• 6-hydroxydopamine;
  • axonopathy;
  • kinase;
  • mitogen-activated protein kinase;
  • Parkinson’s disease;
  • programmed cell death


  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Activation of c-jun N-terminal kinase (JNK) by the mitogen-activated protein kinase cascade has been shown to play an important role in the death of dopamine neurons of the substantia nigra, one of the principal neuronal populations affected in Parkinson’s disease. However, it has remained unknown whether the JNK2 and JNK3 isoforms, either singly or in combination, are essential for apoptotic death, and, if so, the mechanisms involved. In addition, it has been unclear whether they play a role in axonal degeneration of these neurons in disease models. To address these issues we have examined the effect of single and double jnk2 and jnk3 null mutations on apoptosis in a highly destructive neurotoxin model, that induced by intrastriatal 6-hydroxydopamine. We find that homozygous jnk2/3 double null mutations result in a complete abrogation of apoptosis and a prolonged survival of the entire population of dopamine neurons. In spite of this complete protection at the cell soma level, there was no protection of axons. These studies provide a striking demonstration of the distinctiveness of the mechanisms that mediate cell soma and axon degeneration, and they illustrate the need to identify and target pathways of axon degeneration in the development of neuroprotective therapeutics.

Abbreviations used



C/EBP homologous protein


endoplasmic reticulum


c-jun N-terminal kinase


mitogen-activated protein kinase


medial forebrain bundle


phosphate buffer


programmed cell death


Parkinson’s disease




post-lesion day


substantia nigra


SN pars compacta


tyrosine hydroxylase

Accumulating knowledge of the molecular mechanisms of programmed cell death (PCD) offers the promise that it may be possible to therapeutically target specific mediators, and forestall progressive neuron death, the defining feature of adult-onset neurodegenerative diseases such as Parkinson’s disease (PD). Among the multitude of PCD pathways identified in the death of neurons, the mitogen-activated protein kinase (MAPK) signaling cascade has been demonstrated to be important by an abundance of experimental evidence. Many of these observations suggest that MAPK signaling plays a role in the death of substantia nigra (SN) dopamine neurons, one of the principal populations to degenerate in PD (Silva et al. 2005a).

However, little is known about the mechanisms by which the c-jun N-terminal kinases (JNKs), critical mediators of MAPK signaling, mediate the death of these neurons. Homozygous null mutations of the jnk2 and jnk3 genes provide substantial protection of SN dopamine neurons in an acute MPTP neurotoxin model of PD (Hunot et al. 2004), but how this protection is achieved, and its generality, are unknown. How jnk2 and jnk3 deletions may affect the diverse, canonical pathways of PCD cannot be delineated in this model, because neither morphologic apoptosis (Jackson-Lewis et al. 1995), nor biochemical evidence of PCD, such as caspase activation, has been identified within dopamine neurons. Indeed, Furuya et al. have shown that a dominant negative inhibitor of Apaf-1 is not neuroprotective in the acute MPTP model, indicating that the intrinsic pathway of PCD is unlikely to play a role (Furuya et al. 2004).

Another unresolved issue related to the role of the JNKs in neurodegeneration, and specifically in models of PD, is the cellular compartment in which they mediate their effects, and, specifically, whether they participate in axonal degeneration. There is now substantial evidence that the molecular pathways underlying destruction of the cell soma are separate and distinct from those which mediate destruction of axons (Raff et al. 2002; Coleman 2005). A number of prior investigations of MAPK signaling have suggested that it may play a role in axonal degeneration in PD models (Crocker et al. 2001; Ganguly et al. 2004), but others have not (Chen et al. 2008).

To address these issues related to the mechanisms, generality, and cellular location of the participation of the JNKs in degeneration of dopamine neurons, we have examined the effects of genetic deletion of the jnk2 and jnk3 genes on neurotoxicity because of 6-hydroxydopamine (6OHDA). We have utilized an intrastriatal injection model (Sauer and Oertel 1994), because definitive ultrastructural and light microscope evidence of apoptosis has been identified within phenotypically defined SN dopamine neurons (Marti et al. 2002). We did not evaluate the role of the JNK1 isoform because Hunot et al. had previously shown that the JNK2 and JNK3 isoforms were the principal ones involved in dopamine neuron death (Hunot et al. 2004). Furthermore, Coffey and colleagues had previously shown that neuron cell stress specifically activates JNK2 and 3 (Coffey et al. 2002).


  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Experimental animals

The generation of jnk2 and jnk3 null mice has been previously reported (Yang et al. 1997, 1998a), and methods for determination of genotype have been described (Kuan et al. 1999). All mice used in this study had been backcrossed to the C57BL/6 background.

6-hydroxydopamine lesion

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 in 0.9% NaCl/0.02% ascorbate) was injected by microliter syringe at a rate of 0.5 μL/min by pump for a total dose of 15.0 μg/3 μL. Injection was performed into the left striatum at coordinates AP: +0.09 cm; ML: +0.22 cm; DV: −0.25 cm relative to bregma. After a wait of 2 min, the needle was slowly withdrawn. The 6OHDA model in postnatal mice was performed as previously described (Silva et al. 2005b). Briefly, mouse pups at postnatal day 6–7 were pre-treated with 25 mg/kg desipramine, anesthetized by hypothermia, and placed prone on an ice pack. 6OHDA was prepared at 20 μg (total weight)/μL in 0.9% NaCl/0.02% ascorbic acid, and infused by pump (Harvard Apparatus, Holliston, MA, USA) at a rate of 0.25 μL/min for 2 min, for a total dose of 10 μg. For experiments in postnatal mice, littermate wildtype and heterozygote animals were examined in comparison to nulls. All injection procedures, as described, were approved by the Columbia University Animal Care and Use Committee.


For immunostaining of tyrosine hydroxylase (TH) mice were perfused intracardially with 0.9% NaCl followed by 4.0% paraformaldehyde in 0.1 M phosphate buffer (PB) (pH 7.1). The brain was carefully removed and blocked into midbrain and forebrain regions. The region containing the midbrain was post-fixed for one week, cryoprotected in 20% sucrose overnight, and then rapidly frozen by immersion in isopentane on dry ice. A complete set of serial sections was then cut through the SN at 30 μm. Beginning with a random section between 1 and 4, every 4th section was processed, in keeping with the fractionator method of sampling (Coggeshall and Lekan 1996). Sections 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). Following immunoperoxidase staining, sections were thionin counter-stained. The forebrain region containing the striatum was post-fixed for 48 h, then frozen without cryoprotection, and processed as described (Kholodilov et al. 2004). The optical density of striatal TH immunostaining was determined with an Imaging Research Analytical Imaging Station (St. Catherines, Ontario, Canada).

To perform immunohistochemistry for phosphorylated c-jun, mice were anesthetized with ketamine/xylazine solution and hypothermia, and provided with supplemental oxygen at the rate of 4 L/min for 6 min. Mice were then perfused immediately with 4% paraformaldehyde in 0.1 M PB containing 0.2 mM Na-orthovanadate at a rate of 5 mL/min for 10 min. The brains were removed, post-fixed for 24 h, and then cryoprotected with 20% sucrose in 0.1 M PB overnight at 4°C. SN sections were cut and processed according to the methods described above. Primary antibody was rabbit polyclonal anti-phospho-c-jun (Ser73) (Cell Signaling, Beverly, MA, USA) at a dilution of 1 : 50. For the determination of the number of phosph-c-jun-positive profiles, n = 4 mice were studied for each genotype.

Immunostaining for C/EBP homologous protein (CHOP) was performed as previously described (Silva et al. 2005b) with rabbit anti-CHOP antibody (kindly provided by D. Ron) at 1 : 500. For the determination of the number of CHOP-positive profiles, n = 4 mice were studied for each genotype.

Immunostaining for caspase cleaved β-actin was performed with the fractin (‘fragment of actin’) antibody at a dilution of 1 : 100 (Yang et al. 1998b) (BD PharMingen, San Jose, CA, USA) as previously described (El-Khodor and Burke 2002). Immunostaining for a spectrin cleavage product of calpain was performed with Ab38 (kindly provided by R. Siman) (Roberts-Lewis et al. 1994) at a dilution of 1 : 1000.

Determination of SN dopamine neuron numbers by stereologic analysis

Stereologic analysis was performed under blinded conditions on coded slides. For each animal, the SN on both sides of the brain was analyzed. For each section the entire SN was identified as the region of interest. Using StereoInvestigator software (MicroBright Field, Inc, Williston, VT, USA) a fractionator probe was established for each section. The number of TH-positive neurons in each counting frame was then determined by the optical disector method. Our criterion for counting an individual TH-positive neuron was the presence of its nucleus either within the counting frame, or touching the right or top frame lines (green), but not touching the left or bottom lines (red). For stereologic analysis, n = 6–9 mice for each genotype were used.

Determination of the number of apoptotic profiles within the SN

Apoptotic profiles were identified in the SN on TH-immunoperoxidase stained and thionin counter-stained sections as previously described (Oo et al. 2003). For the determination of the number of apoptotic profiles, n = 5–7 mice for each genotype were used.

Suppressed silver stain

Mice were deeply anesthetized and perfused through the left ventricle with 0.9% saline at 4°C for 1 min, followed by 4% paraformaldehyde (PF) in 0.1 M PB, pH 7.1, for 20 min at 4°C. The brain was removed carefully and post-fixed in 4% PF/0.1 M PB for at least 1 week at 4°C. Each brain was cryoprotected with 20% w/w sucrose in 4% PF/0.1 M PB for 48 h at 4°C and then frozen rapidly in 2-methylbutane on dry ice and sectioned horizontally at 30 μm. The silver staining protocol was based on the procedure of Gallyas et al. (Gallyas et al. 1980). Sections were maintained in serial order and processed freely floating in custom-made plastic grids with nylon mesh bottoms. Sections were collected into cold fixative, washed three times in distilled water, and then immersed in pre-treating solution (equal volumes 9% NaOH and 1.2% NH4NO3) twice for 5 min each. They were then immersed in impregnating solution (60 mL of 9% NaOH, 40 mL of 16% NH4NO3, 0.5 mL of 50% AgNO3) for 10 min. Sections were then washed three times in washing solution (1 mL of 1.2% NH4NO3 added to 100 mL of a solution containing 5 g anhydrous Na2CO3, 300 mL 95% ethanol, brought to 1 L with distilled water), followed by immersion in developing solution (1 mL of 1.2% NH4NO3 and 100 mL of a solution consisting of 0.5 g citric acid in 15 ml of 37% formalin, 100 mL 95% ethanol, 700 mL water brought to pH 5.8–6.1 with 9% NaOH, and finally brought to 1 L with water). Sections were kept in developing solution for 1 min. Sections were then mounted on subbed slides, air dried, and immersed in 0.5% acetic acid three times for 10 min each. Sections were then dehydrated through alcohols, cleared in xylene, and coverslipped under Permount.

Northern and Western analysis of JNK1 and JNK2 expression

Probes for hybridization were designed based upon published mouse sequences. For JNK1 (Acc# NM_016700): forward primer- TCCCGGACAAGCAGTTAGAT; reverse- TTGATCATTGCTGCACCTGT. For JNK2 (Acc# AF052469): forward- TGGACTGGGAAGAAAGAAGC, reverse- CGCCGTCTCCAACATACATA. The sequence of each probe was confirmed. 32P-labeled anti-sense riboprobe was prepared using SP6 RNA polymerase or T7 RNA polymerase (Promega, Madison, WI, USA). Membranes were hybridized overnight in ULTRAhyb Buffer (Ambion, Austin, TX, USA) with a standard probe concentration and specific activity, and then washed. The membrane was dried and exposed to phosphorimager cassettes, scanned and analyzed by Image Quant software (Molecular Dynamics, Indianapolis, IN, USA).

For Western analysis SN tissue was microdissected from postnatal day 6 ventral mesencephalon. The protein concentration was determined using a micro BCA kit (Pierce, Rockford, IL, USA). Aliquots of protein were electrophoresed in a sodium dodecyl sulfate-polyacrylamide gel and transferred onto a Hybond-P membrane (Amersham Biosciences, Arlington Heights, IL, USA). The membrane was then probed with either anti-JNK1 (Pharmingen #554286) or anti-JNK1/2 (Pharmingen #554285) primary antibodies, treated with appropriate secondary antibodies, conjugated with horseradish peroxidase, and detected with a chemiluminescent substrate (Pierce). Densitometric analysis of band intensity was performed using a FluorChem 8800 Imaging System (Alpha Innotech, San Leandro, CA, USA).


  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Combined homozygous jnk2/jnk3 null mutations abrogate apoptosis in SN dopamine neurons

Intrastriatal injection of 6OHDA in adult wildtype mice induced the gradual and increasing appearance of apoptosis in dopamine neurons of the SN (Fig. 1a), with the appearance of a maximal effect at post-lesion day (PLD) 6, as previously reported in the adult rat (Marti et al. 2002). The absence of the JNK2 isoform resulted in a striking 95% reduction in the number of apoptotic profiles in the SN pars compacta (SNpc) at PLD6 (Fig. 1a). Similarly, absence of the JNK3 isoform resulted in a marked and somewhat greater (98%) reduction, although this small difference was not significant (Fig. 1a). Double null mice lacking both the JNK2 and -3 isoforms showed a complete absence of apoptosis among dopamine neurons of the SNpc (Fig. 1a). A total of 42 SN sections (six from each of seven animals), representing posterior, central and anterior planes of the SN were scanned at 600× throughout the entire dorsal-ventral and medial-lateral extent of TH immunostained SNpc, and not a single apoptotic profile was identified.


Figure 1.  Inhibition of apoptosis in homozygous jnk2/jnk3 double null mice. (a) Adult mice of different genotypes were lesioned by intrastriatal injection of 6OHDA and killed after 6 days for the counting of apoptotic profiles in the SNpc. A typical example of an apoptotic profile, with characteristic chromatin clumps, is demonstrated by thionin counter-stain in the inset. The homozygous jnk2 and jnk3 single null mutations suppressed apoptosis by 95% and 98%, respectively, and the homozygous jnk2/3 double null mutations completely abrogated apoptosis [***p < 0.001, anova, Wildtype (WT) in comparison to each of the jnk null genotypes; n = 7 each group, with the exception of jnk3 single null, n = 5]. (b) The homozygous jnk2 and jnk3 single null mutations did not result in an increased survival of SN dopamine neurons, based on stereologic counts of TH-immunostained neurons at 28 days post-lesion. The decreased number of TH-positive neurons in the SN ipsilateral to the 6OHDA injection in comparison to the contralateral, non-injected Control (CON) was not different between Wildtype (WT) mice and either jnk2 or jnk3 single null animals (n = 6, both genotypes in the JNK2 study; n = 6 Wildtype, n = 9 jnk3 null in the JNK3 study). (c) The homozygous jnk2/3 double null mutations provided almost complete protection of SN dopamine neurons, based on stereologic counts at 28 days post-lesion (n = 7, both genotypes). Among Wildtype (WT) mice, there was a 63% loss of dopamine neurons, typical for this model, whereas among jnk2/3 nulls, there was only a 4% decrease, not a significant difference in comparison to the contralateral, non-injected Control (CON) side. The adjacent panel shows low power photomicrographs of representative TH-immunostained and thionin-counter-stained anterior SN sections from Wildtype (top) and jnk2/3 nulls (bottom) at 28 days following unilateral 6OHDA injection. (d) Although there is an almost complete preservation of SN TH-positive neurons in the jnk2/3 null mice, they demonstrate a 20% loss of mean neuronal cross-sectional area [p < 0.001, n = 100 neurons contralateral Control (CON) and ipsilateral 6OHDA Experimental (EXP) sides].

Download figure to PowerPoint

We have previously observed that genetic disruption of important cell death mediators may diminish apoptosis acutely in neurotoxin-induced Parkinson models, but not necessarily provide lasting protection of SN dopamine neurons, presumably because of the compensatory utilization of alternative cell death pathways (Silva et al. 2005b). Therefore, we examined the effect of single and combined deletion of the JNK2 and -3 isoforms on the surviving numbers of dopamine neurons at four weeks post-lesion. In addition, at this late post-lesion timepoint, the acute suppression of the TH phenotype following 6OHDA injection has resolved, such that immunolabeling of TH becomes a reliable indicator of the number of surviving neurons (Sauer and Oertel 1994). In spite of the marked suppression of apoptosis by elimination of the individual JNK2 and -3 isoforms, neither null genotype demonstrated an increased surviving number of SN dopamine neurons at 4 weeks post-lesion (Fig. 1b). In marked contrast, homozygous double null mice lacking both JNK2 and -3 isoforms demonstrated a virtually complete preservation of SN dopamine neurons (Fig. 1c). A minimally decreased number (4%) of TH-positive neurons on the side of the 6OHDA injection was not significantly different from the contralateral, non-injected control side. In spite of this robust protection because of the double null mutation, there was a persistent atrophy of dopamine neurons following toxin injection, identified as a 20% decrease in their cross-sectional area defined by TH immunostaining, at 4 weeks post-lesion (Fig. 1d). Based on this analysis, the atrophy appears to affect the entire population of TH-positive neurons, because the distribution of their sizes did not become bimodal; rather, it remained unimodal, as shown in Fig. 1(d), and shifted to the left. Thus, the atrophy component of the neurotoxin injury phenotype does not require the JNK2 or -3 isoforms.

In order to determine the general nature of a role for the JNK2 and -3 isoforms in mediating apoptosis in SN dopamine neurons, we also examined the effects of homozygous single null mutations on the magnitude of apoptosis in a postnatal model in which 6OHDA-mediated destruction of striatal dopamine terminals results in apoptosis because of the loss of developmental target-derived support, i.e., an ‘axotomy’ effect (Marti et al. 1997; Silva et al. 2005b). In this context both homozygous single jnk2 and jnk3 null mutant mice demonstrated suppression of apoptosis, by 40% and 38%, respectively (data not shown). Thus, JNK2 and JNK3 play a contributory role in this form of apoptosis as well as in that because of neurotoxicity in adults.

Two principal interpretations of the finding that neither jnk2 nor jnk3 homozygous single null mutations provide a lasting protection of dopamine neurons, and yet the jnk2/3 double null mutations do so, are that loss of one isoform is compensated for by up-regulation of the other, or, alternatively, that even in the absence of a compensatory change, the remaining isoform subsumes all of the death-mediating function of the missing isoform. To distinguish between these two possibilities, we examined the expression of JNK1 and JNK2 mRNA and protein in jnk3 homozygous null mice. These studies revealed that neither steady-state mRNA nor protein levels of JNK1 and -2 isoforms were affected by the jnk3 null mutation (Fig. 2). We have previously shown, similarly, that the jnk2 null mutation does not result in increased expression of JNK1 or -3 protein (Kuan et al. 2003). We therefore conclude that the persistence of dopamine neuron death in the jnk3 and jnk2 null mice is not because of a compensatory up-regulation of JNK2 or JNK3 respectively.


Figure 2.  The homozygous jnk3 null mutation is not associated with a compensatory increase in JNK1 or JNK2 mRNA or protein expression in the SN. (a) Northern analysis of JNK1 and JNK2 mRNA derived from SN of postnatal 6 homozygous jnk3 null mice demonstrates that there is no change in the level of expression of mRNA for either JNK isoform in the presence of the null mutation. In the top two panels, representative blots are shown; in the top panel the blot has been probed for JNK3 mRNA to confirm genotype; in the middle panel blots have been probed for JNK1 and JNK2. The absence of any alteration in levels of expression of either isoform is shown quantitatively by phosphorimager counts over the principal bands (n = 4 animals for each determination). (b) A representative western blot reveals no alteration in the level of JNK1 or JNK2 protein isoforms in the jnk3 null mice.

Download figure to PowerPoint

Combined homozygous jnk2/jnk3 null mutations suppress phosphorylation of c-jun

The ability of the JNKs to induce cell death is not exclusively mediated through c-jun phosphorylation; other death-mediating JNK targets have been identified, and diminished neuron death in jnk3 null mice has been observed even in the absence of down-regulation of c-jun phosphorylation (Keramaris et al. 2005). We therefore examined the effect of combined jnk2/jnk3 null mutations on c-jun phosphorylation status in the adult 6OHDA model. We found that the homozygous jnk2/jnk3 double null mutation results in an almost complete (96%) suppression of c-jun phosphorylation (Fig. 3a). To further examine the relationship between c-jun phosphorylation and cell death, we asked the converse, whether phosphorylation still occurs in jnk3 single null mice, in which longterm survival is not achieved. We find that phosphorylation does in fact persist in these animals (Fig. 3b). Therefore, in this model, we do not find evidence for an important JNK2/3-mediated death pathway that acts independently of c-jun phosphorylation.


Figure 3.  Decreased phosphorylation of c-jun in homozygous jnk2/3 double null mice. (a) Mice received intrastriatal 6OHDA and were killed for immunohistochemical examination of phospho-c-jun at 6 days post-lesion. In wildtype (WT) mice, there was a striking induction of phospho-c-jun-positive profiles, demonstrated by cell counts (top panel), and shown in a representative section at low power (bottom panel). The induced staining was exclusively nuclear (bottom panel inset). Absence of both the jnk2 and jnk3 genes led to a dramatic suppression of the number of these profiles (96%, p < 0.001, n = 4, each genotype). (b) To further assess the relationship between cell death and c-jun phosphorylation, we examined phospho-c-jun in mice with a homozygous jnk3 single null mutation, in which lasting protection was not observed. In these mice, there was a persistent phosphorylation of c-jun, observed on the 6OHDA-injected Experimental (EXP) side (B′). The Control (CON), injected side (A′), shows no staining. The persistent phosphorylation of c-jun in the homozygous jnk3 single null mice is likely to be because of JNK2 activity, given the virtually complete abrogation of phosphorylation in the homozygous double nulls. (***p < 0.001).

Download figure to PowerPoint

Expression of CHOP/GADD153 following 6OHDA is diminished in mice with combined homozygous jnk2/jnk3 null mutations

We have previously shown that CHOP/GADD153 is an essential mediator of dopamine neuron death in the adult 6OHDA model. Given that a virtually complete absence of apoptosis was observed in jnk2/3 combined null animals, we therefore sought to determine the effect of the null mutations on CHOP expression. Intrastriatal injection of 6OHDA resulted in the expression of CHOP protein within the nuclei of SNpc neurons, as previously reported (Silva et al. 2005b) (Fig. 4). In the jnk2/3 double null mice, the number of these CHOP-positive profiles was reduced by 85%. We therefore conclude that in this model, the JNK2 and -3 isoforms appear to play an important upstream role in regulating CHOP expression.


Figure 4.  CHOP-positive profiles are reduced in homozygous jnk2/3 double null mice following intrastriatal 6OHDA injection. At 6 days following intrastriatal injection of 6OHDA, jnk2/3 null mice were killed and SN sections processed for CHOP immunostaining. An example of a thionin-counter-stained neuronal profile, with intranuclear CHOP demonstrated by brown immunoreaction product, is shown in the inset. The number of such profiles was reduced by 85% in the jnk2/3 nulls in comparison to wildtype (WT) controls (n = 4, each genotype; p = 0.003, t-test). (**p = 0.003).

Download figure to PowerPoint

The absence of JNK2 and JNK3 isoforms, either singly or in combination, does not prevent axonal degeneration induced by 6OHDA

Given the striking inhibition of apoptosis at the cell body level in mice with single or combined null mutations in jnk2 and jnk3, we sought to examine effects on axonal degeneration induced by 6OHDA. We first characterized the morphologic, temporal and biochemical characteristics of axonal degeneration in the intrastriatal 6OHDA model. Following injection of 6OHDA, neurodegeneration is first observed by suppressed silver staining on PLDs1–2 in the striatum in the form of punctate and fibrous silver-impregnated structures, corresponding to degenerating intra-striatal axons and terminals (Fig. 5a). The degenerative process then proceeds retrograde, and on PLDs3–4 is evident primarily as silver-impregnated fibrous and spheroid structures in the medial forebrain bundle (MFB) (Fig. 5a). These spheroid structures correspond in size, shape and location to dystrophic axonal swellings identified by TH-immunostaining following axotomy (El-Khodor and Burke 2002). By PLDs5–6 degenerating fibers are observed immediately rostral to the SNpc, and apoptosis is observed in cell bodies (Fig. 5a). In addition to revealing the anatomical sequence of progressive degeneration within dopaminergic projections, this analysis also demonstrated that the large majority of degenerating neurons in the SNpc were apoptotic in their morphology.


Figure 5.  Degeneration of nigrostriatal dopaminergic axons induced by intrastriatal 6OHDA. (a) Intrastriatal injection of 6OHDA induces retrograde degeneration of nigrostriatal axons, as demonstrated by suppressed silver stain. At PLDs 1–2, numerous silver-impregnated punctate structures are observed mainly in the striatum. By PLDs 3–4, the degenerative process has progressed retrograde to the level of fibers traversing the globus pallidus. By PLDs 5–6, degenerating fibers are observed in the MFB, proximal to the SNpc. At that time post-lesion, a peak in the number of apoptotic profiles (inset, left panel, bottom) is observed. Such profiles are demonstrated by silver staining as argyrophilic (black) chromatin clumps within the nucleus. (b) Following intrastriatal injection of 6OHDA, axonal fragmentation (arrowheads) and spheroids (arrows) are observed within the TH-positive fibers projecting to the striatum. These axonal degenerative changes are identical to those described following axotomy of the MFB (El-Khodor and Burke 2002). (c) Activation of neither caspases nor calpain is observed in degenerating fibers in adult mice following intrastriatal 6OHDA. In the postnatal period, intrastriatal injection of 6OHDA induces caspase activation in degenerating fibers in both the striatum and MFB, demonstrated by immunostaining on PLD1 for a caspase cleavage fragment of β-actin (‘fractin’) (upper left panels). However, in adult mice fractin staining is not observed in either structure (upper right panels). In the bottom panels the spectrin cleavage product of calpain is observed following cortical injury (left), as described (Bartus et al. 1999). However, striatal sections from 6OHDA-treated adult mice show no staining PLD1 (right).

Download figure to PowerPoint

On PLD1 following 6OHDA injection, immunostaining for TH revealed two principal abnormalities in dopaminergic axons. Many axons, rather than having a normal thin and uniform caliber, showed serial discrete axonal swellings (Fig. 5b). In some instances, adjacent swellings remained attached by thin axonal remnants, but in other instances, no adjacent attachments remained, and the swellings therefore appeared as axonal fragments. A second morphologic abnormality identified in TH-positive, degenerating axons were larger (10–15 μm) (Fig. 5b) rounded axonal swellings, resembling the rounded structures observed by silver staining. These structure were identical to ‘spheroids’ described following MFB axotomy (El-Khodor and Burke 2002).

While some investigators have failed to demonstrate activation of caspases in degenerating axons in some contexts (Finn et al. 2000); others have succeeded (Buki et al. 2000), and we have shown that they are activated in nigrostriatal dopaminergic axons following MFB axotomy during the postnatal period (El-Khodor and Burke 2002). We therefore examined the expression of a caspase cleavage product of β-actin in the nigrostriatal projection following intrastriatal 6OHDA in both postnatal and adult mice. In postnatal mice, intense positive immunostaining for the β-actin caspase cleavage product neoepitope was observed in numerous fibers in both the striatum and MFB (Fig. 5c). However, such staining was not observed in adult mice following intrastriatal 6OHDA. We therefore conclude that while there is evidence for caspase activation in immature animals in this model, in adults there is not.

Another protease commonly implicated in axon degeneration is calpain. We therefore sought to detect a spectrin neoepitope generated by calpain cleavage (Roberts-Lewis et al. 1994), utilizing traumatic cortical injury as a positive control (Bartus et al. 1999). While immunolabeling successfully demonstrated calpain cleavage products of spectrin in fibers following traumatic brain injury, as described (Bartus et al. 1999), none were identified in the nigrostriatal pathway following 6OHDA (Fig. 5c). We therefore conclude that neither the canonical caspase or calpain proteolysis pathways appear to play a role in axonal degeneration in this model.

Following intra-striatal injection of 6OHDA in wildtype mice, at 4 weeks post-lesion, there was a complete loss of striatal dopaminergic innervation, as assessed by TH immunoperoxidase staining (Fig. 6a). There was a comparable extent of loss of dopaminergic innervation among mice lacking JNK2, or -3, or both (Fig. 6). We therefore conclude that while a combined deficiency of JNK2 and JNK3 provides virtually complete protection from apoptosis at the cell body level, it provides no protection from axonal degeneration.


Figure 6.  Homozygous jnk2/3 double null mice are not resistant to retrograde degeneration of nigrostriatal dopaminergic axons induced by intrastriatal 6OHDA. (a) Following 6OHDA, there is a virtually complete loss of TH-positive fiber staining in homozygous jnk2/3 null mice, as in wildtype, throughout the striatum. The panels to the right demonstrate TH-positive staining at the level of the anterior commisure (AC); striatum (STR) is dorsal and ventral pallidum (VP) is ventral to the AC. (b) The graphs in the right panels show quantitative measures of striatal TH immunostain optical densities. The top graph shows absolute values of optical densities following 6OHDA in wildtype (n = 7) and jnk2/3 null mice (n = 7). The bottom graph shows a similar degree of OD loss as a percent of the non-injected side, in single jnk2 [p = 0.8, NS, t-test; n = 6 jnk2 null and wildtype (WT)] and jnk3 nulls (p = 0.2, NS, t-test; n = 9 jnk3 nulls; n = 6 WT), and the double jnk2/3 nulls (p = 0.2, NS, t-test; n = 7 jnk2/3 nulls and WT), in comparison to WT processed in parallel.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

There is extensive experimental evidence that signaling through the MAPK cascade plays a role in PCD in neurons. Some of the earliest evidence indicated an essential role of c-jun phosphorylation (Estus et al. 1994), and many subsequent in vitro studies have confirmed these observations and established roles for upstream signaling kinases (Silva et al. 2005a). Importantly, these in vitro observations have been confirmed by observations in living animals. In relation to dopamine neurons of the SN, the possible role of MAPK signaling has been confirmed at the level of the mixed lineage kinases (Saporito et al. 1999; Ganguly et al. 2004; Chen et al. 2008), the JNKs (Xia et al. 2001) and c-jun (Crocker et al. 2001).

While these investigations have clearly established a role for MAPK signaling in neuron death, they have left unanswered critical questions about the neurobiology of the JNKs in PCD of dopamine neurons. Whether any of the JNK isoforms, either singly or in combination, is essential for any component of the death of these neurons in any model in which apoptosis occurs, has remained unknown. We have therefore examined the effect of homozygous jnk2 and jnk3 null mutations, singly and in combination, on dopamine neuron cell death in a neurotoxin model in which apoptosis occurs (Marti et al. 2002), and, as shown herein, is the predominant form of cell death.

We find that homozygous jnk2 and jnk3 single null mutations result in a marked suppression of apoptosis in dopamine neurons in the adult intrastriatal 6OHDA model. To explore the generality of their role in mediating apoptosis in post-mitotic SN dopamine neurons, we examined the effect of these null mutations on the induction of apoptosis in a postnatal model (Marti et al. 1997). We have previously shown that the mechanisms of PCD in the two different age contexts are different (Silva et al. 2005b). In the postnatal model, the time course (Marti et al. 1997) and the cellular pattern of caspase activation (Oo et al. 2002), suggest that the apoptosis is due primarily to loss of target-derived trophic support, or an ‘axotomy’ effect. Thus, we interpret the ability of single homozygous null mutations in jnk2 or -3 to suppress apoptosis at these two different ages as indicative of a general role for these molecules in apoptosis in post-mitotic dopamine neurons.

However, neither of these single homozygous null mutations was sufficient to provide a lasting protection for dopamine neurons in the adult model. Since both single null mutations suppressed apoptosis, we considered the possibility that the loss of one isoform may have been compensated for by an up-regulation of expression of the other. However, loss of the JNK3 isoform does not result in up-regulation of mRNA or protein expression of JNK2. In spite of the absence of up-regulation of JNK2 in JNK3 null mice, it is nevertheless likely that JNK2 compensated for the loss of JNK3 in a chronic context, because in the homozygous JNK2/3 double null mice there was virtually complete preservation of dopamine neurons. Given this observation, it is also likely that JNK3 compensates for the loss of the JNK2 isoform. We conclude therefore, that neither single JNK isoform is necessary for apoptosis in dopamine neurons, but the combination of the two is essential.

The JNKs have other substrates, in addition to c-jun, that participate in cell death, including both anti- and pro-apoptotic members of the Bcl-2 family (Maundrell et al. 1997; Donovan et al. 2002). Thus, it is possible that complete abrogation of dopamine neuron death by jnk2 and jnk3 null mutations may not be associated with suppression of c-jun phosphorylation. Indeed, Keramaris and colleagues have previously reported that protection of facial nerve motor neurons following axotomy in JNK3 null mice was not accompanied by a reduction in c-jun activation (Keramaris et al. 2005). However, we have observed that abrogation of apoptosis in the homozygous double null mice is associated with an almost complete suppression of c-jun phosphorylation. Thus, we were unable, in this model, to identify a component of JNK2/3-mediated death that could not be attributed to c-jun phosphorylation. We therefore conclude that in this model the JNK2 and -3 isoforms are likely to mediate death by c-jun activation. This conclusion is supported by our previous investigations of the ability of dominant negative inhibition of the dual leucine zipper kinases to suppress cell death in this model (Chen et al. 2008).

Previous studies in vitro have shown that 6OHDA induces oxidative modifications of proteins and endoplasmic reticulum (ER) stress (Ryu et al. 2002; Holtz et al. 2006). We have shown in living animals that an important mediator of apoptosis initiated by ER stress, CHOP/GADD153, is essential for a component of dopamine neuron death in the intrastriatal 6OHDA model (Silva et al. 2005b). The present observations demonstrating complete abrogation of apoptosis in the jnk2/3 null mice suggest that cell death mediated by the JNKs and by CHOP are unlikely to be mediated by separate, parallel mechanisms; if such were the case, we would predict that the component of death mediated by CHOP should be resistant to the jnk null mutations. While PCD mediated by the canonical intrinsic pathway can be initiated independently of cell death mediated by ER stress (Rao et al. 2002), nevertheless, there are many interactions between these pathways at the pre- and post-cytochrome c release levels (Momoi 2004). Therefore, to examine the molecular ordering of the JNKs and CHOP as upstream cell death mediators, we determined the effect of the homozygous jnk double null mutations on CHOP protein expression at the cellular level. We find that in the absence of the JNK2 and -3 isoforms, there is a 85% reduction in the number of CHOP-positive profiles. We therefore conclude that the JNKs may be acting upstream to CHOP in this model. This observation is in keeping with the known presence of an AP-1 element in the human CHOP promoter (Guyton et al. 1996), and by the induction of CHOP by JNK-mediated signaling in several cellular contexts (Gomez-Santos et al. 2005; Woo et al. 2007). It may be argued that the jnk 2/3 null mutations block 6OHDA toxicity and thereby inhibit CHOP induction indirectly, rather than through a direct upstream signaling relationship. However, it is apparent from the complete inability of the jnk2/3 null mutations to protect at the axonal level that they do not block primary 6OHDA neurotoxicity. Nevertheless, although our data are consistent with an upstream regulation of CHOP by JNK2 and -3, they do not provide direct evidence of such a relationship and further studies will be required.

There is a great deal of evidence that the molecular mechanisms of axon degeneration are distinct from the canonical pathways of PCD which mediate destruction of the cell body (Raff et al. 2002; Coleman 2005). Evidence of a role for MAPK signaling in axon degeneration in vivo has been mixed. Crocker et al. demonstrated inhibition of nigrostriatal axon degeneration in an adult MFB axotomy model by a dominant negative form of c-jun (Crocker et al. 2001). Using a pharmacologic approach, we found that mixed lineage kinase inhibition protected axons in a postnatal intrastriatal 6OHDA model (Ganguly et al. 2004). However, we have recently found that although dominant negative forms of the dual leucine zipper kinases provide substantial protection in the intrastriatal 6OHDA model at the cell body level, they provide none at the axonal level (Chen et al. 2008). These mixed results are likely to be because of a number of variables among these studies, particularly the age of the animals and the nature of the axonal lesion.

We have found that axonal injury induced by intrastriatal 6OHDA proceeds retrograde from striatal terminals to the SN over a period of 5–6 days. Age has an important influence on the mechanisms of axonal degeneration in this model. In postnatal mice, intrastriatal 6OHDA induces the activation of caspases in degenerating fibers, similar to that observed in postnatal MFB axotomy (El-Khodor and Burke 2002). However, in adult mice no evidence of caspase activation is observed. In addition, although activation of calpain is observed in adult models of acute axonal injury (Bartus et al. 1999), we do not observe its activation in this more slowly progressive model. Strikingly, we find that although homozygous jnk2/3 double null mice demonstrate a complete abrogation of apoptosis at the cell body level, they reveal no protection of the nigrostriatal axonal projection. We therefore conclude that neither JNK2 nor JNK3, singly or in combination, is necessary for retrograde axonal degeneration in this model.

The importance of this observation in relation to neuroprotection for PD is that axonal, rather than cell body, pathology is predominant throughout the course of this disease. The possibility that progressive axonal, rather than cell body, degeneration may underlie clinical progression may provide an explanation for the lack of efficacy of a mixed lineage kinase inhibitor to forestall disease progression (The Parkinson Study Group PRECEPT Investigators 2007).

In conclusion, the present investigation provides evidence that MAPK signaling, and the JNKs in particular, are therapeutic targets for preventing neuronal death in PD. However, the JNKs do not appear to participate in axonal degeneration, so other therapeutic targets will need to be identified to forestall that component of the degenerative process.


  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

This work was supported by NS26836, NS38370, The Parkinson’s Disease Foundation, and the RJG Foundation. VR and RMS were supported by The Michael J Fox Foundation Fellowship at Columbia University. We are grateful to Ms Rebecca Greene and Ms Amy Baohan for morphometric analyses.


  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  • Bartus R. T., Chen E. Y., Lynch G. and Kordower J. H. (1999) Cortical ablation induces a spreading calcium-dependent, secondary pathogenesis which can be reduced by inhibiting calpain. Exp. Neurol. 155, 315326.
  • Buki A., Okonkwo D. O., Wang K. K. and Povlishock J. T. (2000) Cytochrome c release and caspase activation in traumatic axonal injury. J. Neurosci. 20, 28252834.
  • Chen X., Rzhetskaya M., Kareva T., Bland R., During M. J., Tank A. W., Kholodilov N. and Burke R. E. (2008) Antiapoptotic and trophic effects of dominant-negative forms of dual leucine zipper kinase in dopamine neurons of the substantia nigra in vivo. J. Neurosci. 28, 672680.
  • Coffey E. T., Smiciene G., Hongisto V., Cao J., Brecht S., Herdegen T. and Courtney M. J. (2002) c-Jun N-terminal protein kinase (JNK) 2/3 is specifically activated by stress, mediating c-Jun activation, in the presence of constitutive JNK1 activity in cerebellar neurons. J. Neurosci. 22, 43354345.
  • Coggeshall R. E. and Lekan H. A. (1996) Methods for determining numbers of cells and synapses: A case for more uniform standards of review. J. Comp. Neurol. 364, 615.
  • Coleman M. (2005) Axon degeneration mechanisms: commonality amid diversity. Nat. Rev. Neurosci. 6, 889898.
  • Crocker S. J., Lamba W. R., Smith P. D., Callaghan S. M., Slack R. S., Anisman H. and Park D. S. (2001) c-Jun mediates axotomy-induced dopamine neuron death in vivo. Proc. Natl Acad. Sci. USA 98, 1338513390.
  • Donovan N., Becker E. B., Konishi Y. and Bonni A. (2002) JNK phosphorylation and activation of BAD couples the stress-activated signaling pathway to the cell death machinery. J. Biol. Chem. 277, 4094440949.
  • El-Khodor B. F. and Burke R. E. (2002) Medial forebrain bundle axotomy during development induces apoptosis in dopamine neurons of the substantia nigra and activation of caspases in their degenerating axons. J. Comp. Neurol. 452, 6579.
  • Estus S., Zaks W. J., Freeman R. S., Gruda M., Bravo R. and Johnson E. M. (1994) Altered gene expression in neurons during programmed cell death identification of c-jun as necessary for neuronal apoptosis. J. Cell Biol. 127, 17171727.
  • Finn J. T., Weil M., Archer F., Siman R., Srinivasan A. and Raff M. C. (2000) Evidence that Wallerian degeneration and localized axon degeneration induced by local neurotrophin deprivation do not involve caspases. J. Neurosci. 20, 13331341.
  • Furuya T., Hayakawa H., Yamada M., Yoshimi K., Hisahara S., Miura M., Mizuno Y. and Mochizuki H. (2004) Caspase-11 mediates inflammatory dopaminergic cell death in the 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine mouse model of Parkinson’s disease. J. Neurosci. 24, 18651872.
  • Gallyas F., Wolff J. R., Bottcher H. and Zaborsky L. (1980) A reliable and sensitive method to localize terminal degeneration and lysosomes in the central nervous system. Stain Technol. 55, 299306.
  • Ganguly A., Oo T. F., Rzhetskaya M., Pratt R., Yarygina O., Momoi T., Kholodilov N. and Burke R. E. (2004) CEP11004, a novel inhibitor of the mixed lineage kinases, suppresses apoptotic death in dopamine neurons of the substantia nigra induced by 6-hydroxydopamine. J. Neurochem. 88, 469480.
  • Gomez-Santos C., Barrachina M., Gimenez-Xavier P., Dalfo E., Ferrer I. and Ambrosio S. (2005) Induction of C/EBP beta and GADD153 expression by dopamine in human neuroblastoma cells. Relationship with alpha-synuclein increase and cell damage. Brain Res. Bull. 65, 8795.
  • Guyton K. Z., Xu Q. and Holbrook N. J. (1996) Induction of the mammalian stress response gene GADD153 by oxidative stress: role of AP-1 element. Biochem. J. 314, 547554.
  • Holtz W. A., Turetzky J. M., Jong Y. J. and O’Malley K. L. (2006) Oxidative stress-triggered unfolded protein response is upstream of intrinsic cell death evoked by parkinsonian mimetics. J. Neurochem. 99, 5469.
  • Hunot S., Vila M., Teismann P., Davis R. J., Hirsch E. C., Przedborski S., Rakic P. and Flavell R. A. (2004) JNK-mediated induction of cyclooxygenase 2 is required for neurodegeneration in a mouse model of Parkinson’s disease. Proc. Natl Acad. Sci. USA 101, 665670.
  • Jackson-Lewis V., Jakowec M., Burke R. E. and Przedborski S. (1995) Time course and morphology of dopaminergic neuronal death caused by the neurotoxin 1-methyl-4-phenyl-1,2,3,6,-tetrahydropyridine. Neurodegeneration 4, 257269.
  • Keramaris E., Vanderluit J. L., Bahadori M., Mousavi K., Davis R. J., Flavell R., Slack R. S. and Park D. S. (2005) c-Jun N-terminal kinase 3 deficiency protects neurons from axotomy-induced death in vivo through mechanisms independent of c-Jun phosphorylation. J. Biol. Chem. 280, 11321141.
  • Kholodilov N., Yarygina O., Oo T. F., Zhang H., Sulzer D., Dauer W. T. and Burke R. E. (2004) Regulation of the development of mesencephalic dopaminergic systems by the selective expression of glial cell line-derived neurotrophic factor in their targets. J. Neurosci. 24, 31363146.
  • Kuan C. Y., Yang D. D., Samanta R. D., Davis R. J., Rakic P. and Flavell R. A. (1999) The Jnk1 and Jnk2 protein kinases are required for regional specific apoptosis during early brain development. Neuron 22, 667676.
  • Kuan C. Y., Whitmarsh A. J., Yang D. D. et al. (2003) A critical role of neural-specific JNK3 for ischemic apoptosis. Proc. Natl Acad. Sci. USA 100, 1518415189.
  • Marti M. J., James C. J., Oo T. F., Kelly W. J. and Burke R. E. (1997) Early developmental destruction of terminals in the striatal target induces apoptosis in dopamine neurons of the substantia nigra. J. Neurosci. 17, 20302039.
  • Marti M. J., Saura J., Burke R. E., Jackson-Lewis V., Jimenez A., Bonastre M. and Tolosa E. (2002) Striatal 6-hydroxydopamine induces apoptosis of nigral neurons in the adult rat. Brain Res. 958, 185191.
  • Maundrell K., Antonsson B., Magnenat E. et al. (1997) Bcl-2 undergoes phosphorylation by c-Jun N-terminal kinase/stress-activated protein kinases in the presence of the constitutively active GTP-binding protein Rac1. J. Biol. Chem. 272, 2523825242.
  • Momoi T. (2004) Caspases involved in ER stress-mediated cell death. J. Chem. Neuroanat. 28, 101105.
  • Oo T. F., Siman R. and Burke R. E. (2002) Distinct nuclear and cytoplasmic localization of caspase cleavage products in two models of induced apoptotic death in dopamine neurons of the substantia nigra. Exp. Neurol. 175, 19.
  • Oo T. F., Kholodilov N. and Burke R. E. (2003) Regulation of natural cell death in dopaminergic neurons of the substantia nigra by striatal GDNF in vivo. J. Neurosci. 23, 51415148.
  • Raff M. C., Whitmore A. V. and Finn J. T. (2002) Axonal self-destruction and neurodegeneration. Science 296, 868871.
  • Rao R. V., Castro-Obregon S., Frankowski H., Schuler M., Stoka V., Del Rio G., Bredesen D. E. and Ellerby H. M. (2002) Coupling endoplasmic reticulum stress to the cell death program. An Apaf-1-independent intrinsic pathway. J. Biol. Chem. 277, 2183621842.
  • Roberts-Lewis J. M., Savage M. J., Marcy V. R., Pinsker L. R. and Siman R. (1994) Immunolocalization of calpain I-mediated spectrin degradation to vulnerable neurons in the ischemic gerbil brain. J. Neurosci. 14, 39343944.
  • Ryu E. J., Harding H. P., Angelastro J. M., Vitolo O. V., Ron D. and Greene L. A. (2002) Endoplasmic reticulum stress and the unfolded protein response in cellular models of Parkinson’s disease. J. Neurosci. 22, 1069010698.
  • Saporito M. S., Brown E. M., Miller M. S. and Carswell S. (1999) CEP-1347/KT-7515, an inhibitor of c-jun N-terminal kinase activation, attenuates the 1-methyl-4-phenyl tetrahydropyridine-mediated loss of nigrostriatal dopaminergic neurons In vivo. J. Pharmacol. Exp. Ther. 288, 421427.
  • Sauer H. and Oertel W. H. (1994) Progressive degeneration of nigrostriatal dopamine neurons following intrastriatal terminal lesions with 6 hydroxydopamine a combined retrograde tracing and immunocytochemical study in the rat. Neuroscience 59, 401415.
  • Silva R. M., Kuan C. Y., Rakic P. and Burke R. E. (2005a) Mixed lineage kinase-c-jun N-terminal kinase signaling pathway: a new therapeutic target in Parkinson’s disease. Mov. Disord. 20, 653664.
  • Silva R. M., Ries V., Oo T. F. et al. (2005b) CHOP/GADD153 is a mediator of apoptotic death in substantia nigra dopamine neurons in an in vivo neurotoxin model of parkinsonism. J. Neurochem. 95, 974986.
  • The Parkinson Study Group PRECEPT Investigators (2007) Mixed lineage kinase inhibitor CEP-1347 fails to delay disability in early Parkinson disease. Neurology 69, 14801490.
  • Woo K. J., Lee T. J., Lee S. H., Lee J. M., Seo J. H., Jeong Y. J., Park J. W. and Kwon T. K. (2007) Elevated gadd153/chop expression during resveratrol-induced apoptosis in human colon cancer cells. Biochem. Pharmacol. 73, 6876.
  • Xia X. G., Harding T., Weller M., Bieneman A., Uney J. B. and Schulz J. B. (2001) Gene transfer of the JNK interacting protein-1 protects dopaminergic neurons in the MPTP model of Parkinson’s disease. Proc. Natl Acad. Sci. USA 98, 1043310438.
  • Yang D. D., Kuan C. Y., Whitmarsh A. J., Rincon M., Zheng T. S., Davis R. J., Rakic P. and Flavell R. A. (1997) Absence of excitotoxicity-induced apoptosis in the hippocampus of mice lacking the Jnk3 gene. Nature 389, 865870.
  • Yang D. D., Conze D., Whitmarsh A. J., Barrett T., Davis R. J., Rincon M. and Flavell R. A. (1998a) Differentiation of CD4+ T cells to Th1 cells requires MAP kinase JNK2. Immunity 9, 575585.
  • Yang F., Sun X., Beech W., Teter B., Wu S., Sigel J., Vinters H. V., Frautschy S. A. and Cole G. M. (1998b) Antibody to caspase-cleaved actin detects apoptosis in differentiated neuroblastoma and plaque-associated neurons and microglia in Alzheimer’s disease. Am. J. Pathol. 152, 379389.