SEARCH

SEARCH BY CITATION

Keywords:

  • axonal transport;
  • peripheral nerve;
  • prion protein;
  • regeneration

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animal surgery
  5. Experiment 1: degeneration studies
  6. Experiment 2: studies of regenerating parent axons.
  7. Experiment 3: studies of regenerating daughter axons
  8. Two-site immunometric assay for PrPc
  9. Western blotting
  10. β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons
  11. Results
  12. PrPc in the degenerating sciatic nerve
  13. PrPc in regenerating parent axons
  14. PrPc in regenerating daughter axons
  15. Western blot analysis of regenerating nerve
  16. Enzyme digestion
  17. Discussion
  18. Acknowledgements
  19. References

The cellular prion protein, PrPc, is a glycosylphosphatidylinositol-anchored cell surface glycoprotein and a protease-resistant conformer of the protein may be the infectious agent in transmissible spongiform encephalopathies. PrPc is localized on growing axons in vitro and along fibre bundles that contain elongating axons in developing and adult brain. To determine whether the growth state of axons influenced the expression and axonal transport of PrPc, we examined changes in the protein following post-traumatic regeneration in the hamster sciatic nerve. Our results show (1) that PrPc in nerve is significantly increased during nerve regeneration; (2) that this increase involves an increase in axonally transported PrPc; and (3) that the PrPc preferentially targeted for the newly formed portions of the regenerating axons consists of higher molecular weight glycoforms. These results raise the possibility that PrPc may play a role in the growth of axons in vivo, perhaps as an adhesion molecule interacting with the extracellular environment through specialized glycosylation.

Abbreviations used
PrP

prion protein

SDS

sodium dodecyl sulfate

PAGE

polyacrylamide gel electrophoresis

TSE

transmissible spongiform encephalopathy

Transmissible spongiform encephalopathies are infectious diseases in which the infectious agent consists largely or exclusively of aggregates of the prion protein (PrP) (Prusiner 1998). PrPres is a protease-resistant conformer of the host-encoded cellular protein PrPc whose physiological functions are unknown. PrPc, a sialoglycoprotein strongly expressed in neurones, is abundant in the adult brain where some of the protein is localized at synapses and boutons (Salès et al. 1998; Haeberléet al. 2000; Lainéet al. 2001; Ford et al. 2002; Mironov et al. 2003; Bailly et al. 2004). In previous studies we observed little, if any, PrPc along mature axonal pathways (Salès et al. 1998). However, the olfactory nerve, a nerve tract that contains growing axons throughout the life of mammals, has abundant PrPc suggesting that the levels of the protein in axons may depend on the growth state of the fibres. In support of this, we have recently reported that PrPc is localized along axon tracts in the developing brain at the time of axon growth, and that the protein is present on the surface of elongating axons (Salès et al. 2002). In vitro studies have showed that antibody cross-linking of PrPc on neurone-like cells activates p59fyn, a member of the p60src tyrosine kinase family, which is implicated in neurite growth (Mouillet-Richard et al. 2000). Additional in vitro evidence for a possible role of PrPc in axon growth comes from its apparent interaction with the neural cell adhesion molecule and its ability to bind laminin (Graner et al. 2000; Schmitt-Ulms et al. 2001).

Using a nerve cut and accumulation paradigm combined with a sensitive two-site immunometric assay, we demonstrated a significant anterograde accumulation of axonally transported PrPc in intact hamster sciatic nerve (Rodolfo et al. 1999; Moya et al. 2004; see also Borchelt et al. 1994). Western blot analysis showed that a novel 38-kDa PrPc form accounted for most or all of the rapidly transported PrPc whereas 36- and 31-kDa forms were stationary or slowly transported in the peripheral nerve. In the present study, we tested the hypothesis that axonal PrPc is regulated by the growth state of the axons in a nerve crush-induced regeneration model. Furthermore, in light of the role of rapidly transported proteins in axon growth and nerve regeneration, we predicted that the 38-kDa form is likely to be particularly increased after nerve crush.

Sciatic nerve crush axotomizes the axon fibres but leaves the surrounding nerve sheath intact. Within minutes of axotomy the cut end of the axons close and neuritic sprouts begin to appear within several hours (summarized in Fawcett and Keynes 1990). Distal to the axotomy the axons disconnected from their cell bodies begin to degenerate about a day after the transection and this Wallerian degeneration proceeds so that most axons have completely degenerated within a day or two later (Donat and Wisniewski 1973; Lubinska 1977; Bendszus et al. 2004). Schwann cells in the nerve sheath distal to the site of injury lose contact with intact axons and de-differentiate whereas macrophages infiltrate the nerve sheath starting 2 days after axotomy (Shamash et al. 2002; Bendszus and Stoll 2003; Bendszus et al. 2004; see Fawcett and Keynes 1990). Sprouts can form soon after crush, and metabolic labelling studies of newly synthesized and transported proteins have been used to estimate the front of the ensemble of regenerating axons in the nerve. These studies have reported that following nerve crush there is a delay of 1.5–2 days before growth initiation and that the growth rate is 2–4 mm/day for the front of the wave of regenerating axons (Bisby 1985; McQuarrie and Lasek 1989; Jacob and Croes 1998). In studies in rat, initial functional recovery is seen starting about 3 weeks after nerve crush, with full recovery and restoration of voluntary movement at 7 weeks (Dijkstra et al. 2000; Bendszus et al. 2004).

To test our hypothesis, we crushed the sciatic nerve, and then examined levels and electrophoretic patterns of PrPc during regeneration. Our results show that when axons degenerate PrPc is lost from the sciatic nerve. If regeneration is allowed to occur, axonally transported PrPc is significantly increased during the regrowth of axons and then levels of the protein decrease at about the time of complete functional recovery. In addition to these quantitative changes, our results show that a higher molecular weight form is prominent in regenerating axons. Analysis of the PrPc in regenerating axons reveals that the higher molecular weight forms carries sulfated β-glucuronides of the HNK1 glycodomain as well as terminal sialic acid residues.

Animal surgery

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animal surgery
  5. Experiment 1: degeneration studies
  6. Experiment 2: studies of regenerating parent axons.
  7. Experiment 3: studies of regenerating daughter axons
  8. Two-site immunometric assay for PrPc
  9. Western blotting
  10. β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons
  11. Results
  12. PrPc in the degenerating sciatic nerve
  13. PrPc in regenerating parent axons
  14. PrPc in regenerating daughter axons
  15. Western blot analysis of regenerating nerve
  16. Enzyme digestion
  17. Discussion
  18. Acknowledgements
  19. References

Adult male hamsters (130–140 g) were anaesthetized with pentobarbital (10 mg per 100 g bodyweight) and the leg opened at the mid-thigh. For the axon degeneration studies the exposed sciatic nerve was cut whereas in regeneration studies the nerve was crushed twice in the same place using a locking surgical needle holder. The crush site was marked by a loose loop of no. 5 suture silk. The wound was closed and the animals were allowed to recover in their home cages. Hamsters from matched lots served as controls and were maintained under the same conditions as the operated experimental animals. On the final day of the experiment experimental animals and the matched control animals were killed with an overdose of pentobarbital.

We used the test/section accumulation model to study axonally transported PrPc (Hässig et al. 1991; Rodolfo et al. 1999; Moya et al. 2004). In this model, a 2-mm basal segment of sciatic nerve is removed from anaesthetized adult hamsters. The 2-mm basal segment contains mainly stationary axonal proteins as well as slowly transported proteins. Twenty-four hours later a 2-mm anterograde accumulation segment proximal to the nerve cut is removed. This segment contains anterogradely transported proteins that accumulate during the 24-h period in addition to stationary and slowly transported proteins. The difference between the proximal and basal segments represents the anterogradely transported proteins over that time period.

The procedures used here were designed to minimize animal suffering and were carried out in accordance with the recommendation of the EEC (86/609/EEC) and the French National Committee for the use of laboratory animals.

At the time of nerve cut a 2-mm basal segment was removed and frozen (1 in Fig. 1, upper panel). Five days later a 6-mm segment distal to the cut was removed and frozen (2 in Fig. 1, upper panel).

image

Figure 1. Experimental conditions. The temporal order and relative positions of the nerve cut, crush and nerve segments analysed are shown. The numbers indicate the order of surgical procedures. B, basal segment; A, anterograde accumulation segment.

Download figure to PowerPoint

The nerve was crushed below mid-thigh (1 in Fig. 1, middle panel). Between 3 and 50 days later hamsters were anaesthetized, and a test section was made about 5–8 mm proximal to the crush site and a 2-mm basal section was removed (2 in Fig. 1, middle panel). Twenty-four hours later a 2-mm anterograde accumulation segment situated proximal to the test section (3 in Fig. 1, middle panel) was removed and frozen.

The nerve was crushed above mid-thigh (1 in Fig. 1, lower panel). Between 3 and 50 days later hamsters were anaesthetized, and a test section was made about 5–8 mm distal to the crush site and a 2-mm basal section was removed (2 in Fig. 1, lower panel). Twenty-four hours later a 2-mm accumulation segment proximal to the test section (3 in Fig. 1, lower panel) was removed and frozen. For western blot analysis of the regenerating nerve, a length of nerve beginning 2–3 mm proximal to the crush site and extending to the nerve cut distal to the crush site was removed 24 h after the removal of a basal segment (3 in Fig. 1, lower panel) and frozen. The frozen nerve was subsequently cut into 2-mm segments for analysis. For enzymatic digestion experiments, 10 days after nerve crush an 8-mm length of nerve distal to the crush site was removed and frozen; 2-mm segments were used for each digestion condition.

Two-site immunometric assay for PrPc

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animal surgery
  5. Experiment 1: degeneration studies
  6. Experiment 2: studies of regenerating parent axons.
  7. Experiment 3: studies of regenerating daughter axons
  8. Two-site immunometric assay for PrPc
  9. Western blotting
  10. β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons
  11. Results
  12. PrPc in the degenerating sciatic nerve
  13. PrPc in regenerating parent axons
  14. PrPc in regenerating daughter axons
  15. Western blot analysis of regenerating nerve
  16. Enzyme digestion
  17. Discussion
  18. Acknowledgements
  19. References

Quantitative evaluation of PrPc in sciatic nerve was performed as described previously (Moya et al. 2004). Briefly, 2-mm nerve segments were homogenized in 1 mL 10 mm Tris-HCl buffer, pH 7.4, 2% sodium dodecyl sulfate (SDS) and 15% saccharose. Nerve homogenates were diluted 1 : 16 in modified SDS EIA buffer (0.1 m sodium phosphate, pH 7.4, 150 mm NaCl, 0.1% bovine serum albumin, 0.01% NaN3, 2% SDS). Adult hamster brain was used as an internal standard; it was homogenized as above (10% w/v) and diluted 1 : 50 in SDS EIA buffer to give the same total protein concentration as that of the nerve segments. Diluted samples (100 µL) in duplicate were transferred to wells of microtitre plates pre-coated with anti-PrP monoclonal SAF-15 (10 µg/µL; Service de Pharmacologie et d’ Immunologie, CEA, Gif-sur-Yvette, France) and left for 4 h at room temperature (18°C).

The plates were then washed with 10 mm potassium phosphate buffer, pH 7.4, containing 0.05% Tween-20. Some 100 µL of a second anti-PrP monoclonal (Pri 308; SPI, CEA) coupled to the G4 subunit of acetylcholinesterase was added to each well and the plates were incubated overnight at 4°C. After washing in phosphate buffer, 200 µL of Ellman's reagent in 100 mm phosphate buffer, pH 7.4, containing 0.25 mm acetylthiocholine, 0.5 mm 5,5′dithiobis (2-nitrobenzoic acid) and 0.1 mm iso ompa (an inhibitor of non-specific cholinesterase), was added and the optical density of individual wells was measured after 30 min of enzyme reaction.

The optical density values of the duplicate dilutions for each nerve sample were first averaged. For comparisons of stationary PrPc, the optical density value of the basal segment from the experimental groups was normalized to the mean optical density observed in intact, non-regenerating matched control animals. Axonally transported PrPc was determined by subtracting the value of the basal segment from the proximal segment for each animal. The accumulated PrPc value in the regenerating animals was then normalized to the mean value of accumulated PrPc in the matched control group. Data from 5–8 hamsters per group at each time point were plotted as a mean ± SEM percentage of control and analysed statistically using Student's two-tailed t-test (StatView version 4.02; Abacus Concepts, Berkeley, CA, USA). Differences were considered significant at p < 0.05.

Western blotting

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animal surgery
  5. Experiment 1: degeneration studies
  6. Experiment 2: studies of regenerating parent axons.
  7. Experiment 3: studies of regenerating daughter axons
  8. Two-site immunometric assay for PrPc
  9. Western blotting
  10. β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons
  11. Results
  12. PrPc in the degenerating sciatic nerve
  13. PrPc in regenerating parent axons
  14. PrPc in regenerating daughter axons
  15. Western blot analysis of regenerating nerve
  16. Enzyme digestion
  17. Discussion
  18. Acknowledgements
  19. References

For western blotting, 2-mm nerve segments were homogenized in gel sample buffer (200 µL) using a glass–glass homogenizer. The solubilized proteins were separated by SDS–polyacrylamide gel electrophoresis (PAGE) and transferred to Immobilon P polyvinylidene fluoride membranes. PrPc was detected with monoclonal antibody 3F4 (1 : 500–1000; Senetek, St Louis, MO, USA) or SAF-32 (2 µg/mL; SPI, CEA, Saclay, France), a horseradish peroxidase-conjugated anti-mouse antiserum (Sigma, Poole, UK) and enhanced chemiluminescence (Amersham Biosciences, Amersham, UK) as described previously (Salès et al. 1998; Moya et al. 2004).

β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animal surgery
  5. Experiment 1: degeneration studies
  6. Experiment 2: studies of regenerating parent axons.
  7. Experiment 3: studies of regenerating daughter axons
  8. Two-site immunometric assay for PrPc
  9. Western blotting
  10. β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons
  11. Results
  12. PrPc in the degenerating sciatic nerve
  13. PrPc in regenerating parent axons
  14. PrPc in regenerating daughter axons
  15. Western blot analysis of regenerating nerve
  16. Enzyme digestion
  17. Discussion
  18. Acknowledgements
  19. References

Segments from 10-day regenerating nerve distal to the crush were homogenized in buffer [2 mm in 100 µL 50 mm Tris, pH 7.4, 150 mm NaCl, 5 mm EDTA, 2% (w/v) SDS]. β-Glucuronidase (0–2000 mU; Roche Diagnostics, Mannheim, Germany) was added and homogenates were incubated overnight at 37°C for oligosaccharide digestion. The reaction was stopped by adding 100 µL SDS sample buffer and boiling for 5 min; 125 µL of each sample was separated by SDS–PAGE and transferred to polyvinylidene fluoride membranes as above and detected by western blot analysis using the SAF-32 antibody.

For neuraminidase digestion, regenerating nerve 10 days after crush was homogenized [2 mm in 200 µL 50 mm Tris, pH 7.4, 2% (w/v) SDS] and protein was precipitated with 800 µL methanol for 2 h at − 20°C. The precipitated proteins were centrifuged at 15 000 g for 10 min, resuspended in 20 µL distilled water and adjusted to 2% (v/v) NP-40, 155 mm NaCl, 1% bovine serum albumin, and 50 units α2–3 neuraminidase (New England Biolabs, Hitchin, Hertfordshire, UK) or 50 units total neuraminidase (New England Biolabs) in a final reaction volume of 50 µL. Digestion was carried out at 37°C overnight and the reaction stopped by addition of 50 µL 2 × SDS sample buffer. Samples were boiled for 5 min before SDS–PAGE and resolved proteins were transferred to polyvinylidene fluoride membranes.

PrPc in the degenerating sciatic nerve

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animal surgery
  5. Experiment 1: degeneration studies
  6. Experiment 2: studies of regenerating parent axons.
  7. Experiment 3: studies of regenerating daughter axons
  8. Two-site immunometric assay for PrPc
  9. Western blotting
  10. β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons
  11. Results
  12. PrPc in the degenerating sciatic nerve
  13. PrPc in regenerating parent axons
  14. PrPc in regenerating daughter axons
  15. Western blot analysis of regenerating nerve
  16. Enzyme digestion
  17. Discussion
  18. Acknowledgements
  19. References

In light of a recent report that PrPc is localized in Schwann cells in mice overexpressing the protein (Follet et al. 2002), we first examined PrPc when axons were eliminated from the nerve. Nerve cut results in Wallerian degeneration of the axons distal to the section, which typically starts about a day after the transection and proceeds so that most axons have degenerated a day or two later (see Introduction). In the present studies, 5 days after transection the signal for PrPc had virtually disappeared from the nerve distal to the cut compared with that a basal segment of nerve removed at the time of nerve cut (Fig. 2). Similar results were obtained in a separate experiment using the SAF-32 antibody (data not shown). The loss of PrPc occurs even though Schwann cells remain within the axonless nerve sheath and macrophages have infiltrated into the area at this time (Shamash et al. 2002; Bendszus and Stoll 2003; see Fawcett and Keynes 1990). These results show that at least in wild-type hamster sciatic nerve PrPc is in or on the axons.

image

Figure 2. PrPc in the degenerating sciatic nerve. Five days after nerve section PrPc detected with 3F4 antibody was greatly reduced in the segments distal to the cut (D1–4) compared with a basal segment (Bas) removed at the time of the section. Some 50 µg hamster cortex (Ctx) homogenate was run for comparison. The black arrow indicates diglycosylated PrPc, the grey arrow indicates monoglycosylated PrPc and the white arrow indicates the position of non-glycosylated PrPc revealed on longer exposures of the film. The position of the 31-kDa molecular weight marker is indicated to the left.

Download figure to PowerPoint

PrPc in regenerating parent axons

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animal surgery
  5. Experiment 1: degeneration studies
  6. Experiment 2: studies of regenerating parent axons.
  7. Experiment 3: studies of regenerating daughter axons
  8. Two-site immunometric assay for PrPc
  9. Western blotting
  10. β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons
  11. Results
  12. PrPc in the degenerating sciatic nerve
  13. PrPc in regenerating parent axons
  14. PrPc in regenerating daughter axons
  15. Western blot analysis of regenerating nerve
  16. Enzyme digestion
  17. Discussion
  18. Acknowledgements
  19. References

We assessed changes in PrPc levels in the nerve proximal to the crush site (Fig. 1, top). This portion of the nerve contains the parent axons from which regenerative sprouts grow out. Quantitative two-site immunometric assay for PrPc was carried out on a basal segment that contains primarily stationary proteins i.e. those distributed along the axons as well as slowly transported proteins (Hässig et al. 1991; Rodolfo et al. 1999; Moya et al. 2004). Levels of PrPc were significantly increased by around 40% in the basal segment of parent axons 3 days after nerve crush compared with levels in non-crushed control animals (Fig. 3a). At 10 days after nerve crush the PrPc level was more than double the control value. At 20 and 50 days after crush the levels in the basal segment had returned to control values.

image

Figure 3. Changes in PrPc in regenerating parent axons. (a) Two-site immunometric assay shows significant increases in PrPc in the basal segment of regenerating axons at 3 and 10 days after crush compared with levels in control non-crushed nerve. (b) Levels of axonally transported PrPc are significantly increased at 10 and 20 days after crush. Values represent the mean ± SEM of 5–8 animals per time point. Student's two-tailed t -test was used for statistical analysis. (c) Western blot analysis of basal and proximal segments in intact sciatic nerve and in 10-day regenerating parent axons using the 3F4 antibody. The black arrow indicates diglycosylated PrPc, the grey arrow indicates monoglycosylated PrPc and the white arrow indicates non-glycosylated PrPc. The asterisk indicates the higher molecular weight PrPc centred at about 38 kDa which accumulates in intact and regenerating axons. The western blot image is representative of the results from five different animals.

Download figure to PowerPoint

In order to evaluate changes in anterogradely transported PrPc independent of the changes in stationary PrPc, we subtracted the PrPc signal in the basal segment from that in the accumulation segment. The level of accumulated PrPc was slightly, but non-significantly increased 3 days after crush (Fig. 3b). At 10 days after nerve crush, anterogradely accumulated PrPc was significantly increased by about 50% compared with the level in controls and at 20 days the level was about twice that observed in intact animals. Thus, during regeneration, levels of anterogradely transported PrPc were significantly increased in the parent axons over and above the increase in protein in the ‘stationary’ axon compartment (i.e. basal segment). By 50 days after crushing the accumulated PrPc in parent axons had diminished to near control levels.

To determine whether the increase of PrPc in parent axons was due to an increase in the 36- and 33-kDa stationary forms or the axonally transported higher molecular weight forms, we examined the western blot pattern in basal and proximal segments in intact nerve and parent axons after regeneration for 10 days. On western blots of the basal segment from intact nerve we observed the 36- and 33-kDa bands and a faint signal corresponding to a 28-kDa band, which probably correspond to diglycosylated, monoglycosylated and non-glycosylated forms of the protein respectively (Fig. 3c). In the accumulation segment of an intact animal we also observed an anterogradely transported form with an apparent molecular weight centred around 38 kDa (Fig. 3c) (see also Rodolfo et al. 1999). Interestingly, PrPc forms with a similar migration rate were apparent in the basal segment containing parent axons 10 days after nerve crush (Fig. 3c). The western blot signal for higher molecular weight PrPc forms was more intense in the accumulation segment of regenerating nerve than in the intact nerve (Fig. 3c). These results are consistent with the quantitative increase in PrPc in the basal and proximal segments demonstrated by the two-site immunometric assay, and show that the higher molecular weight 38-kDa PrPc is particularly abundant in the accumulation segment of regenerating nerve.

PrPc in regenerating daughter axons

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animal surgery
  5. Experiment 1: degeneration studies
  6. Experiment 2: studies of regenerating parent axons.
  7. Experiment 3: studies of regenerating daughter axons
  8. Two-site immunometric assay for PrPc
  9. Western blotting
  10. β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons
  11. Results
  12. PrPc in the degenerating sciatic nerve
  13. PrPc in regenerating parent axons
  14. PrPc in regenerating daughter axons
  15. Western blot analysis of regenerating nerve
  16. Enzyme digestion
  17. Discussion
  18. Acknowledgements
  19. References

After a delay to initiate growth of 1.5–2 days, sciatic nerve axons regrow at a rate of 2–4 mm/day (see Introduction). Thus, 3 days after nerve crush, the daughter axons will have grown out a maximum of 4–6 mm from the crush site and the nerve beyond this distance will contain mostly degenerating axon debris and few, if any, regenerating daughter axons. Accordingly, the two-site immunometric assay for PrPc in basal segments at about 5–8 mm distal to the crush showed a significant decrease in PrPc 3 days after nerve crush (Fig. 4a). By 10 days after crush many daughter axons had advanced far enough to permit reliable analysis of regenerating axons in basal and proximal segments. At this time overall levels of PrPc in the basal segment distal to the crush site had increased significantly compared with the levels observed 3 days after crush (p = 0.025) and were elevated, although not significantly, compared with levels in matched control animals (Fig. 4a). Levels of stationary and/or slowly transported PrPc in regenerating daughter axons were similar to those in intact control nerve at 20 and 50 days after nerve crush.

image

Figure 4. Changes in PrPc in regenerating daughter axons. (a) After an initial decrease due to degeneration of the axons, PrPc increases in the basal segment of regenerating axons at 10 days after crush and remains indistinguishable from control levels at 20 and 50 days after crush. (b) Anterogradely accumulating PrPc is significantly decreased relative to control values at 3 days, its levels reach control values at 10 days and are significantly raised by 20 days after nerve crush. Values represent the mean ± SEM of 5–8 animals per time point. Student's two-tailed t -test was used for statistical analysis.

Download figure to PowerPoint

Distal to the nerve crush the level of axonally transported PrPc was significantly decreased at 3 days after injury, consistent with the loss of axons by Wallerian degeneration (Fig. 4b). By 10 days after crush, the amount of PrPc in the anterograde accumulation segment containing regenerating daughter axons had reached levels seen in intact control animals. Twenty days after crush, the level of anterogradely accumulating PrPc in the daughter axons was about twice that in controls. It had returned to near control levels by 50 days after crush.

Western blot analysis of regenerating nerve

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animal surgery
  5. Experiment 1: degeneration studies
  6. Experiment 2: studies of regenerating parent axons.
  7. Experiment 3: studies of regenerating daughter axons
  8. Two-site immunometric assay for PrPc
  9. Western blotting
  10. β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons
  11. Results
  12. PrPc in the degenerating sciatic nerve
  13. PrPc in regenerating parent axons
  14. PrPc in regenerating daughter axons
  15. Western blot analysis of regenerating nerve
  16. Enzyme digestion
  17. Discussion
  18. Acknowledgements
  19. References

Western blot analysis of the regenerating nerve revealed important differences in the pattern of PrPc between regenerating and intact nerve. Three days after nerve crush the PrPc signal in the nerve segment adjacent to the crush site was greatly increased in intensity (Fig. 5). The intense signal in the 3-day regenerating nerve was particularly evident in the molecular weight range of 38-kDa axonally transported PrPc (asterisk, left lane of control nerve in Fig. 5) as well as spreading to higher apparent molecular weights. This probably reflects the accumulation of transported PrPc at the site of axon lesion (see below). Distally, PrPc could be detected up to about 4 mm from the crush site, which corresponds well with the estimated distance reached by the fastest regrowing axons at this time. What is striking is that higher molecular weight forms (asterisk) were abundant in these segments, whereas there was comparatively less of the predominant stationary form observed in the basal segment from a control nerve. In segments further distal at the location of the anterograde accumulation and basal segments the level of PrPc was very low compared with that in intact nerve. This is consistent with the semiquantitative results that showed a significant decrease in PrPc concurrent with the loss of axons at this time (Fig. 4b).

image

Figure 5. Western blot analysis of PrPc in the intact and regenerating sciatic nerve. The left panel shows the pattern of PrPc in the basal segment (Bas) and anterograde accumulating segment (Ant) in an intact nerve. The asterisk indicates the higher molecular weight PrPc signal centred at about 38 kDa that accumulates after nerve section. The 3F4 antibody was used to probe western blots of nerves that were cut into sequential 2-mm segments 3, 10 and 20 days after crush. Along the bottom of each panel, a schematic representation of the nerve pieces analysed in each lane is shown. The nerve is orientated with the dorsal root ganglion/spinal cord to the left. The schematic shows the crush site, the parent portion of the nerve to the left (black), the daughter axon containing nerve to the right of the crush, the estimated extent of the regenerating axons indicated in grey, and the position of the test section and basal segment. The positions of the 45- and 31-kDa markers are indicated to the left of each panel. Similar results were observed in western blots of at least two different animals at the three times after nerve crush.

Download figure to PowerPoint

Ten days after nerve crush higher molecular weight PrPc was prominent in nerve segments with regenerating axons. At the crush site and just proximal to it the PrPc signal was more intense than that at the first distal segment, consistent with an accumulation of the protein at the crush site. The pattern of PrPc in basal segments shows that higher molecular weight forms are stationary or slowly transported in the regenerating daughter axons. The segment adjacent to the test section showed a stronger signal for most or all PrPc bands compared with the basal segment, demonstrating the anterograde accumulation of these forms in the regenerating daughter axons.

Twenty days after nerve crush the pattern and intensity of PrPc in the nerve proximal to the crush site started to resemble that seen in the basal segment of an intact control nerve. At the crush site the signal was stronger than that in the proximal nerve and the higher molecular weight PrPc bands were obvious. Distal to the crush, the basal segment of the daughter axons contained the high molecular weight PrPc and the accumulation segment showed strong anterograde accumulation of higher molecular weight forms of PrPc. The accumulation of PrPc in the daughter axons corroborated the significant increase in axonally transported PrPc demonstrated by two-site immunometric assay (see Fig. 4b).

Enzyme digestion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animal surgery
  5. Experiment 1: degeneration studies
  6. Experiment 2: studies of regenerating parent axons.
  7. Experiment 3: studies of regenerating daughter axons
  8. Two-site immunometric assay for PrPc
  9. Western blotting
  10. β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons
  11. Results
  12. PrPc in the degenerating sciatic nerve
  13. PrPc in regenerating parent axons
  14. PrPc in regenerating daughter axons
  15. Western blot analysis of regenerating nerve
  16. Enzyme digestion
  17. Discussion
  18. Acknowledgements
  19. References

Nerve segments with newly formed daughter axons contain PrPc forms that have a slower electrophoretic migration than the PrPc in intact nerve (Fig. 5). In order to partially characterize these higher molecular weight forms, we enzymatically digested proteins contained within solubilized 10-day regenerating daughter axon segments. β-Glucuronidase digestion induced a progressive loss of the highest molecular weight band with increasing concentrations of enzyme (Fig. 6a), demonstrating that this PrPc form carries the HNK-1 glycodomain. Digestion with either α-2,3-neuraminidase or total neuraminidase also reduced the abundance of the slowest migrating PrPc, showing that it carries sialic acid (Fig. 6b).

image

Figure 6. Western blot analysis of PrPc in regenerating daughter axons after β-glucuronidase and neuraminidase digestion. (a) 2-mm nerve equivalents were incubated in increasing concentrations of β-glucuronidase as indicated or solubilized directly in sample buffer (SB). The western blot was probed with the SAF-32 monoclonal antibody. The black arrow indicates the slowest migrating PrPc band that is greatly reduced after digestion in increasing concentrations of enzyme (white arrow). Similar results were observed in two separate experiments. The position of the 31-kDa molecular weight marker is indicated to the left. (b) Western blot analysis with SAF-32 of PrPc in daughter axons after incubation with no enzyme (No), or digestion with α2,3-neuraminidase (α2,3) or total neuraminidase (Total). The black arrow indicates the PrPc band that is reduced after neuraminidase digestion (white arrow).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animal surgery
  5. Experiment 1: degeneration studies
  6. Experiment 2: studies of regenerating parent axons.
  7. Experiment 3: studies of regenerating daughter axons
  8. Two-site immunometric assay for PrPc
  9. Western blotting
  10. β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons
  11. Results
  12. PrPc in the degenerating sciatic nerve
  13. PrPc in regenerating parent axons
  14. PrPc in regenerating daughter axons
  15. Western blot analysis of regenerating nerve
  16. Enzyme digestion
  17. Discussion
  18. Acknowledgements
  19. References

Based upon our observations that PrPc is abundant on elongating axons during development and in adult axon tracts that contain growing fibres (Salès et al. 1998, 2002) we hypothesized that the levels of this protein would be increased in regenerating axons compared with intact mature axons. To test this we used a nerve crush to induce massive coordinated axon regrowth in the adult hamster sciatic nerve. Hamster was chosen for these studies because many reagents react poorly or not at all with rat PrPc and the rat is not readily susceptible to experimental transmissible spongiform encephalopathy, precluding future studies of disease progression and peripheral nerve damage. Our results show highly significant changes in PrPc during nerve regeneration.

Three days after crush there is a significant increase in PrPc levels in the basal segment of parent axons even though there is little or no change in the amount of PrPc anterogradely transported. This increase in the basal segment is probably due to a back-up of the transported protein at the damaged axon endings during the 3 days after injury, as has been described for other transported proteins such as acetylcholinesterase (Hässig et al. 1991). A week later the axonal transport of PrPc is significantly up-regulated and the PrPc in the basal segment continues to increase. The presence of 38-kDa PrPc on western blots of the basal segment at this time is consistent with the backing up of anterogradely transported PrPc in the parent axons.

Distal to the injury, PrPc is greatly diminished in the nerve at 3 days after nerve crush, coincident with Wallerian degeneration of the axons. If axons are prevented from regenerating into the distal nerve by a nerve cut (experiment 1) PrPc remains low or absent from the nerve sheath at 5 days even though Schwann cells and macrophages are present (Shamash et al. 2002; Bendszus and Stoll 2003; Bendszus et al. 2004; see Fawcett and Keynes 1990). If nerve regeneration is allowed to proceed, by 10 days after crush newly formed daughter axons have grown out far enough to be in the region of the basal segment dissection and the test section. At this time the level of PrPc in the stationary compartment in daughter axons is comparable to that in intact axons. Similarly, the level of axonally transported PrPc in the daughter axons is also comparable to control values at this time. The western blot analyses of the daughter axons show that higher molecular weight forms of PrPc accumulate in the daughter axons as well as being present in basal segment. Thus, in axons that are immature but have not yet reached their target and re-established connections, PrPc is transported and deposited along the axons as they grow out.

Our experiments show a significant doubling of axonally transported PrPc in the parent axons at 20 days after nerve crush. PrPc levels in the basal segment were near control values at this time, suggesting that most or all of the axonally transported protein is destined for the daughter axons of the regenerating nerve. This is directly demonstrated by the comparable two-fold increase (relative to the same control group) in anterogradely transported PrPc in the daughter axons at 20 days. In studies in rat, many sciatic nerve axons have regenerated to their target and formed functional connections starting about 3 weeks after nerve crush and full recovery is observed at 7 weeks (Dijkstra et al. 2000; Bendszus et al. 2004). If the rat data can be extrapolated to hamster, then the time of the largest increase in anterogradely transported PrPc coincides with the initial period of functional recovery after nerve injury and, by the time full recovery is achieved, PrPc in the regenerated nerve has returned to control values (i.e. at 50 days).

The increase during nerve regeneration and subsequent decrease is not a general phenomenon for glycosylphosphatidylinositol (GPI)-linked glycoproteins as acetylcholinesterase follows a very different time course. Nerve crush leads to rapid decrease in axonally transported acetylcholinesterase activity within 1–3 days, and it remains significantly diminished for 2 weeks before recovering slightly at 3 weeks (Filliatreau et al. 1994). Rather the time course for axonally transported PrPc in the regenerating nerve resembles that of GAP-43, the prototypical growth-associated protein. GAP-43 synthesis and anterograde transport begin to increase within a few days of sciatic nerve crush, reach half-maximal levels at the end of the first week, and then double to maximal values between 2 and 3 weeks before returning to baseline values between 40 and 84 days (Bisby 1988). Thus, the quantitative and temporal changes in axonally transported PrPc in the peripheral nerve reported here are consistent with the protein being a growth-associated protein.

Our western blot analysis shows that it is the higher molecular weight forms in particular that are sent to the growing axon segments. This raises the question as to the nature of the higher molecular weight forms of the protein. In light of the considerable heterogeneity in PrPc glycosylation (Endo et al. 1989; Rudd et al. 1999) one possibility is a difference in glycosylation. α2–8-Linked polysialic acid is one N-linked glycan modification of adhesion/recognition molecules that play an important role in axon growth and/or guidance, although this is almost exclusively expressed on the neural cell adhesion molecule (see Brusés and Rutishauser 2001). Another glycomodification implicated in cell recognition, adhesion and guidance is a terminal sulfated glucuronyl motif that constitutes the HNK-1 epitope (Voshol et al. 1996; Schachner and Martini 1995).

Recent studies show that the cell lines GT1-7 and PC12 express higher molecular weight glycoforms of PrPc and that in other cells the expression of various glycoforms depends on cell density and the level of differentiation (Monnet et al. 2003). Additionally, mouse brain PrPc has been shown to carry the HNK-1 glycodomain (Chen et al. 2003). Such results have led to the proposition that PrPc may play a role in cell differentiation and/or cell–cell interactions such as neural recognition. Our results now provide in vivo evidence for PrPc glycoform complexity that is growth dependent.

Enzymatic digestion shows that PrPc in regenerating daughter axons contains α2–3-linked sialic acid. The similar change observed after total neuraminidase digestion suggests that it is unlikely that the PrPc forms up-regulated during nerve regeneration carry α2–8-linked sialic acid, consistent with studies of the protein from adult brain (Rudd et al. 1999). Enzymatic removal of sulfated glucuronides demonstrates that the PrPc anterogradely transported to growing daughter axons carries the HNK-1 glycodomain. Receptors for the HNK-1 glycodomain include laminins and sulfoglucuronyl carbohydrate binding protein-1 (SBP-1), which appears to be identical to amphoterin, also known as high-mobility group protein-1 (HMG-1) and P30 (Daston and Ratner 1991; Hall et al. 1997; Nair et al. 1998; Fages et al. 2000; Rauvala et al. 2000; Chou et al. 2001).

It is interesting to note that these two HNK-1 receptors are present in the nerve sheath during axon regeneration. Laminin is a major constituent of the basal lamina of the nerve, which remains intact after crush, and has been shown to have neurite growth-promoting effects (Sandrock and Matthew 1987; Agius and Cochard 1998). However, the laminin-binding molecule on growing sciatic nerve axons has not been identified definitively. Amphoterin (SBP-1/HMG-1/P30) has major neurite growth-promoting effects and is expressed in Schwann cells during development (Rauvala and Pihlaskari 1987; Daston and Ratner 1991). Although mature Schwann cells in the intact nerve do not appear to express P30/amphoterin, when contact with axons is lost following a nerve crush, for example, Schwann cells express abundant P30 suggesting a role for the protein in the interaction between axons and Schwann cells (Daston and Ratner 1991).

Our results show a significant increase in the higher molecular weight β-glucuronidase-sensitive forms of PrPc in regenerating axons distal to the crush where the basal lamina and P30-expressing Schwann cells are found. These higher molecular weight PrPc forms are axonally transported to the daughter axons where they appear to be deposited along the fibres as they are readily apparent in the basal segment. Thus, PrPc in growing daughter axons may act in axon adhesion or guidance along the basal lamina through the HNK-1 domain or perhaps by direct polypeptide backbone interactions with laminin as shown in cultured embryonic cells (Graner et al. 2000). Alternatively, but not mutually exclusively, PrPc on daughter axons may interact with P30 expressed in the Schwann cells. The signal transduction pathway(s) activated by such an interaction are unknown. However, it has been shown that PrPc can activate p59fyn kinase in a neuronal-like cell line in vitro and that regenerating axons increase expression of the related p60src (Mouillet-Richard et al. 2000; Zhao et al. 2003).

In summary, our results show that axonally transported PrPc in the peripheral nerve is significantly increased during axon regeneration after nerve crush with a time course similar to a growth-associated protein profile. During the regrowth of axons, anterogradely transported PrPc is destined for the newly formed daughter axons. In the daughter axons, PrPc is predominantly of a higher molecular weight and this PrPc contains sulfated glucuronides and sialic acid consistent with a role in adhesion or guidance of the regenerating axons.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animal surgery
  5. Experiment 1: degeneration studies
  6. Experiment 2: studies of regenerating parent axons.
  7. Experiment 3: studies of regenerating daughter axons
  8. Two-site immunometric assay for PrPc
  9. Western blotting
  10. β-Glucuronidase and neuraminidase digestion of PrPc in regenerating daughter axons
  11. Results
  12. PrPc in the degenerating sciatic nerve
  13. PrPc in regenerating parent axons
  14. PrPc in regenerating daughter axons
  15. Western blot analysis of regenerating nerve
  16. Enzyme digestion
  17. Discussion
  18. Acknowledgements
  19. References