Susanna J. Dodgson

Topiramate: Preclinical Evaluation of a Structurally Novel Anticonvulsant

Summary: Topiramate [TPM, 2,3:4,5‐his‐O‐(1‐methyl‐ethylidene)‐β‐D‐fructopyranose sulfamate] (RWJ‐17021‐000, formerly McN‐4853) is a structurally novel antiepileptic drug (AED). The preclinical anticonvulsant profile suggests that TPM acts primarily by blocking the spread of seizures. TPM was highly effective in the maximal electroshock (MES) seizure test in rats and mice. Activity was evident 0.5. h after oral administration and lasted at least 16 h. The ED50 values 4 h after oral dosing were 13.5 and 40.9 mg/kg in rats and mice, respectively. TPM blocked pentylenetetrazol (PTZ)‐induced clonic seizures at high doses in mice (ED50= 1,030 mg/kg orally, p.o.). With motor incoordination and loss of righting reflex used as indicators of neurologic impairment, the neuroprotective index (TD50/MES ED50) for TPM was equivalent or superior to that of several approved AEDs. In mice pretreated with SKF‐525A (a P450 enzyme inhibitor), the anticonvulsant potency was either increased or unaffected when TPM was tested 0.5, 1, or 2 h after i.p. administration, suggesting that TPM rather than a metabolite was the active agent. In mice pretreated with reserpine or tetrabenazine, the activity of TPM in the MES test was markedly reduced. TPM was inactive in a variety of receptor binding, neurotransmitter uptake, and ion channel tests. TPM weakly inhibited erythrocyte carbonic anhydrase (CA) activity. However, the anticonvulsant activity of TPM appears to differ mechanistically from that of acetazolamide.

Topiramate as an Inhibitor of Carbonic Anhydrase Isoenzymes

Purpose: This study investigated the effectiveness of topiramate (TPM) as an inhibitor of six isozymes of carbonic anhydrase (CA).

The Carbonic Anhydrases

Carbonic anhydrase (CA) (EC 4.2.1.1.) was first characterized in erythrocytes in 1933 directly as a result of a search by several laboratories for a catalytic factor in the erythrocytes that had been theoretically determined as necessary for rapid transit of the HCO 3 − from the erythrocyte to the pulmonary capillary. Two laboratories simultaneously published papers describing a catalytic factor. From Dr. Roughton’s laboratory at Cambridge University came a paper with an elegant title including a name for what they thought was a single enzyme (“Carbonic Anhydrase: Its Preparation and Properties”)61; from Dr. Stadie’s laboratory in Philadelphia came another paper with a less concise title (“The Catalysis of the Hydration of Carbon Dioxide and Dehydration of Carbonic Acid by the Enzyme Isolated from Red Blood Cells”; the other author was Dr. Helen O’Brien).73 Perhaps the more frequent citation of the Cambridge paper is due to the shorter title with scholarly colon as well as the confident naming of this newly discovered enzyme. The adjective “late” before the senior author’s name describes the tragically premature end of N. M. Meldrum’s life by his own hand not long after a crippling accident. Dr. Roughton remained in Cambridge until his death in 1971; he worked with many stellar white male scientists from both sides of the Atlantic Ocean (e.g., Sir Joseph Barcroft, Sir Hans Krebs, P E Scholander, Britton Chance, Quentin Gibson, Robert Forster, and J. W. Severinghaus). Dr. Roughton continued his interest in CO2 transport until his death; his life and career have been discussed and his list of publications documented by Dr. Quentin Gibson, a fellow member of the Royal Society and one-time colleague.31 Dr. Stadie continued in the field of metabolism, and by his death in the 1940s he was a noted researcher in diabetes. Dr. O’Brien also left the CA field.

The roles of carbonic anhydrase in metabolism, cell growth and cancer in animals.

Viewed from the standpoint of chemical reactivity, CO2 would appear to be a more appropriate substrate than bicarbonate for carboxylation reactions, since it would be more susceptible to nucleophilic attack. However, in aqueous medium, the equilibrium between dissolved CO2 and bicarbonate is such that, at physiological pH, bicarbonate is present at some 20-fold higher concentration. Furthermore, bicarbonate probably has greater potential for binding to enzymes since it is a more polar molecule than CO2. It is perhaps not surprising then, that although the product of decarboxylation reactions in catabolic processes is CO2, several carboxylating enzymes have evolved to employ bicarbonate, not CO2, as their substrate. The carboxylating enzymes in animals, which are known to bind bicarbonate as substrate, are the biotin-dependent carboxylases and the carbamoyl phosphate synthetase isozymes. These enzymes bind bicarbonate, but then generally convert it either to CO2 (biotin-dependent carboxylases) or to an activated form of CO2 (carbamoyl phosphate synthetases). Carbonic anhydrase (CA) is perhaps alone among enzymes in being able to bind either of these substrates. (For reviews see Rubio, 1986; Knowles, 1989 and O’Leary, 1992).

The Role of Carbonic Anhydrase in Hepatocyte Metabolism

The existence of carbonic anhydrase inside mammalian hepatic mitochondria has been suspected since 1959.8,23,25,28*29,33 Rossi” determined the carbonic anhydrase activity of mitochondria that had their inner membranes permeabilized by Triton X-100, but found no activity in intact mitochondria. He concluded that this was evidence for the existence of carbonic anhydrase inside the inner mitochondrial membrane. Work from two laboratories later showed that acetazolamide blocked the uptake of Ca2+ and HCO? by respiring mit~chondria;’~J~.~O these researchers concluded that carbonic anhydrase activity situated within the inner mitochondrial membrane was blocked by the drug. Subsequent work from our laboratory has shown that the hepatic mitochondrial carbonic anhydrase is a soluble matrix enzyme; we have quantitated its activity by monitoring the I8O disappearance from C18016010,18 with intact hepatic mitochondria. Separation of fractionated mitochondria into submitochondrial particles and matrix-derived soluble supernatant by ultracentrifugation resulted in finding that all carbonic anhydrase activity was in the supernatant fraction. Our first goal in this work was to confirm the existence of mitochondrial carbonic anhydrase and to quantitate its activity. Our initial evidence of its presence within intact hepatic mitochondria is given in FIGURE 1. When freezethawed mitochondria were added to the reaction chamber (Curve A) there was an immediate acceleration in the decrease of C160’s0 from NaHC03 (25 mM 2% labeled with I8O). This accelerated decrease was eliminated by the inclusion of 1 pM concentrations of acetazolamide in the solution. When intact mitochondria (Curve B) were added to the reaction chamber, the accelerated decrease in C180’60 was biphasic; this pattern indicates that the carbonic anhydrase is not in immediate contact with the reaction solution; that is, that it is contained within a membrane that is not freely permeable to HCO?.l8 We know from Rossi’s work initially33 and from our fractionation of the mitochondria1° that the carbonic anhydrase is contained within the inner mitochondrial membrane, not merely the outer. The outer membrane is known to be permeable to HCO,;’ our assay would give identical traces for intact and broken mitochondria if there were any carbonic anhydrase in the membrane space. We assayed for carbonic anhydrase activity with intact and cholate-treated mitochondria using a changing pH technique; we found activity by this method only with the broken mitochondria, not with the

Inhibition of CA V decreases glucose synthesis from pyruvate

The carbonic anhydrase inhibitor acetazolamide reduces citrulline synthesis by intact guinea pig liver mitochondria and also inhibits mitochondrial carbonic anhydrase (CA V) and the more lipophilic carbonic anhydrase inhibitor ethoxzolamide reduces urea synthesis by intact guinea pig hepatocytes in parallel with its inhibition of total hepatocytic carbonic anhydrase activity. Intact hepatocytes from 48-h starved male guinea pig livers were incubated at 37 degrees C in Krebs-Henseleit with 95% O2/5% CO2 at pH 7.1 with 5 mM pyruvate, 5 mM lactate, 3 mM ornithine, 10 mM NH4Cl, 1 mM oleate; with these inclusions both urea and glucose synthesis start with HCO3- -requiring enzymes, carbamyl phosphate synthetase I and pyruvate carboxylase, respectively. Urea and glucose synthesis were inhibited in parallel by increasing concentrations of ethoxzolamide, estimated Ki for each approximately 0.1 mM. In other experiments hepatocytes were incubated at 37 degrees C in Krebs-Henseleit with 95% O2/5% CO2 at pH 7.1 with 10 mM glutamine, 1 mM oleate; with these inclusions glucose synthesis no longer starts with a HCO3- -requiring enzyme. Urea synthesis was inhibited by ethoxzolamide with an estimated Ki of 0.1 mM, but glucose synthesis was unaffected. Intact mitochondria were prepared from 48-h starved male guinea pig livers. Pyruvate carboxylase activity of intact mitochondria was determined in isotonic KCl-Hepes buffer, pH 7.4, 25 degrees C, with 7.5 mM pyruvate, 3 mM ATP, and 10 mM NaHCO3. Inclusion of ethoxzolamide resulted in reduction in the rate of pyruvate carboxylation in intact mitochondria, but not in disrupted mitochondria. It is concluded that carbonic anhydrase is functionally important for gluconeogenesis in the male guinea pig liver when there is a requirement for bicarbonate as substrate.

The value of inherited deficiencies of human carbonic anhydrase isozymes in understanding their cellular roles.

Very little light has been shed on the role of the low-activity CA I isozyme in humans by studies on CA I-deficient individuals. On the other hand, CA II-deficient individuals exhibit abnormalities of bone, kidney and brain, implicating a functional role for the high-activity CA II isozyme in cells from these tissues and organs. It also appears that the CA II-deficient red cell is capable of normal respiratory function under unstressed conditions. In addition, there is some preliminary evidence that those organs such as the eye which primarily contain the CA II isozyme, may be able to function effectively in the absence of CA II.

Failure of Acetazolamide to Decrease Intraocular Pressure in Patients with Carbonic Anhydrase II Deficiency

The effect of the carbonic anhydrase inhibitor acetazolamide on intraocular pressure was studied in two patients with carbonic anhydrase II deficiency and in six control subjects. The deficient patients had the autosomal recessive syndrome of osteopetrosis with renal tubular acidosis and cerebral calcification. A dose of 125 mg of intravenous acetazolamide caused a significant (P less than .01) decrease in intraocular pressure from baseline (15.0 +/- 1.5 mm Hg) in the control subjects one hour (11.3 +/- 1.5 mm Hg) and four hours (13.8 +/- 1.2 mm Hg) after drug administration. In contrast, the patients with carbonic anhydrase deficiency showed no such decrease in intraocular pressure; baseline intraocular pressure (19.2 +/- 0.2 mm Hg) was significantly unchanged (P greater than .5) at one hour (20.0 +/- 0.1 mm Hg) and four hours (19.2 +/- 0.2 mm Hg).

Differential regulation of hepatic carbonic anhydrase isozymes in the streptozotocin-diabetic rat

Most work with the male rat liver carbonic anhydrase isozymes in the past decade has centered on the cytosolic CA III and the mitochondrial CA V. This paper reports that the relative activity of both isozymes is altered in streptozotocin-diabetes. Carbonic anhydrase activity of perfused liver homogenates and disrupted, isolated mitochondria was measured by the mass spectrometric 18O decay technique at 37 degrees C. The contributions of the different isozymes were determined based on intracellular location and sensitivity to acetazolamide inhibition. Diabetes resulted in a twofold increase in the activity of CA V but a halving in the activity of CA III. This is the first time that liver CA V has been shown to be altered by physiological stress. The total carbonic anhydrase activity in the diabetic rat liver was unaltered compared with control rats; however, CA III never accounted for more than 50% of this activity. Since CA isozymes I, II, and IV together account for 30% of the CA activity in control rats and 70% in diabetic rats it is concluded that one or more of these isozymes is subject to regulation in the diabetic male rat. The increase in CA V during diabetes is in accord with this isozyme having an important function in provision of substrate for hepatic gluconeogenesis and ureagenesis.

Liver Mitochondrial Carbonic Anhydrase (CA V), Gluconeogenesis, and Ureagenesis in the Hepatocyte

In the 1970s, there was great interest in the functioning of isolated mitochondria, which led in 1978 to the awarding of the Nobel Prize for Physiology and Medicine to Dr. Peter Mitchell for his chemiosmotic theory. A few weeks later, the reviewer arrived in Philadelphia as the postdoctoral fellow of Dr. Robert E. Forster II. Dr. Forster has had a successful career devoted to studying anomalies in the theories of gas diffusion, pH disequilibrium, and carbonic anhydrase (CA)18,22; he believed that there was something wrong in Dr. Mitchell’s theory and that a mitochondrial CA would dissipate the proton gradient required. A collaboration was started with two experts in mitochondrial bioenergetics: Dr. Leena Mela, who knew how to isolate mitochondria from several organs, and Dr. Bayard Storey, who knew how to isolate skeletal muscle mitochondria.42 By 18O mass spectrometric CA analysis,18,30 the reviewer found abundant CA activity in guinea pig liver and skeletal muscle but none in brain, kidney, or heart.15 The work on skeletal muscle that has been done since then is reviewed elsewhere,42 as is the work with the CA V containing rat kidney mitochondria.6 Exhaustive literature searches indicated that liver fractionation by previous workers had given some evidence of CA activity in mitochondria, but several believed that this was the result of contamination from the cytosol of hepatocytes or erythrocytes (reviewed in reference 21). Work in which sulfonamide inhibition reduced HCO3−-linked Ca2+ transport across the mitochondrial membrane did, however, convince the investigators of the existence of a mitochondrial CA.17,24 With one notable exception,45 there has been until recently little interest outside of this laboratory in studying this lovely isozyme directly but considerable interest in studying the effects of rendering it nonfunctional by sulfonamide CA inhibition. Mitochondrial CA inhibition has been concluded to be responsible for decreased urea and glucose synthesis by alligators and chameleons in vivo,5 decreased urea synthesis by isolated perfused rat livers,25,34 decreased urea and glucose synthesis by isolated rat hepatocytes,3,35,39 and by isolated guinea pig hepatocytes,7,10,11,14 and decreased citrulline synthesis by intact isolated liver mitochondria from guinea pigs7,12 and rats46 (S. J. Dodgson and A. J. Meijer, unpublished results). Reviews of ongoing work in the field have appeared in the past decade.14,21,22