activated p38mapk is a novel component of the intracellular

113

Journal of Neuropathology and Experimental Neurology Vol. 63, No. 2Copyright q 2004 by the American Association of Neuropathologists February, 2004

pp. 113 119

Activated p38MAPK Is a Novel Component of the Intracellular Inclusions Found in HumanAmyotrophic Lateral Sclerosis and Mutant SOD1 Transgenic Mice

CATERINA BENDOTTI, PHD, CRISTIANA ATZORI, PHD, ROBERTO PIVA, PHD, MASSIMO TORTAROLO, PHD,MICHAEL J. STRONG, MD, SILVIA DEBIASI, PHD, AND ANTONIO MIGHELI, MD, PHD

Abstract. Cytoskeletal abnormalities with accumulation of ubiquilated inclusions in the anterior horn cells are a pathologicalhallmark of both familial and sporadic amyotrophic lateral sclerosis (ALS) and of mouse models for ALS. Phosphorylatedneurofilaments besides ubiquitin and dorfin have been identified as one of the major components of the abnormal intracellularperikaryal aggregates. As we recently found that p38 mitogen-activated protein kinase (p38MAPK) colocalized with phos-phorylated neurofilaments in spinal motor neurons of SOD1 mutant mice, a model of familial ALS, we investigated whetherthis kinase also contributed to the inclusions found in ALS patients and SOD1 mutant mice. Intense immunoreactivity foractivated p38MAPK was observed in degenerating motor neurons and reactive astrocytes in ALS cases. The intracellularimmunostaining for activated p38MAPK appeared in some neurons as filamentous skein-like and ball-like inclusions, withan immunohistochemical pattern identical to that of ubiquitin. Intracellular p38MAPK-positive aggregates containing ubiquitinand neurofilaments were also found in the spinal motor neurons of SOD1 mutant mice. Our observations indicate thatactivation of p38MAPK might contribute significantly to the pathology of motor neurons in ALS.

Key Words: Amyotrophic lateral sclerosis; Inclusion; MAP kinase; p38MAPK; SOD1; Ubiquitin.

INTRODUCTION

Amyotrophic lateral sclerosis (ALS) is a late-onset neu-rodegenerative disease characterized by selective degener-ation and death of motor neurons in spinal cord, brainstem,and cerebral cortex. About 10% of ALS cases are familial,with one fifth linked to various dominantly inherited mu-tations of the gene encoding the copper-zinc superoxide dis-mutase (SOD1) (1, 2). The clinical and pathological ex-pression of ALS is similar between familial and sporadicforms. A hallmark of all types of ALS is the cytoplasmicaccumulation of ubiquitylated skein-like inclusions in motorneurons of the spinal cord (3–5). Lewy body-like inclu-sions, although not typical for ALS, are also found. Besidesphosphorylated neurofilaments and ubiquitin, the latter in-clusions contain Cdk5 kinase (6, 7). Neurofilament proteins,together with peripherin and kinesin, are also detected inthe majority of spheroids, i.e. the axonal swellings of motorneurons (8–10). More recently, dorfin, a RING finger-typeE3 ubiquitin ligase, has also been localized in the inclusionbodies found in the motor neurons and neuronal processesof familial and sporadic ALS (11, 12) Altogether, theseinclusions are likely to be responsible for the dysfunctionand progressive degeneration of motor neurons (13). Data

From Department of Neuroscience (CB, MT), Istituto di RicercheFarmacologiche Mario Negri, Milan, Italy; Rita Levi Montalcini Centerfor Brain Repair, Department of Neuroscience (CA, RP, AM), Univer-sity of Turin, Italy; Cell Biology Research Group (MJS), Robarts Re-search Institute, London, Ontario, Canada; Department of BiomolecularSciences and Biotechnologies (SD), University of Milan, Milan, Italy.

Correspondence to: Dr. Caterina Bendotti, Laboratory of MolecularNeurobiology, Dept. Neuroscience, Istituto di Ricerche FarmacologicheMario Negri, Via Eritrea, 62, 20157, Milano, Italy. E-mail: [email protected]

The financial support of Telethon, Italy (GP0222Y01), fondazioneMonzino and MND Association, UK is greatly acknowledged.

from various in vitro and in vivo models suggest that phos-phorylation of neurofilaments is critical in their aggregationprocess; however, the kinase(s) involved remain to be iden-tified.

The pathology of ALS, including formation of the ab-normal intracellular inclusions, is strikingly mimicked byseveral lines of transgenic mice overexpressing the humanALS-linked SOD1 gene mutations (14, 15). Several hy-potheses have been proposed to explain the selective motorneuron toxicity linked to mutant SOD1, including elevatedoxidative stress, glutamate excitotoxicity and cytoskeletalabnormalities (13, 16). Recently, we found a remarkableactivation of p38 mitogen-activated protein kinase(p38MAPK) in the spinal cord of transgenic mice carryingthe human G93A SOD1 gene mutation (TgSOD1G93A),during the progression of disease (17). p38MAPK is animportant mediator of signal transduction in response to avariety of extracellular stimuli and is implicated in diversefunctions, including phosphorylation of cytoskeletal pro-teins (18, 19). Interestingly, in motor neurons of Tg-SOD1G93A mice at the initial phase of disease, we ob-served increased labeling of p38MAPK that colocalizedwith phosphorylated neurofilaments in degenerating peri-karya and neurites, suggesting a role of this kinase in neu-rofilament phosphorylation (17). In this report we used im-munohistochemical tools to investigate the localization ofactivated p38MAPK in the intracellular inclusions of spo-radic human ALS and SOD1 mutant mice.

MATERIALS AND METHODS

Human Tissues

Spinal cords from sporadic ALS (n 5 5, mean age at death59 years) and control patients affected by non-neurological dis-eases (n 5 5, mean age at death 66 years) were used (Table).

Downloaded from https://academic.oup.com/jnen/article-abstract/63/2/113/2916485by gueston 31 March 2018

114 BENDOTTI ET AL

J Neuropathol Exp Neurol, Vol 63, February, 2004

TABLEClinical Data of Human Cases Selected for the Study

Case DiseaseAge atdeath

Duration(months)

PMI(hours)

12345678

sALSsALSsALSsALSsALSsALSsALSsALS

5364566065506960

2136442518601014

12118

1513121210

910111213141516

sALSNCNCNCNCNCNCParkinson

5275566870715564

48——————

No motor pathology

1218152012161317

Abbreviations: sALS 5 sporadic amyotrophic lateral sclero-sis; NC 5 normal control; PMI 5 postmortem interval.

→

Fig. 1. A: Normal human spinal cord (a). Only weak p38MAPK immunoreactivity is present in motor neurons. ALS spinalcord (b–d). Low magnification of anterior horn showing strong p38MAPK immunoreactivity (b). p38MAPK is found in bothnucleus and cytoplasm of reactive astrocytes (c). p38MAPK is present in the cytoplasm of an atrophic motor neuron and withinits swollen axon (d, arrow). Magnifications: a, b, 3400; c, d, 31,000. B: Examples of skein-like p38MAPK-positive inclusionsin ALS motor neurons. In (a), the filamentous inclusions are located in the peripheral region of the perikaryon and inside theproximal neurite (arrow). A highly immunoreactive small cell, presumably a glial cell, is also observed (arrowhead). Other skein-like inclusions (b, c) are disordered in atrophic motor neurons around the nucleus. In (d), the aggregates appear as tangled skeinsfilling the entire cell. Magnification: 31,000. C: Comparative immunostaining of ubiquitin (a) and p38MAPK (b) in ALS motorneurons. Note the striking similarity in the morphology of the skein-like inclusions. Magnification: 31,000.

Duration of illness in ALS cases ranged between 10 and 60months. All cases had been autopsied within 8 to 15 hours fromdeath. Tissues were fixed in formalin and embedded in paraffin.

Mouse Tissues

Female C57BL/6 and transgenic mice were maintained at atemperature of 21 6 18C with relative humidity 55% 6 10%and 12 hours of light. Food (standard pellets) and water weresupplied ad libitum. Transgenic mice originally obtained fromJackson Laboratories (Bar Harbor, ME) and expressing highcopy number of mutant human SOD1 with a Gly 93 Ala sub-stitution (TgSOD1G93A) were bred and maintained on aC57BL/6 mice strain at the Consorzio Mario Negri Sud (S.Maria Imbaro, Italy). Transgenic mice were identified by poly-merase chain reaction (PCR) (1). These mice become symptom-atic at around 11 weeks and die at about 21 to 22 weeks ofage. A 52% 6 5% (SE) motor neuron loss is found in the oldestmice compared to age-matched controls (20). Female mice werekilled at 10 and 14 weeks of age corresponding respectively toa presymptomatic and early symptomatic stage of the progres-sion of the motor dysfunction. Tg SOD1 G93A (n 5 18) andnon-transgenic (n 5 6) mice combined in age groups of 9, 14,

and 19 weeks were fixed by perfusion with 4% paraformalde-hyde and their spinal cords embedded in paraffin for light mi-croscopy or in hydrophilic resin (LR white acrylic resin, Sigma,Poole, UK) for electron microscopy.

Procedures involving animals and their care were conductedin conformity with the institutional guidelines that are in com-pliance with national and international laws and policies and allefforts were made to minimize the number of animals used andtheir suffering.

Immunocytochemistry

All light microscopic analyses were performed on 5-mm sec-tions from cervical and lumbar spinal cord. To detectp38MAPK, we used a polyclonal antibody that selectively rec-ognizes its activated (phosphorylated) form (1/100 dilution,New England Biolabs, Beverly, MA or 1/500 dilution, Calbi-ochem, Nottingham, UK). Other primary antibodies used in thestudy were a polyclonal anti-phosphorylated JNK (1/100, NewEngland Biolabs), a monoclonal anti-ubiquitin (1/5000, Chem-icon International, Harrow, UK), a mouse monoclonal anti-phosphorylated neurofilaments (SMI 31, 1/5000, Sternberger,Lutherville, MD). In some cases, sections were treated withtrypsin before immunostaining. Specificity of p38MAPK im-munoreactivity was tested by preadsorbing the primary anti-body with the immunizing peptide (kindly provided by Dr. A.Nelsbach, New England Biolabs). Complete abolishment of thestaining was achieved using 0.5 mg/ml of the antibody and 50mg/ml of the peptide. Immune reactions were revealed bystreptABC using either DAB or AEC as peroxidase substrate.Sections were counterstained with hematoxylin. For double la-beling studies, serial sections were used or, alternatively, sec-tions were first immunostained for p38MAPK with AEC as sub-strate and photographed. After ethanol decoloration and elutionof anti- p38MAPK antibody with HCl, sections were immu-nostained with another primary antibody using DAB as sub-strate. For electron microscopy, thin sections of the ventral hornwere collected on Formvar-coated nickel grids and processedwith a standard post-embedding immunogold staining protocol.Briefly, the grids were incubated overnight in either 1) a poly-clonal anti-phosphorylated p38MAPK antiserum (1/250, Cal-biochem, Nottingham, UK), 2) a monoclonal anti-ubiquitin an-tiserum (1/1000, Chemicon International, Harrow, UK), or 3) amixture of the 2 primary antisera. After being thoroughly rinsedin buffers, the grids were incubated for 1 hour at room tem-perature in a goat-anti rabbit secondary antiserum conjugatedto 15 nm gold particles (British Biocell International, Cardiff,UK) (to detect p38MAPK) or in a goat-anti mouse secondary


115p38 LABELED INCLUSIONS IN ALS



116 BENDOTTI ET AL


Fig. 2. A: p38MAPK immunoreactivity in TgSOD1G93A mice. Examples of p38MAPK in motor neurons with normallooking appearance (a), or with severe (c) vacuolar changes (b). Reactive astrocytes at the border of large vacuoles are alsostrongly p38MAPK-positive (c). Magnification: 31,000. B: Examples p38MAPK-positive inclusions in the ventral horn ofTgSOD1G93A mice. Large p38MAPK-positive inclusions are found in the ventral horn (a) and in the ventral tracts (b). Doublelabeling for p38MAPK (c) and ubiquitin (d) in the same inclusion. A swollen p38MAPK-positive neurite (e, arrow) is also labeledfor phosphorylated neurofilaments (f). Magnification: 31,000.

antiserum conjugated to 15 nm gold particles (to detect ubiq-uitin), whereas for double labeling the grids were first incubatedin a biotinylated goat-anti rabbit secondary antiserum and thenin a mixture of streptavidin conjugated to 5 nm gold particles(to detect p38MAPK) and of a goat-anti mouse secondary an-tiserum conjugated to 15 nm gold particles (to detect ubiquitin).Control grids were processed in the same way, with omissionof the primary antiserum. All grids were counterstained withaqueous uranyl acetate and lead citrate and then observed andphotographed with a Jeol T8 electron microscope.

RESULTS

Some of the present series of ALS cases had been pre-viously analyzed for the expression of the JNK family ofMAPKs and of the transcription factor NFkB (21). Toexamine if p38MAPK had the same expression patternof JNK, serial sections of spinal cord were consecutivelyimmunostained for the 2 MAPKs. In control tissues, im-munoreactivity for p38MAPK reflected that of JNK,showing light cytoplasmic staining in motor neurons (Fig.1Aa). Striking upregulation of p38MAPK occurred inALS spinal cord (Fig. 1Ab). Similar to JNK, p38MAPKwas abundantly expressed by reactive astrocytes in the

anterior horns and pyramidal tracts, i.e. the areas char-acterized by prominent gliosis in ALS (Fig. 1Ac). At var-iance with JNK, however, strong p38MAPK immunore-activity was also found in the majority of motor neurons,with the highest staining intensity in atrophic and degen-erating neurons.

As shown in Figure 1B, some motor neurons showeda perikaryal distribution of p38MAPK immunoreactivitythat appeared as filamentous aggregates of various com-plexities. In some cases, the labeled filamentous inclu-sions were organized in a circular pattern around the nu-cleus, or in arrays located in the peripheral region of theneuron and in neurites. A comparison between ubiquitinand p38MAPK immune reactivity showed a striking sim-ilarity in the morphology of the labeled inclusions (Fig.1C).

As we have previously shown (17), p38MAPK wasconstitutively activated in the substantia gelatinosa ofdorsal horns in both wild type and Tg SOD1 G93A mice,while weak staining for p38MAPK was found in motorneurons of control mice. On the contrary, Tg SOD1G93A mice showed a striking activation of p38MAPK in




Fig. 3. Ultrastructural immunogold localization of p38MAPK and ubiquitin in filamentous intracellular aggregates in theventral horn of TgSOD1G93A mice. A: Single labeling for p38MAPK (revealed by 15 nm gold particles). B: Single labeling forubiquitin (revealed by 15 nm gold particles). C: Colocalization of p38MAPK (revealed by 5 nm gold particles, some indicatedby arrows) and ubiquitin (revealed by 15 nm gold particles) in the same aggregate. Magnification: A, B, 364,000; C, 3120,000.

motor neurons (Fig. 2A) that was more pronounced indegenerating ones. Vacuolated motor neurons showedprominent activation of p38MAPK that persisted even inend stages of destruction (Fig. 2Ab). Reactive glial cellslocated at the border of large vacuoles were also stronglyp38MAPK-immunoreactive (Fig. 2Ac).

More advanced stages in Tg SOD1 G93A mice arecharacterized by progressive accumulation of cytoplasmicinclusions. Analysis of sections revealed a large numberof p38MAPK-positive inclusions in the ventral horns(Fig. 2Ba) and less frequently in the intraspinal portionof the ventral roots (Fig. 2Bb).

To evaluate p38MAPK colocalization with other mol-ecules found in the aggregates, we performed double la-beling studies using antibodies to p38MAPK, ubiquitin,and phosphorylated neurofilaments. The analysis re-vealed that p38MAPK-positive inclusions also containedubiquitin (Fig. 2B, c–d) and phosphorylated neurofila-ments (Fig. 2B, e–f).

To confirm the light microscopic data, we performedan ultrastructural investigation on the ventral horn of TgSOD1 G93A mice. Selective deposition of gold particles,indicating p38MAPK immunolabeling, was found on cy-toplasmic aggregates consisting of irregularly oriented fil-aments (Fig. 3A). In adjacent sections, similar filamen-tous aggregates appeared labeled by the anti-ubiquitinantiserum (Fig. 3B). Double labeling on the same thinsection revealed that p38MAPK-positive aggregates also

contained ubiquitin (Fig. 3C). Immunogold labeling wasabsent in sections from non-transgenic and control mice(not shown).

DISCUSSION

This study shows that activated p38MAPK accumu-lates in the cytoplasm of degenerating motor neurons andastrocytes in both human ALS and transgenicSOD1G93A mice. The most striking observation is thepresence of activated p38MAPK in the intracellularskein-like inclusions of human ALS. These are hetero-geneous inclusions with an overall filamentous appear-ance that so far have been only detected with antibodiesto ubiquitin and dorfin (4, 6, 11, 12, 22). Using immu-noelectron microscopy techniques, phosphorylated neu-rofilament epitopes were found at the periphery of theskeins, suggesting that neurofilaments are a componentof the inclusions (23). Phosphorylated neurofilaments arenormally distributed along the axon or in the axon ter-minals. When they become phosphorylated their trans-port is impaired, leading to accumulation in the perikaryaor in the proximal axon (24, 25). Indeed, direct measure-ments of axonal transport in several transgenic mousemodels of ALS have revealed that slowing of neurofila-ment transport is a common and early pathological fea-ture (26, 27). Among the various kinases that are able tophosphorylate neurofilaments in vitro, members of theMAPK family play a relevant role (24). Recently, we


118 BENDOTTI ET AL


have demonstrated that activated p38MAPK is increasedand colocalizes with phosphorylated neurofilaments invacuolized perikarya and neurites of mutant SOD1 trans-genic mice at the very early stages of the disease (17).This observation suggested that abnormal activation ofp38MAPK contributes to the development of degenera-tive changes in motor neurons (17). Here, we demonstratethat 1) p38MAPK is activated in degenerating motor neu-rons and 2) p38MAPK is a component of the intracellularinclusions that characterize the pathology of both humanand mouse ALS. In particular, these data clearly showthat p38MAPK is a new marker of the skein-like arraysthat until now had been only recognized by ubiquitin,dorfin, and neurofilament antibodies (12, 23).

It has been suggested that MAPK-dependent phos-phorylation may lead to protein stabilization (28) and re-duced degradation by proteases (29). Therefore, by phos-phorylating neurofilaments and perhaps othercytoplasmic proteins, p38MAPK might promote proteinaccumulation in cytoplasmic aggregates, leading to motorneuron toxicity and dysfunction(14, 24).

Accumulation of p38MAPK in motor neurons is atvariance with the distribution of JNK, another componentof the MAPK family, that is selectively activated in as-trocytes but not in motor neurons of ALS patients (21).The diverse pattern of JNK and p38MAPK localizationin pathologically affected regions appears to be charac-teristic of ALS with respect to other neurodegenerativediseases. In fact, we have recently found that in Alzhei-mer disease, as well as in other tau-related diseases, bothkinases are activated and constantly colocalize with hy-perphosphorylated tau in all neuronal and glial inclusions(30).

Both JNK and p38 MAP kinases play major roles inlinking membrane receptors to the transcriptional ma-chinery and to cell death by apoptosis (31), although itis still controversial whether apoptosis is involved in mo-tor neuron death in ALS (20). Moreover, both kinases areinvolved in regulating nitric oxide synthase (NOS) ex-pression (32), which has been suggested to play a rele-vant role in the pathology of ALS (33). However, ourpresent study indicates that p38MAPK, rather than JNK,is implicated in the mechanism of motor neuron degen-eration in ALS. This observation is in line with a recentstudy showing selective involvement of p38MAPK path-way in the death of cultured primary motor neurons trig-gered by Fas activation, which is potentiated by ALS-linked SOD1 mutations (34).

Transcriptional upregulation of neuronal NOS (nNOS)has been proposed as the main mechanism downstreamto activation of p38MAPK in the cascade leading to mo-tor neuron death induced by Fas activation (34). In-creased production of nitric oxide (NO), particularly inmutant SOD1-linked ALS, may lead to formation ofhighly reactive molecules such as peroxynitrite (35).

Neurofilament light proteins are susceptible to peroxy-nitrite-mediated nitration, which may contribute to theircytoplasmic accumulation (36). However, although neu-rofilament inclusions in ALS are stained with anti-nitro-tyrosine antibodies, the extent of nitration of tyrosine res-idues in the NF-L protein from the spinal cord of sporadicALS cases did not differ from age-matched controls (37).

In conclusion, we suggest that activated p38MAPK isa novel marker of the neuronal skein-like inclusions inALS. Its accumulation in motor neurons with alteredmorphology, but not in healthy looking motor neurons,suggests that this pathological change may be detrimentalfor the cells. In view of the potential neurotoxicity ofp38MAPK in both human and mouse ALS, this kinasemay be considered as a candidate target for therapeuticinterventions with p38MAPK inhibitors in ALS patients.

REFERENCES

1. Rosen DR, Siddique T, Patterson D, et al. Mutations in Cu/Zn su-peroxide dismutase gene are associated with familial amyotrophiclateral sclerosis. Nature 1993;362:59–62

2. Shaw CE, Enayat ZE, Chioza BA, et al. Mutation in all five exonsof SOD-1 may cause ALS. Ann Neurol 1998;43:390–94

3. Leigh PN, Anderton BH, Dodson A, Gallo JM, Swash M, PowerDM. Ubiquitin deposits in anterior horn cells in motor neuronedisease. Neurosci Lett 1988;93:197–203

4. Lowe J, Aldrige F, Lennox G, et al. Inclusion bodies in motor cortexand brainstem of patients with motor neurone disease are detectedby immunocytochemical localisation of ubiquitin. Neurosci Lett1989;105:7–13

5. Schiffer D, Autilio-Gambetti L, Chio A, et al. Ubiquitin in motorneuron disease: Study at the light and electron microscope. J Neu-ropathol Exp Neurol 1991;50:463–73

6. Migheli A, Autilio-Gambetti L, Gambetti P, Mocellini C, Vigliani MC,Schiffer D. Ubiquitinated filamentous inclusions in spinal cord of pa-tients with motor neuron disease. Neurosci Lett 1990;114:5–10

7. Nakamura S, Kawamoto Y, Nakano S, Ikemoto A, Akiguchi I, Ki-mura J. Cyclin-dependent kinase 5 in Lewy body-like inclusions inanterior horn cells of a patient with sporadic amyotrophic lateralsclerosis. Neurology 1997;48:267–70

8. Corbo M, Hays AP. Peripherin and neurofilament protein coexist inspinal spheroids of motor neuron disease. J Neuropathol ExpNeurol 1992;51:531–37

9. Migheli A, Pezzulo T, Attanasio A, Schiffer D. Peripherin immu-noreactive structures in amyotrophic lateral sclerosis. Lab Invest1993;68:185–91

10. Toyoshima I, Sugawara M, Kato K, et al. Kinesin and cytoplasmicdynein in spinal spheroids with motor neuron disease. J Neurol Sci1998;159:38–44

11. Niwa J, Ishigaki S, Hishikawa N, et al. Dorfin ubiquitylates mutantSOD1 and prevents mutant SOD1-mediated neurotoxicity. J BiolChem 2002;277:36793–98

12. Hishikawa N, Niwa J, Doyu M, et al. Dorfin localizes to the ubi-quitylated inclusions in Parkinson’s disease, dementia with Lewybodies, multiple system atrophy, and amyotrophic lateral sclerosis.Am J Pathol 2003;163:609–19

13. Julien JP. Amyotrophic lateral sclerosis: Unfolding the toxicity ofthe misfolded. Cell 2001;104:581–91

14. Bruijn LI, Houseweart MK, Anderson KL, et al. Aggregation andmotor neuron toxicity of an ALS-linked SOD1 mutant indipendentfrom wild-type SOD1. Science 1998;281:1851–54




15. Stieber A, Gonatas JO, Gonatas NK. Aggregates of mutant proteinappear progressively in dendrites, in periaxonal processes of oli-godendrocytes, and in neuronal and astrocytic perikarya of miceexpressing the SOD1 G93A mutation of familial amyotrophic lat-eral sclerosis. J Neurol Sci 2000;177:114–23

16. Cleveland DW. From Charcot to SOD1: Mechanisms of selectivemotor neuron death in ALS. Neuron 1999;24:515–20

17. Tortarolo M, Veglianese P, Calvaresi, et al. Persistent activation ofp38 mitogen-activated protein kinase in a mouse model of familialamyotrophic lateral sclerosis correlates with disease progression.Mol Cell Neurosci 2003;23:180–92

18. Ono K, Han J. The p38 signal transduction pathway: Activationand function. Cell Signal 2000;12:1–13

19. Mielke K, Herdrgen T. JNK and p38 stresskinases-degenerative ef-fectors of signal-transduction-cascades in the nervous system. ProgNeurobiol 2000;61:45–60

20. Migheli A, Atzori C, Piva R, et al. Lack of apoptosis in mice withALS. Nature Med 1999;5:966–67

21. Migheli A, Piva R, Atzori C, Troost D, Schiffer D. c-Jun, JNK/SAPK kinases and transcription factor NF-kappa B are selectivelyactivated in astrocytes, but not motor neurons, in amyotrophic lat-eral sclerosis. J Neuropathol Exp Neurol 1997;56:1314–22

22. Leigh PN, Whitewell H, Garofalo O, et al. Ubiquitin-immunore-active intraneuronal inclusions in amyotrophic lateral sclerosis.Morphology, distribution, and specificity. Brain 1991;114:775–88

23. Migheli A, Attanasio A, Schiffer D. Ubiquitin and neurofilamentexpression in anterior horn cells in amyotrophic lateral sclerosis:Possible clues to the pathogenesis. Neuropathol App Neurobiol1994;20:282–89

24. Miller CC, Ackerley S, Brownlees J, Grierson AJ, Jacobsen NJ,Thornhill P. Axonal transport of neurofilaments in normal and dis-ease states. Cell Mol Life Sci 2002;59:323–30

25. Ackerley S, Thornhill P, Grierson AJ, et al. Neurofilament heavychain side arm phosphorylation regulates axonal transport of neu-rofilaments. J Cell Biol 2003;161:489–95

26. Williamson TL, Cleveland DW. Slowing of axonal transport is avery earjy event in the tixicicty of ALS-linked SOD1 mutants tomotor neurons. Nature Neuroscience 1999;2:50–56

27. Zhang B, Tu P, Abtahian F, Trojanowski JQ, Lee VM. Neurofila-ments and orthograde transport are reduced in ventral root axonsof transgenic mice that express human S0D1 with a G93A muta-tion. J Cell Biol 1997;139:1307–15

28. Reynolds CH, Nebreda A, Gibb GM, Utton MA, Anderton BH.Reactivating kinase/p38 phosphorylates tau protein in vitro. J Neu-rochem 1997;69:191–98

29. Musti AM, Treier M, Bohmann D. Reduced ubiquitin-dependentdegradation of c-Jun after phosphorylation by MAP kinases. Sci-ence 1997;275:400–402

30. Atzori C, Ghetti B, Piva R, et al. Activation of the JNK/p38 path-way occurs in diseases characterized by tau protein pathology andis related to tau phosphorylation but not to apoptosis. J NeuropatholExp Neurol 2001;60:1190–97

31. Gallo KA, Johnson GL. Mixed-lineage kinase control of JNK andp38 MAPK pathways. Nat Rev Mol Cell Biol 2002;3:663–72

32. Hua LL, Zhao ML, et al. Role of mitogen-activated protein kinasesin inducible nitric oxide synthase and TNFalpha expression in hu-man fetal astrocytes. J Neuroimmunol 2002;126:180–89

33. Urushitani M, Shimohama S. The role of nitric oxide in amyotro-phic lateral sclerosis. Amyotroph Lateral Scler Other Motor NeuronDisord 2001;2:71–81

34. Raoul C, Estevez AG, Nishimune H, et al. Motoneuron death trig-gered by a specific pathway downstream of Fas potentiation byALS-linked SOD1 mutations. Neuron 2002;35:1067–83

35. Beckman JS, Carson M, Smith CD, Koppenol WH. ALS, SOD andperoxynitrite. Nature 1993;364:584

36. Crow JP, Ye YZ, Strong M, Kirk M, Barnes S, Beckman JS. Su-peroxide dismutase catalyzes nitration of tyrosines by peroxynitritein the rod and head domains of neurofilament-L. J Neurochem1997;69:1945–53

37. Strong MJ, Sopper MM, Crow JP, Strong WL, Beckman JS. Nitra-tion of the low molecular weight neurofilament is equivalent insporadic amyotrophic lateral sclerosis and control cervical spinalcord. Biochem Biophys Res Commun 1998;248:157–64

Received July 7, 2003Revision received October 8, 2003Accepted October 13, 2003


activated p38mapk is a novel component of the intracellular

Documents