TOXICOLOGYANDAPPLIEDPHARMACOLOGY 88,141-152(1987)

Regulation of Intracellular Glutathione in Rat Embryos and Visceral Yolk Sacs and Its Effect on 2-Nitrosofluorene-Induced Malformations

in the Whole Embryo Culture System’

CRAIGHARRIS,MOSESJ.NAMKUNG,ANDMONT R. JUCHAU~

Department of Pharmacology. School ofbfedicine. University of U’ashington, Seattle. U’ushington 98195

Received June 17. 1986; accepted November 17. I986

Regulation of Intracellular Glutathione in Rat Embryos and Visceral Yolk Sacs and Its Effect on 2-Nitrosofluorene-Induced Malformations in the Whole Embryo Culture System. HARRIS, C., NAMKUNG, M. J., AND JUCHAU, M. R. (1987). Toxicol. Appl. Pharmacol. 88, 141-152. The dysmorphogenic effects of 2-nitrosofluorene (NF) in vitro were modulated in Day 10 rat em- bryos by agents which regulate intracellular glutathione (GSH) levels. The incidence of abnor- mal axial rotation caused by NF alone increased in a dose-dependent manner at NF concentra- tions in excess of 25 WM. No effects were observed at 15 fiM NF and doses of 100 pM resulted in a 100% incidence of mortality. L-Buthionine-S,R-sulfoximine (BSO), an inhibitor of GSH synthesis, produced malformations (50%) in embryos exposed to 15 pM NF but produced no additional effects on embryos at higher NF concentrations. BSO treatment alone resulted in a greater than 50% decrease in GSH content in visceral yolk sacs and had a lesser but likewise significant effect (15% decrease) on the GSH content of embryos. Protein content was inversely affected as embryonic levels were increased by 20% and yolk sac levels were unchanged. When BSO was added in combination with NF at the onset of the culture period, embryonic GSH decreased in a dose-dependent manner, suggesting a relatively low rate of embryonic GSH tum- over that could be increased by addition of an exogenous substrate capable of forming adducts with and removing GSH from the cells. 2-Oxothiazolidine-4-carboxylate (OTC). a compound which is enzymatically modified to provide an additional source of intracellular cysteine and increase GSH synthesis, produced no significant changes in embryonic or yolk sac GSH when added alone to the culture medium. When OTC (5 mM) was added in combination with NF, however, NF-elicited malformations were eliminated. This was also the case at 100 pM NF in which OTC not only prevented malformations but completely protected embryos against the loss in viability. The GSH and protein levels were indistinguishable from controls when OTC and NF were added simultaneously except for the 41 gM NF dose at which a highly significant increase in both embryonic and yolk sac protein was observed. This study clearly demonstrates the potential importance of GSH in the modulation of chemical dysmorphogenesis and provides an important new tool for the study of mechanisms of developmental toxicity. ID 1987 Academic

Press. Inc.

Progress in the elucidation of mechanisms in pattern of maternal-embryonic xenobiotic chemical teratogenesis has been restricted disposition and metabolism over an already due to the superimposition of an intricate complex and poorly understood process of

’ A portion of this work was presented at the 1986 An- nual Meeting of the Society of Toxicology in New Or- ’ Affiliate of the Child Development and Mental Re- leans, LA and have appeared in the Toxicologist 6, 97 tardation Center, University of Washington, Seattle, ( 1986). This research was supported by NIH Grants ES- WA. To whom correspondence and reprint requests 0404 1 and ES-07032. should be addressed.

141 004 1-008X/87 $3.00 Copyright GZ 1987 by Academic Press, Inc. All rights of reproduction in any form reserved.

142 HARRIS, NAMKUNG, AND JUCHAU

normal development. Recently it has become clear that mammalian embryos acquire the ability to bioactivate xenobiotics at early stages in their development and that reactive metabolic intermediates can be directly or in- directly responsible for the observed chemi- cal dysmorphogenesis (Juchau et al., 1978, 1985a,b). As with toxicity in other organs and tissues, the manifestations of toxicity in the developing conceptus should ultimately de- pend on the balance between bioactivation and detoxication reactions. Virtually nothing is known concerning intrinsic protective mechanisms available to embryos during early critical stages of their development or the role that modification of these processes would have on the normal development of the conceptus. Most conjugative enzyme ac- tivities normally associated with xenobiotic detoxication are extremely low except at late stages of gestation in experimental animals (Lucier et al., 1979). Low-molecular-weight thiols such as glutathione (GSH4.5)3 and its metabolites are undoubtedly present in pre- natal cells during all stages of development, however, and could exert significant protec- tive effects both nonenzymatically and via enzymatic pathways other than those used by the conjugative enzymes (Harris et al., 1986; Ketterer et al., 1983). Also of considerable importance is the potential of many of these thiols to regulate important metabolic and physiologic functions such as protein and DNA synthesis, microtubule assembly, maintenance of cellular structure and integ- rity, and several others (Kosower and Ko- sower, 1978; Zehavi-Willner et al., 197 1).

The whole embryo culture system provides a means to study embryonic growth without the influence of potentially confounding ma- ternal influences during the teratogen-sensi-

3 Abbreviations used: NF, 2nitrosofluorene; GSH, glutathione (reduced form); BSO, L-buthionine-5%sul- foximine; OTC, 2-oxothiazolidine-4-carboxylate; HBSS, Hanks’ balanced salt solution; DMSO. dimethyl sulfox- ide; PO,, sodium phosphate buffer (pH 8.0), EDTA, ethylenediaminetetraacetic acid, HPLC, high-perfor- mance liquid chromatography.

tive period of early organogenesis. This sys- tem also allows the convenient investigation of both embryo and visceral yolk sac which, because of their anatomic and physiologic in- terdependence, should be considered as a sin- gle developmental entity. Chemicals thought to produce malformations via different classes of reactive species also give rise to some qualitatively distinct morphological de- fects in vitro (Faustman-Watts et al., 1984; Greenaway et al., 1984). 2-Nitrosofluorene (NF), a deacetylated reactive meabolite ofthe procarcinogen and promutagen 2-acetylami- nofluorene, is a direct acting dysmorphogenic agent in the embryo culture system and pro- duces a striking defect characterized by in- complete axial rotation and abnormal trunk morphology (Faustman-Watts et al., 1984). NF is also rapidly and nonenzymatically con- verted to water-soluble, excretable, covalent adducts via conjugation with intracellular GSH (Mulder et al., 1982). In addition, me- tabolism of NF could lead to products capa- ble of undergoing redox cycling, in which case GSH could play an essential role in pro- tecting the conceptus from the reactive radi- cal species produced. We have, therefore, uti- lized NF as a model embryotoxin to assess the manner in which alterations in intracellular GSH levels can influence the incidence and severity of dysmorphogenesis produced in whole embryo cultures. Depletion of GSH in embryos and visceral yolk sacs by L-buthio- nine-S,R-sulfoximine (BSO) has been dem- onstrated previously (Harris et al., 1986). The 5-oxoproline analog, 2-oxothiazolidine-4- carboxylate (OTC), reportedly increases the rate of new GSH synthesis by providing intra- cellular cysteine, the normal rate-limiting component in GSH synthesis. OTC is be- lieved to be cleaved intracellularly by 5-0x0- prolinase, an enzyme of the y-glutamyl cycle, to form a decyclized intermediate that will undergo spontaneous decarboxylation to form L-cysteine (Williamson and Meister, 198 1). BSO and OTC have been employed in this study to modulate intracellular GSH lev- els of rat conceptuses grown in culture and to

GLUTATHIONE AND NITROSOFLUORENE MALFORMATIONS 143

evaluate the effects of altered GSH levels on NF-induced dysmorphogenesis.

METHODS

Animds. Primagravida Sprague-Dawley rats (Wistar- derived) were obtained on Day 5 or 6 of gestation from Tyler Laboratories (Bellevue, WA). Animals were al-

lowed free access to food and water and were housed un- der conditions described previously (Faustman-Watts cl a/.. 1984). The morning following copulation was desig- nated as Day 0 of pregnancy. Dams were anesthetized

with ether at 0900 hr on Day 10 and blood used for prep- aration of serum, a component of the culture medium, was collected from the abdominal aorta.

Embryo cz~lt~freandassessment. Conceptuses were ex- planted and cultured using procedures described by New ( 1973) as modified by Fantel rr al. (1979). Details of this

system have been published elsewhere (Faustman-Watts c( (I/.. 1984). Embryos had approximately 10 somites at the beginning of the culture period. For experiments in

vitro. L-BSO dissolved in Hanks’ balanced salt solution (HBSS) was added directly to the culture medium. OTC was dissolved in redistilled dimethyl sulfoxide (DMSO)

and added directly to the culture medium. Care was taken to keep the volume of DMSO small (20 &I5 ml of culture medium) and no adverse effects on embryo growth could be seen from adding DMSO alone.

Following the 24-hr culture period. conceptuses were placed in petri dishes containing HBSS and examined with a dissecting microscope. Conceptuses were regarded

as nonviable in the absence of an active vitelline circula- tion and were neither further evaluated nor included in

statistical comparisons except for those in which embry- olcthality was evaluated. Embryonic crown-rump length and somite number were recorded for each viable em-

bryo. Malformations induced by NF were not scored on the basis of severity due to the lack of a uniform scoring system. Embryos reported as “malformed” exhibited dysmorphogenesis ranging from complete lack of axial rotation to those which appeared normal other than a

ventro-lateral displacement of the trunk near the level of the forelimb bud. Embryos and yolk sacs were placed individually in 0.5 ml 0.1 M Pod-5 mM EDTA buffer, pH

8.0. (PHOS-EDTA) at the completion of the assessment and frozen at -70°C.

.~.ssa.r~. Thawed embryos and yolk sacs were ultrasoni- cally disrupted and utilized immediately for the detenni-

nation of tissue levels of GSH. Assays were performed essentially as described by Hissn and Hilf (1976). with minor modifications. Reaction mixtures for the determi-

nation of GSH contained 0.1 ml of tissue sample, 0. I ml of 0.1 M formic acid, 1 .O ml PHOS-EDTA. and 0.1 ml of 0phthaldialdehyde (Sigma Chemical Co., St. Louis, MO). I mg/ml. Fluorescence was measured at 345 nm (excitation) and 425 nm (emission). Amounts ofGSH as

low as 30 pmol could be accurately determined under

these conditions. Protein content was determined by the method of

Bradford (1976). DNA content was measured fluoro-

metrically as described by Labarca and Paigen, 1980). Student’s 1 tests were used to evaluate differences be- tween sample means. Significance was based on a 95%

confidence level (p < 0.05) and the numbers of determi- nations made in each case are indicated in the appropri- ate figures.

Chemicals. Synthesis of OTC was as described by Ka- neko et al. (1964) and modified by Shah ef al. ( 1979). Phosgene and cysteine were purchased from Fluka Bio-

chemicals (Hauppauge, NY). The OTC was recrystal- lized until a constant melting point was obtained. IR spectroscopy was used to confirm the identity of the product.

Synthesis of NF was performed according to the methods reported by Poirier et al. ( 1963) and the chemi-

cal was recrystallized until a constant melting point was obtained. The purity was checked with reverse-phase HPLC on a DuPont Zorbax C-8 column and found to be

greater than 99%. The chemical was stored dry at -70°C and was dissolved in redistilled DMSO immediately prior to each experiment.

RESULTS

The effects of adding BSO or OTC directly to the culture medium on viability and com- mon growth parameters (crown-rump length and somite number) are summarized in Ta- ble 1. Neither BSO (1 mM) nor OTC (5 mM) alone had any detectable adverse effects on the normal growth or viability of the em- bryos. NF, added to the culture medium at the beginning of the 24-hr culture period, produced embryos exhibiting abnormal trunk morphology. The particular defect is characterized by failure of the embryo to un- dergo the normal dorsal-to-ventral axial rota- tion that ordinarily occurs during this stage in development ( 10 to 11 days of gestation). Displacement of the trunk at the level of the upper limb bud was also frequently observed and occurred in the presence or absence of axial rotation. We have referred to this dis- placement as “kinking.” The defect arising from NF exposure in vitro ranges in severity from embryos with otherwise normal rota- tion and morphology (save a displacement of

144 HARRIS, NAMKUNG. AND JUCHAU

TABLE 1

EMBRYONIC GROWTH PARAMETERS AND GLUTATHIONE CONTENT (EXPRESSED PER mg OF PROTEIN) IN EM- BRYOSANDYOLKSACSOFCULTURED (24 ~~)CONCEPTU~ESEXPOSEDTONITROSOFLUORENEANDMODULATORS

OFGLUTATHIONECONTENTINTHEEMBRYOCULTURESYSTEM

GSH (nmol/mg protein)

Embryo Yolk sac Crown-rump (mm) Somites

Control 19.1 + 4.1 28.9 k 8.6 2.81 + 0.32 20.6 + 1.6

(77)” (62) (63) (63) BSO 12.3 + 3.3’+++ 12.6 + 4.9+++ 2.86 + 0.33 20.5 + 2.0

(36) (51) (44) (42) OTC 19.7 + 5.6 32.3 + 14. I 2.94 + 0.24+ 20.9 + 1.2

(36) (35) (39) (32) 25 pMNF 22.3 + 4.5”’ 38.8 + 14.3+++ 2.76 + 0.4 I 19.8 + 2.4

(35) (35) (29) (24)

+BSO 12.5 + 13.6”** 13.9 + 6.8*** 2.76 + 0.5 1 19.8 + 1.9

(33) (32) (24) (20) +OTC 22.0 + 5.9 40.5 + 17.1 2.96 + 0.26* 21.0 + 1.1*

(44) (41j (43) (38) 41 FMNF 16.2 + 4.8++ 24.6 + 3.0 2.51 + 0.31++ 20.1 t 0.8

(14) (14) (14) (81 +BSO 7.8 + l.l*** 5.9 + 1.2*** 2.53 + 0.30 20.7 + 1.0

(13) (13) (13) (6) +OTC 11.6 + 1.5** 15.8 + 2.7*** 2.73 + 0.26* 21.3 + 1.2*

(15) (14) (15) (15)

a Mean + SE (n). b ‘0.05, +‘O.Ol, and ‘+‘O.OOl different from media control.

’ *0.05, **O.O I, and ***O.OO 1 different from nitrosofluorene treatment alone.

the trunk near the level of the forelimb bud, Fig. Id) to those embryos which exhibit a complete lack of axial rotation as seen in Fig. le. It should be emphasized that the mor- phology of the cephalic region remained un- affected, except at the highest concentrations. NF-induced dysmorphogenesis in vitro has been described in detail in a previous publica- tion (Faustman-Watts et al., 1984). The mal- formation incidence reported here thus in- cludes all embryos with abnormal axial rota- tion morphology grouped as a common endpoint without regard to severity. No in- stances of incomplete neural tube closure were observed in any of the embryos exam- ined although some of the embryos exposed to doses of NF greater than 41 PM exhibited prosencephalic hypoplasia. This was likely due to generalized toxicity and growth retar- dation because control embryos removed

from culture at earlier time points exhibited very similar prosencephalic morphology (data not shown).

At NF concentrations as high as 15 PM, no abnormal embryos were observed. When NF (15 FM) was administered in combination with BSO (1 mM>. however, the incidence of malformations rose to nearly 50% (Fig. 2). Fi- nal NF concentrations of 25 I.LM resulted in an increased incidence of observed malfor- mations (to over 50%) without any loss of via- bility. When 1 mM BSO and 25 I.LM NF were added together at the beginning of the 24-hr culture period, the incidence of malforma- tions did not differ from that observed with 25 FM NF alone, suggesting that BSO has no potentiative effects under these conditions. Embryos exposed to 4 1 PM NF were found to be malformed in 90% of the cases and again, BSO produced no apparent effect on the per-

GLUTATHIONE AND NITROSOFLUORENE MALFORMATIONS 145

FIG. 1. Day 11 rat embryos after 24 hr in culture. (a) Embryo represents the media control, (b) embryo ex- posed to 1 mM BSO alone, (c) embryo exposed to 5 mM OTC alone. (d) embryo exposed to 4 1 pM NF alone. (e) embryo exposed to 41 pM NF + 1 mM BSO, and (f) em- bryo exposed to 41 +M NF + 5 mM OTC. The amniotic membranes have been left intact to eliminate possible mechanical distortion of the flexure defect. (d) Repre- sents a less severely malformed embryo while (e) repre- sents an embryo which has completely failed to undergo axial rotation.

centage of embryos exhibiting abnormal ax- ial rotation (Fig. 2). Additions of 100 PM NF or NF plus 1 mM BSO to the culture medium each resulted in 100% mortality based on our

,’ 75. ,/ ,” / z 50.

/I 2,’ ,’ I’ F? 7 2 ,’ /’

l No additfons A + InlM as0 0 +5mM OTC

Z-NITROSOFLUORENE (y Ml

FIG. 2. Embryos exhibiting abnormal trunk morphol- ogy following exposure to NF and BSO (1 mM) or OTC (5 mM) for 24 hr in the whole embryo culture system. All conceptuses exposed to 100 JLM NF and 100 pM NF + 1 mM BSO were determined nonviable by our criteria and thus were not included in the malformation incidence data.

Embryo

CTL 25 41

NF (p/M)

Yolk Sac

0 No Additions + I mM BSO

tZ7 +5mM OTC

CTL 25 41

NF (,/I/M)

FIG. 3. The total GSH content of individual embryos and yolk sacs at the end of the culture period measured as described under Methods. Error bars represent the SE and corresponding numbers represent the number of embryos or yolk sacs included in the analysis. values sta- tistically different from those of media controls (CTL, no additions): +p < 0.05. ++p < 0.0 1, and +‘+p < 0.00 1. Val- ues statistically different from those of NF treatment alone: *p < 0.05, **p i 0.0 I, and ***n < 0.00 I.

criteria for viability and thus were not as- sessed for malformations.

Simultaneous inclusion of NF at any con- centration ( 15- 100 PM) and OTC (5 mM) in the medium at the beginning of the culture period resulted in a nearly complete protec- tion against the NF-induced axial rotation defect. Comparisons are shown in Figs. Id and 1 f for embryos exposed to 4 1 PM NF. The protective effects of OTC are also reflected in the growth parameters shown in Table 1. At 100 PM NF, OTC was effective, not only in protection against the axial rotation defect, but also against the loss of viability. Embryos exposed to NF (100 PM) plus OTC (5 mM) exhibited 100% viability with only a 10% inci- dence of abnormal axial rotation but, based on common growth parameters such as crown-rump length, these embryos were slightly smaller than normal controls.

BSO, an inhibitor of GSH synthesis, was added to the medium at the beginning of the culture period and resulted in lower but non- significant decreases in embryonic GSH at the end of the 24-hr culture period (Fig. 3).

HARRIS. NAMKUNG, AND JUCHAU 146

:TL 25 41 I C

NF (,uM)

Yolk Sac

0 No addltlons +lmM ES0 ,

C]+5mMOTC : 15

CTL 25 41

NF (,uM)

FIG. 4. Protein content of embryos and yolk sacs at the end of the culture period. Error bars represent the SE and corresponding numerals represent the number of em- bryos or yolk sacs included in the analysis. Values statisti- cally different from those of media controls (CTL. no ad- ditions): +p < 0.05. “‘p 4 0.01, and +‘+p < 0.00 1. Values statistically different from those of NF treatment alone: *p < 0.05. **p < 0.01. and ***p < 0.001.

Yolk sacs from the same treatment group showed a differential GSH depletion with lev- els decreasing to 45% of control values as we reported earlier (Harris et al., 1986). Treat- ment with OTC alone failed to alter GSH lev- els significantly in either the embryo or the yolk sac (Fig. 3).

Concentrations of 25 to 4 1 PM NF failed to reduce GSH levels significantly below control values in either the embryo or the yolk sac (Fig. 3). A dose-dependent decrease in GSH was observed, however, in both the embryo and the yolk sac when NF was added simulta- neously with BSO at the beginning of the cul- ture period. Administration of OTC in com- bination with NF failed to cause significant deviations in total GSH levels when com- pared to control values.

The protein contents of embryos and yolk sacs were differentially affected by NF treat- ment as well as by NF plus BSO or OTC. BSO caused a differential, statistically significant (p < 0.01) increase of nearly 20% in embry- onic protein and a slight increase in DNA as we have reported earlier (Harris et al., 1986). Protein levels in embryos exposed to NF plus

BSO were also high relative to controls but increases in DNA were only marginally sig- nificant (Figs. 4 and 5). Protein values for em- bryos and yolk sacs treated with NF (25 and 4 1 /IM) alone or NF (25 PM) plus OTC (5 mM) did not differ from control values. When NF (41 PM) plus OTC was used, however, a greater than 30% increase in embryonic and yolk sac protein was observed (Fig. 4). DNA levels likewise were elevated in embryos and yolk sacs under these conditions (Fig. 5).

Standardization of GSH content on a per milligram protein basis provided a different picture of treatment effects on the embryo and yolk sac. The data presented in Table 1 represent effects of treatment on GSH levels that are reflective of the differential alter- ations of protein levels caused by the same treatment. This was especially evident for cases in which embryos were treated with any combination of NF and BSO. In those experi- ments, embryonic protein was differentially increased. NF (41 PM) plus OTC (5 mM> treatment resulted in a similar phenomenon but in both embryos and yolk sacs, GSH/pro- tein values were significantly lower than those of controls (NF alone) due to the large increases in protein in these conceptuses (Ta- ble 1).

Yolk Sac

I5 0 No additions &! + 1 mM 130

IO [3 +5mM OTC

2 0

5

0 CTL 25 41 CTL 25 41

NF (JIM) NF Q/M)

FIG. 5. DNA content of embryos and yolk sacs at the end of the culture period. Error bars represent the SE and corresponding numerals represent the number of em- bryos or yolk sacs included in the analysis. Values statisti- cally different from those of media controls (CTL, no ad- ditions): +p < 0.05, “p < 0.01, and +++p < 0.001. Values statistically different from those of NF treatment alone: *p < 0.05, **p < 0.0 1, and ***p < 0.00 1.

GLUTATHIONE AND NITROSOFLUORENE MALFORMATIONS 147

Embryo

‘1 Yolk Sac

FIG. 6. Glutathione content of embryos (0) and yolk sacs (A) throughout the 24-hr culture period. Control (open symbols) and 1 mM BSO (closed symbols) added at the onset of the culture period. Error bars represent the SE and each data point represents the mean of at least four determinations using embryos and yolk sacs from a minimum of three different litters.

Because the nature and severity of malfor- mations produced by a given compound may be influenced profoundly by the time in the developmental sequence in which the com- pound is administered, it was of interest to know the temporal nature of GSH modula- tion in the embryo and yolk sac in vitro. The rate of GSH depletion due to inhibition of GSH synthesis by BSO was much more rapid in the yolk sac than in the embryo proper (Fig. 6). GSH levels remained unchanged in the embryo during the first 4 hr in culture and then increased gradually during the culture period. Levels of GSH in yolk sacs, on the other hand, increased linearly over the entire culture period (Fig. 6). Addition of BSO to the culture medium at the onset of the culture period resulted in a trend of decreased GSH

that began as early as 1 hr for yolk sacs and not until approximately 4 hr for embryos (Fig. 6). However, changes were not found to be significant (p < 0.05) until 3 and 5 hr for yolk sacs and embryos, respectively.

When expressed on a per milligram protein basis, differences in GSH levels in the embryo vs the yolk sac were more dramatic (Fig. 7). Control values decreased in the conceptuses over the first 3 hr in culture then assumed a saltatory pattern for the remaining culture period. Although reproducible, these fluctu- ations may lack biological significance. Again, BSO-elicited reductions in yolk sac GSH were evident before 1 hr of culture had elapsed while embryonic decreases were not

Embryo

FIG. 7. Glutathione content of embryos (0) and yolk sacs (A) throughout the culture period as expressed per milligram of embryonic or yolk sac protein. Control (open symbols) and 1 mM BSO (closed symbols) added at the onset of the culture period. Error bars represent the SE and each data point represents the mean of at least four determinations using embryos and yolk sacs from a minimum of three different litters.

148 HARRIS, NAMKUNG. AND JUCHAU

Yolk Sot

Time in Culture (hr)

FIG. 8. Glutathione content of Day 10 rat embryos and yolk sacs cultured in vitro. The controls (0) contained media only. The time of addition of OTC (5 mM) alone (a) was designated as time 0. NF (41 pM) was given to the conceptuses in the complete culture medium 1 hr prior to the time of addition of OTC and was cultured as NF alone (0) or NF + OTC (A). Each point represents the mean of at least three embryos or yolk sacs from a total of six different litters. With the exception ofthe time point of interest (in the yolk sac at 3 hr), the SE were omitted for clarity. Overall SE values varied between 40 and 160 pmol.

apparent until after at least 4 hr. Differences between control and BSO-treated samples were much more pronounced for yolk sacs than for embryos over the total time in cul- ture (Fig. 7). The time course of OTC effects was not significantly different from control values at any of the time points tested, when given alone at the onset of the culture period (Fig. 8). When added at the onset of the cul- ture period, NF alone (41 PM) did not cause significant changes in GSH levels in either embryos or yolk sacs during the first 5 hr of culture. However, if OTC (5 mM) was added to the culture medium 1 hr after addition of NF, an apparent increase in yolk sac GSH

was observed at 3 hr (Fig. 8). This change was not statistically significant, however, and it has not been determined whether this rise in GSH may represent a general trend for the entire yolk sac or possibly an increase in GSH in specialized cell types.

The time required to deplete GSH in em- bryos and yolk sacs suggested a possible rea- son for the failure of BSO to further increase the malformation incidence in NF-exposed embryos when NF and BSO were added si- multaneously to the culture medium at the beginning of the culture period. When suffi- cient time was allowed to elapse (3 hr after the addition of BSO at the beginning of the culture period) to result in maximal GSH de- pletion in the yolk sac, potentiation of NF- induced malformations could be demon- strated (Table 2). These effects could be seen even at NF concentrations as high as 4 1 PM, whereas we could detect no differences be- tween embryos treated with NF alone and embryos given NF plus BSO when both were added simultaneously to the culture medium (Fig. 2).

DISCUSSION

Obvious similarities between xenobiotic biotransformation in embryonic and adult tissues suggest that reactive intermediates may play an important role in the etiology of chemical teratogenesis. Likewise the degree of damage caused by any potentially toxic compound is directly related to the capabili- ties of the cells to inactivate and/or remove the reactive species through means such as those involving GSH and related low-molec- ular-weight thiols. For these reasons we have undertaken a project to determine the extent to which modulation of GSH levels can in- fluence the incidence and/or severity of mal- formations caused by NF, an agent known to elicit dysmorphogenesis in vitro. Because of their high degree of interdependence during the critical period of early organogenesis, both the embryo and visceral yolk sac have

GLUTATHIONE AND NITROSOFLUORENE MALFORMATIONS 149

TABLE 2

Embryo Yolk sac

Protein GSH (pmol/ Protein GSH (pmol/ Malformation Treatment bg/embryo) embryo) (rgholk sac) yolk sac) (%)’

Control 179.9 f 15.0 3418t-481 86.6 + 10.5 1527? 187 (4)b (4) (4) (4) 0

1 mM BSO 220.4 f 25.9* 2923 f 397 106.2 -t 13.5 13182264 (6) (6) (6) (6) 0

20 PM NF 210.1 i 39.8 4229 F 683 105.0 f 24.1 2597 F 397*** C-9 (8) (8) (8) 13

~OHMNF+ 1 mMBS0 180.7 f 22.8 2250 rt 240** 97.5 + 13.0 1075t 72 (8) (8) (8) 03) 75***

41 fin NF 75.1 f 35.8*** 1856 Z!I 706*** 51.4& 18.3** 2164 t 546* (8) (8) (8) (8) 63***

41 WMNF+ 1 mMBS0 85.7 -+ 35.7*** 979 ?z 525*** 55.7 f 19.9** 606 k 289*‘* (8) (8) (8) (8) 15***

’ BSO ( 1 mM) was added to the culture medium at the beginning of the culture period. After 3 hr NF was added to the appropriate bottles (those with and without BSO) and conceptuses were cultured as usual for the remainder of the 24 hr period.

’ Mean + SD(N). ’ Malformations include all embryos exhibiting incomplete axial rotation or abnormal trunk morphology as de-

scribed under Methods. ii *0.05. **O.O I, and ***O.OO I different from media control.

been considered as possible targets in the pathway leading to chemically elicited abnor- mal morphology. Direct and/or indirect modulation of GSH levels has been accom- plished through utilization of an agent that specifically produces decreased GSH levels by inhibiting new synthesis (BSO) (Griffith et at., 198 1; Griffith and Meister, 1979a) as well as through use of a separate agent (OTC) that specifically accelerates GSH synthesis by pro- viding the rate-limiting precursor, cysteine (Williamson and Meister, 198 1). The rate and degree of depletion of GSH by BSO is de- pendent on the turnover of GSH in a particu- lar tissue (Griffith and Meister, 1979b). The differential effects of BSO on embryonic vs yolk sac GSH may partially reflect the ability of the yolk sac to turn over GSH at a higher rate. This conjecture is supported by unpub- lished observations from our laboratory that

the yolk sac has a two- to threefold higher y- glutamyl transpeptidase (GGT) activity than does the embryo proper. GGT is recognized as highly important in GSH turnover (Wil- liamson et al., 1982). Depletion of GSH in the embryo was evident only when an addi- tional substrate (NF) for GSH utilization was provided, allowing for increased GSH tum- over through possible adduct formation (Fig. 3). Although NF-SG adducts have been char- acterized (Mulder et al., 1982) we have not, as yet, attempted to quantitate these products in cultured embryos. The effects of OTC did not occur differentially in the embryo vs yolk sac. This is probably due to the proposed di- rect effects of cysteine in all cells capable of synthesizing GSH (Williamson et al., 1982).

It was expected that depletion of GSH by BSO would serve to potentiate the fetotoxic and/or dysmorphogenic effects of NF. This

150 HARRIS, NAMKUNG, AND JUCHAU

was apparent only at very low doses of NF ( 15 pM) under conditions in which NF alone produced no abnormal morphology (Fig. 2). It is perhaps not surprising that the incidence of malformation in embryos cultured at higher doses of NF was not changed by simul- taneous administration of BSO. This lack of expectation is based on previous work in which we showed that the NF-induced axial rotation defect can be produced under condi- tions in which conceptuses were preincu- bated for 2 hr with NF in HBSS prior to trans- fer into NF-free culture medium (Faustman- Watts et al., 1986). Because of the temporal nature of GSH depletion by BSO, we could easily expect the NF-elicited damage to have occurred before there were decreases in GSH sufficient to potentiate the effects. When BSO was added to the culture medium at the be- ginning of the culture period and allowed sufficient time (3 hr) to deplete GSH to its lowest observed levels in the yolk sac (Table 2, Fig. 7) subsequent additions of NF were accompanied by a greater incidence of mal- formations than was seen with NF given alone at the same time point (Table 2). These results suggest that the maintenance of yolk sac GSH levels can be of critical importance in the protection of the embryos from xenobi- otics and that factors which disturb GSH ho- meostasis also could act to increase embry- onic susceptibility to xenobiotic insult.

The simultaneous administration of OTC and NF resulted in a dramatic protection of embryos from NF-induced dysmorphogen- esis. It is uncertain at this point whether the effects of OTC represent a recovery of the em- bryo from the NF-induced insult or whether OTC prevented the occurrence of the lesion. It seems unlikely that OTC would interact di- rectly with NF to exert its protective effects although the cysteine produced by the action of 5-oxoprolinase may form adducts with NF and thereby serve as a detoxication mecha- nism. No information is currently available to confirm the extent of NF-cysteine or NF- GSH adduct formation in the embryo culture system. Any effects of OTC would, in con-

trast to BSO administration, be rapid and di- rect due to the intracellular production of cysteine and the potential for immediate uti- lization in cellular processes. The potential for OTC both to protect from and repair newly elicited damage has already been shown in adult liver in response to paraceta- mol-induced hepatic injury (Williamson et al., 1982). Further studies have suggested that the ability to provide precursors for new GSH synthesis rather than the direct action of cys- teine is the basis for the protective action of OTC. (Hazelton et al., 1986). Attempts to show significant decreases in embryonic or yolk sac GSH during the first 5 hr of culture following administration of NF (4 1 pM) were unsuccessful. This does not, however, pre- clude the possibility of selective depletion in only a relatively small population of suscepti- ble cells. Higher doses of NF, which could re- sult in measurable decreases in intracellular GSH, were not possible due to the overt tox- icity of this compound. OTC alone did not elevate GSH levels above control values in ei- ther embryo or yolk sac during the same pe- riod, presumably because of the feedback in- hibition of y-glutamyl cysteinyl synthetase activity by GSH. Under conditions in which conceptuses had been exposed to NF for 1 hr prior to addition of OTC, a small but tran- sient increase in GSH was observed in the yolk sac (Fig. 8). This suggests that sufficient GSH may have been utilized by NF to relieve the feedback inhibition but that NF concen- trations may not have been high enough to elicit the overshoot in GSH concentrations observed in other systems in which consider- able depletion was achieved prior to OTC ad- dition (Williamson et al. 1982).

The differential increases in embryonic protein in response to BSO treatment may re- flect regulatory influences of any of a number of intermediates of GSH synthesis in the y- glutamyl cycle. It has been suggested that low-molecular-weight thiols such as cysteine or cysteinyl-glycine play a regulatory role in lysosomal proteolysis (Kooistra et al.. 1982; Mego, 1985). Intracellular increases in these

GLUTATHIONE AND NITROSOFLUORENE MALFORMATIONS 151

thiols, produced by BSO blockade of the GSH synthetic pathway by BSO, could result in increased proteolytic activity in the yolk sac and, consequently, in the subsequent de- livery of larger quantities of amino acid pre- cursors to the embryo. These proteolytically derived amino acids would then be utilized for new protein synthesis. It has also been proposed that GSH functions directly in the regulation of both protein and DNA synthe- sis (Kosower and Kosower, 1978) yet virtu- ally nothing is known about such effects on the embryo or yolk sac during embryogen- esis. The nonselective effects of OTC on GSH, with or without NF, on increasing pro- tein levels (yolk sac vs embryo) tend to invoke mechanisms independent of those proposed for BSO. The OTC-elicited increases in pro- tein and DNA were observed only when higher NF concentrations were used. This suggested that new macromolecule synthesis may be a result of the protective functions of OTC and thus resemble a rebound resynthe- sis phenomenon. A similar response with re- spect to protein occurred when conceptuses were preincubated with NF in HBSS contain- ing 1 mM GSH prior to culture in NF-free medium (Faustman-Watts et al., 1986).

The effects of two methods of GSH modu- lation in embryos and yolk sacs on NF-in- duced dysmorphogenesis have been demon- strated. The reproducibility of effects and the low toxicity resulting from BSO or OTC ad- ministration to conceptuses in the whole em- bryo culture system suggests that these com- pounds may be utilized as tools to provide valuable insights into the mechanisms of chemical teratogenesis and into the roles of low-molecular-weight thiols in the regulation of normal growth and differentiation.

ACKNOWLEDGMENT

The authors thank Joan Russell for her assistance in the preparation of this manuscript.

REFERENCES

BRADFORD, M. M. (1976). A rapid, sensitive method for quantitation of microgram quantities of protein utiliz-

ing the principle of protein-dye binding. Anal. Eio-

them. 12,248-X4.

FANTEL, A. G., GREENAWAY, J. C., JUCHAU, M. R., AND SHEPARD, T. H. (1979). Teratogenic bioactiva- tion of cyclophosphamide in vitro. Life Sci. 25,6?-72.

FAUSTMAN-WATTS, E. M., NAMKUNG, M. J.. AND Ju- CHAU, M. R. (1986). Modulation of the embryotoxi- city in vitro of reactive metabolites of 2-acetylami- nofluorene by reduced glutathione and ascorbate and via sulfation. Toxicol. .4ppl. Pharmacol. 86,400-4 IO.

FAUSTMAN-WATTS. E. M., GREENAWAY, J. C.. NAM- KUNG, M. J.. FANTEL, A. G., AND JUCHAU, M. R. ( 1984). Teratogenicity in vitro of two deacetylated me- tabolites of N-hydroxy-2-acetylaminofluorene. To.% co/. .4ppl. Pharmacol. 76, 16 I- 17 1,

GREENAWAY. J. C., BARK, D. H., AND JUCHAU, M. R. ( 1984). Embryotoxic effects of sahcylates: Role of bio- transformation. To.xicol. Appl. Pharmacol. 74, 14 I - 149.

GRIFRTH, 0. W. ( 198 I ). Depletion of glutathione by in- hibition of biosynthesis. In Methods in Enz.vmo1og.v (W. B. Jakoby, Ed.), Vol. 77, pp. 59-63. Academic Press, New York.

GRIFFITH, 0. W., AND MEISTER, A. (1979a). Potent and specific inhibition of glutathione synthesis by buthio- nine sulfoximine (S-n-butylhomocysteine sulfoxi- mine). J. Biol. Chem. 254.7558-7560.

GRIFL~TH, 0. W., AND MEISTER. A. (1979b). Glutathi- one interorgan translocation, turnover. and metabo- lism. Proc. Natl. Acad. Sci. USA 76,5606-56 10.

HARRIS, C., FANTEL, A. G., AND JUCHAU, M. R. (1986). Differential glutathione depletion in rat embryo and visceral yolk sac in viva and in vifro by L-buthionine- S.R-sulfoximine. Biochem Pharmacol., 35. 4431- 4441.

HAZLETON, G. A., HJELLE. J. J., AND KLAASSEN, C. D. ( 1986). Effects of cysteine pro-drugs on acetamino- phen-induced hepatotoxicity. J Pharmacol. E.rp. Ther. 237, 341-349.

HISSN, P. J.. AND HILF, R. (1976). A fluorometric method for determination of oxidized and reduced glutathione in tissues. ,4nal. Biochem. 74,2 14-226.

JUCHAU, M. R., BARK, D. H., SHEWEY. L. M.. AND GREENAWAY. J. C. (1985a). Generation of reactive dysmorphogenic intermediates by rat embryos in cul- ture: Effects of cytochrome P-450 inducers. To.xicol.

.4ppl. Pharmacol. 82,533-544.

JUCHAU, M. R.. GIACHELLI, C. M.. FANTEL. A. G., GREENAWAY, J. C., SHEPARD, T. H.. AND FAUST- MAN-WATTS, E. M. (1985b). Effects of 3-methylcho- lanthrene and phenobarbital on the capacity of em- bryos to bioactivate teratogens during organogenesis. Toxicol. .4ppl. Pharmacol. 80, 137- 146.

JUCHAU, M. R.. NAMKUNG. M. J., JONES, A. A., AND DIGIOVANNI. J. (1978). Biotransformation and bioac- tivation of 7.12-dimethylbenzo(a) anthracene in hu-

152 HARRIS, NAMKUNG, AND JUCHAU

man fetal and placental tissues. Drug Metab. Dispos. 6,273-28 1.

KANEKO, T., SHIMOKOBE, T.. OTA, Y., TOYOKAWA, E., INUI, T., AND SHIBA, T. (1964). Syntheses and proper- ties of 2-oxothiazolidine-4-carboxylic acid and its de- rivatives. Bull. Chem. Sot. (Japan) 37,242-244.

KETTERER, B., COLES, B., AND MEYER. D. J. (1983). The role of glutathione in detoxication. Environ. Health Perspeet. 49,59-69.

KOOISTRA, T.. MILLARD, P. C., AND LLOYD, J. B. ( 1982). Role of thiols in degradation of proteins by ca- thepsins. Biochem. J. 204,47 l-471.

KOSOWER, N. S., AND KOSOWER, E. M. (1978). The glu- tathione status of cells. ht. Rev. Cytol. 54, 109- 160.

LABARCA, L., AND PAIGEN, K. (1980). A simple rapid and sensitive DNA assay procedure. Anal. Biochem. 102,344-352.

LUCIER, G. W., LUI, E. M. K., AND LAMARTINIERE, C. A. (1979). Metabolic activation/deactivation reac- tions during perinatal development. Environ. Health Perspect. 29,7- 16.

MEGO, J. L. (1985). Stimulation of intralysosomal prote- olysis by cysteinyl-glycine, a product of the action of y-glutamyl transpeptidase on glutathione. Biochim. Biophys. Acta 841, 139-144.

MULDER, G. J., UNRUH, L. E., EVANS, F. E., KETTERER. B., AND KADLUBAR, F. F. (1982). Formation and

identification of glutathione conjugates from 2-nitro- sofluorene and N-hydroxy-2-aminofluorene. Chem.- Biol. Interact. 39, 1 I - 127.

NEW, D. A. T. (1973). Studies on mammalian fetuses in vitro during the period of organogenesis. In The Mam- malian Fetus in Vitro (C. R. Austin, Ed.), pp. 16-66. Wiley, New York.

POIRIER, L. A., MILLER, J. A., AND MILLER, E. C. ( 1963). The N- and ring hydroxylation of 2-acetylami- nofluorene and the failure to detect N-acetylation of 2- aminofluorene in the dog. Cancer Rex 23,790-797.

SHAH. H., HARTMAN, S. P., AND WEINHOUSE, S. (1979).

Formation of carbonyl chloride in carbon tetrachlo- ride metabolism by rat liver in vitro. Cancer Rex 39, 3942-3947.

WILLIAMSON, J. M., BOETTCHER, B., AND MEISTER, A. (1982). Intracellular cysteine delivery system that pro- tects against toxicity by promotingglutathione synthe- sis. Proc. Natl. Acad. Sci. USA 79.6246-6249.

WILLIAMSON, J. M., AND MEISTER, A. (1981). Stimula- tion of hepatic glutathione formation by administra- tion of L-2-oxothiazolidine-4-carboxylate, a S-OXO-L- prolinase substrate. Proc. Natl. Acad. Sci. LISA 78, 936-939.

ZEHAVI-WILLNER. T.. KOSOWER, E. M.. HUNT, T., AND KOSOWER, N. S. (1971). Glutathione V. The effects of the thiol-oxidizing agent diamide on initiation and translation in rabbit reticulocytes. Biochem. Biophys. Acta 228.245-25 I.