Meet&g, New Orleans, LA, Abstr. 351.4, 1262. Cummings S., Sharp B. and Elde R. (1988) Corticotrouin-releasing factor in cerebellar afferent svstems. A combined immunoKistochemi&y and retrograde transport study: J. Neurosz. 8, 543-554. De Camilli P., Cameron R. and Greengard P. (1983) Synapsin I (Protein I), a nerve terminal-specific phosphoprotein. 1. Its general distribution in synapses of the central and peripheral nervous system demonstrated by immunofluorescence in frozen and plastic sections. J. Cell Biol. %, 1337-1354. De Camilli P., Miller P. E., Levitt P., Walter U. and Greengard P. (1984) Anatomy of cerebellar Purkinje cells in the rat determined by a specific immunohistochemical marker. Neuroscience 11, 761-817. De Camilli P. and Navone F. (1987) Regulated secretory pathways in neurons and their relation to the regulated secretory pathway of endocrine cells. In Cellular and Molecular Biology of Hormone and Neurotransmitter Containing Secretory Vesicles (ed. Johnson R. G. Jr). Ann. N.Y. Acad. Sci. 493, 461-479. Eiden L. E. (1987) Is chromogranin a prohormone? Nature 325, 301. Eiden L. E., Huttner W. B., Mallet J., O’Connor D. T., Winkler H. and Zanini A. (1987) A nomenclature proposal . for the chromogranin/secretogranin proteins. Neuroscience 21, 1019-1021. Fischer-Colbrie R., Haan C. and Schober M. (1987) Chromogranins A. B and C: widesuread constituents of secretorv vesicles. In Cellular and-Molecular Biology of Hormone and Ne~rotransmitter Containing Secretory Vesicles (ed. Johnson R. G. Jr). Ann. N. Y. Acad. Sci. 493, 120-134. Hagn C., Schmid K. W., Fischer-Colbrie R. and Winkler H. (1986) Chromogranin A, B and C in human adrenal medulla and endocrine tissues. Lab. Invest. 55, 405411. Hearn S. A. (1987) Electron microscopic localization of chromogranin A in osmium-fixed neuroendocrine cells with a protein A-gold technique. J. Histochem. Cytochem. 35, 795-801. Hockfield S. (1987) A Mab to a unique cerebellar neuron generated by immunosuppression and rapid immumzatlon. Science 237, 67-70.
14. Hiikfelt T., Johansson 0. and Goldstein M. (1984) Chemical anatomy of the brain. Science 225, 132&1334. 15. Huttner W. B. and Benedum U. M. (1987) Chromogranin A and pancreastatin. Nature 325, 305. 16. Huttner W. B., Benedum U. M. and Rosa P. (1988) Biosynthesis, structure and function of the secretogranins/chromogranins. In Molecular Mechanisms in Secretion, Alfred Benzon Symposium No. 25, Munksgaard, Copenhagen (eds Thorn N., Treiman M., Petersen 0. H. and Thaysen J. H.), pp. 380-389. 17. Johansson O., Hijkfelt T. and Elde R. P. (1984) Immunohistochemical distribution of somatostatin-like immunoreactivity in the central nervous system of the adult rat. Neuroscience 13, 265-339. 18. Kawai Y., Emson P. C., Hillyard C. J., Girgis S., Maclntyre I., Oertel W. H. and Tohyama M. (1987) Immunohistochemical evidence for the coexistence of calcitonin gene-related peptide and glutamate decarboxylase-like immunoreactivities in the Purkinje cells of the rat cerebellum. Brain Res. 409, 371-373. 19. King J. S., Cummings S. L., Young W. S. III and De Souza E. B. Anatomical evidence for corticotropin releasing factor in the olive cerebellar system. In The Olivocerebellar System in Motor Control (ed. Strata P.). Expl Brain Res. (in press). 20. King J. S., Ho R. H. and Bishop G. A. (1986) Cholecystokinin-like immunoreactivity in the cerebellum of the opossum. Society for Neuroscience 16th Annual Meeting, Washington D.C.. Abstr. 127.2, 461. 21. King J. S., Ho R. H. and Bishop G. A. (1986) Anatomical evidence for enkephalin immunoreactive climbing fibers in the cerebellar cortex of the opossum. J. Neurocytol. 15, 545-559. 22. Konecki D. S., Benedum U. M., Gerdes H. and Huttner W. B. (1987) The primary structure of human chromogranin A and pancreastatin. J. biol. Chem. 262, 1702617030.
Secretogranin II in the rat cerebellum
441
23. Korte G. E., Reiner A. and Karten H. J. (1980) Substance P-like immunoreactivity in cerebellar mossy fibers and terminals in the red-eared turtle Chrisemys Scrip[a Elegans. Neuroscience 5, 903-914. 24. Kubota Y., Inagaki S., Shimada S., Kito S., Zaidi M., Girgis S. I., Macintyre I. and Tohyama M. (1987) Transient appearance of calcitonin gene-related ~ptide-like immunoreactive fibers in the developing cerebellum of the rat. Brain Res. 415, 385-388. 25. Lassmann H., Hagn C., Fischer-Colbrie R. and Winkler H. (1986) Presence of chromogranin A, B and C in bovine endocrine and nervous tissues: a comparative immunohistochemical study. Histochem. J. 18, 38&386. 26. Lee R. W. H. and Huttner W. B. (1983) Tyrosine-O-sulfated proteins of PC12 pheochromocytoma cells and their sulfation by a tyrosylprotein sulfotransferase. 1. biol. Cbem. 258, 11326-l 1334. 27. Mugnaini E. and Dahl A, (1975) Mode of distribution of aminergic fibers in the cerebellar cortex of the chicken. j. camp. Neural. 162, 417-432. factor (CRF) and the feline olivary hypertrophy. In The 28. Mugnaini E. and Nelson B. Corticotropin-releasing Olivocerebellar System in Motor Con@ol (ed. Strata P.). Expl Brain Res. (in press). 29. Navone F., Di Gioia G., Matteoli M. and De Camilli P. (1988) Small synaptic vesicles and large dense-core vesicles of neurons are related to two distinct types of vesicles of endocrine cells. In Molecular ~echon~sms in Secretion. Alfred Benzon Sym~sium No. 25, Munksgaard, Copenhagen (eds Thorn N., Treiman M., Petersen 0. H. and Thaysen J. H.), pp. 433450. 30. Neuman B., Wiedermann C. J., Fischer-Colbrie R., Schober M., Sperk G. and Winkler H. (1984) Biochemical and functional moperties of large and small dense-core vesicles in sympathetic nerves of rat and ox vas deferens. Neuroscience lj, 921-931. 31. Nilaver G., Defendini R., Zimmerman E. A., Beinfeld M. C. and O’Donohue T. L. (1982) Motilin in the Purkinje cell of the cerebellum. Nature 295, 597-598. 32. Palay S. L. and Chan-Palay V. (1974) The Cerebetlar Corfex: Cytology and Organization. Springer, New York. 33. Palkovitz M., L&nth C., Garcs T. and Young W. S. III (1987) Corticotropin-releasing factor in the olivocerebellar tract of rats: demonstration by light- and electron-microscopic immunohistochemistry and in situ hybridization histochemistry. Proc. natn. Acud. Sci. U.S.A. 84, 3911-3915. 34. Rindi G., Buffa R., Sessa F., Tortora 0. and Solcia E. (1986) Chromogranin A, B and C immunoreacti~ties of ma~alian endocrine cells. Dist~bution, distinction from costored ho~on~/proho~on~ and relationship with the argyrophil component of secretory granules. Hisrochemistry 85, 19-28. 35. Rosa P. and Zanini A. (1981) Characterization of adenohypophyseal polypeptides by two-dimensional gel electrophoresis. II. Sulfated and glycosylated polypeptides. Molec. cell Endocrinol. 24, 181-193. 36. Rosa P. and Zanini A. (1983) Purification of a sulfated secretory protein from the adenohypophysis. Immunochemical evidence that similar macromolecules are present in other glands. Eur. J. Ceii Biol. 31, 94-98. 37. Rosa P., Fumagalli G., Zanini A. and Hut&r W. B. (1985)The major tyrosine-sulfate protein of the bovine anterior pituitary is a secretory protein present in gonadotrophs, thyrotrophs, mammotrophs and corticotrophs. J. Cell Biol. 100, 928-937. 38. Rosa P., Hille A., Lee R. W. H., Zanini A., De Camilli P. and Huttner W. B. (1985) Secretogranin I and II: two tyrosine-sulfated secretory proteins common to a variety of cells secreting peptides by the regulated pathway. J. Cell Biol. 101, 1999-201 I. factor-like immunor~cti~ty in the rat brain 39. Sakanaka M., Shibasaki T. and Lederis K. (1987) Corticotropin-rel~sing as revealed by a modified cobalt-glucose oxidastiiaminobenzidine method. J. camp. Neuroi. 240, 256-298. 40. Schulman J. A., Finger T. E., Brecha N. C. and Karten H. J. (1981) Enkephalin immunoreactivity in Go&i cells and mossy fibers of mammalian, avian, amphibian and teleost cerebellum. Neuroscience 6, 2407-2416. 41. Somogyi P., Hodgson A. J., De Potter R. W., Fischer-Colbrie R., Schober M., Winkler H. and Chubb I. W. (1984) Chromogranin immunoreactivity in the central nervous system. Immunochemical characterization, distribution and relationship to catecholamine and enkephalin pathways. Brain Res. Rev. 8, 193-230. 42. Sovenv C., Mercuri J. and Hansky J. (1984) Distribution of bombesin- and cholecystokinin-like immunoreactivjty in rat anb dog brain and gastrointeitinai traci. Reguf. Pepr. 9, 61-68. 43. Winkler H., Apps D. K. and Fischer-Colbrie R. (1986) The molecular function of adrenal chromaffin granules: established facts and unresolved topics. Neuroscience 18, 261-290. 44. Young W. S. III, Walker L. C., Powers R. E., DeSouza E. B. and Price D. L. (1986) Corticotropin-releasing factor mRNA is expressed in the inferior olives of rodents and primates. Molec. Brain Res. 1, 189-192. (Accepted 29 June 1988)