| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Laboratory of Cell and Developmental Genetics, Department of Medicine, Université Laval and CHUL Research Center, Ste-Foy, Quebec, Canada G1K 7P4
1Correspondence: Laboratory of Cell and Developmental Genetics, Department of Medicine, Pav. CE. Marchand, Université Laval, Ste-Foy, Qc, Canada, G1K 7P4. E-mail: Robert.Tanguay{at}rsvs.ulaval.ca
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
m). These events were prevented by the general
caspase inhibitor z-VAD-fmk, whereas G2/M arrest and subsequent
apoptosis were abolished by GSH-monoethylester or
N-acetylcysteine. Other tyrosine metabolites,
maleylacetoacetate and succinylacetone, had no antiproliferative
effects and induced only very low levels of apoptosis. These results
suggest a modulator role of GSH in FAA-induced cell cycle disturbance
and apoptosis where activation of cyclin B-dependent kinase and
caspase-1 are early events preceding mitochondrial cytochrome
c release, caspase-3 activation, and m
loss.—Jorquera, R., Tanguay, R. M. Cyclin B-dependent kinase and
caspase-1 activation precedes mitochondrial dysfunction in
fumarylacetoacetate-induced apoptosis.
Key Words: cell cycle • caspase • cytochrome c • glutathione • cell death
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Accumulation of FAA ensues the disruption of this hydrolytic step in
hereditary tyrosinemia type I (HT I), an autosomal recessive metabolic
disease caused by FAH deficiency (McKusick No. 276700)
(1
2
3
4)
. Severe hepatic alterations are observed in HT I
patients, ranging from acute liver dysfunction at early infancy to
cirrhosis/hepatocellular carcinoma at mid-childhood (5
, 6)
. Although the etiopathogenesis of HT I-associated liver
pathologies remains unknown, a direct hepatotoxic activity of tyrosine
metabolites accumulating in this disease has been suggested but not
demonstrated (1
, 7
, 8)
. This hypothesis is based on the
fact that FAA and maleylacetoacetate (MAA) possess 
,ß-unsaturated
carbonyl compound structures (see Fig. 1
) that confer electrophilic
properties and potential biological activities. Alkylation of cellular
macromolecules such as DNA and/or disruption of essential sulfhydryl
reactions by complexation with proteins or glutathione (GSH) are among
the mechanisms suggested to underlie the toxic effects of FAA and MAA
(1
, 8)
.
|
Recently, four studies—two in cultured cells (9
, 10)
and
two in an animal model of HT I (11
, 12)
—provided evidence
supporting the hypothesis that causally links the toxic activity of
tyrosine metabolites with HT I-associated liver pathologies. Endo et
al. (11)
created a double mutant mouse model
(FAH-/-/HPD-/-) by
crossing lethal albino deletion c14CoS
mice, i.e., FAH-deficient mice (13)
with mice from a
strain lacking p-hydroxyphenylpyruvate dioxygenase (HPD), the enzyme
that converts p-hydroxyphenylpyruvic acid into homogentisic acid (see
Fig. 1
). Although homozygous c14CoS
mice (FAH-/-) die at birth (13)
,
the homozygous null mutation of the HPD gene in the double mutant can
completely rescue the lethal phenotype of FAH deficiency
(11)
. Retrieval of the tyrosine catabolic pathway in the
liver of FAH-/-/HPD-/-
mice by administration of homogentisic acid or by adenovirus-mediated
HPD gene transfer resulted in severe acute liver damage, rapid
apoptotic death of hepatocytes, and animal death within 16–30 h after
retrieval (11)
. These observations led the investigators
to suggest a direct relationship between liver injury (induction of
apoptosis in hepatocytes) and the accumulation of homogentisic
acid-derived metabolites, but without identifying any particular
metabolite as inducer of apoptosis. In subsequent work, Kubo et al.
(12)
proposed that at least one of the signals for cell
death in hepatocytes of the double mutant mice was FAA on the basis of
the observation that in a cell-free system, cytochrome c was
released from mitochondria after the addition of either a cytosolic
fraction from the liver of homogentisic acid-treated
FAH-/-/HPD-/- mice or
purified FAA from this fraction. In these mice, cytochrome c
release occurred prior to apoptotic hepatocyte death and liver failure,
and these events could be prevented by caspase inhibitors
(12)
. However, other than the release of cytochrome
c from isolated mitochondria after FAA treatment, this study
did not provide direct evidence demonstrating that FAA per se is
cytotoxic or an inducer of apoptosis either in vitro or
in vivo. In previous studies using cultured cells, we showed
that among the tyrosine metabolites that accumulate in HT I, i.e., FAA,
MAA, and their derivative succinylacetone (SA), only FAA displayed a
mutagenic activity on V79 Chinese hamster lung fibroblasts
(9)
. Moreover, we demonstrated that FAA was an efficient
depletor of cellular glutathione (GSH) and that GSH depletion with
L-buthionine-(S,R)-sulfoximine (BSO) enhanced the mutagenicity of FAA
(10)
.
The present study was therefore performed to extend our knowledge about
the intrinsic cytotoxic activity of FAA and to understand the molecular
mechanisms that could underlie the toxic effects of FAA in HT I. Since
a cytotoxic activity of FAA (i.e., mutagenicity, which we believe to be
directly involved in HT I liver pathogenesis) was originally
demonstrated in V79 cells (10)
, we used this cell line. We
show that FAA, unlike MAA and SA, inhibits cell proliferation by
arresting the cell cycle in G2/M and that arrested cells then undergo
apoptotic death. Similar responses were observed in FAA-treated human
HepG2 cells. Among the molecular events that accompany these cytotoxic
effects of FAA, activation of cyclin B-dependent kinase and caspase-1
are both early events and precede others such as caspase-3 activation
and mitochondrial dysfunction-related events such as cytochrome
c release and mitochondrial transmembrane potential
(m) loss. In addition, we report that as
previously observed in our study of the mutagenicity of FAA, both the
antiproliferative and apoptosis-inducing activities of FAA are also
markedly enhanced by cellular GSH depletion. These cytotoxic activities
can be prevented by GSH replenishment agents and caspase inhibitors.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and kept frozen at -80°C until use. SA and BSO
were purchased from Sigma Chemical Co. (St. Louis, Mo.). FAA, MAA, SA,
or BSO were dissolved in HBSS (GIBCO BRL Products; Grand Island, N.Y.)
supplemented with NaHCO3 (0.35 g/l), pH 7.3. The
concentration of FAA or MAA solutions were determined by spectral
analysis (14)
. All solutions were filter sterilized before
treatment of cells.
Cell culture and drug treatments
Chinese hamster V79 cells or human HepG2 cells (ATCC; Rockville,
Md.) were grown at 37°C with 5% CO2 in
Dulbecco’s modified Eagle medium (DMEM, high glucose; GIBCO BRL
Products) supplemented with 5% or 10%, respectively, fetal bovine
serum (FBS; Immunocorp Sciences), and penicillin (100
U/ml)/streptomycin (100 µg/ml)/amphotericin B (0.25 µg/ml), pH 7.4.
Cells were initially seeded at a density of 2.3 x
104 cells/ml of medium, which allowed exponential
growth at the start of each treatment. For GSH depletion, BSO (0.2 mM)
was added to medium 6 h after seeding and left for 18 h. The
medium was replaced by HBSS (controls) or HBSS containing FAA, MAA, or
SA at the stated concentrations. Treatments were terminated 2 h
later by replacing HBSS with fresh medium. When caspase inhibitors or
GSH replenishment agents were used, cells were preincubated for 2 h with N-acetyl-Tyr-Val-Ala-Asp-aldehyde (Ac-YVAD-CHO),
N-benzyloxycarbonyl-Asp-Glu-Val-Asp-fluoromethylketone
(z-DEVD-fmk), or
N-benzyloxycarbonyl-Val-Ala-Asp-fluoromethylketone
(z-VAD-fmk) (Calbiochem; La Jolla, Calif.), GSH-monoethylester
(GSH-MEE), or N-acetylcysteine (NAC) (Sigma) before FAA
treatment. Cells were harvested at different times post-treatment and
processed as indicated below.
Survival and growth studies
Cells were seeded in 6-well plates (Nunc) with 2 ml of medium.
The FAA concentration required to achieve 50% cell killing
(IC50) was determined 48 h after FAA removal
by trypsinization of cells and counting in the presence of 0.1% trypan
blue solution, using a hemacytometer. Cell growth was assessed by
counting trypan blue-excluding cells from 24 to 96 h
post-treatment.
Cell cycle analysis by flow cytometry
Cells were seeded in T-80 flasks (Nunc) with 10 ml of medium.
After treatment, floating and trypsinized cells were pooled,
centrifuged (1000xg for 5 min), resuspended in
phosphate-buffered saline (PBS), fixed in cold 70% ethanol, and kept
at -20°C for at least 4 h. Fixed cells
(0.5x106) were collected by centrifugation
(1000xg for 10 min) and resuspended in PBS/0.1% Triton
X-100 containing RNase (1 mg/ml) and propidium iodide (50 µg/ml).
Measurement of DNA content of nonapoptotic cells as well as
ethanol-extracted low molecular weight DNA from apoptotic cells (sub-G1
population) was done using a Epics Elite Esp flow cytometer (Coulter;
Miami, Fla.). The percentage of cells in G1, S, and G2/M phases of the
cycle (among nonapoptotic cell population) was calculated using the
Multicycleav program (Phoenix Flow Systems Inc.; San Diego, Calif.).
Cell synchronization and detection of apoptosis-associated DNA
strand breaks by flow cytometry
Cells were synchronized at the G1/S transition by mimosine
treatment (15)
. Briefly, cells grown for 24 h in
medium were treated with mimosine (200 µM) for an additional 18 h period. After release into medium without mimosine, cells progress
synchronously through S and G2/M phases. Synchronized cells were
treated with FAA (100 µMx2 h) either 3.5 h (mid-S) or
6 h (G2/M) after mimosine release. At defined time points after
FAA treatment, cells were harvested and prepared for flow cytometry for
either DNA content analysis (see above) or in situ TdT assay
as described by Gorczyca et al. (16)
, using TdT and
fluorescein-dUTP (In Situ Cell Death Detection Kit; Boehringer
Mannheim; Mannheim, Germany) and simultaneous counterstaining of DNA
with propidium iodide.
Detection of apoptosis by fluorescence microscopy and agarose gel
electrophoresis
Apoptosis was assessed morphologically by dual staining of cells
with the DNA intercalating fluorescent dyes acridine orange and
ethidium bromide, which allow simultaneous visualization of the nuclear
chromatin pattern and evaluation of cell membrane integrity
(17)
. After treatment, floating and attached trypsinized
cells were pooled, stained (5 µg/ml of each dye), mounted, and
examined in a fluorescence microscope (Leitz DMRB). The percentages of
apoptotic, nonapoptotic, viable, or nonviable cells were determined
by analyzing 100 cells/slide. To detect apoptosis-associated
internucleosomal DNA fragmentation, low molecular weight DNA from
1 x 106 ethanol-fixed cells was extracted
as described by Gong et al. (18)
, with the addition of a
final conventional DNA precipitation step using ammonium acetate 5
M/anhydrous ethanol (0.6/2 v/v). DNA was analyzed by agarose gel
electrophoresis (1.5% agarose, 58 V for ~2 h) using 0.5x
Tris-Borate-EDTA running buffer and visualized by ethidium bromide
staining.
Measurement of p34cdc2 kinase activity in immunoprecipitated
cyclin B complexes
Cells were seeded in 15 cm Petri dishes (Sarstedt) with 30 ml of
medium. At different times post-FAA treatment, cells were harvested by
rapidly rinsing twice with cold PBS containing 1 mM PMSF and 0.4 mM
Na3VO4 and scraped. The
immunoprecipitation procedure was basically as described by Faure et
al. (19)
, with some minor modifications. Lysis and
dilution buffers included EDTA (20 mM)/NaF (50 mM) and NaF (2.5 mM),
respectively. Anti-cyclin B (i.e., cyclin B1; 2 µg/mg protein;
Neomarkers Inc.; Fremont, Calif.) was added to supernatants (200 µg
protein) and incubated overnight at 4°C. Immunocomplexes were
precipitated with 40 µl of protein A-Sepharose beads (Sigma) for
2 h at 4°C. The p34cdc2 kinase activity in cyclin B
immunoprecipitates was measured using histone H1 as substrate
(19)
after incubation (10 min at room temperature) of
beads in a reaction mixture containing 80 µCi
[
-32P] ATP (NEN; Boston, Mass.) and 0.1
mg/ml histone H1 (Boehringer Mannheim). Radioactivity was counted on an
LKB rack beta counter.
Measurement of caspase-1 and caspase-3 activities
Cytosolic extracts were prepared basically as described by Kluck
et al. (20)
except the extraction buffer contained 250 mM
sucrose. Proteolytic reactions were carried out in extraction buffer
(500 µl) containing 40 µg of cytosolic protein extracts and 160
µM N-acetyl-Tyr-Val-Ala-Asp-p-nitroanilide
(YVAD-pNA) or
N-acetyl-Asp-Glu-Val-Asp-p-nitroanilide
(DEVD-pNA) (Calbiochem) for assaying caspase-1 or caspase-3
activity, respectively. The reaction mixtures were incubated at room
temperature for 4 h and the formation of p-nitroanilide
was measured at 405 nm using a Pharmacia LKB Ultrospec III
spectrophotometer.
Gel electrophoresis and immunoblotting
For analysis of mitochondrial cytochrome c release
and procaspase-3 activation, cytosolic extracts were resuspended in
loading buffer [62.5 mM Tris-HCl, pH 6.8, 2.3% sodium dodecyl sulfate
(SDS), 10% glycerol, 5% ß-mercaptoethanol] and proteins (20 µg)
were separated by SDS-PAGE (polyacrylamide gel electrophoresis) in a
15% gel. For analysis of cyclin B and MPM-2 protein levels, cells were
directly scraped in loading buffer and proteins (20 µg) were
separated by SDS-PAGE in a 12% gel. After blotting, nitrocellulose
membranes were blocked overnight at 4°C with 1% blocking solution
(ECL kit; Boehringer Mannheim) and probed with anti-cytochrome
c (1/250; monoclonal antibody, clone 7H8.2C12; PharMingen,
Ont., Canada), anti-caspase-3 (1/1000; polyclonal antibody, p-17 MF393,
kindly provided by D. Nicholson, Merck Frosst, Qc, Canada), anti-cyclin
B (i.e., cyclin B1; monoclonal antibody, 1/300; Neomarkers Inc.;
Fremont, Calif.) or anti-MPM-2 (1/500; monoclonal antibody; Dako
Diagnostics Inc., Ont., Canada). An anti-rabbit or anti-mouse
immunoglobulin G-horseradish peroxidase was used as secondary
conjugated antibody. Signals were revealed using the enhanced
chemiluminescence detection system (ECL; Boehringer Mannheim).
Measurement of the mitochondrial transmembrane potential
(m) by flow cytometry
To measure m, the lipophilic dye
3,3'-dihexyloxacarbocyanine iodide (DiOC6[3];
Molecular Probes, Inc., Eugene, Oreg.) was used as described by Petit
et al. (21)
. Briefly, trypsinized cells (2 x
105 cells) were rinsed with PBS and incubated
with DiOC6[3] (0.2 µM) for 15 min at 37°C.
As a positive control for m loss, control
cells were sequentially incubated (5 min at 37°C) with the uncoupling
agent carbonyl cyanide m-chlorophenylhydrazone (mClCCP, 100 µM;
Sigma). Fluorescence was measured by flow cytometry using excitation at
488 nm and emission at 525 nm.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Figure 2A
shows that survival of FAA-treated cells decreased from
IC50 = 65 µM to IC50 = 20
µM when treatment was done on GSH-depleted cells. GSH depletion had a
marked potentiating effect on FAA-induced cytotoxicity at doses 
100 µM FAA, but this effect was lost at higher doses. As shown in
Fig. 2B
, cell proliferation was also inhibited in a
dose-dependent manner and GSH depletion prior to FAA treatment
potentiated the cell growth inhibitory effects of FAA. Cell recovered
rapidly after treatment with 35 µM FAA/-BSO. In contrast, resumption
of cell proliferation was observed only 48–72 h after treatment with
35 µM FAA/+BSO and 100 µM FAA/-BSO or was not observed (100 µM
FAA/+BSO or 200 µM FAA/±BSO). Depletion of GSH by BSO treatment did
not by itself affect the normal growth of cells (Fig. 2B
).
The colony-forming ability of cells treated with 35 µM FAA/+BSO or
100 µM FAA/-BSO, as measured 10 days after recovery, was 35% and
15%, respectively, vs. 91% in control cells (data not shown). Neither
MAA (100 µM) nor SA (1000 µM) induced cell toxicity or inhibition
of cell proliferation when assayed on normal or BSO-treated cells (data
not shown). Thus, both cell survival and proliferation after FAA
treatment are dose dependent and susceptible to modulation by
manipulating intracellular GSH levels, i.e., GSH depletion decreasing
cell survival and proliferation after FAA treatment.
|
FAA induces cell cycle arrest in exponentially growing cells
To further examine the inhibitory effect of FAA on cell
proliferation, exponentially growing cells were treated with FAA and
analyzed by flow cytometry to determine the percentage of cells in
every phase of the cell cycle by their DNA content. No cell cycle
alterations were observed immediately after FAA removal (data not
shown). However, Fig. 3A
shows that at 24 h post-treatment, cells treated with
35 µM FAA/+BSO (but not 35 µM FAA/-BSO) or 100 µM FAA/± BSO
partially accumulated in G2/M (4N DNA). The percentages of G2/M
accumulation were 20% (35 µM FAA/+BSO), 37% (100 µM FAA/-BSO),
and 15% (100 µM FAA/+BSO) vs. 9–10% in control cells. In cells
treated with 100 µM FAA/-BSO or 100 µM FAA/+BSO, a decrease in the
G1 population (2N DNA; 22–24% vs. 50–53% in control cells) and a
small sub-G1 peak (<2N DNA; 10–13% vs. 0–2% in control cells) were
also observed. The decrease of the G1 peak suggests a block in the
progression of G2/M cells to G1, whereas the sub-G1 population is
indicative of cell death and is considered to be a marker of apoptosis
(22)
. The apoptotic death of these cells was confirmed by
other methods, such as acridine orange/ethidium bromide staining and
the ‘DNA ladder’ technique (see below). Cells treated with 100 µM
FAA/+BSO also showed an increase in S-phase (63% vs. 38% in control
cells) at 24 h post-treatment. This suggests a second block in the
cell cycle—inhibition or slowdown of progression through
S-phase—which could explain why the G2/M fraction in these cells did
not increase to the extent observed after 100 µM FAA/-BSO treatment.
When a shorter recovery interval (e.g., 12 h) was examined, cell
cycle alterations were observed only after treatment with 100 µM FAA
(+/-BSO), with the G2/M and sub-G1 peaks being no greater than 12%
and the fraction of G1- or S-phase cells moderately decreasing (G1:
35% vs. 51% in controls) or increasing (S: 52% vs. 39% in controls)
(data not shown). Treatment with BSO alone did not alter cell cycle
progression (Fig. 3A
).
|
When cells were left to recover 48 h after FAA treatment (Fig. 3B
), a decline in the fraction of cells accumulated in G2/M
was observed and was coincident with an increase of the sub-G1 peak.
The decrease in G2/M-phase cells was 5% (i.e., from 20% at 24 h
to 15% at 48 h) in cells treated with 35 µM FAA/+BSO and 18%
(i.e., from 37% at 24 h to 19% at 48 h) in cells treated
with 100 µM FAA/-BSO. Meanwhile, the decrease in G2/M correlated
with the increase in the sub-G1 peak, which was 4% (i.e., from 2% at
24 h to 6% at 48 h) in cells treated with 35 µM FAA/+BSO
and 11% (i.e., from 10% at 24 h to 21% at 48 h) in cells
treated with 100 µM FAA/-BSO. This indicates that a major fraction
(~2/3) of FAA-treated cells was arrested in the G2/M-phase before
undergoing cell death. In cells treated with 100 µM FAA/+BSO, the
majority of cells were dead at 48 h post-treatment, as indicated
by a prominent sub-G1 peak (71%; Fig. 3B
, last column).
This low DNA content population could be assigned mostly to apoptotic,
nonviable cells (dying by a secondary necrosis process), as confirmed
by other methods (see below).
Treatment with MAA (100 µM) caused a slight decrease in G1-phase
(-BSO: 44% vs. 50% in controls; +BSO: 40% vs. 53% in controls) and
an increase in S-phase cells (-BSO: 43% vs. 40% in controls; +BSO:
49% vs. 38% in controls) at 24 h post-treatment (Fig. 3C
, two first panels). However, a normal cell cycle
distribution (and proliferation) was found 48 h after MAA
treatment (Fig. 3D
, two first panels), as similarly observed
at any time period after SA (1000 µM) treatment (Fig. 3C, D
, last two panels).
When human hepatocarcinoma-derived HepG2 cells were treated with FAA (100 µM/-BSO), cells also accumulated in G2/M (45% vs. 13% in controls) at 24 h post-treatment; the fraction of G2/M-arrested cells declined (from 45% to 24%) at 48 h post-treatment with the concomitant appearance of a sub-G1, apoptotic population (from 22% to 47%; < 8% in control cells) (data not shown).
In summary, these results show that FAA, but not MAA nor SA, induced sequential G2/M arrest and apoptosis in exponentially growing cells, events that do not seem to be cell type specific. Moreover, the FAA-induced cell cycle arrest and apoptosis are dose and GSH dependent, GSH depletion potentiating these FAA effects.
FAA delays S-phase progression in synchronized cells and arrests
cells in G2/M resulting in apoptosis
To better visualize the cell cycle effects of FAA, we
synchronized cells at the G1/S boundary by using mimosine (200
µMx18 h) (15)
. After removal of mimosine, cells
progressed synchronously through the cell cycle, reaching the mid-S and
G2/M phases 3.5 h and 6 h after mimosine release (Fig. 4A
). At each of these times, synchronous cells were treated
with FAA (100 µMx2 h), and their cell cycle progression after
FAA removal was compared with that of untreated cells by flow
cytometric analysis of DNA content. As shown in Fig. 4B
, in
contrast to control cells that massively entered the subsequent
G1-phase by 9.5 h after mimosine release
(see Fig. 4A
), treatment of cells in mid-S-phase with FAA
resulted in a delay in completion of the S-phase, followed by G2/M
arrest. Most cells attained a G2/M DNA content by 24 h and
remained in G2/M for an extended period (>32 h). A sub-G1 peak
was evident at 32 and 48 h after mimosine release, suggesting that
cell death by apoptosis was occurring. As shown in Fig. 4C
,
similar accumulation of cells with G2/M DNA content and appearance of
the sub-G1 peak was observed when synchronous cells were treated in
G2/M with FAA. Apoptotic cells were identified by labeling their broken
DNA ends with fluorescein-dUTP via TdT and simultaneous DNA
counterstaining with propidium iodide, which allows identification of
their cell cycle position (16)
. From the ‘contour‘
graphs shown in Fig. 4B, C
, it was clear that apoptosis
occurred maximally 32 h after mimosine release in FAA-treated
cells whether cells were treated in mid-S or G2/M phases. Moreover, the
majority of apoptotic cells were G2/M-phase cells.
|
Thus, cell cycle disturbance and apoptosis induced by FAA seem to be related events, with accumulation of cells in G2/M preceding their apoptotic death.
Cell death induced by FAA occurs mainly by apoptosis
To quantitatively assess apoptosis and cell viability,
exponentially growing cells were treated with FAA and examined at
different recovery times by fluorescence microscopy using acridine
orange and ethidium bromide. As shown in Fig. 5A
, after a 2 h exposure to FAA, only cells treated with
100 µM FAA/+BSO (0 h, lane 6) showed a low but significant level of
apoptosis (8% vs. 2% in controls). Viability (as assessed by ethidium
bromide exclusion) of these cells remained high (94%), which is in
line with the maintenance of cellular membrane integrity that
characterizes (early) apoptosis. Twenty-four hours after FAA treatment,
the percentage of apoptotic cells reached 10% for cells treated with
35 µM FAA/-BSO (Fig. 5A
; 24 h, lane 3), and ~20% for
treatments with 35 µM FAA/+BSO (24 h, lane 4) and 100 µM FAA/±BSO
(24 h, lanes 5 and 6). Apoptotic, nonviable cells (~20%) were
observed after treatment with 100 µM FAA/+BSO (Fig. 5A
;
24 h, lane 6), which coincided with a fall of cell viability to
87%. At 48 h post-treatment, the percentages of apoptotic, viable
cells had increased to 25% and 34% in cells treated with 35 µM FAA
alone (Fig. 5A
; 48 h, lane 3) or pretreated with BSO
(48 h, lane 4), respectively. In the latter cells, 8% were apoptotic
and nonviable. After a 100 µM FAA/-BSO treatment (48 h, lane 5),
apoptotic, viable cells increased to 54%. In contrast, in cells
treated with 100 µM FAA/+BSO, the percentages of apoptotic, nonviable
cells as well as necrotic cells increased, reaching levels of 30% and
14%, respectively (Fig. 5A
; 48 h, lane 6). As
expected, a marked fall in viability (to 77%) was observed in these
cells.
|
Analysis of the DNA isolated from FAA-treated cells by agarose gel
electrophoresis confirms the apoptotic death of these cells.
Fragmentation of DNA into nucleosome-sized pieces characteristic of
apoptosis was barely detectable 24 h after all FAA treatments
(data not shown). However, 48 h after the removal of FAA, a
substantial fragmentation of DNA into a ladder of nucleosome-sized
pieces (~256 bp) occurred in cells treated with 35 µM FAA/±BSO
(Fig. 5B
, lanes 3 and 4) and 100 µM FAA/-BSO (lane 5).
The DNA ladder induced by treatment of V79 cells with the
apoptosis-inducing agent etoposide (23)
is shown for
comparison (Fig. 5B
, lane 7). Cells treated with 100 µM
FAA/+BSO showed a DNA ‘smear’ instead of a DNA ladder (Fig. 5B
, lane 6), likely reflecting the higher levels of necrosis
(including apoptotic, nonviable cells) induced by this treatment.
Internucleosomal DNA fragmentation was not observed or only barely
detectable after treatment with BSO alone (Fig. 5B
, lane 2),
MAA (100 µM), or SA (1 mM) (data not shown).
In summary, these results show that cell death after FAA treatment occurs mainly by apoptosis and that the extent of FAA-induced apoptosis is dose- and GSH dependent.
FAA causes an early activation of cyclin B-dependent kinase
Since cyclin B-p34cdc2 is a key regulatory complex that controls
cell cycle progression from the G2-phase into mitosis
(24)
, we next measured the kinase activity associated with
this complex, using histone H1 as substrate. After treatment of
asynchronous cells with 100 µM FAA/-BSO, cell lysates were prepared
at different time intervals after FAA removal and immunoprecipitated
with anti-cyclin B. Figure 6
shows the relative histone H1 kinase activity of these
immunoprecipitates. Immediately after FAA removal, cyclin B-dependent
kinase activity was 40% lower than the activity in control cells. The
kinase activity gradually recovered, exceeding control values by 10%
and 60% at 1 h and 6 h post-treatment, respectively, and
reached a maximum 24 h after FAA removal, when it was 3.3-fold
higher than control values. For comparison, cells blocked in mitosis
with nocodazole had a kinase activity 8 times greater than that of a
population of exponentially growing cells (data not shown). The cyclin
B-p34cdc2 kinase activity of FAA-treated cells decreased 48 h
after treatment, but still showed levels higher (by 80%) than control
values. When total cellular levels of cyclin B were analyzed by
immunoblotting, FAA-treated cells showed a slight increase in cyclin B
protein at 24–48 h post-treatment (Fig. 7
, upper panel), thus suggesting an accumulation of cyclin B, which
occurs late after removal of FAA.
|
|
Since cyclin B-p34cdc2 kinase is activated at the transition from
prophase to metaphase (25)
, the observed G2/M arrest
induced by FAA in the flow cytometric analysis seems to reflect an
early mitotic arrest. Twenty-four and 48 h after treatment of
cells with 35 µM FAA/+BSO, the mitotic index was 1% and 4%,
respectively, vs. 5% in control cells, whereas no metaphase or
postmetaphase figures could be observed in cells treated with 100 µM
FAA/-BSO; however, a mitotic index similar to control cells was
observed in these cells 72–96 h after FAA treatment (data not shown).
The early mitotic arrest induced by FAA was further confirmed by the
strong reactivity of the antibody MPM-2, which reacts with
mitosis-specific phosphoproteins (26)
and lysates of cells
treated with 100 µM FAA/-BSO, particularly at 24 h
post-treatment (Fig. 7
, lower panel).
Thus, these results show that one of the early events induced by FAA is a cell cycle-related event (activation of cyclin B-dependent kinase), and the kinetics of this activation closely follows that of the G2/M arrest.
FAA causes an early activation of caspase-1, followed by
mitochondrial cytochrome c release, caspase-3
activation, and a late reduction in m
To examine the sequence of molecular events that took place
in FAA-induced apoptosis, we simultaneously monitored caspase activity,
mitochondrial cytochrome c release, and mitochondrial
transmembrane potential (m) status at
different recovery times after treatment of asynchronous cells with 100
µM FAA/-BSO. Caspase-associated proteolytic activities were measured
by testing the ability of cytosolic extracts from untreated and
FAA-treated cells to cleave the colorimetric peptide substrates YVAD-
and DEVD-pNA, correspondingly indicating caspase-1 and caspase-3
activation. As shown in Fig. 8A
(upper panel), caspase-1 activity was already increased
1 h after FAA removal, reached maximal levels (twofold the control
values) at 3 h post-treatment, and then declined to control values
at 32–48 h post-treatment. Caspase-3 activity was also increased
1 h after FAA treatment, but contrary to caspase-1, its activity
remained elevated between 3 and 24 h and then peaked to its
maximal level (fivefold the control values) 32 h after FAA
treatment (Fig. 8A
, lower panel). Then the activity of
caspase-3 declined, but remained higher than control levels at 48 h post-treatment. A sustained activation of caspase-3 beyond 24 h
post-treatment was also confirmed by the decrease of its pro-form by
immunoblot analysis (see Fig. 8B
, upper panel). The cleavage
of PARP, an endogenous substrate for caspases-3 and -7
(27)
, also occurred in FAA-treated cells and was maximal
at 24–48 h post-treatment (data not shown).
|
Cytochrome c protein levels were measured by immunoblot
analysis of cytosolic extracts prepared under conditions that keep
mitochondria intact. As shown in Fig. 8B
(lower panel),
cytochrome c in FAA-treated cells accumulated slightly in
the cytosol at 3–6 h post-treatment. However, the release of
cytochrome c from mitochondria was maximal 24–32 h after
FAA exposure and lower levels were observed at 48 h
post-treatment. Cytochrome oxydase was absent in cytosolic extracts,
thus discarding the possibility of mitochondrial contamination (data
not shown).
The m was monitored in intact cells by
fluorescence of the lipophilic dye DiOC6[3], a
mitochondrial transmembrane potential-sensitive dye (21)
.
Compared to untreated cells, no reduction in
m was observed either before a recovery
period of 12 h after FAA treatment (data not shown) or between
12 h and 32 h post-treatment (Fig. 8C
). However,
at 48 h after FAA treatment, a population of cells with low
m was easily observed, thus indicating
mitochondrial transmembrane depolarization and dysfunction as usually
observed in apoptotic cells. This shift to weaker fluorescence was
similar to the decrease in fluorescence observed in cells treated with
the mitochondrial uncoupling agent mClCCP.
In summary, caspase-1 activation is an early event in the apoptotic
process induced by FAA, whereas the kinetics of caspase-3 activation
indicates a late involvement in this process, accompanying cytochrome
c release from mitochondria and preceding
m reduction.
Caspase inhibitors abrogate FAA-induced apoptosis and mitochondrial
transmembrane depolarization
To confirm that active caspases are involved in FAA-induced
apoptosis, exponentially growing V79 cells were pretreated or not with
the caspase inhibitors Ac-YVAD-CHO, z-DEVD-fmk, or z-VAD-fmk before FAA
treatment. The specificity of these inhibitors was for caspase-1,
caspase-3, and caspase proteases from a broad spectrum, respectively
(28
, 29)
. As shown in Table 1
, all the caspase inhibitors tested reduced or eliminated the fraction
of cells dying by apoptosis 48 h after FAA treatment (100 µM),
as indicated in the FACS analysis by the reduction of the cell
population with a sub-G1 DNA content. Both Ac-YVAD-CHO and z-DEVD-fmk
were more effective at the highest concentration used (800 µM), in
contrast to z-VAD-fmk, which already abrogated the sub-G1 peak induced
by FAA at a concentration of 100 µM. Similar results were obtained by
preincubating cell with the serine protease inhibitor TLCK (data not
shown). The caspase inhibitor z-VAD-fmk was used to test the
possibility that active caspases may induce the reduction in
m observed in the apoptotic cell population
after FAA treatment. As shown in Fig. 9
, preincubation of cells with z-VAD-fmk (100 µMx2 h) markedly
reduced the loss of m induced by FAA at
48 h post-treatment. Thus, active caspases are involved in the
cell death process induced by FAA; they impinge on mitochondria to
induce mitochondrial permeability transition and a subsequent loss in
m during FAA-induced apoptosis.
|
|
GSH-monoethylester or N-acetylcysteine abrogate FAA-induced
apoptosis
Since both the extent of G2/M arrest and subsequent apoptosis
induced by FAA appeared to be GSH dependent, we investigated whether
these effects of FAA could be modulated by GSH replenishment agents
such as GSH-MEE or NAC. Exponentially growing cells were preincubated
with these agents at different concentrations for 2 h and then
treated with FAA. As shown in Table 2
, both GSH-MEE and NAC markedly reduced the accumulation of cells in
G2/M induced by FAA at 24 h post-treatment and the sub-G1 peak
associated with apoptosis at 48 h post-treatment. These effects
were observed at the highest concentration of GSH-MEE (20 mM) or NAC (1
mM) used. Similar effects were obtained by treating cells concomitantly
with FAA and GSH (1 mM), but not when GSH was added immediately after
FAA treatment (data not shown). Thus, GSH has a protective role on cell
cycle disturbance and apoptosis induced by FAA, effects that seem to
result from a direct conjugation activity of GSH on FAA.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The present study shows that FAA is cytotoxic in
a dose- and GSH-dependent manner and that FAA-induced cytotoxicity
1) correlates with cell cycle arrest at G2/M and
2) occurs through an apoptotic pathway. The sequence of
molecular events that occur between the end of FAA treatment, the cell
cycle arrest at the G2/M transition at 24 h post-treatment, and
the late manifestation of apoptotic cell death (digestion of genomic
DNA at 48 h post-treatment) may be ordered as early events, which
include caspase-1 and cyclin B-dependent kinase activation, and late
events, which include cytochrome c release, caspase-3
activation, and m reduction.
The present data are consistent with the recent observation by
Kubo et al. (12)
that FAA induces the release of
cytochrome c from mitochondria in a cell-free system, but
they do not support the hypothesis proposed by these investigators that
the FAA-induced cytochrome c release from mitochondria as an
initial event by itself triggers the activation of the apoptotic
caspase cascade involving caspases-1 and -3. These investigators showed
the overall involvement of caspases-1 and -3 in apoptosis induced by
homogentisic acid in
FAH-/-/HPD-/-
hepatocytes or mice, but without directly assaying FAA in their
experiments. Assuming that FAA was the homogentisic acid-derived
metabolite responsible for caspase activation in the study by Kubo et
al. (12)
, our kinetic data point instead to a role of
caspase-1 activation, which peaks early (3 h) post-treatment, in the
upstream phase of apoptosis induced by FAA, prior to cytochrome
c release and mitochondrial dysfunction. Activated caspase-1
may function as the activator of downstream caspase-3, as already
shown, for example, in glucocorticoid-treated human T cells undergoing
apoptosis (30)
. Certainly, a fraction of FAA may attack
the mitochondria directly, causing cytochrome c release and
subsequent activation of a caspase cascade, but according to recent
work by Slee et al. (31)
, this cascade would not involve
caspase-1 activation. These investigators demonstrated that cytochrome
c in cytosol can activate through caspase-9 activation the
effector caspase-7 or initiate the hierarchical activation of caspase-3
as well as caspases-2, -6, -8, and -10, activation of the latter
caspases being dependent on caspase-3 activation. Apaf-1, a cytosolic
factor, is required for this cytochrome
c/caspase-3-dependent caspase cascade, which excludes the
ICE subfamily caspases (caspases-1, -4, and –5) (31)
.
The release of cytochrome c from mitochondria into the
cytosol and other mitochondrial changes may play a central role in the
regulation and activation of the executioner phase of apoptosis induced
by FAA. This is supported by the fact that the kinetics of cytochrome
c release closely follows that of caspase-3 activation, both
events peaking relatively late in the death process (at 24 and 32 h post-treatment, respectively) and preceding the time of genomic DNA
fragmentation. The dependency of the executioner phase of apoptosis,
which includes caspase-3 activation, on cytochrome c release
from mitochondria has been demonstrated in other studies (32
, 33)
. Since a reduction in m was
observed only late (at 48 h post-treatment) in the apoptotic
process induced by FAA, it seems that cytochrome c is
released from coupled mitochondria. In fact, the breakdown in
m, which indicates mitochondrial
dysfunction, was not found until the time of DNA fragmentation, well
after the first morphological evidence of apoptosis obtained by
fluorescence microscopy. Thus, mitochondrial transmembrane
depolarization is not a critical step in the commitment of cells to die
after FAA treatment. Instead, mitochondrial dysfunction seems to
reflect rather than predict the apoptotic death of cells treated with
FAA. A late decrease in m either during or
after DNA fragmentation has been observed in other situations, such as
after UVB irradiation or staurosporine treatment of CEM or HeLa cells,
or in nerve growth factor-deprived neurons (34
, 35)
.
Certainly, the block of FAA-induced m
reduction after preincubation of cells with the general caspase
inhibitor z-VAD-fmk may be due to a primary action of this inhibitor on
upstream caspases such as caspase-8 (36)
. However, the
involvement of early caspase activation events other than caspase-1 or
that of other apoptosis-transducing signals in the upstream phase of
the apoptotic process induced, directly or indirectly, by FAA remains
to be elucidated.
The early induction by FAA of cyclin B-p34cdc2 kinase activity, whose
peak correlated with maximal accumulation of cells in G2/M, and the
sequential apoptotic death of G2/M-arrested cells suggest the
involvement of cell cycle-related events in FAA-induced apoptosis. Both
cell cycle arrest and cyclin-dependent kinase activation have been
demonstrated to occur and to be a requisite step in several, but not
all, instances of apoptotic cell death. This requirement was observed
in the apoptotic pathway induced by the glucocorticoid dexamethasone or
APO-1(Fas/CD95), which involves serial activation of ICE-related
caspases and caspase-3, or by granzyme B, a direct activator of
caspase-3 (30
, 37
, 38)
. However, in all these cases the
molecular signaling process that links cell cycle-related events with
apoptosis remains unclear. Unscheduled or premature activation of
cyclin B-p34cdc2 kinase leading to apoptosis has been observed after
treatment of cells with agents known to damage DNA, such as cisplatin
or etoposide (39
, 40)
. Another DNA-damaging agent,
nitrogen mustard, has been shown to hyperactivate cyclin B-p34cdc2
kinase after its initial and transient inactivation (41)
.
Similar sequential events were observed after FAA treatment of cells.
The relatively high levels of cyclin B-p34cdc2 kinase activity observed
48 h after FAA treatment may reflect the slight accumulation of
cyclin B observed at this time, but whether this results from increased
synthesis of cyclin B or inhibition of its degradation was not
examined. Although we have not yet been able to demonstrate that FAA
directly damages DNA, its mutagenic activity in V79 cells
(10)
supports its action on DNA. Activation of the G2/M
transition checkpoint is a common mechanism by which cells respond to
genotoxic stress. The resulting G2/M arrest is thought to allow repair
of DNA damage prior to cell division and ensure that DNA replication
will proceed with fidelity to avoid segregation of defective
chromosomes and potential tumorigenesis (42)
. The specific
tyrosine phosphatase cdc25C involved in the activation of p34cdc2 has
been identified as one main target of the DNA damage checkpoint
(43
, 44)
. However, as reported recently, it seems that
multiple pathways may be involved in initiating the G2/M checkpoint and
modulating cyclin B-p34cdc2 kinase activity depending on types of DNA
damage (45)
. Which types of DNA damage or G2/M checkpoint
pathway are induced by FAA remains to be determined.
Cellular GSH levels seem to modulate the cytotoxic effects of FAA (G2/M
arrest and apoptosis) in a manner similar to that previously suggested
for its mutagenicity, that is, through a buffering-like activity
(10)
. This means that the greater the free/effective FAA
dose that depends directly on the amount of cellular GSH available for
FAA conjugation, the greater the cytotoxicity of FAA. This mechanism is
suggested by the similarities between the cytotoxic/cytostatic effects
induced by a relative low dose of FAA (35 µM) with BSO pretreatment
and by a high dose of FAA (100 µM) without BSO pretreatment. A direct
conjugation activity of GSH on FAA is also supported by the fact that
the addition to cells of free GSH concomitantly with FAA abrogates the
cytotoxic/cytostatic effects of FAA (R. Jorquera and R. M.
Tanguay, unpublished observations). However, there is also the
possibility that FAA-induced apoptosis could be mediated by a
GSH-dependent process. A similar potentiating effect of cellular GSH
depletion on apoptosis induction has been reported in cells treated
with diamide, a thioloxidant agent (46)
. In this
situation, the effective dose of diamide necessary to induce apoptosis
and related events, such as mitochondrial cytochrome c
release and caspase-3 activation, was markedly reduced in
BSO-pretreated cells. Other reports demonstrate that GSH or other
thiol-reducing agents may influence the apoptotic response through, for
example, its participation in the redox regulation of caspase-3
activity (46
, 47)
. At the dose used here (0.2 mM), BSO per
se (i.e., GSH depletion) does not have cytotoxic effects or alter cell
growth or cell cycle progression. Moreover, it has been reported that
preculture of human cells with BSO (5 mM) itself caused neither the
induction of apoptosis nor activation of caspase-3 (46)
.
The requirement of maintenance of a cellular reducing environment for
counteracting the cytotoxic/cytostatic effects of FAA is well
demonstrated in our study by the beneficial action of preincubating
cells with GSH-MEE or NAC. In this case, both cell cycle block and
apoptosis induced by FAA are abrogated. We believe that these agents
exert their anti-apoptotic effect through their GSH replenishment
activity rather than by acting as ROS scavengers, since we have
previously shown that ROS are not generated in FAA-treated cells
(10)
.
Taken together, our previous (10)
and present results show
dual but opposite activities of FAA as far as the cancer induction
process is concerned. On one side, FAA has a mutagenic activity that
could be involved in the initiation of the carcinogenic process in the
liver of HT I patients. On the other hand, FAA induces apoptosis, which
in theory should help to eliminate mutated cells. However, this
cytotoxic activity of FAA may trigger a regenerative process in the
liver and, after multiple, chronic rounds of hepatocyte
loss/proliferation, contribute to cirrhosis establishment, which may
favor fixation of oncogenic mutations. Through its mutagenic activity
or via a metabolic effect, FAA may target genes or proteins essential
for defense mechanisms such as DNA repair. A recent study has shown
that SA, a derivative of FAA which accumulate in HT I, has an
inhibitory activity on DNA-ligase, a DNA repair enzyme. Moreover, the
activity of this enzyme was found to be low in HT I fibroblasts
(48)
. In line with these observations, we have previously
suggested (10)
that FAA, through its GSH-depleting
activity, could impair DNA repair processes, as reported to occur in
human cells with low GSH levels due to a genetic deficiency in GSH
synthetase (49)
. Thus, abnormal DNA repair in HT I
patients could predispose them to cancer induction, where FAA acts as
an hepatocarcinogen. Although the risk of hepatocellular carcinoma such
as that occurring in ‘chronic’ HT I could be decreased by
therapeutically increasing the level of apoptotic death of hepatocytes,
the risk of hepatic dysfunction, such as occurs in ‘acute’ HT I,
might also increase. Thus, we believe that any potential HT I therapy
focused to manipulate the apoptotic process in the liver should be
postponed until the identification of molecular/biochemical markers
that will render the pathophysiological manifestations of the disease
more predictable.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
