Growth-associated protein GAP-43 and L1 act synergistically to promote regenerative growth of Purkinje cell axons in vivo

  1. Yi Zhang*,†,‡,
  2. Xuenong Bo*,†,§,
  3. Ralf Schoepfer§,
  4. Anthony J. D. G. Holtmaat,∥,
  5. Joost Verhaagen,
  6. Piers C. Emson**,
  7. A. Robert Lieberman*, and
  8. Patrick N. Anderson*
  1. *Department of Anatomy and Developmental Biology and §Wellcome Laboratory for Molecular Pharmacology, University College London, Gower Street, London WC1E 6BT, United Kingdom; †Neuroscience Centre, Queen Mary, University of London, 4 Newark Street, London E1 2AT, United Kingdom; Netherlands Institute for Brain Research, Meibergdreef 33, 1105 AZ, Amsterdam, The Netherlands; and **The Babraham Institute, Cambridge CB2 4AT, United Kingdom

  1. Edited by Fred H. Gage, The Salk Institute for Biological Studies, San Diego, CA, and approved August 25, 2005 (received for review June 20, 2005)

 
Next Section

Abstract

Neuronal expression of growth-associated protein 43 (GAP-43) and the cell adhesion molecule L1 has been correlated with CNS axonal growth and regeneration, but it is not known whether expression of these molecules is necessary for axonal regeneration to occur. We have taken advantage of the fact that Purkinje cells do not express GAP-43 or L1 in adult mammals or regenerate axons into peripheral nerve grafts to test the importance of these molecules for axonal regeneration in vivo. Transgenic mice were generated in which Purkinje cells constitutively express L1 or both L1 and GAP-43 under the Purkinje cell-specific L7 promoter, and regeneration of Purkinje cell axons into peripheral nerve grafts implanted into the cerebellum was examined. Purkinje cells expressing GAP-43 or L1 showed minor enhancement of axonal sprouting. Purkinje cells expressing both GAP-43 and L1 showed more extensive axonal sprouting and axonal growth into the proximal portion of the graft. When a predegenerated nerve graft was implanted into double-transgenic mice, penetration of the graft by Purkinje cell axonal sprouts was strongly enhanced, and some axons grew along the entire intracerebral length of the graft (2.5-3.0 mm) and persisted for several months. The results demonstrate that GAP-43 and L1 coexpressed in Purkinje cells can act synergistically to switch these regeneration-incompetent CNS neurons into a regeneration-competent phenotype and show that coexpression of these molecules is a key regulator of the regenerative ability of intrinsic CNS neurons in vivo.

Although intrinsic CNS neurons of adult mammals are normally unable to regenerate their axons after injury (1), some can be induced to regrow axons into and along segments of peripheral nerve grafted into the brain or spinal cord (2). Retinal ganglion cells and thalamic reticular nucleus neurons regenerate well into grafts (3, 4), but neocortical and neostriatal projection neurons do so very poorly, and cerebellar Purkinje cells do not do so at all (5). Successful regeneration of axons along nerve grafts occurs only from neurons that show prolonged up-regulation of growth-associated molecules, including growth-associated protein 43 (GAP-43), cell adhesion molecule L1, and close homologue of L1 CHL1 (6).

GAP-43 was the first neuronal growth-associated protein to be identified (7). It is concentrated in the cortical cytoskeleton of axonal growth cones, interacts with the cell membrane and actin filaments, and overexpression results in enhanced neurite outgrowth in vitro and enhanced axonal sprouting in vivo (8, 9). Although the nervous system of GAP-43 knockout mice is grossly normal, there are pathway and circuit abnormalities in such mice (10-12); the current consensus view is that GAP-43 is involved in axonal sprouting and in the response of growth cones to guidance cues.

L1 is the prototype of a widely distributed family of cell adhesion molecules (CAMs). It is present at the surface of growing axons during development (13, 14), and there is good evidence that it is important in axon growth and guidance (14). Its effects are mediated by homophilic binding to L1 or heterophilic binding to ligands, including integrins, on other cells, or by binding to extracellular matrix-associated molecules, including chondroitin sulfate proteoglycans, neurocan, and phosphacan (15). The intracellular domain of L1 binds to ezrin-radixin-moesin membrane-cytoskeleton linker proteins that are involved in growth cone morphology and motility and second messenger pathways (16, 17). L1 also interacts with FGF receptors, and L1-induced neurite outgrowth requires the tyrosine kinase activity of FGF receptors to activate intracellular signaling cascades, including Ras/mitogen-activated protein kinase and phospholipase-Cγ/diacyl-glycerol/Ca2+/PKC pathways (15, 18, 19). There is evidence that the growth-promoting effects of L1 depend on GAP-43 (20).

Although intrinsic CNS neurons that do not regenerate axons into nerve grafts do not up-regulate GAP-43 and L1 expression, it is not known whether the expression of these molecules would be sufficient to allow them to do so. Our aim was to directly assess the importance of neuronal GAP-43 and L1 for axonal regeneration in vivo. Purkinje cells do not regenerate axons into peripheral nerve grafts and do not express GAP-43 or L1 either normally or after injury and peripheral nerve grafting (21). Transgenic mice in which Purkinje cells express GAP-43 alone display enhanced sprouting of axotomized Purkinje cell axons, but the sprouts do not grow into Schwann cell or peripheral nerve grafts implanted at the lesion site (5, 9). We have therefore generated mice in which Purkinje cells express either L1 or both L1 and GAP-43 and grafted segments of peripheral nerve into the cerebellum to determine whether expression of these molecules would enable Purkinje cells to regenerate axons.

Previous SectionNext Section

Methods

Generation of L7/L1 and L7/GAP-43 Transgenic Mice. The L7/L1 hybrid transgene construct (Fig. 1) was generated from the full-length neuronal form of mouse L1 cDNA containing 248 base pairs in the 3′ UTR (a gift from M. Schachner, University of Hamburg, Hamburg, Germany) inserted in a vector (pL7ΔAUG) with deleted translation start sites. The pL7ΔAUG contains 1 kb of the Purkinje cell-specific L7 promoter, four exons, three introns, and 200 base pairs downstream of the TGA stop codon of the L7 gene. L1 cDNA was inserted into the BamHI site in exon 4 of the L7 gene, and the L7/L1 junctions were sequenced to confirm correct orientation of the insert. The hybrid transgene was digested from the vector by EcoRI and HindIII. The resulting 7.0-kb fragment was electroeluted, purified (with a G50 NICK column, Amersham Pharmacia Biotech), ethanol-precipitated, and dissolved in injection buffer (10 mM Tris·HCl, pH 7.4/0.1 mM EDTA in embryo-tested water), resuspended to a final concentration of 1-5 ng/ml, and injected into (C57BL6 × CBA/Ca) F1 hybrid eggs according to standard methods. Transgenic mice were identified by Southern blots of genomic DNA from tail snips. Founder mice were crossbred with C57BL/6J mice to establish transgenic lines. Expression of L1 in Purkinje cells was confirmed by in situ hybridization (see below) and L1 immunostaining with affinity-purified rabbit polyclonal antibodies against mouse L1 (21) or rat L1 (a gift from Vance Lemmon, Case Western Reserve University, Cleveland). Three founder mice with the highest levels of L1 transgene expression in Purkinje cells were used; all generated similar results.

Fig. 1.
View larger version:
Fig. 1.

Schematic illustration of the L1 transgene construct under the control of the Purkinje cell-specific promoter L7. L1 cDNA was inserted into the fourth exon of the L7 gene, and the start codon of L7 was mutated to prevent the translation of L7 protein.


L7/GAP-43 mice generated by Buffo et al. (9) were crossbred with L7/L1 mice to produce double-transgenic mice, identified by PCR of genomic DNA from tail snips. Expression of L1 and GAP-43 proteins in Purkinje cells in these mice was also confirmed by immunostaining with affinity-purified rabbit polyclonal antibodies against L1 and GAP-43.

In Situ Hybridization. The preparation of digoxigenin-labeled antisense and sense L1, GAP-43, c-Jun, and CHL1 (close homologue of L1) cRNA probes and hybridization procedures are described in ref. 22. Specificity of hybridization was verified by probing sections of cerebellum from transgenic and nontransgenic littermates, using sense and antisense probes under identical conditions. No signal was seen when antisense probes were replaced by sense probes.

Surgical Procedures. Surgery was performed on deeply anaesthetized transgenic mice aged 6-12 weeks, with WT littermates of L1 transgenic mice as controls. A 7-mm segment of tibial nerve was removed from the left thigh. A craniotomy was made through the occipital bone, the dura was opened, and the autograft was inserted 3 mm vertically into the region of the deep cerebellar nuclei, using a fine glass pipette and stereotaxic coordinates from Paxinos and Watson (23). The graft was glued to the dura, the distal was left free on the surface of the skull, and the scalp incision was closed. In a second set of experiments, the left sciatic nerve was cut 1 week before transplantation of the distal stump. Mice were housed, fed, and handled to minimize stress and postoperative pain and in strict accordance with United Kingdom government legislation on animal care and genetic manipulation.

Immunohistochemical Procedures. Animals were perfused transcardially with 30 ml of 0.1 M phosphate buffer (PB) followed by 60 ml of 4% paraformaldehyde in 0.1 M PB. The brains were removed and postfixed in the same fixative for 2 h and then immersed in 30% sucrose in 0.1 M PB (pH 7.4) at 4°C overnight. For detection of L1, frozen coronal or sagittal sections of the cerebellum were cut at 40 μm, collected in 0.1 M PBS, treated with 0.3% H2O2, rinsed in PBS, and incubated in blocking buffer (1% BSA/2% normal goat serum/0.25% Triton X-100) for 1 h at room temperature. Sections were then incubated for 48 h at 4°C in primary polyclonal antibodies against L7 (1:3,000 dilution, a generous gift from J. Oberdick, Ohio State University, Columbus) and/or calbindin D-28K (1:2,000 dilution, Chemicon) to visualize Purkinje cells and their axons. Immunohistochemical staining was performed by using the avidin-biotin-peroxidase method (Vectastain ABC Elite kit, Vector Laboratories) with 3,3′-diaminobenzidine as chromogen. Reacted sections were mounted on gelatin-coated slides, air-dried, dehydrated, coverslipped, and viewed with a microscope (Leica, Deerfield, IL).

Axon Counts. From each animal, a single section in the coronal plane of the brain that passed through the approximate center of the longitudinally sectioned graft was selected, and the number of L7 or calbindin-positive axons clearly within the graft was counted at three proximodistal levels along the graft.

Previous SectionNext Section

Results

The L7 Promoter Directs Expression of L1 and/or GAP-43 in Cerebellar Purkinje Cells. Both the L1 and GAP-43 cDNAs were inserted into the fourth exon of the L7 gene, the start codon of which had been mutated as illustrated in Fig. 1. In the three L1 transgenic lines kept for breeding, in situ hybridization with L1 riboprobe showed strong signal in Purkinje cells [compare Fig. 2 B and A (WT)]. There was strong L1 immunoreactivity in Purkinje cell bodies and axons [compare Fig. 2 D and C (WT)]. In GAP-43 transgenic mice, Purkinje cells were strongly GAP-43 immunoreactive [compare Fig. 2 F and E (WT)]. In double-transgenic animals, expression of both L1 and GAP-43 mRNAs and proteins in Purkinje cells was confirmed by in situ hybridization and immunocytochemistry.

Fig. 2.
View larger version:
Fig. 2.

Transgene expression in the Purkinje cells of L1 and GAP-43 transgenic mice and in a WT mouse. (A and B) The expression of L1 mRNA in the cerebellum of aWT(A) and an L1 transgenic (B) mouse examined by in situ hybridization. No L1 mRNA transcripts are detected in Purkinje cells of the WT mouse. Strong expression of L1, driven by the cell-specific L7 promoter, is detected in Purkinje cells of the L1 transgenic mouse. (C and D) Immunohistochemical detection of L1 expression in the cerebellum of an adult WT (C) and an L1 transgenic (D) mouse. Strong L1 immunoreactivity is apparent in the Purkinje cell bodies, axons, and dendrites in the L1 transgenic mouse but not in the Purkinje cells of the WT mouse, although L1 immunoreactivity is apparent in the molecular layer of the WT cerebellum. (E and F) Immunohistochemical detection of GAP-43 expression in the cerebellum of an adult WT (E) and a GAP-43 transgenic (F) mouse. GAP-43 immunoreactivity is absent in the Purkinje cells in the WT mouse but present in Purkinje cell bodies, axons, and dendrites in the transgenic mouse. (Scale bars, 100 μm.)


Response of Purkinje Cell Axons to Peripheral Nerve Graft in WT Mice. In WT mice, the response was similar to the response of these cells to simple axotomy (1, 9, 24). The injured Purkinje axons developed irregularly shaped, torpedo-like swellings along their initial course through the granule cell layer and terminated around the graft, commonly with spherical or more irregularly shaped end bulbs (Fig. 3A). Very few axonal sprouts were observed to emanate from the end bulbs; none was seen to extend into the graft.

Fig. 3.
View larger version:
Fig. 3.

Responses of Purkinje cell axons to axotomy and peripheral nerve graft implantation in WT, GAP-43, and L1 transgenic mice. The graft is indicated by “g.” (A) Calbindin-immunoreactive Purkinje cell axons terminating close to the graft in a WT mouse at 7 dpo. Axons formed end bulbs (arrowheads) around the graft, and there is no sign of axonal sprouting. (B) L7-immunoreactive Purkinje cell axons and end bulbs around a graft at 3 dpo in a GAP-43 transgenic mouse. Note that small sprouts (e.g., at arrowhead) emerge from some end bulbs. (C) Calbindin-positive Purkinje cell axonal sprouts close to the graft at 25 dpo in a GAP-43 transgenic mouse. No sprouts are present within the graft, and some of the sprouting axons appear to turn back at the boundary of the graft (e.g., at arrowheads; see the enlargement in Inset). (D) L7-immunoreactive Purkinje cell axonal sprouts growing toward the graft at 3 wpo (arrows) in an L1 transgenic mouse. None of the sprouts has penetrated beyond the brain-graft interface. (E) L7-immunoreactive Purkinje cell axons extending into the peripheral portion of a graft (arrows) at 8 wpo in an L1 transgenic mouse. (Scale bars, 50 μm.)


Purkinje Cell Axons Sprout but Do Not Regenerate into Peripheral Nerve Grafts in GAP-43 Transgenic Mice. After peripheral nerve transplantation in L7/GAP-43 single-transgenic mice, many injured Purkinje cell axons, identified by either L7 or calbindin immunostaining, formed end bulbs around the graft (Fig. 3B). Fine axonal sprouts were observed to emerge from such end bulbs 3-6 days postoperation (dpo) (n = 2; Fig. 3B). The number of axonal sprouts was greater at 25 dpo (n = 2; Fig. 3C). The sprouts grew toward the graft but appeared to turn back at the brain-graft interface (arrowhead in Fig. 3C), and very few were present within the graft, even after longer survival times (2-3 months after operation; n = 7). These observations are similar to those of Buffo et al. (9) on cell transplants in these mutant mice.

Purkinje Cell Axons Transiently Sprout into Peripheral Nerve Grafts in L1 Transgenic Mice. L7/L1 single-transgenic mice showed modest sprouting of Purkinje cell axons after graft implantation. Many sprouts were present at or just beyond the brain-graft interface by 25 dpo but did not extend deep into the graft (n = 14; Fig. 3D). By 2 months, however, regenerating axons had entered and were found at the periphery and occasionally deeper in the graft, but none had grown along the full length of the graft (n = 5; Fig. 3E). No labeled axons were detected in the graft 12 weeks postoperation (wpo) (n = 2). GAP-43, c-jun, and CHL1 (close homologue of L1) mRNA were not detected in Purkinje cells after graft transplantation in L7/L1 single-transgenic mice (data not shown).

Purkinje Cell Axons Show Enhanced Sprouting and Transient Regeneration into Peripheral Nerve Grafts in GAP-43/L1 Double-Transgenic Mice. At 3-6 dpo in L7/GAP-43/L1 double-transgenic mice (n = 4), there were numerous Purkinje cell axonal sprouts around the graft, and some penetrated the graft periphery (Fig. 4A). By 25 dpo (n = 2), although many regenerating axons appeared to have turned back at the brain-graft interface, some had penetrated up to 100 μm into the graft (Fig. 4 B and C). At 12 wpo (n = 5), very few labeled axons were present in the grafts (data not shown).

Fig. 4.
View larger version:
Fig. 4.

Responses of Purkinje cell axons to axotomy and peripheral nerve graft implantation in L1/GAP-43 double-transgenic mice. (A) Numerous L7-immunoreactive Purkinje cell axonal sprouts around a graft at 6 dpo. Some of the sprouts (arrows) have penetrated a short distance into the peripheral part of the graft. (B) A calbindin-immunoreactive growth cone-like structure (arrow; see the enlargement in Inset) within a graft in an L1/GAP-43 double-transgenic mouse at 25 dpo. (C) A section adjacent to the one shown in B, immunostained with L7 antibody. Purkinje cell axonal sprouts (arrows) are visualized deep within the graft. Some axons (arrowhead) appear to turn back at the brain-graft interface. (Scale bars, 50 μm.)


Purkinje Cell Axons Regenerate into and Persist in Predegenerated Peripheral Nerve Grafts in GAP-43/L1 Double-Transgenic Mice. To minimize the possible inhibitory effects of myelin debris in the peripheral nerve graft on regenerating Purkinje cell axons, the left sciatic nerve was cut 1 week before transplantation to induce Wallerian degeneration of the nerve fibers and increase expression of L1 and other CAMs at the surface of Schwann cells (25). Twenty-five days after transplantation of a predegenerated graft (n = 3), many regenerating Purkinje axons were found growing into the graft (Fig. 5A), although some turned back at the brain-graft interface. By 8-12 wpo (n = 7), many Purkinje axons were present deep within the graft (Fig. 5 B-D), and some had grown ≈2-3 mm into the most distal portion of the graft preserved in the sections (Fig. 5 E and F). In the sections selected for counts, between 2 and 29 axons were present in the proximal portion of the graft, between 2 and 19 were present in the middle portion, and between 0 and 13 were present in the distal portion of the graft (Fig. 6). In contrast, predegeneration was insufficient to induce Purkinje cell axons to grow far into the grafts in L7/GAP-43 single-transgenic (n = 5) or WT (n = 5) animals (Fig. 6). However, in two L7/L1 single-transgenic mice at 12 wpo, calbindin-positive axons were found deep within the graft and extending to its middle portion but not beyond (Fig. 6).

Fig. 5.
View larger version:
Fig. 5.

Responses of Purkinje cell axons to implantation of a predegenerated peripheral nerve graft in L1/GAP-43 double-transgenic mice. (A) Calbindin-immunoreactive Purkinje cell axonal sprouts (arrows) extending well into a graft at 24 dpo. (B) L7-immunoreactive Purkinje axonal sprouts (arrows) in the middle portion of a graft at 9 wpo. (C and D) Calbindin-immunoreactive Purkinje cell axons (arrows) in the middle portion of a graft 8 wpo. D is an enlargement of part of C. (E and F) Calbindin-immunoreactive Purkinje cell axons (arrows) in the central portion of the middle and distal graft of another two individual double-transgenic mice at 12 (E) and 10 (F) wpo. (Scale bars: A, B, and C, 50 μm; F, which applies to D and E, 25 μm.)


Fig. 6.
View larger version:
Fig. 6.

Numbers of L7 or calbindin-positive axons in the proximal, middle, and distal portions of predegenerated peripheral nerve grafts implanted into the cerebellum. All immunoreactive axons within the graft in the selected section (see Methods) were counted at three proximodistal levels: ≈500 μm from the graft tip, 1,500 μm from the graft tip, and 2,500 μm from the graft tip (≈200-400 μm from the most distal portion of the graft present in the sections). Each bar for the L1+ and GAP-43+/L1+ groups shows the mean ± SD of axon numbers for three or four individual animals. No axons were identified at the counting levels in any animal from the GAP-43+ and WT groups.


Previous SectionNext Section

Discussion

This study was designed to explore the importance of neuronal GAP-43 and L1 for axonal regeneration in vivo and to determine whether these two molecules act synergistically to promote regenerative axonal growth. Specifically, we asked whether adult cerebellar Purkinje cells, which do not normally regenerate axons into peripheral nerve grafts or express either protein, would be able to regenerate axons if they were engineered to express GAP-43 or L1 or both GAP-43 and L1. We found that Purkinje cells made to express GAP-43 or L1 showed enhanced axonal sprouting in response to the insertion of a peripheral nerve graft and that Purkinje cell axons made to express L1 showed some limited and transient penetration of the proximal portion of the graft. Purkinje cells made to express both GAP-43 and L1 showed more extensive sprouting and axonal regrowth into the proximal part of the graft in the first few weeks after implantation. Preliminary observations, based on relatively small numbers of animals and short survival times,†† suggested that there was only modest enhancement of regeneration by Purkinje cell axons in the transgenic mice. However, more extensive observations on larger numbers of double-transgenic animals at longer survival times have shown consistent and interesting effects. Most importantly, we found that when the graft was predegenerated before transplantation, Purkinje cell axons in the double-transgenic animals expressing both GAP-43 and L1 were able to penetrate the graft, grow along its entire length, and persist there for several months. Thus, we have shown that GAP-43 and L1 can act synergistically within adult intrinsic CNS neurons to convert cells that show an inadequate cell body response to axotomy and that will not normally regenerate axons into cells able to regenerate axons over considerable distances when provided with a substrate of predegenerate nerve tissue.

Purkinje cells in normal adult rodents do not express GAP-43 or L1 when intact or after axotomy (9, 21, ‡‡) and show little ability to produce regenerative axonal sprouts in the first 3 months after injury (5, 9, 21, 27). In contrast, axons in injured optic nerves sprout extensively in the first few days after axotomy (28), accompanied by strong up-regulation of GAP-43 and other growth-associated genes (5, 29, 30). Similarly, the central processes of dorsal root ganglion (DRG) neurons regenerate more successfully into nerve grafts, and ascending axons of DRG neurons in the transected dorsal columns of adult rats only sprout extensively into the lesion site if there is also a conditioning injury to their peripheral process, which causes the up-regulation of GAP-43 and other growth-related genes (31, 32). GAP-43 overexpression in Purkinje cells enhances sprouting but is insufficient to induce injured Purkinje cell axons to penetrate grafts of Schwann cells (9) or segments of peripheral nerve, perhaps because the L1-deficient axonal sprouts are not able to interact appropriately with the surfaces of Schwann cells: Although it is now recognized that there is little or no homophilic binding between the L1 isoforms expressed by Schwann cells and those expressed by neurons (33, 34), axonal L1 is important for axon/Schwann cell interactions (35). Alternatively, the axonal sprouts may detect repulsive cues from the Schwann cells or other nonneuronal cells of the grafts. If the hypothesis that GAP-43 helps to amplify the signal from guidance cues (36, 37) is correct, the expression of the GAP-43 transgene would not aid regeneration into grafts containing repulsive cues such as myelin-associated glycoprotein. In previous studies, it has been shown that injured Purkinje cell axons begin to sprout 3 months after axotomy (38), associated with a slight up-regulation of GAP-43 (39). We have not detected GAP-43 mRNA in the Purkinje cells of rats or mice up to 3 months after graft implantation (ref. 21 and unpublished data), making it unlikely that endogenous GAP-43 played a role in the axonal regrowth we observed.

L1 expression by Purkinje cells improves the ability of their axons to regenerate into nerve grafts in the cerebellum, but the extent of regrowth is limited. L1 on the surface of the axons would be expected to bind to surface molecules of graft Schwann cells, causing the transduction of signals to enhance axonal elongation. However, there is good evidence that Ig superfamily CAMs, including neural CAM (N-CAM) and L1, require the presence of GAP-43 within the axons to stimulate neurite outgrowth (20). Dunican and Doherty (36) suggest that when N-CAM, L1, or N-cadherin on a growth cone binds a suitable ligand, clustering and activation of FGF receptors leads to phosphorylation of GAP-43 and, as a result, to neurite outgrowth-promoting changes in the actin cytoskeleton. Thus, although the neurite growth-promoting effect of CAMs may be mediated by multiple intracellular signal transduction pathways (16, 37, 40), the Dunican and Doherty model would explain why coexpression of L1 and GAP-43 by neurons is associated with greater axon sprouting than occurs with the expression of either L1 or GAP-43 alone.

The observation that axons of Purkinje cells of double transgenics only grow extensively into nerve grafts that are predegenerate raises interesting questions. The effects of predegeneration may be mediated through the reduced amount of myelin (41, 42) or other inhibitory or repulsive molecules present and by enhanced expression of L1 ligands or other growth-promoting molecules in the predegenerated grafts. Wallerian degeneration occurs in nerve grafts implanted into the brain (43), and it might be expected that axons would grow into the grafts after it had occurred. However, Wallerian degeneration in grafts implanted into the brain is slower than in peripheral nerves in situ (unpublished data) and may not be completed during the time window imposed on the regeneration of axons into nerve grafts by the reaction of the host tissue. A glia limitans forms around the graft during the first week (43), and inhibitory proteoglycans (44) accumulate at the graft-brain interface. In contrast, predegenerate grafts are presumably able to support regeneration from the time they are implanted, before the host reaction becomes too extensive.

There is also the question of why some axons turn back at the brain-graft interface, even in the double transgenics. It has been widely assumed that peripheral nerve grafts provide a conducive environment for the regenerative growth of all types of axons, but it is now clear that several major categories of CNS neuron are reluctant to grow axons into such grafts (5). Axons may have turned away because of repulsive molecules in the developing scar tissue around the grafts (45). However, it seems likely that diffusible influences from fresh grafts may inhibit or repel Purkinje cell axons and account for both the turning behavior and the retraction of many double-transgenic axons from non-predegenerate nerve grafts.

Finally, when CNS neurons regenerate axons into nerve grafts, they normally up-regulate an array of growth-associated molecules (5). Presumably, (some of) these molecules play important roles in axonal regeneration. Bomze et al. (26) have shown that overexpression of GAP-43 and cyclase-associated protein-23 together in dorsal root ganglion (DRG) neurons greatly enhances their ability to regenerate their central axons into peripheral nerve grafts, but DRG neurons are peripheral nervous system neurons, which are more readily induced to regenerate than Purkinje cells and have an intrinsic ability to interact with Schwann cells. To our knowledge, the present findings are the first demonstration that intrinsic CNS neurons can be changed from a nonregenerating to a regenerating phenotype by expression of identified transgenes and suggest that a combination of GAP-43 and L1 in neurons may be of central importance to the regenerative response in the brain and spinal cord.

Previous SectionNext Section

Acknowledgments

We thank J. Oberdick for the pL7ΔAUG vector and J. Winterbottom and G. Campbell for help with parts of the study. This work was supported by the Wellcome Trust. Y.Z. is a Wellcome Trust Research Career Development Fellow, and R.S. is a Wellcome Trust Senior Research Fellow.

Previous SectionNext Section

Footnotes

  • ↵ ‡ To whom correspondence should be sent at the Neuroscience Centre, Queen Mary, University of London, 4 Newark Street, London E1 2AT, United Kingdom. E-mail: yi.zhang{at}qmul.ac.uk.

  • ↵ ∥ Present address: Howard Hughes Medical Institute, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 11724.

  • This paper was submitted directly (Track II) to the PNAS office.

  • Abbreviations: CAM, cell adhesion molecule; GAP-43, growth-associated protein 43; dpo, days postoperation; wpo, weeks postoperation.

  • ↵ †† Zhang, Y., Holtmaat, A. J. G., Verhaagen, J., Emson, P. C., Winterbottom, J., Campbell, G., Lieberman, A. R. & Anderson, P. N. (2002) Abstract Viewer/Itinerary Planner (Soc. Neurosci., Washington, DC), Program no. 635.6.

  • ↵ ‡‡ Dooley, J. M. & Aguayo, A. J. (1982) Ann. Neurol. , (abstr.).

Previous Section
 

References

  1. ↵
    Ramon y Cajal, S. (1928) Degeneration and Regeneration of the Nervous System (Oxford Univ. Press, London).
  2. ↵
    Richardson, P. M., McGuinness, U. M. & Aguayo, A. J. (1980) Nature 284 , 264-265.pmid:7360259
  3. ↵
    Berry, M., Rees, L. & Sievers, J. (1986) in Neural Transplantation and Regeneration, eds. Das, G. D. & Wallace, R. B. (Springer, New York), pp. 63-79.
  4. ↵
    Morrow, D. R., Campbell, G., Lieberman, A. R. & Anderson, P. N. (1993) Exp. Neurol. 120 , 60-69.pmid:8477829
    CrossRefMedline
  5. ↵
    Anderson, P. N., Campbell, G., Zhang, Y. & Lieberman, A. R. (1998) Prog. Brain Res. 117 , 211-232.pmid:9932411
  6. ↵
    Chaisuksunt, V., Campbell, G., Zhang, Y., Schachner, M., Lieberman, A. R. & Anderson, P. N. (2000) J. Comp. Neurol. 425 , 282-292.
    CrossRefMedlineISI
  7. ↵
    Skene, J. H. & Willard, M. (1981) J. Cell Biol. 89 , 96-103.pmid:6164683
    Medline
  8. ↵
    Aigner, L., Arber, S., Kaphammer, J. P., Laus, T., Schneider, C., Botteri, F., Brenner, H.-R. & Caroni, P. (1995) Cell 83 , 269-278.pmid:7585944
  9. ↵
    Buffo, A., Holtmaat, A. J., Savio, T., Verbeek, J. S., Oberdick, J., Oestreicher, A. B., Gispen, W. H., Verhaagen, J., Rossi, F. & Strata, P. (1997) J. Neurosci. 17 , 8778-8791.pmid:9348347
    Abstract/FREE Full Text
  10. ↵
    Maier, D. L., Mani, S., Donovan, S. L., Soppet, D., Tessarollo, L., McCasland, J. S. & Meiri, K. F. (1999) Proc. Natl. Acad. Sci. USA 96 , 9397-9402.pmid:10430954
    CrossRefMedlineISI
  11. Shen, Y., Mani, S., Donovan, S. L., Schwob, J. E. & Meiri, K. F. (2002) J. Neurosci. 22 , 239-247.pmid:11756507
    Abstract/FREE Full Text
  12. ↵
    Strittmatter, S. M., Fankhauser, C., Huang, P. L., Mashius, H. & Fishman, M. C. (1995) Cell 80 , 445-452.pmid:7859286
    Abstract/FREE Full Text
  13. ↵
    Linnemann, D., Edvardsen, K. & Bock, E. (1988) Dev. Neurosci. 10 , 34-42.pmid:3371232
    Abstract/FREE Full Text
  14. ↵
    Persohn, E. & Schachner, M. (1987) J. Cell Biol. 105 , 569-576.pmid:3301870
    CrossRefMedlineISI
  15. ↵
    Kiryushko, D., Berezin, V. & Bock, E. (2004) Ann. N.Y. Acad. Sci. 1014 , 140-154.pmid:15153429
    Medline
  16. ↵
    Haas, M. A., Vickers, J. C. & Dickson, T. C. (2004) Eur. J. Neurosci. 20 , 1436-1444.pmid:15355311
    Abstract/FREE Full Text
  17. ↵
    Dickson, T. C., Mintz, C. D., Benson, D. L. & Salton, S. R. (2002) J. Cell Biol. 157 , 1105-1112.pmid:12070130
    Abstract/FREE Full Text
  18. ↵
    Doherty, P., Williams, G. & Williams, E. J. (2000) Mol. Cell. Neurosci. 16 , 283-295.pmid:11085868
    CrossRefMedlineISI
  19. ↵
    Saffell, J. L., Williams, E. J., Mason, I. J., Walsh, F. S. & Doherty, P. (1997) Neuron 18 , 231-242.pmid:9052794
    Abstract/FREE Full Text
  20. ↵
    Meiri, K. F., Saffell, J. L., Walsh, F. S. & Doherty, P. (1998) J. Neurosci. 18 , 10429-10437.pmid:9852580
    CrossRefMedlineISI
  21. ↵
    Chaisuksunt, V., Zhang, Y., Anderson, P. N., Campbell, G., Vaudano, E., Schachner, M. & Lieberman, A. R. (2000) Neuroscience 100 , 87-108.pmid:10996461
    CrossRefMedlineISI
  22. ↵
    Zhang, Y., Campbell, G., Anderson, P. N., Martini, R., Schachner, M. & Lieberman, A. R. (1995) J. Comp. Neurol. 361 , 193-209.pmid:8543658
    Abstract/FREE Full Text
  23. ↵
    Paxinos, G. & Watson, C. (1986) The Rat Brain in Stereotaxic Coordinates (Academic, Sydney).
  24. ↵
    Dusart, I. & Sotelo, C. (1994) J. Comp. Neurol. 347 , 211-232.pmid:7814665
    CrossRefMedlineISI
  25. ↵
    Martini, R. & Schachner, M. (1988) J. Cell Biol. 106 , 1735-1746.pmid:2453520
  26. ↵
    Bomze, H. M., Bulsara, K. R., Iskandar, B. J., Caroni, P. & Pate Skene, J. H. (2001) Nat. Neurosci. 4 , 38-43.pmid:11135643
    CrossRefMedlineISI
  27. ↵
    Oestreicher, A. B. & Gispen, W. H. (1986) Brain Res. 375 , 267-279.pmid:2942220
    Abstract/FREE Full Text
  28. ↵
    Zeng, B. Y., Anderson, P. N., Campbell, G. & Lieberman, A. R. (1994) J. Anat. 185 , 643-661.pmid:7649800
    CrossRefMedlineISI
  29. ↵
    Doster, S. K., Lozano, A. M., Aguayo, A. J. & Willard, M. B. (1991) Neuron 6 , 635-647.pmid:1826603
    CrossRefMedlineISI
  30. ↵
    Schaden, H., Stuermer, C. A. & Bahr, M. (1994) J. Neurobiol. 25 , 1570-1578.pmid:7861120
  31. ↵
    Bavetta, S., Hamlyn, P. J., Burnstock, G., Lieberman, A. R. & Anderson, P. N. (1999) Exp. Neurol. 158 , 382-393.pmid:10415144
    CrossRefMedlineISI
  32. ↵
    Neumann, S. & Woolf, C. J. (1999) Neuron 23 , 83-91.pmid:10402195
    CrossRefMedlineISI
  33. ↵
    Jacob, J., Haspel, J., Kane-Goldsmith, N. & Grumet, M. (2002) J. Neurobiol. 51 , 177-189.pmid:11984840
    CrossRefMedlineISI
  34. ↵
    De Angelis, E., Watkins, A., Schafer, M., Brummendorf, T. & Kenwrick, S. (2002) Hum. Mol. Genet. 11 , 1-12.pmid:11772994
    CrossRefMedlineISI
  35. ↵
    Haney, C. A., Sahenk, Z., Li, C., Lemmon, V. P., Roder, J. & Trapp, B. D. (1999) J. Cell Biol. 146 , 1173-1184.pmid:10477768
    CrossRefMedlineISI
  36. ↵
    Dunican, D. J. & Doherty, P. (2000) Mol. Cell. Biol. Res. Commun. 3 , 255-263.pmid:10964748
    Abstract/FREE Full Text
  37. ↵
    Povlsen, G. K., Ditlevsen, D. K., Berezin, V. & Bock, E. (2003) Neurochem. Res. 28 , 127-141.pmid:12587671
    Abstract/FREE Full Text
  38. ↵
    Dusart, I., Morel, M. P., Wehrlé, R. & Sotelo, C. (1999) J. Comp. Neurol. 408 , 399-418.pmid:10340514
    CrossRefMedline
  39. ↵
    Wehrlé, R., Caroni, P., Sotelo, C. & Dusart, I. (2001) Eur. J. Neurosci. 13 , 857-870.pmid:11264659
    CrossRefMedlineISI
  40. ↵
    Rougon, G. & Hobert, O. (2003) Annu. Rev. Neurosci. 26 , 207-238.pmid:12598678
    CrossRefMedlineISI
  41. ↵
    Mukhopadhyay, G., Doherty, P., Walsh, F. S., Crocker, P. R. & Filbin, M. T. (1994) Neuron 13 , 757-767.pmid:7522484
    Medline
  42. ↵
    Shen, Y. J., DeBellard, M. E., Salzer, J. L., Roder, J. & Filbin, M. T. (1998) Mol. Cell. Neurosci. 12 , 79-91.pmid:9770342
    Medline
  43. ↵
    Campbell, G., Lieberman, A. R., Anderson, P. N. & Turmaine, M. (1992) J. Neurocytol. 21 , 755-787.pmid:1279130
    CrossRefMedlineISI
  44. ↵
    Fitch, M. T. & Silver, J. (1997) Cell Tissue Res. 290 , 379-384.pmid:9321701
    CrossRefMedlineISI
  45. ↵
    Davies, S. J. A., Goucher, D. R., Doller, C. & Silver, J. (1999) J. Neurosci. 19 , 5810-5822.pmid:10407022
    CrossRefMedlineISI
« Previous | Next Article »Table of Contents

This Article

  1. Published online before print September 29, 2005, doi: 10.1073/pnas.0505164102 PNAS October 11, 2005 vol. 102 no. 41 14883-14888
  1. AbstractFree
  2. Figures Only
  3. » Full Text
  4. Full Text (PDF)

Classifications

Social Bookmarking