Cell Death Study

(1)

A. M. A. Nasirudeen, Yap Eu Hian, Mulkit Singh and Kevin S. W. Tan

Correspondence Correspondence Kevin S. W. Tan Kevin S. W. Tan [email protected] [email protected]

Department of Microbiology, Faculty of Medicine, National University of Singapore, 5 Science

Drive 2, Singapore 117597

Received

Received 16 May 200316 May 2003 Revised

Revised 26 August 200326 August 2003 Accepted

Accepted 15 September 200315 September 2003

Previous studies by the authors have shown that the protozoan parasite

Previous studies by the authors have shown that the protozoan parasite Blastocystis hominisBlastocystis hominis

succumbed to a cytotoxic monoclonal antibody with a number of cellular and biochemical

features characteristic of apoptosis in higher eukaryotes. The present study reports that

apoptosis-like features are also observed in growing cultures of axenic

apoptosis-like features are also observed in growing cultures of axenic B. hominisB. hominisupon exposureupon exposure

to metronidazole, a drug commonly used for the treatment of blastocystosis. Upon treatment

with the drug,

with the drug, B. hominisB. hominis cells displayed key morphological and biochemical features ofcells displayed key morphological and biochemical features of

programmed cell death (PCD), viz. nuclear condensation and nicked DNA in nucleus, reduced

cytoplasmic volume, externalization of phosphatidylserine and maintenance of plasma membrane

integrity with increasing permeability. This present study also supports the authors’ previously

postulated novel function for the

postulated novel function for the B. hominisB. hominiscentral vacuole in PCD; it acts as a repository wherecentral vacuole in PCD; it acts as a repository where

apoptotic bodies are stored before being released into the extracellular space. The implications

and possible roles of PCD in

and possible roles of PCD in B. hominisB. hominisare discussed.are discussed.

INTRODUCTION

Programmed cell death (PCD) has been noted in all

inver-tebrate and verinver-tebrate multicellular organisms studied so

tebrate and vertebrate multicellular organisms studied so

far, including nematodes, insects, amphibians and

mam-mals (Ellis

mals (Elliset al et al ., 1991; Raff, 1992; Vaux, ., 1991; Raff, 1992; Vaux, 1993; Steller, 1995).1993; Steller, 1995).

It

It is is ththouought ght to to hahave ve evevololveved d to to reregugulalate te grgrowowth th anandd

development in multicellular organisms (Evan, 1994; Vaux

development in multicellular organisms (Evan, 1994; Vaux et al

et al ., 1994) and as a defence mechanism against viral and., 1994) and as a defence mechanism against viral and

bacte

bacterial infectiorial infections ns (She(Shen n & & ShenkShenk, , 1991995; 5; WeinWeinrauch & rauch &

Zychl

Zychlinskyinsky, , 19991999). ). UnneUnnecessacessary, ry, damagdamaged, ed, infecinfected ted andand

pot

potententialially ly harharmfumful l cecells lls are are deldeleteted ed frofrom m sursurrouroundindingng

healthy ones to ensure structural and functional

homeo-sta

stasissis. . ApoApoptoptosissis, , a a forform m of of PCDPCD, , is is chacharacracterterizeized d by by aa

unique pattern of morphological changes in both the cell

nucleus and the cytoplasm. Although different cell types

do not necessarily display all the hallmarks of apoptosis,

cha

characracterterististic ic feafeaturtures es appappear ear to to be be coconsenserverved d in in celcellsls

undergoing this form of cell death. These include shrinkage

of

of ththe e cecellll, , prpreseserervavatition on of of memembmbrarane ne inintetegrgritity y wiwithth

increasing permeability, chromatin condensation and DNA

fragm

fragmentatientation, on, and and exteexternalizrnalization ation of of plasmplasma a membmembranerane

phosphatidylserine (PS) residues (Saraste & Pulkki, 2000).

Alt

Althouhough gh it it was long was long assassumeumed d thathat t PCD evolPCD evolved ved wiwithth

multicellularity, several recent reports have suggested that

similar cell-death machinery exists in protozoans such as

similar cell-death machinery exists in protozoans such as Leishmania

Leishmania ((Leishmania Leishmania ))amazonensis amazonensis (Moreira(Moreiraet al et al ., 1996),., 1996),

Trypano

Trypanosoma soma cruzi cruzi (Ameisen(Ameisen et al et al ., ., 1991995),5), Trypanosoma Trypanosoma

brucei

brucei rhoderhodesiense siense (Welburn(Welburn et al et al ., ., 19961996),), DictyosteliumDictyostelium

discoideum

discoideum(Cornillon(Cornillonet al et al ., 1994),., 1994), Plasmodium falciparumPlasmodium falciparum

(Picot

(Picotetal etal .,.,19199797))andandPlasmodPlasmodiumiumberghei berghei (Al-Olayan(Al-Olayanetal etal .,.,

2002).

In our earlier study (Nasirudeen

In our earlier study (Nasirudeen et al et al ., 2001a), we showed., 2001a), we showed

that

that the the human intestihuman intestinal nal protoprotozoan parasitezoan parasite Blastocystis Blastocystis

hominis

hominis undundergergoeoes s PCD PCD whewhen n tretreateated d witwith h a a sursurfacface-

e-rea

reactictive ve cytcytotootoxic xic mAb mAb (1D(1D5)5), , witwith h typtypicical al feafeaturtures es of of

apo

apoptoptotic tic celcells. ls. AgeAgeing ing agaagar r culculturtures es of of B. B. homhominis inis alsoalso

displayed ultrastructural features of apoptosis (Tan

displayed ultrastructural features of apoptosis (Tan et al et al .,.,

2001). The presence of an apoptotic response to

undesir-ab

able le exexteternarnal l ststimimululi i leled d us us to to ininveveststigigatate e if if a a sisimimilarlar

phenot

phenotype is ype is obseobserved upon rved upon drug exposurdrug exposure. e. DrugDrugs s suchsuch

as staurosporine have been shown to cause cell death in

mammalian cell lines (Nakazono-Kusaba

mammalian cell lines (Nakazono-Kusaba et al et al ., 2002; Koh., 2002; Koh et al

et al ., 1995). In protozoan parasites such as., 1995). In protozoan parasites such as P. P. falcipfalciparumarum,,

chloroquine has been reported to induce cell death (Picot

chloroquine has been reported to induce cell death (Picot et al

et al ., ., 1991997).7). T. T. brubrucei cei rhorhodesdesiensiense e underundergoes goes apoptapoptosisosis

when treated with concanavalin A (Welburn

when treated with concanavalin A (Welburn et al et al ., 1996).., 1996).

Free-living unicellular organisms have also been reported to

undergo apoptosis. Davis

undergo apoptosis. Davis et al et al . (1992) observed apoptotic-. (1992) observed

apoptotic-like nuclear degeneration in conjugating

like nuclear degeneration in conjugating Tetrahymena Tetrahymena cellscells

while Ludovico

while Ludovico et al et al . (2001a, b) reported that PCD can be. (2001a, b) reported that PCD can be

induced by acetic acid in

induced by acetic acid in Saccharomyces cerevisiae Saccharomyces cerevisiae . Taken. Taken

together, these observations suggest that many unicellular

organisms have the unusual capacity to activate a complex

cell-cell-deathdeathprogrprogramme samme similimilarartototheitheirrmultimulticellcellularularcountcounter-

er-parts. This feature has been postulated to aid these

parts. This feature has been postulated to aid these

micro-org

organianisms sms in in reregulgulatiation on of of grogrowth wth and and devdeveloelopmepmentantall

Abbreviations:

Abbreviations: PCD, programmed cell death; PI, propidium iodide; PS,PCD, programmed cell death; PI, propidium iodide; PS, phos

phosphatiphatidylsdylserinerine; e; TUNELTUNEL, , termterminal inal deoxdeoxynuynucleotcleotidyl idyl transtransferasferase- e-mediated deoxyuridine triphosphate-biotin nick-end labelling.

(2)

processes, and may also provide us with clues on host– parasite interactions and pathogenesis (Knight, 2002). The present study demonstrates characteristic features of PCD in the human intestinal protozoan parasite B. hominis

after exposure to 561027 M metronidazole.

Metronida-zole is a nitroimidaMetronida-zole antibiotic used clinically to treat

B. hominis infections as well as for the treatment of anaero-bic infections in humans (Garavelli, 1991; Hager & Rapp, 1992). Dunn & Boreham (1991) developed an assay to measure the sensitivity of drugs against B. hominis and arrived at an ID50value of3?346102

7

M for metronidazole. The pathogenicity of B. hominis is a controversial issue: some authors consider it to be a pathogen (Lee, 1991; Llibre

et al ., 1989; Ok et al ., 1997, 1999), whereas others conclude that it is a harmless commensal (Rosenblatt, 1990; Sun et al ., 1989). Symptoms reported in B. hominis infection include diarrhoea, abdominal pain, ﬂatulence, nausea and con-stipation (Ok et al ., 1999). Several reports show the parasite to be more pathogenic in immunosuppressed patients such as those with AIDS (Llibre et al ., 1989). Although metronidazole has been used clinically as a drug to treat

B. hominis infections, little effort has been made to evaluate the pattern of death induced by this drug. In the present study, we have employed ﬂow cytometry and electron micro-scopy to study the effects of metronidazole on B. hominis .

METHODS

Parasite and growth conditions. B. hominis isolate B was obtained from a local patient and maintained axenically in Iscove’s modiﬁed Dulbecco’s medium (IMDM) containing 10% horse serum (Ho et al ., 1993). Cells were grown anaerobically (Concept Plus Anaerobic Workstation; Ruskinn Technology) at 37uC.

Metronidazole.Metronidazole, 2-methyl-5-nitroimidazole-1-ethanol, is a 5-nitroimidazole drug used in the treatment of blastocystosis and other anaerobic infections. Stock solutions of metronidazole M-1547 (Sigma) were prepared in double distilled water and further diluted in IMDM to obtain the desired concentration.

Treatment with metronidazole to induce cell death. Metro-nidazole (561027_{M) was used to induce cell death in B. hominis .}

Cells were inoculated into pre-reduced IMDM to give a ﬁnal con-centration of 26106cells ml21. Metronidazole was then added to a

ﬁnal concentration of 561027M. Two controls were included in

the experiments: (i) cells grown in normal culture conditions (IMDM with 10 % horse serum); (ii) necrotic cell control of B. hominis was obtained by incubating cells with 0?1 % sodium azide

and harvesting 24 h post-treatment. Cells were harvested at 3, 6, 9 and 12 h for ﬂow cytometry analysis and phase-contrast microscopy. Similar conditions were followed for ultrastructural studies except that necrosis was induced by heating the cells in an 80uC water bath for 15 min.

Membrane integrity analysis and detection of PS externaliza-tion. The detection of loss of membrane permeability and exposure of PS was detected using an Annexin V/FITC kit (BenderMed Systems) as described previously (Nasirudeen et al ., 2001a). Brieﬂy, 26106cells incubated in the presence and absence of metronidazole

were harvested at 3, 6, 9 and 12 h. Cells were washed twice in 1 ml PBS (pH 7?4). Then, 190 ml calcium-containing binding buffer,

0?21 mg ml21 FITC–Annexin V and 2?5mg ml21 propidium iodide

(PI) were added sequentially. The samples were measured by ﬂow cytometry (Coulter Epics Elite ESP) for FITC/PI ﬂuorescence and results were displayed using the WINMDI 2.7 software program.

Exclusion of PI from these cells was taken as a quantitative marker for membrane integrity (Cornillon et al ., 1994).

DNA ladder assay. In the presence and absence of metronidazole, 56106cells ml21 were incubated and harvested at 3, 6, 9 and 12 h

intervals. An apoptotic DNA ladder kit (Boehringer Mannheim) was used to extract DNA from apoptosis-induced and uninduced cells according to the manufacturer’s instructions. Apoptotic U937 cells (histiocytic lymphoma cells), included in the kit, were used as a positive control. DNA was electrophoresed in 2 % agarose gels at 100 V for 2 h and visualized by using an UV transilluminator; the gels were photographed with a Polaroid camera.

Nick-labelling of internucleosomal DNA fragments. Terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate-biotin nick-end labelling (TUNEL) was performed using the ApoAlertTM

DNA Fragmentation Assay kit (Clontech). Metronidazole-treated and unMetronidazole-treated cells (26106cells ml21) were washed twice

with 1 ml PBS and ﬁxed with 4 % formaldehyde/PBS for 25 min at 4uC. After two washes with PBS, the pellet was resuspended in 5 ml permeabilization solution (0?2 % Triton X-100 in PBS) and

incu-bated on ice for 5 min. Eighty microlitres of equilibration buffer were added and the cells were incubated at room temperature for 5 min. The cells were labelled by adding 50 ml TUNEL mix followed

by incubation for 60 min at 37uC in a dark, humidiﬁed incubator. One millilitre of 20 mM EDTA was then added to terminate the tailing reaction. The samples were washed with PBS and the pellet was resuspended in 250ml PBS prior to ﬂow cytometry using an

argon-ion laser tuned to 488 nm. Green fluorescence was detected using a 525 nm band-pass filter. A region (M1) was defined to exclude background fluorescence by unstained cells. Stained para-sites that fell within the M1 region were represented as a percentage of total cells analysed.

Transmission electron microscopy analysis. In the presence and absence of metronidazole, 16107 cells were incubated and

harvested at 3, 6, 9 and 12 h intervals. Necrosis was induced by heating the cells in an 80uC water bath for 15 min. Cells were washed twice in 0?1 M sodium cacodylate buffer with 5 % sucrose

and ﬁxed sequentially in each of the following ﬁxatives: 1 % glutaral-dehyde in 0?5 M sodium cacodylate buffer pH 7?2 for 3 h, followed

by two washes in buffer; 2 % osmium tetroxide in 0?1 % potassium

ferrocyanide solution for 1 h and then two washes in 0?5 M

cacody-late buffer. The cells were then processed according to the procedure used by Moe et al . (1999) and examined using a Philips EM280S electron microscope.

Reproducibility of results. All experiments were repeated at least twice, except for transmission electron microscopy analysis, which was performed once.

RESULTS

Cell shrinkage in metronidazole-treated cells

Phase-contrast microscopy revealed that B. hominis cells grown in normal culture conditions displayed healthy morphology, as indicated by a spherical shape, clear central vacuole and peripheral cytoplasm (Fig. 1a). Metronidazole-treated B. hominis cells, however, appeared smaller and darker (Fig. 1b). Cell shrinkage was conﬁrmed using ﬂow

(3)

Fig. 1. Effect of metronidazole on B. hominis cell size as determined by light microscopy and ﬂow cytometry. (a) Cells grown in normal culture conditions; (b) cells exposed to 5610”7M metronidazole. Note cell shrinkage, condensed

cyto-plasm and darkening of cells after exposure to metronidazole. Bars, 20mm. (c) Flow cytometry analysis of cell size and

granu-larity. Dot plots show two distinct populations of cells. The population on the left (R2; green) is due to dead cells and debris while the population on the right (R1; red) could be a mixture of healthy cells and cells undergoing PCD.

(4)

cytometry where metronidazole-treated cells showed pro-gressive reduction in cell size (Fig. 1c). Dot plot displayed two distinct populations of cells. The population on the right (R1-gated) represents larger cells (healthy and early apoptotic cells) while the population on the left (R2-gated) represents late apoptotic and necrotic cells, and cellular debris.

Preservation of membrane integrity and

externalization of PS in cells undergoing PCD

Loss of cell viability is most often measured as loss of membrane integrity. A short incubation with PI was used to investigate membrane integrity in B. hominis cells under-going PCD. PI can only enter cells with an altered plasma membrane and then stain the nucleic acid, giving red ﬂuorescence. Exclusion of PI by membrane-intact cells was scored as preservation of membrane integrity. This dye was used together with Annexin V, which binds PS exposed during apoptosis. The results were analysed by ﬂow cyto-metry (Fig. 2): 2?87 % of the metronidazole-treated cells

were PI-positive at 3 h and 4?60 % were PI-positive at 12 h.

Cells grown in normal culture media showed 1?45%

PI-positive cells at 3 h and 1?13 % PI-positive cells at 12 h.

Flow cytometry data revealed a gradual increase in dying cells (increase in cell permeability) due to metronidazole treatment. Approximately 30 % of the B. hominis cells were PI-positive when treated with 0?1 % sodium azide for 24 h.

In normal cells, under physiological conditions, PS is segregated to the inner leaﬂet of the plasma membrane. During the early stages of apoptosis/PCD, this asymmetry collapses and PS is exposed onto the outer surface of cells

(Gatti et al ., 1998) while maintaining membrane integrity. Annexin V (a PS-binding protein) preferentially binds PS in a calcium-dependent manner. An Annexin V assay was used to detect the presence of PS on the plasma membrane, while the simultaneous presence and absence of PI accu-mulation indicates apoptotic or necrotic states, respectively. Increase in PS translocation was detected by ﬂow cytometry in metronidazole-treated cells over a period of 9 h. As shown in Fig. 3, before treatment with metronidazole, only 2?57 % of the cells were Annexin V-positive. However,

within 3 h of metronidazole treatment, a population of cells with dramatically increased Annexin V binding began to emerge in these cultures and increased rapidly thereafter. Annexin V binding in these cells increased from 32?39% at

3 h to 51?09 % at 9 h as compared to 2?72 % of control cells

at 9 h. At 12 h, the Annexin V binding of metronidazole-treated cells dropped to 30?78 %. Necrotic cells showed only

9?90 % Annexin V binding.

Absence of DNA laddering in metronidazole-treated B. hominis _cells

DNA fragmentation and the display of DNA ladders, often of 200 bp and multiples thereof, in agarose gel electro-phoresis are common features of apoptosis. Agarose gel electrophoresis of B. hominis DNA in the presence and absence of metronidazole and in the presence of 0?1 %

sodium azide is shown in Fig. 4. No DNA ladder was observed in any of the experiments. A positive control consisting of apoptotic U937 cells (Boehringer Mannheim) showed a DNA-laddering proﬁle typical of apoptosis (Fig. 4).

Presence of in situ _{DNA fragmentation in the}

nucleus of metronidazole-treated cells

Endonuclease activity was evaluated with the TUNEL assay. TUNEL relies on the speciﬁc binding of terminal deoxy-nucleotidyl transferase to exposed 39-OH ends of DNA

followed by the synthesis of a labelled polydeoxynucleotide molecule. Nuclear DNA is first exposed by proteolytic treatment; this is followed by the incorporation of biotiny-lated deoxyuridine triphosphate by terminal deoxynucleo-tidyl transferase into the sites of DNA breaks. The signal is amplified by avidin–peroxidase, enabling conventional histochemical identification of PCD by flow cytometry. Approximately 12?4 % of metronidazole-treated cells were

TUNEL labelled by 3 h and 58?8 % were maximally labelled

by 9 h. Control cells in normal growth media showed minimal DNA fragmentation. Formation of DNA strand breaks as a consequence of endonuclease activity can be detected by the TUNEL assay and expressed as a percent-age of TUNEL-positive cells. Such activity in parasites cultured in normal growth media was approximately 2?0

and 3?1 % at 3 and 9 h, respectively. At 12 h, TUNEL

of metronidazole-treated B. hominis dipped to 34?8 %.

B. hominis cells that were exposed to 0?1 % sodium azide

(necrotic control) showed 11 % TUNEL (Fig. 5).

Fig. 2. Membrane integrity assay of B. hominis cells by ﬂow cytometry. Note gradual increase in PI uptake in metronidazole-treated cells. Necrosis was achieved by treating cells with 0?1 % sodium azide for 24 h and the PI-positive value is

repre-sented as a horizontal line for comparison. X, Metronidazole-treated cells; &, untreated cells. Results are based on two replicates and are shown ±SD.

(5)

Ultrastructural features of apoptosis in metronidazole-treated cells

Transmission electron microscopy micrographs showed that cultures in normal growth media had classical B. hominis morphology (Fig. 6a). Normal DNA chromatin, which is usually seen as a crescentic mass at the nuclear periphery, was visible within the nucleus (Fig. 6a). In contrast,B. hominis cells treated with metronidazole showed

morphological changes suggestive of apoptosis. In the presence of metronidazole, the nuclei appear distinctively smaller and the nuclear chromatin had segregated as distinct clumps along the nuclear periphery (Fig. 6b). The presence of large cytoplasmic vacuoles, appearing empty or containing some inclusions, the invagination of the cyto-plasm into the central vacuole (Fig. 6c) and the presence of membrane-bound electron-dense particles (apoptotic bodies) in the central vacuole (Fig. 6d) were noted. The apoptotic bodies in the central vacuole appeared to be released into the extracellular space. At 12 h, a number of cells showed a break in the central vacuole, suggesting that most or all of the apoptotic bodies had been released (Fig. 6e). Cells that underwent necrosis presented a morphology distinctively different from that of apoptotic

B. hominis . Necrotic cells were electron-lucent and showed swelling of organelles such as mitochondria and nucleus, and the central vacuole seemed devoid of any electron-dense particles (Fig. 6f).

DISCUSSION

Metronidazole is an antibiotic used for the treatment of anaerobic microbes including B. hominis . Dunn & Boreham (1991) showed that metronidazole treatment of B. hominis

impaired its growth at an ID50value of 3?346102 7

M. We therefore reasoned that the concentration of 561027 M

used in this study should be sufﬁcient to induce cell death in a majority of the parasites in culture.

Metronidazole enters cells and mitochondria by simple diffusion. In mitochondria, metronidazole competes efﬁ-ciently with the natural electron acceptor for electrons.

1 2 3 4 5 6 7 8 9 10 11 12

1000

500 bp

Fig. 4. Agarose gel electrophoresis of B. hominis DNA. Lanes: 1, 100 bp marker; 2, apoptotic U937 cells; 3–7, respectively, control untreated parasites 0, 3, 6, 9 and 12 h; 8–11, respec-tively, 3, 6, 9 and 12 h metronidazole treatment; 12, 0?1 %

sodium azide treatment. Note the absence of a DNA ladder in lanes 3–12.

Fig. 3. Representative dot plots and graphical representation of similar experiments showing externalization of PS in B. hominis cells grown in normal culture conditions ( &) and those exposed to metronidazole ( X). Metronidazole-treated cells show a signiﬁcantly higher percentage of Annexin V-labelled cells relative to the control cells. In the lower graph, results are given based on two replicates and are shown ±SD.

(6)

Organisms susceptible to 5-nitroimidazoles transfer elec-trons generated by their electron-transport systems to the nitro group of the drugs and not to their natural electron acceptor (Kulda, 1999). The reduction of the metronidazole nitro group results in the synthesis of cytotoxic radicals (R-NO2

2) (Kulda, 1999). Hence, the antimicrobial effect of

metronidazole depends on its metabolic reduction within the target cell resulting in the release of cytotoxic radicals (Kulda, 1999).

The present study shows that metronidazole causes apoptotic-like death in B. hominis . In an earlier study, we described similar apoptotic-like features in this organism in response to a cytotoxic antibody exposure (Nasirudeen

etal ., 2001a). When treated with 561027 M metronidazole,

light microscopic data showed cell shrinkage and darken-ing of the cytoplasm in B. hominis . Cell shrinkage due to compaction of organelles in the cytoplasm represents important morphological evidence of apoptosis. This could be due to loss of cytoplasmic ﬂuids and denaturation of proteins in apoptotic cells (Huppertz et al ., 1999). Flow cytometry data further supported cell shrinkage observed under light microscopy.

In apoptosis research, electron microscopy is often employed to observe key ultrastructural features. Trans-mission electron micrographs of metronidazole-treated

B. hominis clearly showed ultrastructural characteristics of apoptosis. Nuclear condensation, cell shrinkage, deposi-tion of membrane-bound apoptotic bodies, maintenance of cytosolic organelle structure and size (unlike necrosis where the organelles swell and appear disoriented) and heavy vacuolization clearly provide ultrastructural evidence of apoptosis in metronidazole-treated B. hominis .

Flow cytometry analysis revealed that metronidazole-treated B. hominis preserved its membrane integrity with PI-exclusion levels comparable to those of normal

B. hominis cells (<10 % of cells PI-positive). These were much lower than those observed in control necrotic cells, which gave PI levels of approximately 30 %. Upon exposure to metronidazole, only a small percentage of cells became permeable to PI while most of the cells maintained membrane integrity. In cells grown in normal media, there was no signiﬁcant increase in the uptake of PI, indicating normal growth and an intact plasma membrane.

Fig. 5. Representative histograms and graph showing in situ DNA fragmentation analysis (TUNEL) of B. hominis cells by ﬂow cytometry. Note signiﬁcant TUNEL intensity increase in metronidazole-treated cells. &, Metronidazole-treated cells;

X, untreated cells. In the lower graph, results are given based on three replicates and are shown ±SD.

Fig. 6.Transmission electron micrographs of B. hominis cells exposed to metronidazole for 3–12 h. (a) HealthyB. hominis cell displaying normal morphology and DNA chromatin seen as a crescentic mass (arrow). (b) Segregation of condensed chromatin to the nuclear periphery and condensation of nucleus (arrow), resulting in a large intramembranous space. (c) Large cytoplasmic vacuoles (V), cytoplasm pinching inwards into central vacuole (arrow). (d) Membrane-bound organelle-containing vesicles within the central vacuole. (e) At 12 h, the cells have apparently released most or all of their apoptotic bodies. (f) Necrotic cells with swollen mitochondria and nucleus and chromatin seen in large clumps in necrotic cells. Note clear central vacuole in necrotic cells. N, nucleus; CV, central vacuole; M, mitochondria; V, vacuole. Bars, 1 mm.

(7)

(a) (b)

(c) (d)

(8)

In higher eukaryotes, apoptotic cells exhibit externalization of PS in the early stages of cell death while maintaining membrane integrity (Gatti et al ., 1998). This can easily be detected by Annexin V staining. At 9 h exposure to metronidazole, Annexin V-positive B. hominis cells had increased 10-fold from ~5 to 50%, while membrane integrity was largely intact. At 12 h of metronidazole treatment, the number of B. hominis cells stained with Annexin V decreased. This decrease in Annexin V staining could be largely due to the release of apoptotic bodies where part of the cell membrane might be lost in the pro-cess (Nasirudeenet al ., 2001a). Necrotic B. hominis showed low Annexin V staining. In this study, cells were treated with 0?1 % sodium azide for 24 h to induce necrosis. As

such, PS residues may have already been degraded by cytosolic lipases, which may explain the relatively low Annexin V staining in necrotic cells. These results suggest that PS exposure precedes the loss of membrane integrity by several hours. The preservation of membrane integrity and the simultaneous exposure of PS to the outer leaﬂet of the plasma membrane (in cells treated with metronidazole) clearly demonstrate the apoptotic effect of the drug on

B. hominis .

In mammalian apoptotic machinery, externalization of PS is necessary to signal neighbouring immune cells (Adayev

et al ., 1998). The immune cells then phagocytose the apop-totic cells and/or apopapop-totic bodies to minimize damage to the neighbouring healthy cells. The role of PS externaliza-tion in Blastocystis has yet to be determined. Perhaps,

in vivo , it could be used for attracting neighbouring phagocytic cells. Penfold & Provis (1986) reported that most of the cellular debris resulting from cell death of the retina is taken up by adjacent cells rather than by macro-phages. Without proper disposal of dying cells, the cells or the apoptotic bodies could undergo secondary necrosis. Hence, the externalization of PS residues during apoptotic death of B. hominis cells could have a similar function in the clearance of apoptotic bodies.

The appearance of the nucleosomal DNA ladder on an agarose gel was regarded as a hallmark of apoptosis (Collins et al ., 1992). DNA extracts of metronidazole-treated B. hominis did not result in any apparent ladder-like fragmentation such as that seen in many apoptotic cells. But it has also been shown that internucleosomal DNA fragmentation cannot always be regarded as a hall-mark of apoptosis since certain cells display morphological and biochemical features of apoptosis without ladder-like DNA fragmentation (Howell & Martz, 1987; Barbieri et al ., 1992; Collins et al ., 1992; Mesner et al ., 1992; Falcieri et al ., 1993; Cornillon et al ., 1994; Vaux et al ., 1994; Hirata

et al ., 1998). Furthermore, Darzynkiewicz et al . (1992) showed that DNA degradation, in many cell types, does not proceed to nucleosomal-sized fragments but rather results in 50–300 kb range DNA fragments which do not readily generate a characteristic ‘ladder’ pattern during agarose gel electrophoresis. In our previous studies, on the exposure

of a cytotoxic antibody to B. hominis (Nasirudeen et al ., 2001a), no DNA ladder was noted in apoptotic B. hominis

except for two distinct fragments. However, when cells were treated with 10 mg ml21 Rnase A for 3 h at 30uC,

these two fragments were absent after electrophoresis (data not shown). Hence, the two bands from the earlier run correspond to contaminating rRNA fragments.

Though agarose gel electrophoresis of B. hominis DNA did not show typical ladder-like DNA fragmentation pattern, TUNEL indicated that metronidazole-treated cells do indeed undergo in situ DNA fragmentation. Flow cytometry data showed increasing TUNEL staining of metronidazole-treated B. hominis up to 9 h. At 12 h post-treatment, TUNEL intensity dipped. This could be due to secondary necrosis where much of the nuclear material may have been lost in apoptotic bodies. Normal B. hominis

cells and those exposed to 0?1 % sodium azide (necrosis

control) showed only minimal TUNEL ﬂuorescence. TUNEL detects 39-OH groups at the end of single-strand

and double-strand DNA cuts. Didenkoetal . (2003) reported that DNA cleavage in early necrosis is characterized by selective generation of 59 overhangs, but not 39 overhangs.

This may be the reason why necroticB. hominis cells showed low TUNEL. Another reason could be that much of the nuclear material may have degraded during the 24 h necrosis induction and therefore have been lost during the washing of cells.

What accounts for the inconsistency between the TUNEL and agarose gel electrophoresis methods? Apoptosis may have occurred in an asynchronous fashion so that distinct accumulation of DNA fragments was not observed. Another possible reason for the lack of DNA laddering in dying

B. hominis cells could be the low sensitivity in the method of detection. In apoptotic chloroquine-sensitive P. falciparum, DNA laddering was reported using autoradio-graphy methods when no visible DNA ladder pattern was seen using conventional agarose gel electrophoresis (Picot

et al ., 1997). Another reason for the inconsistency could be that the fragments could have been of larger sizes and hence not resolved in a 2 % agarose gel. Furthermore, Linfert et al . (1997) demonstrated that of the 21 DNA samples extracted from ischaemic tissue and subsequently electrophoresed, only three showed DNA laddering – despite the immuno-histochemical in situ DNA fragmentation detected in all tissue samples.

In our earlier study, we postulated that the central vacuole acts as a repository for the storage of apoptotic bodies during apoptosis (Nasirudeen et al ., 2001a). The current research provides more evidence for this role. Membrane-bound apoptotic bodies containing portions of the frag-mented nucleus and an array of intact organelles such as mitochondria were noted in the central vacuole of metronidazole-treatedB. hominis cells. The central vacuole of B. hominis grown in normal culture media under normal physiological conditions may contain lipid granules or electron-dense particles. But the central vacuole

(9)

unique to cell death by apoptosis.

The effects observed by metronidazole treatment suggest that its efﬁcacy in the treatment of B. hominis infections could be attributed to activation of an apoptotic machinery. The characteristics of mammalian apoptosis are clearly reﬂected in metronidazole-treated B. hominis , namely pres-ervation of plasma membrane integrity, in situ fragmenta-tion of nuclear DNA, externalizafragmenta-tion of PS and formafragmenta-tion of apoptotic bodies. In our earlier studies on apoptosis, we described the presence and activity of caspase 3-like pro-teins inB. hominis undergoing apoptosis (Nasirudeen et al ., 2001b). The potential to undergo apoptosis appears to be present in many protozoans and more work is warranted before we can better understand this process at the mole-cular level. Interestingly, earlier reports have suggested that the genes involved in apoptosis in protozoa are not homo-logous to those in higher eukaryotes (Welburn & Murphy, 1998; Welburn et al ., 1999).

Vaux & Strasser (1996) remarked that cells might detect early changes caused by drugs or other agents and respond by activating physiological death mechanisms. In the case of B. hominis , it appears that the parasite activates its apoptotic machinery when exposed to undesirable stimuli such as cytotoxic mAb 1D5 (Nasirudeen et al ., 2001a) and metronidazole. We believe that B. hominis undergoes apoptosis for its own survival as a population of cells. We postulate that, as in the case of T. cruzi , some cells undergo apoptosis in vivo , significantly decreasing their population size, so as to evade the immune response. When the immune response is minimal, the surviving parasites re-emerge to infect the host (DosReis & Barcinski, 2001). Some authors have reported resistance of B. hominis cells to metronidazole treatment (Haresh et al ., 1999; Zaman & Zaki, 1996). These resistant strains could be a result of an efficient and effective apoptotic response to, and (or) an insufficient dose of, metronidazole. The existence of apoptosis in protozoa may also be relevant to host–parasite interactions, as it has been suggested that effective apoptosis would reduce the host immune response (e.g. inflammation) and favour overall parasite survival (Knight, 2002).

In all the assays performed in this study, the metronidazole concentration was 561027 M. Further research needs to

be done to investigate the effects of higher concentrations of the drug so that dose-dependent effects can be studied. For example, it has been shown in other systems that a higher drug concentration could induce necrosis (Anselmi

et al ., 2002). In summary, the observations described here strongly suggest the apoptotic effect of metronidazole on

B. hominis . The results of this research lead one to specu-late that cell-death mechanisms in B. hominis and higher

ACKNOWLEDGEMENTS

This work was supported by a National University of Singapore grant (R-182-000-021-112). A. M. A. N. is a postdoctoral fellow funded generously by the Singapore Millennium Foundation. The assistance of Mrs Josephine Howe i n transmission electron microscopy and that of Miss Ng Bee Ling and Miss Jeanie Tan Li Hui (National University Medical Institutes, NUS) in ﬂow cytometry is greatly appreciated.

REFERENCES

Adayev, T., Estephan, R., Meserole, S., Mazza, B., Yurkow, E. J. & Banerjee, P. (1998). Externalization of phosphatidylserine may not be an early signal of apoptosis in neuronal cells, but only the phosphatidylserine-displaying apoptotic cells are phagocytosed by microglia. J Neurochem 71_{, 1854–1864.}

Al-Olayan, E. M., Williams, G. T. & Hurd, H. (2002). Apoptosis in the malaria protozoan, Plasmodium berghei : a possible mechanism for limiting intensity of infection in the mosquito. Int J Parasitol 32_,

1133–1143.

Ameisen, J. C., Idziorek, T., Billant-Mulot, O., Loyens, M., Tissner, J., Potentier, A. & Omaissi, A. (1995). Apoptosis in a unicellular eukaryote (Trypanosoma cruzi ): implications for the evolutionary origin and role of programmed cell death in the control of cell proliferation, differentiation and survival. Cell Death Differ 2_,

285–300.

Anselmi, C., Ettorre, A., Andreassi, M., Centini, M., Neri, P. & Di Stefano, A. (2002). In vitro induction of apoptosis vs. necrosis by widely used preservatives: 2-phenoxyethanol, a mixture of isothiazolinones, imidazolidinyl urea and 1,2-pentanediol. Biochem Pharmacol 63_{, 437–453.}

Barbieri, D., Troiano, L., Grassilli, E., Agnesini, C., Cristofalo, E. A., Monti, D., Capri, M., Cossarizza, A. & Franceschi, C. (1992).

Inhibition of apoptosis by zinc: a reappraisal. Biochem Biophys Res Commun 187_{, 1256–1261.}

Collins, R. J., Harmon, B. V., Gobe, G. C. & Kerr, J. F. R. (1992).

Internucleosomal DNA cleavage should not be the sole criterion for identifying apoptosis. Int J Radiat Biol 61_{, 451–453.}

Cornillon, S., Foa, C., Davoust, J., Buonavista, N., Gross, J. D. & Golstein, P. (1994). Programmed cell death in Dictyostelium. J Cell Sci 107_{, 2691–2704.}

Darzynkiewicz, Z., Bruno, S., Del Bino, G., Gorczyca, W., Hotz, M. A., Lassota, P. & Traganos, F. (1992). Features of apoptotic cells measured by ﬂow cytometry. Cytometry 13_{, 795–808.}

Davis, M. C., Ward, J. G., Herrick, G. & Allis, C. D. (1992).

Programmed nuclear death: apoptotic-like degradation of speciﬁc nuclei in conjugating Tetrahymena. Dev Biol 154_{, 419–432.}

Didenko, V. V., Ngo, H. & Baskin, D. S. (2003). Early necrotic DNA degradation: presence of blunt-ended DNA breaks, 39 and 59

overhangs in apoptosis, but only 59overhangs in early necrosis. Am

(10)

DosReis, G. A. & Barcinski, M. A. (2001). Apoptosis and para-sitism: from the parasite to the host immune system. Adv Parasitol

49_{, 133–161.}

Dunn, L. A. & Boreham, P. F. L. (1991). The in-vitro activity of drugs against Blastocystis hominis . J Antimicrob Chemother 27_,

507–516.

Ellis, R. E., Yuan, J. Y. & Horvitz, H. R. (1991). Mechanisms and functions of cell death. Annu Rev Cell Biol 7_{, 663–698.}

Evan, G. (1994). Why we live and why we die. Chem Biol 1_,

137–141.

Falcieri, E., Martelli, A. M., Bareggi, R., Cataldi, A. & Cocco, L. (1993). The protein kinase inhibitor staurosporine induces mor-phological changes typical of apoptosis in MOLT-4 cells without concomitant DNA fragmentation. Biochem Biophys Res Commun

193_{, 19–25.}

Garavelli, P. L. (1991). The therapy of blastocystosis. J Chemother 3_,

245–246.

Gatti, R., Belletti, S., Orlandini, G., Bussolati, O., Dall’Asta, V. & Gazzola, G. C. (1998). Comparison of Annexin V and Calcein-AM as early vital markers of apoptosis in adherent cells by confocal laser microscopy. J Histochem Cytochem 46_{, 895–900.}

Hager, W. D. & Rapp, R. P. (1992). Metronidazole. Obstet Gynecol Clin North Am 19_{, 497–510.}

Haresh, K., Suresh, K., Anuar, K. & Saminathan, S. (1999). Isolate resistance of Blastocystis hominis to metronidazole. Trop Med Int Health 4_{, 274–277.}

Hirata, H., Hibasami, H., Yoshida, T., Morita, A., Ohkaya, S., Matsumoto, M., Sasaki, H. & Uchida, A. (1998). Differentiation and apoptosis without DNA fragmentation in cultured Schwann cells derived from wallerian-degenerated nerve. Apoptosis 3_{, 353–360.} Ho, L. C., Singh, M., Suresh, G., Ng, G. C. & Yap, E. H. (1993).

Axenic culture of Blastocystis hominis in Iscove’s modiﬁed Dulbecco’s medium. Parasitol Res 79_{, 614–616.}

Howell, D. M. & Martz, E. (1987). The degree of CTL-induced DNA solubilization is not determined by the human vs mouse origin of the target cell. J Immunol 138_{, 3695–3698.}

Huppertz, B., Frank, H. G. & Kaufmann, P. (1999). The apoptosis cascade – morphological and immunohistochemical methods for its visualization. Anat Embryol 200_{, 1–18.}

Koh, J. Y., Wie, M. B., Gwag, B. J., Sensi, S. L., Canzoniero, L. M., Demaro, J., Csernansky, C. & Choi, D. W. (1995). Staurosporine-induced neuronal apoptosis. Exp Neurol 135_{, 153–159.}

Knight, R. A. (2002). The archaeology of apoptosis. Cell Death Differ

9_{, 1–2.}

Kulda, J. (1999). Trichomonads, hydrogenosomes and drug resistance. Int J Parasitol 29_{, 199–212.}

Lee, M. (1991). Pathogenicity of Blastocystis homini s. J Clin Microbiol

29_{, 2089.}

Linfert, D. R., Chen, C., Ma, L., Lai, T. & Tsongalis, G. J. (1997).

Internucleosomal DNA fragmentation in apoptotic myocytes. Clin Chem 43_{, 2431–2434.}

Llibre, J. M., Tor, J., Manterola, J. M., Carbonell, C. & Foz, M. (1989). Blastocystis hominis chronic diarrhoea in AIDS patients. Lancet 1_{, 221.}

Ludovico, P., Sansonetty, F. & Corte-Real, M. (2001a). Assessment of mitochondrial membrane potential in yeast cell population by ﬂow cytometry. Microbiology 147_{, 3335–3343.}

Ludovico, P., Sousa, M. J., Silva, M. T., Lea ¨ o, C. & Co ¨ rte-Real, M. (2001b). Saccharomyces cerevisiae commits to a programmed cell death process in response to acetic acid. Microbiology 147_,

2409–2415.

Mesner, P. W., Winters, T. R. & Green, S. H. (1992). Nerve growth factor withdrawal-induced cell death in neuronal PC12 cells resembles that in sympathetic neurons. J Cell Biol 119_,

1669–1680.

Moe, K. T., Singh, M., Howe, J., Ho, L. C., Tan, S. W., Chen, X. Q. & Yap, E. H. (1999). Development of Blastocystis hominis cysts into vacuolar forms in vitro . Parasitol Res 85_{, 103–108.}

Moreira, M. E. L., Portillo, H. A. D., Milder, R. V., Balanco, J. M. F. & Barcinski, M. A. (1996). Heat shock induction of apoptosis in promastigotes of the unicellular organism Leishmania (Leishmania ) amazonensis . J Cell Physiol 167_{, 305–313.}

Nakazono-Kusaba, A., Takahashi-Yanaga, F., Morimoto, S., Furue, M. & Sasaguri, T. (2002). Staurosporine-induced cleavage of alpha-smooth muscle actin during myoﬁbroblast apoptosis. J Invest Dermatol 119_{, 1008–1013.}

Nasirudeen, A. M. A., Tan, K. S. W., Singh, M. & Yap, E. H. (2001a).

Programmed cell death in a human intestinal parasite, Blastocystis hominis . Parasitology 123_{, 235–246.}

Nasirudeen, A. M. A., Singh, M., Yap, E. H. & Tan, K. S. W. (2001b). Blastocystis hominis : evidence for caspase-3-like activity in cells undergoing programmed cell death. Parasitol Res 87_,

559–565.

Ok, U. Z., Cirit, M., Uner, A., Ok, E., Akcicek, F., Basci, A. & Ozcel, M. A. (1997). Cryptosporidiosis and blastocystosis in renal transplant recipients. Nephron 75_{, 171–174.}

Ok, U. Z., Girginkardesler, N., Balcioglu, C., Ertan, P., Pirildar, T. & Kilimcioglu, A. A. (1999).Effect of trimethoprim-sulfamethaxazole in Blastocystis hominis infection. Am J Gastroenterol 94_{, 3245–3247.} Penfold, P. L. & Provis, J. M. (1986). Cell death in the develop-ment of the human retina: phagocytosis of pyknotic and apoptotic bodies by retinal cells. Graefes Arch Clin Exp Ophthalmol 224_,

549–553.

Picot, S., Burnod, J., Bracchi, V., Chumpitazi, B. F. F. & Ambroise-Thomas, P. (1997). Apoptosis related to chloroquine sensitivity of the human malaria parasite Plasmodium falciparum. Trans R Soc Trop Med Hyg 91_{, 590–591.}

Raff, M. C. (1992). Social controls on cell survival and cell death. Nature 356_{, 397–400.}

Rosenblatt, J. E. (1990). Blastocystis homini s. J Clin Microbiol 28_,

2379–2380.

Saraste, A. & Pulkki, K. (2000). Morphologic and biochemical hallmarks of apoptosis. Cardiovasc Res 45_{, 528–537.}

Shen, Y. & Shenk, T. E. (1995). Viruses and apoptosis. Curr Opin Genet Dev 5_{, 105–111.}

Steller, H. (1995). Mechanisms and genes of cellular suicide. Science

267_{, 1445–1449.}

Sun, T., Katz, S., Tanenbaum, B. & Schenone, C. (1989).

Questionable clinical signiﬁcance of Blastocystis hominis infection. Am J Gastroenterol 84_{, 1543–1547.}

Tan, K. S., Howe, J., Yap, E. H. & Singh, M. (2001). Do Blastocystis hominis colony forms undergo programmed cell death? Parasitol Res

87_{, 362–367.}

Vaux, D. L. (1993). Towards an understanding of the molecular mechanisms of physiological cell death. Proc Natl Acad Sci U S A

90_{, 786–789.}

Vaux, D. L. & Strasser, A. (1996). The molecular biology of apoptosis. Proc Natl Acad Sci U S A 93_{, 2239–2244.}

Vaux, D. L., Haecker, G. & Strasser, A. (1994). An evolutionary perspective on apoptosis. Cell 76_{, 777–779.}

Weinrauch, Y. & Zychlinsky, A. (1999). The induction of apoptosis by bacterial pathogens. Annu Rev Microbiol 53_{, 155–187.}

(11)