Tetramethylcyclopropyl analogue of the leading antiepileptic drug, valproic acid: Evaluation of the teratogenic effects of its amide derivatives in NMRI mice†
Conflict of Interest: Meir Bialer received speaking or consultancy fees from The American Epilepsy Society (AES), BIAL, Bioline, Desitin, Gerson Lehrman Group Councils, Janssen-Cliag, Jazz Pharmaceuticals, Johnson and Johnson, NovoNordisk, Ovation, NeuroAdjuvants, Neurocrine Biosciences, Shire, Teva, UCB Pharma, and Valeant. In the last 3 years, the author received research grants from Jazz Pharmaceuticals (together with Boris Yagen), Johnson and Johnson, Teva, and The Epilepsy Therapy Development Project (together with Boris Yagen), and has been involved in the design and development of new antiepileptics and CNS drugs as well as new formulations of existing drugs. All other authors declare of no conflict of interest.
Abstract
BACKGROUND: Although valproic acid (VPA) is used extensively for treating various kinds of epilepsy, it causes hepatotoxicity and teratogenicity. In an attempt to develop a more potent and safer second generation to VPA drug, the amide derivatives of the tetramethylcyclopropyl VPA analogue, 2,2,3,3-tetramethylcyclopropanecarboxamide (TMCD), N-methyl-TMCD (MTMCD), 4-(2,2,3,3-tetramethylcyclopropanecarboxamide)-benzenesulfonamide (TMCD-benzenesulfonamide), and 5-(TMCD)-1,3,4-thiadiazole-2-sulfonamide (TMCD-thiadiazolesulfonamide) were synthesized and shown to have more potent anticonvulsant activity than VPA. Teratogenic effects of these CNS-active compounds were evaluated in Naval Medical Research Institute (NMRI) mice susceptible to VPA-induced teratogenicity by comparing them to those of VPA. METHODS: Pregnant NMRI mice were given a single sc injection of either VPA or TMC-amide derivatives on gestation day 8.5, and then the live fetuses were examined to detect any external malformations on gestation day 18. After double-staining for bone and cartilage, their skeletons were examined. RESULTS: In contrast to VPA, which induced NTDs in a high number of fetuses at 2.4–4.8 mmol/kg, TMCD, TMCD-benzenesulfonamide, and TMCD-thiadiazolesulfonamide at 4.8 mmol/kg and MTMCD at 3.6 mmol/kg did not induce a significant number of NTDs. TMCD-thiadiazolesulfonamide exhibited a potential to induce limb defects in fetuses. Skeletal examination also revealed that fetuses exposed to all four of the tetramethylcyclopropanecarboxamide derivatives developed vertebral and rib abnormalities less frequently than those exposed to VPA. Our results established that TMCD, MTMCD, and TMCD-benzenesulfonamide are distinctly less teratogenic than VPA in NMRI mice. CONCLUSIONS: The CNS-active amides containing a tetramethylcyclopropanecarbonyl moiety demonstrated better anticonvulsant potency compared to VPA and a lack of teratogenicity, which makes these compounds good second-generation VPA antiepileptic drug candidates. Birth Defects Research (Part A), 2008. © 2008 Wiley-Liss, Inc.
INTRODUCTION
Antiepileptic drugs (AEDs) are well known for their teratogenicity in patients as well as in animal models and the laboratory setting (Dansky and Finnell, 1991; Finnell and Dansky, 1991). Valproic acid (2-propylpentanoic acid; VPA; Fig. 1) is an established AED, and, due to its wide spectrum of antiepileptic activity, is one of the most prescribed (Levy et al., 2002). VPA is also an effective and approved drug for migraine prophylaxis and the treatment of bipolar disorder (Levy et al., 2002). Despite its extensive clinical effectiveness and utilization, the therapeutic use of VPA is restricted due to two major side effects: teratogenicity and hepatotoxicity (Levy et al., 2002).
Chemical structures of VPA, VPD, and four amide derivatives of the VPA analogue TMCA: TMCD, MTMCD, TMCD-benzenesulfonamide, and TMCD-thiadiazolesulfonamide.
NTD is the most serious, life-threatening malformation caused by VPA therapy (Lindhout and Schmidt, 1986). Therefore, there is a substantial need to develop new potent nonteratogenic second-generation to VPA drugs that can be given without restriction to women of childbearing age (Nau et al., 1991; Isoherranen et al., 2003; Okada and Fujiwara, 2006; Bialer and Yagen, 2007). The teratogenic activity of VPA has been established in mice; a single injection with VPA on day 8 of gestation caused a high percentage of exencephaly, which is the murine anterior NTD equivalent of human anencephaly (Kao et al., 1981; Menegola et al., 1996). Because other animal species, such as rats and rabbits, are not susceptible to VPA-induced NTDs (Petrere et al., 1986; Menegola et al., 1996), the mice model system is used for the in vivo evaluation of AED-induced NTD potency. Another commonly observed teratogenic effect caused by VPA is skeletal abnormality, which has been detected in many species, including humans, rhesus monkeys, mice, rats, and rabbits (Kao et al., 1981; Ong et al., 1983; Petrere et al., 1986; Vorhees, 1987; Binkerd et al., 1988; Hendrickx et al., 1988). In rodents, the region of the axial skeleton affected by VPA treatment depends on when in the developmental stage the rodent is exposed to VPA (Collins et al., 1991; Menegola et al., 1998).
Structure-teratogenicity relationship studies conducted in mouse strains susceptible to VPA-induced teratogenicity indicated that, in order to be teratogenic, a VPA derivative or analogue had to possess the following three structural requirements: (1) an underivatized carboxylic group, (2) a hydrogen atom at C-2 (the α-position to the carboxylic group), and (3) branching at C-2 (Nau et al., 1991; Nau and Siemens, 1992; Bojic et al., 1996, 1998). Thus, a VPA derivative or analogue that does not possess one of these structural requirements has the potential to become a nonteratogenic AED.
In our ongoing attempt to develop second-generation drugs to VPA, several amide derivatives of tetramethylcyclopropanecarboxylic acid (TMCA; Fig. 1) have been developed. This acid is a cyclopropyl analogue of VPA with two quaternary carbons at the β-position to the carboxylic group; therefore, it cannot be biotransformed into a hepatotoxic metabolite with a terminal double bond analogous to 4-ene-VPA or 2,4-diene-VPA (Sussman and McLean, 1979; Baillie, 1988; Isoherranen et al., 2003; Bialer, 2006). However, TMCA showed low anticonvulsant activity, and no separation of its anticonvulsant activity and neurotoxicity was apparent in rodents (Table 1; Winkler et al., 2005). In contrast to TMCA, its amide derivatives were found to be CNS-active compounds with a broad spectrum of antiepileptic activity in various animal models (Isoherranen et al., 2002; Sobol et al., 2004). Unlike VPA, 2,2,3,3-tetramethylcyclopropanecarboxamide (TMCD) and N-methyl-TMCD (MTMCD) do not possess a free carboxylic group, and thus have the potential to become nonteratogenic AEDs. Consequently, their ability to induce NTDs or skeletal abnormalities in fetuses is of special interest. In addition, they cannot be biotransformed into unsaturated hepatotoxic metabolites analogous to VPA (Isoherranen et al., 2002, 2003; Sobol et al., 2004). Thus, TMCD and MTMCD have the potential to become second-generation to VPA antiepileptics and CNS drugs.
| Compound | MES* (ED50, mg/kg) | scMet† (ED50, mg/kg) | Neurotox‡ (TD50, mg/kg) | PI§ (MES) | PI§ (scMet) |
|---|---|---|---|---|---|
| VPA∥ | 485 (324–677) | 646 (466–869) | 784 (503–1,176) | 1.6 | 1.2 |
| TMCA¶ | >200 | >150 | 181 (102–294) | – | – |
| TMCD# | >250 | 52 (42–63) | 381 (355–418) | – | 7.3 |
| MTMCD# | 82 (64–102) | 45 (31–55) | 163 (138–179) | 2 | 3.6 |
| TMCD-benzenesulfonamide** | 26 (14–42) | >100 | >500 | >19 | – |
| TMCD-thiadiazolesulfonamide | 9 (5–14) | >250 | >500 | >50 | – |
Zonisamide and acetazolamide are AEDs in current use that contain a sulfonamide group in their structure (Restor et al., 1995; Shah et al., 2002). An extensive structure-activity relationship of various aromatic and heterocyclic sulfonamide derivatives of VPA demonstrated the influence of the sulfonamide group on the anticonvulsant activity of these VPA derivatives (Masereel et al., 2002). Recently, TMCD-benzenesulfonamide was synthesized. After its anticonvulsant activity was evaluated, it was found to have a wide safety margin and to be more potent than VPA (Table 1; Shimshoni et al., 2008). The TMCD heterocyclic sulfonamide derivative 5-(2,2,3,3-tetramethylcyclopropanecarboxamide)-1,3,4-thiadiazole-2-sulfonamide demonstrated potent anticonvulsant activity and a protective index (PI = TD50/ED50) above 50 in the maximal electroshock seizure model (Table 1).
The primary objectives of this teratology study were to investigate the potential of the four CNS-active TMCA amide derivatives, TMCD, MTMCD, TMCD-benzenesulfonamide and TMCD-thiadiazolesulfonamide, to induce neural tube and skeletal abnormalities in the well-established Naval Medical Research Institute (NMRI)-mouse model for VPA-induced teratogenicity, and to compare them to VPA (Nau, 1986; Okada et al., 2004). The results of this structure-teratogenicity relationship study could lead the development of better and safer CNS-active drugs.
MATERIALS AND METHODS
Chemicals
TMCA and sodium valproate were purchased from Sigma (St. Louis, MO). TMCD, MTMCD, and TMCD-benzenesulfonamide were synthesized as previously described (Isoherranen et al., 2002; Shimshoni et al., 2008). TMCD-thiadiazolesulfonamide was synthesized as follows: 22.2 g acetazolamide was dissolved in 100 mL concentrated HCl solution (35%) and refluxed on a water bath for 2 h. The solvent was evaporated, and the white precipitate was dissolved in 50 mL of water and neutralized with NaHCO3 until pH 7 was reached. The product (90% yield) was recrystallized from methanol and identified as 5-amino-1,3,4-thiadiazole-2-sulfonamide by TLC, NMR, and elemental analysis. TMCA was converted to its acyl chloride by SOCl2, to yield TMC-Cl (Sobol et al., 2004). TMC-Cl (0.055 mol) dissolved in 87 mL of dry dichloromethane was added to 5-amino-1,3,4-thiadiazole-2-sulfonamide (0.055 mol) and pyridine (0.11 mol) at 0°C. The reaction mixture was kept at 4°C for 16 h. The solvent was evaporated under vacuum, and 50 mL of dichloromethane was added and evaporated again. The residue was dissolved in ethyl acetate, washed with 2 N HCl and water, dried over magnesium sulfate, and evaporated. The product was recrystallized from a mixture of ethanol and water, followed by a second recrystallization from cold ethanol (40% yield, m.p. = 237°C). Its structure was identified using TLC, NMR, and elemental analysis. The tested compounds were crystalline and pure according to elemental microanalyses. For the teratology study, the tested compounds were dissolved or suspended in a vehicle solvent containing 10% DMSO (Sigma) and 90% distilled water for injection.
Animals
Animals were maintained and treated in accordance with the Institutional Guides for the Care and Use of Laboratory Animals. The NMRI mouse strain, which is known to be susceptible to AED-induced NTDs (Nau, 1986; Okada et al., 2004), was used in this study, in accordance with the previously published protocol (Okada et al., 2004). Male and female NMRI strain mice (8 weeks of age) were purchased from Charles River Laboratories France and Germany (via Charles River Japan, Inc., Kanagawa, Japan) and housed in plastic cages under controlled conditions (temperature: 23 ± 2°C, relative humidity: 55 ± 10%) with a 12 h light/12 h dark cycle (07:00–19:00). Pellet feed (CRF-1; Oriental Yeast Co., Ltd., Tokyo, Japan) and municipal tap water were freely available. Females were cohabited with males of the same strain overnight, and subsequently examined for the presence of vaginal plugs. When a vaginal plug was present, the onset of gestation for each mouse was designated as 10 p.m. the previous night (GD 0).
Animal Treatment
The teratology study was conducted in two separate experiments (Experiments 1 and 2). In each experiment, pregnant NMRI mice were randomly assigned to either treatment group (8–10 mice per group) and injected subcutaneously once with solvent (control), VPA, or one of the TMCA derivatives on GD 8.5 (10 a.m.) at a dosing volume of 10 mL/kg of body weight. In Experiment 1, pregnant mice received MTMCD (2.4 and 4.8 mmol/kg) or TMCD-thiadiazolesulfonamide (2.4 and 4.8 mmol/kg). In Experiment 2, pregnant mice received TMCD (1.8, 3.6, and 4.8 mmol/kg), MTMCD (1.8 and 3.6 mmol/kg), or TMCD-benzenesulfonamide (1.8, 3.6, and 4.8 mmol/kg). Both experiments included the vehicle control group and the corresponding VPA group in the identical molar dosage. Because fetuses with NTDs were found at 4.8 mmol/kg MTMCD in Experiment 1, dams were injected at other dose levels in Experiment 2 in order to determine the maximum no-observed-adverse-effect-level of MTMCD with regard to NTD induction. The VPA dose levels used in this study have been known to be teratogenic in mice, and a dose level of around 3.6 or 4.8 mmol/kg was the maximum tolerated dose level (Kao et al., 1981; Menegola et al., 1996, 1998; Okada et al., 2004, 2006). In a preliminary study using nonpregnant mice, none of the animals died after dosing at 4.8 mmol/kg TMCD, MTMCD, TMCD-benzenesulfonamide, or TMCD-thiadiazolesulfonamide. Thus, 4.8 mmol/kg was selected as the maximum dose level for the investigation of the teratogenicity of the new compounds in NMRI mice and its comparison to that of VPA.
Teratology Studies
After chemical injection, the dams were monitored to examine clinical signs and body weight changes. On GD 18, the treated dams were sacrificed by exsanguination from the abdominal aorta under ether anesthesia. The uteri and ovaries were excised, and the numbers of corpora lutea, implantations, live fetuses, and fetal losses were recorded. The live fetuses were weighed and examined for external macroscopic abnormalities, and then double-stained with alizarin red S (Kanto Chemical Co., Inc., Tokyo, Japan) and alcian blue 8 GX (Sigma) for bone and cartilage, respectively, using the double staining method proposed by Kimmel and Trammell (1981) with some modifications (Okada et al., 2004). After staining, the axial skeletons (vertebra and ribs) of the double stained fetuses with no external malformations were examined under a dissecting microscope. However, results from fetuses exposed to 4.8 mmol/kg (Experiment 1) and 3.6 and 4.8 mmol/kg VPA (Experiment 2) were not included because the malformations caused by these high dose levels of VPA were too severe to be categorized and compared with those of the derivatives. The skeletons of fetuses at 1.8 and 3.6 mmol/kg TMCD, 1.8 mmol/kg MTMCD, and 1.8 and 3.6 mmol/kg TMCD-benzenesulfonamide were also not included because, compared with the control findings, no teratogenic changes were noted, even at each compound's next higher dose level. According to the characteristics of the control fetuses, the axial segments comprised seven cervical, thirteen thoracic, six lumbar, and four sacral vertebrae. Each had 13 ribs attached to the thoracic vertebrae. The 6th cervical (C6) and 2nd thoracic (T2) vertebrae had ventral processes and a prominent spinous process, respectively. The 1st to 7th cartilaginous ribs (vertebrosternal ribs or true ribs) met the sternum to form six sternebrae. Supernumerary ribs (SNRs) at the first lumbar position were categorized as extra or rudimentary ribs according to their length, which was determined by visual approximation (more than or less than 1/3 the length of the 13th rib, respectively).
Statistical Analysis
Differences in the number of corpora lutea, implantations, and live fetuses, as well as body weight gain and fetal body weights between the control group and each dose group were analyzed using Dunnett's multiple comparison test. The differences in fetal loss, external abnormalities, and skeletal abnormalities between the control group and each treated group were analyzed using the Wilcoxon rank sum test (Mann-Whitney's U-test). For fetal data, the litter served as the unit of statistical analysis, and the significance level was set at p < .05 or p < .01 (two-sided test).
RESULTS
Effects on Dams
Pregnant NMRI mice were monitored for their survival, general condition, and body weight change after an sc injection of VPA or one of the TMCA amide derivatives on GD 8.5. Table 2 lists the findings for dams, which included decreased locomotor activity after injection of VPA, even at the lowest dose level of 1.8 mmol/kg, on GD 8. Dams receiving 3.6 and 4.8 mmol/kg VPA exhibited severe sedation and a prone or lateral position. One out of eight dams at 3.6 mmol/kg (Experiment 2), and one out of seven dams or three out of seven dams at 4.8 mmol/kg (Experiments 1 and 2, respectively) died sometime between immediately after dosing and the next morning. The changes in the surviving animals were still observable 8 h after treatment, but they recovered by the next morning (GD 9). VPA at dose level of 4.8 mmol/kg induced a body weight decrease in dams, which was already apparent by the day after dosing.
| Compound | Dose (mmol/kg) | No. of dams | Dead dams | Dams with DLA | Dams with PP/LP | Body weight gain: GD 8 to 9 | Body weight gain: GD 8 to 18 | Experiment no. |
|---|---|---|---|---|---|---|---|---|
| Control 1 | 8 | 1.06 ± 0.25 | 24.3 ± 7.3 | 1 | ||||
| Control 2 | 8 | 1.31 ± 1.26 | 27.6 ± 2.6 | 2 | ||||
| VPA | 1.8 | 8 | 5 | 0.12 ± 1.40 | 24.1 ± 3.0 | 2 | ||
| 2.4 | 7 | 3 | 0.00 ± 1.03* | 23.4 ± 3.7 | 1 | |||
| 3.6 | 8 | 1 | 8 | 2 | 0.40 ± 1.05 | 18.8 ± 5.1## | 2 | |
| 4.8 | 7 | 1 | 7 | 5 | −1.12 ± 0.93** | 18.7 ± 5.2 | 1 | |
| 4.8 | 7 | 3 | 7 | 7 | −1.80 ± 2.75## | 5.9 ± 5.4## | 2 | |
| TMCD | 1.8 | 6 | 3 | 0.65 ± 1.09 | 26.9 ± 4.6 | 2 | ||
| 3.6 | 7 | 7 | 1 | 0.43 ± 0.89 | 21.3 ± 2.8## | 2 | ||
| 4.8 | 7 | 1 | 6 | 3 | −0.30 ± 1.19# | 24.0 ± 2.0 | 2 | |
| MTMCD | 1.8 | 7 | 0.57 ± 0.67 | 23.9 ± 2.9 | 2 | |||
| 2.4 | 7 | 3 | 0.23 ± 0.91 | 27.3 ± 3.0 | 1 | |||
| 3.6 | 8 | 4 | 0.80 ± 2.06 | 23.5 ± 3.8# | 2 | |||
| 4.8 | 10 | 1 | 6 | 1 | −1.33 ± 1.88** | 21.8 ± 7.9 | 1 | |
| TMCD-benzenesulfonamide | 1.8 | 7 | 0.51 ± 0.65 | 28.4 ± 4.0 | 2 | |||
| 3.6 | 8 | 0.30 ± 1.14 | 25.0 ± 4.1 | 2 | ||||
| 4.8 | 7 | 0.56 ± 0.53 | 23.8 ± 6.7 | 2 | ||||
| TMCD-thiadiazolesulfonamide | 2.4 | 10 | 10 | 4 | −2.13 ± 0.91** | 18.5 ± 6.2 | 1 | |
| 4.8 | 8 | 8 | 6 | −1.44 ± 1.11** | 16.6 ± 7.1 | 1 |
- DLA, decreased locomotor activity; PP/LP, prone position or lateral position.
- Mean ± S.D.
- ** p < .01.
- * p < .05 vs. Control 1.
- ## p < .01.
- # p < .05 vs. Control 2.
Abnormal activity and behavior similar to that noted with VPA were reproduced in pregnant NMRI mice after sc injection of TMCD, MTMCD, and TMCD-thiadiazolesulfonamide. TMCD induced a decrease in locomotor activity and a prone position in dams treated with more than 1.8 mmol/kg, and one dam died at 4.8 mmol/kg. The body weight of dams decreased the day after treatment with TMCD at 4.8 mmol/kg. The dams treated with MTMCD at above 2.4 mmol/kg exhibited a decrease in locomotor activity, and, at 4.8 mmol/kg, a body weight decrease and a death occurred. MTMCD at a dose level of 1.8 mmol/kg had no effect on general condition or body weight in pregnant NMRI mice. TMCD-thiadiazolesulfonamide caused a decrease in locomotor activity in all dams treated at 2.4 and 4.8 mmol/kg. In addition, a prone or lateral position was exhibited in 4 out of 10 and six out of eight dams at 2.4 and 4.8 mmol/kg, respectively, on GD 8. Dam body weight decreased the day after treatment at both 2.4 and 4.8 mmol/kg. In contrast, no abnormal clinical signs or body weight changes were noted in pregnant mice treated with TMCD-benzenesulfonamide at up to 4.8 mmol/kg.
Overall, the toxicological profiles of TMCD, MTMCD, and TMCD-thiadiazolesulfonamide as they affect pregnant NMRI mice were similar to that of VPA, which induced death as well as decreases in locomotor activity and body weight in dams. However at the same dose level, the toxic effects of TMCD and MTMCD were relatively less severe than those of VPA. Meanwhile, TMCD-thiadiazolesulfonamide at 2.4 mmol/kg was a much more potent inducer of decreases in dam activity and body weight than VPA; however, no deaths occurred even at 4.8 mmol/kg.
Effects on Embryo Lethality and Growth
It is well known that VPA shows a high embryo lethality and growth retardation effect upon cesarean section at GD 18 (Table 3). Treatment with VPA at 3.6 and 4.8 mmol/kg induced a significant increase in the rate of fetal loss (p < .01) and a significant decrease in fetal weight (p < .01) compared with those in the corresponding control group. No effects on these parameters were noted at 1.8 or 2.4 mmol/kg VPA.
| Compound | Dose (mmol/kg) | No. of dams | No. of implants$ | No. of fetal loss (%)$ | Fetal weight† | No. of fetuses with NTD (%)$ | No. of fetuses with limb defects (%)$ | Experiment no. |
|---|---|---|---|---|---|---|---|---|
| Control 1 | 8 | 109 (13.6 ± 5.2) | 6 (4.8 ± 5.7) | 1.43 ± 0.14 | 1 | |||
| Control 2 | 8 | 116 (14.5 ± 1.6) | 4 (3.3 ± 3.5) | 1.42 ± 0.11 | 2 | |||
| VPA | 1.8 | 8 | 107 (13.4 ± 2.5) | 13 (11.5 ± 19.0) | 1.41 ± 0.14 | 2 | ||
| 2.4 | 7 | 95 (13.6 ± 3.3) | 12 (11.4 ± 11.1) | 1.31 ± 0.12 | 8 (11.0 ± 16.9*) | 1 | ||
| 3.6 | 7 | 88 (12.6 ± 2.0) | 21 (25.7 ± 22.2##) | 1.16 ± 0.12## | 11 (22.1 ± 31.8#) | 2 | ||
| 4.8 | 6 | 91 (15.2 ± 1.6) | 42 (46.0 ± 20.5**) | 1.18 ± 0.07** | 12 (19.3 ± 19.9**) | 1 | ||
| 4.8 | 4 | 62 (15.5 ± 2.5) | 57 (92.4 ± 11.9##) | 0.88 ± 0.28## | 1 (50.0 ± 70.7#) | 2 | ||
| TMCD | 1.8 | 6 | 85 (14.2 ± 2.2) | 6 (7.2 ± 5.0) | 1.43 ± 0.09 | 2 | ||
| 3.6 | 7 | 83 (11.9 ± 2.3) | 10 (11.2 ± 9.7) | 1.38 ± 0.08 | 2 | |||
| 4.8 | 6 | 89 (14.8 ± 1.5) | 8 (8.8 ± 8.2) | 1.38 ± 0.13 | 1 (1.2 ± 2.9) | 2 | ||
| MTMCD | 1.8 | 7 | 93 (13.3 ± 1.5) | 5 (5.1 ± 5.2) | 1.44 ± 0.07 | 2 | ||
| 2.4 | 7 | 107 (15.3 ± 1.4) | 9 (8.5 ± 3.4) | 1.47 ± 0.07 | 1 (1.0 ± 2.5) | 1 | ||
| 3.6 | 8 | 108 (13.5 ± 1.7) | 14 (12.9 ± 11.8#) | 1.34 ± 0.12 | 2 | |||
| 4.8 | 9 | 121 (13.4 ± 1.4) | 24 (20.5 ± 26.1) | 1.36 ± 0.26 | 12 (21.3 ± 37.1*) | 1 | ||
| TMCD-benzenesulfonamide | 1.8 | 7 | 106 (15.1 ± 2.0) | 6 (5.4 ± 7.7) | 1.45 ± 0.08 | 2 | ||
| 3.6 | 8 | 118 (14.8 ± 2.7) | 11 (6.1 ± 10.1) | 1.38 ± 0.10 | 2 | |||
| 4.8 | 7 | 90 (12.9 ± 4.8) | 7 (10.4 ± 11.4) | 1.42 ± 0.08 | 2 | |||
| TMCD-thiadiazolesulfonamide | 2.4 | 10 | 153 (15.3 ± 2.1) | 21 (13.6 ± 9.4) | 1.21 ± 0.16** | 2 (1.7 ± 3.6) | 30 (23.4 ± 21.8**) | 1 |
| 4.8 | 8 | 111 (13.9 ± 2.9) | 20 (18.5 ± 33.1) | 1.22 ± 0.11** | 24 (27.5 ± 22.8**) | 1 |
- ** p < .01.
- * p < .05 vs. Control 1.
- ## p < .01.
- # p < .05 vs. Control 2.
- $ Total number (mean ± S.D.).
- † Mean ± S.D.
The embryo lethality and growth retardation effects observed upon cesarean section on GD 18 in dams treated with each derivative compound are compared to those of VPA in Table 3. Among the tested TMCA derivatives, MTMCD and TMCD-thiadiazolesulfonamide induced embryo lethality and/or growth retardation in NMRI fetuses. Treatment with MTMCD at 3.6 mmol/kg (p < .05) and 4.8 mmol/kg (no statistical significance) induced an increase in the fetal loss rate. However, MTMCD had no effect on fetal survival at 1.8 and 2.4 mmol/kg and no effect on fetal weight up to 4.8 mmol/kg. TMCD-thiadiazolesulfonamide at 2.4 and 4.8 mmol/kg induced a significant decrease in fetal weight compared with that in the control group (p < .01, respectively). The rate of fetal loss was higher at both dose levels of TMCD-thiadiazolesulfonamide than that in the control group, although the difference was not statistically significant. TMCD and TMCD-benzenesulfonamide had no adverse effects on fetal survival or fetal growth at doses up to 4.8 mmol/kg.
External Malformations in Fetuses
Incidences of fetuses having NTDs and limb defects found upon cesarean section conducted on GD 18 in dams treated with VPA or one of the TMCA amide derivatives are presented in Table 3. In this study, VPA disrupted neural tube closure in NMRI embryos, induced NTDs in fetuses at the rate of 11.0% at 2.4 mmol/kg (Experiment 1), 22.1% at 3.6 mmol/kg (Experiment 2), and 19.3 and 50.0% at 4.8 mmol/kg (Experiments 1 and 2, respectively). These VPA-induced NTDs were predominantly exencephaly and some meningoencephalocele (Fig. 2), which are both the result of anterior neural tube closure failure. No fetuses with NTDs were noted at the dose level of 1.8 mmol/kg VPA.
Neural tube closure defects observed in fetuses at the GD 18 cesarean section. Exencephaly (B) and meningoencephalocele (C) induced by VPA and TMCD, respectively. The control fetus is shown in (A). Control and exencephalic skulls are shown in (D) and (E), respectively. Note the absence of frontal, parietal, interparietal, and supraoccipital bones and their intermediate cartilage in the exencephalic fetus (E).
Of the TMCD-treated fetuses, only one (1.2%) had an NTD (meningoencephalocele); this fetus was from a dam treated at 4.8 mmol/kg (Fig. 2). None of the fetuses from dams given up to 3.6 mmol/kg TMCD showed any NTDs. Although MTMCD at 4.8 mmol/kg induced exencephalic fetuses at a rate of 21.3% (Experiment 1), none of the fetus from dams given up to 3.6 mmol/kg showed any NTDs (Experiments 1 and 2). Two fetuses (1.7%) at 2.4 mmol/kg TMCD-thiadiazolesulfonamide were found to be exencephalic, but 91 of the live fetuses whose dams treated at 4.8 mmol/kg had no NTDs. However, TMCD-thiadiazolesulfonamide did induce limb defects in fetuses at rates of 23.4 and 27.5% at 2.4 and 4.8 mmol/kg, respectively. These limb defects included polydactyly and ectodactyly, which occurred predominantly in the forelimbs (Fig. 3). In contrast, no external malformations, including NTDs, were apparent in fetuses from dams treated with up to 4.8 mmol/kg TMCD-benzenesulfonamide.
Limb malformations observed in fetuses upon cesarean section on GD 18. Postaxial forelimb polydactyly and ectodactyly induced by TMCD-thiadiazolesulfonamide ([A], middle and right, respectively). Left in (A) shows the forelimb of the control fetus. The forelimb bones of the control and ectodactylic fetuses are shown in (B) and (C), respectively.
Effects on Fetal Skeletons
In order to compare the potency of the amide derivatives of the VPA analogue TMCA to that of VPA with regard to induction of axial skeleton malformations, the bone and cartilage of fetuses treated with VPA or one of the TMCA amide derivatives were double-stained, and their vertebrae and ribs examined (results summarized in Tables 4–6). As was shown in previous reports (Menegola et al., 1998; Okada et al., 2004, 2006), VPA at 1.8 and 2.4 mmol/kg induced significant increases in the rates of a variety of skeletal malformations, including fusion and transformation of the vertebrae, fusion of the ribs, the presence of SNRs, and an extra (8th) pair of vertebrosternal (true) ribs (Table 4, Fig. 4). Ribs were fused partially or completely at the bone and cartilage levels, and an increased incidence of lumbar SNRs in a category of the longer “extra” was also noted. The extra (8th) pair of vertebrosternal (true) ribs was specified as cartilaginous ribs, which connected to the sternum to bring the total number of sternebrae to 7; this indicated anterior transformation. A detailed region-by-region examination was conducted to identify the areas of vertebral fusion and transformation, and found to be localized predominantly in the cervical and thoracic vertebrae (Table 5). The anterior vertebral transformation of T3 to T2 and S1 to L6 was frequently observed (Table 6), and the posterior transformation of C5 to C6 was also identified.
Control (A–C) and abnormal (D–F) fetal skeletons stained with alizarin red S for bone and alcian blue for cartilage. (A) (Control) and (D) (VPA 2.4 mmol/kg): Dorsal views of the cervicothoracic region. Note the fusion of the vertebral arches and ribs (arrows) in the treated fetus. (B) (Control) and (E) (2.4 mmol/kg TMCD-thiadiazolesulfonamide): Dorsal views of the lumbosacral region. Note the misshapen (unilateral transformation) 1st left sacral arch and transverse process (arrow) in the treated fetus. Both the left and right 1st caudal transverse processes are connected to the 4th sacral transverse processes (arrowheads). (C) (Control) and (F) (4.8 mmol/kg MTMCD): Ventral views of the sternum. Note the approach of the 8th pair of cartilaginous ribs to the sternum (arrows) in the treated fetus.
| Compound | Dose (mmol/ kg) | No. of dams | No. of fetuses | Vertebra | Rib | Exp. no. | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Fusion | Transformation | Fusion | CSNR | LSNR, rudimentary | LSNR, extra | 8 V-S ribs | |||||
| Control 1 | 8 | 103 | 6 | 5 | 2 | 17 | 14 | 26 | 1 | 1 | |
| (5.1 ± 6.0) | (4.2 ± 6.0) | (1.6 ± 4.4) | (14.5 ± 24.2) | (11.6 ± 12.4) | (22.9 ± 22.1) | (1.0 ± 2.7) | |||||
| Control 2 | 8 | 110 | 1 | 20 | 1 | 11 | 31 | 20 | 7 | 2 | |
| (0.9 ± 2.5) | (18.4 ± 23.5) | (0.9 ± 2.5) | (9.7 ± 25.0) | (27.5 ± 26.4) | (19.3 ± 16.8) | (6.9 ± 12.8) | |||||
| VPA | 1.8 | 8 | 94 | 52 | 41 | 52 | 21 | 18 | 66 | 66 | 2 |
| (54.8 ± 18.2##) | (42.3 ± 20.4#) | (53.3 ± 32.3##) | (25.0 ± 23.3#) | (21.3 ± 16.3) | (69.9 ± 19.3##) | (68.8 ± 24.9##) | |||||
| 2.4 | 7 | 74 | 44 | 32 | 30 | 8 | 3 | 44 | 40 | 1 | |
| (55.5 ± 41.9*) | (48.5 ± 19.4**) | (35.2 ± 38.2*) | (9.4 ± 6.8) | (6.8 ± 12.9) | (60.6 ± 13.4**) | (55.4 ± 15.4**) | |||||
| TMCD | 4.8 | 6 | 81 | 5 | 17 | 8 | 13 | 18 | 34 | 15 | 2 |
| (5.7 ± 9.2) | (21.0 ± 14.9) | (9.4 ± 14.1) | (15.5 ± 24.5) | (22.6 ± 16.1) | (42.2 ± 35.5) | (17.8 ± 36.9) | |||||
| MTMCD | 2.4 | 7 | 98 | 5 | 2 | 1 | 26 | 15 | 20 | 2 | 1 |
| (5.1 ± 5.1) | (1.8 ± 4.7) | (0.9 ± 2.4) | (26.0 ± 24.6) | (15.3 ± 11.0) | (20.2 ± 20.5) | (2.1 ± 3.6) | |||||
| 3.6 | 8 | 93 | 4 | 31 | 2 | 3 | 27 | 27 | 2 | 2 | |
| (4.1 ± 8.4) | (31.6 ± 30.7) | (1.9 ± 5.4) | (2.9 ± 8.2) | (29.9 ± 16.2) | (29.9 ± 31.9) | (2.2 ± 4.1) | |||||
| 4.8 | 9 | 84 | 7 | 2 | 11 | 11 | 20 | 33 | 23 | 1 | |
| (11.1 ± 22.2) | (13.4 ± 35.1) | (16.2 ± 29.7) | (12.6 ± 11.8) | (31.3 ± 32.5) | (34.3 ± 26.2) | (40.0 ± 45.6*) | |||||
| TMCD- benzene-sulfonamide | 4.8 | 7 | 83 | 4 | 9 | 1 | 4 | 22 | 23 | 2 | 2 |
| (4.4 ± 7.7) | (10.9 ± 16.6) | (1.3 ± 3.4) | (10.6 ± 18.6) | (29.4 ± 16.6) | (29.3 ± 16.9) | (1.8 ± 3.1) | |||||
| TMCD- thiadiazolesulfonamide | 2.4 | 10 | 103 | 10 | 12 | 4 | 25 | 8 | 57 | 1 | 1 |
| (8.6 ± 9.1) | (9.5 ± 15.3) | (4.0 ± 8.8) | (27.9 ± 22.7) | (8.9 ± 7.1) | (56.8 ± 18.9**) | (1.3 ± 4.0) | |||||
| 4.8 | 8 | 72 | 4 | 9 | 22 | 18 | 1 | 1 | |||
| (5.9 ± 10.1) | (14.5 ± 24.8) | (23.3 ± 22.4) | (30.6 ± 32.5) | (2.9 ± 7.6) | |||||||
- CSNR, cervical supernumerary rib; LSNR, lumbar supernumerary rib; V-S, vertebrosternal.
- ** p < .01.
- * p < .05 vs. Control 1.
- ## p < .01.
- # p < .05 vs. Control 2.
- Total number (mean ± S.D.).
| Compound | Dose (mmol/kg) | No. of fetuses | Vertebral fusion | Exp. no. | |||||
|---|---|---|---|---|---|---|---|---|---|
| Cervical | C7 – T1 | Thoracic | Lumbar | Sacral | Caudal | ||||
| Control 1 | 103 | 6 (5.1 ± 6.0) | 1 | ||||||
| Control 2 | 110 | 1 (0.9 ± 2.5) | 2 | ||||||
| VPA | 1.8 | 94 | 20 (21.5 ± 13.4##) | 12 (13.6 ± 11.4##) | 26 (24.9 ± 22.2##) | 2 | |||
| 2.4 | 74 | 17 (27.2 ± 33.4**) | 13 (13.7 ± 13.7) | 25 (26.4 ± 35.8*) | 1 | ||||
| TMCD | 4.8 | 81 | 4 (4.5 ± 6.9) | 1 (1.2 ± 2.9) | 2 | ||||
| MTMCD | 2.4 | 98 | 1 (1.0 ± 2.7) | 3 (2.9 ± 5.1) | 1 | ||||
| 3.6 | 93 | 3 (3.2 ± 6.1) | 1 (1.0 ± 2.7) | 2 | |||||
| 4.8 | 84 | 1 (2.5 ± 7.1) | 7 (11.1 ± 22.2) | 1 | |||||
| TMCD-benzenesulfonamide | 4.8 | 83 | 4 (4.4 ± 7.7) | 2 | |||||
| TMCD-thiadiazolesulfonamide | 2.4 | 103 | 1 (0.7 ± 2.1) | 3 (2.8 ± 5.9) | 2 (2.3 ± 4.8) | 4 (2.9 ± 6.7) | 1 (1.3 ± 4.0) | 1 | |
| 4.8 | 72 | 3 (3.1 ± 8.1) | 1 (2.9 ± 7.6) | 1 | |||||
- ** p < .01,
- * p < .05 vs. Control 1.
- ## p < .01.
- #p < .05 vs. Control 2.
- Total number (mean ± S.D.).
| Compound | Dose (mmol/kg) | No. of fetuses | Vertebral transformation | Exp. no. | |||||
|---|---|---|---|---|---|---|---|---|---|
| C5 to C6 | C7 to C6 | T3 to T2 | L6 to S1 | S1 to L6 | C1 to S4 | ||||
| Control 1 | 103 | 2 (1.7 ± 3.3) | 2 (1.7 ± 4.7) | 1 (0.8 ± 2.2) | 1 | ||||
| Control 2 | 110 | 1 (1.1 ± 3.2) | 2 | ||||||
| VPA | 1.8 | 94 | 4 (3.6 ± 5.3) | 1 (0.8 ± 2.4) | 22 (22.2 ± 26.2##) | 7 (7.1 ± 8.7) | 2 | ||
| 2.4 | 74 | 10 (15.0 ± 19.5) | 16 (25.6 ± 28.0*) | 1 (1.2 ± 3.2) | 7 (9.3 ± 9.1*) | 1 (1.0 ± 2.7) | 1 | ||
| TMCD | 4.8 | 81 | 1 (1.2 ± 2.9) | 2 | |||||
| MTMCD | 2.4 | 98 | 1 (0.9 ± 2.4) | 1 (0.9 ± 2.4) | 1 | ||||
| 3.6 | 93 | 1 (0.9 ± 2.5) | 2 | ||||||
| 4.8 | 84 | 1 (12.5 ± 35.4) | 1 (0.9 ± 2.5) | 1 | |||||
| TMCD-benzenesulfonamide | 4.8 | 83 | 2 (2.6 ± 6.9) | 1 (0.9 ± 2.4) | 2 | ||||
| TMCD-thiadiazolebenzenesulfonamide | 2.4 | 103 | 1 (1.0 ± 3.2) | 8 (5.9 ± 12.6) | 5 (3.9 ± 6.9) | 1 | |||
| 4.8 | 72 | 9 (14.5 ± 24.8) | 3 (4.6 ± 8.6) | 1 | |||||
- **p < .01.
- * p < .05 vs. Control 1.
- ## p < .01.
- #p < .05 vs. Control 2.
- Total number (mean ± S.D.).
MTMCD at 4.8 mmol/kg significantly increased the incidence of fetuses with an 8th pair of vertebrosternal ribs (Table 4), as well as the incidence of fetuses with vertebral fusion and rib fusion (no statistical significance). This effect was not observed at 2.4 or 3.6 mmol/kg. At 3.6 mmol/kg of MTMCD, a high incidence of vertebral transformation (31.6%, no statistical significance) was noted, but there was no clear dose dependency.
Fetuses exposed to TMCD-thiadiazolesulfonamide at 2.4 and 4.8 mmol/kg had longer lumbar SNRs (extra) than those in the control (Table 4). No induction of rib fusion or generation of an 8th pair of vertebrosternal ribs was evident. Even though there were no statistically significant increases in total incidence of vertebral fusion or transformation by TMCD-thiadiazolesulfonamide, the vertebral regions affected by this compound were distributed more caudally to the lumbar and sacral vertebrae than those affected by VPA (Tables 5 and 6).
No statistically significant increases in vertebral or rib abnormalities were detected in this study at the highest dose level (4.8 mmol/kg) of TMCD or TMCD-benzenesulfonamide, respectively. Although TMCD induced slight increases in rib fusion, extra ribs, and 8th V-S ribs when compared to those in the control group, the increases were clearly less than that induced by VPA.
DISCUSSION
More potent and safer CNS-active VPA-derivatives have been developed using structure-activity relationship and structure-teratogenicity relationship studies (Nau et al., 1991; Isoherranen et al., 2003; Bialer and Yagen, 2007). Changes in the chemical structure of the alkane moiety of VPA or conversion of its carboxylic group to form amide derivatives led to dramatic differences in the pharmacologic and teratogenic properties of these compounds (Isoherranen et al., 2003; Sobol et al., 2004; Bialer and Yagen, 2007; Shimshoni et al., 2007). In this study, we tested the teratogenic activity of four amide derivatives of TMCA: TMCD, MTMCD, TMCD-benzenesulfonamide, and TMCD-thiadiazolesulfonamide in pregnant NMRI mice susceptible to VPA-induced NTDs and skeletal abnormalities. Our experimental findings indicate that the CNS-active amides containing a tetramethylcyclopropanecarbonyl moiety demonstrate a better anticonvulsant potency and a wider safety margin than VPA in animal models of epilepsy. This suggests the potential of these amide derivatives of TMCA to become second-generation to VPA drugs.
Cesarean section revealed that MTMCD caused embryo lethality, and TMCD-thiadiazolesulfonamide caused both embryo lethality and fetal growth retardation when administered subcutaneously on GD 8.5. MTMCD caused a significant increase in fetal loss rate starting at 3.6 mmol/kg. A similar dose level of VPA induced embryo lethality. However, it is important to note that MTMCD was more potent than VPA in MES and scMet tests and has a wider safety margin (Isoherranen et al., 2002). In contrast, the fetal loss rate and fetal body weights of dams treated with TMCD and TMCD-benzenesulfonamide were comparable to those observed in the control group, which indicates that these compounds imposed no risk to embryo-fetal survival or growth rate in utero at these dose levels. Our experimental findings indicated that the embryo lethality of the tested compounds was dose-related and dependent on the mouse strain used. In a previous study conducted in SWV/Fnn mice, both MTMCD and TMCD contributed to embryo lethality at doses above 350 mg/kg (2.5 mmol/kg) (Isoherranen et al., 2002), and an ip injection of MTMCD at 300 mg/kg (1.9 mmol/kg) on GD 8.5 caused 15% of embryos to be resorbed. Although the embryo lethality of TMCD in that study did not agree with the results obtained in the present study, an increased fetal loss rate was detected at 450 mg/kg (3.2 mmol/kg, ip), which was a dose potent enough to cause maternal death in SWV/Finn mice (Isoherranen et al., 2002). This discrepancy between the dose causing embryo lethality in the previous and present studies may be due to a difference in injection route (ip or sc) and/or mouse strain (SWV/Fnn or NMRI). Nonetheless, compared to VPA, the amide derivatives of the VPA analogue TMCA had a lower risk of embryo lethality and growth retardation in mice.
In contrast to VPA, the rate of exencephaly in NMRI mice dosed with TMCD, TMCD-benzenesulfonamide, and TMCD-thiadiazolesulfonamide was low to none. TMCD at 4.8 mmol/kg induced NTD in only one fetus (1.2%). At the same molar dose, VPA induced 19.3 and 50.0% of NTDs in Experiments 1 and 2, respectively. In Experiment 1, the incidence of NTDs in mice dosed with MTMCD at 4.8 mmol/kg (21.3%) was comparable to that induced by VPA (19.3%) at the same level. However, at doses below 3.6 mmol/kg, MTMCD induced no NTD fetuses, while VPA induced 11% (2.4 mmol/kg) and 22% (3.6 mmol/kg) in Experiments 1 and 2, respectively. Although TMCD-thiadiazolesulfonamide induced two exencephalic fetuses (1.7%) at 2.4mmol/kg, there were no NTD fetuses at 4.8 mmol/kg of TMCD-thiadiazolesulfonamide. At 2.4 mmol/kg, VPA caused exencephaly in 11% of fetuses. TMCD-benzenesulfonamide induced no NTDs, even at 4.8 mmol/kg. In another well-established SWV/Fnn mouse model of VPA-induced NTDs (Finnell et al., 1988), TMCD or MTMCD demonstrated lower teratogenicity and induced fewer NTDs than VPA (Isoherranen et al., 2002). Moreover, because these derivatives are potent anticonvulsants, their activity-to-NTD formation ratio appears to be much more beneficial than that of VPA.
Although 2.4 and 4.8 mmol/kg of TMCD-thiadiazolesulfonamide did not increase significant NTDs when administered on GD 8.5, both doses produced limb defects (predominantly forelimb ectodactyly and polydactyly) in 23.4 and 27.5% of fetuses, respectively. It is well known that AEDs, including VPA, cause forelimb ectodactyly in mice, and that this effect depends on the time at which the chemicals are administered during embryo development (Collins et al., 1991). Administration of VPA later than day 9 plus 8 h (approximately 21 somites) induced a high percentage of this malformation, even though earlier administration of VPA (day 9 plus 0 h, approximately 17 somites) produced no limb defects. Although, in this study, TMCD-thiadiazolesulfonamide was injected at GD 8.5 (approximately 6 somites), it was suggested that fetuses were exposed to TMCD-thiadiazolesulfonamide at the developmental stage (more than 21 somites) during the critical period for development of forelimb ectodactyly. Thus, it was suspected that TMCD- thiadiazolesulfonamide suspended in 10% DMSO had a different pharmaco/toxicokinetic profile than other analogues, and that this resulted in a high level of exposure to TMCD-thiadiazolesulfonamide at later stages of fetal development (at least 21 somites). It is also possible that the fetuses were exposed to teratogenic metabolites of TMCD-thiadiazolesulfonamide at a later time (delayed teratogenicity).
During the skeletal examination of fetal vertebra and ribs, VPA induced a number of malformations at higher frequencies than that seen in the control group. These severe skeletal abnormalities included fusion and transformation of vertebrae as well as fused ribs and SNRs, as reported previously (Okada et al., 2004, 2006). The axial abnormalities could be explained by abnormal segmental identity specification during the anteroposterior patterning of the embryo. At 1.8 and 2.4 mmol/kg, approximately half of the fetuses exposed to VPA exhibited vertebral fusion and transformation as well as rib fusion, and, at 3.6 and 4.8 mmol/kg, these skeletal abnormalities became severely complex. In contrast to VPA, all four amide derivatives of the tetramethylcyclopropyl analogue of VPA tested in this study showed significantly fewer teratogenic effects on the axial skeleton of NMRI mouse fetuses. Even at the highest dose level of 4.8 mmol/kg, TMCD-benzenesulfonamide did not induce any axial skeleton abnormalities in mouse fetuses. At 4.8 mmol/kg of TMCD, no statistical significant increases in vertebral or rib abnormalities were detected. Although TMCD induced slight increases in rib fusion, extra ribs, and 8th V-S ribs when compared to the control group, these increases were clearly less than that induced by VPA. MTMCD significantly induced the formation of an 8th pair of vertebrosternal ribs, one of the characteristics of anterior transformation, which was also observed in VPA-treated fetuses. Although MTMCD at 4.8 mmol/kg also increased the incidences of fetuses with vertebral fusion and rib fusion, these changes were slight and not statistically significant. Furthermore, these effects of MTMCD were not observed at 2.4 or 3.6 mmol/kg, which indicates that this compound has lower skeletal teratogenicity than VPA (same molar dose). Significant induction of extra lumbar SNRs was detected at a high frequency when NMRI mice were treated with both TMCD-thiadiazolesulfonamide and VPA at 2.4 mmol/kg. Analysis of the vertebral regions affected showed that vertebral fusion and transformation caused by TMCD-thiadiazolesulfonamide were more caudal to the lumbar and sacral vertebrae than those induced by VPA (cervical and thoracic regions). It is known that the axial skeleton malformations in rats produced by VPA are stage-dependent, and the effected regions are those formed during the treatment period. Exposure to VPA at the six-somite stage showed severe abnormalities localized at the 4th to 7th cervical segment and at the level of the 1st and 2nd thoracic segments (Menegola et al., 1998), which agrees with the results of this study. Lumbosacral abnormalities were induced by exposure to VPA mainly at the 18- to 22-somite stage (Menegola et al., 1998). Despite the stage-dependency of VPA's teratogenesis along the axial skeleton, administration of TMCD-thiadiazolesulfonamide caused fusion and transformation of vertebrae in posterior regions. As mentioned above, exposure to TMCD-thiadiazolesulfonamide or its metabolites at a later time, when posterior axial skeleton abnormalities would be induced, was suspected. Nonetheless, TMCD-thiadiazolesulfonamide did not statistically significantly increase the total incidence of vertebral fusion, transformation, or rib fusion in fetuses, thereby indicating that TMCD-thiadiazolesulfonamide is less teratogenic than VPA. Taken together, these results for all four amide derivatives of TMCA, a tetramethylcyclopropyl analogue of VPA, indicate significantly fewer teratogenic effects on embryonic anteroposterior patterning in a well-established NMRI mouse model of VPA-induced axial skeleton teratogenicity. The evidence of a reduced risk of skeletal teratogenicity strongly suggests that the four amide derivatives of TMCA investigated in this study pose lower risks of inducing fetal skeleton abnormalities.
In conclusion, our results indicate that TMCD, MTMCD, and TMCD-benzenesulfonamide are distinctly less teratogenic than VPA with regard to the induction of abnormalities in the developing neural tube and skeleton in NMRI mice. This evidence suggests that amide derivatives of TMCA, a cyclopropane analogue of VPA, reduce their teratogenic potency compared to VPA. Thus, in addition to the improved anticonvulsant profiles of TMCD, MTMCD, and TMCD-benzenesulfonamide, the results of this study strengthen the potential of TMCA amide derivatives to become new antiepileptics and CNS drugs with a more potent and safer profile, as the second-generation to VPA.
Acknowledgements
The authors gratefully acknowledge Ms. Hiroko Noyori, Mr. Seiki Matsuo, Ms. Yuko Sakai, Mr. Shin Ito, and Mr. Tadashi Saegusa for their assistance with the teratology study and the double staining and examination of the fetal skeleton.
