@Copyright 1986 by The Humana Press Inc. All rights of any nature whatsoever reserved. 0163-4984/86/1001-0025502.20
Effect of Valproate on Zinc Metabolism in Fetal and Maternal Rats Fed Normal and Zinc-Deficient Diets J/~IRGEN VORMANN,1 VERA HOLLRIEGL, ~ HANS-JOACHIM MERKER, 2 AND THEODOR GCINTHER*'1
tinstitute of Molecular Biology and Biochemistry, Free University of Berlin, Amimallee 22, D-1000 Berlin 33, FRG; and 2Institute of Toxicology and Prenatal Pharmacology, Free University of Berlin, Garystr. 5, D-1000 Berlin 33, FRG Received October 21, 1985; Accepted December 31, 1985
ABSTRACT Pregnant Wistar rats fed control and Zn-deficient diets received daily oral doses of 0, 100, and 300 mg/kg sodium valproate from d 16 to 20 of gestation. Only the highest valproate doses induced a small reduction in fetal body weight in the normally fed group. Zinc deficiency caused a drastic reduction in maternal and only a small reduction in fetal serum Zn concentrations. Valproate treatment had no effect on maternal and fetal serum Zn concentrations. Valproate reduced fetal liver Zn content only in the normally fed group. The reduction of liver Zn content resulted from the reduction of Zn-metallothionein. Valproate did not affect total Zn and Znmetallothionein in kidneys. Three percent of the Zn-deficient fetuses developed hydronephrosis and hydrops. Valproate treatment drastically enhanced the occurrence of fetal hydronephrosis and hydrops. Valproate induced fetal liver necroses, independent of Zn nutrition. Index Entries- Zn-metaltothionein; effect of valproate and Zn deficiency on hydronephrosis; induction by Zn deficiency and valproate; hydrops, induction by Zn deficiency and valproate; fetal liver; fetal kidney. *Author to whom all correspondence and reprint requests should be addressed.
Biological Trace Element Research
25
Vol. 10, 1986
26
Vormann et aL
INTRODUCTION Va|proic acid, an anticonvulsant drug, can induce various malformations, preponderantly defects of the neural tube and other toxic effects, e.g., hepatotoxicity (1,4). Zinc deficiency produces similar malformations (5). Several side effects of valproic acid are similar to those observed after zinc or selenium deficiency, and administration of valproic acid to adult rats reduces plasma Zn and Se concentration and liver Se content (6,7). Therefore, an interaction of Zn and valproate or its metabolites are suggested. However, in adult rats, hepatotoxicity of valproic acid does not appear to be caused by Zn deficiency, but probably by an interaction with coenzyme A (CoASH) (8). The teratogenic effects of valproate are as yet unexplained. Therefore, we investigated fetal and maternal Zn metabolism of normal and Zn deficient rats treated with valproate. As in similar experiments with salicylate (9), the drug was given from d 16 to 20 of gestation, when most of the fetal rat body mass is synthesized.
MATERIALS AND METHODS Pregnant Wistar rats, weighing 200 g, were fed a control or Zndeficient diet and distilled water ad libitum from d 0 to 21 of gestation. For preparing the diets, a commercial diet (Ssniff, Soest, FRG) without ZnCO3 and MgSO4 in the mineral mix was used. Casein, separately received, was extracted with 1% EDTA (disodium salt) for 4 h and 0.5% EDTA for 3 h with stirring, and washed twice with distilled water by suction. The Zn-deficient diet was supplemented with MgC12 to 1000 ppm Mg. Its Zn content was measured after ashing by atomic absorption spectrophotometry and amounted to 1.5 ___ 0.2 ppm Zn. Control rats were fed the same diet supplemented with ZnC12 to 100 ppm Zn. Food intake by the control and Zn-deficient rats was measured daily. [For more details see ref. (10).] From d 16 to 20, the rats in both dietary groups received daily oral doses of 0, 100, or 300 mg/kg sodium valproate. Other authors (6,7) gave 300 (6) or up to 750 mg/kg (7) valproic acid (VPA) for 7 d to nonpregnant rats. Thus, our doses were somewhat lower, although the elimination of VPA is increased during pregnancy (11). Daily therapeutic doses of VPA are lower (4) than those applied in experiments with rats. However, the elimination rate of VPA in rats is higher than in humans because of the lower binding of VPA to rat serum proteins (12). At d 21, the uteri were dissected under nembutal anesthesia (50 mg/kg sc). Blood was taken from the fetal hearts by means of heparinized capillaries and from the mothers by heart puncture and was centrifuged for 5 rain at 1300g. Fetal and maternal livers were removed and cleaned in cold 10% sucrose. Sections (0.5 g) of livers were used for preparing a Biological Trace Element Research
Vot I O, 1986
Effect of Valproate on Zn Metabolism
27
10% homogenate and a 100,000g supernatant, according to Thiers and Vallee (13). Protein content in the supernatants was determined according to Lowry et al. (14). The rest of the livers was frozen in liquid nitrogen, freeze-dried, and pulverized in a plastic mortar. Freeze-dried, powdered liver was ashed overnight in the PlasmaProcessor 200-E (Technics, GmbH, Munich, FRG). The ash was dissolved in 0.1N HC1. Aliquots were taken for measurement of Na and K by flame photometry (Beckman), and for measurement of Mg and Zn by atomic absorption spectrophotometry (Perkin-Elmer, model 303). In fetal and maternal serum, Na, K, Mg, and Zn were measured using the same methods. Ultrapure water (Millipore) was used for dilution and calibration curves. Glycogen of all livers was measured by hydrolysis with amyloglucosidase, and the liberated glucose was determined in an optical test (15). For determination of Zn-metallothionein (Zn-MT), 2 mL of the 100,000g supernatant of maternalor pooled fetal liver were separated on a column 90 x 1.5 cm (Pharmacia, Uppsala, Sweden) filled with Sephadex G-75. The column was eluted with a solution containing 0.1M NaCI, 50 mM TrisC1, ph 7.4, and 0.02% NAN3, at 20~ The flow rate amounted to 15 mL/h. Fractions of 2 mL were sampled (Fraction Collector Frac-300, Pharmacia) for measurement of Zn by atomic absorption spectrophotometry and for identification of metallothionein by UV absorption. Calibration of the columns was performed using the Gel Filtration Calibration Kit, according to the instructions of the producer (Pharmacia). Fetal livers and kidneys from all groups were fixed in 3.5% formaldehyde, embedded in paraplast, sectioned, and stained with hematoxylin and eosin.
RESULTS General Observations Under our experimental conditions (up to d 20), there was no significant difference in food intake between the Zn-deficient and normally fed groups, as was found in preceeding papers (9,10), in which the effects of single Zn deficiency were described in detail. Body weights of the Zn-deficient dams and fetuses were 20 and 25% lower compared to the normally fed controls. Only the highest valproate doses caused a small reduction in body weight of the fetuses from normally fed dams (Table 1). The resorption rate in Zn deficiency was 10%, as found in other experiments (10). Valproate application had no effect on the resorption rate. In the Zn-deficient groups treated with 0, 100, or 300 mg/kg valproate, 3, 10, or 40% of the fetuses developed hydrops, whereas in the normally fed rats, only 6% of the fetuses treated with 300 mg/kg Biological Trace Element Research
Vol. I O, 1986
28
Vormann et aL
{',1
~,
oo +i +i +I
V
+1 +1 +1 +1
I
N
{N
8 rr
v
o
~c3o
~; o
oo +1 +1 +l
r,.,)
t~j"
+1 +1 +1 §
mm
m
V
t-xl t,-,l ~-.~ tr t'N . ~ rr c-,i ~ o~
+1 +1 +1 +1
+
a
m
B N
N
~'~ N
Effect of Valproate on Zinc Metabolism in Fetal and Maternal Rats Fed Normal and Zinc-Deficient Diets J/~IRGEN VORMANN,1 VERA HOLLRIEGL, ~ HANS-JOACHIM MERKER, 2 AND THEODOR GCINTHER*'1
tinstitute of Molecular Biology and Biochemistry, Free University of Berlin, Amimallee 22, D-1000 Berlin 33, FRG; and 2Institute of Toxicology and Prenatal Pharmacology, Free University of Berlin, Garystr. 5, D-1000 Berlin 33, FRG Received October 21, 1985; Accepted December 31, 1985
ABSTRACT Pregnant Wistar rats fed control and Zn-deficient diets received daily oral doses of 0, 100, and 300 mg/kg sodium valproate from d 16 to 20 of gestation. Only the highest valproate doses induced a small reduction in fetal body weight in the normally fed group. Zinc deficiency caused a drastic reduction in maternal and only a small reduction in fetal serum Zn concentrations. Valproate treatment had no effect on maternal and fetal serum Zn concentrations. Valproate reduced fetal liver Zn content only in the normally fed group. The reduction of liver Zn content resulted from the reduction of Zn-metallothionein. Valproate did not affect total Zn and Znmetallothionein in kidneys. Three percent of the Zn-deficient fetuses developed hydronephrosis and hydrops. Valproate treatment drastically enhanced the occurrence of fetal hydronephrosis and hydrops. Valproate induced fetal liver necroses, independent of Zn nutrition. Index Entries- Zn-metaltothionein; effect of valproate and Zn deficiency on hydronephrosis; induction by Zn deficiency and valproate; hydrops, induction by Zn deficiency and valproate; fetal liver; fetal kidney. *Author to whom all correspondence and reprint requests should be addressed.
Biological Trace Element Research
25
Vol. 10, 1986
26
Vormann et aL
INTRODUCTION Va|proic acid, an anticonvulsant drug, can induce various malformations, preponderantly defects of the neural tube and other toxic effects, e.g., hepatotoxicity (1,4). Zinc deficiency produces similar malformations (5). Several side effects of valproic acid are similar to those observed after zinc or selenium deficiency, and administration of valproic acid to adult rats reduces plasma Zn and Se concentration and liver Se content (6,7). Therefore, an interaction of Zn and valproate or its metabolites are suggested. However, in adult rats, hepatotoxicity of valproic acid does not appear to be caused by Zn deficiency, but probably by an interaction with coenzyme A (CoASH) (8). The teratogenic effects of valproate are as yet unexplained. Therefore, we investigated fetal and maternal Zn metabolism of normal and Zn deficient rats treated with valproate. As in similar experiments with salicylate (9), the drug was given from d 16 to 20 of gestation, when most of the fetal rat body mass is synthesized.
MATERIALS AND METHODS Pregnant Wistar rats, weighing 200 g, were fed a control or Zndeficient diet and distilled water ad libitum from d 0 to 21 of gestation. For preparing the diets, a commercial diet (Ssniff, Soest, FRG) without ZnCO3 and MgSO4 in the mineral mix was used. Casein, separately received, was extracted with 1% EDTA (disodium salt) for 4 h and 0.5% EDTA for 3 h with stirring, and washed twice with distilled water by suction. The Zn-deficient diet was supplemented with MgC12 to 1000 ppm Mg. Its Zn content was measured after ashing by atomic absorption spectrophotometry and amounted to 1.5 ___ 0.2 ppm Zn. Control rats were fed the same diet supplemented with ZnC12 to 100 ppm Zn. Food intake by the control and Zn-deficient rats was measured daily. [For more details see ref. (10).] From d 16 to 20, the rats in both dietary groups received daily oral doses of 0, 100, or 300 mg/kg sodium valproate. Other authors (6,7) gave 300 (6) or up to 750 mg/kg (7) valproic acid (VPA) for 7 d to nonpregnant rats. Thus, our doses were somewhat lower, although the elimination of VPA is increased during pregnancy (11). Daily therapeutic doses of VPA are lower (4) than those applied in experiments with rats. However, the elimination rate of VPA in rats is higher than in humans because of the lower binding of VPA to rat serum proteins (12). At d 21, the uteri were dissected under nembutal anesthesia (50 mg/kg sc). Blood was taken from the fetal hearts by means of heparinized capillaries and from the mothers by heart puncture and was centrifuged for 5 rain at 1300g. Fetal and maternal livers were removed and cleaned in cold 10% sucrose. Sections (0.5 g) of livers were used for preparing a Biological Trace Element Research
Vot I O, 1986
Effect of Valproate on Zn Metabolism
27
10% homogenate and a 100,000g supernatant, according to Thiers and Vallee (13). Protein content in the supernatants was determined according to Lowry et al. (14). The rest of the livers was frozen in liquid nitrogen, freeze-dried, and pulverized in a plastic mortar. Freeze-dried, powdered liver was ashed overnight in the PlasmaProcessor 200-E (Technics, GmbH, Munich, FRG). The ash was dissolved in 0.1N HC1. Aliquots were taken for measurement of Na and K by flame photometry (Beckman), and for measurement of Mg and Zn by atomic absorption spectrophotometry (Perkin-Elmer, model 303). In fetal and maternal serum, Na, K, Mg, and Zn were measured using the same methods. Ultrapure water (Millipore) was used for dilution and calibration curves. Glycogen of all livers was measured by hydrolysis with amyloglucosidase, and the liberated glucose was determined in an optical test (15). For determination of Zn-metallothionein (Zn-MT), 2 mL of the 100,000g supernatant of maternalor pooled fetal liver were separated on a column 90 x 1.5 cm (Pharmacia, Uppsala, Sweden) filled with Sephadex G-75. The column was eluted with a solution containing 0.1M NaCI, 50 mM TrisC1, ph 7.4, and 0.02% NAN3, at 20~ The flow rate amounted to 15 mL/h. Fractions of 2 mL were sampled (Fraction Collector Frac-300, Pharmacia) for measurement of Zn by atomic absorption spectrophotometry and for identification of metallothionein by UV absorption. Calibration of the columns was performed using the Gel Filtration Calibration Kit, according to the instructions of the producer (Pharmacia). Fetal livers and kidneys from all groups were fixed in 3.5% formaldehyde, embedded in paraplast, sectioned, and stained with hematoxylin and eosin.
RESULTS General Observations Under our experimental conditions (up to d 20), there was no significant difference in food intake between the Zn-deficient and normally fed groups, as was found in preceeding papers (9,10), in which the effects of single Zn deficiency were described in detail. Body weights of the Zn-deficient dams and fetuses were 20 and 25% lower compared to the normally fed controls. Only the highest valproate doses caused a small reduction in body weight of the fetuses from normally fed dams (Table 1). The resorption rate in Zn deficiency was 10%, as found in other experiments (10). Valproate application had no effect on the resorption rate. In the Zn-deficient groups treated with 0, 100, or 300 mg/kg valproate, 3, 10, or 40% of the fetuses developed hydrops, whereas in the normally fed rats, only 6% of the fetuses treated with 300 mg/kg Biological Trace Element Research
Vol. I O, 1986
28
Vormann et aL
{',1
~,
oo +i +i +I
V
+1 +1 +1 +1
I
N
{N
8 rr
v
o
~c3o
~; o
oo +1 +1 +l
r,.,)
t~j"
+1 +1 +1 §
mm
m
V
t-xl t,-,l ~-.~ tr t'N . ~ rr c-,i ~ o~
+1 +1 +1 +1
+
a
m
B N
N
~'~ N
