Hydroxamic acid and fluorinated derivatives of valproic acid: Anticonvulsant activity, neurotoxicity and teratogenicity

Neurotoxicology and Teratology 30 (2008) 390–394

Contents lists available at ScienceDirect

Neurotoxicology and Teratology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / n e u t e r a

Hydroxamic acid and fluorinated derivatives of valproic acid: Anticonvulsant activity, neurotoxicity and teratogenicity Ute Gravemann, Jutta Volland, Heinz Nau ⁎ Department of Food Toxicology, University of Veterinary Medicine Foundation, D-30173 Hannover, Germany

A R T I C L E

I N F O

Article history: Received 17 December 2007 Received in revised form 26 February 2008 Accepted 10 March 2008 Available online 19 March 2008 Keywords: Anticonvulsant activity Fluorination Hydroxamic acids Neurotoxicity Teratology Valproic acid

A B S T R A C T Purpose: Fluorinated and non-fluorinated valproic acid (VPA) analogues with hydroxamic acid moieties were tested for their teratogenic, anticonvulsant and neurotoxic potencies in mice. Methods: Compounds were synthesized from their corresponding acids. The induction of neural tube defects (exencephaly) of the resulting hydroxamates (applied on day 8.25 of gestation) was tested in the offspring of pregnant animals (Han:NMRI mice). The anticonvulsant activity was evaluated in the subcutaneous pentylenetetrazole (PTZ) seizure threshold test and neurotoxicity in the rotorod neurotoxicity test. Results: All tested hydroxamates showed no or greatly reduced teratogenic potency in mice compared to the free acids. Furthermore all compounds exhibited anticonvulsant activity with ED50 doses ranging from 0.16 mmol/kg to 0.59 mmol/kg (VPA 0.57 mmol/kg). Neurotoxicity of the hydroxamates was increased compared to VPA. TD50 doses range from 0.70 mmol/kg to 1.42 mmol/kg (VPA 1.83 mmol/kg). Conclusion: Hydroxamic acid derivatives of VPA with improved protective index and little or undetectable teratogenic potency compared to the free acids are described. α-fluorination of VPA also resulted in loss of teratogenic activity. Such fluorination of the hydroxamic acids also led to compounds with an improved anticonvulsant profile compared to non-fluorinated hydroxamates. The non-chiral 2-Fluoro-VPA-hydroxamic acid was the most promising compound with a protective index (ratio of TD50 to ED50) of 4.4 compared to 3.2 for VPA. This compound combines an improved ratio of anticonvulsant potency/neurotoxicity with the advantage of not being teratogenic in the mouse neural tube defect model used. © 2008 Published by Elsevier Inc.

1. Introduction Valproic acid (VPA, 1) is one of the major drugs for the treatment of epilepsy, for the treatment of bipolar disorders and in the prophylaxis of migraine [1,19,35]. More recently antitumor activity [5] and cognition-enhancing properties [10,22] of valproic acid and its derivatives have been discussed in addition to inhibition of angiogenesis and differentiation of leukemic cells [5,21]. VPA was also shown to modulate differentiation of stem cells into cardiomyocytes [23]. Some of these effects, especially neural tube defect induction and inhibition of cancer cell growth and differentiation, are likely to be induced via inhibition of histone deacetylases [9], and induction of DNA-demethylation [6]. The teratogenic mechanism of VPA may also include the induction of the homeotic gene Hoxa1 [32]. Anticonvulsant activity of these compounds is apparently not mediated by inhibition of histone deacetylation, which enabled us now – in a

⁎ Corresponding author. Department of Food Toxicology, University of Veterinary Medicine Foundation, Bischofsholer Damm 15, D-30173 Hannover, Germany. Tel.: +49 511 856 7600; fax: +49 511 856 7680. E-mail address: [email protected] (H. Nau). 0892-0362/$ – see front matter © 2008 Published by Elsevier Inc. doi:10.1016/j.ntt.2008.03.060

mechanistically-driven approach – to develop additional analogs with the desired pharmacological property, but avoiding teratologic effects. Despite its broad use in the therapy of several diseases, treatment with VPA is associated with different side effects, with the most serious of those being its hepatotoxicity and teratogenicity [15]. In man, as well as in mice VPA, when administered during early organogenesis, causes neural tube defects (NTD) in addition to other teratogenic and postnatal neuroteratogenic effects. The most common VPA-induced neural tube defect is spina bifida aperta in humans [27,29], and exencephaly in mice, which can be induced by the injection of VPA on day 8.25 of gestation [2,3,24,25]. Higher doses and later treatment (day 9) produced spina bifida aperta in mice [7]. Several studies [2,3,24,25,28] on different valproic acid derivatives revealed clearly that teratogenicity of the compounds is strictly structure-dependent. Different structural aspects such as further branching of the aliphatic side chain as well as modification of the carboxyl-function lead to greatly reduced or undetectable teratogenicity of resulting compounds [3,26,28,30]. Thus, valpromide (2), the amide of valproic acid, is a compound of low or undetectable teratogenic potency in the mouse model [28]. However in humans valpromide (2), an epoxide hydrolase inhibitor [31], serves as prodrug

U. Gravemann et al. / Neurotoxicology and Teratology 30 (2008) 390–394

for VPA. Valpromide may therefore not be a valid alternative anticonvulsant drug in the human. The exchange of one hydrogen in the amide by a hydroxyl group leads to so-called hydroxamic acids. Hydroxamic acids are a class of compounds closely related to amides. They are weak acids with a pKavalue of ∼ 9 which lies between the pKa of weak carboxylic acids (pKa = 4–5) and amides (pKa ∼ 17). Due to their similarities with amides the hydroxamic acid derivatives may also exhibit a low teratogenic potential. The hydroxamic acid derivative of VPA (VPA-HA, 3) has previously been shown to have improved anticonvulsant activity in the electrically induced shock (MES) test compared with valproic acid [4,12,14,16,17]. The teratogenicity of this compound has not yet been investigated. Therefore the aim of this study was to synthesize VPA-HA (3) and several other structural analogues such as fluorinated derivatives [34], and to investigate their teratogenic risk as well as their neurotoxic and anticonvulsant potency. 2. Materials and methods 2.1. VPA-derivatives All compounds were numbered in the order of appearance in the text as indicated in Figs. 1 and 3. 2.2. Chemistry Melting points were determined using the Büchi melting point apparatus B-450 and are uncorrected. Flash chromatography was carried out on Macherey and Nagel silica gel 60 (230–400 mesh) and thin layer chromatography was performed on precoated silica gel plates (Macherey and Nagel, silica gel 60 / UV254). NMR spectra were recorded with a Bruker AM 300-MHz spectrometer using tetramethylsilane as the internal standard. 2.2.1. Synthesis Valproic acid was purchased from Sigma-Aldrich. Hydroxamic acids were synthesized starting from the corresponding acids which were prepared as described previously [4,12,14,16]. Compound 8 was

391

prepared from the corresponding acid 7 using standard procedure for amidation. 2.2.1.1. 2-Fluoro-2-propyl-pentanoic acid (2-F-VPA) (7). 1H-NMR (300 MHz): δ = 0.94 (t, J = 7.3 Hz), 1.24–1.44, 1.48–1.66 (2 m, 4H), 1.72–2.01 (m, 4H), 7.84 (bs, 1H). 13C-NMR (75 MHz): δ = 14.0, 16.6 (JCF = 2 Hz), 39.3 (JCF = 22 Hz), 97.9 (JCF = 187 Hz), 176.8 (JCF = 28 Hz). 2.2.1.2. 2-Fluoro-2-propyl-pentanamide (2-F-VPD) (8). 1 H-NMR (300 MHz): δ = 0.89–0.95 (m, 6H), 1.28–1.1.56 (m, 4H), 1.66–1.98 (m, 4H), 6.25, 6.4 (bs, 2H). 13C-NMR (75 MHz): δ = 14.0 (p, 2C, C-3′, C-5), 16.40, 16.44 (s, 2C, C-4, C-2′), 39.07, 39.29 (p, 2C, C-3, C1′), 100.9 (q, C-2, JCF = 181.1 Hz), 175.1 (q, C-1, JCF = 20.11 Hz). Compound 6 was prepared via fluorination of the methyl ester, subsequent saponification and conversion to the hydroxamate (Fig. 3). 2.3. Fluorination procedure Buthyl lithium (34.4 ml, 1.6 M in hexan, 0.05 mol) was added at 0 °C to a solution of diisopropylamine (7.2 ml, 0.05 mol) in dry THF (120 ml). The mixture was cooled to −70 °C and a solution of 13 (7.2 g, 0.045 mol) in dry THF (20 ml) was added slowly. The mixture was warmed up to −20 °C within 2 h then cooled down again to −70 °C. After the addition of N-Fluorobenzensulfonimide (21 g, 0.05 mol) in dry THF (40 ml), the reaction mixture was warmed up to room temperature over 12 h and was subsequently quenched with hydrochloric acid (10%, 100 ml). The water layer was extracted with diethyl ether, dried over Na2SO4 and evaporated in vacuo. Distillation of the crude product affords the fluorinated ester 14 (6.8 g, 0.039 mol, 85%) as a clear colourless oil. 2.3.1. (±)-Methyl-2-fluoro-2-ethyl-4-methyl-pentanoate (14) 1 H-NMR (300 MHz): δ = 0.79–1.00 (m, 9H), 1.64–1.99 (m, 5H), 3.79 (s, 3H). 13C-NMR (75 MHz): δ = 7.52 (p, C-2′), 23.30, 23.60 (p, 2C, C-5, C-5′), 24.39 (t, C-4), 31.58 (s, C-3, JCF =22.9 Hz), 45.63 (s, C-1′, JCF = 21.2 Hz), 52.13 (p, −OCH3), 98.27 (q, C-2, JCF =188.8 Hz), 172.50 (q, C-1, JCF =26.1 Hz). 2.4. Alkaline hydrolysis Hydrolysis was carried out as described before [16] and the resulting acid was purified via fractionated distillation. 2.4.1. (±)-2-Fluoro-2-ethyl-4-methyl-pentanoic acid (15) 1 H-NMR (300 MHz): δ = 0.81–1.10 (m, 9H), 1.67–2.06 (m, 5H), 9.83 (bs, 1H). 13C-NMR (75 MHz): δ = 7.4, 23.4, 23.6, 24.4, 31.3 (JCF = 23.1 Hz), 45.2 (JCF = 21.1 Hz), 98.0 (JCF = 188.9 Hz), 177.7 (JCF = 26.9 Hz). 2.5. General procedure for the synthesis of hydroxamic acids Carboxylic acids were converted to the corresponding acid chlorides by thionylchloride and used without further purification. Hydroxylamine hydrochloride (13.55 g, 0.2 mol) was dissolved in water (50 ml), then triethylamine (27.8 ml, 0.2 mol) was added at 0 °C and the solution was stirred for 30 min. After the addition of the crude acid chloride (10 mmol) in dry THF (20 ml) the reaction mixture was stirred an additional 60 min. Layers were separated and the aqueous solution was extracted with CH2Cl2. The combined organic layers were dried over Na2SO4 and evaporated in vacuo to afford the crude hydroxamic acids as colourless solids, which were further purified via recrystallization from hexane / CH2Cl2.

Fig. 1. Chemical structures of VPA and different analogues.

2.5.1. N-hydroxy-2-propyl-pentanamide (VPA-HA, 3) mp 126–127 °C. 1H-NMR (300 MHz): δ = 0.89 (t, 6H, J = 7.1 Hz), 1.16– 1.48 (m, 6H), 1.54–1.72 (m, 2H), 1.98 (dd, 1H), 8.18 (bs, −NHOH) 13C-NMR (75 MHz): δ = 14.0, 20.7, 34.7, 44.0, 174.3.

392

U. Gravemann et al. / Neurotoxicology and Teratology 30 (2008) 390–394

Table 1 Teratogenicity of VPA and analogues Compounda Dose

Fetal weight

[mmol/ kg] 3 4 5 6 8 9 10 VPA (1) Controlsd 11e Controlsf

[g]

3.0 1.11 ± 0.11 3.0 1.14 ± 0.11 3.0 1.18 ± 0.11 3.0 1.19 ± 0.09 3.0 1.20 ± 0.08 3.0 1.13 ± 0.15 1.5 1.09 ± 0.11 2.0 1.11 ± 0.10 3.0 1.06 ± 0.11 10 ml/kg 1.08 ± 0.10 1.25 0.93 ± 0.13 10 ml/kg 1.11 ± 0.11

Number Number of Embryo- Exencephalyc of litters live fetuses lethalityb [n]

[n]

[%]

[%]

7 6 8 9 9 9 10 7 8 15 8 6

89 77 90 106 111 101 108 67 91 185 67 71

6 8 5 9 6 9 13 18 20 11 36 8

0 0 0 0 0 1 14° 6 39° 0 81⁎ 0

° significantly different from Cremophor controls (⁎p b 0.05, Fischer Exact Test); ⁎ significantly different from NaCl controls (p b 0.05, Fischer Exact Test). a b c d e f

Compounds were suspended in a mixture of Cremophor EL 25% in water. Percentage of resorptions and dead fetuses of all implants. Percentage of live fetuses. Controls received 10 ml/kg Cremophor EL 25% in water. dosed as solution of the sodium salt in water. controls received 10 ml/kg NaCl 0.9% in water.

2.5.2. (±)-N-hydroxy-2-ethyl-4-methyl-pentanamide (4) mp 111–112 °C. 1H-NMR (300 MHz): δ = 0.78–0.96 (m, 9H), 1.18– 1.31 (m, 1H), 1.41–1.69 (m, 4H), 1.92–2.06 (m, 1H), 8.22 (bs, 1H, −OH). 13 C-NMR (75 MHz): δ = 12.0, 22.2, 23.1, 25.8,26.0, 41.3, 43.9, 174.2. 2.5.3. (±)-N-hydroxy-2-propyl-4-hexynamide (5) mp 87–89 °C. 1H-NMR (300 MHz): δ = 0.91 (t, 3H, J = 7.2 Hz), 1.12– 1.40 (m, 2H), 1.46–1.74 (m, 2H), 1.78 (t, 3H, J = 2.2 Hz), 2.16–2.48 (m, 3H), 8.68 (bs, 1H, −OH). 13C-NMR (75 MHz): δ = 3.5, 13.9, 20.4, 21.8, 33.6, 43.7, 76.0, 78.2, 173.1. 2.5.4. (±)-N-hydroxy-2-ethyl-2-fluoro-4-methyl-pentanamide (6) mp 114–115 °C. 1H-NMR (300 MHz): δ = 0.79–1.03 (m, 9H), 1.58– 2.04 (m, 5H), 8.88 (bs, −NHOH). 13C-NMR (75 MHz): δ = 7.2, 23.7, 23.8, 24.1, 30.9 (JCF = 22.2 Hz), 45.0 (JCF = 20.5 Hz), 101.5 (JCF = 101.5 Hz), 169.5 (JCF = 21.7 Hz). 2.5.5. N-hydroxy-2-fluoro-2-propyl-pentanamide (9) mp 128–129 °C. 1H-NMR (300 MHz): δ = 0.91 (t, J = 7.3 Hz), 1.20– 1.36, 1.38–1.56 (m, 4H), 1.65–1.99 (m, 4H), 9.08 (bs, 1H, −OH). 13C-NMR (75 MHz): δ = 14.0, 16.3, 39.1 (JCF = 21.4 Hz), 100.9 (JCF = 182.9 Hz), 169.5 (JCF = 22.1 Hz). 2.5.6. (±)-N-hydroxy-2-(2-propynyl)-heptanamide (Pentyl-4yn-HA, 10) mp 65–66 °C. 1H-NMR (300 MHz): δ = 0.87 (t, 3H, J = 6.3 Hz), 1.27 (m, 6H), 1.48–1.75 (m, 2H), 2.06 (s, 1H), 2.23 (m, 1H), 2.42–2.53 (m, 2H), 8.70 (bs, −NHOH). 13C-NMR (75 MHz): δ = 14.0, 21.6, 22.4, 26.8, 31.5, 31.6, 43.5, 70.6, 81.3, 172.6. 2.5.7. (±)-2-(2-Propinyl)-heptanoic acid (Pentyl-4-yn-VPA) (11) 1H-NMR (300 MHz): δ = 0.89 (t, J = 6.2 Hz), 1.18–1.45 (m, 6H), 1.59– 1.80 (m, 2H), 2.01 (t, J = 2.2 Hz), 2.35–2.67 (m, 3H). 13C-NMR (75 MHz): δ = 13.9, 20.7, 22.4, 26.4, 30.9, 31.6, 44.3, 70.0, 81.2, 180.7. 3. Teratology studies Approval for the study was obtained from the government in Hannover, Germany (Az. Nr. 509-42502-00/269). The animals were kept under controlled conditions: room temperature (21 ± 1 °C), relative humidity (50 ± 5%), and 12-h light–

dark-cycle. The animals were given free access to food (Altromin 1324 diet; Lage, Germany) and tap water. All compounds except 11 were suspended in a mixture of Cremophor EL 25% in water. Compound 11 was dissolved in equimolar amounts of sodium hydroxide. Animal experiments were carried out as described previously [2,3,8,25,28]. The induction of exencephaly was determined using NMRI mice (Harlan-Winckelmann, 33176 Borchen, Germany). The compounds were tested at the dose level indicated in Table 1. They were injected subcutaneously on the morning of day 8.25 of gestation. On day 18 of gestation the dams were sacrificed, the uteri removed, and the number of implantation sites was recorded. Each fetus was weighted and inspected for the presence of external malformations. 4. Anticonvulsant and neurotoxicity assessment The procedures for the animal experiments used for the assessment of anticonvulsant activity and neurotoxicity were reported previously [11,20,33]. Anticonvulsant activity was investigated in mice in the subcutaneous PTZ seizure threshold test [20,33]. Compounds were injected intraperitoneally (ip) (10 ml/kg) 15 min before injection of PTZ (65 mg/kg). The animals were observed for 30 min and the number of animals that were protected against seizures was recorded. Neurotoxicity was assessed by the rotorod toxicity test. Compounds were suspended in cremophor EL (25% in water) and the suspension was injected intraperitoneally. After 15 min the mice were placed for 1 min on a rod rotating at 15 rpm (Rotorod, Ugo Basile, Italy). The number of animals that was not able to maintain equilibrium on the rod for at least 1 min in each of three trials was recorded. For the calculation of median anticonvulsant and neurotoxicity doses (ED50 and TD50 values, Table 2) compounds were tested at least at three different dose levels (data not shown) and the data were analysed according to Litchfield and Wilcoxon [18,20]. 5. Results The structures of the investigated compounds are presented in Fig. 1. A variety of different compounds were chosen to represent several structural aspects. Compounds were synthesized using standard methods. As an example the synthesis of compound 6 is described in Fig. 3. All hydroxamic acid derivatives were investigated for their ability to induce exencephaly in the mouse model. The results of these studies are presented in Table 1. None of the hydroxamic acid derivatives, except for compound 10 (Pentyl-4-yn-HA), induced exencephaly in the mouse model at a dose level of 3 mmol/kg. Pentyl-4-yn-HA (10) could not be investigated at the same dose level due to its high toxicity. At 1.5 mmol/kg this derivative produced 14%

Table 2 Median anticonvulsant and neurotoxic doses and protective indices Compound ED50 [mmol/kg] 3 4 5 6 8 9 10 VPA

0.38 (0.23–0.64)b 0.59 (0.44–0.78) 0.44 (0.31–0.64)b 0.37 (0.28–0.50)b,c 0.16 (0.10–0.25)b,c 0.16 (0.06–0.42)b,c 0.54 (0.45–0.66) 0.57 (0.46–0.71)

Slope TD50 function [mmol/kg] 1.91 1.43 1.87 1.42 1.97 3.30 1.18 1.30

0.82 (0.67–1.00)b 1.29 (1.14–1.46)b 1.42 (1.28–1.57) 1.00 (0.91–1.10)b,c 0.57 (0.49–0.68)b,c 0.70 (0.64–0.75)b,c 0.87 (0.72–1.06)b 1.83 (0.8–3.78)

Slope PI function 1.20 1.17 1.13 1.13 1.29 1.13 1.34 1.21

2.2 2.2 3.2 2.7 3.6 4.4 1.6 3.2

Results were calculated according to Litchfield and Wilcoxon [18]. a Lipophilicity calculated according to Crippen's fragmentation method [13]. b significantly different to VPA, ⁎p b 0.05. c significantly different to non-fluorinated compounds, ⁎p b 0.05.

clogPa 2.10 2.01 1.95 1.82 1.74 1.91 2.19 2.58

U. Gravemann et al. / Neurotoxicology and Teratology 30 (2008) 390–394

393

TD50 values range from 0.70 mmol/kg to 1.42 mmol/kg (VPA 1.83 mmol/kg). Similar to the amidated VPA-derivatives, neurotoxicity of hydroxamic acid derivatives was increased compared to the parent acids. The protective index (PI, ratio between TD50 and ED50) of the hydroxamic acids was between 1.6 and 4.4. The index of compound 5 (PI = 3.2) is comparable with the one for VPA. The fluorinated VPA-HA (9) showed good anticonvulsant results and the most favourable ratio between wanted and unwanted effects (i.e. PI = 4.4) (Fig. 2). 6. Discussion

Fig. 2. Comparison of the protective index (PI), the median anticonvulsant (ED50) and neurotoxic doses (TD50) of different VPA-related amides and hydroxamic acids.

exencephaly. The teratogenicity of 10 could be explained by its partial metabolic conversion to the corresponding carboxylic acid 11 [8]. The embryolethality rate induced by the hydroxamic acids including Pentyl-4-yn-HA was not significantly different from the control group. Also fetal weights were not affected by treatment with hydroxamic acids. Since VPA-HA (3) is known to have improved anticonvulsant activity compared to VPA [17], all hydroxamic acids were subjected to an examination of their anticonvulsant activity. All substances revealed anticonvulsant activity in the subcutaneous pentylenetetrazole (PTZ) seizure threshold test (Table 2) which is considered to be a valuable model in predicting antiepileptic activity against generalized minor seizures in humans [15,27]. The sedative activity of the compounds was assessed on the rotorod 15 min after administration (Table 2) which was determined to be the time period of peak neurotoxic effect (sedation and ataxia) for valproic acid derivatives [11]. Median anticonvulsant and neurotoxicity doses (ED50 and TD50 values, Table 2) were calculated according to Litchfield and Wilcoxon [18]. ED50 values range from 0.16 mmol/kg to 0.59 mmol/kg (VPA 0.57 mmol/kg). All hydroxamic acid derivatives tested were effective anticonvulsants.

In the present study hydroxamic acids structurally related to VPA were shown to be derivatives with reduced teratogenic potency compared to the respective carboxylic acids. In the literature there are some examples of potentially teratogenic hydroxamic acid derivatives including acetohydroxamic acid [4,14,16,17], but rather high doses were necessary to observe teratogenic effects. Higher homologues of acetohydroxamic acid like propio- and valerohydroxamic acid were non-teratogenic in rats [16]. In our investigations, only Pentyl-4-ynHA 10 exhibited teratogenic action at a dose level of 1.5 mmol/kg. Compared to VPA, Pentyl-4-yn-HA is a more potent teratogen. However, the teratogenic potency of Pentyl-4-yn-HA (14%, 1.5 mmol/kg, Table 1) is lower than that of the corresponding acid Pentyl-4-yn-VPA 11 (81%, 1.25 mmol/kg, Table 1). The teratogenic potency of Pentyl-4yn-HA 10 therefore might result from its metabolism to the acid (Pentyl-4-yn-VPA, 11) which is one of the most potent teratogens of the VPA-series [2,5]. Eikel et al. [8] have shown that Pentyl-4-yn-HA is partially metabolized to the carboxylic acid which could explain the teratogenic effects of this hydroxamate in mice. In dogs VPA-HA is considered to be a metabolically stable analogue of VPA and is not metabolized to the acid [17]. However, for other short chain aliphatic hydroxamates like aceto- or propiohydroxamate, a reductive metabolism to the amide and also hydrolysis to the acid in mice has been described in the literature [4,12,14,16,17]. Further studies on the metabolism and pharmacokinetics of Pentyl-4-yn-HA therefore should be carried out to observe its intrinsic teratogenic potency and its metabolism to the corresponding carboxylic acid (Fig. 3). In addition to their reduced teratogenicity, hydroxamates exhibit improved anticonvulsant activity in the subcutaneous pentylenetetrazole (PTZ) seizure threshold test. However the neurotoxicity of these compounds is also elevated. Compounds 5 and 9 show the same or even an improved protective index (PI) as compared to VPA. Consistent with literature [17] VPA-HA 3 had a slightly decreased PI for chemically induced seizures than VPA. Lipophilicity values of the different compounds are very similar (Table 2), and therefore do not correlate with the anticonvulsant

Fig. 3. Synthesis of N-Hydroxy-2-fluoro-2-ethyl-4-methyl-pentanamide (6)(a) MeOH, H+ (b) LDA, THF, N-Fluorobenzensulfonimide (c) LiOH, MeOH / H2O (d) SOCl2 (e) NH2OH × HCl, Et3N.

394

U. Gravemann et al. / Neurotoxicology and Teratology 30 (2008) 390–394

activity values [13]. It is therefore reasonable to assume that the anticonvulsant activity is an intrinsic property of these compounds and are not due to pharmacokinetic variables. A comparison of the fluorinated and non-fluorinated amides and hydroxamates (Fig. 2) clearly shows, that the fluorination of these compounds leads to chemical structures with a favourable anticonvulsant profile. The same effect can be seen for compounds 4 and 6. In contrast to the fluorination of amides and hydroxamates, the fluorination of acids leads to much weaker anticonvulsants. 2-Fluorovalproic acid (7), for example, is a compound with reduced anticonvulsant activity [34]. These results may provide some evidence that fluorinated amides and hydroxamates are not rapidly metabolized to their corresponding acids in mice but exhibit an intrinsic anticonvulsant activity. Fluorination of C-2 of amides and hydroxamic acids does not only improve the protective index, but will also render the corresponding carboxylic acids – should they be metabolically formed – nonteratogenic. Amide derivatives of VPA have previously been shown to induce reduced teratogenic effects in mice [25,26,28,30]. Furthermore, VPAHA, just like Valpromide (2), has been shown to be an inhibitor of epoxide hydrolase [31]. It is unclear if this enzyme inhibition affects teratogenicity or plays a role in human therapy. In conclusion the newly synthesized hydroxamates are a class of VPA-derivatives with greatly reduced or undetectable teratogenic potency in the neural tube defect model and with improved anticonvulsant activity compared to the corresponding acids such as VPA. In particular, 2-Fluoro-VPA-HA (9) is a promising, new anticonvulsant agent. Further studies therefore should be carried out on the anticonvulsant profile and the metabolism/pharmacokinetics of this new compound. Acknowledgements This work was supported by the European Commission (Research Training Network HPRN-CT-2002-00268 and the Framework 6 project ReProTect). We also acknowledge the support of the Federal Ministry for Education and Research, BMBF/PTJ/BIO, Project 0313070D. We would like to thank Georgina Zivkoviv and Hubert Haarstrich for their technical assistance. U. Gravemann and J. Volland were supported by programs of the European Commission, and by the American Biogenetic Sciences. References [1] R.A. Blaheta, H. Nau, M. Michaelis, J. Cinatl, Valproate and valproate analogues: potent tools to fight against cancer, Curr. Med. Chem. 9 (2002) 1417–1433. [2] U. Bojic, K. Ehlers, U. Ellerbeck, C.L. Bacon, E. O'Driscoll, C. O'Connell, V. Berezin, A. Kawa, E. Lepekhin, E. Bock, C.M. Regan, H. Nau, Studies on the teratogen pharmacophore of valproic acid analogues: evidence of interactions at a hydrophobic centre, Eur. J. Pharmacol. 354 (1998) 289–299. [3] U. Bojic, M.M. Elmazar, R.S. Hauck, H. Nau, Further branching of valproate-related carboxylic acids reduces the teratogenic activity, but not the anticonvulsant effect, Chem. Res. Toxicol. 9 (1996) 866–870. [4] J.M. DeSesso, C.F. Jacobson, A.R. Scialli, G.C. Goeringer, Hydroxylamine moiety of developmental toxicants is associated with early cell death: a structure-activity analysis, Teratology 62 (2000) 346–355. [5] H. Deubzer, B. Barbara, G. Rönndahl, D. Eikel, M. Michaelis, J. Cinatl, S. Schulze, H. Nau, O. Witt, Novel valproic acid derivatives with potent differentiation-inducing activity in myloid leukemia cells, Leuk. Res. 30 (2006) 1167–1175. [6] N. Detich, V. Bovenzi, M. Szyf, Valproate induces replication-independent active DNA demethylation, J. Biol. Chem. 278 (2003) 27586. [7] K. Ehlers, H. Sturje, H.J. Merker, H. Nau, Valproic acid-induced spina bifida: a mouse model, Teratology 45 (1992) 145–154. [8] D. Eikel, K. Hoffmann, K. Zoll, A. Lampen, H. Nau, S-Pentyl-4-pentynoic hydroxamic acid and its metabolites S-2-Pentyl-4-pentynoic acid in the NMRI-exencephaly mouse model: pharmakokinetic profiles, teratogenic effects, and histone deacetylase inhibition abilities of further valproic acid hydroxamates and amides, Drug Metab. Dispos. 34 (2006) 612–620.

[9] D. Eikel, A. Lampen, H. Nau, Teratogenic effects mediated by inhibition of histone deacetylases: evidence from quantitative structure activity relationships of 20 valproic acid derivatives, Chem. Res. Toxicol. 19 (2006) 272–278. [10] K. Gottfryd, S. Owczarek, K. Hoffmann, B. Klementiev, V. Berezin, E. Bock, P.S. Walmond, Multiple effects of pentyl-4-yn-VPA enantiomers: from toxicity to short-term memory enhancement, Neuropharmacology 52 (2007) 764–778. [11] M.M. Elmazar, R.S. Hauck, H. Nau. Anticonvulsant and neurotoxic activities of twelve analogues of valproic acid, J. Pharm. Sci. 82 (1993) 1255–1258. [12] W.N. Fishbein, Hydroxamic acids as urease inhibitors for medical and veterinary use, in: H. Kehl, Kehl (Eds.), Chemistry and Biology of Hydroxamic Acids. Proceedings of the First International Symposium on Chemistry and Biology of Hydroxamic Acids, New York, 1, 1982, pp. 94–103. [13] A.K. Ghose, G.M. Crippen, Atomic physicochemical parameters for three-dimensional-structure-directed quantitative structure-activity relationships. 2. Modeling dispersive and hydrophobic interactions, J. Chem. Inf. Comput. Sci. 27 (1987) 21–35. [14] N. Isoherranen, H.S. White, R.H. Finnell, B. Yagen, J.H. Woodhead, G.D. Bennett, K.S. Wilcox, M.E. Barton, M. Bialer, Anticonvulsant profile and teratogenicity of Nmethyl-tetramethylcyclopropyl carboxamide: a new antiepileptic drug, Epilepsia 43 (2002) 115–126. [15] E. Jaeger-Roman, A. Deichel, S. Jakob, A.-M. Hartmann, S. Koch, S. Rating, R. Steldinger, H. Nau, H. Helge, Fetal growth, major malformations and minor anomalies in infants born to women receiving valproic acid, J. Pediatr. 108 (1996) 997–1004. [16] T. von Kreybig, R. Preussmann, W. Schmidt, Chemical constitution and teratogenic effect in rats. I. Carbonic acid amides, carbonic acid hydrazides and hydroxamic acids], Arzneimittelforschung 18 (1968) 645–657. [17] M. Levi, B. Yagen, M. Bialer, Pharmacokinetics and antiepileptic activities of valproyl hydroxamic acid derivatives, Pharm. Res. 14 (1997) 213–217. [18] J.T. Litchfield, F.A. Wilcoxon, A simplified method of evaluating dose-effect experiments, J Pharmacol. Exp. Ther. 257 (1949) 20–31. [19] W. Löscher (Ed.), Valproate, Birkhäuser Verlag Basel, Boston, 1999. [20] W. Loscher, H. Nau, Pharmacological evaluation of various metabolites and analogues of valproic acid: anticonvulsant and toxic potencies in mice, Neuropharmacology 24 (1985) 427–435. [21] U.R. Michaelis, I. Fleming, T. Suhan, J. Cinatl, R.A. Blaheta, K. Hoffmann, R. Kotchetkov, R. Busse, H. Nau, J. Cinatl, Valproic acid inhibits angiogenesis in vitro and in vivo, Mol. Pharmacol. 65 (2004) 520–527. [22] K.J. Murphy, G.B. Fox, A.G. Foley, H.C. Gallagher, A. O'Connell, A.M. Griffin, H. Nau, C.M. Regan, Pentyl-4-yn-valproic acid enhances both spatial and avoidance learning, and attenuates age-related NCAM-mediated neuroplastic decline within the rat medial temporal lobe, J. Neurochem. 78 (2001) 704–714. [23] L. Na, M. Wartenberg, H. Nau, J. Hescheler, H. Sauer, Anticonvulsant valproic acid inhibits cardiomyocyte differentiation in embryonic stem cells by increasing intracellular levels of reactive oxygen species, Birth Defects Res. Part A: Clin. Mol. Teratol. 67 (2003) 174–180. [24] H. Nau, R.S. Hauck, K. Ehlers, Valproic acid-induced neural tube defects in mouse and human: aspects of chirality, alternative drug development, pharmacokinetics and possible mechanisms, Pharmacol. Toxicol. 69 (1991) 310–321. [25] H. Nau, W. Loscher, Pharmacologic evaluation of various metabolites and analogs of valproic acid: teratogenic potencies in mice, Fundam. Appl. Toxicol. 6 (1986) 669–676. [26] Okada, H. Kurihara, Y. Aoki, M. Bialer, M. Fujiwara, Amidic modification of valproic acid reduces skeletal teratogenicity in mice, Birth Defect Res. Part B Dev. Reprod. Toxicol. 71 (2004) 47–53. [27] J.G. Omtzigt, F.J. Los, D.E. Grobbee, L. Pijpers, M.G. Jahoda, H. Brandenburg, P.A. Stewart, H.L. Gaillard, E.S. Sachs, J.W. Wladimiroff, The risk of spina bifida aperta after first-trimester exposure to valproate in a prenatal cohort, Neurology 42 (1992) 119–125. [28] M. Radatz, K. Ehlers, B. Yagen, M. Bialer, H. Nau, Valnoctamide, valpromide and valnoctic acid are much less teratogenic in mice than valproic acid, Epilepsy Res. 30 (1998) 41–48. [29] E. Robert, F. Rosa, Valproate and birth defects, Lancet II (1983) 1142. [30] O. Spiegelstein, M. Bialer, M. Radatz, H. Nau, B. Yagen, Enantioselective synthesis and teratogenicity of phenylisopropyl acetamide, a CNS-active chiral amide analogue of valproic acid, Chirality 11 (1999) 645–650. [31] O. Spiegelstein, D.L. Kroetz, R.H. Levy, B. Yagen, S.I. Hurst, M. Levi, A. Haj-Yehia, M. Bialer, Structure activity relationship of human microsomal epoxide hydrolase inhibition by amide and acid analogues of valproic acid, Pharm. Res. 17 (2000) 216–218. [32] C.J. Stodgell, J.L. Ingram, M. O'Bara, B.K. Tistale, H. Nau, P.M. Rodier, Induction of the homeotic gene Hoxal through valproic acid's teratogenic mechanism of action, Neurotox. Teratol. 28 (2006) 617–624. [33] E.A. Swinyard, F.H. Woodhead, General principles, experimental detection, quantification, and evaluation of anticonvulsants, in: D.M. Woodbury, J.K. Penry, C.E. Pippenger (Eds.), Antiepileptic Drugs, Raven Press, New York, 1982, pp. 111–126. [34] W. Tang, J. Palaty, F.S. Abbott, Time course of alpha-fluorinated valproic acid in mouse brain and serum and its effect on synaptosomal gamma-aminobutyric acid levels in comparison to valproic acid, J. Pharmacol. Exp. Ther. 282 (1997) 1163–1172. [35] M. Tohen, S. Grundy, Management of acute mania, J. Clin. Psychiatry 60 (Suppl 5) (1999) 31–34.