Neurite-promoting factors and extracellular matrix components accumulating in vivo within nerve regeneration chambers

Nerve regeneration chamber: evaluation of exogenous agents applied by multiple injections

In this laboratory, a silicone chamber model for peripheral nerve regeneration in adult rats has been developed and used to define basic principles of the regenerative events, such as the sequential stages being followed during 'spontaneous' regeneration in vivo and the role of neuronotrophic- and neurite-promoting factors as well as extracellular matrix molecules. Each of the defined stages seems amenable to experimental modulation. Previous attempts to enhance regeneration included increasing the volume of the nerve chambers along with the modification of fibrin matrix formation by prefilling with saline (PBS) or matrix precursors. We present here the results of a series of experiments on the effects of exogenous biochemical agents applied by multiple injections into these in vivo chambers. Out of a variety of agents screened, a mixture of laminin (L), testosterone (T), ganglioside GM 1 (G), and catalase (C) was shown to advance substantially the progress of regeneration in 16 day chambers, as compared to PBS-prefilled and PBS-injected controls. LTGC-treatment at day 0, 6, and 10 postimplantation caused an increasingly frequent occurrence of cellular elements in cross-sections obtained from the middle (S5) of the chambers (i.e. 5 mm from the proximal stump), which was 2-fold for vessels, 3-fold for Schwann cells, and 10-fold for axons. When only sections containing axons 3 mm from the proximal stump (S3) were compared in experimental and control groups, computerized area measurements also revealed an average 2-fold difference for the cross-sectional size of the whole regenerate, the endoneurium and the space occupied by blood vessels.

Response of sensory neurites and growth cones to patterned substrata of laminin and fibronectin in vitro

During neurite elongation in the developing peripheral nervous system, the distribution of laminin and fibronectin may provide preferred substrates for neurite elongation. In this study, the response of sensory neurites and growth cones to patterns of laminin or fibronectin applied to a background substrate of Type IV collagen was studied to determine any possible substrate preference. Neurites exhibited elongation restricted to a laminin pattern, but not a fibronectin pattern, indicating that sensory neurites prefer to elongate on laminin compared to Type IV collagen. When polylysine is included in the background substrate, neurite preference for laminin is decreased. Laminin also enhances neurite elongation and defasciculation and stabilizes growth cone protrusions. These results suggest an adhesive as well as a cytoskeletal involvement in the response to laminin, but direct adhesion estimates indicate that laminin decreases overall adhesion, arguing against an adhesive involvement. Regardless of the mechanism involved, the observed neurite preference for laminin is consistent with the hypothesis that spatial and temporal laminin distributions provide preferred pathways for peripheral neurite elongation.

Human amnion membrane serves as a substratum for growing axons in vitro and in vivo

The epithelial cell layer of human amnion membrane can be removed while the basement membrane and stromal surfaces remain morphologically intact. Such a preparation has been used as a substratum for the in vitro culture of dissociated neurons. Embryonic motor neurons from chick ciliary ganglion attached to both surfaces but grew extensive neurites only on the basement membrane. On cross sections of rolled amnion membranes, regenerating axons of cultured neurons were guided along pathways of basement membrane that were immunoreactive with an antibody to laminin. In addition, when rolled amnion membranes were implanted into a lesion cavity between the rat septum and hippocampus, cholinergic neurons extended axons through the longitudinally oriented implant into the hippocampus. Thus, this amnion preparation can serve as a bridge to promote axonal regeneration in vivo in damaged adult brain.

Distribution of laminin and fibronectin along peripheral trigeminal axon pathways in the developing chick

The trigeminal region of the chick was studied with indirect immunofluorescence in order to determine whether extracellular matrix components might be distributed in such a way as to guide trigeminal axons to their peripheral targets in the mandibular arch. Tissue sections from stages 13-15 and 21/22 were immunolabeled indirectly with affinity-purified antibodies raised against fibronectin and laminin, two extracellular matrix glycoproteins that support axon growth in vitro. Fibronectin was distributed ubiquitously throughout the head mesenchyme prior to and during initial axon growth from the brainstem (stages 13-15). Shortly after trigeminal axons reached their target tissues (stage 21/22), fibronectin immunolabeling was distributed throughout the head mesenchyme, but was present only at low levels in the trigeminal ganglion and motor nerve. Laminin immunolabeling was distributed in the lateral head mesenchyme at stage 13 as small specks and patches. At stage 14, when the motor axons first exit from the brainstem, short, linear arrays of laminin immunostaining were present from the basement membrane of the neural tube to the core of the mandibular arch, and many were parallel to the direction of axon growth. By stage 21/22 the trigeminal ganglion and motor root showed intense antilaminin immunofluorescence as did the central core of the mandibular arch. These studies suggest that the distribution of fibronectin within the head mesenchyme cannot give directional information to the growing trigeminal axons because of its homogeneous distribution. However, the initial distribution of laminin during the earliest stages of axon outgrowth may provide an extracellular matrix pathway that permits trigeminal axons to reach their targets.

Permeable tubes increase the length of the gap that regenerating axons can span

After placement of stumps of transected rat sciatic nerve in an impermeable tube, the maximum gap the axons can span is 10 mm. The present study shows that the regenerating axons cross much longer gaps if the tube is made permeable. This improvement does not require another nerve as a transplant nor the preplacement of extracellular materials in the tube. Possible mechanisms for this improvement are discussed.

Evoked muscle action potentials from regenerated rat tibial and peroneal nerves: synthetic versus autologous interfascicular grafts

To compare peripheral nerve regeneration across bridging synthetic tubular implants with ordinary autologous transplantation, we recorded evoked muscle action potentials 3 months after transection, following direct stimulation of rat tibial and peroneal nerves. Significantly, with autologous transplantation we were able to evoke compound muscle action potentials in all but one case (this having a 15-mm gap). Pooling the groups together, compound action potentials could be recorded in 15 of 45 cases that had regenerated through 13- to 16-mm gaps. Potentials in the synthetic implant-treated group tended to show more temporal "scattering" in the late phases of the EMG response. No differences were found in compound muscle action potential amplitudes and duration, or terminal motor conduction velocity, between both experimental groups, between different artificial tubuli, between different gap lengths, or between the peroneal and tibial nerves. The overall values were remarkably lower than in nontransected controls, about 25% of the compound muscle action potential normal amplitude voltage and 60% of the terminal motor conduction velocity.

Immunoelectron microscopic localization of neural cell adhesion molecules (L1, N-CAM, and MAG) and their shared carbohydrate epitope and myelin basic protein in developing sciatic nerve

The cellular and subcellular localization of the neural cell adhesion molecules L1, N-CAM, and myelin-associated glycoprotein (MAG), their shared carbohydrate epitope L2/HNK-1, and the myelin basic protein (MBP) were studied by pre- and post-embedding immunoelectron microscopic labeling procedures in developing mouse sciatic nerve. L1 and N-CAM showed a similar staining pattern. Both were localized on small, non-myelinated, fasciculating axons and axons ensheathed by non-myelinating Schwann cells. Schwann cells were also positive for L1 and N-CAM in their non-myelinating state and at the onset of myelination, when the Schwann cell processes had turned approximately 1.5 loops. Thereafter, neither axon nor Schwann cell could be detected to express the L1 antigen, whereas N-CAM was found in the periaxonal area and, more weakly, in compact myelin of myelinated fibers. Compact myelin, Schmidt-Lanterman incisures, paranodal loops, and finger-like processes of Schwann cells at nodes of Ranvier were L1-negative. At the nodes of Ranvier, the axolemma was also always L1- and N-CAM-negative. The L2/HNK-1 carbohydrate epitope coincided in its cellular and subcellular localization most closely to that observed for L1. MAG appeared on Schwann cells at the time L1 expression ceased. MAG was then expressed at sites of axon-myelinating Schwann cell apposition and non-compacted loops of developing myelin. When compaction of myelin occurred, MAG remained present only at the axon-Schwann cell interface; Schmidt-Lanterman incisures, inner and outer mesaxons, and paranodal loops, but not at finger-like processes of Schwann cells at nodes of Ranvier or compacted myelin. All three adhesion molecules and the L2/HNK-1 epitope could be detected in a non-uniform staining pattern in basement membrane of Schwann cells and collagen fibrils of the endoneurium. MBP was detectable in compacted myelin, but not in Schmidt-Lanterman incisures, inner and outer mesaxon, paranodal loops, and finger-like processes at nodes of Ranvier, nor in the periaxonal regions of myelinated fibers, thus showing a complementary distribution to MAG. These studies show that axon-Schwann cell interactions are characterized by the sequential appearance of cell adhesion molecules and MBP apparently coordinated in time and space. From this sequence it may be deduced that L1 and N-CAM are involved in fasciculation, initial axon-Schwann cell interaction, and onset of myelination, with MAG to follow and MBP to appear only in compacted myelin. In contrast to L1, N-CAM may be further involved in the maintenance of compact myelin and axon-myelin apposition of larger diameter axons.

Neuronal growth factors produced by adult peripheral nerve after injury

Dorsal root ganglion neurons from embryonic rats, co-cultured with endoneurial explants from transected, adult rat sciatic nerve, extended neurites in the absence of exogenous nerve growth factor (NGF). The effect was seen with endoneurial explants from normal adult sciatic nerves or from nerves which had been permanently transected up to 51 days prior to explantation. The rate of outgrowth decreased at 5 and 7 days and reached a minimum at 14 days after transection. A second phase of increased neurite-promoting activity appeared in 28-, 35-, 41- and 51-day posttransection tissue. The early phase, but not the late phase, was partially inhibited by antisera to NGF.

The effect of inhibiting Schwann cell mitosis on the re-innervation of acellular autografts in the peripheral nervous system of the mouse

Reactive gliosis in the zone immediately proximal to transection of the sciatic nerve has been inhibited by intraneural injection of mitomycin C, an anti-mitotic agent known to arrest Schwann cell division after transection, crush or demyelination. Mitomycin C-pretreated proximal stumps were subsequently sutured to cellular or acellular autografts (0.5 cm long) and neurite growth into and within the grafts was examined during a 5-week post-operative period. Neurites grew into cellular autografts and became associated with the resident population of Schwann cells within the grafts, to the extent that remyelination was well established in the majority of Schwann cell basal lamina tubes by week 5 post-suture. In marked contrast, very few neurites grew into acellular grafts during this time, and where axons and Schwann cells were seen they tended to be grouped in 'minifascicles'. The results suggest that neurite outgrowth from proximal stumps is dependent upon active Schwann cell participation.

Induction of a neurite-promoting factor in rat brain following injury or deafferentation

Ablation of the entorhinal/occipital cortex in young adult rats caused a several-fold increase in the neurite-promoting activity in extracts of the tissue surrounding the wound and in areas that had been deafferented by the lesion. The time course of induction closely paralleled reactive axon sprouting in the deafferented hippocampus, with maximal levels of neurite-promoting activity reached between 9 and 15 days post-lesion. Aged animals, in which reactive sprouting is deficient, showed no increase in activity by 12 days after deafferentation of the hippocampus. The neurite-promoting activity of brain extracts was non-diffusible, heat-labile, and sensitive to proteolysis. All of the activity bound to diethylaminoethyl (cellulose) and was eluted at 200 mM NaCl. The apparent molecular weight (by gel filtration) of the activity in extracts of uninjured brain was 9-17 kilodaltons, whereas the extracts of injured brain also had peaks or shoulders at 30, 70 and greater than or equal to 200 kilodaltons. These data suggest that the brain neurite-promoting activity resides in one or more proteins. Both the injury-induced and basal activities were different from laminin, nerve growth factor, and polyornithine-bindable neurite-promoting factors. The injury-induced activity was sensitive to repeated freezing and thawing, but this inactivation was reversed by thiol reagents such as glutathione, thioglycerol, and mercaptoethanol. We report a neurite-promoting factor that is induced following brain injury or denervation, and may also be important for reactive axon sprouting after brain injury. The induction of this factor is abnormal in aged animals, as is the reactive sprouting response. The properties of the injury-induced activity distinguish it from the basal activity (found in uninjured brain) and from other characterized neurite-promoting factors.

Regeneration in cellular and acellular autografts in the peripheral nervous system

The regeneration that occurs in cellular autografts of sciatic nerve has been compared with that seen in acellular models prepared either by cycles of alternating freezing and thawing, or by detergent-extraction. The responses to either fresh or pre-degenerate grafts (cellular and acellular) have been examined electron microscopically. It was found that whereas neurites grew into a fresh autograft and rapidly re-established functional relationships with vital Schwann cells lying in bands of Büngner within the graft, penetration of acellular grafts was less efficient. Many basal lamina tubes in the acellular grafts remained either empty or filled with debris-laden macrophages for the first 2 weeks after suture, although subsequently reinnervation did occur. The roles of Schwann cells, macrophages and basal laminae during reinnervation are discussed.

Enhanced graft survival in the hippocampus following selective denervation

The trophic effects of denervation on the survival of fetal cholinergic neuronal cell suspensions grafted to the hippocampal formation of the rat were assessed in the present study. Young adult female rats were injected with cell suspensions of neurons obtained from the fetal basal forebrain region into the hippocampal formation simultaneously with (or without) a fimbria-fornix transection, which removes the hippocampal cholinergic afferents. Four to six months later, one group of grafted animals was evaluated histochemically for: transplant volume; number of acetylcholinesterase-positive cells, and size of acetylcholinesterase-positive cells in the graft. A parallel study was conducted to determine the total number and size of the acetylcholinesterase-positive cells in the septal-diagonal band-substantia innominata complex of the adult rat, to match with the cell survival and growth in the grafts. A second group of grafted rats was taken in parallel for biochemical analysis of choline acetyltransferase activity in the grafted hippocampus. The transplant volume in the rats with fimbria-fornix transection was greater than twice the volume seen in animals without fimbria-fornix lesion. In addition, the number of acetylcholinesterase-positive cells in the transplant was twice as great in the denervated animals as in the non-denervated ones. However, the number of acetylcholinesterase-positive cells per mm3 of graft volume did not differ between the two groups, suggesting that the trophic effect of the denervation was not specific for the cholinergic neurons, but affected the entire grafted tissue. The hippocampal choline acetyltransferase activity of the animals that received the fimbria-fornix lesion simultaneously with transplantation was about three times higher than that of the rats that received grafts but no simultaneous fimbria-fornix transection. A control experiment with animals that received an aspirative lesion of the retrosplenial cortex, transecting the perforant path input, revealed no enhancing effect of hippocampal choline acetyltransferase activity over non-lesioned grafted animals. Thus, the denervation-enhancing effects of the fimbria-fornix lesion appear to be selective and not the result of a general wound-induced mechanism. These results strongly support the contention that neurotrophic factors are released as a result of denervation in the adult hippocampal formation, and that these neurotrophic factors can support survival and growth of central cholinergic neurons. However, the factors involved do not appear to be specific for the cholinergic neurons, but rather have their trophic effects on many types of cells.

Nerve regeneration through biodegradable polyester tubes

One approach to repair of transected nerves is to attempt extrinsic guidance of axons across the gaps. We inserted the proximal and distal stumps of severed mouse sciatic nerves into opposite ends of biodegradable polyester tubes. The nerves and ensheathing tubes were examined after postoperative survival times of as long as 2 years. Myelinated fiber number in each successfully regenerated nerve was measured and correlated with the tube's residual lumen size. In selected regenerated nerves axonal sizes and myelin sheath widths were sampled and compared with control values. Swelling and deformation of tube walls occurred in nearly all tubes. Successful regeneration was obtained through more than half of the implants, and was more probable in tubes with larger initial lumens. Myelinated fiber number in regenerated nerves ranged from 231 to 3561 (normally 3900 to 4200); larger values again were found in tubes with larger initial lumens. Mean axonal areas in regenerated nerves were roughly half of normal, though myelin sheaths became appropriately thick. We concluded that the more biodegradable a tube, the more likely it was to incur distortion and luminal narrowing. Tube composition per se seemed of importance mainly as it related to maintenance of adequate luminal size over the length of the degrading tubes; luminal adequacy, not tube composition, seemed paramount in determining the extent of nerve regeneration.

Parameters of neuritic growth from ciliary ganglion neurons in vitro: influence of laminin, schwannoma polyornithine-binding neurite promoting factor and ciliary neuronotrophic factor

Ciliary ganglion neurons extend neuritic processes when cultured for 24 h in medium containing ciliary neuronotrophic factor (CNTF) and on a polyornithine substratum precoated with either laminin or a Schwannoma-derived neurite promoting factor (PNPF). We have examined the roles of laminin, PNPF and CNTF for each of four parameters of neuritic growth, including: initiation time, neuronal polarity, neuritic branching and average neurite output (lengths) with time. Increasing laminin and PNPF levels were found to advance the time of neurite initiation as well as shift the majority (70-80%) of the neurons from a unipolar to multipolar neuritic morphology. The polarity imposed by any given concentration of either neurite promoting factor remained constant over the 24 h culture period examined. The average lengths from the longest neurites per neuron over a 10-28 h culture interval were not affected by increasing levels of laminin or PNPF, but total neuritic output per neuron was increased. This increased total neuritic output could be attributed to a combination of earlier neuritic initiation time and an increased neuronal polarity at high laminin or PNPF levels. CNTF at threshold survival levels did not promote initiation time, neuronal polarity or total neuritic output. However, cultures receiving less CNTF than that required for maximal neuronal survival displayed an increased neuronal polarity and a reduced neuritic output before any apparent loss of neurons. Neuritic branching was not affected by either the neurite promoting or trophic factors after 24 h of culture. Laminin and PNPF were found to be indistinguishable in their effects on the ciliary ganglion neurons in each of the four parameters studied.

Comparison of the effects of laminin and the polyornithine-binding neurite promoting factor from RN22 Schwannoma cells on neurite regeneration from cultured newborn and adult rat dorsal root ganglion neurons

We have investigated the effects of two neurite promoting factors (NPFs)--laminin and the semipurified polyornithine-binding neurite promoting factor (PNPF-1) from RN 22 Schwannoma cells--on neurite regeneration from dissociated newborn and adult rat dorsal root ganglion (DRG) neurons during 24 and 48 h culture periods in the absence of exogenous neuronotrophic factors. Both laminin and PNPF, when used to pretreat the polyornithine substratum, significantly enhanced neurite recruitment from surviving newborn and adult DRG neurons as compared to an untreated polyornithine substratum. However, the responses of newborn neurons at saturating concentrations of laminin and PNPF were consistently greater (46% neurite-bearing cells at 24 h, 81% at 48 h) than those of adult neurons (14 and 45%, respectively). The responsive neurons of both newborn and adult DRG displayed extensive neuritic networks at 48 h. The ED50 of laminin, or PNPF was 0.15-0.2 micrograms/ml for both newborn and adult neurons. The similarities in the responses of newborn and adult DRG neurons to NPFs validate the use of neurons from embryonic and newborn animals for the in vitro assays of NPFs that can be collected from injured and regenerating adult peripheral nervous tissues.