点击显示 收起
【关键词】 Cyclopropyl
Valproic acid (VPA) is an effective antiepileptic drug with an additional activity for the treatment of bipolar disorder. It has been assumed that both activities arise from a common target. At the molecular level, VPA targets a number of distinct proteins that are involved in signal transduction. VPA inhibition of inositol synthase reduces the cellular concentration of myo-inositol, an effect common to the mood stabilizers lithium and carbamazepine. VPA inhibition of histone deacetylases activates Wnt signaling via elevated -catenin expression and causes teratogenicity. Given the VPA chemical structure, it may be possible to design VPA derivatives and analogs that modulate specific protein targets but leave the others unaffected. Indeed, it has been shown that some nonteratogenic VPA derivatives retain antiepileptic and inositol signaling effects. In this study, we describe a further set of VPA analogs and derivatives that separate anticonvulsant activity from effects on neuronal growth cone morphology. Lithium, carbamazepine, and VPA induce inositol-dependent spread of neuronal growth cones, providing a cell-based assay that correlates with mood-stabilizing activity. We find that two constitutional isomers of VPA, propylisopropylacetic acid and diisopropylacetic acid, but not their corresponding amides, and N-methyl-2,2,3,3-tetramethyl-cyclopropanecarboaxamide are more effective than VPA in increasing growth cone spreading. We show that these effects are associated with inositol depletion, and not changes in -catenin-mediated Wnt signaling. These results suggest a route to a new generation of central nervous system-active VPA analogs that specifically target bipolar disorder.
Valproic acid (VPA) is an eight-carbon, branched side chain carboxylic acid with many clinical uses. It has both antiepileptic and mood-stabilizing activity, and it also provides effective prophylaxis for neuropathic pain and migraine (Bourgeois, 2002). The relationship between the antiepileptic and mood-stabilizing activity is particularly interesting because it may reveal information about the molecular origins of bipolar mood disorder, a chronic and disabling illness that affects 1% of the population and when untreated carries in 10 to 20% of patients a lifetime risk of suicide (Muller-Oerlinghausen et al., 2002).
Bipolar disorder and epilepsy have some common features, such as their episodic nature and associated kindling phenomena, and this led to the use of the antiepileptic drugs (AEDs) carbamazepine (CBZ) and subsequently, valproic acid (VPA) in the treatment of bipolar disorder (Macdonald and Young, 2002). Although pharmacological treatment proves effective in many patients, relapse rates of bipolar disorder are 73% during 5 or more years, and approximately 30% of epileptic patients are resistant to multiple AEDs (Gitlin et al., 1995; Kwan and Brodie, 2005). Furthermore, comorbidity and severe adverse effects can further complicate treatment in both disorders (Macdonald and Young, 2002). Hence, there is a pressing need for developing better and safer mood-stabilizing agents.
The clinical use of VPA is limited by two rare but potentially life-threatening, adverse effects: teratogenicity and hepatotoxicity (Dreifuss et al., 1987; Tomson and Battino, 2005). A second generation of AEDs has been developed based on the structure of VPA that improve antiepileptic activity but avoid these side effects. These second generation drugs include specific amide and alkyl amide derivatives of VPA and amide derivatives of a cyclopropyl analog of VPA, 2,2,3,3-tetramethylcyclopropanecarboxylic acid (TMCA) (Sobol et al., 2004; Bialer, 2006).
The situation with bipolar drugs is more complex. Here, a major limitation for the development of drugs with improved mood-stabilizing activity is our poor knowledge of the mechanism of treatment for bipolar disorder. In contrast to epilepsy, no animal models in bipolar disorder are universally accepted, and no model is able to exhibit the characteristic mood swings (Machado-Vieira et al., 2004). Although most AEDs have now been investigated for their mood-stabilizing effects, only a few demonstrated clinical efficacy in patients (Macdonald and Young, 2002). This suggests the mechanism of drug action in mood disorder and in epilepsy only partially overlaps.
One approach to identify the targets of VPA in the treatment of bipolar disorder is to seek common effects with other mood stabilizers. Williams et al. (2002) have recently found that, in common to VPA, both lithium and CBZ increase spreading of growth cones from rat dorsal root ganglia (DRG). This effect can be prevented by exogenous myoinositol. Furthermore, lithium and VPA lower in vivo concentrations of inositol and inositol-1,4,5-triphosphate (InsP3) in a range of cell systems. These findings fit the hypothesis originally proposed by Berridge et al. (1989) for lithium that proposes mood stabilizers function by attenuation of an overactive inositol phosphate signal transduction pathway. Use of this neuronal system offers a rapid cell-based assay to investigate and screen potential new mood disorder drugs (Eickholt et al., 2005).
A number of the molecular VPA targets have been identified. VPA inhibits inositol synthase, thereby leading to depletion of the cellular concentration of myo-inositol (Shaltiel et al., 2004). At first, VPA was also proposed to inhibit GSK-3, a protein kinase important for cell signaling. This report has not been substantiated (Phiel et al., 2001, Ryves et al., 2005), and it was further shown that a range of VPA derivatives also had no direct effect on GSK-3 activity (Eickholt et al., 2005). It was subsequently shown that in some cell types VPA mimics the loss of GSK-3 in the Wnt signaling pathway by inhibition of histone deacetylases. This direct inhibition may explain the effect of VPA on gene expression and animal development.
VPA increases transcription of the -catenin gene and mimics the effect of Wnt stimulation (Phiel et al., 2001). Wnt stimulation blocks GSK-3 activity, leading to a build up of -catenin, which in turn binds to TCF family transcription factors and mediated changes in gene expression. Lithium inhibits GSK-3 elevating -catenin. Consequently, VPA and lithium may have similar effects via different molecular mechanisms (Harwood and Agam, 2003). Given the existence of an alternative signaling target in addition to inositol phosphate signaling, it is important to examine which pathway is targeted by VPA-related compounds.
In this report, we address the question of whether the antiepileptic and mood-stabilizing properties of VPA arise from the same molecular targets or represent distinct effects that overlap with the VPA molecule. To do this, we evaluated the effects of a number of CNS-active VPA constitutional isomers, cyclopropyl analog and its amide derivatives on DRG growth cones spreading and on the cellular concentration of InsP3. The compounds were also evaluated for their ability to modulate the canonical Wnt signaling pathway. All tested compounds possess eight carbon atoms in their major chemical template and have previously been shown to possess antiepileptic activity (Isoherranen et al., 2003). We find little correlation between those drugs that have antiepileptic activity and those that cause growth cones spreading and inositol depletion. This suggests that the antiepileptic and mood-stabilizing activities of VPA may arise through different mechanisms and that these results indicate a potential route to specific mood-stabilizing drugs structurally based on VPA analogous and derivatives.
Materials. Solvents and drugs were purchased from Sigma-Aldrich (Rehovot, Israel). VPA was a gift from Teva Pharmaceutical Industries (Petach Tikva, Israel). Analogs and derivatives of VPA were synthesized according to the synthetic procedures described in the following references: propylisopropylacetic acid (PIA) (Bojic et al., 1996), diisopropylacetic acid (DIA) (Haj-Yehia and Bialer, 1990), valnoctic acid (VCA) and valnoctamide (VCD) (Radatz et al., 1998), 2,2,3,3,-tetramethylcyclopropanecarboxamide (TMCD) and N-methyl-2,2,3,3-tetramethylcyclopropanecarboxamide (MTMCD) (Bialer et al., 1996), and 2,2,3,3-tetramethylcyclopropanecarbonylurea (TMCU) (Sobol et al., 2004). All VPA derivatives used in the in vitro experiments were dissolved in water or ethanol to result in stock solutions of 0.2 M.
Dorsal Root Ganglion Explant Culture. Dorsal root ganglia (DRG) were dissected from the spinal cord area of Sprague-Dawley rat pups and cultured individually on polyornithine/laminin-coated coverslips in serum-free medium at 37°C with 5% CO2 (Bottenstein et al., 1979). Media were supplemented with 25 ng/ml mouse nerve growth factor (NGF)-7s (the optimal concentration for growth stimulation was determined for each batch) to promote neuron survival and axon outgrowth (Williams et al., 2002). After allowing attachment for 24 h, the antimitotic agent cytosine -arabinofuranoside (10 µM) was added for 24 h to kill non-neuronal cells. DRG explants were then changed to fresh serum-free media for a further 24 h. After the development of an extensive axonal outgrowth in this 3-day period, drugs were added at 1.5 times the therapeutic level [i.e., 1 mM sodium chloride (control), 2 mM lithium chloride, or 3 mM VPA (with or without 2 mM myoinositol)]. After 48-h exposure, the explants were fixed using 4% paraformaldehyde in phosphate-buffered saline for 20 min at room temperature. Growth cones images were recorded using an inverted microscope, and spread areas were measured using NIH Image analysis software (http://rsb.info.nih.gov/nih-image/). Growth cones were identified by their distinctive morphology, made up of membrane extension beyond the boundary of the extending axon. DRG growth cones were not routinely stained, but antibody staining for tubulin and phalloidin staining for F-actin demonstrated that the delineation between axonal microtubule and F-actin exactly coincided with the morphology of the measured growth cone (data not shown). Data were collected and analyzed using a double-blind protocol. Between 10 and 50 growth cones were formed per DRG. Drug concentrations that reduced growth cone numbers below this number were considered toxic, and the data were rejected. Each experiment was made up of one to two DRGs and was repeated in triplicate; giving between 50 and 150 growth cone measurements per drug. In the experiments described here, the dose of NGF caused a generally higher degree of growth cone spreading; hence, sizes are slightly higher than previously reported. All chemicals were supplied by Sigma Chemical (Poole, Dorset, UK) unless otherwise stated.
Dictyostelium Cell Culture and InsP3 Assays. Wild-type Dictyostelium discoideum cells (Ax2) were grown at 22°C for 20 h in AX media at 1 x 106 cells/ml in the presence of drugs at indicated concentrations or with vehicle-only control. Cells were washed in the presence of the drug and lysed in perchloric acid. InsP3 levels were measured by isotope dilution binding assay (Amersham Biosciences UK, Ltd., Little Chalfont, Buckinghamshire, UK) as described previously (Williams et al., 1999). Protein was measured by Bradford assay (Bio-Rad, Hemel Hempstead, UK).
TCF Promoter Activity. Activation of TCF promoter activity was monitored using the TOPflash assay (Fahnert et al., 2004). Human embryonic kidney epithelial cells transfected with TCF-luciferase reporter plasmid were seeded at 1 x 104 cells per well of 96-well black-walled dishes in 50 µl of DMEM and incubated at 37°C for 48 h. On the third day 30 µl of DMEM was added to each well, and the drugs, dissolved in pure ethanol, were added in triplicates to give a final concentration of 3 mM. The luminescence assay was performed by removing the medium and washing the cells twice with phosphate-buffered saline. The cells were incubated with the drugs for 48 h, and the luminescence was read after 5-min incubation with a luminescence substrate. For control wells, 10 µl of 1% ethanol in DMEM was added.
Statistical Analysis. The p values were calculated by analysis of variance with Dunn's multiple comparison test for post hoc pairwise comparison with the control value. All statistical analyses were performed with GraphPad InStat, version 3.01 (GraphPad Software Inc., San Diego, CA). Data are expressed as mean ± S.D. A p value of 0.05 was considered significant.
Fig. 1. Chemical structures of the tested compounds: VPA, 4-ene-VPA, 4-yne-VPA, DIA, PIA, VCA, VCD, VPD, DID, PID, TMCA, TMCD, MT-MCD, and TMCU.
Effect of VPA Constitutional Isomers and Their Amide Derivatives on the Morphology of DRG Neuronal Growth Cones. We used a modified version of the DRG cell-based assay described by Williams et al. (2002) to compare the action of VPA analogs. In this assay, DRG excised from newborn rats are cultured as explants for 72 h in NGF-supplemented medium. Under these conditions, explants project axonal outgrowths that extend through migration of their growth cones. Cultures were treated with drugs for a further 48 h and then fixed and observed under phase contrast microscopy. Growth cone morphologies were scored by measuring the spread area of each growth cone and analyzed to give a numerical value. The advantage of this modified method over that reported (Williams et al., 2002) is that we were able to collect data from all growth cone morphologies. This complete range of size measurements made possible a statistical analysis.
Lithium chloride was used as a positive control, and all treatments were compared with drug-free, solvent controls. The effects of VPA were compared with the VPA constitutional isomers VCA, DIA, and PIA, three constitutional isomers of VPA with modified aliphatic moieties, and their amide derivatives (Fig. 1). A substantial change in the distribution in growth cone sizes was observed following treatment with 2 mM lithium chloride or 3 mM VPA (Fig. 2), with cultures showing fewer small growth cones and an increased proportion of growth cones with large spread areas. Very little effect was seen in cultures treated with 1 mM VPA.
Fig. 2. Effects of aliphatic VPA constitutional isomers on rat DRGs growth cone spreading. These compounds demonstrate low anticonvulsant potency in animal models of epilepsy (Table 1) (Haj-Yehia and Bialer, 1990). A histogram shows the percentage of growth cones with spread areas within 50-µm2 class intervals. The abbreviations for the various compounds are given in Fig. 1. a, untreated control experiment, making up 119 growth cones derived from nine DRG explants, where b denotes where a small growth cone would be seen and c denotes where a spread growth cone would be. b, differential interference contrast image of a small growth cone (64x). c, differential interference contrast image of a spread growth cone (64x). d, 2 mM LiCl-treated experiment, making up 79 growth cones derived from three DRG explants. e, 3 mM VPA-treated experiment, consisting of 146 growth cones derived from six DRG explants. f, 1 mM VPA-treated experiment, showing a suboptimal dose for growth cone spreading (110 growth cones from six DRG explants). g, 3 mM VCA-treated experiment, composed of 89 growth cones derived from four DRG explants. h, 1 mM VCA-treated experiment, comprised of 124 growth cones derived from four DRG explants. i, 1 mM VCA treatment with 2 mM myo-inositol. Data comprises 106 growth cones derived from three DRG explants. j, 1 mM DIA treatment experiment, made up of 75 growth cones derived from three DRG explants. k, 0.5 mM DIA treatment experiment, made up of 61 growth cones derived from four DRG explants. l, 1 mM DIA treatment with 2 mM myo-inositol. Data comprises 60 growth cones derived from three DRG explants. m, 1 mM PIA-treated experiment, made up of 99 growth cones derived from four DRG explants. n, 0.5 mM PIA-treated experiment, made up of 73 growth cones derived from three DRG explants. o, 1 mM PIA treatment with 2 mM myo-inositol. Data represent 61 growth cones derived from three DRG explants. Addition of myo-inositol reverses the effect of DIA- and PIA-induced growth cone spreading. Addition of myoinositol also reduces the median growth cone size of VCA-treated experiment significantly (p < 0.05). Total data of (a and d-o) was analyzed by a Kruskal-Wallis test, followed by Dunn's multiple comparison test of each experiment versus control: *, p < 0.05; **, p < 0.01; and ***, p < 0.001.
PIA and DIA strongly increased growth cone spreading at 1 mM concentrations (Fig. 2). VCA gave a strong effect at 3 mM, but a weak effect at 1 mM. The stronger effect of PIA and DIA than observed for 1 mM VPA suggested a higher potency of these isomers. To examine this further, we tested both drugs at 0.5 mM. PIA again exhibited a strong increase in growth cone spread area; however, no effect was observed for DIA. These results suggested an order of potency with PIA > DIA > VCA.
TABLE 1 Activity of VPA derivatives and analogs as anticonvulsants (MES test in mice; i.p.) and their effects on the median growth cone area from rat dorsal root ganglia
The abbreviations for the various compounds are given in Fig. 1.
The common effects of mood stabilizers on growth cone size are reversed by addition of myo-inositol to the medium, consistent with the proposed mechanism via inositol depletion (Williams et al., 2002). Inositol reversal has previously been used to identify constitutional isomers of VPA that possess inositol-depleting activity (Eickholt et al., 2005). In all cases, the effects of PIA, DIA, and VCA can be reversed by addition of myo-inositol to the medium (Fig. 2).
Valpromide (VPD), the corresponding amide of VPA, has antiepileptic activity but does not increase growth cone spreading (Shaltiel et al., 2004). We therefore tested the corresponding amides of the constitutional VPA isomers (PIA, DIA, and VCA), which have been shown to have significant antiepileptic activity (Isoherranen et al., 2003; Table 1). Again, these compounds have no effect on growth cone spreading at 1 or 3 mM (Fig. 3). Finally, we tested two unsaturated VPA derivatives 4-ene-valproic acid (4-ene-VPA) and 4-yne-valproic acid (4-yne-VPA). These compounds were toxic to the DRG at 3 mM, but they increased spreading at 1 mM (data not shown).
Fig. 3. Effects of VPA and aliphatic VPD constitutional isomers on rat DRGs growth cones spreading. The VPD constitutional isomers exhibit potent anticonvulsant activity in animal models of epilepsy (Table 1) (Haj-Yehia and Bialer, 1990; Winkler et al., 2005b). A histogram shows the percentage of growth cones with spread areas within 50-µm2 class intervals. The abbreviations for the various compounds are given in Fig. 1. a, untreated control experiment, as shown in Fig. 2. b, 3 mM VPA-treated experiment, as shown in Fig. 2. c, 3 mM VPD-treated experiment, made up of 69 growth cones derived from three DRG explants. d, 3 mM DID-treated experiment, made up of 110 growth cones derive from three DRG explants. e, 3 mM PID-treated experiment, made up of 75 growth cones derived from four DRG explants. f, 3 mM VCD-treated experiment, made up of 60 growth cones derived from three DRG explants. Total data of a-f were analyzed by a Kruskal-Wallis test, followed by Dunn's multiple comparison test of each experiment versus control: ***, p < 0.001.
Effect of Cyclopropyl VPA Analog and Its Amide Derivatives on the Morphology of DRG Neuronal Growth Cones. The cyclopropyl VPA analog TMCA and its amide derivatives TMCD, MTMCD, and TMCU all possess anticonvulsant activity in rodent models (Isoherranen et al., 2003; Sobol et al., 2004; Winkler et al., 2005b). Neither TMCA, TMCD, nor TMCU showed an effect on growth cone morphology at concentrations of 3 mM, and they were toxic at higher concentrations (Fig. 3). In contrast, the N-methylamide of TMCA (MTMCD) was more potent than VPA, causing a substantial increase in growth cone area at 1 mM, a result consistent with previous observations (Shaltiel et al., 2004). We observed a strong effect on growth cones when the concentration of MTMCD was lowered further to 0.5 mM. We find that addition of 2 mM myo-inositol to the medium of growth cones treated with 1 mM MTMCD (Fig. 4) reverses its effect, consistent with the mode of action of VPA and suggests an inositol depletion mechanism.
Fig. 4. Effects of VPA analog TMCA and its amide derivatives on rat DRGs growth cones spreading. The amide derivatives of TMCA demonstrate potent anticonvulsant activity in animal models of epilepsy (Table 1), whereas TMCA shows only weak antiseizure activity. A histogram shows the percentage of growth cones with spread areas within 50-µm2 class intervals. The abbreviations for the various compounds are given in Fig. 1. Untreated control experiment as well as 3 mM VPA and 2 mM LiCl-treated experiments as in Fig. 2. a, 3 mM TMCA-treated experiment, made up of 97 growth cones derived from four DRG explants. b, 3 mM TMCD-treated experiments, made up of 63 growth cones derived from three DRG explants. c, 1 mM TMCU-treated experiment, made up of 73 growth cones derived from four DRG explants. d, 0.5 mM MTMCD-treated experiment, made up of 50 growth cones derived from two DRG explants. e, 1 mM MTMCD-treated experiment, made up of 79 growth cones derived from four DRG explants. f, 1 mM MTMCD treatment with 2 mM myo-inositol. Data represents 60 growth cones derived from three DRG explants. Addition of myo-inositol reverses the effect of MTMCD-induced growth cone spreading. Total data of a to f were analyzed by a Kruskal-Wallis test, followed by Dunn's multiple comparison test of each experiment versus control: *, p < 0.05; **, p < 0.01; and ***, p < 0.001.
Effects of VPA Isomers and Derivatives on InsP3. In all cases, the increase in growth cone area is reversed by addition of myo-inositol, indicative of an inositol depletion mechanism. It is not possible to directly measure inositol or inositol phosphate concentrations in DRG explants, so we examined InsP3 in an alternative system, Dictyostelium amoeba. This has previously proven useful to measure inositol depletion caused by VPA and other related molecules (Williams et al., 2002; Eickolt et al., 2005). To test the effects of the VPA derivatives described above on inositol-based signaling, we measured their effects on InsP3 concentrations in Dictyostelium. For technical reasons cells were treated at 0.5 mM for 20 h. PIA, the most potent branched constitutional isomer, caused a significant decrease in InsP3 (Fig. 5). Propylisopropylacetamide (PID), the corresponding amide of PIA, had no effect on InsP3 concentration, an observation consistent with its lack of effect on growth cone spreading. All of the other compounds tested [DIA, VCA, DID, TMCA, TMCD, TMCU, MTMCD, and 4-ene-VPA), except 4-yne-VPA, had no significant effect on InsP3 (Fig. 5).
Fig. 5. Effects of VPA derivatives, its analogs, and their amides on Dictyostelium InsP3 levels. Dictyostelium discoideum cells were grown overnight in complete medium in the presence of VPA or one of the tested compounds at 0.5 mM or with lithium at 2 mM. Changes in InsP3 levels were subsequently measured by isotope dilution assay and compared with vehicle only. Comparison of relative efficacy of VPA derivatives and their analogs to cause InsP3 depletion expressed as a percentage of that found for control. Results represent three experiments assayed in triplicate ± S.D; *, p < 0.05.
The Effect of VPA Derivatives and Analog on the Wnt Signal Transduction Pathway. Given the potential for VPA to stimulate TCF transcriptional activity via histone deacetylase inhibition (Phiel et al., 2001), we examined whether the compounds identified above stimulate this pathway. We used a TCF-luciferase-based reporter assay in human embryonic kidney 293 cells (a TOPflash assay) to examine the effects of each VPA isomer, analog, and their derivatives. As a control, we used lithium to stabilize -catenin protein via inhibition of GSK-3. This caused a significant increase in TCF-mediated gene expression as measured by luminescence at 3 mM, a concentration only slightly higher than its Ki of 2 mM against GSK-3 (Fig. 6). VPA treatment (3 mM) induces a significant but smaller induction of luciferase gene expression. At this concentration, none of the three VPA isomers (i.e., neither TMCA nor its amide derivatives) caused a significant increase in luminescence compared with an untreated control.
Fig. 6. Screening for Wnt pathway modulators. Cells were grown for 48 h in complete medium in the presence of lithium, VPA, or one of its derivatives and analogs at 3 mM. Luminescence was read after 5-min incubation with luminescence substrate. The histograms show the relative light unites for the corresponding drugs assayed in triplicates (mean ± S.D). The abbreviations for the various compounds are given in Fig. 1.
These experiments map the structural elements of the VPA molecule that affect neuronal growth cone behavior and inositol depletion. We show that structurally different groups of VPA-related compounds cause growth cone spreading. First, there is a class of branched constitutional isomers of VPA, which include PIA, DIA, and VCA. The most potent isomer was PIA, which caused growth cone spreading at 0.5 mM, a significantly lower concentration than seen with VPA. Second, we found that one of the amide derivatives of a cyclopropyl analog of VPA, MTMCD, had activity on DRG growth cones. This was previously noted by Shaltiel et al. (2004), but here we show that this property is unique among the cyclopropyl analogs that have been tested
The effects of all compounds that increased growth cone spreading were reversed by addition of myoinositol, consistent with the inositol depletion mechanism reported for VPA. The inositol reversal of MTMCD reported here is also consistent with the strong inhibition of inositol synthase activity reported by Shaltiel et al. (2004). Finally, we show that PIA reduces InsP3 in the Dictyostelium-based assay, again consistent with a mechanism of inositol depletion. It is unclear why the other branched constitutional isomers DIA, VCA, and the VPA amide analog MTMCD failed to significantly reduce InsP3 in this assay. This may arise from differences in cellular uptake, pharmacokinetics, or inositol synthase structure; furthermore, many of the compounds exhibited toxic effects on Dictyostelium above 0.5 mM.
We therefore have found two independent structural determinants, modification of the branching at C-2 of VPA and the derivatization by methylamine of the carboxylic group of TMCA, that cause growth cone spreading via an inositol depletion mechanism. These can be added to those reported in the earlier study of Eickholt et al. (2005), who also identified another branched constitutional isomer of VPA, 2-ethyl-4-methylpentanoic acid, that possesses growth cone spreading and inositol-depleting activity. This earlier study also suggested a third structural feature of unsaturated C-C bonds, as seen for the VPA analogs 2-methyl-2-pentenoic acid, (S)-2-pentyl-4-pentynoic acid, and (S)-2-propyl-4-hexynoic acid, which contribute to inositol depletion; however, these molecules were also teratogenic (Eickholt et al., 2005). Consistent with this structural feature, we found that the two unsaturated VPA derivatives 4-ene-VPA and 4-yne-VPA also caused growth cone spreading, and 4-yne-VPA lowered InsP3 in Dictyostelium (Fig. 5).
It is noteworthy that these structural determinants show little overlap with those of VPA derivatives and analogs possessing potent anticonvulsant activity (Table 1). The corresponding amide of VPA, VPD (Shaltiel et al., 2004), and its CNS-active constitutional isomers VCD, PID, and DID (Isoherranen et al., 2003; Winkler et al., 2005a) had no effect on growth cone morphology. In contrast, PIA, DIA, and VCA, the constitutional isomers of VPA that increase growth cone spreading have higher MES-ED50 value in mice compared with their corresponding amides (Table 1). There is also little correlation between growth cone spreading and MES-ED50 of TMCA, the cyclopropyl analog of VPA, and its amide derivatives. All have lower (more potent) MES-ED50 values than VPA and that of VPA is >16 times higher than that measured for TMCU; however, only MTMCD caused growth cone spreading in these and previous experiments (Shaltiel et al., 2004).
Due to similarities between bipolar disorder and epilepsy, it has been proposed that they could have a common origin. Consistent with this hypothesis, the AEDs VPA and CBZ are effective mood stabilizers. New AEDs are now routinely evaluated for mood-stabilizing properties (Gajwani et al., 2005). The results presented here show no significant correlation between MES-ED50 values and growth cone spreading at therapeutically used concentrations (Table 1). One caveat is that amides of several VPA analogs can reach higher concentrations in the brain than their corresponding acids (Blotnik et al., 1996); thus, we cannot exclude the possibility that the amides of several VPA analogs tested by us (Fig. 1) achieve higher brain concentrations than could be tested in these cell-based assays. Taken together, these results demonstrate that anticonvulsant activity and growth cone effects are likely to arise from different molecular targets that interact by distinct pharmacophore molecular conformations of VPA and its isomers. Given the correlation between increased growth cone spreading and other mood stabilizers, we suggest by extrapolation that the anticonvulsant and moodstabilizing properties of VPA may arise through different molecular mechanisms.
Because lithium inhibits GSK-3, a protein kinase central component of the Wnt signaling pathway, it has been proposed that this signaling pathway may be a target of mood stabilizer drugs. Indeed, the action of lithium shows many similarities to Wnt stimulation on regenerating neurons (Lucas et al., 1998). Furthermore, because VPA has been shown to elevate expression of -catenin and hence mimic the effect of lithium on the Wnt pathway, it has been proposed that VPA may also act on the Wnt signaling pathway (Phiel et al., 2001). To test the possibility that the VPA analogs could elicit a Wnt signaling response, we looked at their ability to activate TCF promoter activity, the nuclear target of -catenin. We found no stimulation of TCF promoter activity of any of the VPA constitutional isomers and derivatives used in this study. These results further indicate that VPA can cause growth cone spreading independent of Wnt pathway activation, an activity also seen with CBZ (Williams et al., 2002; Ryves et al., 2005). GSK-3 has also been associated with rapid changes in growth cone structure during axon guidance. Here, inhibition of GSK-3 blocks Sem3A-induced growth cone collapse (Eickholt et al., 2002). Although unrelated to the changes of growth cone morphology examined in this study, we can exclude a similar mechanism of VPA action, because a reduction in cellular inositol would reduce cellular concentrations of phosphatidyl-(3,4,5)-trisphosphate and lead to elevated GSK-3 activity; the opposite effect to that seen in the Sem3A collapse studies.
In conclusion, we have shown that structural modifications of the alkyl side chain moieties on the VPA core structure, to form its constitutional isomers PIA and DIA, potentiate the activity on the DRG. The structural-activity relationships described here indicate the different pharmacological activity seen with the VPA molecule and its constitutional isomers, may relate to different structural requirements or conformations adopted by the side chain moieties of these isomers in their interactions with cellular VPA targets. In addition, MTMCD, an amide derivative of cyclopropyl VPA analog, was the only amide analog demonstrating growth cone spreading effect, suggesting further investigation in elucidating the exact structural requirements for amides of VPA analogs for the growth cone spreading effect. Our results provide new strategies for future design of second generation to VPA CNS-active drugs with potential mood-stabilizing activities, based on structural modifications of the alkyl side chain moieties of the VPA core structure.
Acknowledgements
This work is derived from the Ph.D. thesis of J.A.S. in partial fulfillment of the Ph.D. degree requirements for The Hebrew University of Jerusalem.
ABBREVIATIONS: AED, antiepileptic drug; CBZ, carbamazepine; VPA, valproic acid; TMCA, 2,2,3,3-tetramethylcyclopropanecarboxylic acid; DRG, dorsal root ganglia; InsP3, inositol-1,4,5-trisphosphate; GSK, glycogen synthase kinase; CNS, central nervous system; PIA, propylisopropylacetic acid; DIA, diisopropylacetic acid; VCA, valnoctic acid; VCD, valnoctamide; VPD, valpromide; DID, diisopropylacetamide; PID, propylisopropylacetamide; TMCD, 2,2,3,3-tetramethylcyclopropanecarboxamide; MTMCD, N-methyl-2,2,3,3,-tetramethylcyclopropane carboxamide; TMCU, 2,2,3,3-tetramethylcyclopropanecarbonylurea; 4-ene-VPA, 4-ene-valproic acid; 4-yne-VPA, 4-yne-valproic acid; NGF, nerve growth factor; MES, maximal electroshock; TCF, T cell factor.
【参考文献】
Berridge MJ, Downes CP, and Hanley MR (1989) Neural and developmental actions of lithium: a unifying hypothesis. Cell 59: 411-419.
Bialer M (2006) New antiepileptic drugs that are second generation to existing antiepileptic drugs. Exeprt Opin Investig Drug 15: 637-647.
Bialer M, Hadad S, Kadry B, Abdul-Hai A, Haj-Yehia A, Sterling J, Herzig Y, and Yagen B (1996) Pharmacokinetic analysis and antiepileptic activity of tetramethylcyclopropane analogues of valpromide. Pharm Res (NY) 13: 284-289.
Blotnik S, Bergman F, and Bialer M (1996) Disposition of valpromide, valproic acid, and valnoctamide in the brain, liver, plasma, and urine of rats. Drug Metab Dispos 24: 560-564.
Bojic U, Elmazar MM, Hauck RS, and Nau H (1996) Further branching of valproate-related carboxylic acids reduces the teratogenic activity, but not the anticonvulsant effect. Chem Res Toxicol 9: 866-870.
Bottenstein J, Hayashi I, Hutchings S, Masui H, Mather J, McClure DB, Ohasa S, Rizzino A, Sato G, Serrero G, et al. (1979) The growth of cells in serum-free hormone-supplemented media. Methods Enzymol 58: 94-109.
Bourgeois BFD (2002). Valproic acid-clinical efficacy and use in epilepsy, in Antiepileptic Drugs (Levy RH, Mattson RH, Meldrum BS, and Perucca E eds) pp 808-817, Lippincot Williams & Wilkins, Philadelphia, PA.
Dreifuss FE, Santilli N, Langer DH, Sweeney KP, Moline KA, and Menaander KB (1987) Valproic acid hepatic fatalities: a retrospective review. Neurology 37: 379-395.[Abstract/Free Full Text]
Eickolt BJ, Towers GJ, Ryves WJ, Eikel D, Adley K, Ylinen LMJ, Chadborn NH, Harwood AJ, Nau H, and Williams RSB (2005) Effects of valproic acid derivatives on inositol trisphosphate depletion, teratogenicity, glycogen synthase kinase-3 inhibition, and vial replication: a screening approach for new bipolar disorder drugs derived from the valproic acid core structure. Mol Pharmacol 67: 1426-1433.[Abstract/Free Full Text]
Eickholt BJ, Walsh FS, and Doherty P (2002) An inactive pool of GSK-3 at the leading edge of growth cones is implicated in Semaphorin 3A signaling. J Cell Biol 157: 211-217.[Abstract/Free Full Text]
Fahnert B, Veijola J, Roel G, Karkkainen MK, Railo A, Destree O, Vainio S, and Neubauer P (2004) Murine Wnt-1 with an internal c-myc tag recombinantly produced in Escherichia coli can induce intracellular signaling of the canonical Wnt pathway in eukaryotic cells. J Biol Chem 276: 47520-47527.
Gajwani P, Forsthoff A, Muzina D, Amann B, Gao K, Elhaj O, Calabrese JR, and Grunze H (2005). Antiepileptic drugs in mood-disordered patients. Epilepsia 46 (Suppl 4): 38-44.
Gitlin MJ, Swendsen J, Heller TL, and Hammen C (1995) Relapse and impairment in bipolar disorder. Am J Psychiatry 152: 1635-1640.[Abstract/Free Full Text]
Haj-Yehia A and Bialer M (1990) Structure-pharmacokinetic relationships in a series of short fatty acid amides that possess anticonvulsant activity. J Pharm Sci 79: 719-724.
Harwood AJ and Agam G (2003) Search for a common mechanism of mood stabilizers. Biochem Pharmacol 66: 179-189.
Isoherranen N, Yagen B, and Bialer M (2003) New CNS-active drugs which are second-generation valproic acid. Can they lead to the development of a magic bullet? Curr Opin Neurol 16: 203-211.
Kwan P and Brodie MJ (2005) Potential role of drug transporters in the pathogenesis of medically intractable epilepsy. Epilepsia 46: 224-235.
Loscher W and Nau H (1985) Pharmacological evaluation of various metabolites and analogues of valproic acid. Anticonvulsant and toxic potencies in mice. Neuropharmacology 24: 427-435.
Lucas FR, Goold RG, Goordon-Weeks PR, and Salinas PC (1998) Inhibition of GSK-3beta leading to the loss of phosphorylated MAP-1B is an early event in axonal remodeling induced by WNT-7a or lithium. J Cell Sci 111: 1351-1361.
Macdonald KJ and Young LT (2002) Newer antiepileptic drugs in bipolar disorder: rationale for use and role in therapy. CNS Drugs 16: 549-562.
Machado-Vieira R, Kapczinski F, and Soares JC (2004) Perspectives of the development of animal models of bipolar disorder. Prog Neuropsychopharmacol Biol Psychiatry 28: 209-224.
Muller-Oerlinghausen B, Berghofer A, and Bauer M (2002) Bipolar disorder. Lancet 359: 241-247.
Phiel CJ, Zhang F, Huang EY, Guenther MG, Lazar MA, and Klein PS (2001) Histone deacetylase is a direct target of valproic acid, a potent anticonvulsant, mood stabilizer, and teratogen. J Biol Chem 276: 36734-36741.[Abstract/Free Full Text]
Radatz M, Ehlers K, Yagen B, Bialer M, and Nau H (1998) Valnoctamide, valpromide and valnoctic acid are much less teratogenic in mice than valproic acid, Epilepsy Res 30: 41-48.
Ryves JW, Dalton EC, Harwood AJ, and Williams RS (2005) GSK-3 activity in neocortical cells is inhibited by lithium but not carbamazepine or valproic acid. Bipolar Disord 7: 260-265.
Shaltiel G, Shamir A, Shapiro J, Ding D, Dalton E, Bialer M, Harwood AJ, Belmaker RH, Greenberg ML, and Agam G (2004) Valproate decreases inositol biosynthesis. Biol Psychiatry 56: 868-874.
Sobol E, Bialer M, and Yagen B (2004) Tetramethylcyclopropyl analogues of a leading antiepileptic drug-valproic acid. Synthesis and evaluation of the anticonvulsant activity of its amide derivatives. J Med Chem 47: 4316-4326.
Spiegelstein O, Yagen B, Levy RH, Finnell RH, Bennett GD, Rowder M, Schurig V, and Bialer M (1999) Stereoselective pharmacokinetics and pharmacodynamics of propylisopropyl acetamide, a CNS-active chiral amide analog of valproic acid. Pharm Res 16: 1582-1588.
Tomson T and Battino D (2005) Teratogenicity of antiepileptic drugs: state of the art. Curr Opin Neurol 18: 135-140.
White HS, Woodhead JH, Wilcox KS, Stables JP, Kupferberg HJ, and Wolf HH (2002). Clinical efficacy and use in other neurological disorders, in Antiepileptic Drugs (Levy RH, Mattson RH, Meldrum BS, and Perucca E eds) pp 36-48, Lippincot Williams & Wilkins, Philadelphia, PA.
Williams RS, Cheng L, Mudge AW, and Harwood AJ (2002) A common mechanism of action for three mood-stabilizing drugs. Nature (Lond) 417: 292-295.
Williams RS, Eames, M, Ryves WJ, Viggars J, and Harwood AJ (1999) Loss of prolyl oligopeptidase confers resistance to lithium by elevation of inositol (1,4,5)trisphosphate. EMBO (Eur Mol Biol Organ) J 18: 2734-2745
Winkler I, Blotnik S, Shimshoni J, Yagen B, Devor D, and Bialer M (2005a) Efficacy of antiepileptic isomers of valproic acid and valpromide in a rat model of neuropathic pain. Br J Pharmacol 1 46: 198-208.
Winkler I, Sobol E, Yagen B, Steinman A, Devor M, and Bialer M (2005b) Efficacy of antiepileptic tetramethylcyclopropyl analogs of valproic acid amides in a rat model for neuropathic pain. Neuropharmacology 49: 1110-1120.
作者单位：Department of Pharmaceutics, School of Pharmacy, Faculty of Medicine, Hebrew University of Jerusalem, Jerusalem, Israel (J.A.S., S.E., N.P., M.B.); School of Biosciences, Cardiff University, Cardiff, United Kingdom (E.C.D., K.E., A.J.H.); Department of Biology, University College London, London, Uni