Richard P. Bunge

Expanding roles for the Schwann cell: ensheathment, myelination, trophism and regeneration

Schwann cells show remarkable versatility in undertaking a broad repertoire of functions. It is now clear that the well known functions of ensheathment and myelination are specifically regulated by contact with axons, that the Schwann cell is centrally involved in extracellular matrix production in the peripheral nerve trunk, and that the Schwann cell plays a critical role in promoting axonal regeneration in the peripheral nervous system. The Schwann cells ability to promote regenerative efforts in many central neurons has led to an increasing interest in using Schwann cell autografts for central nervous system repair.

Schwann cells genetically modified to secrete human BDNF promote enhanced axonal regrowth across transected adult rat spinal cord.

The infusion of BDNF and NT‐3 into Schwann cell (SC) grafts promotes regeneration of brainstem neurones into the grafts placed in adult rat spinal cord transected at T8 ( Xu et al. 1995b ). Here, we compared normal SCs with SCs genetically modified to secrete human BDNF, grafted as trails 5 mm long in the cord distal to a transection site and also deposited in the transection site, for their ability to stimulate supraspinal axonal regeneration beyond the injury. SCs were infected with the replication‐deficient retroviral vector pL(hBDNF)RNL encoding the human preproBDNF cDNA. The amounts of BDNF secreted (as detected by ELISA) were 23 and 5 ng/24 h per 106 cells for infected and normal SCs, respectively. Biological activity of the secreted BDNF was confirmed by retinal ganglion cell bioassay. The adult rat spinal cord was transected at T8. The use of Hoechst prelabelled SCs demonstrated that trails were maintained for a month. In controls, no SCs were grafted. One month after grafting, axons were present in SC trails. More 5‐HT‐positive and some DβH‐positive fibres were observed in the infected vs. normal SC trails. When Fast Blue was injected 5 mm below the transection site (at the end of the trail), as many as 135 retrogradely labelled neurones could be found in the brainstem, mostly in the reticular and raphe nuclei (normal SCs, up to 22, mostly in vestibular nuclei). Numerous neurones were labelled in the ventral hypothalamus (normal SCs, 0). Also, following Fast Blue injection, a mean of 138 labelled cells was present in dorsal root ganglia (normal SCs, 46) and spinal cord (39 vs. 32) rostral to the transection. No labelled spinal neurones rostral to the transection were seen when SCs were not transplanted. Thus, the transplantation of SCs secreting increased amounts of BDNF improved the regenerative response across a transection site in the thoracic cord. Moreover, the enhanced regeneration observed with infected SCs may be specific as the largest response was from neurones known to express trkB.

The Ability of Human Schwann Cell Grafts to Promote Regeneration in the Transected Nude Rat Spinal Cord

Advances in the purification and expansion of Schwann cells (SCs) from adult human peripheral nerve, together with biomaterials development, have made the construction of unique grafts with defined properties possible. We have utilized PAN/PVC guidance channels to form solid human SC grafts which can be transplanted either with or without the channel. We studied the ability of grafts placed with and without channels to support regeneration and to influence functional recovery; characteristics of the graft and host/graft interface were also compared. The T9-T10 spinal cord of nude rats was resected and a graft was placed across the gap; methylprednisolone was delivered acutely to decrease secondary injury. Channels minimized the immigration of connective tissue into grafts but contributed to some necrotic tissue loss, especially in the distal spinal cord. Grafts without channels contained more myelinated axons (x = 2129 +/- 785) vs (x = 1442 +/- 514) and were larger in cross-sectional area ( x = 1.53 +/- 0.24 mm2) vs (x = 0.95 +/- 0.86 mm2). The interfaces formed between the host spinal cord and the grafts placed without channels were highly interdigitated and resembled CNS-PNS transition zones; chondroitin sulfate proteoglycans was deposited there. Whereas several neuronal populations including propriospinal, sensory, motoneuronal, and brainstem neurons regenerated into human SC grafts, only propriospinal and sensory neurons were observed to reenter the host spinal cord. Using combinations of anterograde and retrograde tracers, we observed regeneration of propriospinal neurons up to 2.6 mm beyond grafts. We estimate that 1% of the fibers that enter grafts reenter the host spinal cord by 45 days after grafting. Following retrograde tracing from the distal spinal cord, more labeled neurons were unexpectedly found in the region of the dextran amine anterograde tracer injection site where a marked inflammatory reaction had occurred. Animals with bridging grafts obtained modestly higher scores during open field [(x = 8.2 +/- 0.35) vs (x = 6.8 +/- 0.42), P = 0.02] and inclined plane testing (x = 38.6 +/- 0. 542) vs (x = 36.3 +/- 0.53), P = 0.006] than animals with similar grafts in distally capped channels. In summary, this study showed that in the nude rat given methylprednisolone in combination with human SC grafts, some regenerative growth occurred beyond the graft and a modest improvement in function was observed.

Improved method for harvesting human Schwann cells from mature peripheral nerve and expansion in vitro.

The use of cellular prostheses containing large populations of Schwann cells (SC) has been proposed as a future therapeutic approach in the repair of neural tissue. We have sought to define an efficient protocol for the harvest and expansion of human SC from mature human peripheral nerve. We evaluated SC proliferation occurring within fresh explants and studied the relationship between certain parameters (cell yield, purity, and rate of SC proliferation) and the conditions of maintenance of nerve explants prior to dissociation. In addition, we studied SC proliferation after dissociation in a variety of conditions. We observed that SC within explants divide at a low rate during the first 3 weeks following explantation; this proliferation falls to near zero during the fourth week. The cell yield, SC purity, and proliferation rate following dissociation were all increased when nerve explants were exposed to heregulin/forskolin for 2 weeks prior to dissociation. Electron microscopic analysis showed that heregulin/forskolin exerted trophic effects on SC within explants. Following dissociation, SC growth in heregulin/forskolin‐containing medium was more rapid on laminin or collagen than on poly‐L‐lysine. These results provide new insights into human SC biology and suggest several procedural improvements for harvesting and expanding these cells. The new method we describe shortens our previous procedure by 4–6 weeks and provides a 30–50‐fold increase in the number of SC obtained relative to the earlier procedure.

Acute traumatic central cord syndrome: MRI-pathological correlations

SummaryThe acute traumatic central cord syndrome (ATCCS) is commonly stated to result from an injury which affects primarily the center of the spinal cord and is frequently hemorrhagic. To test the validity of this widely disseminated hypothesis, the magnetic resonance images [MRI] of 11 consecutive cases of ATCCS caused by closed injury to the spine were analyzed and correlated with the gross pathological and histological features of 3 cervical spinal cords obtained at post mortem from patients with ATCCS, including 2 of patients studied by MRI. The MRI studies were performed acutely (18 h to 2 days after injury) in 7 patients and subacutely (3–10 days after injury) in 4. Ten of the 11 patients had pre-existing spondylosis and/or canal stenosis. The 11th suffered a cervical fracture. All patients exhibited hyperintense signal within the parenchyma of the cervical spinal cord on gradient echo MRI. None showed MRI features characteristic of hemorrhage on T1-weighted spin echo or T2*-weighed gradient echo studies. Gross and histological examination of the necropsy specimens showed no evidence of blood or blood products within the cord parenchyma: the primary finding was diffuse disruption of axons, especially within the lateral columns of the cervical cord in the region occupied by the corticospinal tracts. The central gray matter was intact. In patients with ATCCS, the predominant loss of motor function in thedistal muscles of the upper limbs may reflect the importance of the corticospinal tract for hand and finger function in the primate. In this study, the MRI and pathological observations indicate that ATCCS is predominantly a white matter injury and that intramedullary hemorrhage is not a necessary feature of the syndrome; indeed, it is probably an uncommon event in ATCCS. We suggest that the most common mechanism of injury in ATCCS may be direct compression of the cervical spinal cord by buckling of the ligamenta flava into an already narrowed cervical spinal canal; this would explain the predominance of axonal injury in the white matter of the lateral columns.

Influence of IN-1 antibody and acidic FGF-fibrin glue on the response of injured corticospinal tract axons to human Schwann cell grafts

Two strategies have been shown by others to improve CST regeneration following thoracic spinal cord injury: 1) the administration of a monoclonal antibody, IN‐1, raised against a myelin‐associated, neurite growth inhibitory protein, and 2) the delivery of acidic fibroblast growth factor (aFGF) in fibrin glue in association with peripheral nerve grafts. Because autologous transplantation of human Schwann cells (SCs) is a potential strategy for CNS repair, we evaluated the ability of these two molecular agents to induce CST regeneration into human SC grafts placed to span a midthoracic spinal cord transection in the adult nude rat, a xenograft tolerant strain. IN‐1 or control (HRP) antibodies were delivered to the injury/graft region by encapsulated hybridoma cells (“IN‐1 ravioli”) or daily infusion of hybridoma culture supernatant; aFGF‐fibrin glue was placed in the same region in other animals. Anterograde tracing from the motor cortex using the dextran amine tracers, Fluororuby (FR) and biotinylated dextran amine (BDA), was performed. Thirty‐five days after grafting, the CST response was evaluated qualitatively by looking for regenerated CST fibers in or beyond grafts and quantitatively by constructing camera lucida composites to determine the sprouting index (SI), the position of the maximum termination density (MTD) rostral to the GFAP‐defined host/graft interface, and the longitudinal spread (LS) of bulbous end terminals. The latter two measures provided information about axonal die‐back. In control animals (graft only), the CST did not enter the SC graft and underwent axonal die‐back [SI = 1.4 ± 0.1, MTD = 2.0 ± 0.2, LS = 1.3 ± 0.3, (n = 3)]. Results of IN‐1 delivery from ravioli did not differ from controls, but injections of IN‐1‐containing supernatant resulted in a significant degree of sprouting but did not prevent axonal die‐back [SI = 1.9 ± 0.1, MTD = 1.5 ± 0.2, LS = 1.1 ± 0.1, (n = 7)] and traced fibers did not enter grafts. Acidic FGF dramatically reduced axonal die‐back and caused sprouting [SI = 2.0 ± 0.1 (n = 5), MTD = 0.5 ± 0.04 (n = 6), LS = 0.4 ± 0.1 (n = 6)]. Some traced fibers entered SC grafts and in 2/6 cases entered the distal interface. We conclude that 1) human SC grafts alone do not support the regeneration of injured CST fibers and do not prevent die‐back, 2) grafts plus IN‐1 antibody‐containing supernatant support some sprouting but die‐back continues, and 3) grafts plus aFGF‐fibrin glue support regeneration of some fibers into the grafts and reduce die‐back. J. Neurosci. Res. 50:888–905, 1997. © 1997 Wiley‐Liss, Inc.

Schwann cells infected with a recombinant retrovirus expressing myelin-associated glycoprotein antisense RNA do not form myelin

To elucidate the role of myelin-associated glycoprotein (MAG) in the axon-Schwann cell interaction leading to myelination, neonatal rodent Schwann cells were infected in vitro with a recombinant retrovirus expressing MAG antisense RNA or MAG sense RNA. Stably infected Schwann cells and uninfected cells were then cocultured with purified sensory neurons under conditions permitting extensive myelination in vitro. A proportion of the Schwann cells infected with the MAG antisense virus did not myelinate axons and expressed lower levels of MAG than control myelinating Schwann cells, as measured by immunofluorescence. Electron microscopy revealed that the affected cells failed to segregate large axons and initiate a myelin spiral despite having formed a basal lamina, which normally triggers Schwann cell differentiation. Cells infected with the MAG sense virus formed normal compact myelin. These observations strongly suggest that MAG is the critical Schwann cell component induced by neuronal interaction that initiates peripheral myelination.

Studies of Myelin Formation after Transplantation of Human Schwann Cells into the Severe Combined Immunodeficient Mouse

We have previously demonstrated (J. Neurosci., 14: 1309-1319) that Schwann cells (SCs) isolated from adult human peripheral nerve in tissue culture and then transplanted into an immune-deficient rat can enhance axonal regeneration and myelinate regenerating peripheral axons. We have now (a) compared the capacity of both primary and expanded populations of cultured human SCs to form myelin around regenerating mouse axons when transplanted into a gap within the sciatic nerve of severe combined immunodeficiency (scid) mice and (b) also compared the myelinating capability of these cultured SCs to their counterparts in the native human peripheral nerve xenograft. Schwann cells were isolated from adult human peripheral nerve. Semipermeable guidance channels were filled with a 30% Matrigel solution mixed with either primary human SCs or human SCs expanded with mitogens both at a density of 120 million cells/ml. These channels or a human peripheral nerve xenograft were implanted within a 5-mm gap in the transected sciatic nerve of the scid mice and analyzed after a period of 6 weeks. The presence of human myelin segments was confirmed in both the guidance channels containing human SCs and the xenografts by immunostaining with a monoclonal antibody (592) which specifically recognizes a prominent myelin component, P0, in the human but not in the mouse. Within both the guidance channels and the xenografts there was an invasion of the transplant by host SCs which went on to form myelin around regenerating mouse axons. In this report, we also demonstrate that human SCs that have been expanded in culture with mitogens are capable of forming myelin after transplantation in this experimental paradigm.

Clinical syndromes associated with disproportionate weakness of the upper versus the lower extremities after cervical spinal cord injury

Patients with cervical spinal cord injuries who present with weakness or paralysis of the hands and arms with relative preservation of lower extremity strengths are often categorized as having two clinical syndromes, cruciate paralysis and acute central cervical spinal cord injury. The explanation for the pathophysiological findings of the dissociated strength in the upper versus the lower extremities has relied on the assumption that there is a localized injury within a somatotopically organized corticospinal tract. This article summarizes the evidence that there is no somatotopic organization within the corticospinal tract in the medulla or cervical spinal cord in primates. An alternative hypothesis for these two syndromes is presented and is based on evidence that has demonstrated that the corticospinal tract in primates is critical for hand function but not for locomotion. Other prevailing theories are reviewed. Thus, we propose that a syndrome consisting of relatively greater hand and arm weakness compared with leg weakness can occur after an injury to the corticospinal tracts in the medulla or the cervical cord. The proposed mechanism, based on the function of the corticospinal tract, unifies a spectrum of injuries of the lower medulla and cervical spinal cord, which produce similar clinical syndromes (cruciate paralysis and acute central cervical spinal cord injury).

Studies of the Initiation of Myelination by Schwann Cells

The rapid morphologic changes in Schwann cells and in their relationships to axons during the transition from the premyelinating to the myelinating state have been known for more than 15 years. The sorting of axons by dividing Schwann cells, the establishment of a 1:1 relationship between a postmitotic Schwann cell, and the onset of myelin sheath formation have all been described in detail. However, the chain of molecular events and mechanisms by which these morphologic changes are regulated has not been elucidated. In this chapter we have reviewed results that strongly suggest that the adhesion molecule L1 is one of the important determinants that mediate the elongation of the Schwann cell along the axon, and the extension of Schwann processes to engulf axons. Thus, L1 functions to promote the spreading of the Schwann cell process over the surface of the axon. L1 does not appear to be exclusively involved in the adhesion of Schwann cells to axons, in the activation of Schwann cell proliferation by axons, or in the induction of synthesis of extracellular matrix proteins. The results from the anti-L1 blocking experiments further provided clues for an understanding of how the expression of GalC and MAG, which are both likely to be involved in the initiation of myelination, are regulated. These results imply that the overall regulation of expression of these early myelin components could require controls other than a single signaling mechanism derived from contact with axons. We propose that the deposition of basal lamina or one of its components could also be involved. Finally, the results from anti-GalC-blocking experiments indicated that GalC is involved in the mechanism of early growth of the myelin spiral.