|Home | About | Journals | Submit | Contact Us | Français|
For many neurodegenerative disorders, such as Parkinson’s disease, there is evidence that the disease first affects axons and terminals of neurons that are selectively vulnerable. This would suggest that it may be possible to forestall progression by targeting the cellular mechanisms of axon degeneration. While it is now clear that these mechanisms are distinct from the pathways of programmed cell death, they are less well known. Compelling evidence of the distinctiveness of these mechanisms has derived from studies of the WldS mutation, which confers resistance to axon degeneration. Little is known about how this mutation affects degeneration in dopaminergic axons, those that are affected in Parkinson’s disease. We have characterized the WldS phenotype in these axons in four models of injury: two that utilize the neurotoxin 6-hydroxydopamine or axotomy to induce anterograde degeneration, and two that use these methods to induce retrograde degeneration. For both 6-hydroxydopamine and axotomy, WldS provides protection from anterograde, but not retrograde degeneration. This protection is observed as preserved immunostaining for tyrosine hydroxylase in axons and striatum, and by structural integrity visualized by GFP in tyrosine hydroxylase-GFP mice. Therefore, WldS offers axon protection, but it reveals fundamentally different processes underlying antero- and retrograde degeneration in this system.
Adult-onset neurodegenerative diseases such as Alzheimer’s disease and Parkinson’s disease (PD) are characterized neuropathologically by the loss of selectively vulnerable neuronal populations and their axonal projections. While the primary cellular locus of dysfunction is unknown for these diseases, many authorities have proposed that the process may begin in axons or their terminals (Hornykiewicz 1998; Selkoe 2002). In PD, this possibility is supported by postmortem studies that have demonstrated a greater loss of measures of striatal dopaminergic terminal projections (Scherman et al. 1989) than nigral neuron numbers, estimated to be about 30% at the time of disease onset (Fearnley and Lees 1991; Ma et al. 1997; Greffard et al. 2006). A greater relative loss of striatal dopaminergic terminal projections is also suggested by in vivo neuroimaging techniques that have demonstrated a loss of 50–60% of dopaminergic terminal markers at the time of onset (Lee et al. 2000).
The importance of the concept that axonopathy may be an early and predominant feature of these diseases is that there is now substantial evidence that the molecular and cellular mechanisms of axon degeneration are distinct from the canonical pathways of programmed cell death that mediate destruction of the neuron soma. Indeed, while many experimental therapeutic approaches based on inhibition of programmed cell death have been quite effective in preserving neuron numbers in models of PD, they have often failed to preserve axon projections (Eberhardt et al. 2000; Chen et al. 2008; Ries et al. 2008).
Many of the most striking experimental observations to support the concept that the mechanisms of axon degeneration are distinct from those of cell soma destruction derive from studies of the WldS mouse (Coleman and Perry 2002). The WldS mutation arose spontaneously in a colony of C57Bl/6 mice, and it was found to delay Wallerian degeneration of axons following peripheral nerve axotomy (Lunn et al. 1989). The WldS mutation has been shown to protect axons in models of both anterograde and retrograde degeneration, in diverse species, because of a variety of insults, including toxin exposure (Wang et al. 2001) and genetic abnormalities (Mi et al. 2005) (See Coleman 2005 and Luo and O’Leary 2005 for reviews). Thus, characterization of the molecular and cellular basis of the WldS phenotype promises to provide important insights into the mechanisms of axon degeneration.
Relatively little is known about the WldS phenotype in dopaminergic axons of the nigro-striatal projection, one of the major systems affected in PD. Sajadi and colleagues demonstrated that WldS provides remarkable axon protection in a model of anterograde degeneration induced by injection of the neurotoxin 6-hydroxydopamine (6OHDA) into the medial forebrain bundle (MFB) (Sajadi et al. 2004). However, WldS provided no protection in a model of retrograde degeneration of these axons induced by 6OHDA injection into the striatum. This result was unexpected because the MFB injection is generally considered to be more devastating. Sajadi et al. proposed that in the nigro-striatal projection, WldS may provide protection from anterograde, but not retrograde, degeneration. This explanation, however, seems difficult to reconcile with observations that WldS protects from retrograde degeneration in other models (Kerschensteiner et al. 2005; Mi et al. 2005). Additionally, in the nigro-striatal system, WldS was subsequently found to protect from another toxin, MPTP, in the striatal terminals (Hasbani and O’Malley 2006). Therefore, to better delineate the WldS phenotype in the nigro-striatal dopaminergic projection we have characterized its effect in four models of axon injury, two anterograde and two retrograde. In these investigations we monitor not only the expression of dopaminergic phenotype in axons by immunostaining for tyrosine hydroxylase (TH), but also their structural integrity by use of a novel optical technique that employs confocal optical dissection of the MFB in mice expressing green fluorescent protein (GFP) under the TH promoter (Sawamoto et al. 2001).
Adult (8-week) C57Bl/OlaHsd-Wlds (Wlds) male and female mice were obtained from Harlan-Olac (Bicester, Oxfordshire, UK). Wlds mice were crossed with TH-GFP mice (Sawamoto et al. 2001), which express green fluorescent protein driven by tyrosine hydroxylase promoter, to obtain the Wlds/wt : TH-GFP genotype. Both lines were maintained on a C57Bl/6 background. The Wlds genotype was determined as described (Mi et al. 2002).
For all experiments, adult male C57Bl/6 mice were obtained from Charles River Laboratories Inc., Wilmington, MA, USA. Four different methods of lesion of the nigro-striatal axonal projection were used (Fig. 1). The neurotoxin 6OHDA was used to induce either retrograde degeneration by intra-striatal injection, or anterograde degeneration by injection into the MFB. For both lesion models, adult mice were pre-treated with desipramine, anesthetized with ketamine/xylazine solution and placed in a stereotaxic frame. A solution of 6OHDA hydrobromide (Regis, Morton Grove, IL, USA) (5.0 μg/μL or 3.0 μg/μL in 0.9% NaCl/0.02% ascorbate) was injected by microliter syringe at a rate of 0.5 μg/min by pump for a dose of 15.0 μg/3.0 μL in intra-striatal model or a 6OHDA solution (3.0 μg/μL) at a rate of 0.25 μg/min by pump for a dose of 3.0 μg/1.0 μL in MFB. Injections were performed into the left striatum at coordinates AP: +0.09 cm; ML: +0.22 cm; DV: −0.25 cm or into the left MFB at the coordinates AP: −0.121 cm; ML: +0.105 cm; DV: −0.375 cm relative to bregma. After a wait of 2 min, the needle was slowly withdrawn.
Axotomy of the MFB was performed either distal to the substantia nigra (SN) to induce retrograde degeneration (‘Anterior Axotomy’), or proximal to the SN to induce anterograde degeneration (‘Posterior Axotomy’). Posterior Axotomy was performed essentially as previously described (El-Khodor and Burke 2002) by use of a retractable knife (Kopf Instruments, Tujunga, CA, USA) at coordinates AP: −0.10 cm; ML: +0.20 cm; DVdorsal: −0.25 cm; DVventral: −0.475 cm. Axotomy distal to the SN was performed in a similar fashion at AP: −0.030 cm, ML: +0.20 cm. All surgical procedures were approved by the Columbia University Animal Care and Use Committee.
For immunostaining of TH, mice were perfused intracardially with 0.9% NaCl followed by 4.0% paraformaldehyde in 0.1 M phosphate buffer (pH 7.1). The brain was post-fixed for 1 week, cryoprotected in 20% sucrose overnight, and then rapidly frozen by immersion in isopentane on dry ice. A set of serial horizontal sections was then cut at 30 μm. Sections containing MFB or striatum were processed free-floating. The primary antibody was rabbit anti-TH (Calbiochem, La Jolla, CA, USA) at 1 : 750. Sections were then treated with biotinylated protein A and avidin-biotinylated horseradish peroxidase complexes (ABC, Vector Labs, Burlingame, CA, USA). The optical density of striatal TH immunostaining was determined with an Image Research Analytical Image Station (St. Catherines, ON, Canada).
For Wlds immunofluorescent staining, the primary antibody was rabbit anti-Wlds (gift of M.P. Coleman) at 1 : 500. Sections then were treated with goat-anti-rabbit Texas Red conjugated secondary antibody.
All quantification of axons was performed in TH-GFP transgenic mice. Mice were killed and perfused intracardially with 0.9% NaCl followed by 4.0% paraformaldehyde in 0.1 M phosphate buffer (pH 7.1). Brains were post-fixed for 24 h and then cryoprotected in 20% sucrose/4%paraformaldehyde (PF)/phosphate buffer solution for 24–48 h. They were serially sectioned horizontally at 30 μm in a cryostat. A section containing the posterior third ventricular recess and the A13 dopamine cell group was selected for analysis (Fig. 2). Confocal microscopy (Leica TCS SP5 AOBS MP System, Wetzlar, 35578 Germany) was used to acquire images through the entire medial-to-lateral extent of the MFB, by use of a 20× objective with a zoom factor of 8× applied. In the retrograde axonal degeneration models, images of axons were acquired from medial to lateral starting from a rostro-caudal point adjacent to the posterior recess of the third ventricle, and just medial to the first appearance of axons of the MFB. Six contiguous fields (97 × 97 × 2 μm) were acquired (total tissue volume of 113 × 103 μm3). In the anterograde axonal degeneration models, images were acquired from a rostro-caudal point adjacent to the fornix and just medial to the first appearance of axons. Five contiguous fields (97 × 97 × 2 μm) were acquired (total tissue volume of 94 × 103 μm3). Each field was scanned in the Z-axis with twenty 0.1 μm thickness optical planes from dorsal to ventral in a 2.0 μm physical plane in the center of the section. These twenty optical planes were then merged to obtain a single maximal projection of the sampled volume. In order to count the number of axons passing in the rostro-caudal dimension through each sample volume, two horizontal lines were drawn on the image at a separation distance of 10 μm in the center of the maximal projection. Every intact axon crossing both lines was counted as positive.
We have previously shown in adult rats that following intranigral injection of 6OHDA, degeneration of intra-striatal dopaminergic axons is first demonstrated at post-lesion day (PLD) 1 by the suppressed silver stain technique, and that it continues through PLD7 (Jeon et al. 1995). Following injection of 6OHDA into the MFB (adjacent to the SN) in mice, a similar time course is observed for the loss of TH immunostaining of the striatum (Fig. 3). By PLD2 a loss of staining is evident, and progressive depletion occurs through PLD9, the last time point examined. By PLD9, the optical density of striatal TH staining on the lesioned side has been reduced to about 20% of the contralateral control side (Fig. 3). As previously reported by Sajadi et al. for this model (Sajadi et al. 2004), the WldS mutation provided striking preservation of TH-positive dopaminergic innervation of the striatum. The effect was most evident at PLD2, when there was only a 10% loss of striatal TH optical density in homozygous WldS mice, as compared to a 63% loss in wildtype controls (Fig. 3). This protective effect remained significant through PLD6, and at PLD9 it was still apparent, but only as a trend.
We have shown that following injury, dopaminergic axons lose expression of the TH phenotype while remaining structurally intact (Cheng et al. 2008). Therefore, following injury it is possible that an apparent protective effect is because of preservation of phenotype in axons that have remained structurally intact. To address this issue, we have examined the ability of WldS to protect nigro-striatal axons in TH-GFP mice (Sawamoto et al. 2001). We have shown that in these mice, following axon injury, although dopaminergic axons lose expression of TH, their number and structural integrity in the MFB can be monitored by fluorescence microscopy (Cheng et al. 2008). For these experiments WldS/WldS mice were crossed with mono-transgenic TH-GFP mice to obtain the WldS/Wt : TH-GFP genotype (hereafter designated as WldS : TH-GFP). Using these mice, we confirmed that WldS protein is expressed in dopamine neurons of the substantia nigra pars compacta (SNpc); double-label immunofluorescence revealed that the large majority of GFP-positive, dopaminergic neurons of the SNpc also expressed WldS protein (Fig. 4a). As previously reported for WldS, the most prominent labeling was in the nucleus (Mack et al. 2001), but cytoplasmic straining was also observed as reported more recently (Beirowski et al. 2009).
Injection of 6OHDA into the MFB of TH-GFP mice not carrying the WldS gene resulted in a 62% loss of MFB axons at PLD3 (Fig. 4c). At this time point, characteristic features of acute axon degeneration, included fragmentation and spheroid formation (El-Khodor and Burke 2002) were also observed (Fig. 4b). A substantial degree of axon protection was observed in WldS : TH-GFP mice. At PLD3, these mice showed only a 32% loss of axons, a highly significant difference (p = 0.002) in comparison to the TH-GFP group. Minimal axon pathology was observed (Fig. 4b). We therefore conclude that in this neurotoxin model of anterograde degeneration within the nigro-striatal system, the WldS mutation provides protection not only from loss of TH expression phenotype, but also from structural axonal degradation.
Following intra-striatal injection of 6OHDA in mice, suppressed silver staining reveals degeneration of striatal neurites at PLD1–2, anterior MFB axons at PLD3–4, and posterior MFB (proximal to the SNpc) at PLD5–6 (Ries et al. 2008). In order to optimize the likelihood of observing a protective effect of WldS in this model, we therefore performed morphologic analysis at PLD3. At this post-lesion time, we observed no protection of striatal TH-positive immunostaining in WldS in comparison to wildtype mice (Fig. 5a). We investigated the possibility that although WldS provided no apparent protection of expression in the TH phenotype within the striatum, it may nevertheless provide preservation of the structural integrity of dopaminergic axons. However, examination of the MFB in these sections revealed as much axonal spheroid and fragmentation pathology in the WldS homozygous mice as in wildtype (Fig. 5b). The inability of WldS to preserve axon structural integrity in this model was confirmed in WldS : TH-GFP mice. These mice showed a comparable degree of axon loss, and axonal spheroid pathology, as TH-GFP mice not carrying the WldS allele (Fig. 5b).
The MFB and intra-striatal lesion models differ in other respects besides the direction of induced axon degeneration. They differ in the local tissue concentration of toxin to which dopaminergic axons and terminals are exposed, and in the time course of induced degeneration (Sauer and Oertel 1994; Jeon et al. 1995). Therefore, to further investigate whether there is a fundamental difference between antero- and retrograde axon degeneration in the nigro-striatal projection in their regulation by WldS, we examined its effect on these two types of axon degeneration when induced by the same injury, that of knife-cut axotomy. In doing these experiments, we also sought to determine whether the ability of WldS to protect in a model of anterograde degeneration was a general phenotype, rather than particular to neurotoxin injury.
Following induction of anterograde degeneration in the MFB by a posterior axotomy, we observed preservation of striatal TH immunostaining in both homozygous and heterozygous WldS mice (Fig. 6a). This protective effect was also observed in these sections at the level of the MFB. At 3 days following posterior axotomy, a loss of TH-positive axons, and an abundance of axon spheroid pathology was observed in the MFB of wildtype mice, whereas in WldS mice, axons were relatively preserved (Fig. 6b).
The ability of WldS to preserve the structural integrity of MFB axons following posterior axotomy was confirmed by studies in WldS : TH-GFP mice. These mice showed only a 29% loss of GFP-positive MFB axons, whereas TH-GFP control mice showed a 69% loss, a highly signification difference (p = 0.001) (Fig. 6c).
In contrast to this ability of WldS to provide protection of both axon integrity and expression of the TH phenotype in the posterior MFB axotomy model, WldS failed to provide protection in the anterior MFB axotomy model of retrograde degeneration (Fig. 7).
Adult-onset neurodegenerative diseases of the central nervous system cause the loss of selectively vulnerable neurons and their axons, and these losses underlie their diverse and devastating clinical manifestations. In many prior efforts to understand the pathogenesis of these diseases, and to develop therapies that forestall their progression, it has apparently been tacitly assumed that the loss of axons is a manifestation of the neuron loss. Such an assumption would underlie the prevailing concept that prevention of neuron death should be the primary goal in the development of neuroprotective therapeutics. However, there is now abundant evidence that the cellular processes that mediate axon destruction are separate and distinct from the diverse pathways of programmed cell death that mediate destruction of the neuron soma (Raff et al. 2002; Coleman 2005). In addition, for PD, there is now broad evidence that the disease process may begin in the axons, because it is most prominently manifested throughout its course by axon and terminal loss (Cheng et al. 2009). Thus, neuroprotective approaches aimed exclusively at prevention of neuron loss are unlikely to forestall clinical progression of disease.
Since the first demonstrations of the molecular basis of programmed cell death in the nematode C. elegans (Ellis et al. 1991), there has been an extraordinary growth in our knowledge of the diverse pathways of programmed cell death (Hotchkiss et al. 2009), whereas our knowledge of the mechanisms of axon degeneration have lagged. Nevertheless, based on the information we do have, it is already clear that diverse mechanisms of axon destruction exist. Some of the clearest evidence that mechanisms of axon degeneration are distinct from those of programmed cell death, and that they are diverse, derive from studies of WldS. A particularly clear example of this diversity was the demonstration by Hoopfer and colleagues that although WldS slows Wallerian degeneration in Drosophila and mice following injury, it does not influence developmental axon pruning in either species (Hoopfer et al. 2006).
Given that the ability of WldS to provide protection, or not, may define distinct mechanisms of axon degeneration, the observation of Sajadi and colleagues that WldS provides protection in an anterograde, but not a retrograde, model of mesencephalic dopaminergic axon degeneration was of great interest (Sajadi et al. 2004). However, these observations were made exclusively by use of immunohistochemical techniques to identify protein markers of cell phenotype. We have previously shown that, following injury, axons of the MFB quickly lose expression of phenotype while remaining structurally intact (Cheng et al. 2008). We have therefore used a technique of confocal optical dissection of the MFB in TH-GFP mice to visualize and quantify dopaminergic axons following injury. By use of this technique, in conjunction with immunostaining for TH, we confirm the observations of Sajadi and colleagues, that the WldS mutation protects dopaminergic axons following injection of 6OHDA into the MFB, with ensuing anterograde degeneration, but not following injection into the striatum, with ensuing retrograde degeneration. The protection afforded by WldS in the anterograde model is demonstrated by relative preservation of striatal TH immunostain optical density, by preservation of axon number in TH-GFP mice, and by diminished axonal spheroid pathology in the MFB. The ability of WldS to protect axons following direct injection of 6OHDA into the MFB, but not the striatum, is especially remarkable considering that the MFB lesion is generally considered to be more destructive: it results in a sudden onset and abrupt destruction of adjacent neurons of the SNpc (Jeon et al. 1995) whereas the striatal lesion induces a delayed-onset (Ries et al. 2008) and gradual loss (Sauer and Oertel 1994).
The 6OHDA MFB and intra-striatal models of dopaminergic axon degeneration differ in other respects besides the direction of the ensuing axon degeneration. In the MFB model, the dopaminergic axons are subject to a direct exposure to high concentrations of the toxin injected in their vicinity, whereas in the intra-striatal model, many dopaminergic elements are at a distance of several millimeters from the injection site, and are therefore exposed to lower concentrations. Therefore, while the intra-striatal model would be induced mainly by a neurotoxin effect, the MFB model could be due, at least in part, to a local, physically destructive axotomy effect. In addition, as mentioned, the MFB model induces axon degeneration with a different time course than the striatal model. Therefore, to compare the ability of WldS to protect from antero- and retrograde degeneration in models that eliminate these differences, we examined its effects in models of posterior and anterior MFB axotomy. These models are comparable in both the nature and the timing of axon injury. In these models as well, we observe an ability of WldS to protect from anterograde (Wallerian) degeneration, but not retrograde degeneration. The difference in the ability of WldS to protect is observed by relative preservation of striatal TH immunostain optical density, by preservation of axon number in TH-GFP mice, and by diminished axonal spheroid pathology. We therefore conclude that, as suggested by the observations of Sajadi et al. (Sajadi et al. 2004), that there is a definitive, fundamental difference between mechanisms of anterograde and retrograde degeneration in the dopaminergic axons of the MFB: the WldS mutation affords protection in the former, but not the latter.
This difference in response to the mutant WldS protein therefore identifies for the first time in the adult central nervous system a difference at the cellular level between the mechanisms of anterograde and retrograde degeneration in an axonal projection. So little is known about the mechanisms of axon degeneration in any context that it is difficult to speculate about what the basis for this difference might be. Emerging evidence suggests that WldS does not act at the level of the nucleus to mediate protection, as originally believed, but rather at the level of the axon itself. Thus WldS may influence local axonal degenerative processes that function independently of the cell soma and nucleus, as must occur in Wallerian degeneration following axotomy. However, in retrograde degeneration, where the axon remains in continuity with the cell soma, different mechanisms may play a role. We have recently found that the kinase Akt blocks this form of axon degeneration via mTor signaling and inhibition of macroautophagy (Cheng et al. 2008; and submitted). It is possible that in this context, the cell body utilizes macroautophagy in its known role to ‘recycle’ cellular contents, to attempt to do so for axonal cellular material. At the present time, there is no evidence that the WldS protein would regulate such a process.
The existence of different mechanisms underlying anterograde and retrograde axon degeneration in the nigro-striatal projection may be important for understanding the pathogenesis of PD. Some authorities have suggested that the disease begins in the striatal dopaminergic terminals, and then proceeds as a ‘dying back’ axonopathy (Hornykiewicz 1998). This possibility is supported by observations made on postmortem PD brains and by in vivo neuroimaging techniques. Overall, these data suggest that, at the time of disease onset, while there has been about a 30% loss of SN neurons, there has been a 50–60% loss of striatal dopaminergic markers (reviewed in Cheng et al. 2009). However, while these observations are compatible with the possibility of a ‘dying back’ process, they are also not incompatible with an anterograde degenerative process. While the available evidence suggests that in PD axonal injury is out of proportion to neuron loss, it does not elucidate whether the axonal pathology has evolved in the retrograde or anterograde direction. Based on our observations indicating that these processes have fundamentally different cellular mechanisms, it will be important, in efforts to understand and treat the disease, that the nature of the axon degenerative process be defined. In this regard, it will be important to determine whether the axonopathy recently described in a novel bacterial artificial chromosome (BAC) transgenic model based on the hLRRK2 (R1441G) mutation proceeds in retrograde or anterograde fashion (Li et al. 2009). It will also be of interest to determine whether or not the WldS mutation protects in this model.
In conclusion, identification of the mechanism of WldS to forestall axon degeneration will have broad and important implications for understanding the pathogenesis and treatment of PD and other neurodegenerative disorders.
This work was supported by NS26836, NS38370, the RJG Foundation, and the Parkinson’s Disease Foundation. We are grateful to Dr. M.P. Coleman for his gift of anti-Wlds antibody, and to Drs. K. Kobayashi and H. Okano for the use of their TH-GFP mice.