Thiol reducing compounds prevent human amylin-evokedcytotoxicityBarbara Konarkowska1,2, Jacqueline F Aitken1,2, Joerg Kistler1, Shaoping Zhang1,2 andGarth J S Cooper1,2,3

1 The School of Biological Sciences, University of Auckland, New Zealand

2 Centre for Research Excellence in Molecular Biodiscovery, Faculty of Science, University of Auckland, New Zealand

3 Department of Medicine, Faculty of Medical and Health Sciences, University of Auckland, New Zealand

Human amylin (hA) is a small protein that is cosecreted

with insulin from pancreatic islet b-cells upon stimula-

tion by glucose or other chemical signals [1,2]. It is the

main proteinaceous constituent of the amyloid deposits

commonly found in the islets of human subjects with

type-2 diabetes mellitus (T2Dm) [3,4], as well as in dia-

betic cats and primates [5,6], and has also been

referred to as islet amyloid polypeptide (IAPP) [5]. The

toxicity of hA towards cultured pancreatic b-cells is a

well-documented phenomenon [7–10]. Post mortem

examination of pancreatic tissue from T2Dm patients

has shown a correlation between the presence of amy-

loid deposits in pancreatic islets and b-cell loss [11–13].Evidence from longitudinal studies in spontaneously

diabetic primates has also indicated an inverse rela-

tionship between b-cell number and the extent of islet

amyloidosis [6]. Studies of hA transgenic animals

have shown that the occurrence of islet amyloid is

Keywords

amylin; b-cell apoptosis; diabetes; oxidative

stress; reactive oxygen species; N-acetyl-

L-cysteine

Correspondence

G. J. S. Cooper, School of Biological

Sciences, University of Auckland, Private

Bag 92–019, Auckland, New Zealand

Fax: +64 9 373 7045

Tel. +64 9 373 7599 ext. 87394

E-mail: g.cooper@auckland.ac.nz

(Received 7 June 2005, revised 1 August

2005, accepted 4 August 2005)

doi:10.1111/j.1742-4658.2005.04903.x

Human amylin (hA) is a small fibrillogenic protein that is the major con-

stituent of pancreatic islet amyloid, which occurs in most subjects with

type-2 diabetes mellitus (T2Dm). There is growing evidence that hA toxic-

ity towards islet b-cells is responsible for their gradual loss of function in

T2Dm. Preventing hA-mediated cytotoxicity has been proposed as a route

to halt the progression of this disease, although this has not yet been dem-

onstrated in vivo. The aim of our studies, in which we show that a small

number of hA-treated cells exhibit intracellular accumulation of reactive

oxygen species (ROS), was to evaluate the role of oxidative stress in the

mechanism of hA-mediated cytotoxicity. Here we report that catalase and

n-propyl gallate, antioxidants that are thought to act mainly as free radical

scavengers, afford RINm5F cells only limited protection against hA-medi-

ated toxicity. By contrast, the thiol antioxidants, N-acetyl-L-cysteine

(NAC), GSH and dithiothreitol, which not only react with ROS, but also

modulate the cellular redox potential by increasing intracellular levels of

GSH and ⁄or by acting as thiol reducing agents, afford almost complete

protection and inhibit the progression of hA-evoked apoptosis. We also

show that hA treatment is not associated with changes in intracellular

GSH levels and that inhibition of GSH biosynthesis has no effect on either

hA-mediated cytotoxicity or NAC-mediated protection. These results indi-

cate that, in addition to the induction of oxidative stress, hA appears to

mediate cytotoxicity through signalling pathways that are sensitive to the

actions of thiol antioxidants.

Abbreviations

BSO, buthionine-(S,R)-sulfoximine; carboxy-H2DCFDA, 5-(and-6)-carboxy-2¢,7¢-dichlorodihydrofluorescein diacetate; EthD-1, ethidium

homodimer-1; hA, human amylin; JNK, c-Jun NH2-terminal kinase; KRH, Hepes-balanced Krebs-Ringer bicarbonate buffer; NAC, N-acetyl-

L-cysteine; n-PG, n-propyl gallate; rA, rat amylin; RIN, rat insulinoma; ROS, reactive oxygen species; T2DM, type-2 diabetes mellitus.

FEBS Journal 272 (2005) 4949–4959 ª 2005 FEBS 4949

associated with the degeneration of pancreatic b-cells[14–16]. To summarise, there is strong support for the

hypothesis that hA-mediated toxicity towards pancre-

atic b-cells occurs in vivo and plays a role in the patho-

physiology of T2Dm.

Preventing hA-induced loss of pancreatic b-cells has

been proposed as a route to attenuate the gradual

decline in endogenous production of insulin in T2Dm

[17]. However, at this time knowledge of the mecha-

nisms by which hA evokes pancreatic b-cell death is

not complete.

Electron microscopic examination of hA-treated

RINm5F cells provided morphological evidence of cell

death through apoptosis [18]. In addition, DNA frag-

mentation, a hallmark of apoptosis, as well as the

up-regulation of the apoptosis-specific genes, p53 and

p21WAF1 ⁄CIP1, were detected in RINm5F cells follow-

ing treatment with hA [9]. Recent data point towards

the role of c-Jun, caspases 8 and 3, and the c-Jun

NH2-terminal kinase (JNK) signalling pathway in the

hA-induced apoptosis of RINm5F cells [19–21].

The signalling pathways found to be implicated in

the apoptosis of pancreatic b-cells are generally

believed to include phenomena such as (a) increases

in intracellular calcium [22,23] (b) overproduction of

reactive oxygen species (ROS) [24] (c) increased pro-

duction of ceramide [25], and (d) activation of mito-

gen-activated protein kinases [26]. Previous results

from our laboratory excluded the elevation of intracel-

lular calcium as a signalling mechanism in the

hA-mediated death of RINm5F cells [27]. Reports in

the literature implicating hA-mediated induction of

oxidative stress in pancreatic b-cells are contradictory

[7,8,28]. Lorenzo et al. [7] reported that hA toxicity

towards primary cultures of b-cells could not be pre-

vented with the antioxidant vitamin E. However,

Janciauskiene and Ahren [8,28] reported hA treatment

of RINm5F cells to be associated with elevated levels

of cellular lipid peroxidation, and increased activity of

glutathione reductase and membrane NADPH oxidase.

These authors interpreted their findings as indicative of

the occurrence of oxidative stress in hA-treated

RINm5F cells. In support of this proposition, they

correlated the sensitivity of RINm5F cells to hA-medi-

ated toxicity with sensitivity to hydrogen peroxide

(H2O2), while another pancreatic b-cell line, hamster

insulinoma HIT-T15, was found to be resistant to both

hA and H2O2 [8].

Furthermore, there is growing evidence for the

involvement of oxidative stress in the neurotoxicity of

amyloidogenic peptides thought to be responsible for

neuronal cell loss in various neurodegenerative dis-

orders [29]. In addition, it has been shown that the

cytotoxic effects of fibril-forming peptides, which are

not implicated in any amyloid diseases, are also associ-

ated with increased ROS production [30,31]. Indeed, it

has been proposed that oxidative stress represents a

common mechanism of toxicity shared by all amyloido-

genic peptides, including human amylin [29,30,32].

Much of this work has been performed on neuronal

cells, but little with b-cells.Here, our aim was to determine whether oxidative

stress might play a role in hA-mediated b-cell death.

We report that thiol antioxidants act on hA-induced

toxicity to prevent apoptosis. While we show that hA

treatment is associated with a certain degree of intra-

cellular oxidative activity, our results do not support

induction of oxidative stress as the only mechanism

through which hA mediates its cytotoxicity in these

cells. Instead our data indicate that toxicity of hA also

involves a signalling pathway that is regulated by the

redox status of intracellular thiol-containing com-

pounds.

Results

Antioxidants protect pancreatic b-cells against

hA-mediated toxicity

Several antioxidants were evaluated for their protect-

ive potential against hA-mediated toxicity of b-cells:catalase, n-propyl gallate (n-PG), N-acetyl-L-cysteine

(NAC), reduced glutathione (GSH) and dithiothreitol.

These compounds have diverse antioxidant actions

and their protective effects against oxidative damage

have been reported in other cell systems [33–37]. Our

selection of antioxidants included those that act enzy-

matically (catalase [38]) or as cofactors of endo-

genous antioxidant enzymes (NAC and GSH [39]),

and others that are thought to act nonenzymatically

(n-PG and dithiothreitol). Catalase acts in the extra-

cellular space, while the other antioxidants are cell

membrane permeable (n-PG [40], NAC and dithio-

threitol [41]; GSH ethyl ester [42]). In addition, these

antioxidants scavenge different types of ROS, e.g.

catalase decomposes H2O2 [38], n-PG scavenges

superoxide anion [34,43], NAC and GSH scavenge

hydroxyl radical [44]. They are also thought to pre-

vent different types of oxidative damage, e.g. n-PG

prevents lipid peroxidation [45], whereas NAC and

GSH prevent thiol oxidation [46]. Catalase and n-PG

are thought to act purely as free radical scavengers,

while the thiol antioxidants, NAC, GSH and dithio-

threitol, can also act by replenishing intracellular

stores of the endogenous antioxidant, GSH, or as

thiol-reducing agents [46,47].

Thiol antioxidants suppress hA cytotoxicity B. Konarkowska et al.

4950 FEBS Journal 272 (2005) 4949–4959 ª 2005 FEBS

The effect of these antioxidants on hA-mediated

cytotoxicity was determined by assessing the viability

of RINm5F cells following a 24 h exposure to peptide

in the presence or absence of the antioxidant. Antioxi-

dants were used at the highest possible concentrations

not affecting cell viability.

Catalase was used at a final concentration of

50 UÆmL)1, i.e. at twice the concentration which was

shown to protect cultured rat hippocampal neurons

against H2O2 toxicity [33]. As shown in Fig. 1A,

catalase provided RINm5F cells with a small but

nevertheless statistically significant protection against

hA-mediated toxicity (74.5 ± 2.2% cell survival, as

opposed to 66.2 ± 1.8% following treatment with hA

alone).

Similar to catalase, n-PG also afforded RINm5F

cells partial protection against hA-mediated toxicity

(Fig. 1B). The protective action of n-PG was dose-

dependent and statistically significant at 500 lm(79.3 ± 2.6% cell survival, as opposed to 58.6 ± 3.2%

following treatment with hA alone). At concentrations

higher than 500 lm, n-PG was toxic to RINm5F cells.

In contrast to catalase and n-PG, NAC provided

almost complete protection against hA toxicity

(95.8 ± 0.5% cell survival, as opposed to 50.3 ±

4.4% following treatment with hA alone) (Fig. 1C).

Similarly, GSH was very potent in protecting RINm5F

cells against hA (89.5 ± 1.9% cell survival, as opposed

to 50.3 ± 4.4% following treatment with hA alone)

(Fig. 1D). Both NAC- and GSH-mediated protection

was dose-dependent, with the maximal effect observed

at 20 mm and 10 mm, respectively (lower concentra-

tions not shown).

Dithiothreitol also protected RINm5F cells against

hA toxicity (Fig. 1E). Protection by 0.25 mm dithio-

threitol was statistically significant, but 0.5 mm dithio-

threitol afforded the highest level of protection

(95.8 ± 0.7% cell survival in the presence of dithio-

threitol, as opposed to 60.5 ± 1.6% following treat-

ment with hA alone).

Our results show that the thiol antioxidants, NAC,

GSH and dithiothreitol, are highly effective in protect-

ing RINm5F cells against hA-mediated toxicity,

whereas those antioxidants thought to act purely as

Fig. 1. Antioxidants protect RINm5F cells against hA-mediated toxicity Antioxidants. (A) catalase (B) n-PG (C) NAC (D) GSH and (E) dithiothre-

itol, were tested at the indicated concentrations for their ability to protect RINm5F cells against hA (10 lM) toxicity. Viability was assessed

by live ⁄ dead assay following 24-h coincubation. Survival corresponds to the number of live cells as a percentage of total cells. Values are

mean ± SEM of duplicates from 3 (A, B, E) or 4 (C, D) independent experiments. Data were analysed by one-way ANOVA with Tukey’s

Multiple Comparisons test. *P < 0.001, **P < 0.01 vs. vehicle; +P < 0.001, ++P < 0.01 vs. hA treatment in the absence of antioxidant; n.s.

vs. veh ¼ nonsignificant vs. vehicle control.

B. Konarkowska et al. Thiol antioxidants suppress hA cytotoxicity

FEBS Journal 272 (2005) 4949–4959 ª 2005 FEBS 4951

free radical scavengers, catalase and n-PG, afford only

limited protection.

hA-treated cells show evidence of intracellular

ROS accumulation

Given that a variety of antioxidants protected against

hA toxicity, we determined whether hA treatment

evokes increased intracellular ROS, using the oxi-

dation-sensitive probe, H2DCFDA, for fluorescence

microscopy-based detection. This method can be

applied to cell cultures with partially compromised

viability, as dead cells can be identified by staining

with EthD-1 and excluded from the analysis. Staining

of RINm5F cells with H2DCFDA was conducted at

two time-points during hA treatment: early (6 h), when

cell viability remained almost normal (3.2 ± 0.6% cell

death vs. 1.0 ± 0.2% following vehicle treatment),

and later (12 h), when a small, but significant propor-

tion of cells had died (10.5 ± 0.3% death compared

to 1.6 ± 0.1% after vehicle). These time points

enabled assessment of relative levels of intracellular

ROS in RINm5F cells just before and shortly after ini-

tiation of hA-induced death. At both time points,

numbers of viable DCF-positive cells in hA-treated

cultures were low (mean ± SEM of 3 independent

experiments; 0.5 ± 0.1% of DCF-positive ⁄EthD-1-

negative cells at 6 h, and 2.8 ± 0.7% at 12 h, as

opposed to 0% and 0.2 ± 0.2%, respectively, in vehi-

cle-treated cultures). Representative micrographs of

H2DCFDA-stained cells that were exposed to hA or

vehicle for 12 h are shown in Figs 2A,C, respectively.

Differences between hA treatment and vehicle control

were significant at both time points (P < 0.05), indica-

ting that a small proportion of hA-treated cells exhib-

ited intracellular ROS accumulation. While these

observations are consistent with the low level of pro-

tection afforded by free radical scavengers, they also

indicate that the high efficiency of thiol antioxidant-

mediated protection may be attributed to mechanisms

other than free radical scavenging.

Thiol antioxidants do not act through the

reduction of hA disulfide bond

NAC, GSH and dithiothreitol possess thiol-reducing

properties, and we considered the possibility that their

protective action against hA toxicity was mediated

through the reduction of amylin’s intramolecular disul-

fide bridge causing a structural alteration that could

reduce cytotoxic activity. However, recently published

data from our laboratory showed that the presence of

the disulfide bridge is not necessary for cytotoxicity.

The hA variant, hA(8–37), lacks the first 7 residues and

the intramolecular disulfide bridge between cysteine res-

idues 2 and 7, but is still toxic to RINm5F cells [19].

Here we show that thiol antioxidants protected

RINm5F cells against hA(8–37)-mediated toxicity.

NAC, GSH and dithiothreitol promoted the survival

of hA(8–37)-treated RINm5F cells to an extent similar

to that observed with the full-length peptide (Table 1).

These results indicate that the protective effects of

thiol antioxidants against hA toxicity are not mediated

via the reduction of the disulfide bond.

NAC and GSH prevent hA-induced apoptosis

We further investigated whether thiol antioxidants act

directly on cells to prevent hA-induced apoptosis, by

A B

C D

Fig. 2. hA-treated RINm5F cells show intracellular ROS accumula-

tion. Intracellular ROS in RINm5F cells were assessed by staining

with H2DCFDA after 12-h treatment with (A, B) hA (15 lM) or (C,

D) vehicle. Representative groups of cells are shown. Each set of

micrographs contains the following images of the same field: (A, C)

a fluorescein-optics image showing DCF-positive (fluorescent

green) cells, combined with EthD-1-stained (fluorescent red) areas

from a rhodamine-optics image and (B, D) a phase-contrast image

showing all cells in the field. Only those DCF-positive cells (contain-

ing ROS) that were also EthD-1-negative (viable) were scored as

undergoing oxidative stress. White arrows in image (A) indicate

DCF-positive ⁄ EthD-1-negative cells. Scale bar ¼ 50 lm.

Thiol antioxidants suppress hA cytotoxicity B. Konarkowska et al.

4952 FEBS Journal 272 (2005) 4949–4959 ª 2005 FEBS

assessing the effects of NAC and GSH on the levels

of caspase-3 activity in hA-treated RINm5F cells. Our

results show that both NAC and GSH suppress

hA-induced increases in caspase-3 activity of RINm5F

cells to levels similar to those in vehicle-treated cells,

indicating that these thiol antioxidants inhibit progres-

sion of hA-evoked apoptosis (Fig. 3A,B).

In addition to measurements of caspase-3 activity,

we assessed the effect of NAC and GSH on the extent

of the genomic DNA fragmentation that accompanies

hA-induced apoptosis. Both compounds greatly

diminished relative DNA fragmentation in hA-treated

b-cells. A representative experiment illustrating the

effect of GSH is shown (Fig. 3C). These results also

support the conclusion that thiol antioxidants inhibit

hA-induced apoptosis in b-cells.

Prevention of apoptosis by thiol antioxidants is

not mediated via inhibition of cell proliferation

Thiol antioxidants were reported to inhibit prolifer-

ation of cultured cells [48–50]. On the other hand, sen-

sitivity of pancreatic b-cells to hA-mediated toxicity

reportedly correlated with cell proliferation [9,51]. We

investigated the possibility that the protective actions

of the thiol antioxidants, NAC, GSH and dithiothrei-

tol, on hA-treated RINm5F cells are mediated through

their inhibitory effects on cell growth. The results of

total cell counts following 26-h incubation indicate

that thiol antioxidants had no significant effect on the

growth of RINm5F cells (mean ± SEM of duplicates

from 3 independent experiments; NAC: 11.3 ± 8.7%

growth inhibition vs. control, which was set at

100%; GSH: 10.8 ± 5.8%; dithiothreitol: 4.9 ± 9.3%;

significance was determined by one-way anova with

Dunnett’s Multiple Comparison test). Therefore, it is

unlikely that the protection afforded by thiol antioxi-

Table 1. Thiol antioxidants protect RINm5F cells against hA(8–37)

toxicity. Thiol antioxidants: NAC (20 mM), GSH (10 mM), and dithio-

threitol (0.5 mM) were tested for their ability to protect RINm5F

cells against toxicity of hA(8–37) (25 lM). Cell viability was

assessed by live ⁄ dead assay following 24-h coincubation. Cell survi-

val corresponds to the number of live cells as a percentage of total

cells. Values are mean ± SEM of duplicates from three (NAC, GSH)

or two (dithiothreitol) independent experiments. Data were ana-

lysed by one-way ANOVA with Tukey’s Multiple Comparisons test.

NS ¼ nonsignificant vs. vehicle control.

Cell survival

[live cells as % of total cells]

Vehicle hA(8–37) hA(8–37) + antioxidant

99.0 ± 0.3 59.2 ± 3.0* NAC; 87.5 ± 1.5**†

99.0 ± 0.2 62.8 ± 4.6* GSH; 88.0 ± 3.2NS†

97.5 ± 0.7 63.8 ± 4.2* dithiothreitol; 93.3 ± 0.5NS†

*P < 0.001; **P < 0.01 vs. vehicle; †P < 0.001 vs. hA treatment in

the absence of antioxidant.

A

B

C 1 2 3 4 5

100 bp

600 bp

Fig. 3. Thiol antioxidants attenuate hA-induced increases in

caspase-3 activity and DNA fragmentation in RINm5F cells.

Caspase-3 activity was measured in extracts from RINm5F cells

following 15-h treatment with hA (12 lM) in the presence or

absence of (A) NAC (20 mM) or (B) GSH (10 mM). Data for all the

samples (including vehicle control) from each experiment were

transformed into units relative to the average vehicle control value,

which was set at one. All values are mean ± SEM of duplicates

from three independent experiments. Data were analysed by one-

way ANOVA with Tukey’s Multiple Comparisons test. *P < 0.001 vs.

vehicle; +P < 0.001 vs. hA-treatment in the absence of antioxidant;

n.s. vs. veh ¼ nonsignificant vs. vehicle control. (C) Agarose gel

electrophoresis was used to verify the integrity of genomic DNA

from cells treated with hA (12 lM) in the presence or absence of

GSH (10 mM) for 24 h. Lane 1: 100 bp DNA ladder; lane 2: vehicle;

lane 3: hA; lane 4: GSH; lane 5: hA + GSH.

B. Konarkowska et al. Thiol antioxidants suppress hA cytotoxicity

FEBS Journal 272 (2005) 4949–4959 ª 2005 FEBS 4953

dants towards hA-treated RINm5F cells is mediated

through inhibition of proliferation.

Treatment with hA is not associated with

changes in intracellular GSH levels

In addition to direct scavenging of ROS, thiol anti-

oxidants augment intracellular levels of the endo-

genous antioxidant GSH and we therefore

determined whether hA treatment is associated with

changes in GSH levels. The complication of GSH

leakage from dead cells was avoided by assessing

intracellular GSH levels at a time point (6 h), when

the viability of treated cells remained largely uncom-

promised. hA treatment did not affect GSH levels in

RINm5F cells (mean ± SEM of duplicates from

three independent experiments; hA: 116 ± 11% vs.

vehicle: 100 ± 3%). The values shown are percent-

ages relative to the average vehicle control value,

which was set at 100%.

We also assessed the ability of NAC to protect

RINm5F cells against hA toxicity in the presence of a

specific inhibitor of GSH biosynthesis, BSO, which has

been reported to deplete intracellular GSH stores in

other cell systems [50,52]. BSO inhibits c-glutamylcy-

steine synthetase, the rate-limiting enzyme of GSH

synthesis [53]. The extent of the BSO-induced

GSH depletion depends on the cell type-specific rate of

GSH turnover [39]. Intracellular GSH levels in

RINm5F cells were decreased by nearly 25% (mean ±

SEM of duplicates from 5 independent experiments;

BSO: 77 ± 3% vs. vehicle: 100 ± 2%; *P < 0.0001)

following 17-h treatment with 5 mm BSO. The effect

of BSO on these cells was maximal under these condi-

tions, as a higher dose of the inhibitor (10 mm) did

not cause any further decrease in GSH levels (data not

shown). Importantly, however, the lowering of GSH

levels by BSO (5 or 10 mm) was not associated with

significant loss of cell viability, indicating that GSH

depletion in itself does not lead to apoptosis in

RINm5F b-cells.As shown in Fig. 4, the presence of BSO did not

interfere with the NAC-mediated protection of

RINm5F cells (95.8 ± 0.8% cell survival following

exposure to hA in the presence of NAC, and

96.0 ± 0.6% following exposure to hA in the presence

of NAC and BSO, as opposed to 49.3 ± 7.3% follow-

ing exposure to hA alone). This observation also indi-

cates that the protective actions of NAC occur

independently of GSH biosynthesis. Moreover, cell

survival following exposure to hA in the presence

of BSO (39.3 ± 5.7%) was not statistically different

from that following treatment with hA alone

(49.3 ± 7.3%). Hence, the BSO-induced GSH deple-

tion did not increase the sensitivity of RINm5F cells

to hA-mediated toxicity.

Discussion

These investigations focussed on whether hA evokes

pancreatic b-cell death through oxidative stress. To

this end, we tested several antioxidants for the ability

to protect RINm5F b-cells against hA-mediated apop-

tosis. RINm5F is a continuous cell-line that retains

many features indicative of pancreatic islet b-cell lin-

eage [19]. Our observations that catalase, n-PG, NAC,

GSH and dithiothreitol protect RINm5F cells from

hA-evoked cell death indicate that hA treatment of

RINm5F cells is associated with increased levels

of intracellular pro-oxidant activity. Protection by

catalase indicates that hA treatment leads to, or is

accompanied by, increased H2O2 production, while

n-PG-mediated protection could occur through inhibi-

tion of lipid peroxidation. Our results are therefore in

agreement with the findings of Janciauskiene and

Ahren [8,28] who reported some evidence of oxidative

stress in hA-treated RINm5F cells. However, we show

that catalase and n-PG, i.e. antioxidants that are

thought to act purely as free radical scavengers, afford

Fig. 4. The hA cytotoxicity and NAC-mediated protection are not

affected by the inhibition of intracellular GSH biosynthesis. The pro-

tective actions of NAC (20 mM) towards RINm5F cells treated with

hA (10 lM) were assessed in the presence or absence of BSO

(5 mM). Cell viability was determined by live ⁄ dead assay at the end

of 24-h coincubations. Cell survival corresponds to the number of

live cells expressed as a percentage of total cells. Values are

mean ± SEM of duplicates from 2 independent experiments. Data

were analysed by one-way ANOVA with Tukey’s Multiple Compari-

sons test. *P < 0.001 vs. vehicle; +P < 0.001 vs. hA treatment

without NAC; n.s. ¼ nonsignificant.

Thiol antioxidants suppress hA cytotoxicity B. Konarkowska et al.

4954 FEBS Journal 272 (2005) 4949–4959 ª 2005 FEBS

only partial protection against hA-mediated toxicity.

Moreover, intracellular ROS accumulation is observed

only in a small number of hA-treated RINm5F cells,

although it is possible that hA-induced increase in

ROS levels is either transient or occurs shortly before

cell death. The suboptimal efficiency with which cat-

alase and n-PG suppress hA-mediated cytotoxicity, is

consistent with the findings of Lorenzo et al. [7], who

did not observe any protection of hA-treated human

and mouse islet b-cells by another free radical scaven-

ger, vitamin E. While our results provide evidence in

support of hA-induced oxidative stress, they also indi-

cate that it is not the only mechanism through which

hA evokes b-cell death. This conclusion is similar to

that drawn by other researchers who showed that the

occurrence of oxidative stress in cells exposed to other

amyloidogenic peptides such as b-amyloid or HypF-N

did not play a major role in the induction of cell death

[31,54,55].

In contrast to catalase and n-PG, the thiol antioxi-

dants, NAC, GSH and dithiothreitol, afford RINm5F

cells almost complete protection against the toxic hA

stimulus. Furthermore, these compounds hinder the

progression of hA-induced apoptosis, as indicated by

their inhibitory effects on hA-induced increase in ca-

spase-3 activity and internucleosomal DNA fragmenta-

tion.

In addition to their ROS-scavenging capabilities,

thiol antioxidants can also modulate the redox poten-

tial of cells by augmenting intracellular GSH levels

and ⁄or by acting as thiol reducing agents [46,47].

However, our experiments indicate that these com-

pounds do not act via the reduction of the hA disulfide

bridge, or through effects on cell proliferation. More-

over, intracellular GSH levels in hA-treated cells were

unchanged. Our observations that the BSO-induced

depletion of GSH failed to affect viability of RINm5F

cells or to increase the severity of hA-mediated toxicity

do not support the proposition that changes in GSH

levels play a decisive role in hA-evoked cell death, and

we conclude that the protective actions of NAC are

independent of GSH biosynthesis. Rather, it appears

that the protective actions of thiol antioxidants may be

mediated through effects on the redox status of cellular

thiols other than GSH.

We also investigated the possibility that thiol

antioxidants exerted their protective action towards

hA-treated RINm5F cells by affecting aggregate for-

mation. However, the combined results of our thiofla-

vin-T fluorescence and circular dichroism experiments

showed that there was only a small inhibitory effect of

thiol antioxidants on peptide aggregation, thus making

this proposal unlikely (data not shown).

In summary, we conclude that ROS are not the only

mediators of hA cytotoxicity. Instead hA-induced cell

death can be prevented by agents that affect the redu-

cing potential of the cell. An important component of

the hA-activated cell death signalling pathway is thus

likely to be regulated via its thiol redox status. There

is growing evidence in the literature that thiol-reduc-

tants can alter the expression and activity of various

cell death-regulating transcription factors [56–59].

Redox-regulated proteins include JNK, a pro-apop-

totic factor, and extracellular signal-regulated kinase

(ERK), which reportedly has antiapoptotic activity

[60]. Both these proteins have been recently implicated

in hA-mediated apoptosis [19,21]. NAC itself, has been

reported to activate ERK [61] and inhibit JNK activity

[58]. Perhaps NAC with its dual ability to activate

ERK and suppress JNK activity may influence the sig-

nalling pathway leading to hA-mediated apoptosis.

Thus, inclusion of thiol antioxidants in the therapy of

T2Dm patients might be beneficial in the prevention of

hA-mediated loss of pancreatic islet b-cells.

Experimental procedures

Peptides and chemicals

hA and hA(8–37) were lyophilized preparations (Bachem

California, Torrance, CA, USA; hA, #0542559 and 0538994;

hA(8–37), #0523544). Aqueous stock solutions were pre-

pared immediately before use, by dissolution in water (18

MW; MilliQ, Millipore) to a final concentration of 500 lm.Catalase (from bovine liver), n-propyl gallate (n-PG), N-ace-

tyl-l-cysteine (NAC), reduced glutathione ethyl ester (GSH)

and buthionine-(S,R)-sulfoximine (BSO) were from Sigma

(St Louis, MO, USA), and dithiothreitol was from Biochemi-

ca (AppliChem GmbH, Germany). Catalase was dissolved

(final concentration 55 UÆmL)1; one unit decomposes

1.0 lmol of H2O2 per min at pH 7.0 at 25 �C) in RPMI 1640

and supplemented with fetal bovine serum [FBS, 10% (v ⁄ v)].Stocks of n-PG, NAC, GSH, dithiothreitol and BSO in water

were prepared (n-PG, 10 mm; NAC, 1000 mm; GSH,

500 mm; dithiothreitol, 100 mm; BSO, 225 mm) and aliquots

used to prepare antioxidant-containing media. Proteinase K

was from Life Technologies (Auckland, New Zealand), and

5-(and-6)-carboxy-2¢,7¢-dichlorodihydrofluorescein diacetate

(carboxy-H2DCFDA) from Molecular Probes (Sunnyvale,

CA, USA). Caspase-3 substrate Ac-DEVD-AFC (Bio-Rad,

Hercules, CA, USA) was dissolved in DMSO. General cell

culture materials were from Invitrogen (Carlsbad, CA,

USA). All other chemicals were of analytical grade or better.

Hepes-balanced Krebs-Ringer bicarbonate buffer (KRH)

was comprised of 119 mm NaCl, 4.74 mm KCl, 2.54 mm

CaCl2, 1.19 mm MgCl2, 1.19 mm KH2PO4, 25 mm NaHCO3,

and 10 mm Hepes, pH 7.4.

B. Konarkowska et al. Thiol antioxidants suppress hA cytotoxicity

FEBS Journal 272 (2005) 4949–4959 ª 2005 FEBS 4955

RINm5F cell culture

RINm5F cells were gifted by H K Oie (NIH, Bethesda,

MD) and maintained as previously described [9,27]. Here,

cells between passages 28 and 39 were grown in 24- or

6-well cell dishes (Greiner, Germany). The mechanism of

hA-evoked islet b-cell apoptosis in RINm5F cells closely

resembles that in human b-cells [19].

Cytotoxicity evoked by hA

Cells were cultured (48 h, 24-well dishes, initially 1.5 · 105

per well); washed (NaCl ⁄Pi, once); then preincubated (fresh

medium with antioxidant or relevant vehicle) for 30 min

(catalase) or 3–4 h (n-PG, NAC, GSH or dithiothreitol). In

experiments to assess the effect of BSO on hA toxicity or

the protective potential of NAC, treatment with 5 mm BSO

(or relevant vehicle) was initiated 2 h before incubation with

hA or preincubation with NAC, respectively. Next, aliquots

of either hA or hA(8–37) stock solution (or vehicle) were

added to final concentrations as follows: hA, 10 lm;hA(8–37), 25 lm; catalase, 50 UÆmL)1; n-PG, 100, 250 and

500 lm; NAC, 20 mm; GSH, 10 mm; and dithiothreitol, 0.1,

0.25 and 0.5 mm. Following hA treatment (24 h), viability

was determined [calcein-AM ⁄ ethidium homodimer-1 (EthD-

1)], as described previously [17,18].

Detection of intracellular ROS accumulation

Intracellular ROS were detected by carboxy-H2DCFDA

staining wherein fresh stock solutions (10 mm in DMSO)

were used for each experiment. RINm5F cells were plated

(24-well dishes, 1.2 · 105 cells per well), grown (48 h), then

treated [6 or 12 h, hA (15 lm) or vehicle]. Three independ-

ent experiments were performed, each of which included

two wells per condition: cells in one well were stained with

carboxy-H2DCFDA and the other well treated with DMSO.

After hA treatment, cells were rinsed (KRH), then serum-

free phenol red-free RPMI 1640 and carboxy-H2DCFDA

(100 lm final) or an equivalent amount of DMSO were

added. Cells were incubated (1 h, 37 �C, dark), rinsed

twice (serum-free phenol red-free RPMI 1640), visualized

[2¢,7¢-dichlorofluorescein, (DCF), Zeiss Axiovert S100] (Carl

Zeiss International, Oberkochen, Germany), and images

digitized (25 · , Zeiss AxioCam). In each experiment, six

fields from each well were recorded (with both fluorescein

optics and phase-contrast). EthD-1 (final concentration

2 lm) was added and each field re-photographed (rhodam-

ine optics for dead cells). DCF-positive cells (i.e. exhibiting

intracellular ROS accumulation) that were also EthD-1-neg-

ative (i.e. viable), were scored as cells undergoing oxidative

stress; numbers were expressed as percentages of total cells.

No fluorescence was observed in control or hA-treated cells

without carboxy-H2DCFDA-staining.

Caspase-3 activity

Caspase-3 activity in cell extracts was quantified by measur-

ing cleavage of a synthetic oligopeptide substrate for

caspase-3, Ac-DEVD-AFC, according to established proto-

cols [19] with the following modifications: (a) cells were

grown (plating density 1.6 · 106 cells per well, 6-well plates)

for 48 h before treatment; (b) following hA-treatment

(15 h, 12 lm), cells were lysed in buffer (100 lL per well,

or 85 lL per well when 2–3 wells were combined), then

combined with lysates (20 lL) of floating cells; (c) following

freeze-thaw cycles, samples were incubated (ice, 15 min)

and centrifuged; (d) 40–60 lL of extracts were added to

reaction mixtures [100 lL final volume; Hepes, 15 mm,

pH 7.4; EDTA, 3 mm; CHAPS, 0.15% (w ⁄ v); dithiothrei-

tol, 9.5 mm; DMSO, 3% (v ⁄ v); Ac-DEVD-AFC, 55 lm];and (e) mixtures were incubated (2 h, 37 �C) before fluores-

cence measurement (emission, 540 nm; excitation, 400 nm)

of duplicates for each condition. Data were trans-

formed into units relative to the average control value, set

at one.

Genomic DNA integrity

RINm5F cells were cultured (48 h; plated at 1.6 · 106 cells

per well in 6-well plates); preincubated [4 h; GSH (10 mm)

or vehicle]; then treated with hA (12 lm) or vehicle

(± GSH, 24 h). Cells were then harvested by trypsiniza-

tion, resuspended (fresh lysis buffer, 10 mm Tris.HCl,

pH 8.0, 10 mm NaCl, 10 mm EDTA, 0.5% SDS,

3 mgÆmL)1 proteinase K), and incubated (24 h, 55 �C).Genomic DNA was extracted (buffer-saturated phenol,

pH 8.0 – chloroform) in three steps: (a) phenol; (b)

phenol ⁄ chloroform ⁄ isoamyl alcohol (25 : 24 : 1; v ⁄ v ⁄ v);and (c) chloroform ⁄ isoamyl alcohol (24 : 1; v ⁄ v). DNA was

precipitated (0.3 mm sodium acetate in ethanol, on solid

CO2, 30 min) or at )20 �C for 24 h. Precipitated DNA was

washed [70% (v ⁄ v) aqueous ethanol], air-dried and resus-

pended (water). Following RNase A treatment (20–45 min,

37 �C), DNA was re-purified using the same procedures for

extraction, precipitation and purification as described

above, except that the first extraction step was carried out

with phenol ⁄ chloroform ⁄ isoamyl alcohol (25 : 24 : 1;

v ⁄ v ⁄ v), and then resuspended [10 mm Tris ⁄HCl

(pH 8.0) ⁄ 1 mm EDTA]. DNA samples (15 lg) were electro-

phoresed (1.5% agarose), stained (ethidium bromide) and

UV-visualized with a 100 bp DNA ladder (Invitrogen).

RINm5F cell growth

RINm5F cells were grown (48 h; 24-well plates initially at

1.5 · 105 cells per well), then treated with antioxidants

(20 mm NAC or 10 mm GSH or 0.5 mm dithiothreitol)

for 26 h. Cells were harvested (trypsinization), combined

Thiol antioxidants suppress hA cytotoxicity B. Konarkowska et al.

4956 FEBS Journal 272 (2005) 4949–4959 ª 2005 FEBS

with the floating cells, centrifuged and counted (haemo-

cytometer), and analysed in duplicate. Counts were

expressed as percentages of control (i.e. vehicle-treated)

numbers.

Changes in intracellular GSH

Intracellular GSH was determined (Glutathione Apoptosis

Detection Kit; Oncogene, San Diego, CA, USA) according

to the manufacturer’s instructions, with minor modifica-

tions. Cells were cultured (48 h; plating density 1.2 · 105

per well in 24-well dishes; or at 6.0 · 105 per well, 6-well

dishes). Experiments were performed in duplicate or tripli-

cate and cells from either one (6-well plate) or two wells (24-

well plate) were used for sample preparation. Following hA-

(15 lm) or vehicle-treatment (6 h), cells were trypsinized

and resuspended (lysis buffer, 60 lL per well in 6-well

plates, or 25 lL per two wells in 24-well plates). Following

debris removal (centrifugation), reaction mixtures were pre-

pared: extract (50 lL) from a 6-well dish was mixed with

monochlorobimane solution (1 lL) and GST reagent

(1 lL); or extract (16 lL) from a 24-well dish, monochloro-

bimane solution (0.5 lL), and glutathione-S-transferase rea-

gent (0.5 lL). Fluorescence values for ‘treated’ and ‘vehicle

control’ samples were normalized to protein concentrations.

Results are expressed as percentages relative to the average

vehicle control value, which was set at 100%.

Statistical analysis

Statistical analysis was conducted using either unpaired

two-tailed t-test (to compare two conditions), or one-way

anova with Dunnett’s or Tukey’s Multiple Comparisons

test (to analyse more than two conditions). Statistical signi-

ficance was determined at P < 0.05.

Acknowledgements

This research was supported by Grants from the Foun-

dation for Research, Science and Technology, New

Zealand, Endocore Research Trust, Lottery Health

(NZ), Maurice & Phyllis Paykel Trust, and the Univer-

sity of Auckland Graduate Research Fund. We thank

J.Z. Bai for technical assistance.

References

1 Cooper GJ (1994) Amylin compared with calcitonin

gene-related peptide: structure, biology, and relevance

to metabolic disease. Endocr Rev 15, 163–201.

2 Moore CX & Cooper GJ (1991) Co-secretion of amylin

and insulin from cultured islet b-cells: modulation by

nutrient secretagogues, islet hormones and hypoglycemic

agents. Biochem Biophys Res Comm 179, 1–9.

3 Bell ET (1959) Hyalinization of the islet of Langerhans

in non-diabetic individuals. Am J Pathol 35, 801–805.

4 Cooper GJ, Willis AC, Clark A, Turner RC, Sim RB

& Reid KB (1987) Purification and characterization of

a peptide from amyloid-rich pancreases of type 2

diabetic patients. Proc Natl Acad Sci USA 84, 8628–

8632.

5 Lukinius A, Wilander E, Westermark GT, Engstrom U

& Westermark P (1989) Colocalization of islet amyloid

polypeptide and insulin in the B cell secretory granules

of the human pancreatic islets. Diabetologia 32, 240–

244.

6 de Koning EJ, Bodkin NL, Hansen BC & Clark A

(1993) Diabetes mellitus in Macaca mulatta monkeys is

characterised by islet amyloidosis and reduction in beta-

cell population. Diabetologia 36, 378–384.

7 Lorenzo A, Razzaboni B, Weir GC & Yankner BA

(1994) Pancreatic islet cell toxicity of amylin associated

with type-2 diabetes mellitus. Nature 368, 756–760.

8 Janciauskiene S & Ahren B (1998) Different sensitivity

to the cytotoxic action of IAPP fibrils in two insulin-

producing cell lines, HIT-T15 and RINm5F cells.

Biochem Biophys Res Comm 251, 888–893.

9 Zhang S, Liu J, Saafi EL & Cooper GJ (1999) Induction

of apoptosis by human amylin in RINm5F islet b-cellsis associated with enhanced expression of p53 and

p21WAF1 ⁄ CIP1. FEBS Lett 455, 315–320.

10 Janson J, Ashley RH, Harrison D, McIntyre S & Butler

PC (1999) The mechanism of islet amyloid polypeptide

toxicity is membrane disruption by intermediate-sized

toxic amyloid particles. Diabetes 48, 491–498.

11 Westermark P (1973) Fine structure of islets of Langer-

hans in insular amyloidosis. Virchows Arch-a: Pathol-

Pathologische Anat 359, 1–18.

12 Westermark P & Wilander E (1978) The influence of

amyloid deposits on the islet volume in maturity onset

diabetes mellitus. Diabetologia 15, 417–421.

13 Garg A, Chandalia M & Vuitch F (1996) Severe islet

amyloidosis in congenital generalized lipodystrophy.

Diabetes Care 19, 28–31.

14 Wang F, Hull RL, Vidal J, Cnop M & Kahn SE (2001)

Islet amyloid develops diffusely throughout the pancreas

before becoming severe and replacing endocrine cells.

Diabetes 50, 2514–2520.

15 Couce M, Kane LA, O’Brien TD, Charlesworth J, Soel-

ler W, McNeish J, Kreutter D, Roche P & Butler PC

(1996) Treatment with growth hormone and dexametha-

sone in mice transgenic for human islet amyloid poly-

peptide causes islet amyloidosis and b-cell dysfunction.Diabetes 45, 1094–1101.

16 Soeller WC, Janson J, Hart SE, Parker JC, Carty MD,

Stevenson RW, Kreutter DK & Butler PC (1998) Islet

amyloid-associated diabetes in obese A(vy) ⁄ a mice

expressing human islet amyloid polypeptide. Diabetes

47, 743–750.

B. Konarkowska et al. Thiol antioxidants suppress hA cytotoxicity

FEBS Journal 272 (2005) 4949–4959 ª 2005 FEBS 4957

17 Aitken JF, Loomes KM, Konarkowska B & Cooper GJ

(2003) Suppression by polycyclic compounds of the

conversion of human amylin into insoluble amyloid.

Biochem J 374, 779–784.

18 Saafi EL, Konarkowska B, Zhang S, Kistler J & Cooper

GJ (2001) Ultrastructural evidence that apoptosis is the

mechanism by which human amylin evokes death in

RINm5F pancreatic islet b-cells. Cell Biol Internat 25,339–350.

19 Zhang S, Liu J, Dragunow M & Cooper GJ (2003)

Fibrillogenic amylin evokes islet b-cell apoptosisthrough linked activation of a caspase cascade and

JNK1. J Biol Chem 278, 52810–52819.

20 Zhang S, Liu J, MacGibbon G, Dragunow M &

Cooper GJ (2002) Increased expression and activation

of c-Jun contributes to human amylin-induced apop-

tosis in pancreatic islet b-cells. J Mol Biol 324, 271–

285.

21 Rumora L, Hadzija M, Barisic K, Maysinger D & Gru-

biic TZ (2002) Amylin-induced cytotoxicity is associated

with activation of caspase-3 and MAP kinases. Biol

Chem 383, 1751–1758.

22 Efanova IB, Zaitsev SV, Zhivotovsky B, Kohler M,

Efendic S, Orrenius S & Berggren PO (1998) Glucose

and tolbutamide induce apoptosis in pancreatic b-cells.A process dependent on intracellular Ca2+ concentra-

tion. J Biol Chem 273, 33501–33507.

23 Wang L, Bhattacharjee A, Zuo Z, Hu F, Honkanen

RE, Berggren PO & Li M (1999) A low voltage-acti-

vated Ca2+ current mediates cytokine-induced pancre-

atic b-cell death. Endocrinology 140, 1200–1204.

24 Dypbukt JM, Ankarcrona M, Burkitt M, Sjoholm A,

Strom K, Orrenius S & Nicotera P (1994) Different

prooxidant levels stimulate growth, trigger apoptosis,

or produce necrosis of insulin-secreting RINm5F cells.

The role of intracellular polyamines. J Biol Chem 269,

30553–30560.

25 Shimabukuro M, Zhou YT, Levi M & Unger RH

(1998) Fatty acid-induced b cell apoptosis: a link

between obesity and diabetes. Proc Natl Acad Sci USA

95, 2498–2502.

26 Mandrup-Poulsen T (2001) b-cell apoptosis: stimuli and

signaling. Diabetes 50 (Suppl. 1), S58–S63.

27 Bai JZ, Saafi EL, Zhang S & Cooper GJ (1999) Role of

Ca2+ in apoptosis evoked by human amylin in pancre-

atic islet b-cells. Biochem J 343, 53–61.

28 Janciauskiene S & Ahren B (2000) Fibrillar islet amy-

loid polypeptide differentially affects oxidative mechan-

isms and lipoprotein uptake in correlation with

cytotoxicity in two insulin-producing cell lines. Biochem

Biophys Res Comm 267, 619–625.

29 Stefani M & Dobson CM (2003) Protein aggregation

and aggregate toxicity: new insights into protein folding,

misfolding diseases and biological evolution. J Mol Med

81, 678–699.

30 Schubert D, Behl C, Lesley R, Brack A, Dargusch R,

Sagara Y & Kimura H (1995) Amyloid peptides are

toxic via a common oxidative mechanism. Proc Natl

Acad Sci USA 92, 1989–1993.

31 Bucciantini M, Calloni G, Chiti F, Formigli L, Nosi D,

Dobson CM & Stefani M (2004) Prefibrillar amyloid

protein aggregates share common features of cytotoxi-

city. J Biol Chem 279, 31374–31382.

32 Mattson MP & Goodman Y (1995) Different amyloido-

genic peptides share a similar mechanism of neurotoxi-

city involving reactive oxygen species and calcium.

Brain Res 676, 219–224.

33 Lockhart BP, Benicourt C, Junien JL & Privat A (1994)

Inhibitors of free radical formation fail to attenuate

direct b-amyloid 25)35 peptide-mediated neurotoxicity in

rat hippocampal cultures. J Neurosci Res 39, 494–505.

34 Reddan JR, Giblin FJ, Sevilla M, Padgaonkar V,

Dziedzic DC, Leverenz VR, Misra IC, Chang JS &

Pena JT (2003) Propyl gallate is a superoxide dismutase

mimic and protects cultured lens epithelial cells from

H2O2 insult. Exp Eye Res 76, 49–59.

35 Wu TW, Fung KP, Zeng LH, Wu J & Nakamura H

(1994) Propyl gallate as a hepatoprotector in vitro and

in vivo. Biochem Pharmacol 48, 419–422.

36 Liu L, Trimarchi JR & Keefe DL (1999) Thiol oxida-

tion-induced embryonic cell death in mice is prevented

by the antioxidant dithiothreitol. Biol Reprod 61, 1162–

1169.

37 Starke PE & Farber JL (1985) Endogenous defenses

against the cytotoxicity of hydrogen peroxide in cul-

tured rat hepatocytes. J Biol Chem 260, 86–92.

38 Halliwell B & Gutteridge JMC (1989) Free Radicals in

Biology and Medicine. Oxford University Press, Oxford,

UK.

39 Meister A & Anderson ME (1983) Glutathione. Annu

Rev Biochem 52, 711–760.

40 Takami M, Preston SL, Toyloy VA & Behrman HR

(1999) Antioxidants reversibly inhibit the spontaneous

resumption of meiosis. Am J Physiol 276, E684–E688.

41 Newton GL, Aguilera JA, Kim T, Ward JF & Fahey

RC (1996) Transport of aminothiol radioprotectors into

mammalian cells: passive diffusion versus mediated

uptake. Radiat Res 146, 206–215.

42 Anderson ME, Powrie F, Puri RN & Meister A (1985)

Glutathione monoethyl ester: preparation, uptake by

tissues, and conversion to glutathione. Arch Biochem

Biophys 239, 538–548.

43 Deeble DJ, Parsons BJ & Phillips GO (1987) Evidence

for the addition of the superoxide anion to the anti-

oxidant n-propyl gallate in aqueous solution. Free Radic

Res Commun 2, 351–358.

44 Aruoma OI, Halliwell B, Hoey BM & Butler J (1989) The

antioxidant action of N-acetylcysteine: its reaction with

hydrogen peroxide, hydroxyl radical, superoxide, and

hypochlorous acid. Free Radic Biol Med 6, 593–597.

Thiol antioxidants suppress hA cytotoxicity B. Konarkowska et al.

4958 FEBS Journal 272 (2005) 4949–4959 ª 2005 FEBS

45 Zhu J, Johnson WJ, Sevilla CL, Herrington JW &

Sevilla MD (1990) An electron spin resonance study of

the reactions of lipid peroxyl radicals with antioxidants.

J Phys Chem 94, 7185–7190.

46 Cotgreave IA (1997) N-acetylcysteine: pharmacological

considerations and experimental and clinical applica-

tions. Adv Pharmacol 38, 205–227.

47 Meister A (1988) Glutathione metabolism and its select-

ive modification. J Biol Chem 263, 17205–17208.

48 Burdon RH, Alliangana D & Gill V (1994) Endogen-

ously generated active oxygen species and cellular glu-

tathione levels in relation to BHK-21 cell proliferation.

Free Radic Res 21, 121–133.

49 Ferrari G, Yan CY & Greene LA (1995) N-acetylcys-

teine (d- and l-stereoisomers) prevents apoptotic death

of neuronal cells. J Neurosci 15, 2857–2866.

50 Yan CY, Ferrari G & Greene LA (1995) N-acetylcys-

teine-promoted survival of PC12 cells is glutathione-

independent but transcription-dependent. J Biol Chem

270, 26827–26832.

51 Ritzel RA & Butler PC (2003) Replication increases

b-cell vulnerability to human islet amyloid polypeptide-

induced apoptosis. Diabetes 52, 1701–1708.

52 Oda T, Iwaoka J, Komatsu N & Muramatsu T (1999)

Involvement of N-acetylcysteine-sensitive pathways in

ricin-induced apoptotic cell death in U937 cells. Biosci

Biotechnol Biochem 63, 341–348.

53 Griffith OW, Anderson ME & Meister A (1979) Inhibi-

tion of glutathione biosynthesis by prothionine sulfoxi-

mine (S-n-propyl homocysteine sulfoximine), a selective

inhibitor of c-glutamylcysteine synthetase. J Biol Chem

254, 1205–1210.

54 Pike CJ, Ramezan-Arab N & Cotman CW (1997)

b-amyloid neurotoxicity in vitro: evidence of oxidative

stress but not protection by antioxidants. J Neurochem

69, 1601–1611.

55 Zhang Z, Rydel RE, Drzewiecki GJ, Fuson K, Wright

S, Wogulis M, Audia JE, May PC & Hyslop PA (1996)

Amyloid b-mediated oxidative and metabolic stress in

rat cortical neurons: no direct evidence for a role for

H2O2 generation. J Neurochem 67, 1595–1606.

56 Sato N, Iwata S, Nakamura K, Hori T, Mori K &

Yodoi J (1995) Thiol-mediated redox regulation of

apoptosis. Possible roles of cellular thiols other than

glutathione in T cell apoptosis. J Immunol 154, 3194–

3203.

57 Abate C, Patel L, Rauscher FJ 3rd & Curran T (1990)

Redox regulation of Fos and Jun DNA-binding activity

in vitro. Science 249, 1157–1161.

58 Park DS, Stefanis L, Yan CY, Farinelli SE & Greene

LA (1996) Ordering the cell death pathway. Differential

effects of BCL2, an interleukin-1-converting enzyme

family protease inhibitor, and other survival agents on

JNK activation in serum ⁄ nerve growth factor-deprived

PC12 cells. J Biol Chem 271, 21898–21905.

59 Meyer M, Schreck R & Baeuerle PA (1993) H2O2 and

antioxidants have opposite effects on activation of

NF-kappa B and AP-1 in intact cells: AP-1 as second-

ary antioxidant-responsive factor. EMBO J 12, 2005–

2015.

60 Xia Z, Dickens M, Raingeaud J, Davis RJ & Greenberg

ME (1995) Opposing effects of ERK and JNK-p38

MAP kinases on apoptosis. Science 270, 1326–1331.

61 Muller JM, Cahill MA, Rupec RA, Baeuerle PA &

Nordheim A (1997) Antioxidants as well as oxidants

activate c-fos via Ras-dependent activation of extracellu-

lar-signal-regulated kinase 2 and Elk-1. Eur J Biochem

244, 45–52.

B. Konarkowska et al. Thiol antioxidants suppress hA cytotoxicity

FEBS Journal 272 (2005) 4949–4959 ª 2005 FEBS 4959