| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Anatomy and Cell Biology, Queen’s University, Kingston, ON, Canada; and
Department of Anatomy and Cell Biology, University of Saskatchewan, Saskatoon, SK, Canada
1 Correspondence: Department of Anatomy and Cell Biology, Queen’s University, Room 926, Botterell Hall, Kingston, ON, Canada K7L 3N6. E-mail: boydg{at}post.queensu.ca
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONTHE MYELINATING CAPACITY OF...THE MYELINATING CAPACITY OF...SUMMARYSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Key Words: OECs • remyelination • axon regeneration • Schwann cell • olfactory axon
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONTHE MYELINATING CAPACITY OF...THE MYELINATING CAPACITY OF...SUMMARYSUMMARY AND CONCLUSIONSREFERENCES |
|---|
2
3
4
5
6
7
8
9
; reviewed in refs 10
11
12
13
). Studies performed by a number of laboratories worldwide over the past decade have established the principle that after implantation into the damaged spinal cord OECs promote axon regeneration, remyelinate axons, and facilitate functional recovery (reviewed in refs 13
14
15
16
17
18
). This principle has recently been disputed by studies that have failed to replicate the original experiments, which reported that OECs can promote axon regeneration across the dorsal root entry zone after dorsal rhizotomy (19
20
21)
. Moreover, recent data have challenged the claims that OECs associate with axons and synthesize a myelin sheath, both in vitro (22)
and in vivo (23
, 24)
. One of the main factors that may be contributing to this controversy is that Schwann cells, which readily myelinate axons after CNS injury, share many antigenic and morphological characteristics of OECs. These Schwann cells not only invade the damaged CNS (23
24
25
26
27
28
29)
, but may also be contaminating the in vitro preparations of OECs before implantation. Thus, the specific role for OECs in promoting axon regeneration and remyelination may require redefinition. The purpose of this review is to provide a critical evaluation of the in vitro and in vivo evidence that addresses the myelinating capacity of OECs. |
THE MYELINATING CAPACITY OF OECs IN VITRO: EVIDENCE FOR |
|---|
|
TOPABSTRACTINTRODUCTIONTHE MYELINATING CAPACITY OF...THE MYELINATING CAPACITY OF...SUMMARYSUMMARY AND CONCLUSIONSREFERENCES |
|---|
2
3
4
5
6
7)
or in vitro (30)
, this may be due to the small caliber of the axons and not to an inherent inability to synthesize myelin (31
32
33)
. This is analogous to Schwann cells, which do not synthesize myelin when in contact with small diameter sympathetic axons, but can be induced to express a myelinating phenotype if the caliber of these axons is enlarged by increasing the size of target tissue (34)
. To test whether OECs could synthesize myelin after contact with larger diameter axons, Devon and Doucette (32)
cocultured rat embryonic day 18 (E18) OECs with E15 dorsal root ganglion (DRG) neurons. Before addition of the OECs, media was supplemented with flurodeoxyuridine/uridine to eliminate possible contaminating cells from the DRG cultures. OECs were isolated from the olfactory nerve layer of the olfactory bulb from E18 Wistar rats. At this developmental stage (Theiler stage 23), OECs are the only neural support cells of the nerve fiber layer. This point is particularly significant, as no additional purification techniques were required. Four to 5 wk after addition of OECs to the DRGs, the cultures were examined for light and electron microscopic evidence of myelin. Spindle-shaped cells immunoreactive for myelin basic protein (MBP) were observed extending processes along the length of DRG neurites. Devon and Doucette (32)
provided the first ultrastructural evidence that embryonic OECs associated with myelinated and unmyelinated axons in a manner virtually identical to Schwann cells in the peripheral nervous system. In addition, each OEC that associated with a myelinated neurite was enclosed within its own basal lamina, a condition that does not normally exist in vivo but is reminiscent of a Schwann cell-like pattern of myelination. The authors explicitly stated that these myelinating cells are not contaminating Schwann cells, as pure neuronal cultures (negative controls) contained no glial cells and showed no ultrastructural evidence of myelination.
Further evidence that embryonic OECs, and not Schwann cells, were the myelinating cell observed in vitro was provided in a subsequent study (35)
in which three sets of cultures were established: 1) DRG-OEC cocultures, 2) DRG with residual glial cells eliminated with fluorodeoxyuridine/uridine, and 3) DRG cultures with some residual clusters of Schwann cells. The three sets of cultures were fed with C10 medium, with or without the addition of L-ascorbic acid, which had initially been shown to be required for Schwann cell myelination and formation of a basal lamina (36
, 37
; cf. ref 38
). Before adding OECs to the DRG cultures, the OECs were prelabeled with PKH26, a fluorescent cell linker that becomes incorporated into cellular membranes. As seen before, cultures treated with anti-mitotic agents showed no evidence of glial cell contamination or remyelination. DRG cultures that had the residual clusters of Schwann cells contained MBP-positive cells, and Schwann cells were evident at the ultrastructural level associating with myelinated or unmyelinated axons, but only when L-ascorbic acid was added to the culture medium. When L-ascorbic acid was omitted from the culture media, Schwann cells did not associate with axons, nor did they synthesize myelin or a basal lamina. In marked contrast, myelinated neurites were observed in DRG-OEC cocultures whether or not L-ascorbic acid was added to the culture media. Furthermore, there was colocalization of the PKH26 fluorescent cell linker and MBP in cells that were observed to be aligned with DRG neurites. These results demonstrate that OECs and Schwann cells require different cellular cues within their microenvironment to initiate the intracellular machinery to synthesize molecules to make the myelin sheath. Similarly, media that support the differentiation of oligodendrocytes into a myelinating phenotype do not induce up-regulation of MBP by OECs in neuron-fee cultures (35)
. Thus, several experiments have demonstrated that, under specific coculture conditions, OECs can associate with axons, synthesize myelin and generate a basal lamina. The cellular cues required to induce this phenotype are, however, unique to OECs and not completely shared by Schwann cells or oligodendrocytes.
Other evidence that OECs can express a myelinating phenotype is derived from observations that these cells express certain myelin-associated proteins under various culture conditions. While these data are inherently more indirect (as will be described below), expression of these myelin-associated proteins is dependent on the culture conditions used and the developmental stage of the OECs. OECs isolated from neonatal rat olfactory bulb and purified using a fluorescence-activated cell sorting protocol with antibodies from the O4 hybridoma (which bind to OECs, oligodendrocytes, and Schwann cells) were found to only weakly express the peripheral myelin protein P0 and galactocerebroside (31)
. Cells with a bipolar and/or multipolar morphology migrating from adult olfactory bulb explants show "weak but unambiguous" expression of the noncompact myelin protein 2',3'-cyclic nucleotide 3'-phosphodiesterase (39)
. In the absence of a definitive marker for OECs, it is not clear whether these cells migrating away from the olfactory bulb are OECs or oligodendrocyte precursors. Moreover, Pixley (8)
failed to detect the same protein in bipolar cells in explant cultures from neonatal olfactory bulbs. A similar controversy exists over the expression of MBP by OECs. In neuron-free cultures, neither unpurified embryonic OECs (35)
nor purified neonatal OECs (31)
expressed MBP in vitro even after intracellular up-regulation of cAMP, a strong stimulus for MBP synthesis by Schwann cells (35)
. In contrast, OECs isolated from the adult rat olfactory bulb and purified on the basis of their expression of the p75 neurotrophin receptor (p75NTR) displayed robust MBP immunoreactivity (40)
.
|
THE MYELINATING CAPACITY OF OECs IN VITRO: EVIDENCE AGAINST |
|---|
|
TOPABSTRACTINTRODUCTIONTHE MYELINATING CAPACITY OF...THE MYELINATING CAPACITY OF...SUMMARYSUMMARY AND CONCLUSIONSREFERENCES |
|---|
, where the authors compared the ability of Schwann cells and OECs from adult rats to myelinate DRG neurites in coculture experiments similar to those described previously (32
, 36)
. OECs were isolated from adult Fischer rat olfactory bulbs and purified based on their expression of p75NTR as described (40)
. The DRG cultures were isolated from E15 rats and treated intermittently with fluorodeoxyuridine to kill the non-neuronal cells. As with previous studies (32
, 35)
, control cultures were maintained in which neither Schwann cells nor OECs were added. The results of this study contrasted with studies by Devon and Doucette (32
, 35)
in two main ways: 1) fluorodeoxyuridine treatment failed to eliminate all Schwann cells from all DRG cultures and there was evidence of Schwann cell myelination in a small number of control cultures that had no additional cells added, and 2) there was no evidence of myelin in OEC-DRG cocultures above what was observed in control cultures. At the ultrastructural level, OECs failed to exhibit the described "Schwann cell-like" pattern of myelination of DRG neurites. Instead, "flat meandering processes" of OECs were observed encircling the DRG neurites (22)
.
As discussed by Plant et al. (22)
, OECs were purified on the basis of their expression of p75NTR. It has been well documented that p75NTR-positive and p75NTR-negative OECs can be isolated from the olfactory nerve layer of the olfactory bulb (5
, 12
, 41)
. This strategy of "immunopanning" with antibodies against p75NTR selects only a subpopulation of OECs, which may not be responsible for the myelinating potential of this cell. Another possible reason for this discrepancy not addressed by Plant et al. (22)
is the differing age of the animals from which the OECs are harvested. Although myelinated DRG neurites were evident at the ultrastructural level when embryonic OECs were added to the cultures (32
, 35)
, similar robust myelination was not evident when adult OECs were added (22)
. It is possible that embryonic OECs represent a more plastic cell and may retain the potential to readily adopt a myelinating phenotype given the appropriate cues in their microenvironment. This hypothesis is supported by a recent study published in abstract form in which the myelinating potential of adult vs. embryonic OECs was compared (42)
. In agreement with their earlier findings, Plant and colleagues failed to detect evidence of myelination when OECs isolated from adult rats were added to DRG cultures, yet numerous MBP immunopositive myelinated segments were detectable at the light microscope in cultures of embryonic OECs.
While these aforementioned studies are consistent with the myelinating capacity of embryonic OECs in the absence of definitive ultrastructural markers, one cannot definitively exclude the possibility that Schwann cells are responsible for the ultrastructural evidence of myelinated axons in OEC-DRG cocultures. Specifically, although Devon and Doucette (32
, 35)
found no residual Schwann cells in their DRG cultures after fluorodeoxyuridine/uridine treatment, an undetectable number of Schwann cells may have remained. Addition of OECs to the coculture conditions may have provided a mitotic signal for residual Schwann cells remaining with the DRGs. Thus, at the end of the 4 wk period of DRG-OEC coculture, sufficient Schwann cells could have been present for myelination. To support the notion that the myelinated axon segments are solely due to contaminating cells, it is only in coculture conditions (e.g., with DRG neurons or olfactory bulb explants) that spindle-shaped cells have been observed to directly associate with axons, synthesize myelin, and/or express MBP (32)
. With the exception of galactocerebroside (which OECs express in the presence of elevated levels of cAMP), OECs fail to express other myelin-associated proteins in neuron-free cultures (33
, 35)
. Nonetheless, the fact that cells prelabeled with PKH26 aligned themselves along the length of neurites, and these same cells express MBP suggests at least some of the myelin sheaths were due to OEC myelination (35)
.
|
SUMMARY |
|---|
|
TOPABSTRACTINTRODUCTIONTHE MYELINATING CAPACITY OF...THE MYELINATING CAPACITY OF...SUMMARYSUMMARY AND CONCLUSIONSREFERENCES |
|---|
The myelinating capacity of OECs in vivo
Several recent and extensive reviews have described the evidence supporting the myelinating capacity of OECs in vivo (15
, 43
, 44)
, and therefore this evidence will be only briefly described here in the context of possible alternative explanations for the conclusions drawn by the authors. Much of the evidence supporting the capacity for OECs to synthesize myelin in vivo comes from experiments using the ethidium bromide-X-irradiation model of experimental demyelination developed by Blakemore (45)
. In this model, ethidium bromide is injected directly into the white matter of the spinal cord to cause an area of focal demyelination because of its cytotoxic effects on resident glia. The endogenous remyelination process is significantly delayed by exposing the damaged spinal cord to high doses of X-irradiation. The region of denuded axons within the spinal cord provides a microenvironment in which to test the remyelinating capacity of transplanted cells. An important consideration that needs to be made in the context of this model of demyelination is the possibility that Schwann cells can migrate into the demyelinated zone and form peripheral myelin-like sheaths around CNS axons. Indeed, more recent studies have confirmed the original reports that endogenous remyelination is significantly delayed (>8 wk) in the absence of implanting myelin forming cells (46
, 47)
. However, implanted cells may provide an additional cue for Schwann cell migration into the demyelinated area, either via release of paracrine chemotactic factors or by providing a permissive substrate for Schwann cell migration into the CNS.
In a recent study, Kocsis and colleagues (48)
compared the myelinating potential of Schwann cells and OECs. Cells were isolated from transgenic adult rats expressing the alkaline phosphatase gene and transplanted into the demyelinated dorsal columns of immunosuppressed rats. The major limitation of this strategy is that the alkaline phosphatase reaction must be processed on frozen sections, preventing ultrastructural identification of transplanted cells. Nonetheless, in animals that received Schwann cell transplants, many of the myelinated axon profiles stained dark blue after the chromogenic reaction, suggesting that the transplanted cells were remyelinating the denuded axons. The authors described similar peripheral-type myelination in the dorsal columns after implantation of transgenic OECs, but this pattern of alkaline phosphatase reactivity was far more diffuse, making it difficult to distinguish between axons that were myelinated by the transplanted OECs, and axons that were myelinated by host Schwann cells. Thus, in absence of an in vivo ultrastructural marker of OECs, no definitive conclusions could be drawn regarding their ability to myelinate axons.
Kocsis and colleagues addressed this issue in a subsequent investigation using OECs isolated from the olfactory nerve layer of the olfactory bulb of adult transgenic Sprague Dawley rats that constitutively express green fluorescent protein (49)
. The OECs were transplanted immediately after transection of the dorsal funiculus. The advantage of this strategy is that green fluorescent protein could be detected immunohistochemically for ultrastructural identification of labeled cells. Using this strategy, approximately half of the axons that displayed compact myelin were myelinated by green fluorescent protein-positive cells. The authors postulated that poor antibody penetration likely accounted for the remaining half of myelinated axons that were associated with green fluorescent protein-negative cells (49)
.
Another line of evidence used to exclude the possibility that remyelination is performed by Schwann cells comes from models of xenotransplantation. Specifically, porcine (50)
and human (51)
OECs have been transplanted into the rat demyelinated spinal cord. In animals that do not receive immunosuppression, remyelination of denuded CNS axons is not observed. Thus, the general conclusion is that the OECs are rejected shortly after implantation and not are able to remyelinate the axons (50
, 51
; see ref 15
). An alternative explanation for this finding is that xenogenic OECs fail to create a permissive conduit into the spinal cord for invading host Schwann cells because of their immunological rejection.
In addition to the ethidium bromide-X irradiation model of spinal demyelination, ultrastructural evidence of Schwann cell-like remyelination has been presented after implantation of neonatal OECs into the spinal cord after electrolytic lesions of the corticospinal tract (52
53
54)
and adult OECs into the transected spinal cord (55)
. In light of the robust invasion of Schwann cells after various models of spinal cord injury (23
24
25
26
27
28
29)
, host Schwann cells migrating into the spinal cord lesion are abundant, and significantly confound the ability to evaluate the myelinating capacity of OECs (23
, 24)
.
To provide the first unequivocal evidence that OECs implanted into the damaged CNS associate with myelinated axons in vivo, we infected E18 rat OECs with a retrovirus carrying the gene lacZ (24)
. This technique was chosen since lacZ-expressing cells will synthesize ß-galactosidase, a bacterial enzyme that cleaves specific substrates into electron dense reaction products that are readily identifiable in labeled cells at the ultrastructural level (56)
. OECs were implanted 1 wk after a mild clip compression injury at the T10 spinal level, and spinal cords were evaluated 3 wk later for evidence of axon regeneration and remyelination. In an attempt to confirm what a decade of research had suggested, what we found was quite the opposite. Labeled cells morphologically identical to OECs in vivo were never observed associating with unmyelinated or myelinated axons. Instead, lacZ-expressing embryonic OECs formed unique tunnel-like structures in the damaged spinal cord (Fig. 1
). Within these tunnels, we observed unlabeled cells, associating with a single myelinated axon or multiple unmyelinated axons. These unlabeled cells were morphologically identical to Schwann cells at the ultrastructural level, with small ovoid nuclei, continuous basal lamina encircling a single cell and its associated axons (57)
. Thus, these cells within the tunnels created by the implanted embryonic OECs likely represent a population of host Schwann cells that have migrated into the cystic cavity. The number of axon-Schwann cell units inside the tunnels created by implanted OECs ranged from one to several dozen. In addition to Schwann cells and axons, the larger tunnels contained capillaries.
|
The morphological characteristics of the lacZ-expressing embryonic OECs after implantation into the damaged spinal cord are strikingly similar to the ultrastructural characteristics of OECs in their native environment in the olfactory nerve (1
, 3
, 4
, 58
; Fig. 1
, Fig. 2
). These cells had large ovoid nuclei, significantly larger than adjacent Schwann cells. They extended long, thin overlapping processes to create these extensive tunnels in the CNS, which share a strong similarity to the fascicles created by OECs in the olfactory nerve in vivo (1
, 3
, 4
, 9
, 58)
. However, in the olfactory nerve there are no intervening cells, such as Schwann cells or endothelial cells of capillaries, as seen after implantation into the damaged CNS.
|
The tunnel-like arrangement of cells is not unique to our study of embryonic OEC implantation into the compressed spinal cord. In fact, many studies of OEC implantation into the damaged or demyelinated spinal cord have provided evidence of larger cells extending overlapping processes that encircle clusters of myelinated and unmyelinated axons (Table 1
). In the absence of definitive labeling of OECs before implantation, these cells on the periphery of the clusters of axons have been referred to as astrocyte-like OECs (52
, 53)
, olfactory nerve fibroblasts (9
, 54)
, olfactory nerve meningeal cells (61)
, and simply cellular elements (49)
.
|
Our data directly contradict the established view that after implantation into the damaged spinal cord OECs transform into a Schwann cell-like morphology, then begin directly associating with axons and synthesizing myelin. We and others (20
, 22
, 23)
have proposed an alternative explanation. Specifically, intraspinal implantation of OECs serves to facilitate host Schwann cell migration into the damaged CNS (Fig. 3
). We think it unlikely that the cells observed ultrastructurally that are morphologically identical to Schwann cells are OECs that have transformed into a Schwann cell-like morphology (see Table 1
). Our observations demonstrate that these cells are indeed Schwann cells that have migrated in from the host. Furthermore, given the striking similarity between the ultrastructural characteristics of OEC ensheathment of olfactory axons, the larger cells that form long, thin overlapping processes around fascicles of axons and their associated glia are likely the implanted OECs.
|
It was recently suggested that OECs must collaborate with additional cell types in order to promote maximum beneficial effects on remyelinating denuded axons within the spinal cord (61
; reviewed in ref 62
). Our data strongly agree with this hypothesis, but we suggest that the principal cell type is the Schwann cell. In collaboration with host Schwann cells, we propose that OECs contribute to a 3-dimensional matrix within the damaged spinal cord (see ref 20
). This matrix may contain growth factors and extracellular matrix molecules that create a substrate conducive for axonal regrowth and remyelination. In addition to providing this scaffolding, OECs may serve to form a protective barrier between host Schwann cells and the otherwise hostile environment of the damaged CNS (Fig. 3)
. For example, there is ample evidence that OECs intermingle with astrocyte processes and induce less chondritin sulfate proteoglycan up-regulation by astrocytes in vitro (63)
and in vivo (23
, 64)
than Schwann cells.
Origin of Invading Schwann cells
The non-neoplastic proliferation of Schwann cells (Schwannosis) in rats and humans after spinal cord injury has been described for several decades (65
66
67
68)
. In rodents, Schwann cells originating from the dorsal roots can migrate into the damaged spinal cord in several models of injury (23
, 25
26
27
28
29
, 57)
. Other sources of Schwann cells often neglected are the nonmyelinating Schwann cells that accompany sympathetic axons that innervate the larger blood vessels supplying the spinal cord (69)
. The importance of this latter source of Schwann cells is supported by the fact that pockets of Schwann cells are most often found adjacent to blood vessels of demyelinated spinal cords, particularly in the ventral aspect of the cord parenchyma (70)
.
The cues that trigger the Schwann cell migration into the damaged spinal cord remain unclear (71)
. In the damaged peripheral nervous system, a dramatic increase in Schwann cell proliferation is directly associated with the macrophage invasion of the distal nerve stump that occurs several days after injury (72)
. With the compromise of the blood-brain barrier that occurs with trauma to the spinal cord, there is significant invasion of macrophages into the spinal cord after injury (73)
. These invading macrophages may release chemotactic factors (e.g., transforming growth factor-ß) that trigger the migration and subsequent proliferation of Schwann cells in the damaged CNS.
Schwann cell migration into the damaged cord may be further enhanced in models of intraspinal implantation of OECs. It is possible that besides factors released by invading macrophages, OECs release chemotactic factors that stimulate the migration of Schwann cells. For example, it is known that OECs synthesize and release a wide variety of growth factors (e.g., nerve growth factor, NGF; 74, 75). NGF has been shown to bind to p75NTR expressed on Schwann cells and promote chemotaxis in vitro (76)
.
Although migration into the damaged spinal cord from host peripheral nerves is a likely source of contaminating Schwann cells, an additional source of Schwann cells may be the original preparations of cells implanted into the spinal cord. One of the most common sources of OECs used in experiments providing the ultrastructural evidence suggestive of their myelinating potential is the olfactory nerve layer of the olfactory bulb of neonatal (e.g., refs 47
, 61
) and adult (e.g., refs 52
, 53
) rodents. An issue with this source of OECs is that nonolfactory axons (myelinated and unmyelinated) can be found in the immediately adjacent meninges (4)
. Most studies describe the great care with which the meninges are removed from the olfactory bulb before isolation of OECs. The possibility cannot be excluded that the Schwann cells associated with these nonolfactory axons persist to contaminate the cultures of OECs, especially given the highly convoluted nature of the pial surface of much of the olfactory bulb (4)
. An example of how this source of contaminating Schwann cells can contribute to the difficulty in interpreting the role of OECs in myelinating axons is the recent study by Kocsis and colleagues (49)
. Although the authors present evidence of green fluorescent protein-labeled cells associating with axons at the ultrastructural level, one must consider the possibility that these cells were contaminating Schwann cells from their original cultures isolated from the olfactory bulb of the transgenic green fluorescent protein-expressing rat. The authors concede the possibility that Schwann cells may be contaminating their cultures used for transplantation, but state that this "minor contamination could not account for the vast majority (>95%) of our cells displaying a p75/S100 phenotype." As Schwann cells and OECs both express these two phenotypic markers, no definitive conclusions should be drawn about the relative contribution of contaminating Schwann cells or their role in the myelination of axons in the CNS.
In addition to the olfactory bulb, several research groups have advocated the use of olfactory lamina propria as a potential source of OECs (77
78
79
80)
. From a clinical perspective, this region is a particularly enticing source of OECs because strips of lamina propria can be harvested from humans in a minimally invasive surgery under local anesthetic, negating the need for intracranial surgery or the use of allogeneic cadaveric sources to obtain OECs (81)
. The olfactory axons with their associated support cells course through the olfactory lamina propria, providing a rich source of OECs. Once again, however, Schwann cells are present throughout the olfactory lamina propria. The lamina propria is richly innervated by myelinated and unmyelinated sensory axons (and their associated Schwann cells) from the trigeminal nerve (see ref 1
; Fig. 2
). As seen surrounding the spinal cord, the medium-sized blood vessels present in the olfactory lamina propria are accompanied by sympathetic axons that are associated with nonmyelinating Schwann cells.
Although it is acknowledged that the olfactory lamina propria may contain Schwann cells that can potentially contaminate OEC cultures, their significance has been considered negligible or even outright dismissed (62
, 77)
. It has been suggested that in cultures of tissue from the lamina propria, Schwann cells may only be a minor contributor, since OECs "constitute the vast majority of glia in the olfactory mucosa where they ensheath several million" axons of olfactory receptor neurons (77)
. The implication of this statement is that since olfactory axons far outnumber sympathetic and sensory axons, then clearly OECs must far outnumber Schwann cells. This is not necessarily true, though, as the relationship between OECs and olfactory axons is dramatically different from the relationship between Schwann cells and sympathetic/sensory axons. Whereas a single OEC may associate with up to hundreds of olfactory axons, Schwann cells will associate only with a single myelinated axon or up to several unmyelinated axons. Therefore, the actual number of cells (not axons) may be much more nearly equivalent than originally suspected (Fig. 2)
.
To add further complication to this already contentious issue, OECs cultured from olfactory lamina propria have been considered a relatively pure population of cells on the basis of various phenotypic markers, including p75NTR, S100ß, O4, and GFAP (for review, see refs 82
, 83
). Unfortunately, since Schwann cells share these phenotypic markers, definitive conclusions cannot be drawn about the relative purity of these OEC cultures.
To address this, we recently undertook a proteomic strategy to identify unique phenotypic markers of OECs. Using 2-dimensional gel electrophoresis, we have provided preliminary evidence that OECs and Schwann cells indeed have distinct proteomic phenotypes (84)
. It remains to be seen, however, whether or not proteomic differences identified in vitro can be used to distinguish between implanted OECs and invading Schwann cells after spinal cord injury.
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONTHE MYELINATING CAPACITY OF...THE MYELINATING CAPACITY OF...SUMMARYSUMMARY AND CONCLUSIONSREFERENCES |
|---|
20
21
, 23)
and remyelination (23
, 24)
of damaged axons. Until now, it has been hypothesized that after isolation and purification, OECs that do not normally myelinate axons in vivo are induced by a yet unidentified signal to transform into a Schwann cell-like phenotype (Fig. 3)
. This premise has been extrapolated from in vitro findings, including the ability of embryonic OECs to myelinate axons in vitro (32
, 35
; cf. 22) and dramatic examples of OEC morphological plasticity (85)
. This myelinating phenotype may be inducible only in embryonic (23
, 24
, 42)
, and not adult (22
, 42)
, OECs under certain culture conditions. In vivo, however, we propose that it is much more likely that after implantation, OECs maintain their ability to ensheath other cellular elements but do not transform into Schwann cell-like cells. (Fig. 3)
. This alternative interpretation of earlier experiments should not detract from the many promising studies demonstrating a beneficial effect of OEC implantation on axon regeneration (24
, 29
, 52
53
54
, 79
, 80
, 86
87
88
89
90
91
92
93
94)
and remyelination (46
47
48
49
50
51
52
53
, 55
, 59
, 61)
. Nonetheless, the exact mechanisms underlying the beneficial effects after spinal cord injury and OEC implantation need to be reevaluated in the context of the unique collaboration between invading host Schwann cells and OECs.
Received for publication July 28, 2004. Accepted for publication December 22, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONTHE MYELINATING CAPACITY OF...THE MYELINATING CAPACITY OF...SUMMARYSUMMARY AND CONCLUSIONSREFERENCES |
|---|
