Pteroylpolyglutamate pool modifications in rat liver after chronic administration of phenobarbitone and valproate

Pharmacological Research, Vol . 25, No . 4, 1992

317

PTEROYLPOLYGLUTAMATE POOL MODIFICATIONS IN RAT LIVER AFTER CHRONIC ADMINISTRATION OF PHENOBARBITONE AND VALPROATE C . BOVINA *, G . FORMIGGINI *, S . SASSI *, M . BATTINOt and M . MARCHETTI* *Department of Biochemistry, University of Bologna, via Irnerio 48, Bologna, Italy ; tlnstitute of Biochemistry, University of'Ancona, via Ranieri-Monte D'Ago, Ancona, Italy Received in final form 20 November 1991

SUMMARY Chronic intraperitoneal administration of low doses of phenobarbitone and valproate caused different alterations in hepatic percentage distribution of pteroylpolyglutamate derivatives without modification of total folate content . Phenobarbitone treatment caused a significant decrease of the percentage content of reduced unsubstituted and methylene-substituted derivatives, while valproate produced an increase of the percentage content of methenyl-, formyl- and formimino-substituted derivatives and a concomitant percent increase of hexaglutamates . The modified ratios of various pteroylpolyglutamates, both in phenobarbitone- and in valproate-treated animals, probably contribute to influencing the partitioning of the one-carbon pool through the various areas of one-carbon metabolism . Kr-:v WORDS :

folate metabolism, pteroylpolyglutamates, phenobarbitone, valproate .

INTRODUCTION A growing number of reports have appeared in recent years concerning folate deficiency induced by long-term treatments with various anticonvulsant drugs . Carl et al . [1-3] provided much evidence in support of the hypothesis that such anticonvulsants as primidone, phenytoin and phenobarbitone can interfere with one-carbon metabolism by inducing pteroylpolyglutamate content and folatedependent enzyme activity modifications . Valproate administration has also been shown to have significant effects on folate metabolism in both the liver and brain, even though qualitatively different from the effects produced by other anticonvulsants [4] . Folate depletion is a very unpleasant secondary effect which has recently heen Correspondence to : Prof. Mario Marchetti . 10-13-6618/92/040317-07/$03 .00/0

C 1992 The Italian Pharmacological Society



318

Pharmacological Research,

Vol. 25, No . 4, 1992

associated with psychiatric morbidity and various other neurological disorders of the epileptic anticonvulsant-treated population [5, 61 in addition to megaloblastic anaemia [7] . As is well known, pteroylpolyglutamates are the preferred substrates for folatedependent enzymes as well as being inhibitors for a number of these [8] with different specificity depending on glutamate chain length . For this reason onecarbon flux might be affected by the modification of the various polyglutamate percentage compositions . Shane in his recent review [9] focused on how the 10formyl-H 4 -PteGlu, :H4 -PteGlu, ratio is important in regulating the availability of one-carbon units . Our previous results [10, 111 showed that acute administration of phenobarbitone and valproate did affect total hepatic folate content in opposite directions : phenobarbitone caused a significant decrease of reduced unsubstituted and methylene-substituted pteroylpentaglutamates, while valproate produced an increase of pteroylpenta- and hexaglutamates of the same kind . In the present study we have investigated pteroylpolyglutamate distribution in rat liver after the i .p . administration of low daily doses of phenobarbitone and valproate for 40 days in order to extend our knowledge of anticonvulsant-folate metabolism interactions .

MATERIALS AND METHODS Materials Synthetic pteroylglutamates (PteGlu . n=1-7) were obtained from Dr B . Schirks Laboratories (Jona, Switerzerland) . Bio-Gel P2 (200-400 mesh) was purchased from Bio-Rad Laboratories (Richmond, CA) . Chemicals and solvents were obtained from Sigma Chemical Co . (St . Louis, MO, USA) and Merck (Darmstadt, FRG) . Animals and treatments Three groups (12 animals each) of male Wistar albino rats were maintained from weaning for 40 days under the controlled feeding schedule of Potter [12] : 12 h light-dark cycle with food, a purified diet containing pteroylmonoglutamate (15 mg/kg of diet), supplied only during the first 8 h of darkness ; water ad libitum . The first group of rats were i .p . injected with phenobarbitone (20 mg/kg body wt) in physiological saline once daily for 40 days ; the second group of rats were injected with valproate (100 mg/kg body wt) ; the third group of rats were injected with appropriate amounts of physiological saline (control) . Each rat was weighed twice weekly and dosages were adjusted . All the rats were killed 24 h after their last injection . Their livers were promptly removed and individually homogenized in 9 vol (w/v) of degassed . argonsaturated 0 .111 N HCI . Pterovlpol.vglutanmate determination Liver homogenates were pooled two by two and subjected to the differential cleavage procedure of Eto and Krumdieck [ 13] . This procedure permits the quantitation of three different pools of folates :



Pharmacological Research, Vol. 25, No. 4, 1992

319

pool 1 is made up of 5,10-methylenetetrahydrofolates, unsubstituted tetrahydroand dihydrofolates ; pool 2 is made up of 5-methyltetrahydrofolates ; pool 3 is made up of 5,10-methenyl-, l0-formyl-, 5-formyl and 5formiminotetrahydrofolates . Selective cleavage of the C9-N 10 bond of the folates of pool 1 (route I), pools 1+2 (route II), pools 1+2+3 (route III) was performed and the resulting p- . aminobenzoylpolyglutamates (pABGlu,) were converted to Azo-derivatives (AzoGlu 0 ) by the Bratton-Marshall reaction ; AzoGlu, purified on a Bio-Gel P2 column (1 .5x14 cm), were concentrated by lyophilization [14] . Standard pteroylglutamates were subjected to the procedure described by Brody et al. [15] . Analytical determinations were performed by HPLC as previously described [10] . Calculations and statistics

Glutamate chain length of AzoGlu,, derivatives of the hepatic folates was determined by comparing the retention times corresponding to each peak with those of the standards . The amount of each polyglutamate of routes I, II and III was evaluated by automatic integration of the area under each peak and external standard quantitation . The absolute amounts of each pteroylpolyglutamate of pools, 1, 2 and 3 were calculated as follows : pool 1=route I pool 2=route II-route I pool 3=route III-route II . Statistical comparisons between groups were carried out using the Student's ttest. A level of significance of P<0 .05 was chosen .

RESULTS The pteroylpolyglutamate distribution in pools 1, 2 and 3 (see Materials and methods) in the liver of control and treated rats is shown in Table I : chronic administration of phenobarbitone causes a significant decrease (P<0 .05-0 .001) of all polyglutamate derivatives of pool 1 compared with controls . For valproate-treated rats the tetraglutamate content in pool 3 is significantly lower (P<0 .05) and the hexaglutamate content of the same pool significantly higher (P<0 .05) compared with controls . Total folate contents (nmol/g liver) both in phenobarbitone- (21 .28±3 .14) and in valproate-treated (26 .18±2 .20) rat livers are not significantly different from the control value (24 .34±2 .05) . In Fig . 1, which shows the hepatic percentage distribution of folate derivatives distinguished by chain length, it can be seen that there are no significant variations in phenobarbitone-treated animals, while a significant decrease of tetraglutamates (P<0 .001) and increase of hexaglutamates (P<0 .01) can be seen in valproate-treated rats . Figure 2 shows the hepatic percentage distribution of folate derivatives distinguished by pteridine ring substitution : for pool 1 derivatives significant



Pharmacological Research, Vol. 25, No. 4, 1992

3 20

Table I Pteroylpolyglutamate distribution in the liver of control and treated rats Control (nmollg liver)

Phenobarbitonetreated (nmol/g liver)

Valproate-treated (nmollg liver)

Pool 1 (Methylene+ Unsubstituted)

G4 G5 G6

1 .62±0 .17 8 .41±0 .21 2 .04±0 .19

0 .55±0 .09** 6 .59±0 .37** 1 .35±0 .09*

1 .14±0 .07 8 .06±0 .45 2 .16±0 .13

Pool 2

G4 G5 G6

0 .68±0 .33 3 .21±0 .67 1 .75±0 .42

1 .13±0.16 2 .67±0 .93 1 .40±0 .34

0 .63±0 .17 3 .72±0 .94 1 .62±0 .44

G4 G5 G6

1 .47±0 .42 3 .97±0 .87 1 .40±0 .38

1 .63±0 .54 3 .84±1 .92 2 .11±0 .46

0 .75±0 .28* 5 .38±1 .23 3 .37±0 .70*

(Methyl) Pool 3

(Methenyl+ Formyl+ Formimino)

The symbols G4 , G 5 and G6 indicate the number of glutamyl residues . Each value represents the mean±sE of results obtained from 12 animals pooled two by two . Each measurement was performed in triplicate . *Significantly different from control, P<0 .05 by Student's t-test . **Significantly different from control, P<0 .001 by Student's t-test .

decreases can be seen both in phenobarbitone- (P<0 .01) and in valproate-treated rats (P<0 .05) compared with controls . Finally the significant increase (P<0 .01) of the percentage content of pool 3 derivatives is seen in the liver of valproate-treated animals .

DISCUSSION The present data substantially agree with those of other authors in showing the interference of phenobarbitone and valproate on liver folate derivative distribution [3, 4] . The chronic drug administrations we performed differ from those of other rat treatment models : the daily doses of both drugs were low and not in the range protective against seizures . Our experiments were performed in this way so that we could investigate the interference of anticonvulsants of different chemical structures on folate metabolism, without taking their pharmacological action into consideration . Compared with gastric gavage, intraperitoneal administration avoids the possible direct interference of the drugs on folate intestinal absorption and transport [ 16] . Moreover, the feeding of a purified diet containing a large quantity of pteroylmonoglutamate allows the animals to have good vitamin availability . Under these conditions, significant modifications in the liver percentage distribution of the various folate derivatives were observed both in phenobarbitone- and in valproate-treated rats without total folate content changes . Indeed, a number of experimental circumstances have been described resulting in



Pharmacological Research, Vol . 25, No . 4, 1992

G4 Control

32 1

G5 I

1 Phenobarbitone -treated

G6 \\\` Valproate-treated

Fig. 1 . Percentage distribution of liver folate derivatives distinguished by chain length . The symbols G 4 , G 5 and G6 indicate the number of glutamyl residues . Values which differ significantly from controls are indicated . **P<0 .01, ***P<0 .001 .

60 50

c 40 o u, v

30 20 10

Pool I Control

Pool 2

L_J Phenobarbitone-treated

Pool 3 \\\ Valproate-treated

Fig. 2 . Percentage distribution of liver folate derivatives distinguished by pteridine ring substitution . Pools 1, 2 and 3 components are indicated in Materials and methods . Values which differ significantly from controls are indicated : *P<0 .05, **P<0.01 .

significant alterations of folate derivative distribution in the absence of any change in the total content [ 17] . After chronic phenobarbitone treatment, as after the acute treatment [10], it can be observed that the group of modified derivatives is pool 1 (reduced unsubstituted and methylene-substituted), which appears decreased both in absolute value and in percentage terms .



322

Pharmacological Research, Vol. 25, No . 4, 1992

It is interesting to note that only pentaglutamate derivatives were significantly lower after the acute treatment while both tetra- and hexaglutamates are significantly decreased after the chronic treatment . We think that the hypothesis we have already proposed [10] applies in this case too : the induction of cytochrome P450 system by phenobarbitone [18] might direct NADPH for hydroxylation reactions indirectly perturbing folate-dependent reactions and as a consequence folate distribution . In fact Carl and Smith [3] show that phenobarbitone chronic treatment affects one-carbon enzyme activities and in particular the NADPH-dependent methylenetetrahydrofolate reductase activity, with significant correlation to the length of the treatment . As for valproate treatment, our results agree with those of Carl [4] in showing the different effects of this drug on the liver contents of various folate pools . However we would like to stress that chronic administration of both phenobarbitone and valproate, even though producing different modifications, leads to a quite similar percentage distribution of folates distinguished by pteridine ring substitution . The significantly different polyglutamate distribution in valproate-treated rats compared with controls probably reflects a secondary effect due to the variation in folate one-carbon distribution, in its turn related to the enzymic activity modifications seen by Carl [4] . In conclusion it is likely that the changed ratios of various polyglutamates observed in the liver both of phenobarbitone- and valproate-treated rats contribute to influencing the partitioning of the one-carbon flux through the various areas of one-carbon metabolism .

REFERENCES 1 . Carl GF, Eto I, Krumdieck CL . Chronic treatment of rats with primidone causes depletion of pteroylpentaglutamates in liver . J Nutr 1987 ; 117 : 970-5 2 . Carl GF, Smith DB . Interaction of phenytoin and folate in the rat . Epilepsia 1983 ; 24 : 494-501 . 3 . Carl GF, Smith DB . Effect of chronic phenobarbital treatment on folates and one-carbon enzymes in the rat . Biochem Pharmacol 1984; 33 : 3457-63 . 4 . Carl GF . Effect of chronic valproate treatment on folate-dependent methyl biosynthesis in the rat . Neurochem Res 1986 ; 11 : 671-85 . 5 . Shorvon SD, Carney MWP, Chanarin I, Reynolds EH . The neuropsychiatry of megaloblastic anaemia. Br Med J 1980; 281 : 1036-8 . 6 . Trimble MR, Corbett JA, Donaldson D . Folic acid and mental symptoms in children with epilepsy . J Neurol Neurosurg Psychiat 1980 ; 43 : 1030-4 . 7 . Badenoch J . The use of labeled vitamin B12 and gastric biopsy in the investigation of anemia. Proc R Soc Med 1954 ; 47 : 426-7 . 8 . Matthews RG, Ghose C, Green JM, Matthews KD, Dunlap RB . Folylpolyglutamates as substrates and inhibitors of folate-dependent enzymes . Adv Enz Reg 1986; 26 : 157-7 1 . 9 . Shane B . In : Aurbach GD, McCormick DB eds . Vitamins and hormones . London : Academic Press, 1989 : 263-335 . 10 . Bovina C, Formiggini G, Sassi S, Battino M, Marchetti M . Changes in hepatic folylpolyglutamate pattern in phenobarbitone-treated rats . Biochem Pharmacol 1990 ; 39 : 1847-51 . I1 . Bovina C, Formiggini G, Sassi S, Marchetti M . Effect of treatments with different anticonvulsant drugs on folate pattern in rat liver . In : Curtius Ch-H, Ghisla S, Blau N



Pharmacological Research, Vol. 25, No. 4, 1992

12 .

13 .

14 .

15 . 16 . 17 .

18 .

323

eds . Proceedings of 9th int symp pteridines and folic acid derivatives, Zurich . 1989 . Berlin : Walter de Gruyter, 1990 ; 934-7 . Potter VE . Baril EF, Watanabe M, Whittle ED . Systematic oscillations in metabolic functions in liver from rats adapted to controlled feeding schedules . Fed Proc 1968 : 27 : 1238-45 . Eto 1, Krumdieck CL . Determination of three different pools of reduced one-carbon substituted folates . 11 . Quantitation and chain-length determination of the pteroylpolyglutamates of rat liver. Anal Biochem 1981 ;115 :138-46. Eto 1, Krumdieck CL . Determination of three different pools of reduced one-carbon substituted folates . III . Reversed-phase high-performance liquid chromatography of the azo dye-derivatives of p-aminobenzoylpoly-y-glutamates and its application to the study of unlabeled endogenous pteroylpolyglutamates of rat liver . Anal Biochem 1982 ; 120 : 323-9 . Brody T, Shane B, Stokstad ELR . Separation and identification of pteroylpolyglutamates by polyacrylamide gel chromatography . Anal Biochem 1979 ; 92 : 501-9 . Darcy-Vrillon B, Selhub J, Rosenberg IH . Analysis of sequential events in intestinal absorption of folylpolyglutamate . Am J Physiol 1988 ; 255 : 361-6 . Krumdieck CL, Eto I. Folates in tissues and cells . Support for a `two-tier' hypothesis of regulation of one-carbon metabolism . In : Cooper BA and Whitehead VM, eds . Chemistry and biology of pteridines 1986 . Pteridines and folic acid derivatives . Berlin : Walter de Gruyter, 1986 ; 447-66 . Strobel HW, Cattaneo E, Adesnik M, Maggi A . Brain cytochromes P-450 are responsive to phenobarbital and tricyclic amines . Pharmacol Res 1989 ; 21 : 169-75 .