Báo cáo khoa học: Effects of thymoquinone on isolated and cellular proteasomes pot

U87 MG and T98G malignant glioma cells were treated with thymoquinone, and 20S and 26S proteasome activity was measured.. Accumulation of p53 and Bax, two proteasome substrates with proa

Valentina Cecarini, Luana Quassinti, Alessia Di Blasio, Laura Bonfili, Massimo Bramucci,

Giulio Lupidi, Massimiliano Cuccioloni, Matteo Mozzicafreddo, Mauro Angeletti and

Anna Maria Eleuteri

School of Biosciences and Biotechnology, University of Camerino, Italy

Introduction

Black cumin seed (Nigella sativa) oil extracts have been

used for many centuries in the treatment of several

human diseases, and thymoquinone (TQ), its active

component, has recently been tested for its efficacy

against several diseases, including cancer [1–3]

In this regard, TQ was found to inhibit proliferation

in a concentration-dependent manner in numerous cell

lines [4,5] It has shown significant antineoplastic

activ-ity against multidrug-resistant human pancreatic

ade-nocarcinoma, uterine sarcoma and leukemic cell lines,

with minimal toxicity for normal cells [6]

In a mouse model, the injection of the essential oil

into the tumor site significantly inhibited solid tumor

development as well as the incidence of liver metasta-sis, thus improving mouse survival [5] These results indicate that the antitumor activity or cell growth inhi-bition could in part be due to the effect of TQ on the cell cycle [5] Furthermore, it has been demonstrated that the growth of prostate cancer cells is highly sensitive to the inhibitory effect of TQ, and that this inhibitory action is extremely selective, showing very little effect on the growth of noncancerous prostate epithelial cells in culture, and preventing the growth of human prostate tumors in nude mice [7]

Despite awareness of these potential antineoplastic effects, the molecular pathways involved are not

Keywords

apoptosis; glioblastoma; p53; thymoquinone;

ubiquitin proteasome system

Correspondence

V Cecarini, School of Biosciences and

Biotechnology, University of Camerino, Via

Gentile III da Varano, 62032 Camerino (MC),

Italy

Fax: +39 0737 403290

Tel: +39 0737 403247

E-mail: valentina.cecarini@unicam.it

(Received 18 November 2009, revised

24 February 2010, accepted 26 February

2010)

doi:10.1111/j.1742-4658.2010.07629.x

Thymoquinone, a naturally derived agent, has been shown to possess anti-oxidant, antiproliferative and proapoptotic activities In the present study,

we explored thymoquinone effects on the proteasomal complex, the major system involved in the removal of damaged, oxidized and misfolded pro-teins In purified 20S complexes, subunit-dependent and composition-depen-dent inhibition was observed, and the chymotrypsin-like and trypsin-like activities were the most susceptible to thymoquinone treatment U87 MG and T98G malignant glioma cells were treated with thymoquinone, and 20S and 26S proteasome activity was measured Inhibition of the complex was evident in both cell lines, but predominantly in U87 MG cells, and was accompanied by accumulation of ubiquitin conjugates Accumulation of p53 and Bax, two proteasome substrates with proapoptotic activity, was observed in both cell lines Our results demonstrate that thymoquinone induces selective and time-dependent proteasome inhibition, both in isolated enzymes and in glioblastoma cells, and suggest that this mechanism could

be implicated in the induction of apoptosis in cancer cells

Abbreviations

AMC, 7-amino-4-methyl-coumarin; BrAAP, branched chain amino acid-preferring; ChT-L, chymotrypsin-like; ECL, enhanced

chemiluminescence; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide; pAB, 4-aminobenzoate; PGPH, peptidyl-glutamyl peptide-hydrolyzing; PVDF, poly(vinylidene difluoride);

Suc, succinyl; T-L, trypsin-like; TQ, thymoquinone; Ub, ubiquitin.

completely clear Recent findings suggest that TQ has

a strong chemopreventive potential for the inhibition

of carcinogenesis by modulating lipid peroxidation and

the cellular antioxidant milieu [8,9] In fact, TQ is

reported to possess strong antioxidant properties,

inhibiting free radical generation [10] Interestingly,

according to Gali-Muhtasib et al., TQ is able to trigger

apoptosis in several cell lines in a p53-independent or

a p53-dependent manner [11,12], and, as recently

shown, its proapoptotic effects are linked to its

pro-oxidant activity [13]

Among the different mechanisms involved in the

induction of apoptotic pathways, the tumor suppressor

protein p53 plays a pivotal role [14] Under

physiologi-cal conditions, p53 is maintained at low steady-state

levels by the MDM2 protein, an E3 ubiquitin (Ub)

ligase, which ubiquitinates and targets p53 for

protea-some-mediated degradation [15] Specific stress agents

make p53 and MDM2 undergo different

post-transla-tional modifications, including phosphorylation, thus

disrupting the interaction and leading to activation of

p53 [16] At this point, p53 induces a series of

down-stream events that regulate the transcription of a

sub-set of genes involved in apoptosis, such as that

encoding Bax, a member of the Bcl-2 family [17]

The Ub–proteasome pathway is a nonlysosomal

pro-tein degradation system responsible for degrading both

damaged/unfolded proteins dangerous for normal cell

growth and metabolism [18], and critical regulatory

proteins involved in apoptosis [19], cell cycle regulation

[20], gene expression [21], carcinogenesis and DNA

repair [22–24] Because of this, studies on the discovery

of molecules that are able to modulate proteasome

activity have recently been gaining great attention The

central core of this system is the 20S proteasome This

is a cylindrical structure with an internal cavity,

com-posed of four rings, each containing seven different

a subunits and b subunits, resulting in the following

arrangement: a1–7b1–7b1–7a1–7 [19] Only three of the

seven b subunits, b1, b2, and b5, located inside the

main chamber, show proteolytic activity Specifically,

b1 is associated with the peptidyl-glutamyl

peptide-hydrolyzing (PGPH) activity and possesses limited

branched chain amino acid-preferring (BrAAP)

activ-ity, b2 is associated with the trypsin-like (T-L) activactiv-ity,

and b5 is associated with the chymotrypsin-like

(ChT-L) activity However, mutational analyses have shown

that b5 also has a tendency to cleave after small

neu-tral and branched side chains; therefore, two other

activities, BrAAP and small neutral amino

acid-prefer-ring (SNAAP), can be assigned to this subunit [25] In

certain conditions, such as in the presence of

c-inter-feron, these three b subunits can be replaced by

homologous subunits, b1i, b2i, and b5i, resulting in a

de novo synthesized proteasomal form, the immuno-proteasome, which produces mainly immunogenic pep-tides in association with major histocompatibility complex class I [19]

Malignant gliomas are the most common and lethal tumors of the central nervous system [26] Treatment outcomes, even with an aggressive approach including surgery, radiation therapy, and chemotherapy, are dis-mal The median survival of treated patients with glio-blastoma multiforme is less than 1 year, with fewer than 20% surviving for 2 years [27] There is therefore

an urgent need to devise alternative therapeutic strate-gies with which to fight gliomas

In the present work, the effects of TQ on protea-some functionality were investigated both in isolated and in cellular complexes For this purpose, constitu-tive and immune-isolated proteasomes and two human glioblastoma cell lines, U87 MG and T98G, differing

in their p53 gene status, were used Specifically, U87 MG cells present the wild-type form of p53, whereas T98G cells harbor a single p53 mutation [28]

Results

Nucleophilic susceptibility analysis

TQ was examined for sites of electrophilic and nucleo-philic susceptibility Computational analysis revealed that TQ possessed two carbons (C1 and C4) with simi-lar nucleophilic susceptibility (Fig 1) that are likely to

be the target of a nucleophilic attack [29]

TQ effects on isolated 20S proteasomes

To test TQ effects on isolated 20S constitutive and immunoproteasome functionality, we incubated purified enzymes with different concentrations of TQ (0.0–100 lm) In particular, the ChT-L, T-L, PGPH and BrAAP activities of the isolated complexes were tested using specific substrates, as described in Experi-mental procedures

As shown in Fig 2, it is possible to highlight that

TQ is able to modulate proteasome functionality indu-cing a subunit and composition-dependent inhibition

Of the two complexes, the immunoproteasome was the most susceptible to the presence of TQ, and the

ChT-L and T-ChT-L activities were the components with the highest degree of inhibition The PGPH component was not particularly altered in the presence of TQ; only 16% inhibition was evident at 30 lm Finally, the BrAAP activity was not significantly influenced by the presence of TQ (data not shown)

Interestingly, the inhibition showed

concentration-dependent behavior only up to 20 lm, when the

maxi-mum detectable rates of inhibition were 30% and 40%

for the ChT-L and T-L components, respectively, of the immunoproteasome Thereafter, an increase in TQ concentrations did not lead to enhanced inhibition Supported by the literature [30], we propose that this U-shaped inhibition depends on the presence of an additional binding site on the proteasomal complex to which TQ binds with a lower affinity than it does to the active site Our model assumes that TQ preferen-tially binds to the active site at low concentrations, resulting in the observed inhibition, whereas at higher concentrations the binding to the additional site becomes significant, allosterically restoring the activity The fraction of TQ bound to the active site is now released, allowing the substrate to enter it and be suc-cessfully degraded, resulting in the activity recovery observed at TQ concentrations higher than 20 lm To verify this hypothesis, we performed an experiment using the peptide aldehyde Z-LLF-CHO, a selective and reversible proteasome inhibitor, with the aim of blocking part of the proteasome active sites [31] After

1 h of incubation of the 20S immunoproteasome with Z-LLF-CHO, TQ at different concentrations was added and the T-L activity was measured (Fig 3) In agreement with the mechanism described above, we observed a recovery in the proteasome activity

The Nitro Blue tetrazolium assay, which monitors the formation of quinone adducts, shows the existence

of additional TQ-binding sites on the proteasome Figure 4 indicates that the formation of b-subunit–TQ adducts increases at TQ concentrations of 5 and

20 lm, whereas it decreases at a concentration of

100 lm (corresponding to the recovery of proteasome activity) At the same time, increases in TQ concentra-tion resulted in clear enhancement in the levels of a-subunit–TQ adducts, confirming our model of the presence of two different TQ-binding sites on the pro-teasome complex

TQ inhibits cell proliferation Two cell lines, U87 MG and T98G, derived from human glioblastomas were used as a model They

Fig 2 Effects of increasing TQ concentrations (0–100 l M ) on

iso-lated 20S complexes The ChT-L, T-L and PGPH activities were

assayed r, constitutive proteasome; , immunoproteasome.

Highest susceptibility

Fig 1 Chemical structure and nucleophilic susceptibility of TQ Chemical structure (A), nucleophilic susceptibility (B) and electrophilic susceptibility (C) of TQ Isosurfaces were calculated with WEBMO C 1 and C 4 carbonyl were found to be nucleophilically attacked by the OH group of Thr1.

carry, respectively, the wild-type and a mutant p53

gene This mutation consists of a single Gfi A

transi-tion in codon 237, resulting in a missense mutatransi-tion of

methionine to isoleucine [32,33] Interestingly, a study

conducted by Van Meir et al on different glioblastoma

lines and their p53 status revealed that this mutation

in the T98G line results in a transcriptionally inactive

form of p53 [34]

A set of dose–response experiments was performed

to compare the effects of TQ on cell viability in U87 MG and T98G cells Cells were incubated in the presence of TQ at concentrations ranging from 0.0 lm

to 200 lm Analysis by light microscopy showed that treatment of glioblastoma cells with increasing amounts of TQ resulted in significant alterations in cell morphology and impaired the ability of the cells

to become confluent (Fig 5A) Data obtained with the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazo-lium bromide (MTT) assay indicated that cell viability was significantly reduced in a dose-dependent and exposure time-dependent manner in both cell lines (Fig 5B) In both cell lines, almost complete loss of viability was seen after exposure to 200 lm TQ At lower concentrations, TQ exerted a stronger inhibitory effect on U87 MG cells than on T98G cells A com-parison of IC50 values, reported in Table 1, showed that, after 48 h of treatment with TQ, IC50 values were 38.82 lm for U87 MG cells and 62.48 lm for T98G cells

TQ effects on the proteasome functionality of glioblastoma cells

Considering the major role of the proteasome in medi-ating numerous cellular pathways, including apoptosis,

we wanted to determine whether TQ was able to mod-ulate its functionality in the two glioblastoma cell lines Cells were treated with TQ at 20 lm, the concentration with the greatest effects on isolated proteasomes, for

12, 24, 48 and 72 h Control cells were cultured in par-allel in the presence of dimethylsulfoxide Both cell lines had a high level of responsiveness to TQ treat-ment, showing compromised activities as compared with controls (Figs 6 and 7) Parallel assays run in the presence of specific proteasome inhibitors, Z-GPFL-CHO and lactacystin, demonstrated that the contribu-tion to the proteolysis was effectively due to the 20S proteasome (data not shown)

Figures 5 and 6 illustrate the presence of time-dependent proteasome inhibition, which assumes par-ticular significance after 48 and 72 h of treatment Interestingly, U87 MG cells showed a higher extent of proteasome inhibition, with relevant differences also at

24 h, as evident for the T-L and BrAAP activities Generally, in this cell line, TQ induced a global and stronger decrease in proteasome functionality than that observed in T98G cells

We also measured the ChT-L component of the 26S proteasome, whose proteolytic activity is ATP-depen-dent, and obtained, at 72 h, similar percentages of inhi-bition in the two lines However, at 48 h, a significant

Fig 3 TQ binding to a secondary site of the proteasome complex.

After Z-LLF-CHO and 20S immunoproteasome preincubation, in

order to partially inhibit the enzyme, the effects of increasing

con-centrations of TQ on the T-L activity were tested Data are reported

as percentages relative to proteasome activity in the presence of

Z-LLF-CHO (mean values ± standard deviations of five independent

determinations).

A

B

Fig 4 Detection of quinone adducts 20S isolated

immunoprotea-somes were treated with different concentrations of TQ and

lacta-cystin (see Experimental procedures), resolved by SDS ⁄ PAGE, and

electroblotted onto PVDF membranes Adducts were visualized

after 45 min of incubation with Nitro Blue tetrazolium Lane C

rep-resents 20S proteasome loaded without pretreatment with TQ and

lactacystin (A) Densitometry related to three different experiments.

(B) A representative membrane after the Nitro Blue tetrazolium

staining.

difference after TQ exposure was evident for U87 MG

cells

To verify the above-mentioned proteasome

inhibi-tion, we conducted western blot analyses, using

anti-bodies against Ub In fact, the abnormal presence of

Ub conjugates is a clear marker of impaired

protea-some activity Our findings demonstrate

time-depen-dent accumulation of Ub–protein aggregates,

confirming the data on proteasome inhibition (Fig 8)

Furthermore, western blot assays performed with an

antibody against 20S suggested that the observed

inhi-bition was really due to compromised complex

func-tionality and not to downregulation of its synthesis

(Fig 9) These results support the findings regarding

the ability of TQ to act directly on the proteasome

activity, and remove the possibility of decreased

syn-thesis of the enzyme

TQ effects on p53 and Bax levels

In order to strengthen the data on proteasome inhibi-tion, we measured the levels of two proteasome substrates, p53 and Bax, that play an important role in the onset of apoptotic events

In both cell lines, time-dependent accumulation of p53 was observed In T98G cells, this increase was sig-nificant even after 24 h of treatment, but was particu-larly evident at 48 h and 72 h (levels that are 2.3-fold and a 2.8-fold higher, respectively, than in controls)

In U87 MG cells, instead, the enhancement in protein levels was delayed, and became consistent only after

48 h of TQ exposure (Fig 10) Bax accumulation was more evident in T98G cells than in U87 MG cells Spe-cifically, the former responded in a shorter time, with significant increases at 48 h and 72 h (1.21-fold and 1.42-fold, respectively, that seen in controls), whereas the latter presented significant enhancement only at

72 h, with a 1.44-fold increase as compared with the respective control (Fig 11)

Discussion

The debate on the use of naturally derived drugs as coadjuvants in the treatment of cancer is of growing interest In fact, owing to concerns about the possible

A

B

Fig 5 TQ effects on U87 MG and T98G cells (A) Morphology of U87 MG and T98G cells grown under standard conditions and treated with 50 l M or 100 l M TQ dissolved

in dimethylsulfoxide Dimethylsulfoxide con-centrations in treated and control cells did not exceed 0.25% per well Cells were observed by using an inverted microscope

24 h post-treatment (B) Dose–response curve for the effect of TQ on cell viability after 24, 48 and 72 h of exposure Cell via-bility was determined by the MTT assay, and is reported as the percentage of viable cells Each value is the mean ± standard deviation of three separate experiments performed in triplicate.

Table 1 Thymoquinone IC50values for glioma cell lines after

incu-bation periods of 24, 48 and 72 h CI, confidence interval.

Incubation

period (h)

IC 50 (l M ) (95% CI)

toxic side effects of conventional medicine, the use of

natural products as alternatives to such treatments has

been increasing TQ is the most abundant constituent

of N sativa, and has pivotal roles in several biological

processes Numerous studies have demonstrated the

antioxidant, antiproliferative and proapoptotic

activities of TQ Most notably, TQ is able to induce

selective apoptosis, discriminating between tumor and

normal cells, in a p53-dependent or p53-independent

way For example, previous published data established

that osteosarcoma cells [4] and neoplastic keratino-cytes [35] are susceptible to TQ treatment, whereas normal cells and mouse primary keratinocytes do not exhibit morphological and⁄ or proliferative alterations [4,35]

The observation that proteasome inhibitors are able

to induce apoptosis in tumor cells opened the possibil-ity of their use as potential drugs, and numerous stud-ies have been conducted with the aim of finding natural, nontoxic and inexpensive compounds [36–38]

Fig 6 20S and 26S proteasome functionality in U87 MG cells treated with 20 l M TQ Activities were assayed as reported in Experimental procedures Data are expressed as percentage of activity relative to control cells in each set (mean values ± standard deviations of five inde-pendent determinations) Fluorescence due to nonproteasomal degradation was subtracted The asterisks indicate data points that are statis-tically significant as compared with the respective untreated control cells (*P < 0.05, **P < 0.01).

In this scenario, we decided to investigate the possible

interaction between TQ and proteasomes in order to

determine whether TQ could modulate the enzyme

functionality

Considering the data obtained from computational

analysis, it is reasonable to think that TQ could

behave as a nucleophilic target, resulting in inhibition

of proteasome activity To confirm this hypothesis, we

tested proteasome functionality after TQ treatment of

both isolated and cellular complexes Interestingly, we

observed subunit-dependent and composition-depen-dent inhibition of both the purified enzymes, with the immunoproteasome being the most sensitive and the ChT-L and T-L components being the most influenced activities We also demonstrated that TQ induces a U-shaped inhibition in proteasome complexes through the binding of two distinct sites with different degrees

of affinity

Exposure of two human glioblastoma cell lines, U87 MG and T98G, to TQ was able to significantly

Fig 7 20S and 26S proteasome functionality in T98G cells treated with 20 l M TQ Activities were assayed as reported in Experimental pro-cedures Data are expressed as percentage of activity remaining relative to control cells in each set (mean values ± standard deviations of five independent determinations) Fluorescence due to nonproteasomal degradation was subtracted The asterisks indicate data points that are statistically significant as compared with the respective untreated control cells (*P < 0.05, **P < 0.01).

compromise proteasome activity The two cell lines are

different with respect to a single mutation in the p53

gene; this characterizes the T98G line, whereas the

U87 MG line maintains the wild-type form of the

pro-tein As previously shown by other authors, this

muta-tion results in a transcripmuta-tionally inactive p53 gene

[34]

Assaying TQ effects on cell viability, we found that

both cell lines showed clear changes in cell

morphol-ogy, although with different degrees of sensitivity In

fact, U87 MG cells were more susceptible to the

treatment, as shown by the different IC50 values

obtained after the treatments

Cells were then treated with TQ at 20 lm, the

con-centration with the highest effect according to the

in vitro data, and both 20S and 26S proteasomes

showed changes in their functionality In particular,

our assays showed significant, time-dependent but

dif-ferential sensitivities of U87 MG and T98G cells to

TQ treatment T-L, BrAAP and PGPH activities were

significantly more affected in U87 MG cells than in T98G cells, with the former showing altered protea-some functionality at 24 h This inhibition was also confirmed by accumulation of Ub–protein conjugates Furthermore, when we tested the 20S expression levels with specific antibodies, we could not detect any differ-ences between control and treated cells, demonstrating the ability of TQ to directly alter proteasome activity without affecting its synthesis

Considering our data, it is clear that TQ is able to modulate proteasome activity, inducing global inhibi-tion in the studied models, although to different extents These results are in line with previously pub-lished data from our laboratory and others reporting

on the ability of small, naturally derived ligands, e.g flavonoids, to inhibit proteasome functionality and selectively modulate its activity, depending on the subunit composition [37,39,40]

It has been widely reported that the proteasome, being responsible for the removal of proapoptotic

A

B

C

Fig 8 Detection of Ub–protein conjugates in U87 MG and T98G cells The densitometric analysis from five separate blots, shown as mean values ± standard deviations, and a representative western blot are shown (A, B) Membranes were reprobed with GAPDH antibody to ensure equal protein loading (C) Detection was performed with an ECL western blotting analysis system The asterisks indicate data points that are statistically significant as compared with the respective untreated control cells (*P < 0.05, **P < 0.01).

proteins, is involved in the induction of programmed

cell death [19] Its inhibition, in fact, triggers the

accu-mulation of proteins such as p53 and Bax [41–43] For

this reason, numerous compounds with the ability to modulate proteasome activity have been used in the treatment of malignancies

A

B

C

Fig 9 Detection of the 20S ‘core’ in U87 MG and T98G cells The densitometric analysis from five separate blots, shown as mean val-ues ± standard deviations, and a representative western blot are shown (A, B) Membranes were reprobed with GAPDH antibody to ensure equal protein loading (C) Detection was performed with an ECL western blotting analysis system.

A

B

C

Fig 10 Detection of p53 in U87 MG and T98G cells The densitometric analysis from five separate blots, shown as mean values ± standard deviations, and a representative western blot are shown (A, B) Membranes were reprobed with GAPDH antibody to ensure equal protein loading (C) Detection was performed with an ECL western blotting analysis system The asterisks indicate data points that are statistically significant as compared with the respective untreated control cells (*P < 0.05, **P < 0.01).

Our results on the accumulation of both p53 and

Bax are in line with the data describing the ability of

TQ to inhibit proteasome activity These two

proapop-totic proteins are proteasome substrates, and their

intracellular levels increase together with proteasome

malfunctions It is therefore likely that one of the

mechanisms through which TQ triggers apoptosis in

cancer cells is the induction of proteasome inhibition

In summary, our data demonstrate that TQ is able

to modulate proteasome functionality, inducing

compo-sition-dependent inhibition both in isolated complexes

and in glioblastoma cells This inhibition leads to

intra-cellular increases in the levels of apoptotic proteins

such as p53 and Bax, and may be linked to the onset of

apoptotic events Such findings represent evidence that

this compound, characterized by very low toxicity,

deserves further clinical analysis and investigation,

mostly for its potential application as an adjuvant in

the treatment of cancer and other diseases

Experimental procedures

Reagents and chemicals

Thymoquinone, substrates for assaying the ChT-L, T-L

and PGPH activities [succinyl

(Suc)-Leu-Leu-Val-Tyr-7-amino-4-methyl-coumarin (AMC),

Z-Leu-Ser-Thr-Arg-AMC, and Z-Leu-Leu-Glu-AMC], proteasome inhibitors (Z-Gly-Pro-Phe-Leu-CHO and lactacystin), Nitro Blue Tet-razolium and MTT were purchased from Sigma-Aldrich S.r.L (Milan, Italy) The substrate Z-Gly-Pro-Ala-Phe-Gly-4-aminobenzoate (pAB), for testing BrAAP activity, and the proteasome inhibitor Z-LLF-CHO (Cbz-Leu-Leu-Phe-CHO) were kind gifts from M Orlowski (Depart-ment of Pharmacology, Mount Sinai School of Medicine, New York, NY, USA) Aminopeptidase N (EC 3.4.11.2) for the coupled assay utilized to detect BrAAP activity [44] was purified from pig kidney as reported elsewhere [45,46] TQ was dissolved in dimethylsulfoxide (Sigma Aldrich S.r.l.) U87 MG and T98G human glioblastoma cell lines were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA) All of the reagents for cell cultures were obtained from Euroclone (Milan, Italy) Rabbit anti-(human 20S proteasome) serum, rabbit anti-(human 20S proteasome b5 subunit) serum and mouse anti-[human 20S a(1, 2, 3, 5, 6, and 7) subunits] serum were purchased from BIOMOL Interna-tional, L.P The mouse monoclonal antibodies against Ub, p53 and Bax were obtained from Santa Cruz Biotechnol-ogy, Inc (Heidelberg, Germany) Membranes for western blot analyses were purchased from Millipore (Milan, Italy) Proteins immobilized on films were detected with the enhanced chemiluminescence (ECL) system (Amersham Pharmacia Biotech, Milan, Italy) All chemicals and sol-vents were of the highest analytical grade available

A

B

C

Fig 11 Detection of Bax in U87 MG and T98G cells The densitometric analysis from five separate blots, shown as mean values ± standard deviations, and a representative western blot are shown (A, B) Membranes were reprobed with GAPDH antibody to ensure equal protein loading (C) Detection was performed with an ECL western blotting analysis system The asterisks indicate data points that are statistically significant as compared with the respective untreated control cells (*P < 0.05, **P < 0.01).