C 2014 Wiley Periodicals, Inc.
Birth Defects Research (Part B) 101:429–437 (2014)
Close Link between Protoporphyrin IX Accumulation and Developmental Toxicity Induced by N-Phenylimide Herbicides in Rats Satoshi Kawamura,1∗ Terushige Kato,1 and Alan G. Fantel2 1 Environmental
Health Science Laboratory, Sumitomo Chemical Co., Ltd., Konohana-ku, Osaka, Japan of Pediatrics, University of Washington, Seattle, Washington
BACKGROUND: S-53482, 7-fluoro-6-[(3,4,5,6-tetrahydro)phthalimido]-4-(2-propynyl)-1,4-benzoxazin-3(2H)-one (flumioxazin), is an N-phenylimide herbicide and developmentally toxic to rats, but not to rabbits. The day of greatest sensitivity to S-53482 is gestational day (GD) 12 in rats. There is a compound-specific difference in developmental toxicity among structurally similar compounds including S-23121 (N-[4-chloro-2-fluoro-5-[(1-methyl2-propynyl)oxy]phenyl]-3,4,5,6-tetrahydrophthalimide; teratogenic) and S-23031 (pentyl 2-chloro-4-fluoro-5-(3,4,5,6tetrahydrophthalimido)phenoxyacetate (flumiclorac pentyl); nonteratogenic). The herbicidal action is due to photodynamic action of accumulating protoporphyrin IX (PPIX), resulting from the inhibition of protoporphyrinogen oxidase (PPO), an enzyme in porphyrin biosynthesis. Species difference in PPIX accumulation in embryos corresponded to those of the developmental toxicity. Our objective in this study was to further investigate a link between PPIX accumulation resulting from PPO inhibition and developmental toxicity. This article is part of a series of studies to be published serially. METHODS: To investigate compound-specific differences, each compound was orally administered to rats on GD 12. To define peak period of PPIX accumulation, single oral treatments of S-53482 were given to rats or rabbits at 19:30 on GD 10 through GD 15. PPIX was extracted from embryos 14 hr after treatment. RESULTS: Remarkable PPIX accumulation was observed when treated with S-53482 or S-23121, but not with S-23031. The greatest accumulation of PPIX was observed when treated with S-53482 at 19:30 on GD 11 or GD 12. No PPIX accumulation was found on any GDs in rabbits. CONCLUSIONS: The developmentally toxic compounds caused PPIX accumulation in embryos. The peak period of PPIX accumulation corresponded to that of developmental effects. This correlation suggests a close link between PPO C 2014 Wiley Periodicals, Inc. inhibition and developmental abnormality. Birth Defects Res (Part B) 101:429–437, 2014.
Key words: N-phenylimide herbicide; compound-specific difference; peak period; porphyrin accumulation; teratogenicity; developmental toxicity; rats; rabbits
INTRODUCTION S-53482, 7-fluoro-6-[(3,4,5,6-tetrahydro)phthalimido]-4(2-propynyl)-1,4-benzoxazin-3(2H)-one (flumioxazin), is an N-phenylimide herbicide that shows a striking effect in particular on broadleaf weeds. The herbicide produced embryolethality, teratogenicity (mainly ventricular septal defects [VSD] and wavy ribs), and growth retardation when administered orally to pregnant rats on gestational days (GDs) 6 through 15 at 30 mg/kg/day. Maternal toxicity was not observed (Kawamura et al., 1995). In previous studies, we conducted a series of investigations on developmental effects of S-53482. The major characteristics of oral developmental toxicity induced by S-53482 are species difference, sensitive period, and compound-specific difference. There is a remarkable
species difference in developmental toxicity between rats and rabbits. In contrast to rats, S-53482 caused no developmental toxicity at a maternal toxic dose of 3000 mg/kg/day in rabbits. To define the period of greatest sensitivity to the herbicide, single oral treatments of S-53482 at 400 mg/kg were given to pregnant rats on different GDs. Results showed that GD 12 is the day of greatest sensitivity (Kawamura et al., 1995). ∗ Correspondence to: Satoshi Kawamura, Environmental Health Science Laboratory, Sumitomo Chemical Co., Ltd., 1-98, Kasugade-naka 3-chome, Konohana-ku, Osaka 554-8558, Japan. E-mail: [email protected]
Received 24 September 2014; Accepted 06 November 2014 Published online in Wiley Online Library (wileyonlinelibrary.com/journal/ bdrb) DOI: 10.1002/bdrb.21133
KAWAMURA ET AL.
Fig. 1. Chemical structures of N-phenylimide herbicides.
We have also found a compound-specific difference in developmental toxicity among N-phenylimide compounds structurally similar to S-53482 (Fig. 1) in rats and rabbits. S-23121, N-[4-chloro-2-fluoro-5-[(1-methyl2-propynyl)oxy]phenyl]-3,4,5,6-tetrahydrophthalimide, at 20 mg/kg induced the same pattern of developmental toxicity in rats as did S-53482 at 30 mg/kg, including embryonic lethality, growth retardation, and teratogenicity, consisting mainly of VSD and wavy ribs. In contrast, developmental toxicity was not observed in rats treated with S-23031, pentyl 2-chloro-4-fluoro-5-(3, 4,5,6-tetrahydrophthalimido)phenoxyacetate (flumiclorac pentyl), even at a dosage level of 1500 mg/kg. Neither S-23121 nor S-23031 exhibited developmental toxicity in rabbits at the highest dose levels tested. Thus, all N-phenylimide herbicides tested do not show teratogenic potential in rabbits (Kawamura et al., 2013). S-23121 and S-23031 show herbicidal effects on weeds similar to S-53482. As a whole, S-53482 and S-23121 are more effective than S-23031. The herbicides inhibit plant protoporphyrinogen oxidase (PPO), the target enzyme of the herbicides. Concentrations of 50% inhibition of radish PPO ranged from 0.49 nM of S-53482 to 1.27 nM of S-23031 (unpublished data). We also investigated whether or not S-53482 and S23121 induce developmental toxicity via the dermal route, which is more relevant to occupational exposure, hence better addressing human health risks. Dermal exposure of rats to S-53482 at 300 mg/kg and to S-23121 at 800 mg/kg produced developmental toxicity similar to that resulting from oral exposure. Investigation of the mechanism and its human relevancy, therefore, becomes more important (Kawamura et al., 2014). In an extensive research program to elucidate mechanism behind the observed developmental toxicity of S-53482, we first studied the herbicidal action and the target enzyme of S-53482. The herbicidal activity of S-53482 is due to photodynamic action of accumulating protoporphyrin IX (PPIX) in plants, resulting from inhibition of PPO that is an enzyme in porphyrin biosynthesis. The enzyme is common to plants and animals as part of chlorophyll and heme biosynthesis (Fig. 2). Studies by Matringe et al. (1989a, 1989b) have shown that plant PPO inhibitors can inhibit animal PPO. Several diphenyl ether-type herbicides that inhibit
Fig. 2. Outline of heme and chlorophyll biosynthesis.
PPO in chlorophyll biosynthesis as does S-53482 inhibit mouse liver mitochondrial PPO. Moreover, the herbicides differently inhibit PPOs from different sources, such as plants, yeasts, and mice. Then, it was conceivable that N-phenylimide herbicides could inhibit rat mitochondrial PPO with different levels of inhibition and that an N-phenylimide herbicide could differently inhibit PPOs from rats and rabbits. As an initial step in examining a possible link between enzyme inhibition and developmental toxicity, we have investigated differences in PPIX accumulation presumably resulting from PPO inhibition between rat and rabbit embryos. The species difference in PPIX accumulation corresponds very well to that of the developmental toxicity caused by S-53482. Remarkable Birth Defects Research (Part B) 101:429–437, 2014
LINK BETWEEN PROTOPORPHYRIN ACCUMULATION AND TERATOGENICITY PPIX accumulation was observed in rat embryos, but not in rabbit embryos following treatment with S-53482 at 1000 mg/kg on GD 12 (Kawamura et al., 1996a). Our objective in this study was to further investigate a possible link between PPIX accumulation resulting from PPO inhibition and developmental toxicity by determining whether there is a compound-specific difference in PPIX accumulation in rat embryos and its correlation with developmental toxicity, and whether there is peak periods of PPIX accumulation and its correspondence with the critical period of sensitivity to S-53482 developmental toxicity.
MATERIALS AND METHODS Chemicals S-53482, S-23121, and S-23031 were supplied as powder by Sumitomo Chemical (Tokyo, Japan). Structures are shown in Figure 1.
Animals Female Sprague-Dawley rats were purchased from Charles River Japan (Kanagawa, Japan). Males of the same strain were used for breeding. The rats were housed in hanging wire cages in rooms where air was exchanged more than 10 times per hour. Temperature, humidity and artificial light cycle were maintained at 22 to 26°C, 45 to 65%, and 12 hr (8:00–20:00 light), respectively. Pelleted rodent diet and tap water were given ad libitum. One female was paired with one male overnight in a cage. The next morning, copulation was confirmed by the presence of vaginal plugs. The observation day of the plug was designated as GD 0. Female JW-NIBS rabbits were purchased from Nisseiken (Tokyo, Japan). The rabbits were housed in hanging wire cages in rooms where air was exchanged more than 10 times per hour. Temperature, humidity, and artificial light cycle were maintained at 20 to 24°C, 45 to 65%, and 12 hr (8:00–20:00 light). Pelleted rabbit diet and tap water were given ad libitum. The semen from a few males of the same strain was pooled and used to inseminate females. Immediately after insemination, females were given an intravenous injection of 50 IU chorionic gonadotropic hormone (Sankyo Zoki, Tokyo, Japan) to promote ovulation. The day of artificial insemination was designated as GD 0. Female animals were evaluated for their health based on clinical signs and body weights during quarantine. Healthy animals were subjected to experiments. Since a single dose of each herbicide at 5000 mg/kg to female rats or repeated doses of 1000 mg/kg S-53482 to female rabbits shows no effect on treated animals, we conducted cage-side observations during experimental periods. There were no adverse signs related to treatment. Rats and rabbits were euthanized 14 hr after treatment (9:30 the next morning) by cervical dislocation and by iv injection with sodium pentobarbital, respectively.
Experimental Design To study compound-specific differences, the test chemicals S-53482, S-23121, and S-23031 were suspended in 0.5% methylcellulose. The dosing suspensions of the test Birth Defects Research (Part B) 101:429–437, 2014
chemicals were administered to four to five pregnant rats per test chemical group at 1000 mg/kg by oral intubation around 19:30 on GD 12, the most sensitive day for toxicity (Kawamura et al., 1995). In standard developmental toxicity studies, the developmental lowest observed effect levels of S-53482 (30 mg/kg) and S-23121 (20 mg/kg) are comparable and very much lower than that of S-23031 (more than 1500 mg/kg; Kawamura et al., 1995, 2013). A single administration of S-53482 at 1000 mg/kg on GD 12 produced a striking accumulation of PPIX in rat embryos (Kawamura et al., 1996a). A dose level of 1000 mg/kg was, therefore, expected to demonstrate obvious compound difference in PPIX accumulation in rat embryos among the test chemicals if such a difference exists. The dose administered to a female was based on individual body weight on GD 6 in the same manner as the previous studies. To determine peak periods, S-53482 was suspended in 0.5% methylcellulose. The dosing suspension was administered to three to five rats and rabbits at 400 and 1000 mg/kg, respectively, by oral intubation at 19:30 on different GDs from GD 10 through GD 15. The dose level of 400 mg/kg for rats was identical to that of the in vivo critical period-finding study (Kawamura et al., 1995). The dose level of 1000 mg/kg for rabbits was set to that of the previous study to find species difference in PPIX accumulation (Kawamura et al., 1996a). The dose administered to a female was based on individual body weight on GD 6. A control group of pregnant rats received 0.5% methylcellulose in the same dosing scheme. A vehicle group of pregnant rabbits was not set up to reduce the number of rabbits, taking animal welfare into account. Because rabbits treated on GD 12 were insensitive to S-53482 and PPIX concentrations in treated fetuses were as low as the detection limit in the previous study, evident PPIX accumulation in treated rabbits was expected to be noticeable if it occurs on any of GDs during a treatment period.
Measurement of PPIX in Tissues To study compound-specific differences, three embryos per litter were removed from the rat uterus and pooled by litter 14 hr after the treatment (9:30 the next morning). The 14-hr time period approximates the peak time point of PPIX accumulation (12 hr after the treatment) shown in a previous study with S-53482 (Kawamura et al., 1996a). Wet weight of each embryo pool was measured. PPIX was extracted from embryo pools as previously described (Kawamura et al., 1996a). Each embryo pool was homogenized in methanol: 0.1N NH4 OH (9:1 v/v, basic methanol). The homogenate was centrifuged at 10,000 × g for 10 min below 4°C. The pellet was resuspended in the basic methanol, sonicated for 30 sec, and centrifuged at 10,000 × g for 10 min below 4°C. The supernatant was pooled for PPIX analysis. PPIX thus extracted was analyzed by high-performance liquid chromatography fitted with a fluorescence spectrophotometer and a 250 × 4.6 mm reversed phase column (Nucleosil 100–5C18 GL-PACK). The solvent was 0.1 M ammonium phosphate:methanol (14:111 v/v) at a flow rate of 1.0 ml/min. Excitation and emission wavelengths for PPIX were 405 and 630 nm, respectively. A sample of maternal liver weighing 0.2–0.4 g was dissected out and PPIX in
KAWAMURA ET AL.
maternal livers was extracted and measured by the same procedure as for embryos. For a peak period finding study, except in cases mentioned below, three embryos per litter were removed from the uterus treated with S-53482 or 0.5% methylcellulose, and pooled by litter 14 hr after the treatment (9:30 the next morning). In rats, each embryo pool of both the control and treated groups on GD 10 consisted of all embryos from a litter. Each embryo pool of the control group on GD 11 consisted of 10 embryos from a litter. In rabbits, GD 10 embryo pools consisted of five embryos from a litter. Wet weight of each embryo pool was measured. PPIX was extracted from embryo pools by the procedure described above. Maternal liver samples weighing 0.16 to 0.4 g were dissected out and PPIX in maternal livers was extracted and measured by the same procedure as for embryos.
RESULTS Compound-Specific Difference in PPIX Accumulation Wet weights of rat embryos were comparable among the control and treated groups (Fig. 3A). Concentrations of PPIX in rat embryos were obviously different among the three herbicides at 14 hr after the treatment. No treatment-related accumulation of PPIX was observed in rat embryos exposed to 1000 mg/kg S-23031, a chemical that has shown no developmental toxicity in rats. The concentration of PPIX per gram of embryonic tissue in rats treated with S-23031 (0.118 g/g) was similar to the control value (0.095 g/g). In contrast, treatment with 1000 mg/kg of either S-23121 or S-53482, chemicals that have demonstrated obvious developmental effects in rats, produced remarkable accumulation of PPIX in treated embryos. The concentration of PPIX in embryos treated with S-23121 (25.850 g/g) or S-53482 (27.619 g/g) was greater than that in the control group embryos by two orders of magnitude (Fig. 3B). Thus, the accumulation of PPIX in rat embryos corresponded very well to the developmental toxicity produced by the chemicals. Compound-related accumulation of PPIX in maternal liver was similar to that noted for embryos. Mean PPIX concentrations in rat liver specimens from animals treated with S-23121 (0.857 g/g) and S-53482 (0.679 g/g) were higher than that of the control (0.236 g/g). Each individual value for PPIX concentration in liver specimens from animals treated with S-23121 or S-53482 was higher than the highest value in the control liver specimens. The average PPIX concentration in liver of animals treated with S-23031 (0.379 g/g) was comparable or slightly higher than the control value (Fig. 3C).
Peak Periods for PPIX Accumulation Wet weight of rat embryos treated with 400 mg/kg of S-53482 on GD 10 or GD 11 and weighed 14 hr later was comparable to the corresponding control values. A single exposure on GD 12, 13, 14, or 15 resulted in slightly lower wet weight of embryos when compared with controls obtained at the same time period (Fig. 4A). As shown in Figure 3A, treatment with 1000 mg/kg S-53482 on GD 12 does not affect embryonic wet weight measured 14 hr later. It is unlikely that these slightly lower wet weights in-
dicate adverse effects on growth 14 hr later. Wet weights of rabbit embryos from does treated with a single dose of 1000 mg/kg S-53482 on 1 day of gestation from GD 10 through GD 15 indicated no effect on fetal growth (Table 1). Hirasawa and Takeiri (1981) reported values for fetal rabbit weights of 53 and 392 mg on GD 12 and GD 15, respectively. Although these data are based on rabbits obtained from a different supplier, they are similar to values of 50 and 384 mg for rabbit embryos exposed to S-53482 on GD 11 and GD 14, and sampled on GD 12 and GD 15, respectively. There was an obvious difference in PPIX concentration between control and treated rat embryos. In controls, PPIX concentrations per gram tissue were low on each treatment day, ranging from 0.055 to 0.183 g/g. In contrast, PPIX concentrations in treated embryos ranged from 9 to 84 times as high as the control concentrations. In particular, on GD 11 or GD 12 the concentration of PPIX reached peak values in both relative (69× and 84×, respectively) and absolute terms (10.214 and 8.701 g/g, respectively; Fig. 4B). A large standard deviation on GD 11 or GD 12 was because two of five pools showed much higher values (13 19 g/g) than others (1 7 g/g). Therefore, the most sensitive period for PPIX accumulation in rat embryos followed dosing with S-53482 on GD 11 and GD 12. Little or no PPIX was detected in rabbit embryos treated on any day during GD 10 through GD 15 (Table 1). In rats, PPIX concentrations in control maternal livers ranged from 0.230 to 0.555 g/g, and in the treated maternal livers from 0.220 to 0.364 g/g (Fig. 4C). In rabbits, PPIX concentrations in maternal liver varied from 0.080 to 0.256 g/g (Table 1). These data indicate no remarkable increase in PPIX concentration in maternal livers of rats exposed to 400 mg/kg S-53482 or rabbits exposed to 1000 mg/kg S-53482.
DISCUSSION In previous studies, we conducted a series of investigations on developmental effects of S-53482. We found a remarkable species difference in the developmental toxicity of S-53482 between rats and rabbits, with the day of the greatest sensitivity to the herbicide being GD 12 in rats (Kawamura et al., 1995), and a compound-specific difference in developmental toxicity among N-phenylimide compounds structurally similar to S-53482 (Fig. 1) in rats (Kawamura et al., 2013). The major characteristics of developmental toxicity induced by S-53482 are species difference, critical period, and compound-specific difference. The mechanism of action of developmental toxicity by S-53482 should account for these characteristics. To elucidate the mechanism of developmental toxicity produced by S-53482, we have investigated a possible link between developmental toxicity and effects on PPO, which is the target enzyme for the herbicidal activity of S-53482 and common to plants and animals as part of chlorophyll and heme biosynthesis (Fig. 2). S-53482 inhibits PPO in plants, resulting in accumulation of PPIX that destroys plants by liberating singlet oxygen on irradiation with light. We have studied species differences in PPIX accumulation presumably resulting from PPO inhibition between rat and rabbit embryos. The species Birth Defects Research (Part B) 101:429–437, 2014
LINK BETWEEN PROTOPORPHYRIN ACCUMULATION AND TERATOGENICITY
Fig. 3. Compound-specific difference in PPIX accumulation in rats. Embryos were weighed and PPIX was extracted from rat embryos or maternal liver at 9:30 the next morning, which is 14 hr after treatment with 1000 mg/kg N-phenylimide herbicides at 19:30 on GD 12. Error bars show standard deviation. (A) Wet weight of embryo, (B) PPIX concentration in embryo, (C) PPIX concentration in maternal liver.
difference in PPIX accumulation corresponds very well to that of the developmental toxicity caused by S-53482. Remarkable PPIX accumulation was observed in rat embryos, but not in rabbit embryos (Kawamura et al., 1996a). Our objective in this study was to further investigate a possible link between PPIX accumulation as an indicator of PPO inhibition and developmental toxicity. The results of this study demonstrated that remarkable PPIX accumulations in rat embryos were caused by developmentally toxic compounds and the greatest accumulation of PPIX Birth Defects Research (Part B) 101:429–437, 2014
in embryos treated with S-53482 was observed when treated during sensitive period of developmental toxicity. Thus, not only species difference, but also compoundspecific differences and peak period of sensitivity in PPIX accumulation corresponded closely to those of developmental toxicity. This correlation demonstrated a close link between PPO inhibition and developmental abnormality. Accumulation of PPIX is considered to be indicative of PPO inhibition rather than a causative factor in teratogenicity. It might be assumed that developmental
KAWAMURA ET AL.
Fig. 4. Sensitive period for PPIX accumulation in rats. Embryos were weighed and PPIX was extracted from rat embryos or maternal liver at 9:30 the next morning, which is 14 hr after single treatments with 400 mg/kg S-53482 at 19:30 on different days from GD 10 through GD 15. Error bars show standard deviation. (A) Wet weight of embryo, (B) PPIX concentration in embryo, (C) PPIX concentration in maternal liver.
toxicity of S-53482 is mediated through the same mode of action as the herbicidal one or any effects of resulting PPIX on embryonic tissues, since PPIX accumulation corresponded to the developmental toxicity of S-53482. However, light would not sufficiently reach embryos through the maternal body wall to induce photodynamic action of accumulating PPIX in embryos. The target tissues for S-53482-induced toxicity are presumed to be erythroblasts as demonstrated by electron and light microscopy (Kawamura et al., 1996b). While PPIX levels
in erythroid cells in PPIX transporter null mice are 10 times higher than wild-type mice, there is no abnormality in reproduction and erythroid development of the null mice (Jonker et al., 2002). Although exposure of red blood cells to PPIX results in abnormal shape of the cells, decreased cell deformability, and swelling of the cells (Langerberg et al., 1996), no such changes were seen in the histopathological study of S-53482 (Kawamura et al., 1996b). PPIX is assumed to be an endogenous ligand to the peripheral benzodiazepine receptor on mitochondria
Birth Defects Research (Part B) 101:429–437, 2014
LINK BETWEEN PROTOPORPHYRIN ACCUMULATION AND TERATOGENICITY
Table 1 Wet Weight per Embryo, PPIX Concentrations in Embryo, and Maternal Liver in Rabbits at 14 hr after Single Oral Administrations of 1000 mg/kg S-53482 on Different Days from GD 10 through GD 15 GD of administration Specimen Rabbit embryo No. of embryo poolsa Wet weight per embryo (mg, mean ± SD) PPIX concentration (g/g tissue, mean ± SD) Rabbit maternal liver No. of specimens PPIX concentration (g/g tissue, mean ± SD)
3b 19.4 ± 1.7
4 50.2 ± 6.0
4 149.6 ± 17.3
3 0.080 ± 0.028
4 0.256 ± 0.166
4 0.175e ± 0.089
3 384.0 ± 78.2
4 641.9 ± 18.7
0.077 ± 0.019
3 0.137 ± 0.040
3 0.121f ± 0.027
4 0.109e ± 0.057
3 231.4 ± 3.8
pool consists of three embryos from a litter except for pools with notes. embryo pool consists of five embryos from a litter. c Not detected. d Value of one pool. PPIX was not detected in the other pools. e Mean value of three pools. PPIX was not detected in the other pool. f Mean value of two pools. PPIX was not detected in the other pool. b Each
since porpyrins displaced radiolabeled benzodiazepines and PPIX was most potent (Krishnamurthy et al., 2007). Nevertheless, many of the benzodiazepoines, such as diazepam, lorazepam, clonazepam, and oxazepam, failed to exhibit teratogenicity in rats (Schardein, 1993). Accumulation of PPIX would be caused by PPO inhibition, but not ferrochelatase inhibition, suppressed excretion of cellular PPIX, or excess biosynthesis of porphyrins. While protoporphyrinogen is the substrate for PPO and expected to accumulate in cells due to an inhibition of PPO, PPIX accumulation following treatment with diphenyl ether type herbicides, another type of PPO inhibiting herbicide, was observed in sensitive organisms, such as mice, rats (Krijt et al., 1992), cultured rat hepatocytes (Jacobs et al., 1992), and rat fetal liver at term (Machemer, 1992). Variegate porphyria is a genetic disease of PPO deficiency and patients with the defect accumulate high levels of PPIX (Kappas et al., 1995; Shaner, 2003). A common manner of PPIX accumulation resulting from PPO inhibition should be involved across organisms. This is a paradoxical situation that the product accumulates when an enzyme is inhibited, and probably because porphyrinogens undergo rapid oxidation under air to become porphyrins (Kappas et al., 1995). Matringe et al. (1989a) postulated that the accumulating protoporphyrinogen resulted from PPO inhibition eventually leaves the mitochondria, enters the cytoplasm, and is oxidized nonenzymatically there to PPIX. Because of the abnormal subcellular location, the resulting PPIX is beyond the reach of ferrochelatase and cannot be transferred to iron porphyrin. According to a model proposed by Ferreira et al. (1988) for the terminal three enzymes in heme synthesis in mouse liver mitochondria, PPO connects with ferrochelatase to channel PPIX in situ. Once PPIX is liberated into cytoplasm, it cannot reach the active site of ferrochelatase through the substrate channel. Recently, accumulation of protoporphyrinogen before PPIX in intact plants treated with PPO inhibitors was reported (Murata et al., 2004). Birth Defects Research (Part B) 101:429–437, 2014
Diphenyl ether herbicides inhibit PPO, but not ferrochelatase in corn, potato, mouse, and yeast (Matringe et al., 1989b). All PPO inhibitors act by competing with protoporphyrinogen (Hao et al., 2011). S-23121 binding to its target site in plastid was displaced by a diphenyl ether herbicide, indicating that S-23121 and diphenyl ether herbicides share the same binding site (Mito et al., 1991). In addition, in in vitro experiments S-53482 and S-23121 inhibited rat PPOs from GD 12 embryos at very low concentration and the inhibitory potency of S-23031 was much lower. N-phenylimide compounds inhibited rabbit PPOs much less than rat PPOs (in preparation), suggesting the compound- and species-specific differences in PPIX accumulation might be due to different levels of inhibitory response. Cell surface receptor for feline leukemia virus subtype C and ATP-binding cassette transporter ABCG2 are implicated in heme efflux in red blood cells (Schultz et al., 2010). While it remains unknown whether feline leukemia virus subtype C transports other porphyrins such as PPIX (Krishnamurthy et al., 2007), ABCG2 decreased the levels of endogenously produced PPIX in erythroid cells (Zhou et al., 2005). The levels of PPIX in erythroid cells in ABCG2 null mice increase 10 times higher than wild-type mice. ABCG2 null mice are, however, normal in reproduction and erythroid development (Jonker et al., 2002). Accordingly ABCG2 would not be a target of S-53482. ABCB6 is located in outer membrane of mitochondria and acts as a porphyrin importer from cytoplasm to mitochondria. Uptake of porphyrin into mitochondria activates porphyrin biosynthesis (Krishnamurthy et al., 2006). Initial histologic changes, such as abnormal iron deposits, however, suggest suppression of heme synthesis, indicating excess biosynthesis of porphyrin is not the case. Thus, PPO inhibition is more likely to be the primary cause of PPIX accumulation by S-53482 than ferrocheratase inhibition, suppressed excretion of cellular PPIX, or excess biosynthesis of porphyrins. Porphyrins carry anionic carboxylate side chains that introduce negative charges to the molecule, limiting their
KAWAMURA ET AL.
Fig. 5. Brief outline of postulated mode of action for developmental toxicity induced by S-53482.
ability to diffuse across membranes, and ABCG2, which exports PPIX, is expressed in the plasma membrane of placental syncytiotrophoblast (Krishnamurthy et al., 2007). Porphyrins could be transported from the placenta into maternal circulation. Therefore, placental transfer of PPIX might be limited. Placental transfer of S-53482 was examined in rats and rabbits on GD 12 and there was only a slight difference between the species (unpublished data). Lower concentrations of PPIX accumulated in maternal liver than those of embryos. This is probably because of rapid excretion of PPIX via bile, since PPIX is poorly water soluble due to its lipophilic nature and excreted only into lipid-rich bile (Kappas et al., 1995). When male mice were treated with oxadiazon, a diphenyl ether herbicide, at 1000 ppm in the diet for 6 days, PPIX accumulation in bile and feces were elevated 50- and 40-fold, respectively, whereas accumulation in liver was double that of controls (Krijt et al., 1992). In previous studies, we conducted histopathological studies to elucidate mechanism of developmental toxicity caused by S-53482 (Kawamura et al., 1996b). As initial histologic changes observed in rat embryos treated with S-53482, electron, and light microscopy demonstrated mitochondrial lesions including abnormal iron deposits in polychromatophilic erythroblasts, which are derived from yolk sac and may actively synthesize hemoglobin (Kawamura et al., 1996b). An abnormal accumulation of iron granules in mitochondria is found in erythroblasts in sideroblastic anemia together with decreased activities of enzymes in porphyrin biosynthesis (Konopka and Hoffbrand, 1979; Aoki, 1980). Iron destined for incorporation into porphyrin to produce heme would accumulate by deficient porphyrin synthesis in mitochondria in erythroblasts (Ghadially, 1997). Following the mitochondrial lesions, degenerated erythroblasts in the embryonic circulation and thinning of the ventricular walls of the heart are observed. These might indicate dilatation of the ventricles as a compensatory reaction to embryonic anemia (Kawamura et al., 1996b). A characteristic of yolk sac hemopoiesis is that erythroid cells synchronously undergo maturation as a relatively homogeneous population (Tavassoli, 1991). It is likely that most erythroid cells are lost synchronously and severe anemia may occur in rat embryos exposed to S-53482. Then, we have postulated the mechanism for VSD as follows: S-53482 inhibits PPO in rat embryos, thereby suppressing normal heme biosynthesis and producing erythroblastic degeneration, resulting in embryonic anemia. Increased stroke volume to compensate for anemia leads to enlargement of the heart and VSD is produced by mechanical distortion of the heart or abnormal cardiac hemodynamics (Kawamura
et al., 1996b). General toxicity studies of S-53482 demonstrate that anemia is the primary toxicity by S-53482 (unpublished data). The results of this study demonstrated a close link between rat developmental toxicity and PPO inhibition, and thus, strongly support our postulated mechanism of S-53482-induced teratogeniciy. Based on all available experimental data, we roughly postulate the mechanism of action of S-53482 developmental toxicity as shown in Figure 5. We will discuss it in more detail by conducting further investigation on fetal morphologic and hematologic changes during mid- and late-gestation, and PPO inhibitions by N-phenylimide herbicides.
ACKNOWLEDGMENTS The authors thank Dr. Nobuaki Mito, Sumitomo Chemical for his technical advice and the staff of Sumitomo’s Developmental and Reproductive Biology Team for their excellent assistance.
CONFLICTS OF INTEREST The authors have no conflicts of interest to declare.
REFERENCES Aoki Y. 1980. Multiple enzymatic defects in mitochondria in hematological cells of patients with primary sideroblastic anemia. J Clin Invest 66:43–49. Ferreira GC, Andrew TL, Karr SW, Dailey HA. 1988. Organization of the terminal two enzymes of the heme biosynthetic pathway. J Biol Chem 263:3835–3839. Ghadially FN. 1997. Intramitochondrial iron inclusions. In: Ghadially FN. Ultrastructural pathology of the cell and matrix. 4th ed. Vol. I. Boston, MA: Butterworth-Heinemann. p 324–327. Hao GF, Zuo Y, Yang SG, Yang GF. 2011. Protoporphyrinogen oxidase inhibitor: an ideal target for herbicide discovery. Chimia 65:961–969. Hirasawa K, Takeiri S. 1981. Introduction of normal data for SPF rabbit (KBL:JW). J Exp Anim Technol 16:103–109. (Japanese.) Jacobs JM, Sinclair PR, Gorman N, Jacobs NJ, Sinclair JF, Bement WJ, Walton H. 1992. Effects of diphenyl ether herbicides on porphyrin accumulation by cultured hepatocytes. J Biochem Toxicol 7:87–95. Jonker JW, Buitelaar M, Wagenaar E, van der Valk MA, Scheffer GL, Scheper RL, Plosch T, Kuipers F, Oude Elferink RPJ, Rosing H, Beijnen JH, Schinkel AH. 2002. The breast cancer resistance protein protects against a major chlorophyll-derived dietary phototoxin and protoporphyria. PNAS 99:15649–15654. Kappas A, Sassa S, Galbraith RA, Nordmann Y. 1995. The porphyrias. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic and molecular bases of inherited disease. 7th ed. New York: McGraw-Hill, Inc., p 2103–2159. Kawamura S, Kato T, Matsuo M, Sasaki M, Katsuda Y, Hoberman AM, Yasuda M. 1995. Species difference in developmental toxicity of an Nphenylimide herbicide between rats and rabbits and sensitive period of the toxicity to rat embryos. Cong Anom 35:123–132. Kawamura S, Kato T, Matsuo M, Katsuda Y, Yasuda M. 1996a. Species difference in protoporphyrin IX accumulation produced by an
Birth Defects Research (Part B) 101:429–437, 2014
LINK BETWEEN PROTOPORPHYRIN ACCUMULATION AND TERATOGENICITY N-phenylimide herbicide in embryos between rats and rabbits. Toxicol Appl Pharm 141:520–525. Kawamura S, Yoshioka T, Kato T, Matsuo M, Yasuda. 1996b. Histological changes in rat embryonic blood cells as a possible mechanism for ventricular septal defects produced by an N-phenylimide herbicide. Teratology 54:237–244. Kawamura S, Kato T, Kannan N, Fantel AG. 2013. Difference in developmental toxicity among structurally similar N-phenylimide herbicides in rats and rabbits. Birth Defects Res B 98:437–444. Kawamura S, Kato T, Nakaoka M, Fantel AG. 2014. Dermal developmental toxicity of N-phenylimide herbicides in rats. Birth Defects Res B 101:162–167. Konopka L, Hoffbrand AV. 1979. Haem synthesis in sideroblastic anaemia. Br J Haematol 42:73–83. Krijt J, Pleskot R, Santitrak J, Janousek V. 1992. Experimental hepatic porphyria induced by oxadiazon in male mice and rats. Pestic Biochem Physiol 42:180–187. Krishnamurthy PC, Du G, Fukuda Y, Sun D, Sampath J, Mercer KE, Wang J, Sosa-Pineda B, Murti KG, Shuetz JD. 2006. Identification of a mammalian mitochondrial porphyrin transporter. Nature 443:586–589. Krishnamurthy PC, Xie T, Shuetz JD. 2007. The role of transporters in cellular heme and porphyrin homeostasis. Pharmacol Ther 114:345– 358. Langerberg JWM, Williams M, Moor ACE, Brand A, van der Zee J, Dubbelman TMAR, VanSteveninck J. 1996. The influence of merocyanine 540 and protoporphyrin on physicochemical properties of the erythrocyte membrane. Biochim Biophys Acta 1278:247–253. Machemer L. 1992. Specific and non-specific developmental effects. In:
Birth Defects Research (Part B) 101:429–437, 2014
Neubert D, Kavlock RJ, Merker HJ, Klein J, editors. Risk assessment of prenatally-induced adverse health effects. Berlin: Springer-Verlag. p P85–100. Matringe M, Camadro JM, Labbe P, Scalla R. 1989a. Protoporphyrinogen oxidase inhibition by three peroxidizing herbicides: oxadiazon, LS82– 556 and M & B 39279. FEBS Lett 245:35–38. Matringe M, Camadro JM, Labbe P, Scalla R. 1989b. Protoporphyrinogen oxidase as a molecular target for diphenyl ether herbicides. Biochem J 260:231–235. Mito N, Sato R, Miyakado M, Oshio H, Tanaka S. 1991. In vitro mode of action of N-phenylimide photobleaching herbicides. Pestic Biochem Physiol 40:128–135. Murata S, Kimura Y, Mabuchi T, Miura Y. 2004. Accumulation of protoporphyrinogen IX prior to protoporphyrin IX in intact plants treated with protoporphyrinogen IX oxidase-inhibiting herbicides. J Pestic Sci 29:195–199. Schardein JL. 1993. Chemically induced birth defects. 2nd ed. New York, NY: Dekker. Schultz IJ, Chen C, Paw BH, Hamza I. 2010. Iron and porphyrin trafficking in heme biogenesis. J Biol Chem 285:26753–26759. Shaner DL. 2003. Herbicide safety relative to common targets in plants and mammals. Pest Manag Sci 60:17–24. Tavassoli M. 1991. Embryonic and fetal hematopoiesis: an overview. Blood Cells 1:269–281. Zhou S, Zong Y, Ney PA, Nair G, Stewart CF, Sorrentino BP. 2005. Increased expression of the Abcg2 transporter during erythroid maturation plays a role in decreasing cellular protoporphyrin IX levels. Blood 105:2571–2576.