Address correspondence and reprint requests to Rashmi Kothary, Ottawa Health Research Institute, 501 Smyth Road, Ottawa, Ontario, Canada K1H 8L6. E-mail: firstname.lastname@example.org
Dystonia musculorum (dt) mice suffer from a severe sensory neuropathy caused by mutations in the gene encoding the cytoskeletal cross-linker protein dystonin/bullous pemphigoid antigen 1 (Bpag1). Loss of function of dystonin/Bpag1 within neurons leads to a loss in the maintenance of cytoskeletal organization and to the development of focal axonal swellings prior to death of the neuron. In the present study, we demonstrate that neurons within the sciatic nerves of dt27J mice undergo axonal degeneration as has been previously reported for the dorsal roots. Furthermore, ultrastructural studies reveal a perturbed organization of the neurofilament and microtubule networks within the axons of sciatic nerves in dt27J mice. The disrupted cytoskeletal organization suggested that axonal transport is affected in dt mice. To address this, we assessed fast axonal transport by measuring the rate of accumulation of acetylcholinesterase (AChE) proximal and distal to a surgically introduced ligature on the sciatic nerves of normal and dt27J mice. Our findings demonstrate that axonal transport of AChE in both orthograde and retrograde directions is markedly affected, and allow us to conclude that axonal transport defects do exist in the sciatic nerves of dt27J mice.
Dystonia musculorum (dt) is a hereditary sensory neuropathy of the mouse in which animals display progressive loss of limb coordination starting in the second week of life (Duchen et al. 1963; Duchen and Strich 1964; Duchen 1976). By the third week of age, dt mice succumb to the disease and die of unknown causes. The original description of the pathology details degeneration of nerve fibers in the sensory roots and ganglia of spinal and cranial nerves (Duchen et al. 1963; Duchen and Strich 1964; Sotelo and Guenet 1988). The dorsal root and dorsal root ganglia (DRG) of dt mice are considerably smaller than those of wild-type mice. Accumulation of organelles and empty vacuoles within focal axonal swellings in sensory neurons is a hallmark of the dt neuropathology (Janota 1972). More recent work has demonstrated that the neurofilament and microtubule network organization within the dorsal roots of dt mice is perturbed (Dalpéet al. 1998), and that this anomaly is already evident at embryonic stage 14 (Bernier and Kothary 1998). Taken together, the above results demonstrate that loss of cytoskeletal organization precedes neurodegeneration in dt mice.
Based on the structure of the neuronal isoform of dystonin/Bpag1, it is reasonable to assume that it is involved in the cross-linking of cytoskeletal elements. The observed loss of cytoskeletal organization in dt neurons is consistent with this presumption. We therefore propose that cytoskeletal disorganization in dt neurons is the primary event in the pathogenesis in dt mice. Furthermore, we suggest that this disruption progressively leads to a perturbation in axonal transport, formation of axonal swellings, and eventually to neurodegeneration. Although axonal transport has been predicted to be affected in dt mice (Janota 1972), it has nonetheless remained to be experimentally tested. Axonal transport can be separated into two categories – slow and fast. For example, neurofilament and tubulin proteins are slowly transported at a rate of 0.2–1 mm/day, whereas membranous organelles and neurotransmitters are rapidly transported at a rate of 50–400 mm/day (Okabe and Hirokawa 1989). Since dt mice do not survive much beyond the third week of life and due to the progressive nature of the disease, steady-state measurements of slow axonal transport are difficult. We have therefore chosen to examine the orthograde and retrograde transport of acetylcholinesterase (AChE), a neurotransmitter-related enzyme that is transported bi-directionally in axons by the mechanism of fast transport. We demonstrate that sciatic nerves from dt mice are smaller, have axonal loss and display impaired fast axonal transport.
Materials and methods
dt27J mice originated from a re-mutation to the dystonia musculorum locus which occurred in the congenic strain B10.PL(73)/Sn. The breeding colony was obtained from The Jackson Laboratory (Bar Harbor, ME, USA). dt27J mice were identified by using a restriction fragment length polymorphism analysis as described previously (Bernier et al. 1995). The experimental protocols were approved by the Animal Care Committee of the University of Ottawa. Care and use of experimental animals followed the guidelines of the Canadian Council on Animal Care.
Homozygous dt27J and wild-type mice at P15 were anesthetized by intraperitoneal injection of tribromoethanol (avertin). Each pup was perfused transcardially with 5 mL of phosphate-buffered saline followed by 10 mL of Karnovsky's fixative (4% paraformaldehyde, 2% glutaraldehyde and 0.1 m cacodylate in phosphate-buffered saline, pH 7.4). Sciatic nerves were collected from each animal under a stereomicroscope, cut into portions of 3 mm thickness and fixed for 4 h in Karnovsky's fixative at 4°C. These portions were subsequently washed twice in 0.1 m cacodylate buffer for 1 h and once overnight at room temperature. Using a surgical blade and under a light microscope, each portion of the nerve was sliced into five or six transverse slices and some into five or six longitudinal slices. Each slice was less than 2 mm in diameter. After 1 h of post-fixation in 2% osmium tetroxide in 0.1 m cacodylate buffer at 4°C, all sections were washed twice for 5 min each in distilled water. Sections were dehydrated, infiltrated in spur monomer (Electron Microscopy Sciences, Fort Washington, PA, USA) with three changes over 24 h at room temperature and embedded in liquid spur resin at 70°C overnight. Ultrathin sections (90 nm) from the sciatic nerve were collected onto 200-mesh copper grids. Sections were pre-treated with 2% aqueous uranyl acetate, stained with Reynold's lead citrate and observed with the use of a transmission electron microscope (GEOL 1010). For light microscopy, sections of 0.5 µm were stained with toluidine blue. All histology and ultrastructural observations were based on at least three homozygous dt27J mice at P15 showing ataxic symptoms and three wild-type mice of the same age.
Ligature of the sciatic nerve
Approximately 30–60 min before surgery, an analgesic drug (buprenorphine) was injected subcutaneous to a dt27J/+ heterozygous mother lactating a P14 litter. In the surgery room, each pup was anesthetized with an inhalant anesthetic (isoflurane with oxygen) using specialized equipment for administration and constant monitoring of the pup. Ophthalmic ointment was applied to each eye to avoid dehydration during the anesthesia. The mouse thigh was shaved and disinfected. Under a stereomicroscope the femur bone was localized by palpation of the thigh. A small skin incision was made with a surgical blade (no. 11) below the femur midway in the length of the external side of the thigh to localize the sciatic nerve (white) followed by a small muscle incision near the nerve. The nerve was gently isolated using two pairs of fine forceps (no. 7). At about the middle of the thigh, a ligature of the sciatic nerve was introduced using a silk suture (no. 6). The muscle was then closed with a silk suture (no. 6) followed by closure of the skin with a prolene suture (no. 7). This procedure was repeated for each pup. After the surgery, the pups were placed in cages with their mother. Pups were killed 4 h post-ligature. Each sciatic nerve was collected in a Petri dish and segmented under a stereomicroscope into equal lengths of 2 mm using a ruler and a surgical blade. Starting from the proximal side of the ligature, seven to nine equal segments (numbered as +9 to +1, with +1 being closest to the ligature) were collected. Similarly, seven to nine equal segments (numbered as −9 to −1, with −1 being closest to the ligature) were collected from the distal side of the sciatic nerve (see schematic representation in Fig. 5a). Each segment was frozen in liquid nitrogen and kept at −80°C until protein extraction was performed.
AChE enzymatic assay
The different tissues (brain, skeletal muscle, DRG, sciatic nerve and spinal cord) from wild-type and dt27J mice were homogenized on ice with a disposable Kontes pellet pestle and microtube homogenizer in 0.2–2.0 mL of a high-salt detergent buffer containing anti-proteolytic agents [10 mm Tris-HCl, pH 7.0; 10 mm EDTA; 1 m NaCl; 1% Triton X-100; 1 mg/mL bacitracin (Sigma, St Louis, MO, USA); and 25 U/mL aprotinin (Sigma)] (Jasmin and Gisiger 1990). Following centrifugation of the homogenates (20 000 × g for 15 min at 4°C), the supernatant was removed and stored immediately at −80°C.
AChE activity was measured using a modified version of the spectrophotometric method of Ellman et al. as described previously (Jasmin and Gisiger 1990; Boudreau-Larivière et al. 2000). Briefly, 25–50 µL aliquots of the protein extracts were incubated in 1 mL of a phosphate buffer solution (pH 7.0) containing 7.5 × 10−4m acetylthiocholine (Sigma) as the substrate and 5 × 10−4m dithiobis(nitrobenzoic acid) (Sigma). AChE activity was measured in the presence of the non-specific cholinesterase inhibitor tetraisopropylpyrophosphoramide. Non-specific hydrolysis was determined by measuring substrate hydrolysis in the presence of both tetraisopropylpyrophosphoramide and the AChE-specific inhibitor 1,5-bis(4-allydimethylammoniumphenyl)pentan-3-one dibromide (Sigma). The total amount of protein present in the extracts was determined by the bicinchoninic acid assay (Pierce Laboratories, Rockford, IL, USA). The specific activity of AChE was calculated by examining the rate of substrate hydrolysis and by correcting the values to take into account the amount of protein present in each sample.
Student's t-tests were performed to evaluate the difference in the specific activity of AChE in the different samples obtained from wild-type and dt27J mouse tissues. The level of significance was set at p < 0.05 using a two-tailed t-test (null hypothesis). For assessing axonal transport, we used a one-tailed t-test in accordance with the fact that we propose a directional hypothesis (see Introduction). Data are expressed as mean ± SE throughout.
dt27J mice display a neuropathy within their sciatic nerves
Histological observations revealed a severe loss of axons in the sciatic nerves of P15 dt27J mice (Fig. 1). In cross-section, sciatic nerves from the mutant mice were smaller than those from their normal littermates, they contained fewer axon profiles, and there was a dramatic reduction in myelinated fibers. The latter was accompanied by an abundance of apparently denervated Schwann cells (Fig. 1b). In addition, the occasional giant axon profile was observed. Of the myelinated fibers still remaining in the sciatic nerves of dt27J mice, a variation of the axon caliber was observed. To perform a quantitative assessment, myelinated axons were counted in the sciatic nerves of wild-type and dt27J mice. Our analysis revealed that the total number of surviving axons was 37% lower (p < 0.01) in the nerves from dt27J mice (Fig. 1c).
The cytoskeleton is disorganized in neurons of the sciatic nerve in dt mice
Although we have previously demonstrated that axonal cytoskeletal organization is perturbed in the DRG neurons of dt mice (Dalpéet al. 1998), a similar study has not been performed for the sciatic nerve. We have therefore fixed dt27J mice and their wild-type littermates, and processed the dissected sciatic nerves for electron microscopy. Ultrastructural analysis revealed normal myelinated axons as well as amyelinated small caliber axons within the normal sciatic nerves (data not shown). A longitudinal section of the wild-type nerves showed normal cytoskeleton organization with the neurofilaments and microtubules oriented along the long axis of the axon (Fig. 2a). In contrast, a similar analysis of the dt27J sciatic nerve revealed that the axons contained disorganized neurofilament and microtubule networks (Fig. 2b). The cytoskeletal filaments were not oriented along the long axis of the axon but rather were aligned in no particular direction. Occasionally, in cross-section, some segmental demyelination and axonal swellings were also observed (Fig. 2c) as described previously for the dorsal root axons (Dalpéet al. 1998). Cross-sections of sciatic nerves from P15 wild-type and dt27J mice were also examined for the presence of mitochondria. From 100 myelinated axons (diameter of 2 µm) and 100 non-myelinated axons (diameter 1 µm), we counted the number of mitochondria present. On average, there were 2.5 mitochondria per section of the myelinated axons and 0.5 mitochondria per section of the non-myelinated axons (data not shown). There was no statistically significant difference in the distribution between wild-type and dt samples. Thus, mitochondrial distribution is not affected in dt sciatic nerves, at least in normal appearing axons. However, once axonal swellings develop, mitochondria as well as other cytoplasmic components accumulate in these regions.
Acetylcholinesterase activity levels in various tissues of dt mice
As a prelude to using AChE as a reporter for fast axonal transport in the sciatic nerve, we examined the expression of AChE in several different tissues obtained from both wild-type and dt27J mice. Specifically, we determined AChE activity in the brain, spinal cord, DRG and skeletal muscle of P1 and P14 animals. As shown in Fig. 3, AChE specific activity was the highest in the brain and spinal cord. The specific activity in these tissues also increased (> fivefold) in both wild-type and dt27J mice between postnatal day 1 and 14. By comparison, the level of AChE specific activity was relatively lower in skeletal muscle and DRG (Fig. 3). In both wild-type and dt27J mice, AChE activity in skeletal muscle remained largely unaffected during the first 2 weeks of life. In DRG, however, AChE activity increased by approximately twofold between postnatal day 1 and 14 in wild-type animals whereas no such increase in AChE activity could be observed in DRG from dt27J mice (Fig. 3).
We next focused on the level of AChE expression in the sciatic nerve, which contains a mixture of axons originating from different sources, i.e. motor neurons, DRG neurons, and sympathetic neurons. As shown in Fig. 4, the level of AChE specific activity increased considerably in both wild-type and dt27J sciatic nerves from P1 to P14. However, it is also clear that there is a significant difference in AChE specific activity between the dt27J and wild-type sciatic nerves at P14. These findings are consistent with the observation that there is axonal degeneration in sciatic nerves during this time period in dt27J mice (see Fig. 1).
Axonal transport is perturbed in sciatic nerves of dt mice
Based on the above observations, we wished to determine whether axonal transport in P14 dt27J mice is affected. To this end, we chose to examine the orthograde and retrograde transport of AChE, used in this case as a marker of fast axonal transport. For these experiments, the sciatic nerves of both wild-type and dt27J mice were ligated for 4 h to bilaterally interrupt axonal transport. Following this time period, the sciatic nerves were excised, cut into 2 mm segments proximal and distal to the ligature, and the specific activity of AChE was determined in each segment thereby providing a pattern of AChE expression along the lengths of these sciatic nerves (see Fig. 5a).
In agreement with the data presented in Fig. 4, the baseline level of AChE specific activity in the proximal (14–6 mm proximal to the ligature) segments of sciatic nerves from wild-type mice was statistically higher than the baseline values seen in the samples obtained from dt27J mice (Figs 5b and c). In these experiments, the bilateral axonal transport of AChE appeared to proceed as expected in wild-type animals (see for example Lubinska and Niemierko 1971; Brimijoin and Wiermaa 1978; Jasmin et al. 1987). Indeed, we observed an accumulation of AChE activity in the first 2 mm segment immediately proximal and distal to the ligature in wild-type mice. In comparison to the appropriate baseline values, these accumulations of AChE represent the orthograde and retrograde axonal transport of AChE in the sciatic nerve of wild-type mice. In contrast to these findings, there was no detectable accumulation of AChE activity in the sciatic nerve segments immediately proximal and distal to the ligature in dt27J mice (Figs 5b and c). To quantitate these findings, we determined the ratios between the AChE specific activities of the segment adjacent to the ligature (+1) and the segment at the proximal end (+7) to arrive at a relative accumulation value. The average ratio is ∼3 in the wild-type samples and ∼1 in the dt27J samples. The difference between the wild-type and dt27J ratios is significant (p < 0.01). It is important to note that the axonal transport measurements are internally controlled, i.e. we are comparing the activities observed at the ligature versus those at the proximal and distal segments of the same sciatic nerves. These results indicate therefore that the axonal transport of AChE in both orthograde and retrograde directions is affected in the sciatic nerves of dt27J mice.
Mutations in dystonin/Bpag1, a cytoskeletal linker protein, lead to a multisystemic disorder affecting epithelial, neuronal, muscle and myelinating cell types (Brown et al. 1995; Guo et al. 1995; Bernier et al. 1998; Dalpéet al. 1999; Saulnier et al. 2002). Consistent with the predicted function of dystonin/Bpag1, loss of cytoskeletal organization within the various cell types leads to cellular dysfunction and ultimately to cellular degeneration (Guo et al. 1995; Yang et al. 1996; Bernier et al. 1998; Dalpéet al. 1998, 1999; Saulnier et al. 2002). dt mice harbor mutations in dystonin/Bpag1, and have historically been studied as neurological mutants (Duchen et al. 1963; Duchen and Strich 1964) since the predominant phenotype is a lack of limb motor coordination. The early pathological studies revealed argyrophilic axonal swellings in peripheral nerves even before the onset of clinically detectable symptoms (Duchen 1976). Furthermore, we have previously shown that axonal swellings are present in sensory nerve fibers of dt embryos as early as E15.5, before myelination and radial axonal growth have begun (Bernier and Kothary 1998). However, it is clear that during the acutely progressive phase of the dt disease, axonal swellings are more numerous (Duchen 1976). The presence of axonal swellings has led to the suggestion that an early feature in dt neurons is probably a disturbance in the normal flow of molecules in affected axons (Janota 1972). In the present study, we have demonstrated that fast axonal transport is indeed impaired in the sciatic nerves of dt27J mice.
Absolute measurements of AChE specific activities demonstrated that there was no difference between wild-type and dt27J mice in most of the tissues analyzed. For example, the levels of AChE activity in skeletal muscle of mutant mice were comparable to those seen in wild-type animals. Given that AChE is a sensitive marker of the state of innervation and differentiation of muscle, these findings indicate that denervation of skeletal muscle is not a feature of the dt pathology. This result is in fact consistent with our previous report showing that the neuropathy in dt mice does not result in muscle denervation, at least at 2 weeks of age (Dalpéet al. 1999; data not shown). Moreover, AChE specific activity was also found to be unchanged in brains from dt27J mice and is in agreement with previous studies showing a lack of pathology in this organ.
The two regions that showed a significant decrease in AChE specific activity were the DRGs and the sciatic nerves of dt27J mice at P14. Correspondingly, these two regions display neurodegeneration in these mice. Motor axon loss has not been observed in the ventral roots of dt mice, although similar studies have not been performed for sciatic nerves. Here, we have shown that the extent of axonal loss in sciatic nerves was significant at the P15 stage. The reduced axonal profiles resulted in smaller sciatic nerves in the mutant mice. The loss of large caliber axons in the sciatic nerves suggests that in addition to sensory neurons, motor neurons may also be affected in dt27J mice. Whether the effect on motor neurons is an intrinsic property of dystonin/Bpag1 loss of function or a secondary effect of sensory neurodegeneration remains to be established.
Our ultrastructural studies revealed that the neurofilament and microtubule network organization was perturbed in the axons of dt27J sciatic nerves, similar to what we had previously observed in the DRG neurons (Dalpéet al. 1998). Since loss of cytoskeletal organization is an early event in dt pathogenesis, and axonal swellings are a hallmark in this disease, we assessed whether axonal transport was impaired. Fast axonal transport in the sciatic nerve was assessed using the ligature method and AChE enzymatic activity was used as the quantitative marker for protein pooling at the ligature. Using AChE as the quantitative marker for axonal transport has the distinct advantage of allowing for the measurement of an endogenous component and is therefore more specific. This method can easily be used to measure changes in both orthograde and retrograde directions. Accordingly, transport of AChE has been widely used in the past to assess the functional state of axonal transport in several genetic or chemically induced disorders that exhibit peripheral nerve damage, including diabetic neuropathy and demyelinating diseases (see for example Oikarinen and Kalimo 1984; Marini et al. 1986; Macioce et al. 1988). Here we have shown that in sciatic nerves of dt27J mice, the process of fast axonal transport, as measured by following AChE transport in both orthograde and retrograde directions, is considerably lower than that in sciatic nerves of wild-type mice. This difference between wild-type and mutant animals likely reflects changes in both rate and amount of transported material that are caused by simultaneous morphological changes in the axons as well as by reductions in AChE levels in the DRGs. Furthermore, the changes we have observed reflect a cumulative effect on sensory and motor neurons within the sciatic nerve. At this point, it is difficult to discern differences in axonal transport between the two types of neurons. Whether or not motor neurons are affected in dt mice is an interesting question and is the subject of our future studies.
In summary, our results show altered AChE transport in the sciatic nerves of dt mice. Our studies raise an important question – is the impaired axonal transport a result of the neurodegeneration in dt mice, or is it causative? We favor the latter scenario based on our knowledge of the protein implicated in the primary defect, cytoskeletal disruption early in dt pathogenesis, and the observation of focal axonal swellings prior to neurodegeneration.
We would like to thank Ian Robb from the Children's Hospital of Eastern Ontario and Louise Pelletier from the University of Ottawa for their technical assistance in the processing of tissues for electron microscopy. This work was supported by operating grants from the Canadian Institutes of Health Research (CIHR) and the Ontario Neurotrauma Foundation (ONF). BJJ is a CIHR Investigator and JD-F was a recipient of a studentship from the ONF.