In Vitro Cell. Dev. Biol.--Animal 39:98-105, January. and February" 2003 �9 2003 Society for In Vitro Biology 1071-2690/03 $18.00+0.00

HOMOCYSTEINE OXIDATION AND APOPTOSIS: A POTENTIAL CAUSE OF CLEFT PALATE

LYNDA KNOTT, l TOM HARTRIDGE, NATHAN L. BROWN, JASON P. MANSELL, AND JONATHON R. SANDY

Division of Child Dental Health, Dental School, University of Bristol, United Kingdom BS1 2LY

(Received 29 January 2003; accepted 11 March 2003)

SUMMARY

Cleft palate is the most common craniofacial anomaly. Affected individuals require extensive medical and psychosocial support. Although cleft palate has a complex and poorly understood etiology, low maternal folate is known to be a risk /actor for craniofacial anomalies. Folate deficiency results in elevated homocysteine levels, which may disturb palatoge- nests by several mechanisms, including oxidative stress and perturbation of matrix metabolism. We examined the effect of homocysteine-induced oxidative stress on human embryonic palatal mesenchyme (HEPM) ceils and demonstrated that biologically relevant levels of homocysteine (20-100 I~M) with copper (10 ~M) resulted in dose-dependant apoptosis, which was prevented by addition of catalase but not superoxide dismutase. Incubation of murine palates in organ culture with homocysteine (100 IxM) and CuSO4 (10 trM) resulted in a decrease in palate fusion, which was not significant. Gelatin gel zymograms of HEPM cell-conditioned media and extracts of cultured murine palates, however, showed no change in the expression or activation of pro-matrix metalloproteinase-2 with homocysteine (20 I~M-1 mM) with or without CuSO4 (10 I~M). We have demonstrated that biologically relevant levels of homocysteine in combination with copper can result in apoptosis as a result of oxidative stress; therefore, homocysteine has the potential to disrupt normal palate development.

Key words: folic acid; oxidative stress; craniofacial development.

INTRODUCTION

Cleft palate represents the most common craniofacial anomaly in man, with an incidence of approximately 1 in 700 live births. Al- though cleft lip and palate have a complex, muhifactorial etiology, it is well documented that low maternal folate status is an important risk factor for neural tube defects and craniofacial anomalies (Fin- nell et al., 1998). Both folic acid requirement and its metabolism are elevated during pregnancy (Christensen and Rosenblatt, 1995). Reduced serum folate is also associated with maternal smoking and anticonvulsant medication for epilepsy, and both groups have a higher incidence of oral clefts (for review, see Hartridge et al., 1999).

The cellular mechanism for the role of folate in palate develop- ment is not clear. Folate deficiency results in elevated levels of set~lm homocysteine, a sulfur-containing amino acid generated dur- ing the cellular remethylation and transulfuration pathways for which folate is required. Homocysteine can enter the amniotie fluid of the developing fetus, at levels correlating with maternal plasma concentrations, and elevated levels of homoeysteine in amniotie flu- id have been found in women bearing fetuses with neural tube de- fects (Steegers-Theunissen et al., 1995; Wang et al., 2000). Ho- moeysteine has been implicated epidemiologieally as a risk factor

To whom correspondence should be addressed at Matrix Biology Group, Department of Clinical Veterinary Science, University of Bristol, Churchill Building, Langford, Bristol, United Kingdom BS40 5DT. E-mail: 1.knon@ bris.ac.uk

for neural tube defects, independent of folate status (Steegers-Theu- nissen et at., 1994). Furthermore, hyperhomocysteinemia due to vi- tamin B6 or B12 deficiency or due to genetic defects of methyle- netetrahydrofolate reductase and cystathione [3-synthase is associ- ated with an increased incidence of nonsyndromic orofacial clefts (Mills et al., 1999; Wong et al., 1999).

Homncysteine is currently the subject of much research and dis- cussion because of its potential link with cardiovascular disease (Refsum et al., 1998). Most of the information on the potential mechanisms of homocysteine action on cellular function has been obtained irom studies using vascular cells and tissues. It is of rel- evance that congenital malformations involving heart septation, neu- ral tube closure, and oral clefting often occur together (e.g., "Catch 22"; Sergi, 1999), the common factor being the origin of the ceils involved. The neural crest ceils that uhimately form these structures derive from adjacent areas of the neural ectoderm and appear to be particularly sensitive to a variety of environmental and genetic fac- tors (Rosenquist et al., 1996; Burgoon et al., 2002). There is little direct research on the effects of homocysteine on embryological events or processes. However, Rosenquist et al. (1996, 2001) and Limpach et al. (2000) have reported homocysteine-induced malfor- mations in the developing chick embryo, including abnormal palate formation. However, development of the avian palate is very differ- ent from that of the mammalian palate, and further work is evidently required in this area to investigate the role of homocysteine in the development of cleft palate.

Many mechanisms of action have been proposed for homocysteine

98

HOMOCYSTEINE AND CLEFT PALATE 99

including hypomethylation, protein homocysteinylation, interaction with receptors, and oxidative stress, one or more of which may be involved in the disruption of palate development. We have concen-

trated our initial investigations on the role of homocysteine- induced

oxidative stress because neural crest cells have a reduced capacity

to deal with reactive Oxygen species (ROS; Davis et al., 1990; Sivan

et al., 1997). The free radicals generated by ethanol, for example, are thought to contribute to the development of cleft palate and

other abnormal i t ies involving cranial neural crest cells in fetal al- cohol syndrome (Chen et al., 1996).

Homoeysteine is readily oxidized to the disulphide honmcystine; oxidation is a process augmented by transit ion metals such as cop-

per that may be elevated in homoeysteinaemie patients (Dudman and Wilcken, 1983). The oxidation of homoeysteine results in the

l iberation of highly reactive hydrogen peroxide (H202) and super-

oxide anions ( 0 2 ) , both of which are directly damaging to cells,

altering their function, inducing apoptosis, or both (Starkebamn and

Harlan, 1986; Hultberg et al., 1997; Davies, 1999). Neural c res t - derived vascular endothelial cells (VEC) in contrast to non-nem'a l

crest-derived vascular cells also have an abnornml oxidative stress

response (Outinen et al., 1998). Homocysteine precipitates a de-

crease in intracellular glutathione peroxidase messenger r ibonueleic

acid (mRNA) and glutathione peroxidase activity, thereby impair ing

the ability of VEC to detoxify H2Oz (Upehurch et al., 1997). Data

from differential display and complementary deoxyribonucleic acid

micromTay analysis further suggest that VEC exposure to homoeys-

teine results in reductive stress, al though the levels of homoeysteine

required to elicit these responses are very high (Austin et al., 1998; Outinen et al., 1999).

Homocysteine and ROS can also disturb the extracellular matrix

and its metabolism. Homocysteine has been variously descr ibed to activate matrix metal loproteinase-2 (MMP-2; Bescond et al., 1999;

Mujumdar et al., 2001), induce t issue inhibi tor of metalloprotei-

nases-1 (Torres et al., 1999), inhibi t and promote AP-1 (an MMP

promotor) activity (Torres et al., 1999; Suzuki et al., 2000), increase

collagen synthesis (Majors et al., 1997), and alter collagen cross- l inking (Jackson, 1973). Oxidative stress in general has been shown

to increase MMP-1 mRNA (Brenneisen et al., 1997) and regulate

both MMP activity and collagen synthesis (Rajagopalan et al., 1996;

Siwik et al., 2001). The extracellular matrix, principally collagen,

and its metabol ism (Morris-Wiman et al., 1999; Mansell et al.,

2000), undergo temporal changes during secondary palate devel-

opment. Normal development of the palate requires a complex series

of events involving mesenehymal proliferation in the maxillary pro-

cesses, elevation of the palatal shelves, and breakdown of the con-

tacting medial edge epi thel ium (MEE) (Brinkley, 1980; Ferguson,

1988). Many of these processes require metabolism of the matrix,

and we and others have previously demonstrated that disruption of MMP function can result in nonfusion of palates in vitro (Blavier et al., 2001; Brown et al., 2002).

We therefore hypothesize that elevated levels of homocysteine during embryogenesis can result in the development of cleft palate.

To test this hypothesis, we first examined the effect of homocysteine-

induced oxidative stress on both human embryonic palatal mesen-

chyme (HEPM) cells and on murine palate organ culture. Second,

we investigated the effects of homocysteine and oxidative stress on matrix turnover, specifically, MMP-2 production.

MATERIALS AND METHODS

All materials were purchased from Sigma Chemical Co. (Poole, U.K.) un- less otherwise stated.

Human embryonic palatal mesenchyme cell culture. Human embryonic pal- atal mesenchyme cells Were purchased from the European Collection Of Cell Culture (Salisbury, U.K.). All cell cultures were carried out at 37 ~ C in a humidified atmosphere containing 5% CO2. Before any experimentation, cells were maintained in modified Eagle medium supplemented with 5% (v/v) fetal calf serum (FCS; GIBCO, Paisley, U.K.), 4 mM L-glutamine, penicillin-strep- tomycin (100 U/ml and 0.1 mg/ml, respectively), and 1% v/v nonessential amino acids.

Human embryonic palatal mesenchyme cells were seeded into 12-well plates (Helena Biosciences, Sunderland, U.K.) and left for 24 h until 50- 60% confluent. The medium was then removed and replaced with medium supplemented with 1% FCS, and the cells were left for a further 24 h to reduce the basal level of activity. Both CuSO4 and DL-homocysteine were prepared as stocks (10 nrM) in 1 urM hydrochloric acid on the d of use. Concentrations of homocysteine between 10 and 100 p~M were used with 10 IxM CuSO4 and 100 IxM homocysteine with 0.2-10 ~M CuSQ. Control ex- periments using the disulphides homocystine and cystine and the thiol cys- teine together with 10 p.M CuSO4 were also conducted. The concentrations of homocysteine and copper are within the ranges observed clinically (Cart- wright and Wintrobe, 1964; Finkelstein and Martin, 2000). To investigate the role of superoxide dismutase (SOD) or catalase (or both), stock solutions were prepared in media and added to wells at a final concentration of 200 and 1000 U/ml.

Each experiment was repeated three times (n = 6 for each condition per experiment). Cell number was quantified using a Coulter counter after dis- association with trypsin-ethylenediamine-tetraacetic acid (GIBCO). The use of colorimetric methods such as the 3-(4,5-dimethyhhiazole-2-yl)-2,5-biphe- nyl tetrazolimn bromide assay could not be considered because of thiol in- terference.

Determining the mechanism of cell death. Cell death can occur by either apoptosis or necrosis, and exposure of cells to oxidative stress can result in cell death by either mechanism, depending on stress level, cell type, and cell environment (Davies, 1999). Morphological and histochemical observa- tions were made after 3 h of incubation with 100 txM homocysteine and 10 txM CuSO4, when the initial effects can be observed by phase contrast mi- croscopy (cell shrinkage), before the dead cells detach and finally lyse.

A fluorescein isothiocyanate (FITC)-labeled cell-permeable caspase inhib- itor, valyl-alanyl-[o-methyl]-fluoromethylketone (VAD-FMK), was used to la- bel activated caspases to allow identification of apoptosing cells with a fluo- rescent microscope. Cells were grown on three-well slides until 50% comqu- ent and challenged with 100 ~M homocysteine per 10 ta3//CuSO4 in media containing 1% FCS for 3 h. The CaspACE ~ FITC-VAD-FMK in situ marker (Promega, Southampton, U.K.) was then added to the media at a final con- centration of 20 txM, and incubation at 37 ~ C continued for 90 min. The cells were then fixed in 10% buffered formalin in the dark at room temper- ature and subsequently washed in phosphate-buttered saline. A slide moun- tant supplemented with 4',6-diamidine-2-phenylindole (DAN) (Vector Lab- oratories, Peterborough, U.K.) enabled nuclear localization, when cells were viewed under the fluorescence microscope.

Gelatin gel zymography of cell-conditioned media. Human embryonic pal- atal mesenchyme ceils wei~ cultured as described above and were challenged with homocysteine and copper levels that did not result in significant cell death (20 tzM homocysteine and 10 txM CuSO4). In addition, HEPM cells were cultured with 100 txM homocysteine alone.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis zymography was performed as previously described (Mansell et al., 1998). Briefly, cell- conditioned media were diluted in nonreducing sample buffer (2• as de- scribed by Laemmli (1970), but containing twice the concentration of SDS (2% w/v). Samples and MMP-2 standards (Calbiochem, Nottingham, U.K.) were eleetrophoresed using 10% polyacrylamide (Bio-Rad, Hertfordshire, U.K.) gels eopolymerized with 0.5 mghul d gelatin (bovine skin; Sigma). Gels were washed in 2.5% Triton X-100 to displace SDS and then incubated overnight at 37 ~ C in proteolysis buffer (500 mM NaC1, 50 mM CaC12, 50 nrM Tris-HC1, pH 7.8) containing 200 txM p-aminophenylmercuric acetate. The gels were then stained with Coomassie brilliant blue and the proteolyt- ically clarified zones quantified using an Agfa Studiostar scanner and Scion Image (Scion Corporation, MD) software.

Palate organ culture. CD1 mice were subjected to a reverse day-night

1 0 0 KNOTT ET AL.

FIG. i. The effect of homocysteine (lacy) and copper (CuSO~) on human embryonic palatal mesenchymc cell nmnber. Human embryonic palatal mesenchyme cells were cultured under standard conditions, challenged for 24 h, and cell number quantified using an automated Coulter counter. Cell number was not significantly affected with 100 p~M homocysteine or 10 p~M CuSQ alone; however, there was a homocysteine concentration-related decrease in cell number when combined with 10 taM CuSO4. Data are presented as percentages of controls and standard deviations. **, P < 0.01 compared with control; n = 6.

cycle and the males introduced to females for a period of 2 h during the night phase. Females with vaginal plugs were isolated and if pregnant sac- rificed by cervical dislocation 13 d postcoitum _+ 2 h. This method ensured accuracy and consistency in obtaining embryonic day 13 mice. The palates were then prepared for organ culture, as described previously (A1-Obdaidi et at., 1995).

Palates were cultured in sialinized roller tubes (Supelco, Sigma, Poole, U.K.) containing BGJ-B medium (GIBCO-BRL) supplemented with 850 t~M ascorbic acid, 10 taM CuSO4, and 0.6% (w/v) bovine serum albumin and were purged with 95% 02 and 5% COs before roller culturing at 37 ~ C for 72 h. The tubes were purged for 3 min at 24-h intervals and the media changed after 48 h. Palates were cultured in the presence of 100 IxM ho- mocysteine or vehicle. In addition, palates were also cultured with 5 taM retinoic acid as a control for clefting.

Palate histology. Palates were fixed in formal saline for 24 h and prepared for histology on a Shandon 2LE processor with graded methylated spirits, xylene, and paraffin wax baths. Wax was removed with 100% xylene, and the palates were air-dried overnight. Palates were photographed using a mi- croscope and an external light source. Palates were then embedded in par- affin wax, sectioned on a microtome (2 ~m), and stained with hematoxylin and eosin.

Gelatin gel zymography of cultured palate extracts. A second series of pal- ates, cultured as described above, were homogenized in 10 mM N-2-hydrox- yethylpiperazine-N'-2-ethane-sulfonic acid buffer at pH 7.4 to a concentra- tion of 50 mg/ml. Homogenates were centrifuged at 5000 rpm for 10 rain at 4 ~ C, and the supernatant was used for zymography. A protein assay (Bio- Rad) was used to measure the protein concentration for each sample, and the equivalent of 5 t~g protein was subjected to electrophoresis as described earlier.

Statistics. The data from the cell culture experiments were analyzed using a one-way analysis of variance after testing the assumptions of a normal population and equal standard deviations by the methods of Kohnogorov and Smimov, and Bartlett, respectively. A Fisher exact test was used to analyze a contingency table of the organ culture data. The tests were carried out using Instat | (Grapbpad Software Inc., San Diego, CA).

RESULTS

The effect of homocysteine and copper on HEPM cell number. In-

cubation of HEPM cells with either 100 v,M homocysteine or lO p,M CuSO4 had no effect on cell number. However, when cells were

incubated with homocysteine together with CuSQ, there was a dose-

related decrease in cell number (P < 0.01; Fig. 1). Cysteine, but not the disulphide homocystine, also resulted in cell death when

incubated with CuSO4 (P < 0.01; Fig. 2), suggesting that copper-

catalyzed oxidation of thiols resulted in cell death. The decrease in cell number was abrogated by the addition of

catalase but not SOD or boiled catalase (P < 0.01; Fig. 3), indi-

cating that the cell death is triggered by the generation of n202 rather than by O2 . Furthermore, the addition of catalase to the cells

significantly increased cell number above controls (P < 0.01) be- cause of the reduction in the background levels of oxidative s t ress -

related cell death in standard media. CaspACE labeling. Observed under a standard phase contrast

microscope, the ceils do not show any morphological changes until

4 - 5 h of incubation with bomocysteine and CuSO4. The cells ini- tially shrink and become rounded with condensed chromatin char-

acteristic of apoptotic cells. After 8 h, the cells swell and become detached from the surface, followed by extensive lysis after 24 h of

incubation. The use of the FITC-labeled caspase inhibitor, VAD-FMK, en-

abled identification of apoptotic ceils after only 3 h of incubation with homocysteine and copper. The majority of cells were positively stained for active caspase, with bright DAPI-fluorescing condensed

nuclei and some blebbing of nuclei (Fig. 4). Organ culture. The addition of CuSO4 alone was found to have

no effect on palate fusion, and therefore 10 p A / C u S Q was routinely

added to the standard media to reduce the number of experimental animals required. The results of the organ culture are summarized

in Table 1. Seventy-six percentage (13 of 17) of palates cultured in control standard medium (containing 10 ~M CuSO~) had evidence

of fusion, whereas only 13% (2 of 16, P < 0.001 compared with

control) fused when cultured in the presence of 5 p.M retinoic acid

(clefting control). In the experimental gToup, addition of 100 p,M homocysteine to the standard medium resulted in a reduction in

HOMOCYSTEINE AND CLEFT PALATE 101

FIG. 2. The effect of thiols and disulphides on human embryonic palatal mesenchyme cell number. Human embryonic palatal mes- enchyme cells were cultured under standard conditions, challenged for 24 h, and cell number quantified using an automated Coulter counter. Cell number was not significantly affected with 100 IxM homocystine or cysteine with or without 10 txM CuSO,. However 100 I-tM cysteine with 10 IxM CuSO4 resulted in a decrease in cell number similar to that produced by 100 p.M homocystine (Fig. 1). Data are presented as percentages of controls and standard deviations. **, P < 0.01 compared with control; n = 6.

palatal fusion, with 52% (11 of 21, not significant compared with control) of palates fused.

Gelatin gel zymography. Pro-metalloproteinase-2 was the major gelatinase produced by HEPM cells in cuhure, with no evidence of MMP-9 and small, but unquantifiable, levels of activated MMP-2. There was no significant change in the production or activation of

MMP-2 in HEPM ceils incubated with 20 IxM homocysteine and 10 p~M CuSO4 for up to 48 h (Fig. 5a). Similarly, there was no change in expression of pro-MMP-2 or in its activation in cells challenged with 100 p~M homocysteine alone (data not shown).

Both pro- and active MMP-2 were present in extracts from cul- tured murine palates. Ahhough there was a decrease in the levels of MMP-2 with 1 mM homocysteine, there were no significant changes in the levels of pro- or active MMP-2 with addition of any of the concentrations of homocysteine compared with controls (Fig.

5b).

DISCUSSION

Given the link between cleft palate and folate deficiency and the reported susceptibility of neural cres t -der ived cells to the effects

FIG. 3. The effect of catalase and superoxide dismutase (SOD) on homocysteine- and copper-induced cell death. Human embryonic palatal mesenchyme cells were cultured under standard conditions, challenged for 24 h, and cell number quantified using an automated Coulter counter. Catalase prevented the cell death induced by homocysteine and CuSO4, whereas SOD had no effect. Inactivation of the catalase by boiling abolished its protective effect. Data are presented as percentages of controls and standard deviations. **, P < 0.01 compared with control; n = 6.

102 KNOTT ET AL.

FIG. 4. Activated easpase labeling in homoeysteine- and copper-treated human embryonic palatal mesenehyme cells. Hmnan embry- onic palatal mesenchyme ceils were cultured on chamber slides, challenged, and after 3 h, labeled with fluorescein isothiocyanate (FITC)- labeled CaspACE, counterstained with 4',6-diamidine-2-phenylindole (DAPI), and then examined using a fluorescence microscope to determine extent of apoptosis. (a, b) Control cells, treated with vehicle only, FITC and DAPI fluorescence, respectively. No CaspACE- labeled, apoptotic cells are present, and all nuclei have normal morphology. (c, d) Cells treated with 100 ~M homocysteine and 10 txM CuSO~, FITC and DAPI fluorescence, respectively, showing five apoptosing cells with strong CaspACE labeling with condensed, brightly fluorescing shrunken nuclei under DAPI fluorescence (large arrow), one normal cell with no labeling and normal nucleus (small arrow). Magnification: X400.

TABLE 1

FREQUENCY OF PALATAL FUSION OF CULTURED MUR1NE PALATE EXPLANTS. PALATES WERE CULTURED IN STANDARD MEDIA

(CONTAINING 10 IxM CuSO~), SUPPLEMENTED WITH 5 txM RETINOIC ACID (CLEFTING CONTROL), 100 ~M HOMOCYSTEINE,

OR VEHICLE

Fused Nonfused Total

Vehicle 13 4 17 Retinoic acid 2 14"** 16 Homocysteine 11 10 21 Total 26 28 54

***P < 0.001 compared with vehicle, as determined using the Fisher exact test.

of homoeysteine and associated oxidative stress, there has been lit- tle investigation into the role of elevated homocysteine and the de- velopment of this common eraniofaeial anomaly or the potential cellular mechanisms involved. This study has demonstrated that the oxidation of homoeysteine can result in palatal mesenchyme cell death in vitro and the possible disruption of normal murine pala- togenesis in organ culture. The mechanism of action that we inves- tigated was the potential for oxidative stress, induced by oxidation of homoeysteine, to disrupt HEPM cell function because neural crest derived-cells such as palatal mesenchyme cells are known to be susceptible to changes in redox status (Outinen et al., 1998). Starkebaum and Harlan (1986) first demonstrated that oxidative

stress as a consequence of homocystine oxidation by copper (free or bound as ceruloplasmin) could result in endothelial cell death. However, it is important to note that in the study by Starkebaum and Harlan (1986) and in several other studies investigating the pathological effect of homocystine, supraphysiologieal levels of ho- mocysteine (100-500 ixM) were used. A study by Hultberg et al. (1997) demonstrated that oxidative stress induced by low levels of homocysteine and copper resulted in cell damage, and in our study, we have demonstrated a dose-related increase in cell death with elevated, but clinically relevant, levels of both homocysteine (10- 100 /xM) and copper (1-10 t~M) (Cartwright and Wintrobe, 1964; Finkelstein and Martin, 2000).

Transition metal-catalyzed autooxidation of free thiols produces H~02 and O~ ROS. Reactive oxygen species can induce a range of cellular events depending on the level of oxidative stress, in- ducing proliferation, growth arrest, apoptosis, and necrosis under both pathologic and nonpathologie conditions (Davies, 1999; Si- mon et al., 2000). In this study, the oxidative stress resulting from incubation of HEPM cells with homocysteine (20-100 ~M) and copper (0.1-10 IxM) led to apoptosis rather than necrosis. Reac- tive oxygen species can induce apoptosis through several different mechanisms, including the triggering of death receptors, mito- ehondrial disruption, lysosomal rupture, oxidative damage to key proteins, and disruption of the cellular redox status (Kehrer, 2000; Simon et al., 2000; Antunes et al., 2001). Although not all apop- toses involve easpases, these intraeellular eysteine proteinases are integral to many of the apoptotie pathways, and as such, the pres-

HOMOCYSTEINE AND CLEFT PALATE 103

FIG. 5. Gelatin gel zylnography of (a) human embryonic palatal mesenchym e (HEPM)-conditioned media and (b) cultured murine palate extracts. (a) Media taken from HEPM cells cultured under standard conditions and challenged for 48 h with either vehicle (v) Or 20 tiM homocysteine and 10 tAM CuSO4 (h). (b) Extracts were taken from routine palates roller cultured for 72 h in media containing 10 tAM CuSO4, with 5, 1, or 0.1 mM homocysteine or controts with vehicle only. There were no significant differences in the levels of pro- or active MMP-2 in either cell or organ culture experiments.

ence of activated caspases is often used to identify apoptosing ceils. Caspase 3 has been previously identified in VEC in which apoptosis was induced by a relatively high level of homocysteine (500 IxM) and 10 IxM copper (Bessede et al., 2001). In this study, the use of the CaspACE in situ marker enabled the identification of early apoptosing cells before (3 h) the morphological changes were observed under phase contrast microscopy after 4-5 h of incubation. After 8 h, few ceils were caspase positive, as the ma- jority of apoptosing cells had progressed to secondary necrosis or disintegrated. Caspases are regulated by ROS; however, as cys- teine proteases, they are themselves sensitive to the redox status of the cell, and at higher levels of oxidative stress, caspases may become inactivated, thereby blocking some apoptotic pathways (Hampton et al., 1998). Apoptosis has been proposed along with several other mechanisms for the loss of the MEE during palatal shelf fusion, in addition to being an important process in devel- opmental processes in general. Recently published data suggest that programmed cell death is required for successful palate shelf fusion (Cuervo et al., 2002), and any disruption in the balance of apoptosis and cell survival would result in disruption of palate formation.

Human embryonic palatal mesenchyme apoptosis in response to homocysteine and copper was abrogated with the addition of catalase but not SOD, indicating that the cell death was triggered by the generation of H202 rather than O2 . Homocysteine also affects other aspects of the cell's ability to respond to oxidative stress therefore potentially exacerbating the effects of the ROS generated by homocysteine autooxidation. Elevated levels of ho- mocysteine are known to inhibit production and activity of glu- tathione peroxidase, which is responsible for the removal of n202 (Nishio and Watanabe, 1997; Outinen et al., 1998). High levels of homocysteine (0.5-5 raM) were used in several of these studies, and it is possible that different mechanisms may be involved at these pharmacological levels; however, it is likely that the mech- anism whereby homocysteine damages cells is multifactorial. In- deed, several cellular mechanisms of action have been proposed

for homocysteine-induced responses, including hypomethylation, protein homoeysteinylation, and interaction with N-mehyl-D-as- partate receptors.

We, and others, have previously described the importance of ge- latinases during palate development (Mansell et al., 2000) and also that disruption of MMP function results in cleft palate (Blavier et al., 2001; Brown et al., 2002). Both homocysteine and ROS have been reported to disrupt MMPs, from transcription to activation; however, in this study, we were unable to demonstrate any signifi- cant changes in MMP-2 levels or activation. Homocysteine and ox- idative stress have been reported to have opposing effects on MMPs, and ROS have been shown to increase MMP transcription and ac- tivity. Homocysteine can inhibit the activity of the transcription fac- tor AP-1, thereby decreasing MMP-1, -3, and -9 production (Suzuki et al., 2000; Siwik et al., 2001), although MMP-2 transcription is controlled by AP-2. The overall expectation for MMP involvement is therefore not clear or predictable, and this aspect of palatal de- velopment may not be affected by oxidative stress in this situation.

Normal human palate development requires a complex and co- ordinated sequence of events including palatal shelf elevation with subsequent tissue reorganization and differentiation into definitive structures. In addition to our study of palatal mesenchyme ceils, we have used a murine organ culture method to test whether homocys- teine can disrupt palate development at the tissue level. Palatoge- nesis differs among species, but human and mouse palate devel- opment is comparable. Both species have individual shelves of the secondary palate that grow vertically and need to reorientate into a position above the tongue, contact, and exclude the MEE to com- plete fusion. This makes mouse a suilable model to study pal, ate development, and murine palatal organ culture is an established technique for examining the potential of teratogenic agents to in- duce cleft palate. Our palatal culture results suggest that 100 IxM homocysteine with 10 IAM CuSQ might be teratogenic, although the findings indicate a reduction in the number of fused palates under these conditions, and the data obtained did not reach stalistical significance. Further studies of the effect of homocysteine on palate

1 0 4 KNOTT ET AL.

culture would be valuable to clarify these data and also to inves-

tigate the possible changes in the levels of apoptosis in the palate

in light of the data from the cell culture experiments. The organ

culture data may suggest, however, that the main effect of homo-

cysteine toxicity demonstrated in cell culture may not be in the

later stages of palate elevation and fusion examined in this study

but during earlier stages of palate formation such as during neural

crest cell migration within the developing palate.

It is apparent from this and other studies that homocysteine does

have the potential to disrupt normal palate development. We have

demonstrated that biologically relevant levels of homocysteine and

copper can result in apoptosis induced by oxidative stress. Clearly,

the role of homocysteine in the development of cleft palate requires

further investigation, and such studies would not only provide in-

sights into the etiology of cleft palate in general but also the com-

plex process of palatogenesis.

REFERENCES

A1-Obdaidi, N.; Kastner, U.; Merker, H.-J.; Klug, S. Development of a sus- pension organ culture of the foetal rat palate. Arch. Toxicoh 69:472-479; 1995.

Antunes, E; Cadenas, E.; Brunk, U. T. Apoptosis induced by exposure to a low steady state concentration of hydrogen peroxide is a consequence of lysosomal rupture. Biochem. J. 356:549-555; 2001.

Austin, R. C.; Sood, S. K.; Dorward, A. M.; Singh, G.; Shaughnessy, S. G.; Pamidi, S.; Outinen, P. A.; Weitz, J. I. Homocysteine-dependent al- terations in mitochondrial gene expression, function and structure. J. Biol. Chem. 273:30808-30817; 1998.

Bescond, A.; Augie~, T.; Chareyre, C.; Garcon, D.; Hornebeck, W.; Charpiot, P. Influence of homocysteine on matrix metallopruteinase-2: activation and activity. Biochem. Biophys. Res. Commun. 263:498-503; 1999.

Bessede, G.; Miguet, C.; Gambert, P.; Neel, D.; Lizard, G. Efficiency of ho- mocysteine plus copper in inducing apoptosis is inversely propor- tional to ",/-glutamyl transpeptidase activity. FASEB J. 15:1927-1940; 2001.

Blavier, L.; Lazaryev, A.; Groffenm, J.; Heisterkamp, N.; DeClerck, Y. A.; Kaartinen, V. TGF-beta3-induced palatogenesis requires matrix me- talloproteinases. Mol. Biol. Cell 12:1457-1466; 2001.

Brenneisen, P.; Briviba, K.; Wlaschek, M.; Wenk, J.; Scharffetter-Kochanek, K. Hydrogen peroxide (H202) increases the steady-state mRNA levels of collagenase/MMP-1 in human dermal fibroblasts. Free Radic. Biol. Med. 22:515-524; 1997.

Brinkley, L. L. In vitro studies of palatal shelf elevation. In: Pratt, R. M.; Christiansen, R. L. ed. Current research trends in craniofacial de- velopment. Elsevier/North-Holland, New York. 1980:203-220.

Brown, N. L.; Yarram, S. J.; Mansell, J. P.; Sandy, J. R. Matrix metallopro- teinases have a role in palatogenesis. J. Dent. Res. 81:826-830; 2002.

Burgoon, J. M.; Selhub, J.; Nadeau, M.; Sadler, T. W. Investigation of the effects of folate deficiency on embryonic development through the establishment of a folate deficient mouse model. Teratology 65:219-227; 2002.

Cartwright, G. E.; Wintrobe, M. M. Levels of copper in serum of normal men and women. Am. J. Clin. Nutr. 14:224-231; 1964.

Chen, S. Y.; Sulik, K. K. Free radicals and ethanol-induced cytotoxicity in neural crest cells. Alcohol Clin. Exp. Res. 20:1071-1076; 1996.

Christensen, B.; Rosenblatt, D. S. Effects of folate deficiency on embryonic development. Baillieres Clin. Haematol. 8:617-637; 1995.

Cuervo, R.; Valencia, C.; Chandraratna, R. A. S.; Covarrubias, L. Programmed cell death is required for palate shelf fusion and is regulated by retinoic acid. Dev. Biol. 245:145-156; 2002.

Davies, K. J. The broad spectrum of responses to oxidants in proliferating cells: a new paradigm for oxidative stress. IUBMB Life 48:41-47; 1999.

Davis, W. L.; Crawford, L. A.; Cooper, O. J.; Farmer, G. R.; Thomas, D. L.; Freeman, B. L. Ethanol induces the generation of reactive free rad- icals by neural crest cells. J. Craniofac. Genet. Dev. Biol. 10:277-293; 1990.

Dudman, N. P.; Wilcken, D. E. L. Increased plasma copper in patients with homocystinuria due to cystathione 13-synthase deficiency. Clin. China. Acta 127:105--113; 1983.

Ferguson, M. W. K. Palate development. Development 103:41-60; 1988. Finkelstein, J. D.; Martin, J. J. Homocysteine. Int. J. Biochem. Cell Biol.

32:385-389; 2000. Finnell, R. H.; Greer, K. A.; Barber, R. C.; Piedrahita, J. A. Neural tube and

craniofacial defects with special emphasis on folate pathway genes. Crit. Rev. Oral Biol. Med. 9:38-53; 1998.

Hampton, M. B.; Fadeel, B.; Orrenius, S. Redox regulation of the caspases during apoptosis. Ann. NY Acad. Sci. 854:328-335; 1998.

Hartridge, T.; Illing, H. M.; Sandy, J. R. The role of folic acid in oral clefting. Br. J. Orthod. 26:115-120; 1999.

Huhberg, B.; Andersson, A.; Isaksson, A. The cell damaging effects of low amounts of homocysteine and copper ions in human cell line cultures are caused by oxidative stress. Toxicology 123:3340; 1997.

Jackson, S. H. The reaction of homocysteine with aldehyde: an explanation of the collagen defects in homocystinuria. Clin. Chim. Acta 45:215-217; 1973.

Kehrer, J. P. Cause-effect of oxidative stress and apoptosis. Teratology 62:235-236; 2000.

Laemmli, U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-682; 1970.

Limpach, A.; Dalton, M.; Miles, R.; Gadson, P. Homocysteine inhibits retinoic acid synthesis: a mechanism for homocysteine-induced congenital de- fects. Exp. Cell Res. 260:166-174; 2000.

Majors, A.; Ehrhart, L. A.; Pezacka, E. H. Homocysteine as a risk factor for vascular disease. Enhanced collagen production and accumulation by smooth muscle cells. Arterioscler. Thromb. Vasc. Biol. 17:2074-2081; 1997.

Mansell, J. P.; Bailey, A. J. Abnormal cancellous bone collagen metabolism in osteoarthritis. J. Clin. Invest. 101:1596-1603; 1998.

Mansell, J. P.; Kerrigan, J.; McGill, J.; Bailey, A. J.; TeKoppele, J.; Sandy, J. R. Temporal changes in collagen composition and metabolism during rodent palatogenesis. Mech. Ageing Dev. 119:49-62; 2000.

Mills, J. L.; Kirke, P. N.; Molloy, A. M., et al. Methylenetetrahydrofolate reductase thermolabile variant and oral clefts. Am. J. Med. Genet. 86:71-74; 1999.

Morris-Wiman, J.; Du, Y.; Brinkley, L. Occurrence and temporal variation in matrix metalloproteinases and their inhibitors during murine second- ary palatal morphogenesis. J. Craniofac. Genet. Dev. Biol. 19:201-212; 1999.

Mujumdar, V. S.; Am, G. M.; Tyagi, S. C. Induction of oxidative stress by homocyst(e)ine impai~ endothelial function. J. Cell Biochem. 82:491-500; 2001.

Nishio, E.; Watanabe, Y. Homocysteine as a modulator of platelet-derived growth factor action in vascular snmoth muscle cells: a possible role for hydrogen peroxide. Br. J. Pharmacoh 122:269-274; 1997.

Outinen, P. A.; Sood, S. K.; Liaw, P. C. Y., et al. Characterisation of the stress-inducing effects of homocysteine. Biochem. J. 332:213-221; 1998.

Outinen, P. A.; Sood, S. K.; Pfeifer, S. I.; Pamidi, S.; Podor, T. J.; Li, J.; Weitz, J. I.; Austin, R. C. Homocysteine-iuduced endoplasmic reticulum stress and growth arrest leads to specific changes in gene expression in human vascular endothelial cells. Blood 94:959-967; 1999.

Rajagopalan, S.; Ping Meng, X.; Ramasamy, S.; Harrison, D. G.; Galls, Z. S. Reactive oxygen spices produced by macrophage-derived foam cells regulate the activity of vascular matrix metalloproteinases in vitro. J. Clin. Invest. 98:2572-2579; 1996.

Refsum, H.; Ueland, P.; Nygard, O.; Vollset, S. Homocysteine and cardio- vascular disease. Ann. Rev. Med. 49:31-62; 1998.

Rosenquist, T. H.; Finnell, R. H. Genes, folate and homocysteine in embry- onic development. Proc. Nutr. Soc. 60:53-61; 2001.

Rosenquist, 3". H.; Ratashak, S. A.; Selhub, J. Homoeysteine induces con- genital defects of the heart and neural tube: effect of folic acid. Pruc. Natl. Acad. Sci. USA 93:15227-15232; 1996.

HOMOCYSTEINE AND CLEFT PALATE 1 0 5

Sergi, C.; Sellai, M.; Muller-Navia, J.; Schnabe], R A.; Hagl, S.; Otto, H. E; Ulmer, H. E. Catch 22 syndrome. Report of 7 infants with follow-up data and review of the recent advancements in the genetic knowledge of the locus 22qi1. Pathologica 91:166-172; 1999.

Simon, H.-U.; Haj-Yehia, A.; Levi-Schaffer, E Role of reactive oxygen species (ROS) in apoptosis induction. Apoptosis 5:415-418; 2000.

Sivan, E.; Lee, Y.; Wu, Y.; tleece, E. A. Free radical scavenging enzyme, s in fetal dysmorphogenesis among offspring of diabetic rats. Teratology 56:343-349; 1997.

Siwik, D. A.; Pagano, E J.; Colucei, W. S. Oxidative stress regulates collagen synthesis and matrix metalloproteinase activity in cardiac fibroblasts. Am. J. Physiol. Cell Physiol. 280:C53-C60; 2001.

Starkebaum, G.; Harlan, J. M. Endothelial cell injury due to copper-catalyzed hydrogen peroxide generation fi'om homoeysteine. J. Clin. Invest. 77:1370-1376; 1986.

Steegers-Theunissen, R. R; Boers, G. H.; Blom, H. J.; Nijhuis, J. G.; Thomas, C. M.; Borm, G. E; Eskes, T. K. Neural tube defects and elevated homoeysteine levels in amniotie fluid. Am. J. Obstet. Gyneeol. 172:1436-1441; 1995.

S~eegers-Theunissen, R. P.; Boers, G. H.; Trijbels, F. 1., et al. Maternal hy- perhomocysteinemia: a risk factor for neural tube defects? Metabo- lism 43:1475-1480; 1994.

Suzuki, Y. J.; Lorenzi, M. V..; Shi, S. S.; Day, B. M.; Blmnberg, J. B. Ho- mocysteine exerts cell type-specific inhibition of AP-1 transcription factor. Free Radic. Biol. Med. 28:39-45; 2000.

Tortes, L.; Garcia-Trevijano, E. R.; Rodriguez, J. A., et al. Induction of TIMP- 1 expression in rat hepatic stellate cells and hepatocytes: a new role for homocysteine in liver fibrosis. Biochim. Biophys. Acta 1455:12-22; 1999.

Upchurch, G. R.; Welch, G.; Fabian, A. J.; Freedman, J. L.; Johnson, J. L.; Keaney, J. E; Loscalzo, J. Homocyst(e)ine decreases bioavailable ni- tric oxide by a mechanism involving glutathione peroxidase. J. Biol. Chem. 272:17012-17017; 1997.

Wang, J.; Trudingeb B. J.; Duarte, N.; Wilcken, D. E.; Wang, X. L. Elevated circulating homocyst(e)ine levels in placental vascular disease and associated pre-eclampsia. Br. J. Obstet. Gynecol. 107:935-938; 2000.

Wong, W. Y.; Eskes, T. K.; Kuijpers-Jagtman, A. M.; Spauwen, P. H.; Steegers, E. A.; Thomas, C. M.; Hamel, B. C.; Biota, H. J.; Steegers-Theunis- sen, R. R Nonsyndromic m'alfacial clefts: association with maternal hyperhomocysteinemia. Teratology 60:253-257; 1999.