THE JOURNAL OF COMPARATIVE NEUROLOGY 312451-466 (1991)

PNS-CNS Transitional Zone of the First Cranial Nerve R. DOUCETTE Department of Anatomy, College of Medicine, University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N OW0

ABSTRACT This study examined the ultrastructure of the region of transition where fascicles of olfactory axons leave the peripheral nervous system (PNS) to enter the central nervous system (CNS), the so-called PNS-CNS transitional zone. Adult rats were transcardially perfused with a solution of 1%glutaraldehyde and 1%paraformaldehyde, decapitated, and the heads decalcified over a period of several weeks in a solution of 1% glutaraldehyde in 0.1 M tetrasodium ethylenediamine tetraacetic acid; the latter solution was changed daily. It was found that astrocytes did not form the glia limitans at the nerve entry zone, unlike the situation that exists in other cranial and spinal nerves. Rather, the glia limitans in this region of the olfactory bulb was formed by a special type of glial cell, referred to as an ensheathing cell. Ensheathing cells are found only in the nerve fiber layer of the olfactory bulb. They possess a mixture of Schwann cell and astrocytic features and are more likely to be of placodal than of CNS origin. The meningeal coverings of the olfactory nerve rootlets and of the olfactory bulb are also described and the functional implications of the findings discussed. Key words: ensheathing cells, astrocytes, olfactory bulb

In the injured adult mammalian nervous system, neurons attempt to replace lost axonal parts by axonal regeneration. In the central nervous system (CNS), this regrowth of axons is usually unsuccessful, whereas in the peripheral nervous system (PNS), the regenerating axons often reinnervate the sensory and effector organs so that almost normal function is restored (Berry, '79; Clemente, '64; Kiernan, '79). This contrast between the PNS and CNS is exemplified by the growth of sensory axons as they approach the dorsal root entry zone of the spinal cord. Namely, the regeneration of dorsal root axons following a transection central to the ganglion occurs only along the PNS portion of their course, with the majority of the injured axons failing to regenerate into the spinal cord (Bignami et al., '84; Carlstedt, '88; Liuzzi and Lasek, '87; Reier, '86; Reier and Houle, '88; Reier et al.; '89; Stensaas et al., '87). This dichotomy between axonal growth in the PNS and CNS does not apply, however, to the axons that constitute the olfactory nerve, because olfactory axons are able to grow within both the PNS and CNS (Barber, '82; Doucette et al., '83;Graziadei and Monti Graziadei, '78a,b; Monti Graziadei and Graziadei, '79a). It has been known for some time that it is the axons of newly generated (i.e., immature) olfactory receptor neurons that grow within the CNS, not the regenerating axons of mature chemoreceptors (Barber, '82; Graziadei and DeHan, '73; Graziadei and Monti Graziadei, '78a,b; Graziadei et al., '80; Monti Graziadei and

o 1991 WILEY-LISS, INC.

Graziadei, '79b). Although the immaturity of these olfactory receptor neurons may be important to the successful growth of their axons within the CNS, Doucette et al. ('83) have shown that olfactory axons are not able to grow through astrocytic scar tissue. Could a difference in the cellular structure of the PNSCNS junction of spinal cord and olfactory nerve rootlets contribute to the successful entry of growing olfactory axons, but not of regenerating dorsal root axons, into the adult CNS? For instance, when nonolfactory primary sensory axons enter the CNS, their ensheathment abruptly changes from a Schwann cell cytoplasmic investment to one involving either astrocytes or oligodendrocytes (Berthold and Carlstedt, '77; Maxwell et al., '69; Steer, '71). Primary olfactory axons, however, are ensheathed by the same glial cell type in both the PNS (i.e., in the olfactory fascicles of the lamina propria) and the CNS (i.e., in the olfactory fascicles of the nerve fiber layer of the bulb). These glial cells, called ensheathing cells (Doucette, '84, '90a,b), exhibit phenotypic features of both Schwann cells and astrocytes (see Discussion). In the accessory olfactory bulb, Raisman ('85) has described what appears to be an homologous type of glial cell, which in that instance ensheaths Accepted June 12,1991 Address reprint requests to Dr.R. Doucette, Dept. of Anatomy, College Medicine, Univ. of Saskatchewan. Saskatoon, Saskatchewan, Canada S7N OWO.

R. DOUCETTE

452 vomeronasal axons in both the PNS and CNS. In the nerve fiber layer of the main olfactory bulb, astrocytes are confined to the spaces between olfactory fascicles, and oligodendrocytes are absent (Doucette, '84, '86, '9Oa,b). Ensheathing cells can be distinguished from astrocytes by virtue of the greater electron density of the cytoplasmic matrix of the former cells and by the fact that the intermediate filaments of ensheathing cells are not grouped into the large filament bundles that are so characteristic of astrocytes. However, astrocytes and ensheathing cells both contribute to the formation of the glia limitans of the bulb (Doucette, '84, '86, '90a,b). The purpose of this study was to examine the ultrastructure of the region of transition where fascicles of olfactory axons leave the PNS to enter the CNS. This transitional zone occurs at the level of a n entire olfactory nerve rootlet, not at the level of individual axons as happens, for example, in dorsal rootlets of the spinal cord. This is the first time that the PNS-CNS transitional zone of the first cranial nerve has been studied with the electron microscope; a preliminary account of data has appeared (Doucette '90b). The major finding is that only these ensheathing cells form the glia limitans where the olfactory nerve rootlets merge with the nerve fiber layer of the olfactory bulb.

MATERIALS AND METHODS Five female, albino Wistar rats, weighing between 150250 grams, were used in this study. The rats were anesthetized with sodium pentobarbital (40 mg/Kg body weight, i.p). Prior to perfusion fixation, the abdominal cavity was opened and 1ml of 1% sodium nitrite solution was injected into the inferior vena cava of each animal. The rats were perfused through the heart with a freshly prepared solution containing 1%glutaraldehyde and 1% paraformaldehyde in a 0.12 M sodium and potassium phosphate buffer (pH 7.4). The animals were decapitated, the bone was removed from the dorsal surface of both olfactory bulbs, and the heads were immersed in the same fixative solution for at least 18 hours (at 4°C). For three of the animals, the fixative solution was replaced the following morning with a freshly prepared decalcifying fluid (Baird et al., '68). This decalcifying fluid consisted of 1%glutaraldehyde in 0.1 M tetrasodium ethylenediamine tetraacetic acid (EDTA, pH 7.3). Decalcification was done in the cold, and the solution was changed every 24 hours on days 1-4, and twice on the fifth day. This sequence of changes was repeated beginning on day 8. Decalcification of the tissue with this technique routinely took about 6 weeks to complete. The decalcified tissue was trimmed to include only the cribriform plate, the adjoining parts of the olfactory mucosae, and the olfactory bulbs. These specimens were cut into serial sagittal or coronal slices approximately 0.5 mm thick; each tissue slice was trimmed further so that it was no more than 3 mm in any dimension. The tissue samples were washed in 0.1 M phosphate buffer, postfixed in 2% osmium tetroxide, and stained en bloc with saturated, aqueous uranyl acetate (at 4°C). They were then dehydrated in a graded series of ethanols, cleared in propylene oxide, and embedded in Epon-Araldite (Marivac Ltd., Halifax, N.S.). To control for the histological quality of the decalcified tissue, two additional animals were perfused through the heart, decapitated, and their heads immersed in fresh fixative overnight as described for the other three animals. The next morning the olfactory bulbs of these two rats were

dissected from the cranial cavity. One olfactory bulb from each animal was cut into serial sagittal slices approximately 0.5 mm thick and embedded in Epon-Araldite according to the protocol described above. The two remaining olfactory bulbs were placed in freshly prepared decalcifying solution; the changing of solutions and the total length of time in the decalcifying fluid were identical to that required to decalcify the entire head. Six weeks later, the bulbs were serially sectioned in the sagittal plane and embedded in EponAraldite. One-micron-thick sections were cut with a glass knife and stained with 1%toluidine blue in 1%borax for light microscopy. Ultrathin sections of selected areas were then cut with a diamond knife, mounted on formvar-coated slot (1 x 2 mm) grids, and stained with lead citrate and uranyl acetate.

RESULTS Each Epon-Araldite embedded block of tissue was trimmed so that its dimensions were no more than 0.75 mm by 1.75 mm. For the tissue that contained part of the cribriform plate, each trimmed block contained one or more foramina of the cribriform plate and the olfactory nerve rootlets, the ventral surface of the olfactory bulb lying on the intracranial side of the cribriform plate, and the olfactory mucosa lining the dorsocaudal aspect of the nasal cavity (see Fig. 1).Those pieces of tissue that were obtained from the lateral and dorsal surfaces of the bulb were cut so that the long axis of the trimmed bock was always parallel to the pial surface; thus the pial surface of these blocks could be studied over a length of almost 2 mm.

Entrance to the cranial cavity Within the lamina propria, each olfactory fascicle was completely covered by a basal lamina, which separated the ensheathing cells (Doucette, '90a) from the collagen fibers and fibroblasts of the surrounding connective tissue sheath (Fig. 2a). The connective tissue sheaths remained closely applied to the surfaces of the olfactory fascicles as the latter structures entered the foramina of the cribriform plate (Fig. 2b). Several of these olfactory fascicles and their connective tissue sheaths fused to form the much larger fascicles that usually almost completely filled each of the many foramina of the cribriform plate; the subarachnoid space of the cranial cavity (see Fig. 3b) extended for a short distance into each foramen (Fig. 3a). The olfactory nerve rootlets comprised those portions of these olfactory fascicles that travelled through this subarachnoid space to bridge the gap between the foramina of the cribriform plate and the nerve fiber layer along the ventral surface of the olfactory bulb. The connective tissue sheaths of these rootlets fused with both the dural and leptomeningeal linings of the cranial cavity.

Fusion of rootlets with ventral surface of olfactory bulb Each olfactory nerve rootlet normally travelled within the subarachnoid space for only a very short distance before entering the nerve fiber layer of the olfactory bulb. The basal lamina that covered the external surface of each of these rootlets was continuous with that covering the external surface of the glia limitans of the nerve fiber layer. Where fissures made deep invaginations in the ventral surface of the olfactory bulb, the cytoplasmic processes of cells of the pia mater faithfully followed the contours of the

FIRST CRANIAL NERVE PNS-CNS TRANSITIONAL ZONE

Fig. 1. These micrographs show either the olfactory nerve rootlets within the foramina of the cribriform plate (a and h) or the olfactory nerve fiber layer on the ventral aspect of the olfactory bulb (c and d). (a) Olfactory nerve rootlet (R) is travelling through one of the foramina of the cribriform plate (CP). The suharachnoid space (SAS)extends for a short distance into this foramen. The large arrow points towards the cranial cavity. (b) Olfactory nerve rootlets (R) of two adjacent bony

453

foramina enter the olfactory nerve fiber layer (ONL) along the ventral surface of the olfactory bulb. ( c ) and d) The olfactory nerve fiber layer (ONL) and glomerular layer (GL) of the bulb a short distance away from an entering olfactory nerve rootlet. The asterisk in (d) denotes a small olfactory nerve rootlet that has not yet entered the olfactory nerve fiber layer. D - dura mater; OE - olfactory epithelium. Bar = 50 pm.

454

R. DOUCETTE

Fig. 2. The peripheral olfactory fascicles of the lamina propria (a) and cribriform plate (b).The ensheathing (En) cells of these fascicles ensheath large numbers of olfactory axons (Ax) within the same mesaxon. These ensheathing cells are apposed to a basal lamina, which

separates them from the fibroblastic (F)connective tissue sheath that surrounds the entire fascicle. Arrowheads point to the basal lamina of the ensheathing cells. Bars = 1 km.

glia limitans, being always separated from the latter structure by a basal lamina and bundles of collagen fibers. This arrangement imparted a highly convoluted appearance to the ventral surface of the bulb. With few exceptions (seebelow), the cells that formed the glia limitans along the ventral surface of the bulb had no compact bundles of intermediate filaments in their cytoplasms and provided ensheathment to the entering olfactory axons (Fig. 4). The cytoplasmic matrices of ensheathing cells were of variable electron density (Figs. 4,5).On the basis of these morphological criteria, these glial cells were indistinguishable from the “ensheathing cells” that could be found in the nerve fiber layer along more lateral, medial, and dorsal portions of the olfactory bulb (Doucette, ’84, ’90b). Glial cells that had an electron lucent cytoplasmic matrix and in which were found large compact bundles of intermediate filaments (i.e., typical astrocytes) did occasionally form part of the glia limitans on the ventral surface of the olfactory bulb (Fig. 5a), but this always occurred some distance away from the fusion of a rootlet with the olfactory nerve fiber layer. Furthermore, there were no astrocytic perivascular end feet surrounding the blood vessels of the nerve fiber layer along the ventral surface of the bulb, although astrocytic cytoplasmic processes were plentiful in and around the glomerular layer (Fig. 5b). The small number of astrocytes that were present in this part of the nerve fiber layer were directly apposed to ensheathing cells with no intervening basal lamina (Fig. 5). Thus ensheathing cells had the almost exclusive role of forming the glia limitans along the ventral surface of the olfactory bulb.

The dural lining of the intracranial side of the cribriform plate was sometimes quite thick, as a result both of the loose arrangement of the constituent cells, and of the number of cell layers (Fig. 6a). Small branches of nonolfactory peripheral nerves and the occasional peripheral olfactory fascicle were embedded amongst the cells of the dura mater (Figs. 3a, 6a). Lining the deep (i.e., subarachnoid) surface of the dura mater and intermingling to some extent with the more superficially situated electron dense dural cells were one or more layers of the less electron dense cells of the arachnoid mater. The latter cells were directly apposed to the more superficially situated cells of the dura mater, although no specializedjunctions were seen between the apposed areas of membrane. Cells of the arachnoid mater also covered the subarachnoid surface of the pia mater, although the former cells did not accompany those of the pia mater into the depths of the many fissures along the ventral surface of the bulb. Furthermore, the cytoplasmic processes of cells of both layers of the arachnoid mater formed trabeculae that stretched across the subarachnoid space (Figs. 6b, c).

Fusion of rootlets with lateral and dorsal surfaces of olfactory bulb The virtual absence of astrocytes contributing to the formation of the glia limitans on the ventral surface of the olfactory bulb was unlikely to be due to the prolonged treatment (6 weeks) with decalcifying fluid, for such cells were always seen within the glomerular layer and within deeper parts of the nerve fiber layer (see Fig. 5b) of the same

FIRST CRANIAL NERVE PNS-CNS TRANSITIONAL ZONE

Fig. 3. (a)Olfactory nerve rootlet (R)within the cribriform plate near the intracranial side of the foramen. This picture was taken from a block of tissue similar to that of Figure la. This micrograph includes that part of the subarachnoid space (SAS)that extends into one of the foramina of the cribriform plate. The subarachnoid space separates the connective tissue sheath of the rootlet from that of the periosteal lining of the ethmoid bone. The large arrow points towards the cranial cavity.

455

(b)Olfactory nerve fiber layer (ONL) on the ventral surface of the olfactory bulb. A large olfactory fascicle occupies the entire upper portion of the micrograph. The ensheathing (En) cells of this fascicle form the glia limitans of the nerve fiber layer. Ax - olfactory axons; F fibroblast; Ob - osteoblast; 0 s - osteoid; PA - pia-arachnoid. Bars = 10 pm.

456

Fig. 4. The glia limitans along the ventral surface of the olfactory bulb. (a)The glia limitans in this region is formed by an ensheathing (En) cell, which is separated from the cells of the pia-arachnoid (PA) by a basal lamina and collagen fibers. (b) This micrograph shows two adjacent ensheathing (En) cells contributing to the formation of the glia

R. DOUCETTE

limitans. Note the absence of large compact bundles of intermediate filaments in the cytoplasm of the cells in both (a) and (b). Arrowheads denote the location of the basal lamina of the glia limitans. Bars = 1 pm.

FIRST CRANIAL NERVE PNS-CNS TRANSITIONAL ZONE

Fig. 5. Astrocytes in the nerve fiber layer of the ventral surface of the bulb. (a)Some distance away from the entering olfactory nerve rootlets (see Fig. lc, d) astrocytes (star)occasionally contributed to the formation of a glialimitans. This micrograph also contains two electron dense ensheathing (En) cells. Arrowheads denote the location of the

457

basal lamina of the glia limitans. (b)The deep portion of the nerve fiber layer adjacent to the glomerular layer. An interfascicular astrocytic process (star) is running through the left hand corner of the micrograph. Ax-olfactory axons; GL-neuron of glomerular layer. Bars = 1 +m.

458

Fig. 6. The meninges that line the intracranial surface of the ethmoid bone. (a)The dura mater comprises multiple overlapping layers of fibroblasts (F), which are surrounded by bundles of collagen fibers and in which are embedded blood vessels and small branches of nonolfactory peripheral nerves. (b and ci In some regions of the subarachnoid space, arachnoid trabeculae (T)can be seen extending

R. DOUCETTE

from the inner surface of the dura mater towards the pial surface of the olfactory nerve fiber layer. The cells of the arachnoid mater have a more electron lucent cytoplasmic matrix than do the fibroblasts of the adjacent dura mater. E = ensheathing cells; Ob - osteoblast. Bars = 10 Km in (a) and (b), and 5 Fm in (c).

FIRST CRANIAL NERVE PNS-CNS TRANSITIONAL ZONE tissue sections. Nevertheless, to confirm that the decalcification technique per se had no adverse effect on the glial cell morphology of this part of the nerve fiber layer, the olfactory bulbs of two transcardially perfused adult rats were removed from the skulls and prepared for viewing in the electron microscope. One olfactory bulb of each rat was cut into small pieces and prepared immediately for embedding in Epon-Araldite (see Materials and Methods). Each remaining whole olfactory bulb was placed into the EDTA/ glutaraldehyde decalcifying fluid for a total of 41 days, during which time the fluid was changed as described in Materials and Methods. A plentiful supply of both astrocytes and ensheathing cells was found in the nerve fiber layers of all four olfactory bulbs (Fig. 7), thus making it unlikely that the morphology of astrocytes in the nerve fiber layer along the ventral surface of the bulb was altered by the prolonged time that the respective pieces of tissue had spent in decalcifying fluid. The pial surface along the lateral aspect of the olfactory bulb was very convoluted, although the fissures usually were not as deep as those found along the ventral surface of the bulb. This convoluted appearance of the pial surface was due to the presence of olfactory nerve rootlets that were only now entering the olfactory nerve fiber layer (Fig. 8), giving the impression that they were only partially fused with the surface of the bulb. There were also many examples of olfactory nerve rootlets that were completely surrounded by their own basal lamina, by bundles of collagen fibers, and by cells of the pia mater (Fig. 9). It is quite likely that these latter rootlets were destined to fuse with more dorsal or caudal portions of the olfactory nerve fiber layer. The glial cells that ensheathed the olfactory axons within both types of rootlet had cytoplasmic matrices of medium electron density, in which the intermediate filaments were not grouped into large compact bundles, i.e., these were ensheathing cells. Pial cells usually intervened between each partially fused olfactory nerve rootlet and the glia limitans of the bulb, at least as far as the fusion of the respective basal laminae (Fig. 8).At the point where these rootlets entered the nerve fiber layer of the bulb, it was always ensheathing cells, not astrocytes, that formed the glia limitans. The pial surface along the dorsal aspect of the bulb was considerably smoother than it was more laterally and was covered by a very thin glia limitans that was usually no more than 1 or 2 cytoplasmic processes thick. It seemed that if only a single layer of glial processes formed the glia limitans, then the cytoplasmic matrix of these processes was always of medium electron density, did not contain bundles of intermediate filaments, and frequently ensheathed adjacent olfactory axons. However, when the cytoplasmic processes of typical astrocytes were involved, there were always at least two such processes piled one on top of the other. Totally separate and partially fused olfactory nerve rootlets were also seen in the subarachnoid space along the dorsal surface of the bulb. Small fascicles of nonolfactory peripheral nerves (containing unmyelinated axons) were occasionally seen in the subarachnoid space near the pial surface of the bulb; each Schwann cell in these fascicles was always individually enclosed within its own basal lamina. An interesting observation that was made in the course of this study was that in one location a small nonolfactory peripheral nerve fascicle had fused with the ventrolateral surface of the olfactory nerve fiber layer (Fig. 10). Neither a glia limitans nor a

459

basal lamina separated the Schwann cell of this peripheral nerve fascicle from the olfactory axons and ensheathing cells of the nerve fiber layer, even though Schwann cells are normally totally enveloped by a basal lamina. Instead, this Schwann cell contributed to the formation of the glia limitans of the bulb. Perhaps the presence of ensheathing cells in the olfactory nerve fiber layer created a favorable environment for the entry of this Schwann cell into the CNS; the origin and termination of the axons of this fascicle are unknown.

DISCUSSION The experimental paradigm that i s usually used to injure olfactory axons is one in which the nerve fibers are transected at the point where they enter the CNS. For this reason, it is important to study both the normal and the pathological anatomy of the PNS-CNS border region in olfactory nerve rootlets. The present study examined the normal anatomy of this transitional zone. It was noted that the glia limitans along the ventral surface of the olfactory bulb was formed almost exclusively by ensheathing cells and that there were not very many astrocytes in this part of the nerve fiber layer. In other words, the cellular structure of the PNS-CNS junction in olfactory nerve rootlets (see Fig. 11)is very different from that of the dorsal root entry zone of the spinal cord. Whether the pathological anatomy of these two CNS regions is also different remains to be determined. The present description is based on the impression that the PNS-CNS border lies at the point of fusion of the olfactory nerve rootlets with the olfactory nerve fiber layer. Barber and Lindsay ('82), in contrast, state that this border actually lies between the glomerular layer and the olfactory nerve fiber layer, although they give no evidence to support their claim. However, the whole olfactory bulb (including the nerve fiber layer) is enclosed within the same meningeal coverings that surround the rest of the CNS. Thus in the olfactory bulb of adult mammals, the nerve fiber layer is actually part of the CNS, not the PNS.

The transitional zone of spinal and cranial nerve rootlets I n spinal cord and cranial nerve rootlets, the PNS-CNS transitional region consists of two main parts: (1) a CNS cone-shaped compartment that is convex peripherally and formed by astrocytes, and (2) a PNS compartment that is concave centrally and is formed by Schwann cells (Berthold and Carlsted, '77; Maxwell et al., '69; Ross and Burkel, '71; Ryu and Kawana, '85). The cytoplasmic processes of subpial astrocytes extend into the CNS compartment forming an external glial limiting membrane that is continuous with that of the spinal cord proper. Each Schwann cell of the PNS compartment is surrounded by a basal lamina and collagen fibers. At the PNS-CNS junction, each Schwann cell basal lamina is continuous with that of the glia limitans of the subpial astrocytes. Axons are the only structures to pass through the transitional region of spinal cord and cranial nerve rootlets; in the PNS compartment they are ensheathed by Schwann cells, and on the CNS side by oligodendrocytes or astrocytes. These axons pass through pores or openings in the glia limitans of the transitional region; for myelinated axons the myelin-forming cell abruptly changes from the Schwann cell to the oligodendrocyte. The relationships of unmyelinated sensory axons to adjacent glial cells also

460

R. DOUCETTE

Fig. 7. The glia limitans on the dorsal surface of the olfactory bulb. This tissue was obtained from one of the olfactory bulbs that had been left in the decalcifying fluid for 6 weeks. Both astrocytes (in a) and ensheathing cells (in b) contributed to the formation of the glia limitans in these tissue samples. En-ensheathing cell. Bars = 1pm.

FIRST CRANIAL NERVE PNS-CNS TRANSITIONAL ZONE

Fig. 8. An olfactory nerve rootlet (R)that has partially fused with the olfactory nerve fiber layer (ONL) is shown in (a).The point of fusion of this fascicle with the nerve fiber layer is shown in (b).Above the arrowheads the fascicle is still covered by its own basal lamina, which is

461

separated from the basal lamina of the glia limitans by the cytoplasmic process of a pial cell. Below the arrowheads the fascicle is continuous with one of the central olfactory fascicles of the nerve fiber layer. SAS subarachnoid space. Bars = 5 pm in (a), 2 pm in (b).

462

Fig. 9. An example of an olfactory nerve rootlet that has not yet fused with the nerve fiber layer of the olfactory bulb. (a)This large rootlet (R) was seen in the subarachnoid space (SAS) along the dorsolateral surface of the olfactory bulb. The olfactory nerve fiber layer (ONL) can be seen in the left hand corner of the micrograph. (b)A

R. DOUCETTE

magnified view of the same rootlet shown in (a) to show how the cytoplasmic process of a pial cell separates the basal lamina of the rootlet from that of the glia limitans. A similar olfactory nerve rootlet was seen previously in Figure Id. E n - ensheathing cell. Bars = 10 wm in (a),1 wm in (b).

FIRST CRANIAL NERVE PNS-CNS TRANSITIONAL ZONE

Fig. 10. The pial surface of the nerve fiber layer along the ventrolateral surface of the olfactory bulb. (a)Small branch of a nonolfactory peripheral nerve (PN) fascicle is present within the subarachnoid space (SAS)in the upper portion of this micrograph, being totally surrounded by its own basal lamina, and at least partially enveloped by the cells of the pia-arachnoid. A similar small peripheral nerve fascicle is embedded within the nerve fiber layer, as can be seen in the lower right hand

463

corner of the micrograph. (b)At higher magnification, the latter fascicle is seen to be covered by a basal lamina (arrowheads) only on the side that faces the subarachnoid space. Note that in contrast to the ensheathment of olfactory axons these unmyelinated nonolfactory axons lie within individual pockets of the Schwann cell cytoplasm. En ensheathing cell; Nu - nucleus of Schwann cell. Bars = 2 pm.

464

R. DOUCETTE

Fig. 11. The PNS-CNS transitional zone of the first cranial nerve. Three olfactory nerve rootlets (R) extend through the subarachnoid space (SAS)to merge with the nerve fiber layer (ONL) along the ventral surface of the olfactory bulb. This diagram shows how the astrocytes (As)are excluded from forming a glia limitans at the nerve entry zone; it

also depicts the finding that there is no basal lamina between the astrocytes and the ensheathing (En) cells of the ONL. However, a basal lamina does separate both astrocytes and ensheathing cells from the leptomeningeal (LM) cells. Not shown here is how the subarachnoid space extends into the foramina of the cribriform plate. D - dura mater.

abruptly changes at the PNS-CNSjunction, the s o n s being ensheathed in single pockets of the Schwann cells in the PNS, whereas centrally large numbers of them are enclosed within the same astrocytic process. Thus the ensheathment of olfactory axons on the CNS side of the transitional zone of the primary olfactory pathway is quite different from that of other sensory nerve rootlets. Upon entering the CNS the olfactory axons are ensheathed by neither oligodendrocytesnor astrocytes (Doucette, '84, '86, '90a,b). In fact, oligodendrocytes are absent from the nerve fiber layer altogether, whereas astrocytes are confined, for the most part, to the interfascicular spaces of the nerve fiber layer along the medial, lateral, and dorsal surfaces of the olfactory bulb. The ensheathment of olfactory axons within the CNS, as in the PNS, is performed by ensheathing cells (Doucette, '84, '90a,b). Therefore, the relationship of unmyelinated olfactory axons to adjacent glial cells does not abruptly change at the PNS-CNS junction. This situation is in marked contrast to the abrupt transition that is seen in spinal and cranial nerve rootlets where the glial cell immediately changes from a Schwann cell to either an oligodendrocyteor an astrocyte. Ensheathing cells have a mixed phenotype, resembling to some extent both astrocytes and Schwann cells (Doucette, '90a,b). On the one hand, the intermediate filaments of ensheathing cells comprise central type glial fibrillary acidic protein (Barber and Dahl, '87), which is not expressed by

nonolfactory Schwann cells (Jessen and Mirsky, '85; Jessen et al., '84; Mirsky and Jessen, '86). At the same time, ensheathing cells express the 217c antigenic epitope (Fields and Dammerman, '851, which is found on the low affinity nerve growth factor receptor (Stemple and Anderson, '911, and the Ll/Ng-CAM cell adhesion molecule (Miragall et al., '88, '89), neither of which are present on the surfaces of astrocytes (Fields and Dammerman, '85; Linnemann and Bock, '89). In contrast, ensheathing cells are derived from glial progenitor cells (Farbman and Squinto, '85; Doucette, '89; Marin-Padilla and Amieva, '89) that are more likely to be of placodal (i.e., olfactory) than of neural crest or neural tube origin.

Implications of the findings The scar tissue that forms when brain tissue is injured (Reier, '86; Reier and Houle, '88; Reier et al., '89) places a potential obstacle in the path of growing olfactory axons. Although this scar tissue does not form an impenetrable barrier along the extracranial portion of the olfactory nerve (Bedini et al., '76; Cancalon, '87; Oley et al., '75) or when it is formed near the PNS-CNS border of the mammalian olfactory bulb (Doucette et al., '83; Graziadei and Monti Graziadei, '80; Monti Graziadei et al., '80), the astrocytic scar that forms when these axons are incised within the CNS is not penetrated by olfactory axons (Doucette et al.,

FIRST CRANIAL NERVE PNS-CNS TRANSITIONAL ZONE '83). It would appear, then, that in spite of the immaturity of olfactory neurons, their axons display a differential ability to grow through tissue that has been injured depending on whether the injury has occurred in the PNS or in the CNS. This inability of the olfactory axons to penetrate a CNS stab wound is likely due to the different cellular andlor extracellular structure of the resulting astrocytic scar, because olfactory axons can grow through regions of the CNS in which reactive astrocytes are busy cleaning up the debris of Wallerian degeneration (Doucette, '86; Doucette et al., '83). The presence of ensheathing cells, but not astrocytes, at the PNS-CNS transitional region of the olfactory nerve rootlets may well be important to the successful entry of growing olfactory axons into the CNS (Doucette '90a,b). In other words, either an astrocytic (glial) scar does not form after intracranial transection of the olfactory nerve or the mere presence of ensheathing cells in this region is sufficient to guide the axons through the scar. Indeed, the presence of Ll/Ng-CAM and N-CAM in the plasma membrane of both ensheathing cells and immature olfactory receptor neurons (Miragall et al., '88, '89) would enable the olfactory axons to use these glial cell surfaces as substrata on which to grow. In addition, glia-derived nexin is synthesized by glial cells, perhaps even by ensheathing cells themselves, in the olfactory nerve fiber layer (Reinhard et al., '88). In addition to being a neurite-promoting factor (Guenther et al., '85; Monard et al., '73; Zurn et al., '881, glia-derived nexin is also able to modulate the degradation of the extracellular matrix (Gloor et al., '86; Knauer et al., '87; Saksela and Rifkin, '88). Ensheathing cells may also synthesize and secrete laminin (Liesi, '85). Thus the ensheathing cell may not only provide an adhesive substrate for growing olfactory axons, but may also secrete chemotropic agents (see Doucette, '90a, for a review). It may be, however, that when the olfactory axons are incised at their point of entry into the CNS, the ensheathing cells of the transitional region are replaced by astrocytes, as they are after a stab wound to more caudal parts of the nerve fiber layer (Doucette et al., '83). If the latter turns out to be true, it will be clear that, if given no other choice, olfactory s o n s can grow through an astrocytic scar to reinnervate the neurons of the olfactory bulb. Perhaps this is what happens when olfactory axons grow out of transplanted pieces of olfactory mucosa and enter the neuropil of the cerebral hemisphere (Morrison and Graziadei, '83; Monti Graziadei and Morrison, '88). In contrast, cells were also observed to grow out of these homografts, and it is likely that at least some of the migrating cells were ensheathing cells. The next step, then, is to study the histological structure of the glial scar that forms after intracranial transection of the olfactory nerve, noting in particular whether it is ensheathing cells, astrocytes, or a combination thereof that comprises the glial scar.

ACKNOWLEDGMENTS This research was supported by a grant from the Medical Research Council of Canada. I thank Drs. P.J. Reier and P. Graziadei for critically reading the manuscript and for providing useful comments and suggestions. Thanks are also due to Ms. S. Williams for her expert technical assistance, to Mr. W. Appl for caring for the animals, and to the secretarial staff of the Department of Anatomy.

465

LITERATURE CITED Baird, I.L., W.B. Winborn, and D.E. Bockman (1968) A technique of decalcification suited to electron microscopy of tissues closely associated with bone. Anat. Rec. 159:281-289. Barber, P.C. (1982) Neurogenesis and regeneration in the primary olfactory pathway of mammals. Bibl. Anat. 23t12-25. Barber, P.C., and D. Dahl (1987) Glial fibrillary acidic protein (GFAP)-like immunoreactivity in normal and transected rat olfactory nerve. Exper. Brain Res. 65t681-685. Barber, P.C., and R.M. Lindsay (1982) Schwann cells of the olfactory nerves contain glial fibrillary acidic protein and resemble astrocytes. Neurosci. 7:3077-3090. Bedini, C., V. Fiaschi, and A. Lanfranchi (1976) Olfactory nerve reconstitution in homing pigeon after resection: Ultrastructural and electrophysiological data. Arch. Ital. Biol. 114:l-22. Berry, M. (1979) Regeneration in the central nervous system. In W.T. Smith and J.B. Cavanagh (eds): Recent Advance in Neuropathology. Edinburgh: Churchill Livingstone, pp. 67-111. Berthold, C.-H., and T. Carlstedt (1977) Observations on the morphology at the transition between the peripheral and the central nervous system in the cat. 11. General organization of the transitional region in S1 dorsal rootlets. Acta Physiol. Scand., Suppl. 446t2342. Bignami, A,, N.H. Chi, and D. Dahl(1984) Regenerating dorsal roots and the nerve entry zone: An immunofluorescence study with neurofilament and laminin antisera. Exper. Neurol. 85:426436. Cancalon, P.F. (1987) Survival and subsequent regeneration of olfactory neurons after a distal axonal lesion. J. Neurocyt. 16:829-841. Carlstedt, T. (1988) Reinnervation of the mammalian spinal cord after neonatal dorsal root crush. J. Neurocyt. 17t335-350. Clemente, C.D. (1964) Regeneration in the vertebrate central nervous system. Int. Rev. Neurobiol. 6:257-302. Doucette, R. (1984) The glial cells in the nerve fiber layer of the rat olfactory bulb. Anat. Rec. 210:385-391. Doucette, R. (1986) Astrocytes in the olfactory bulb. In S. Fedoroff and A. Vernadakis (eds): Astrocytes, Vol. 1. New York: Academic Press, pp. 293-3 10. Doucette, R. (1989) Development of the nerve fiber layer in the olfactory bulb of mouse embryos, J. Comp. Neurol. 285:514-527. Doucette, R. (1990a) Glial influences on axonal growth in the primary olfactory system. Glia 3:433-449. Doucette, R. (1990b) Glial cells in the nerve fiber layer of the main olfactory bulb of embryonic and adult mammals. J. Electr. Micr. Tecbn. (in press). Doucette, J.R., J.A. Kiernan, and B.A. Flumerfelt (1983) The reinnervation of olfactory glomeruli following transection of primary olfactory axons in the central or peripheral nervous system. J. Anat. 137:l-19. Farbman, A.I., and L.M. Squinto (1985) Early development of olfactory receptor cell axons. Dev. Brain Res. 19:205-214. Fields, K.L., and M. Dammerman (1985).A monoclonal antibody equivalent to anti-rat neural antigen-1 as a marker for Schwann cells. Neurosci. 15:877-885. Gloor, S., K. Odink, J. Guenther, H. Nick, and D. Monard (1986) A glia-derived neurite promoting factor with protease inhibitory activity belongs to the protease nexins. Cell 473387-693. Graziadei, P.P.C., and R.S. DeHan (1973) Neuronal regeneration in the frog olfactory system. J. Cell Biol. 59:525-530. Graziadei, P.P.C., and G.A. Monti Graziadei (1978a) Continuous nerve cell renewal in the olfactory system. In M. Jacobson (ed): Handbook of Sensory Physiology, Vol. 9. Berlin: Springer Verlag, pp. 55-82. Graziadei, P., and G. Monti Graziadei (1978b) The olfactory system: Amodel for the study of neurogenesis and axon regeneration in mammals. In C.W. Cotman (ed): Neuronal Plasticity, New York: Raven Press, pp. 131-153. Graziadei, P.P.C., and G.A. Monti Graziadei (1980) Neurogenesis and neuron regeneration in the olfactory system of mammals. 111. Deafferentation and reinnervation of the olfactory bulb following section of the fila olfactoria in rat. J. Neurocyt. 9:145-162. Graziadei, P., M. Karlan, G. Monti Graziadei, and J. Bernstein (1980) Neurogenesis of sensory neurons in the primate olfactory system after section of the fila olfactoria. Brain Res. 186:289-300. Guenther, J., H. Nick, and D. Monard (1985) A glia-derived neuritepromoting factor with protease inhibitory activity. EMBO J. 4t19631966.

466 Jessen, K.R., and R. Mirsky (1985) Glial fibrillary acidic polypeptides in peripheral glia. Molecular weight, heterogeneity and distribution. J. Neuroimmunol. 81377-393. Jessen, K.R., R. Thorpe, and R. Mirsky (1984) Molecular identity, distrihution and heterogeneity of GFAP: An immunoblotting and immunohistochemical study of Schwann cells, satellite cells, enteric glia and astrocytes. J. Neurocyt. 13:187-200. Kiernan, J.A. (1979) Hypotheses concerned with axonal regeneration in the mammalian nervous system, Biol. Rev. 54:155-197. Knauer, D.J., R.A. Orlando, and D. Rosenblatt (1987) The glioma cellderived neurite promoting activity protein is functionally and immunologically related to human protease nexin-1. J. Cell Physiol132:318-324. Liesi, P. (1985) Laminin-immunoreactive glia distinguish regenerative adult CNS systems from non-regenerative ones. EMBO J. 4(10):2505-2511. Linnemann, D., and E. Bock (1989) Cell adhesion molecules in neural development. Dev. Neurosci. 11:149-173. Liuzzi, F.J., and R.J. Lasek (1987) Astrocytes block axonal regeneration in mammals by activating the physiological stop pathway. Science237:642645. Marin-Padilla, M., and M.R. Amieva (1989) Early neurogenesis of the mouse olfactory nerve: Golgi and electron microscopic studies. J. Comp. Neurol. 288:339-352. Maxwell, D.S., L. Kruger, and A. Pineda (1969) The trigeminal nerve root with special reference to the central-peripheral transition zone: An electron microscopic study in the macaque. Anat. Rec. 164:113-126. Miragall, F., G. Kadmon, M. Husmann, and M. Schachner (1988) Expression of cell adhesion molecules in the olfactory system of the adult mouse: Presence of the embryonic form of N-CAM. Dev. Biol. 129:516-531. Miragall, F., G. Kadmon, M. Husmann, and M. Schachner (1989) Expression of L1 and N-CAM cell adhesion molecules during development of the mouse olfactory system. Dev. Biol. 135:272-286. Mirsky, R., and K.R. Jessen (1986) The biology of non-myelin-forming Schwann cells. Ann. N.Y. Acad. Sci. 486:132-146. Monard, D., F. Solomon, M. Rentsch, and R. Gysin (1973) Glia-induced morphological differentiation in neuroblastoma cells. P.N.A.S. 70: 18941897. Monti Graziadei, G.A., and P.P.C. Graziadei (1979a) Studies on neuronal plasticity and regeneration in the olfactory system: Morphologic and functional characteristics of the olfactory sensory neuron. In E. Meisami and M.A.B. Brazier (eds): Neural Growth and Differentiation. New York: Raven, pp. 373-396. Monti Graziadei, G., and P. Graziadei (197913) Neurogenesis and neuron regeneration in the olfactory system of mammals. 11. Degeneration and reconstitution of the olfactory sensory neurons after axotomy. J. Neurocyt. 8:197-213. Monti Graziadei, G.A., and E.E. Morrison (1988) Experimental studies on

R. DOUCETTE the olfactory marker protein. 4.Olfactory marker protein in the olfactory neurons transplanted within the brain. Brain Res. 455401-406. Monti Graziadei, G.A., M.S. Karlan, J.J. Bernstein, and P.P.C. Graziadei (1980) Reinnervation of the olfactory bulb after section of the olfactory nerve in monkey (Suirniri sciureus). Brain Res. 89:343-354. Morrison, E.E., and P.P.C. Graziadei (1983) Transplants of olfactory mucosa in the rat brain. 1. A light microscopic study of transplant organization. Brains Res. 279241-245. Oley, N., R.S. Deham, D. Tucker, J.C. Smith, and P.P.C. Graziadei (1975) Recovery of structure and function following transection of the primary olfactory nerves in pigeons. J. Comp. Physiol. Psych. 88~477-495. Raisman, G. (1985) Specialized neuroglial arrangement may explain the capacity of vomeronasal axons to reinnervate central neurons. Neurosci. 14:237-254. Reier, P.J. (1986) Astrocytic scar formation following CNS injury: Its microanatomy and effects on axonal elongation. In S. Fedoroff and A. Vernadakis ieds): Astrocytes, Vol. 3. New York: Academic Press, pp. 263-324. Reier, P.J., and J.D. Houle (1988) The glial scar: Its bearing on axonal elongation and transplantation approaches to CNS repair. Advances in Neurology 47:87-138. Reier, P.J., L.F. Eng, and L. Jakeman (1989) Reactive astrocyte and axonal outgrowth in the injured CNS: Is gliosis really an impediment to regeneration? In: Neuronal Regeneration and Transplantation. New York Alan R. Liss, Inc., pp. 183-209. Reinhard, E., R. Meier, W. Halfter, G. Rovelli, and D. Monard (1988) Detection of glia-derived nexin in the olfactory system of the rat. Neuron 1:387-394. Ross, M.D., and W. Burke1 (1971) Electron microscopic observations of the nucleus, glial dome, and meninges of the rat acoustic nerve. Amer. J. Anat. 130:73-92. Ryu, K., and E. Kawana (1985) Mesencephalic root fibers of the trigeminal nervein the cat. Acta Anat. 121:197-204. Saksela, O., and D.B. Rifkin (1988) Cell-associated plasminogen activation: Reguation and physiological functions. Ann. Rev. Cell Biol. 4:93- 126. Steer, J.M. (1971) Some observations on the fine structure of rat dorsal spinal nerve roots. J. Anat. 109:467485. Stemple, D.L., and D.J. Anderson (1991) A Schwann cell antigen recognized by monoclonal antibody 217C is the rat low-affinity nerve growth factor receptor. Neuroscience Letters 124:57-60. Stensaas, L.J., L.M. Partlow, P.R. Burgess, and K.W. Horch (1987) Inhibition of regeneration: The ultrastructure of reactive astrocytes and abortive axon terminals in the transition zone of the dorsal root. Progr. Brain Res. 7k457-468. Zurn, A.D., H. Nick, and D. Monard (1988) A glia-derived nexin promotes neurite outgrowth in cultured chick sympathetic neurons. Devel. Neurosci. 10:17-24.