late stage treatment with arimoclomol delays disease progression and prevents protein aggregation in...

*Institute of Neurology, University College London, Queen Square, London, UK

�UCL Institute of Ophthalmology, UCL, London, UK

�Institute of Cytology RAS, St Petersburg, Russia

Amyotrophic lateral sclerosis (ALS) is a progressive neuro-degenerative disorder characterized by the loss of motoneu-rons in the motor cortex, brainstem and spinal cord, resultingin paralysis and death, typically within 1–5 years of diag-nosis. The precise pathogenesis remains unclear and there iscurrently no effective treatment. Approximately, 10% of ALScases are familial, and of these, about 10–20% are due tomutations in the gene of the ubiquitously expressed humanCu/Zn superoxide dismutase (SOD1) protein (Rosen et al.1993). Transgenic mice generated to over-express the mutanthuman SODG93A protein exhibit a phenotype and pathologythat resemble those observed in ALS patients (Gurney et al.1994). These mice have been widely used for both theassessment of potential therapies for ALS and also as anexperimental tool to investigate the pathomechanism of ALS.

Studies of SODG93A mice have implicated disturbances inaxonal transport, excitotoxicity and perturbed protein metab-olism because of proteasome dysfunction in ALS pathologyand therapies that target several of these functional defectshave been investigated (Shaw 2005).

One such therapeutic approach that has been tested inSODG93A mice has involved the manipulation of an

Received March 10, 2008; revised manuscript received June 26, 2008;accepted July 18, 2008.Address correspondence and reprint requests to Dr B Kalmar, Sobell

Department of Motor Neuroscience and Movement Disorders, Instituteof Neurology, University College London, Queen Square, LondonWC1N 3BG, UK. E-mail: [email protected] used: ALS, amyotrophic lateral sclerosis; HSF1, heat

shock factor-1; Hsp, heat shock protein.

Abstract

Amyotrophic lateral sclerosis (ALS) is a progressive neuro-

degenerative disorder characterized by motoneuron degen-

eration, resulting in muscle paralysis and death, typically

within 1–5 years of diagnosis. Although the pathogenesis of

ALS remains unclear, there is evidence for the involvement of

proteasome dysfunction and heat shock proteins in the dis-

ease. We have previously shown that treatment with a

co-inducer of the heat shock response called arimoclomol is

effective in the SODG93A mouse model of ALS, delaying dis-

ease progression and extending the lifespan of SODG93A mice

(Kieran et al. 2004). However, this previous study only

examined the effects arimoclomol when treatment was initi-

ated in pre- or early symptomatic stages of the disease.

Clearly, to be of benefit to the majority of ALS patients, any

therapy must be effective after symptom onset. In order to

establish whether post-symptomatic treatment with arimoclo-

mol is effective, in this study we carried out a systematic

assessment of different treatment regimes in SODG93A mice.

Treatment with arimoclomol from early (75 days) or late

(90 days) symptomatic stages significantly improved muscle

function. Treatment from 75 days also significantly increased

the lifespan of SODG93A mice, although treatment from

90 days has no significant effect on lifespan. The mechanism

of action of arimoclomol involves potentiation of the heat

shock response, and treatment with arimoclomol increased

Hsp70 expression. Interestingly, this up-regulation in Hsp70

was accompanied by a decrease in the number of ubiquitin-

positive aggregates in the spinal cord of treated SODG93A

mice, suggesting that arimoclomol directly effects protein

aggregation and degradation.

Keywords: ALS, heat shock response, motoneuron, neuro-

protection.

J. Neurochem. (2008) 107, 339–350.

JOURNAL OF NEUROCHEMISTRY | 2008 | 107 | 339–350 doi: 10.1111/j.1471-4159.2008.05595.x

� 2008 The AuthorsJournal Compilation � 2008 International Society for Neurochemistry, J. Neurochem. (2008) 107, 339–350 339

endogenous cellular defence mechanism called the heatshock response (Okado-Matsumoto and Fridovich 2002;Kalmar et al. 2005). Members of the heat shock proteinfamily (Hsps) are ubiquitously present in mammalian cellsand their functions include promoting the natural folding ofnascent proteins as well as the refolding of damaged proteins.Members of the Hsp are classified according to theirmolecular weight, and include small Hsps (ranging from 16to 40 kDa), Hsp60, Hsp70, Hsp90, and Hsp110. Thebeneficial effects of increased synthesis of some Hsp familymembers have been described in a number of disease models(for reviews, see Yenari 2002; Franklin et al. 2005;Muchowski and Wacker 2005). In the SODG93A mousemodel of familial ALS, we have previously shown thatarimoclomol, a hydroxylamine derivative that is a co-inducerof the heat shock response, could be a successful candidatefor Hsp-based therapy of ALS (Kieran et al. 2004). Pre-symptomatic or early symptomatic treatment with arimoclo-mol increased lifespan and improved motor function inSODG93A mice (Kieran et al. 2004). Arimoclomol acts via amechanism that involves the prolonged activation of HSF1,resulting in an elevated heat shock response in motoneurons(Kalmar et al. 2002; Hargitai et al. 2003; Kieran et al. 2004).Based on these findings, a Phase II Clinical Trial in ALSpatients has been established (for details, see A Multicenter,Dose Ranging Safety and Pharmacokinetics Study of arimo-clomol in ALS; http://www.clinicaltrials.gov).

However, many successful pre-clinical studies in mutantSOD1 mice have failed to translate into successful humantrials (DiBernardo and Cudkowicz 2006; Benatar 2007). It ispossible that this failure to translate is a consequence of poortrial design, inconsistency and poor control of the pre-clinicalstudies, in which SODG93A mice for example, are oftentreated with the therapeutic agent long before the onset ofdisease symptoms. Since most, if not all ALS patients willonly present to their clinician after the onset of diseasesymptoms, it is clear that any approach that is to be of benefitto ALS patients must be effective when given after symptomonset. Therefore, in this study, we aimed to extend ourunderstanding of the efficacy of arimoclomol in the SODG93A

mouse model of ALS by exploring various treatmentregimes. We have designed and conducted our experimentsaccording to the recommendations by the ENMC group onthe conduct of pre-clinical studies on ALS (Ludolph et al.2007). Thus, we tested the effects of treatment witharimoclomol from mid- and late- stages of disease on largetreatment groups that included treatment groups allocatedrandomly across several litters.

Disease stages when treatment was initiated were definedby published hallmarks of motor function and spinal cordpathology. Thus, disease onset is defined at around 75 days,when the first signs of weight loss appear (Kieran et al. 2004,2005). Further pathological symptoms appear at around90 days when motor deficits occur. Initiation of treatment

was set around these dates. Histologically, these stages canbe characterized by the proportion of motoneurons that arealready lost. Thus, at 75 days, there is already a 27% loss inmotoneuron numbers (Greensmith lab, unpublished obser-vation) and by 90 days nearly 40% of motoneurons are deadin this mouse model of ALS (Sharp et al. 2005; Hegeduset al. 2007).

In view of the known gender differences in disease onsetand survival between male and female SODG93A mice, wealso tested for any specific gender effects of arimoclomolby examining separate gender groups of SODG93A mice. Wealso established whether arimoclomol was effective whengiven orally, since our previous studies have only examinedthe effect of intraperitoneal administration of arimoclomol.The effects of treatment with arimoclomol on diseaseprogression in SODG93A mice was assessed by establishingthe effect on lifespan as well as muscle function in TibialisAnterior (TA) and Extensor Digitorum Longus (EDL)muscles, and motoneuron survival. Furthermore, in orderto gain some understanding of the mechanism by whicharimoclomol may influence disease in SODG93A mice, wealso measured Hsp70 levels in the ventral spinal cord ofthese animals using a sensitive ELISA technique andexamined the effect of arimoclomol on cellular diseasecharacteristics such as inclusion formation that is acharacteristic feature of disease in both animal modelsand ALS patients.

Materials and methods

Breeding and maintenance of transgenic SOD1G93A mouse colonyAll animals were bred and maintained by Institute of Neurology,

UCL Biological Services. Transgenic mice carrying a human SOD1

gene with a G93A mutation (TgN[SOD1-G93A]1Gur; Jackson

Laboratories, Bar Harbour) were maintained by breeding male

heterozygous carriers with female (C57BL/6 · SJL) F1 hybrids.

Mice were identified by genotyping for mutations in the human

SOD1 transgene using DNA extracted from ear snips (Gurney et al.1994). All studies were carried out according to the Guide for the

Care and Use of Laboratory Animals as adopted by the US National

Institutes of Health and under license from the UK Government

[Animals (Scientific procedures) Act 1986], following ethical

approval from the Institute of Neurology.

Drug treatment regimesIn these experiments, separate male and female SODG93A and wild-

type littermate mice were set up and each gender/transgenic group

was divided into the following treatment groups: (i) injected daily

with 10mg/kg arimoclomol (diluted in saline) intraperitoneally (i.p.)

from 75 days of age (early-symptomatic stage); (ii) treated daily

with arimoclomol i.p. from from 90 days (late-symptomatic age);

(iii) vehicle-treated (saline), i.p (iv) one group of female SODG93A

animals were treated orally, from 35 days of age with arimoclomol

(daily dose 25–30 mg/kg) given in drinking water. Each treatment

group was distributed evenly among litters used in this study. Thus,

Journal Compilation � 2008 International Society for Neurochemistry, J. Neurochem. (2008) 107, 339–350� 2008 The Authors

340 | B. Kalmar et al.

each treatment regime had its littermate vehicle-treated transgenic

and wild-type control.

Lifespan studyDisease progression was initially assessed by behavioral observation

and by monitoring body weight. From 75 days of age, each animal

in all experimental groups were weighed twice a week and from

110 days, all animals were observed and weighed daily. In these

experiments, mice were determined as having reached end-stage

when they had either lost 15% of their maximum bodyweight (as

measured at 75 days) or lost the ability to right themselves within

20 s when turned on their back. At this point, animals were

humanely killed. For lifespan analysis, a total of 75 SODG93A mice

were assessed, divided into the following treatment groups: vehicle-

treated SODG93A mice (n = 30), SODG93A mice treated with

arimoclomol from 75 days (n = 21) and SODG93A mice treated

with arimoclomol from 90 days (n = 24). Lifespan analysis was

carried out on single sex groups as well as combined gender groups.

Physiological assessment of muscle function and motor unitnumberFor the physiological assessment of muscle function and motor unit

survival, a separate set of experimental mice were established

(n = 51): Experimental groups included wild-type (WT) mice

(n = 6); SODG93A mice treated with arimoclomol (i.p) from 75 days

(n = 20); SODG93A mice treated with arimoclomol (i.p) from 90 days

(n = 20) as well as SODG93A mice treated with vehicle (saline;

n = 24). Each SODG93A experimental groupwas subdivided intomale

and female gender groups and each gender group consisted of at least

10 mice. In addition, a number of female SODG93A mice were treated

orally with arimoclomol from 35 days (n = 5).

When the mice were 120-day old, they were deeply anesthetized

(4.5% chloral hydrate; 1mL/100 g of body weight, i.p) and prepared

for in vivo analysis of muscle function. The distal tendons of the TA

and the EDL muscles in both hindlimbs were dissected free and

attached by silk thread to isometric force transducers (Dynamometer

UFI Devices, Welwyn Garden City, UK). The sciatic nerve was

exposed and sectioned, and all branches cut except for the deep

peroneal nerve that innervates the TA and EDL muscles. The length

of the muscles was adjusted for maximum twitch tension and the

muscle and nerves were kept moist with saline throughout the

recordings. All experiments were carried out at room temperature

(23�C). Isometric contractions were elicited by stimulating the nerve

to EDL and TA using square-wave pulses of 0.02 ms duration at

supra-maximal intensity, via silver wire electrodes. Contractions

were elicited by trains of stimuli at frequencies of 40, 80, and 100 Hz.

The maximum tetanic tension was measured using a computer and

appropriate software (Scope, Loredan Biomedical, Davis, CA, USA).

The number of motor units innervating the EDL muscles was

determined by stimulating the motor nerve with stimuli of increasing

intensity, resulting in stepwise increments in twitch tension because

of successive recruitment of motor axons. The number of stepwise

increments was counted to give an estimate of the number of motor

units present in each EDL muscles.

Morphological assessment of motoneuron survivalFollowing physiological assessment of muscle function, animals

were perfused transcardially with saline followed by fixative

containing 4% paraformaldehyde. The spinal cords were then

removed and processed for Nissl staining (Kalmar et al. 2002) Inthese experiments, five mice were analyzed from each experimental

group.

Measurement of Hsp70 levelsAt three stages of disease representing pre-symptomatic (60d;

n = 5), symptomatic (100d; n = 4) and end-stage (130D; n = 5),

WT (n = 6), SODG93A vehicle-treated (n = 6) and SODG93A

arimoclomol-treated mice (treated from 75 days and 90 days,

n = 6 in each subgroup) were perfused transcardially with saline

and the lumbar spinal cord and sciatic nerves removed and

homogenized in homogenization buffer (5 mM Tris, 2% SDS,

2 mM EDTA, 2 mM EGTA, 1% protease inhibitor cocktail pH 6.8).

Samples were spun at 17 530 g for 10 min and the supernatant was

collected. Protein concentration was determined using a Bio-Rad

assay system (Bi-Rad, Hemel Hempstead, UK; Cat#: 500-0116) and

a series of BSA standards. Inducible Hsp70 in mouse tissue samples

was measured by ELISA as described elsewhere (Kustanova et al.2006). In brief, ATP was conjugated to ovalbumin and immobilized

onto the surface of a medium binding capacity ELISA plate

(Greiner, Solingen, Germany). Immobilization was performed at

37�C for 1 h in buffer T (20 mM Tris–HCl, pH 7.5, 140 mM NaCl,

10 mM MgCl2). Buffer T, containing 0.2% Tween 20 (T-Tw), was

then used for blocking as well as for all subsequent steps and

washes. Calibration standards of pure Hsp70 (Stressgen, Assay

Designs Inc, Ann Harbor, MI, USA; NSP-555) and mouse tissue

extracts in T-Tw were applied to the wells for 1 h, after which the

wells were washed and anti-Hsp70 rabbit polyclonal antibody, R2

(1 : 1000; this antibody was raised in rabbits immunized with a

complex antigen consisting of native and denatured mixture of

Hsp70/Hsc70), was added, followed by goat anti-rabbit-horseradish

peroxidase antibody (Jackson ImmunoResearch Laboratories, West

Grove, PA, USA). 3,3¢,5,5¢-tetramethyl-benzidine (Sigma, St. Louis,

MO, USA, T-0440) and hydrogen peroxide were used to develop

the calorimetric reaction which was stopped after 15 min by the

addition of 0.2 M sulfuric acid. The optical density was measured at

450 nm using the Safire plate reader (Tecan, Grodig, Austria) and

Hsp70 quantities in samples were determined using the calibration

curve (0.50–250 ng/mL) obtained from the titrated standards.

ImmunohistochemistrySpinal cord sections of 120 day WT, arimoclomol- and vehicle-

treated SODG93A mice age were immunostained for ubiquitin (rabbit

polyclonal by Dako Z0458, 1 : 1000; Dako, Ily, UK) to reveal

ubiquitin inclusions and SOD1 (mouse monoclonal Clone SD-G6,

by Sigma; 1 : 500) in order to visualize SOD1 containing

inclusions. For these experiments, three mice was used from each

experimental group, thus from WT, vehicle treated SODG93A mice

as well as mice treated from 75 and 90 days of age. Staining was

visualized by horseradish peroxidase reaction using 3,3¢-diamino-

benzidine (DAB) as a substrate. Sections were then counterstained

using a Nissl stain to reveal the cellular environment.

Statistical analysisThe results were analyzed using the Mann–Whitney U-test for

comparison of independent samples. Two-tailed tests were used in

all instances, and significance level was set at p < 0.05.

� 2008 The AuthorsJournal Compilation � 2008 International Society for Neurochemistry, J. Neurochem. (2008) 107, 339–350

Arimoclomol is effective in late stage SODG93A mice | 341

Results

Effect of arimoclomol on muscle function in SODG93A miceIn order to assess whether treatment with arimoclomol couldimprove muscle function in SODG93A mice when adminis-tered after disease onset, we first assessed the maximumtetanic force of TA and EDL muscles of arimoclomol andvehicle-treated SODG93A mice at 120 days of age inSODG93A mice that had been treated with arimoclomol fromeither 75 days (early symptomatic stage) or 90 days (late-symptomatic stage). For the physiological assessment ofmuscle function, each experimental group (n = 20) wassubdivided into male and female gender groups and eachgroup consisted of at least 10 mice.

The results are summarized in Fig. 1. In 120-day old maleSODG93A mice, TA and EDL muscles produced 19.5 ± 3.5and 13.9 ± 1.7 g force, respectively. These muscles areapproximately 80% weaker than the corresponding musclesin WT mice (Fig. 1a; WT values are 112 ± 13 g for TA and81 ± 3 g for EDL). However, as can be seen in Fig. 1b, inSODG93A male mice treated with arimoclomol from 75 days,

both TA and EDL were significantly stronger than in vehicletreated SODG93A mice. Thus, by 120 days of age the TA andEDL muscles produced 36.9 ± 9 and 22.4 ± 3 g force,respectively (p < 0.05). Moreover, even late-symptomatictreatment of SODG93A male mice with arimoclomolimproved muscle force, and in SODG93A mice treated from90 days of age, TA and EDL produced 34.8 ± 6 and23.1 ± 4 g force, respectively (p < 0.05).

A similar improvement in muscle force was observed infemale SODG93A mice treated with arimoclomol from either75 or 90 days. Similarly to male SODG93A mice, TA andEDL muscles of untreated female SODG93A mice wereapproximately 80% weaker than the corresponding musclesin WT mice (87 ± 7 g in TA and 62 ± 4 g in EDL). Invehicle-treated female SODG93A mice TA and EDL produce22.6 ± 3 and 11.7 ± 8 g force, respectively. In contrast,SODG93A mice treated with arimoclomol from 75 days, bothTA and EDL were significantly stronger than in vehicle-treated SODG93A mice and TA and EDL produced 39.8 ± 5 g(p < 0.001), and 16.8 ± 1.5 g (p < 0.05), respectively. Thisimprovement in muscle force in arimoclomol-treated mice

(a)

0

20

Males Males

EDL

Females Females

40

60 Fo

rce

(g)

Fo

rce

(g)

80

140

120

100

TA

WT SOD1 Saline

* *

* *

(b)

0

5

10

15

Saline Arimoclomol 75–120 Arimoclomol 90–120 Arim oral 35–120

20

25

30

35

40

45

50TA Males EDL MalesTA Females EDL Females

****

***

**

** **

** **

**

Fig. 1 Treatment with arimoclomol

improves hindlimb muscle force in SODG93A

mice. The maximum force generated by

Tibialis Anterior (TA) and Extensor Digito-

rum Longus (EDL) muscles in vehicle and

arimoclomol-treated SODG93A mice at

120 days of age are summarized in the bar

chart. (a) Shows tetanic force production in

120 day old male and female TA and EDL

muscles compared with normal wild-type

animals of the same age. In (b), tetanic

force production by 120-day old TA and

EDL muscles of male and female SOD1

mice are shown in response to treatment

with arimoclomol (i.p.) from either an early

(75 days) or late (90 days) symptomatic

stage. As can be seen, treatment with ari-

moclomol prevented the decrease in mus-

cle force that normally occurs in both male

and female vehicle-treated SODG93A mice

by 120 days of age. However, earlier

symptomatic treatment was more effective

than late-symptomatic treatment. Pre-

symptomatic, oral treatment with arimoclo-

mol (from 35 days) was also effective in

preventing the decline in muscle force

observed in vehicle-treated SODG93A mice.

Error bars = SEM. *p < 0.001; **p < 0.05.



was also observed even when treatment was initiated at90 days, a late-symptomatic age so that TA and EDLproduced 36.3 ± 5, and 15.9 ± 1.7 g force respectively(p < 0.05).

In a separate experimental group of female SODG93A mice,arimoclomol was supplemented in drinking water from35 days of age in order to assess its effectiveness whenadministered orally. The mice were housed in separate cagesand fluid consumption recorded every 72 h to assess drugdoses for each animal. The mice consumed 25–30 mg/kgarimoclomol daily and the effect on motor performance wasassessed at 120 days of age. The results showed oraltreatment from 35 days is effective and TA and EDL musclesproduced 35.4 ± 5 and 16.4 ± 2 g force, respectively, whichwas significantly more than that produced by TA and EDLmuscles in vehicle-treated SODG93A mice (p < 0.05).

These results show that post-symptomatic treatment witharimoclomol improved muscle strength in TA and EDLmuscles, in both male and female SODG93A mice. However,arimoclomol was more effective in preserving muscle forcein TA than EDL, and more effective in male than femaleSODG93A mice. Thus, arimoclomol-treated TA muscles wereapproximately 80% stronger in male and 60% stronger infemale SODG93A mice than the TA in vehicle-treated

SODG93A mice and EDL was 60% stronger in male and40% stronger than EDL in vehicle-treated female SODG93A

mice. However, these differences between the male andfemale treatment groups did not reach statistical significance.Oral treatment of female SODG93A mice with arimoclomolfrom 35 days improved TA and EDL muscle function byaround 60% and 40%, respectively, which is similar to theimprovement seen following late stage intra-peritonealtreatment in female mice.

Effect of arimoclomol treatment on motor unit survival inSODG93A miceAt 120 days of age, the number of functionally active motorunits that innervate the EDL muscle was assessed in eachanimal. Representative examples of motor unit recordings ofmice from each experimental group are shown in Fig. 2a andthe results of mean motor unit number are summarized inFig. 2b. Normal EDL muscles are innervated by, on average,30 motor units but in vehicle-treated SODG93A male mice by120 days of age only 10 ± 0.7 motor unit survived. Incontrast, in male SODG93A mice that have been treated witharimoclomol from 75 days of age, 14 ± 0.6 motor unitssurvived (p < 0.001). Late-symptomatic treatment however,was less effective at preserving motor units, and on average

(a)

(b)

20 ms

1.8g 0.65g 0.9g 0.8g

0 2 4 6 8

10 12 14 16 18

Saline Arimoclomol 75– 120

Arimoclomol 90– 120

Arim oral 35–120

Males

WT SOD1G93A SOD1G93A+Arim 75-120D

SOD1G93A+Arim90-120D

Females

* ** **

*

**

Fig. 2 The effect of arimoclomol on motor unit survival in EDL mus-

cles of SODG93A mice. Typical examples of motor unit traces from EDL

muscles of female WT, vehicle and arimoclomol-treated 120 day old

SODG93A mice are shown in (a) and the results of mean motor unit

survival in each experimental group is summarised in the bar chart (b).

By 120 days, a significant proportion of motor units have died in

vehicle (saline)-treated SODG93A mice, although more motor units

survive in female compared with male SODG93A mice at this stage.

Although treatment with arimoclomol (i.p.) from 75 days preserves a

significant proportion of motor units in both gender groups, treatment

from 90 days only results in significant improvement in motor unit

survival in female SODG93A mice. Pre-symptomatic (35 days) oral

treatment of female SODG93A mice with arimoclomol, rescues a similar

proportion of motor units as observed following i.p. treatment from

90 days. Error bars = SEM. *p < 0.001; **p < 0.05.



12 ± 0.5 motor units survived in EDL in male SODG93A

mice treated with arimoclomol from 90 days, which was stillstatistically different from the number of motor units presentin vehicle-treated EDL (Fig. 2b; p < 0.05).

Arimoclomol was equally effective in female SODG93A

mice. In vehicle-treated female SODG93A mice by 120 days,slightly more motor units survive than in male mice at thisage, and 11 ± 0.7 motor units innervate the EDL muscle.Treatment with arimoclomol from 75 days, however, signif-icantly improved motor unit survival, and 16 ± 0.6 motorunits innervated EDL (p < 0.001). As observed for maleSODG93A mice, treatment from 90 days was less effective,but in female mice, the improvement in motor unit survivalwas significant and 13 ± 0.5 motor units survived (p < 0.05).Oral treatment of female SODG93A mice with arimoclomolfrom 35 days was also effective in improving motor unitsurvival, and 14.8 ± 5 units innervated EDL muscles oforally treated female SODG93A mice (p < 0.05).

Effect of arimoclomol on the lifespan of SODG93A miceFor the assessment of the effect of arimoclomol on lifespan, atotal of 75 SODG93A mice were assessed: (i) vehicle-treatedSODG93A mice (n = 30); (ii) SODG93A mice-treated witharimoclomol from 75 days (n = 21); (iii) SODG93A micetreated with arimoclomol from 90 days (n = 24).

The effect of treatment with arimoclomol (i.p) from 75 and90 days on the lifespan of separate male and female groupsof SODG93A mice as well as the combined gender group wasexamined and the results are summarized in Fig. 3. Theresults show that vehicle-treated male SODG93A mice(Fig. 3a) had an average lifespan of 123 ± 2 days. Treatmentwith arimoclomol, however, significantly increased survivalof male SODG93A mice even when administered aftersymptom onset, such that following treatment from 75 daysSODG93A mice reached end-stage at 133 ± 3, which is asignificant increase in lifespan of these mice (p < 0.05).Treatment from 90 days also increased lifespan to129 ± 2 days, although this improvement in lifespan didnot reach statistical significance.

As previously observed, female SODG93A mice live forlonger than males, and in this study, they had an averagelifespan of 130 ± 2 days. As can be seen in Fig. 3b, asobserved in male SODG93A mice, treatment of female micewith arimoclomol also increased survival. Thus female micetreated from 75 days reached end-stage at 142 ± 3 days,which was significantly later than vehicle-treated femaleSODG93A mice (p < 0.05). Although treatment with arimo-clomol from 90 days also increased survival to139 ± 3 days, this improvement in lifespan did not reachstatistical significance.

We also examined the overall effect of arimoclomol on thelifespan of SODG93A irrespective of gender. As can be seenin Fig. 3c, in spite of the known difference in lifespanbetween (vehicle treated) male and female SODG93A mice,

which differs, on average, by as much as 7–10 days, theoverall increase in lifespan induced by treatment witharimoclomol from 75 days of age in a mixed gender groupof SODG93A mice was still highly significant (p = 0.002).

The effect of arimoclomol on motoneuron survival inSODG93A miceThe number of motoneurons surviving in the sciatic motorpool of the spinal cord was assessed and the results aresummarized in Fig. 4. In the L3-L5 region of the lumbarspinal cord of WT mice there were on average 379 ± 9motoneurons present in the sciatic motor pool (Fig. 4a and

(a)

0%

20%

40%

60%

80%

100%

120%

15 17 19 21 23 25

(b)

0%

20%

40%

60%

80%

100%

120%

15 17 19 21 23 25

SODA75F (n = 11)

SODA75M (n = 10)

SODA90M (n = 15)

SOD M (n = 17)

SOD90AF (n = 11)

SOD F (n = 13)

SOD (n = 30)

SODA90 (n = 24)

SODA75 (n = 21)

(c) Cumulative survival of SODG93A mice

Survival of female SODG93A mice

Survival of male SODG93A mice

(Males + Females)

–20%

0%

20%

40%

60%

80%

100%

120%

15 17 19 21 23 25

Weeks

Weeks

Weeks

% o

f su

rviv

ing

mic

e%

of

surv

ivin

g m

ice

% o

f su

rviv

ing

mic

e

Fig. 3 The effect of treatment with arimoclomol on lifespan of SODG93A

mice. Male and female SODG93A mice were treated with arimoclomol

daily (10 mg/kg, ip) from 75 or 90 days and the effect on lifespan was

assessed. The cumulative survival in each experimental group is sum-

marised for male (a) and female (b) SODG93A mice. The combined

survival of both male and female SODG93A mice is shown in (c). It can be

seen that arimoclomol treatment from 75 days results in a significant

increase in lifespan for both male and female SODG93A mice.



d). In male SODG93A mice, at 120 days of age, only 125 ± 5motoneurons survived in the same region of the spinal cord.Thus, only 33% of motoneurons survived in the sciatic motorpools of SODG93A mice at this stage. Treatment witharimoclomol from 75 and 90 days of age protected asignificant proportion of motoneurons that would have diedby 120 days, so that in these animals 245 ± 7 (p < 0.05) and217 ± 33 motoneurons survived, respectively. Thus, in maleSODG93A mice treated with arimoclomol from 75 and90 days, 65% and 57% respectively, of motoneuronssurvived compared with WT, although the increased survivalobserved after 90 day treatment failed to reach statisticalsignificance.

The progression of the disease in female SODG93A mice isslower than in male mice, and this is reflected in greatermotoneuron survival in female than male SODG93A mice ofthe same age. Thus, at 120 days, there are still 156 ± 12motoneurons in the sciatic motor pool of female SODG93A

mice, representing 41% of the number of motoneurons inWT spinal cord (Fig. 4b and d). Arimoclomol treatment from

75 and 90 days significantly increased motoneuron survivalin female SODG93A mice, and at 120 days there were still298 ± 11 and 273 ± 15 motoneurons present in the sciaticpool, respectively (Fig. 4c and d; p < 0.05), representing78% and 72% respectively, of the normal number ofmotoneurons.

The effect of arimoclomol on Hsp70 expression in SODG93A

miceWe have previously shown that treatment of SODG93A micefrom 35 days of age results in activation of HSF1 and an up-regulation in the expression of Hsp70 and Hsp90 in spinalcords of end-stage SODG93A mice. In this study, we measuredthe level of inducible Hsp70 using a sensitive ELISA methodin the soluble fraction of spinal cord and sciatic nervesamples of arimoclomol and vehicle-treated SODG93A miceat various disease stages, i.e., 60 days, 100 days and130 days and the results are summarised in Fig. 5.

The results showed that during disease progression inSODG93A mice, there was a progressive loss of inducible

(a) (b)

(c)

0

50

100

150

200

250

300

350

400

450

WT Saline Arimoclomol 75–120

Arimoclomol 90– 120

MN

nu

mb

er

Males Females

**

**

(d)

**

Fig. 4 The effect of treatment with arimoclomol on motoneuron sur-

vival in SODG93A mice. Examples of spinal cord ventral horn sections

stained for Nissl from a WT mouse (a) and 120 day SODG93A mouse

treated with vehicle (b) or arimoclomol (c) from 75 days of age are

shown. The number of motoneurons in the sciatic motor pool (see

insets) was established for each experimental group and the results

are summarised in the bar chart (d). It can be seen that the dramatic

loss of motoneurons that occurs in SODG93A mice by 120 days of age is

reduced in SODG93A mice treated with arimoclomol. Scale bars: main

images = 100 lm; insets = 20 lm. Error bars = SEM. **p < 0.05.



Hsp70 in the soluble fraction of spinal cord homogenates.Thus, in WT spinal cord there is 950 ± 68 ng Hsp70/mgprotein, whereas in SODG93A mice at 60, 100 and 130 days,this was reduced to 864 ± 44, 827 ± 166 and 778 ± 58 ngHsp70/mg protein, respectively (Fig. 5a). However, inSODG93A mice that were treated with arimoclomol from 75and 90 days, there was a significant elevation in Hsp70 levelsassessed at 130 days and as can be seen in Fig. 5b, there was1047 ± 3 and 977 ± 5 ng Hsp70/mg protein in spinal cordsof SODG93A mice treated from 75 and 90 days of age,respectively.

In addition to assessing the effects of arimoclomol onHsp levels within the spinal cord, we also examined theeffects on Hsp70 expression in peripheral nervous tissue,the sciatic nerve. Our results showed that peripheralnervous tissue contains higher levels of Hsp70 than spinalcord, such that in the sciatic nerve of WT mice, there was1354 ± 300 ng Hsp70 present per mg protein. Duringdisease progression in SODG93A mice, as observed in thespinal cord, there was a progressive decrease in Hsp70levels in the sciatic nerve. Thus, in 60, 100, and 130-dayold SODG93A mice, there was 1094 ± 230, 1143 ± 126, and774 ± 40 ng Hsp70 present per mg protein (Fig. 5c).Treatment of SODG93A mice with arimoclomol however,significantly increased Hsp70 levels in the sciatic nerve.Thus, in the sciatic nerve of 130-day old SODG93A micetreated with arimoclomol from 75 and 90 days, Hsp70

levels were elevated to 1584 ± 240 and 1396 ± 426 ngHsp70/mg protein (Fig. 5d).

Arimoclomol reduces the appearance of ubiquitinaggregates in the spinal cord of SODG93A miceOne of the pathological characteristics of disease progressionin both ALS patients and SODG93A mice is the appearance ofubiquitin positive inclusions in the spinal cord. To establishwhether the up-regulation of Hsps induced by treatment witharimoclomol could reduce the number of ubiquitinatedinclusions, we examined the pattern of ubiquitin and SOD1immunoreactivity in spinal cord sections of WT, 120 dayvehicle-treated SODG93A and 120 day SODG93A mice treatedwith arimoclomol from 75 days. Examples of spinal cordsections stained for ubiquitin are shown in Fig. 6. It can beseen that ubiquitin immunoreactivity in normal WT spinalcords was barely detectable (Fig. 6a). In SODG93A mice at120 days, several ubiquitin positive inclusions of differentsizes were observed in the lumbar spinal cord, in and aroundshrunken, degenerating motoneurons, which display reducednumber of processes (Fig. 6b). In contrast, in spinal cords of120-day old SODG93A mice treated with arimoclomol from75 days, there were fewer ubiquitin positive inclusions andmost motoneurons did not contain any detectable ubiquitin-positive inclusions (Fig. 6c). In spinal cords of SOD1G93A

treated from 90 days the appearance of ubiquitin inclusionswas similar to that observed in mice treated from 75 days,

(a) (b)Hsp 70 in SOD1 spinal cord Hsp 70 in SOD1 spinal cord 130D

Hsp 70 in SOD1 sciatic nerves Hsp 70 in SOD1 sciatic nerves 130D

500

600

700

800

900

1000

1100

1200

1000

1500

2000

60 D 130 D

WT

0

200

400

600

800

1200

1000

WT SOD SOD + Arim 75 D

SOD + Arim 75 D SOD + Arim 90 D

SOD + Arim 90 D

hsp

70 (

ng

/mg

pro

t)h

sp70

(n

g/m

g p

rot)

hsp

70 (

ng

/mg

pro

t)h

sp70

(n

g/m

g p

rot)

****

(c) (d)

0

500

1000

1500

2000

500

WT

60 D 100 D 130 D

SOD

100 D

SOD

0WT SOD

** **

Fig. 5 The effect of arimoclomol on Hsp70 levels in CNS and PNS

tissues of SODG93A mice. Hsp70 levels were measured in homogen-

ates of spinal cord (a) and sciatic nerves (b) from WT mice and

SODG93A mice at various stages of disease. In addition the effect of

treatment with arimoclomol from 75 days or 90 days on Hsp70 levels

in spinal cord (c) and sciatic nerves (d) of SODG93A mice was also

assessed. It can be seen that in the spinal cord and sciatic nerve of

SODG93A mice, levels of Hsp70 decrease during disease progression

(a and b). However, treatment with arimoclomol prevented this

decrease in Hsp70 in both spinal cords and sciatic nerves (c and d), so

that following treatment from either 75 or 90 days, Hsp70 levels in

spinal cord and sciatic nerve of 130 day (endstage) SODG93A mice were

not different from normal and were significantly greater than in vehicle-

treated SODG93A mice. (p < 0.05; Error bars = SEM.). **p < 0.05.



but the reduction in the number of inclusions was notobvious. A similar pattern of inclusion formation wasobserved for SOD1 immunoreactivity. Thus, the presenceof SOD1 containing aggregates was less pronounced inarimoclomol-treated spinal cords than in vehicle-treatedcontrols.

Discussion

We have previously described the beneficial effects ofarimoclomol on disease progression and survival inSODG93A mice (Kieran et al. 2004). However, this study

did not address several issues that may be important for theclinical application of arimoclomol in ALS patients. In thepresent study we therefore addressed a number of unan-swered questions, including (i) whether arimoclomol had anyspecific gender effects, (ii) whether arimoclomol was effec-tive when given orally and (iii) whether arimoclomolimproved muscle function and overall survival in SODG93A

mice when given at various stages after the onset ofsymptoms. Our results show that arimoclomol has similarbeneficial effects in male and female SODG93A mice andsignificantly increases muscle function in both gendergroups. These beneficial effects are observed even whenarimoclomol treatment is initiated after disease onset, indeedeven when given at late-symptomatic age. Interestingly, theresults observed following early, pre-symptomatic treatment(35 days) with arimoclomol were not significantly better thanthose observed following late-symptomatic treatment, sug-gesting that arimoclomol is only effective after symptomonset when motoneurons are likely to be under a significantlevel of stress. However, although treatment with arimoclo-mol at both symptom onset (75 days) and at a late stage ofdisease (90 days) both significantly improved muscle func-tion, only SODG93A mice treated from 75 days showed asignificant increase in lifespan, so that treatment from90 days had no beneficial effects on the survival of SODG93A

mice. However, in view of the known gender differences inthe survival of SODG93A mice, where female mice can livefor 7–10 days longer than male mice, it was surprising to findthat the overall improvement in survival of the combinedgroup of male and female SODG93A treated with arimoclo-mol from 75 days, was still highly significant. These resultsindicate that arimoclomol has a significant beneficial effecton the lifespan of SODG93A mice.

Our results, therefore, show that treatment with arimo-clomol provides clear benefits in the SODG93A mousemodel of ALS mice even when treatment is initiated aftersymptom onset. These findings may be significant for theclinical development of arimoclomol for the treatment ofALS, since almost all ALS patients only present in theclinic after symptom onset. Indeed, even when the diseaseis relatively advanced in SODG93A mice, at 90 days, when avery large proportion of motoneurons, around 40%, willalready have died (Sharp et al. 2005; Hegedus et al. 2007),arimoclomol treatment can have clear beneficial effects onmuscle function, although it did not appear to significantlyextend lifespan. Since late stage treatment still had apowerful effect on muscle function, it is possible that theactions of arimoclomol can be mediated through theregeneration process of the motor nerve, inducing reinner-vation and sprouting in these axons, leading to reinnerva-tion of muscle fibers. Increased reinnervation can lead toenhanced muscle activity even in the presence of less motorneurons since newly recruited muscle fibers increase motorunit size and thus, muscle force elicited by the same

(a)

(b)

(c)

Fig. 6 Treatment with arimocomol reduces ubiquitin immunoreactivity

in the spinal cord of SODG93A mice. Low and high power illustrations

(insets) of spinal cord sections of 120-day old (a) WT, (b) vehicle-

treated SODG93A and (c) SODG93A mice treated from 75 days of age

are shown. In the ventral horn of the spinal cord of vehicle-treated

SODG93A mice (b), ubiquitin positive inclusions can be seen, in and

around shrunken, degenerating and surviving motoneurons. However,

in arimoclomol-treated SODG93A mice (c), ubiquitin immunoreactivity

was reduced, and only a few inclusions were present, with most large

motoneurons spared. Scale bars: 200 lm; insets: 20 lm.



stimulus. This would explain the improvement of motorfunction without a positive effect on lifespan that we haveseen in animals treated from a late stage of the disease. Thishypothesis is supported by our finding that arimoclomoltreatment has a strong Hsp70 inducing effect in sciaticnerves even at late disease stages, implying that arimoclo-mol has powerful effects in the periphery. It is possible thatthere is a potential window of opportunity when treatmentwith arimoclomol can be most effective in the disease, afterwhich the benefits of arimoclomol become restricted tomotor function.

We found some gender differences in the effects ofarimoclomol on the muscle function of male and femaleSODG93A mice. For example, EDL muscle of male SODG93A

mice improved more after treatment with arimoclomol thanthe same muscle group of female mice. These differences didnot reach statistical significance they may indicate that malemice have stronger muscles, with larger motor units. Thus,the same improvement in motoneuron/motor unit survivalafter treatment with arimoclomol can result in greater muscleperformance since each male motor unit is stronger.

Since arimoclomol-analogues act as a co-inducer of theheat shock response (Vigh et al. 1997), we also measuredlevels of Hsp70 in the soluble fraction of lumbar spinal cordsand sciatic nerves using a sensitive ELISA method. Theresults showed that in SODG93A mice, Hsp70 levels in thesoluble fraction of both spinal cord and sciatic nerve decreaseduring disease progression. Previous results have shown thatthere is no up-regulation in Hsp70 expression in tissue fromboth ALS patients and SODG93A mice during diseaseprogression (Vleminckx et al. 2002; Batulan et al. 2003).This apparent poor activation of the heat shock response instressed motoneurons in ALS is thought to be the result of anintrinsic inability of motoneurons to activate HSF1 (Batulanet al. 2003). In addition, histological and biochemicalevidence indicates that mutant, but not wild-type SOD1,binds and sequesters cellular Hsp70 into insoluble aggregates(Watanabe et al. 2001; Matsumoto et al. 2005) reducing theamount of available cytosolic Hsp70. However, the resultspresented in this study suggest that mutant SOD1 expressionleads to loss of Hsp70 in the soluble fraction of both centralneurons in the spinal cord as well as distal axons in theperiphery. More importantly, this reduction in soluble Hsp70is detectable at an early stage of the disease and Hsp70 levelscontinue to decline throughout disease progression, indicat-ing a change that is concomitant with disease progression.However, treatment with arimoclomol increased cytosoliclevels of Hsp70 by approximately two-fold. As can be seenin Fig. 5, the tendency to increase in Hsp70 levels inarimoclomol-treated spinal cord and sciatic nerves wastreatment regime-dependant, although these differences donot reach statistical significance. Thus, there were higherHsp70 levels present in tissues that had been treated fromearlier disease stages than in animals in which treatment was

initiated late symptomatically. It is also possible that theincreased soluble Hsp70 levels are, in part a consequence offewer inclusions present in treated spinal cords and thereforeless Hsp70 is sequestered in insoluble fraction as it has beenshown previously for Hsp70 in stressed cells (Beckmannet al. 1990, 1992).

Intracellular inclusions of aggregated protein containingubiquitin, SOD1 and Hsps have been observed in tissuesfrom both ALS patients and SODG93A mice (Bruijn et al.1998; Watanabe et al. 2001). Our results suggest that theincrease in Hsp70 induced by arimoclomol may besufficient to compensate for at least some of the loss ofHsp70 action that normally occurs in this disease. We founda marked reduction in the number of ubiquitin-positivestructures in arimoclomol-treated spinal cords, suggestingthat there was a reduction in aggregate formation. Sincearimoclomol and the family of compounds it belongs tohave been shown to prolong the activation of HSF1, it islikely that the actions of arimoclomol are not limited to theinduction of only Hsp70 (Hargitai et al. 2003). Moreover, ithas been shown that this family of compounds interact withmembrane lipids and modify their dynamic properties uponcellular stress condition so that it leads to even greateractivation of heat shock genes (Torok et al. 2003; Vighet al. 2007). Indeed, we have previously shown that Hsp90is also up-regulated in SODG93A spinal cord followingarimoclomol treatment (Kieran et al. 2004). Therefore, it islikely that arimoclomol acts as a co-inducer of a range ofHsps whose expression is regulated by HSF1. There isample of evidence that in order to have beneficial effects, itis necessary to have several Hsp acting together in acoordinated fashion (McClellan and Frydman 2001; Mayerand Bukau 2005). For example, in cellular models ofneurodegeneration, Hsp70 over-expression is more effectivein protecting neurons from cell death when it is over-expressed with other Hsps, such as Hsp40, which acts as aco-chaperone to Hsp70 (Patel et al. 2005). In view of theseresults it is therefore not surprising that over-expression ofHsp70 alone is not sufficient to protect against mutantSOD1 induced motoneuron toxicity in SODG93A mice (Liuet al. 2005). In other models of mutant SOD1 inducedmotoneuron death, induction of individual Hsps resulted inlimited benefits compared with robust neuroprotectiveeffects observed when the expression of multiple Hsps isachieved, either by over-expressing a constitutively activeform of HSF1 in cells or exposing them to mild pre-conditioning heat shock (Batulan et al. 2006; Krishnanet al. 2006). However, although this strategy proves a proofof principle, i.e. that broad activation of HSF1 offers greaterneuroprotective effects than induction of individual Hsps inisolation, this approach has more scientific than clinicalrelevance due to the possible deleterious side effects of apotential HSF1 therapy that may, for example, causemultiple malignancies. Thus, arimoclomol has the advan-



tage that although it can induce a robust heat shockresponse that involves the up-regulation of a number ofHsps, thereby optimising the neuroprotective effects of thisendogenous cellular defence mechanism, arimoclomol canonly do so in cells already under stress since it is aco-inducer of the heat shock response and not an activator.Thus arimoclomol may be a particularly powerful pharma-ceutical tool for the treatment of neurodegenerative diseasessuch as ALS, since it has the ability to selectivelyup-regulate a number of cytoprotective Hsps where andwhen they are needed.

Acknowledgments

This work was supported by the Brain Research Trust (UK), MRC

and the Muscular Dystrophy Association (USA).

References

Batulan Z., Shinder G. A., Minotti S., He B. P., Doroudchi M. M.,Nalbantoglu J., Strong M. J. and Durham H. D. (2003) Highthreshold for induction of the stress response in motor neurons isassociated with failure to activate HSF1. J. Neurosci. 23, 5789–5798.

Batulan Z., Taylor D. M., Aarons R. J., Minotti S., Doroudchi M. M.,Nalbantoglu J. and Durham H. D. (2006) Induction of multipleheat shock proteins and neuroprotection in a primary culturemodel of familial amyotrophic lateral sclerosis. Neurobiol. Dis. 24,213–225.

Beckmann R. P., Mizzen L. E. and Welch W. J. (1990) Interaction of Hsp70 with newly synthesized proteins: implications for protein fold-ing and assembly. Science 248, 850–854.

Beckmann R. P., Lovett M. and Welch W. J. (1992) Examining thefunction and regulation of hsp 70 in cells subjected to metabolicstress. J. Cell Biol. 117, 1137–1150.

Benatar M. (2007) What zebras and mice can teach us about familialALS. Neuromuscul. Disord. 17, 671–672.

Bruijn L. I., Houseweart M. K., Kato S., Anderson K. L., AndersonS. D., Ohama E., Reaume A. G., Scott R. W. and Cleveland D. W.(1998) Aggregation and motor neuron toxicity of an ALS-linkedSOD1 mutant independent from wild-type SOD1. Science 281,1851–1854.

DiBernardo A. B. and Cudkowicz M. E. (2006) Translating preclinicalinsights into effective human trials in ALS. Biochim. Biophys. Acta1762, 1139–1149.

Franklin T. B., Krueger-Naug A. M., Clarke D. B., Arrigo A. P. andCurrie R. W. (2005) The role of heat shock proteins Hsp70 andHsp27 in cellular protection of the central nervous system. Int. J.Hyperthermia 21, 379–392.

Gurney M. E., Pu H., Chiu A. Y., Dal Canto M. C., Polchow C. Y.,Alexander D. D., Caliendo J., Hentati A., Kwon Y. W. and DengH. X. (1994) Motor neuron degeneration in mice that express ahuman Cu,Zn superoxide dismutase mutation. Science 264, 1772–1775.

Hargitai J., Lewis H., Boros I. et al. (2003) Bimoclomol, a heat shockprotein co-inducer, acts by the prolonged activation of heatshock factor-1. Biochem. Biophys. Res. Commun. 307, 689–695.

Hegedus J., Putman C. T. and Gordon T. (2007) Time course of pref-erential motor unit loss in the SOD1 G93A mouse model of am-yotrophic lateral sclerosis. Neurobiol. Dis. 28, 154–164.

Kalmar B., Burnstock G., Vrbova G., Urbanics R., Csermely P. andGreensmith L. (2002) Upregulation of heat shock proteins rescuesmotoneurones from axotomy-induced cell death in neonatal rats.Exp. Neurol. 176, 87–97.

Kalmar B., Kieran D. and Greensmith L. (2005) Molecular chaperonesas therapeutic targets in amyotrophic lateral sclerosis. Biochem.Soc. Trans. 33, 551–552.

Kieran D., Kalmar B., Dick J. R., Riddoch-Contreras J., Burnstock G.and Greensmith L. (2004) Treatment with arimoclomol, a coin-ducer of heat shock proteins, delays disease progression in ALSmice. Nat. Med. 10, 402–405.

Kieran D., Hafezparast M., Bohnert S., Dick J. R., Martin J., Schiavo G.,Fisher E. M. and Greensmith L. (2005) A mutation in dyneinrescues axonal transport defects and extends the life span of ALSmice. J. Cell Biol. 169, 561–567.

Krishnan J., Lemmens R., Robberecht W. and Van Den B. L. (2006)Role of heat shock response and Hsp27 in mutant SOD1-dependentcell death Exp. Neurol. 200, 301–310.

Kustanova G. A., Murashev A. N., Karpov V. L., Margulis B. A.,Guzhova I. V., Prokhorenko I. R., Grachev S. V. and Evgen’ev M.B. (2006) Exogenous heat shock protein 70 mediates sepsis man-ifestations and decreases the mortality rate in rats. Cell StressChaperones 11, 276–286.

Liu J., Shinobu L. A., Ward C. M., Young D. and Cleveland D. W.(2005) Elevation of the Hsp70 chaperone does not effect toxicity inmouse models of familial amyotrophic lateral sclerosis. J. Neuro-chem. 93, 875–882.

Ludolph A. C., Bendotti C., Blaugrund E. et al. (2007) Guidelines forthe preclinical in vivo evaluation of pharmacological active drugsfor ALS/MND: report on the 142nd ENMC international work-shop. Amyotroph Lateral Scler. 8, 217–223.

Matsumoto G., Stojanovic A., Holmberg C. I., Kim S. and MorimotoR. I. (2005) Structural properties and neuronal toxicity of amyo-trophic lateral sclerosis-associated Cu/Zn superoxide dismutase 1aggregates. J. Cell Biol. 171, 75–85.

Mayer M. P. and Bukau B. (2005) Hsp70 chaperones: cellular functionsand molecular mechanism. Cell. Mol. Life Sci. 62, 670–684.

McClellan A. J. and Frydman J. (2001) Molecular chaperones and the artof recognizing a lost cause. Nat. Cell Biol. 3, E51–E53.

Muchowski P. J. and Wacker J. L. (2005) Modulation of neurodegen-eration by molecular chaperones. Nat. Rev. Neurosci. 6, 11–22.

Okado-Matsumoto A. and Fridovich I. (2002) Amyotrophic lateralsclerosis: a proposed mechanism Proc. Natl. Acad. Sci. USA 99,9010–9014.

Patel Y. J., Payne S., de Belleroche J. and Latchman D. S. (2005) Hsp27and Hsp70 administered in combination have a potent protectiveeffect against FALS-associated SOD1-mutant-induced cell death inmammalian neuronal cells. Brain Res. Mol. Brain Res. 134, 256–274.

Rosen D. R., Siddique T., Patterson D., Figlewicz D. A., Sapp P., HentatiA., Donaldson D., Goto J., O’Regan J. P. and Deng H. X. (1993)Mutations in Cu/Zn superoxide dismutase gene are associated withfamilial amyotrophic lateral sclerosis. Nature 362, 59–62.

Sharp P. S., Dick J. R. and Greensmith L. (2005) The effect of peripheralnerve injury on disease progression in the SOD1(G93A) mousemodel of amyotrophic lateral sclerosis. Neuroscience 130, 897–910.

Shaw P. J. (2005) Molecular and cellular pathways of neurodegenerationin motor neurone disease. J. Neurol. Neurosurg. Psychiatry 76,1046–1057.

Torok Z., Tsvetkova N. M., Balogh G. et al. (2003) Heat shock proteincoinducers with no effect on protein denaturation specificallymodulate the membrane lipid phase. Proc Natl Acad Sci USA 100,3131–3136.



Vigh L., Literati P. N., Horvath I. et al. (1997) Bimoclomol: a nontoxic,hydroxylamine derivative with stress protein-inducing activity andcytoprotective effects. Nat. Med. 3, 1150–1154.

Vigh L., Horvath I., Maresca B. and Harwood J. L. (2007) Can the stressprotein response be controlled by ‘membrane-lipid therapy’?Trends Biochem. Sci. 32, 357–363.

Vleminckx V., Van Damme P., Goffin K., Delye H., Van Den B. L.and Robberecht W. (2002) Upregulation of HSP27 in a trans-

genic model of ALS J. Neuropathol. Exp. Neurol. 61, 968–974.

Watanabe M., Dykes-Hoberg M., Culotta V. C., Price D. L., Wong P. C.and Rothstein J. D. (2001) Histological evidence of proteinaggregation in mutant SOD1 transgenic mice and in amyotrophiclateral sclerosis neural tissues. Neurobiol. Dis. 8, 933–941.

Yenari M. A. (2002) Heat shock proteins and neuroprotection. Adv. Exp.Med. Biol. 513, 281–299.



late stage treatment with arimoclomol delays disease progression and prevents protein aggregation in...

Documents