Neur5sc~e~ee Vol. 34, No. 2, pp. 479489, Printed in Great Britain

030~4522/~ $3.00 + 0.00 Pergamon Press plc Q 1990 IBRO

1990

ULTRASTRUCTURAL LOCALIZATION OF ENKEPHALIN-LIKE IMMUNOREACTIVITY IN DEVELOPING RAT CEREBELLUM 1.X ZAGON and P.J. MCLAUGHLIN Department of Anatomy, The M. S. Hershey Medical Center, The Pennsylvania State University, Hershey, PA 17033, U.S.A. Abstract-Methionine enkephalin, an endogenous opioid peptide, participates in the regulation of growth in the developing brain. In the present study, enkephalin-like immunoreactivity was localized in the cerebellum of developing and adult rats by immunoelectron microscopy. In lo-day-old animals, enkephalin-like immunoreactivity was found in the somata of proliferating, migrating and differentiating neural cells, and was associated with the plasma membrane, microtubuies, filaments, mitochondria, endoplasmic reticulum and nuclear envelope. Both neurons and glia in the cerebellum of the preweaning rat displayed a similar profile of immunoreactivity. Reaction product was also detected in the dendrites and dendritic spines of Purkinje cells where it was concentrated in postsynaptic densities. The majority of internal granule neurons in IO-day-old animals were not immunoreactive, nor were axons, glial processes and postsynaptic elements (with the exception of mossy fiber terminals). At weaning (Day 21), enkephalin-like immunoreactivity was confined primarily to the somata of Purkinje, basket and stellate neurons, and to Purkinje cell dendrites and synaptic spines. Adult rats (day 75) exhibited no enkephalinlike immunoreactivity. These results establish that enkephalin or an enkephalin-like substance can be detected during the ontogeny of both neurons and glia in the cerebellar cortex, and appears to be associated with certain structural elements.

The endogenous opioids and opioid receptors may related to other functions is unclear. We43have found play important roles in the regulation of nervous enkephalin-like immunoreactivity expressed transystem development. 34,35Opioid receptors have been siently in the developing cerebellum; the germinative identified in brain (and body) tissues during onto(i.e. external granule cells) and developing neural geny,4.‘7,31 and endogenous opioids are present in cells (e.g. Purkinje cells) of preweaning rats are plasma, as well as brain and body tissues, of various immunoreactive, their mature (35-day-old) counterdeveloping organisms.i,‘~,3’*43.~Although the presence parts (e.g. internal granule neurons) are not. This of these opioid systems may indicate the development would suggest that some endogenous opioids are of mature neural functions (e.g. neurotransmission), involved specifically in early phases of nervous system the elaboration of certain endogenous opioids/opioid ontogeny. receptors appears to be related specifically to develEndogenous opioids have been localized by elecopment. In experimental paradigms utilizing opioid tron microscopy5~“~23~24 in adult neural elements. agonists,” as well as opioid antagonists,*~9~25~~8~~~ Osamura et ~1.~”and Kapadia and Kapadiai2 have endogenous opioids have been demonstrated to be examined the ultrastructural localization of opioids inhibitory trophic factors that do not alter the basic in human fetal pituitary and gastric antrum, respecultrastructural morphology of the ~41s.‘~ Cell protively. Only light microscopic immunohistochemistry, liferatioqM cell differentiation>9s34,38,39and neurohowever, has been used to reveal endogenous opioid behavioral ontogeny’s,37 are influenced profoundly by localization in the developing brain. If we are to endogenous opioid-opioid receptor interactions. understand how endogenous opioid function moduImmunoreactive endogenous opioid-like sublates nervous system development, we need to know stances are detectable in fetal and neonatal neural the fine structural association of the opioids with cells. ‘5,‘6~20*43 Whether this endogenous activity in particular neural membranes, organelles and/or cytodeveloping neural cells indicates the ontogeny of skeletal structures. In this report we describe for the peptidergic neurons and/or is transient and thus first time the localization of enkephalin-like immunoreactivity by immunoelectron microscopy in the developing brain. Using antiserum to methionine enkephalin, an opioid peptide known to be involved Abbreviations: EGL, external granular layer; IGL, internal in the regulation of neural cell proliferation,19.4R42 granule layer; Met-ENK, methionine enkephalin; these studies were conducted in the developing and MOL, molecular layer; NGS, normal goat serum; PBS, phosphate-buffered serum. adult rat cerebellum. 479

480

I. S. ZAGON and P. EXPERIMENTAL

PROCEDURES

Animals Sprague-Dawley rats (Charles River Labs, Wilmington, MA), approximately 10, 45 and 75 days of age, were used in this study and housed under controlled conditionsXx

Light and eiectron ~mM~~~c.ytochernistr~ Light microscopic procedures for tissue preparation and immunocytochemistry have been described eisewhere,4i,4’.” except that peroxidase was used to visualize immunoreactivity.26 For immunoelectron microscopy, animals were anesthetized (Nembutal) and perfused intracardially with a primary fixative solution containing 4% paraformaidehyde, 0.25% giutaraldehyde and 0.25% sucrose in 0.07 M Sorenson’s phosphate buffer (pH 7.4, 20°C) for 5min at 90.IiOmm Hg. 3* The brain was removed immediately, placed in the primary fixative solution for 3 h at 4”C, and postfixed overnight in 4% paraformald~hyde in 0.07 M Sorenson’s phosphate buffer with 0.25% sucrose at 4°C. The cerebellum was excised, submerged in a solution of Sorenson’s phosphate buffer and 0.25% sucrose, and sectioned (120pm) in sagittai or coronal planes with a Vibratome (Oxford Vibratome, Electron Microscopy Sciences). In general, sections were utilized for immunocytochemistry inlmediately after preparation. Sections not processed immediately for immunoelectron microscopy were stored in 0.07 M Sorenson’s phosphate buffer with 0.25% sucrose and 0.01% sodium azide at 4°C; no differences in immunoreactivity were noted between freshly used sections and those stored in sucrose/sodium azide. For conventional morphological examination, animals were perfused intracardially with 2% giutaraidehyde and 4% paraformaidehyde in 0.07 M Sorenson’s phosphate with 0.25% sucrose (pH 7.4, 20”C).“z Cerebella were removed and postfixed in fresh fixative overnight at 4°C. Vibratome sections were postfixed in 1% osmium tetroxide in Sorenson’s phosphate buffer with 0.25% sucrose for 1h (4’C), and tissues were dehydrated and processed for electron microscopy as described below. For immunocyt~hemistry of electron microscopic specimens, sections were washed in three changes of phosphatebuffered saline (PBS) for I5 min. and preincubated in 2% normal goat serum (NGS) in PBS for 30min at room temperature. Sections were incubated overnight in a humid chamber with antiserum to methionine enkephaiin (INCSTAR Stillwater, MN) at I : 200 dilution in a solution of NGSIPBS. Control specimens were incubated in nonimmune IgG (I :200 dilution) or antiserum to methionine enkephalin preabsorbed with an excess of methionine enkephalin (Sigma, St. Louis, MO). Tissues were washed in NGS/PBS for 90min with three changes of solution. Sections were incubated in a humid atmosphere for 3 h with peroxidase conjugated goat anti-rabbit IgG (Cappei Laboratories, Division of Cooper Biomedical, Inc., Malvern, PA) diluted 1: 100 in NGS/PBS, and washed for 90 min (three changes of PBS solution). Sections were reacted with diaminobenzidine tetrachloride (I5 mgj50 ml, Sigma) containing 10~1 hydrogen peroxide for IOmin. The reaction was terminated by washing in a PBS solution and the reaction product intensified by immersion of the tissues in 1% osmium tetroxide and 6.1 M Sorenson’s phosphate buffer for 1h at 4°C. The tissue sections were dehvdrated in a graded series of alcohols and embedded in Epon. Thin sections were cut with a Reichert OMU-2 uitramicrotome and mounted on uncoated 200 and 300 mesh slotted copper grids. Sections of specimens prepared for conventional electron microscopy were stained with uranyl acetate and lead citrate. Tissues prepared for immunoelectron microscopy were not stained. Specimens were examined with a Philips 400 Electron Microscope at an accelerating voltage of 60 kV.

J. MCLAIJCHLIN

The anti~rum to methionine enkephalin utilized in this study was obtained commercially (INCSTAR, Stillwater, MN) and produced in rabbit against a keyhole limpet hemocyanin conjugate. Using a quantitative immunodot assay,2h we found that a I : 200 dilution of the antisera recognized down to IOOng of methionine enkephalin, but did not recognize up to 50 p(g of leucine enkephalin, human b-endorphin or dynorphin Al-13. In view of the limited characterization of this antibody, the immunoreactive substance identified by the antibody to methionine enkephalin will be referred to as methionine enkephalin-like (MetENK) or enkephalin. RESULTS

Examination of the cerebellar cortex of preweaning rats showed that Met-ENK was localized in the external granular layer (EGL) and molecular layer (MOL) (Fig. 1A). Immunostaining was prominently associated with the cytoplasm of the external granule cells (Fig. IA), whereas cell nuclei were unreactive. Often, this profile of peroxidase deposition presented a honeycombed appearance. Immunoreactivity was diminished markedly in the internal granule layer (IGL) and, when observed, was related to the granule cells. Control specimens exhibited no immunoreactivity (Fig. IB). Adult animals contained no Met-ENK (data not shown). With the exception of using peroxidase as a means of visualizing Met-ENK, these results confirmed earlier observations.4~ Electron microscopy Ten days. Investigation of the lo-day-old rat cerebellum processed with antiserum to Met-ENK (Figs lC4, 2 and 3), and conventionally prepared tissues (Fig. S), revealed a specific pattern of immunoreactivity at the electron microscopic level of resolution. Met-ENK was concentrated in the cytoplasm of external granule cells (Fig. IC-F). Within the cytoplasm of external granule cells, aggregates of immunostaining were located subjacent to the plasma membrane (Fig. IE), and they often extended some distance into the cytoplasm. In general, these peroxidase deposits measured IS-90 nm in diameter (Fig. 1E, F) and resembled polyribosomal clusters (polyribosomes average SO--l 30 nm) in conventional preparations (Fig. 5).‘2 Immunoreactivity was associated with vesicles of SO-70 nm in diameter (Fig. iF), and microtubule-like elements with a diameter of 20-26 nm (Figs 1F and 2A). Reaction product was found to be associated with the cytoplasmic surface of the nuclear envelope {Fig. 1E, F). Cell nuclei were unreactive (Fig. IC-F). The processes of Bcrgmann glial fibers which extended throughout the EGL were not immunoreactive (Fig. 11)). Peroxidase was often associated with the basal lamina (Fig. ID); however, this association was probably an artifact since it also occurred in control (i.e. pre-absorbed) specimens. Even the smallest ramifications of Purkinje cell dendrites in the molecular layer contained peroxidase

Fig. 1. lmmunocytochemical preparations of the developing rat cerebellar cortex using anti-methionine enkephalin lgG and peroxidase-conjugated goat anti-rabbit lgG. (A) and (B) are light micrographs of the cerebellar cortex from 7-day-old rats and (C)-(F) are electron micrographs from the cerebellum o f 10-day-old animals. (A) In this sagittal section, note the immunoreactivity (arrows) associated with cells in the EGL, and in the MOL. Cells in the IGL exhibited little immunoreactivity. Arrowheads, pia x 375. (B) A control specimen processed with anti-methionine enkephalin IgG absorbed with methionine enkephalin. × 375. (C) In this cross-section of the EGL, the process o f a glial cell (gl), presumably a Bergmann gliat fiber, can be seen to terminate subjacent to the basal lamina (arrowheads), Glial cell processes contained little Met-ENK. The nucleus (nu) and cytoplasm o f an external granule cell can be noted. × 20,650. (D) A sagittal section of the EGL recorded at low magnification, lmmunoperoxidase staining of the cytoplasm, but generally not the nucleus (nu), of external granule cells can be observed. The basal lamina (arrowheads) exhibited non-specific peroxidase deposits, × 3130. (E) The cytoplasm of external granule cells, but not cell nuclei (nu), contained Met-ENK. lmmunoperoxidase was often found in aggregates (arrow) subjacent to the plasma membrane (cross-hatched arrow), encapsulating (arrowheads) mitochondria (m), and in association with the nuclear envelopes facing the cytoplasm (double-hatched arrow), x 44,750. (F) The cytoplasm of an external granule cell. Met-ENK was associated with small vesicles (arrowheads) and microtubulelike elements (arrow). nu, nucleus, x 57,750. (G) External granule ceils (nu, nucleus; m, mitochondria) o f a control specimen processed with anti-methionine enkephalin lgG absorbed with methionine enkephalin. × 20,800. 481

Fig. 2. Electron micrographs of the MOL from the cerebellar cortex of IO-day-old rat. All tissues were processed with anti-methionine enkephalin IgG and peroxidase-conjugated goat anti-rabbit IgG. (A) In this transverse section, a Purkinje cell dendrite (den), but not parallel fibers (pf) or axons (ax), exhibited immunoreactivity. Immuno~roxidase staining was associated with microtubule-like and/or neurofilament-like structures (arrow), and extended into the synaptic spines (sp). m, mitrochondria. x 18,500. (B) The dendrite (den) of a Purkinje cell with intense immunoreactivity of postsynaptic densities (psd). Note that parallel fibers (pf) were not immunoperoxidase-positive. x 27,500. (C) An intermediate neuron, believed to be a basket cell, with immunoreaction product in the cytoplasm (*). but not the cell nucleus (nu). Adjacent parallel fibers (pf) were unreactive. x 8300. (D) A migrating externai granule cell with Met-ENK in the advancing process (*) but not in the cell nucleus (nu). Surrounding parallel fibers (pf) were not immunoreactive, but a Purkinje cell dendrite (den) and soma of a presumptive basket cell (BA) were positive for anti-methionine enkepha~in IgG. x 4600. 482

483

Opioids and brain development (Fig. 2A, B, D). As in the cell body, reaction product was localized subjacent to the plasma membrane and surrounded mitochondria. Met-ENK was often associated with long, straight microtubulelike elements that were approximately 27 nm in diameter (Fig. 2A). In some instances, reaction deposits were also associated with filament-like structures of about 1Onm in diameter. Met-ENK was especially prominent in the synaptic spines of the dendrites (Fig. 2A, B) where it was frequently associated with postsynaptic densities (Fig. 2B). Postsynaptic regions of dendritic shafts were also immunoreactive. Intermediate sized neurons believed to be stellate and basket neurons (Fig. 1E) exhibited considerable immunostaining of the cytoplasm, with a pattern of distribution of reaction product similar to EGL cells; the nuclei of these cells were unreactive (Fig. 2C). The soma and cytoplasm of the translocating cell process of EGL cells were also intensively immunoreactive (Fig. 2D). Axons (e.g. parallel fibers), presynaptic areas and nucleoplasm of cells in the molecular layer had no reaction product. The pattern of immunoreactivity in the soma of Purkinje cells (e.g. encapsulating mitochondria, subjacent to the plasma membrane) (Fig. 3A) was similar to that of EGL cells and Purkinje cell dendrites. Perisomatic spines also contained Met-ENK, with considerable immunoreactivity associated with the postsynaptic density (Fig. 3A). The soma of Bergmann glial cells were difficult to identify, precluding determination of immunoreactivity. In the internal granule layer, the pattern of immunoreactivity was a mosaic (Fig. 3B); most internal granule neurons were unreactive. While the cytoplasm of some internal granule neurons contained Met-ENK, the amount relative to that in the EGL cells was reduced qualitatively. The only axons in the internal granule layer that were immunoreactive were the terminals of presumptive mossy fiber endings (Fig. 3C). In these terminals, the reaction product surrounded mitochondria and synaptic vesicles, and was often found in aggregates subjacent to the plasma membrane. Axons in the medullary layer were not Met-ENKpositive (Fig. 3D, E). However, the cytoplasm but not the processes of glial cells, tentatively identified as oligodendrocytes (e.g. nuclei with numerous heterochromatin, cytoplasm rich with ribonucleoprotein), contained Met-ENK (Fig. 3D, E). Immunostaining of glial cells was localized subjacent to the plasma membrane and surrounded the mitochondria and the nuclear envelope. Forty-jive days. At 45 days, the distribution of immunoreactive cells often appeared to be reduced in the cerebellar cortex. The soma of some stellate/ basket cells still exhibited reaction product, but the quantity of immunostaining often appeared subjectively less than that observed in lo-day-old rats. In general, although the Purkinje cell soma exhibited little immunoreactivity, the shafts and spines of deposits

Purkinje cell dendrites were Met-ENK-positive, with considerable reaction product still associated with postsynaptic densities (Fig. 4A, B). Seventy-five days. At 75 days, no Met-ENK was detected in cerebellar cortical tissues (Fig. 4C). Controls. Control specimens processed with nonimmune IgG or anti-methionine IgG preabsorbed with methionine enkephalin were not immunoreactive (Fig. 1F). As mentioned earlier, reaction product was often associated with the basal lamina in control specimens. In addition, the glial cell coverings often contained some randomly distributed immunoreactive deposit. DISCUSSION

In the present study immunoelectron microscopy was used to localize enkephalin-like substances in the developing mammalian brain. Using immunocytochemistry and light microscopy, Zagon et aZ.43found that germinative and developing neural cells in the cerebellum of lo-day-old rats contained enkephalin; differentiated cellular counterparts examined in 35 day-old animals were not immunoreactive. While the present results confirm these earlier findings, they importantly provide the first detailed inspection of the ultrastructural localization of opioids in developing neuronal and glial cells. As summarized in Fig. 6, enkephalin-like activity in the developing cerebellar cortex was localized in the soma of proliferating, migrating and differentiating neural cells, and was associated with the plasma membrane and the cytoplasmic surfaces of mitochondria, endoplasmic reticulum and the nuclear envelope. The profiles of enkephalin distribution in the somata of both neuronal and glial cells were similar. The enkephalinlike molecules detected in the dendrites of Purkinje cells were associated with the plasma membrane, organelles and cytoskeletal matrix (e.g. microtubules). When enkephalin immunoreactivity was present in the dendritic spines of some neurons, it was often associated with postsynaptic densities. Axons, glial processes, and presynaptic elements were not immunoreactive, with the exception of terminals in the IGL resembling those of mossy fibers. Thus, as maturation of the brain occurs, the detection of enkephalin-like substances by immunoelectron microscopy diminishes progressively in the cerebellar cortex and is virtually absent by young adulthood. Enkephalin immunoreactivity was detectable during the ontogeny of cells derived both prenatally (Purkinje neurons) and postnatally (internal granule, basket and stellate neurons), and in neural cells of all sizes and types (e.g. macroneurons, microneurons). The temporal course of Met-ENK disappearance during ontogeny was dependent on cell type. Of course, a number of factors impact upon these observations. Firstly, the antibody to methionine enkephalin may not be recognizing all sites of the peptide because the product is less accessible for detection.

Fig. 3. Electron micrographs of the cerebellar cortex of 10-day-old rat. All lissucs were processed with anti-methionine enkephalin IgG and peroxidase-conjugated goat anti-rabbit IgG. IA) MeI-ENK in the soma of a PurkiI!ie cell (PUR). Cell protuburances (pr) were frequently recorded, with postsynaptic densities (arrowheads) being markedly immunoreactive, x 11,500. (B) Internal granule neurons (nu) often exhibited a mosaic pattern ofimmunoreactivity. Some internal granule neurons (nu*) comained Met-ENK in the cytoplasm, but a far greater proportion of cells were not reactive (nu). × 9200. (C) The terminals of nerve fibers, resembling mossy fiber endings (mf), were positive for Met-ENK in the internal granule layer, lmmunoreactivity could often be observed in aggregates proximal to mitochondria and synaptic vesicles, x 19,000. (D) A glial cell (nu, nucleus), resembling an oligodendrocyle, in the medullary layer. Met-ENK was present in the cytoplasm (*), but not in the cell nucleus (nu), × 9200. (E) Higher magnification micrograph of the cytoplasm of an oligodendrocyte in the medullary layer. Aggregates of reaction product were prominent in the cytoplasm (arrow), and surrounded (arrowhead) mitochondria (m). x 20,000. 484

Opioids and brain development

Fig. 4. Electron micrographs of the MOL from the eerebellar cortex of 45day-old (A, B) and 75-day-old (C) rats. All tissues were processed with anti-methionine enkephalin IgG and peroxjda~-conjugated goat anti-rabbit IgG. (A) MetENK in the dendrite (den) of a Purkinje eeli. Intense immunoreactivity was associated with postsynaptic densities (psd) and the periphery of mitochon~a (m). x 32,500. (B) Synaptic spines related to Purkinje cell dendrites contain Met-ENK; note the intense immunoreactivity of postsynaptic densities (psd). x 32,500. (C) Immunoreactivity was not detected in the MOL of 75-day-old rats. den, Purkinje cell dendrite; psd, ~ostsynaptic density. x 22,200.

Far example, newly synthesized peptide may be in a more diffuse cytoplasmic state, or the packaging (e.g. vesicles) cannot be recognized by the antiserum. Experiments using several fixation conditions and different ~~eability agents could he helpful with regard to differential localization. Secondly, utihxation of immunogold procedures should provide a more specific localization of the antibody in contrast to that seen with peroxidase methodology as used herein. Previous studies led to the hypothesis that methionine enkephdhn is a potent growth-related For example, the administration opioid peptide. 19*40-42 of methionine enkephalin to 6-day-old rats depresses the incorporation of 13H]thymidine by EGL cells, an inhibitory action mediated by opioid receptors since concomitant injection of methionine enkephalin and the opioid antagonist, naloxone, did not affect the incorporation of radiolabeled thymidine. A similar

inhibitory effect on cell repiication is observed in neurobiastoma cells in uiuo4’ and in uitro.‘9id2Therefore, the previous light microscopic results43 and these ~mmun~lectron microscopic observations are consistent with methionine enkephalin’s ~stulat~ role as a peptide mediating cellular replication. While Met-ENK ~mmunoreactiv~ty was abundant in proliferating, migrating (translocating) and differentiating neural cells, it was not present in differentiated neural ceils. A novel finding in these studies was the striking relationship between Met-ENK and the dendritic spines and postsynaptic densities (including those on the dendritic shaft and soma) of some types of neurons. The postsynaptic densities of Purkinje and basket/s~~late neurons, but not internal granule neurons, displayed prominent Met-FNK activity. This immunoreacti~ty did not appear to be an artifact since: (I) no immunostaining was observed

486

1. S. ZAGON and P. J. MCLAUGHLIN

Fig. 5. Electron micrographs of the cerebellar cortex of lo-day-old rats. Conventionally prepared tissues of the cerebellar cortex showing structural details of representative areas of the EGL (A), MOL (B), Purkinje cell layer (C) and IGL/MED border (D). (A) In this sagittal section of the EGL. the nuclei (nu) and cytoplasm (m, mitochondria; er, endoplasmic reticulum) of external granule cells can be seen. x 18,500. (B) In this sagittal section, profiles of parallel fibers and the shaft of a Purkinje cell dendrite (den) can be recognized. Note the microtubules (arrows) in the Purkinje cell dendrite, m, mitochondria; arrowhead, synapse. x 16,500. (C) The cytoplasm and nucleus (nu) of a Purkinje cell. The cytoplasm was rich in mitochondria (m) and endoplasmic reticulum (er). Arrowhead, synapse. x 16,000. (D) An axonal process with myelin (arrowhead), microtubules (arrow), and mitochondria (m). A growth cone (gc) from a fiber can be noted. x 14,500.

in non-immune or preabsorbed controls, (2) the postsynaptic elements of adult rats did not exhibit reaction product, (3) only certain postsynaptic densities in developing rats were Met-ENK-positive, and (4) presynaptic elements (except for mossy fiber terminals) showed no immunoreactivity, eliminating the possible translocation to postsynaptic densities. While the function of methionine enkephalin in specific postsynaptic and presynaptic areas during ontogeny is presently unclear, enkephalins could play a trophic role in establishing the neuronal network. Enkephalin-like activity in the cerebellum has been examined previously. Sar et aL2’ used colchicine-

treated preparations of the adult rat (SpragueDawley) and found immunoreactive Golgi type II cells. Schulman et al.” localized leucine enkephalin in a variety of species and found immunostaining of Golgi type II cells and mossy fibers in the rat (treated and untreated with colchicine). However, the distribution of enkephalin appeared strain specific; BD-IX rats had a homogeneous distribution of enkephalin while Sprague-Dawley rats had immunoreactive cells that were most numerous in the ventral-most folia (I, II and X), and “relatively infrequent in dorsal and posterior regions” (Schulman et al., I9 p. 2408). Finally, King et aLI3 used a polyclonal

Opioids and brain development

EGL

[]

487

= met-enkaphalln

BA/ST~

~

8oms F

............

|xon

mossy fiber

~glis

Fig. 6. Schematic representation of the localization of Met-ENK in the cerebellar cortex of preweaning rats (e.g. 10 days). Met-ENK is indicated by the cross-hatched shading, except for that associated with filament-like structures which are depicted by solid lines. BA/ST, basket/stellate cell; den, dendrite; EGL, external granule cell; er, endoplasmic reticulum; f, filament; IGL, internal granule neuron; nu, nucleus; m, mitochondria; mt, microtubule; pm, plasma membrane; pr, polyribosomes; psd, postsynaptic density; sp, dendritic spine.

antisera to Met-ENK that cross-reacted with leucine enkephalin and found immunoreactive climbing fibers in the developing and adult cerebellar cortex of the opossum; colchicine treatment was reported to "improve staining of cell bodies and dendrites, but did not reveal any populations of immunoreactive cells which were not present in tissue from untreated animals" (Schulman et al., 29 p. 2407). Such immunoreactivity was localized in lobules II-VIII and X of the vermis, but was not in the intermediate cortex or hemispheres. Thus, the detection and localization of enkephalin may be dependent on a variety of factors including the species--and strain--of animal studied, the nature of the antisera used, the procedures employed (e.g. colchicine treated), and the region of the cerebellum examined. In this and an earlier studyY which did not employ colchicine, we did not observe staining of Golgi type II cells in the developing or adult cerebellum of Sprague-Dawley rats. However, mossy fiber terminals of 10- and 35-day-old rats, but not adults, were immunoreactive. Of course, this study did not represent a systematic examination of the cerebellum, and enkephalinergic cells and fibers may have been present in other regions of the adult rat cerebellum. The source of the methionine enkephalin, a neuropeptide derived from proenkephalin A, 3,7'21 in proliferating and differentiating neural cells has not been

defined. The methionine enkephalin present in the plasma and brain tissues of preweaning rats 31 may be internalized--bound or unbound to receptors--in a process that involves a variety of organelles,TM including the interaction of peptides with intracellular opioid receptors, 3° and the uptake and utilization by developing cells. Cultured neuroblastoma cells secrete methionine enkephalin, ~9,42 which serves to control cell proliferation through autocrine mechanisms. If germinative and differentiating neural cells also produce and secrete methionine enkephalin, it would then not be surprising to find immunoreactivity associated with structures (e.g. endoplasmic reticulum, ribosomes) involved in peptide biogenesis.6 Further studies using a variety of techniques (e.g. cDNA probes, electron microscopic autoradiography) are needed to shed light on the precise interfacing of enkephalin with opioid receptors in developing cells, and the traffic patterns involved in the event of autoerine regulation. CONCLUSION

While this study focused on the cerebeilar cortex, it is now clear that a specific endogenous opioid system, most likely Met-ENK-like compounds and related receptors, is involved in the growth of many brain regions. For example, disruption of endogenous opioid systems during postnatal development

488

1. S. ZAGON and P J. MCLAUGHLIN

alters ontogenesis of both the cerebrumx,9,25,34,39and hippocampus.8.9,39 /I-Endorphin-like immunoreactivity has been observed in the soma of germinal cells located at the lateral border of the ventricles at the level of the septum by light microscopic immunocytochemistry techniquesI and may represent the ontogeny of intrinsic peptidergic neurons. During the course of the present study we examined the brains of fetal and neonatal rats and detected MetENK-like immunoreactivity in germinal cells and

neuroblasts throughout the central nervous system. Thus, the transient expression of enkephalin-like activity is observed in other regions of developing brain and may serve as trophic factors regulating the proliferation and differentiation of the nervous system. Acknowledgements-This NS-20500 and NS-20623

was supported by grants from the National Institutes of Health. We thank Jeff Conforti and Steve Prouty for technical assistance. work

REFERENCES I.Bayon A., Shoemaker W. J., Bloom F. E., Mauss A. and Guillemin 2. 3. 4. 5. 6.

7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

18. 19. 20. 21. 22.

23. 24. 25. 26. 27.

R. (1979) Perinatal development of the endorphinand enkephalin-containing systems in the rat brain. Bruin Res. 179, 93-101. Bennett D. B., Spain J. W., Laskowski M. B., Roth B. L. and Coscia C. J. (1985) Stereospecific opiate-binding sites occur in coated vesicles. J. Neurosci. 5, 3010-3015. Comb M., Seeburg P. M., Adelman J., Eiden L. and Herbert E. (1982) Primary structure of the human met- and leu-enkephalin precursor and its mRNA. Nature 295, 663466. Coyle J. and Pert C. B. (1976) Ontogenetic development of opiate receptors in rodent brain. Brain Rex 118, 1577160. DiFiglia M. and Aronin N. (1984) Immunoreactive leu-enkephalin in the monkey hypothalamus including observations on its ultrastructural localization in the paraventricular nucleus. J. camp. Neurol. 225, 3133326. Eipper B. A., May V., Cullen E. I., Sato S. M., Murthy A. S. N. and Mains R. E. (1987) Cotranslational and posttranslational processing in the production of bioactive peptides. In Psychopharmacology; The Third Generalion ;y Progress (ed. Meltzer HT Y.), pp. 385400. Raven Press, New York. Gubler Il., Seeburg P., Seeburg B. J., Gage L. P. and Udenfriend S. (1982) Molecular cloning establishes proenkephalin as precursor of enkephalin-containing peptides. Nurure 295, 206208. Hauser K. F., McLaughlin P. J. and Zagon I. S. (1987) Endogenous opioids regulate dendritic growth and spine formation in developing rat brain. Brain Res. 416, 157-161. Hauser K. F., McLaughlin P. J. and Zagon I. S. (1989) Endogenous opioid systems and the regulation of dendritic growth and spine formation. J. camp. Neural. 281, 13322. Hess G. D. and Zagon I. S. (1983) Endogenous opioid systems in neural development: ultrastructural studies in the cerebellar cortex of infant rats. Brain Res. Bull. 20, 473478. Inagaki S., Kubota Y. and Kito S. (1986) Ultrastructural localization of enkephalin immunoreactivity in the substantia nigra of the monkey. Bruin Res. 362, 171-174. Kapadia S. E. and Kapadia C. R. (1986) Ultrastructure and localization of substance P and met-enkephalin immunoreactivity in the human fetal antrum. Cell Tiss. Res. 243, 289-297. King J. S., Ho R. H. and Bishop G. A. (1986) Anatomical evidence for enkephalin immunoreactive climbing fibres in the cerebellar cortex of the opossum. J. Neurocytol. 15, 5455549. Klein C., Levy R. and Simantov R. (1986) Subcellular compartmentation of opioid receptors: modulation by enkephalin and alkaloids. J. Neurochem. 46, 1137~1144. Knodel E. L. and Richelson E. (1980) Methionine-enkephalin immunoreactivity in fetal rat brain cells in aggregating culture and in mouse neuroblastoma cells. Bruin Res. 197, 565-570. Loughlin S. E., Massamiri T. R., Kornblum H. 1. and Leslie F. M. (1985) Postnatal development of opioid systems in rat brain. Neuropeptides 5, 469472. Marzullo G. and Friedhoff A. J. (1980) Opiate binding and cerebroside sulfates in brain and lung of developing chick. In Endogenous and Exogenous Opiate Agonists and Anrugonists (ed. Way E. L.), pp. 47-50. Pergamon Press, New York. Maseda C.. Aguado E. G., Mena M. A. and De Yevenes J. G. (1983) Ontogenetic development of /?-endorphin immunoreactivity in rat brain regions. Neurosci. Left. Suppl. 14, S237. McLaughlin P. J. and Zagon I. S. (1987) Endogenous opioid systems regulate growth of neurotumor cells in culture. Sot. Neurosci. Abstr. 13, 210. Neale J. H., Barker J. L., Uhl G. R. and Snyder S. H. (1979) Enkephahn-containing neurons visualized in spinal cord cultures. Science 201, 467469. Noda M., Furutani Y., Takashashi H., Toyosato M., Hirose T., Kayonna S., Nakanishi S. and Numa S. (1982) Cloning and sequence of cDNA for bovine adrenal preproenkephalin. Nature 295, 202-206. Osamura R. Y., Tsutsumi Y. and Watanabe K. (1984) Light and electron microscopic localization of ACTH and pro-opiomelanocortin-derived peptides in human developmental and neoplastic cells. J. Histochem. Cvtochem. 32, 885--893. Pickel V. M., Joh T. H., Reis D. J., Leeman S. E. and Miller R. J. (1979) Electron microscopic localization of substance P and enkephalin in axon terminals related to dendrites of catecholaminergic neurons. Brain Res. 160, 387400. Pickel V. M., Sumal K. K., Beckley S. C., Miller R. J. and Reis D. J. (1980) Immunocytochemical localization of enkephalin in the neostriatum of rat brain: a light and electron microscopic study. J. camp. Neural. 189, 721-740. Reyes J., Lewis D. and Diamond M. C. (1987) Regional effects of opioid receptor blockade on cortical thickness. Sot. Neurosci. Absrr. 13, 1118. Riederer B. M., Zagon I. S. and Goodman S. R. (1986) Brain spectrin(240/235) and brain spectrin(240/235E): two distinct spectrin subtypes with different locations within mammalian neural cells. J. Cell Biol. 102, 208882097. Sar M.. Stumpf W. E., Miller R. J., Chang K. J. and Cuatrecasas P. (1978) Immunohistochemical localization of enkephalin in rat brain and spinal cord. J. camp. Neurol. 182, 17-38.

Opioids and brain development

489

28. Schmahl W. G., Plendl J. and Reinohl-Kompa S. (1987) Effects of naltrexone in postnatal rats on the recovery of disturbed brain and lymphatic tissues after X-irradiation or ethylnitrosourea treatment in utero. Res. Commun. Chem. Pathol. Pharmac. 55, 89-99. 29. Schulman J. A., Finger T. E., Brecha N. C. and Karten H. J. (1981) Enkephalin immunoreactivity in Golgi cells and mossy fibres of mammalian, avian, amphibian and teleost cerebellum. Neuroscience 6, 2407-2416. 30. Slotkin T. A., Burwell B. and Lau C. (1980) An intracellular opiate receptor. Life Sci. 27, 1975-1978. 31. Tsang D., Ng S. C., Ho K. P. and Ho W. K. K. (1982) Ontogenesis of opiate binding sites and radioimmunoassayable jI-endorphin and enkephalin in regions of rat brain. Deul Brain Res. 5, 257-261. 32. Zagon I. S. (1972) Light and electron microscopic observations on the postnatal development of rat cerebellum.

Doctoral Dissertation University of Colorado Medical Center, Denver. 33. Zagon I. S., Higbee R., Riederer B. M. and Goodman S. R. (1986) Spectrin subtypes in mammalian brain: an immunoelectron microscopic study. J. Neurosci. 6, 2977-2986. 34. Zagon I. S. and McLaughlin P. J. (1983) Increased brain size and cellular content in infant rats treated with an opiate antagonist. Science 221, 671673. 35. Zagon I. S. and McLaughlin P. J. (1983) Naltrexone modulates growth in infant rats. Life Sci. 33, 2449-2454. 36. Zagon I. S. and McLaughlin P. J. (1984) Naltrexone modulates body and brain development in rats: a role for endogenous opioids in growth. Life Sci. 35, 2057-2064. 37. Zagon I. S. and McLaughlin P. J. (1985) Naltrexone’s influence on neurobehavioral development. Pharmac. Biochem. Behao. 22, 441448. 38. Zagon I. S. and McLaughlin 39. 40. 41. 42.

P. J. (1986) Opioid antagonist (naltrexone) modulation of cerebellar development: histological and morphometric studies. J. Neurosci. 6, 14241432. Zagon I. S. and McLaughlin P. J. (1986) Opioid antagonist-induced modulation of cerebral and hippocampal development: histological and morphometric studies. Devl Brain Res. 28, 233-246. Zagon I. S. and McLaughlin P. J. (1987) Endogenous opioid systems regulate cell proliferation in the developing rat brain. Brain Res. 412, 68-72. Zagon I. S. and McLaughlin P. J. (1989) Opioid antagonist modulation of murine neuroblastoma: a profile of cell proliferation and opioid peptides and receptors. Brain Res. 480, 16-28. Zagon I. S. and McLaughlin P. J. (1989) Endogenous opioid systems regulate growth of neural tumor cells in culture.

Brain Res. 409, 14-25. 43. Zagon I. S., Rhodes R. E. and McLaughlin P. J. (1985) Localization of enkephalin immunoreactivity

in germinative cells of developing rat cerebellum. Science 227, 1049-1051. 44. Zagon I. S., Rhodes R. E. and McLaughlin P. J. (1986) Localization of enkephalin immunoreactivity in diverse tissues and cells of the developing and adult rat. Cell Tiss. Res. 246, 561-565. (Accepted 30 August 1989)