Fordham University Masthead Logo DigitalResearch@Fordham

Chemistry Faculty Publications Chemistry

1991 Fluoroacetate and fluorocitrate: mechanism of action / Donald D. Clarke Donald Dudley Clarke PhD Fordham University, [email protected]

Follow this and additional works at: https://fordham.bepress.com/chem_facultypubs Part of the Biochemistry Commons

Recommended Citation Clarke, Donald Dudley PhD, "Fluoroacetate and fluorocitrate: mechanism of action / Donald D. Clarke" (1991). Chemistry Faculty Publications. 16. https://fordham.bepress.com/chem_facultypubs/16

This Article is brought to you for free and open access by the Chemistry at DigitalResearch@Fordham. It has been accepted for inclusion in Chemistry Faculty Publications by an authorized administrator of DigitalResearch@Fordham. For more information, please contact [email protected]. ..

Neurochemical Research, Vol. 16, No. 9, 1991, pp. 1055-1058

Fluoroacetate and Fluorocitrate: Mechanism of Action*

Donald D. Clarke1

(Accepted April 22, 1991)

The concept of lethal synthesis as suggested by Peters is reviewed in the light of the more recent work in this area. It is suggested that fluorocitrate is a "suicide" substrate for aconitase rather than a competitive inhibitor as originally suggested. The use of these substances to study glial­ neuronal relationships is considered.

KEY WORDS: Fluoroacetate; fluorocitrate; fluoroaconitate.

In a recent paper (22), it is stated that "The bio­ isomers (2R,3S & 2S,3S) could be ruled out as toxic chemical mechanism by which fluorocitrate and/or fluo­ unless fluoroacetyl-CoA behaved differently than acetyl­ roacetate disrupt the central nervous system and heart CoA when acted on by citrate synthase. Despite this well are not adequately defined". While this statement is cor­ known information there have been a few attempts in rect, the main aim of this review is to distinguish what biochemistry textbooks to assume that the C-F bond of is well defined from what is still controversial about this fluorocitrate must be in a position analogous to the C­ subject. The concept of "lethal synthesis", i.e. the con­ OH bond of isocitrate to cause inhibition of aconitase. version of fluoroacetate to fluorocitrate which in turn In an extreme example it was suggested that the initially inhibits citrate metabolism, orginally proposed by Pe­ formed fluorocitrate was converted to fluorooxaloacetate ters, is still generally well accepted. However, the man­ in the Krebs cycle and on the second turn of the cycle ner in which fluorocitrate produces its toxic effect is still was converted to an isomer of fluorocitrate with the ster­ unsettled. eochemistry assumed to be necessary. Not only does That only one stereoisomer, which they defined as such an assumption contradict the results of Kun's lab­ (-)-erythro-fluorocitrate, is toxic was elegantly and oratory which show that fluorooxaloacetate condenses unequivocally demonstrated by Dumel and Kun (8). The with acetyl-CoA to yield a non toxic isomer (8) but it absolute configuration of this isomer was shown to be had been demonstrated by Clarke et al. (5) that fluoro­ 2R,3R using X-Ray diffraction techniques in Glusker's fumarate is a substrate for fumarase and fluoride is elim­ laboratory (23). inated in the process. The mechanism of this process has been established by Marietta et al. (13). Well before the absolute stereochemistry of the toxic The original suggestion of Peters that fluorocitrate isomer had been established it was appreciated by many was the toxic agent was not based on a definitive iso­ workers in this field that, on the basis of the fact that lation and chemical identification of fluorocitrate. Rather, [1- 14C]acetate yields a labeled citrate which in turn is it was based on a partial purification of the toxic agent converted to isocitrate and then a-ketoglutarate labeled which suggested that it was similar to material later syn­ in the -y-carboxyl group, the absolute stereochemistry at thesized chemically by Rivett (21 ), and was primarily position 3 of 2-fluorocitrate was most likely R. The 3S based on the accumulation of citrate in tissues of fluo­ roacetate poisoned animals. It is only relatively recently 1 Department of Chemistry, Fordham University, Bronx, NY 10458, U.S.A. that a definitive identification of erythro-2-fluoro-[2- * Special issue dedicated to Dr. Louis Sokoloff. 3H]citrate was achieved by adding carrier material, de- 1055 0364-3190/91/0900-1055$06.50/0 © 1991 Plenum Publishing Corporation 1056 Clarke rivativizing as the trimethyl ester benzoate and separa­ beling of glutamine. Later work suggested that this small tion by HPLC (20). In that study [29-3 H]-29- pool was located in glial cells. On this basis it would fluorostigmasterol was fed to tobacco hornworms (Man­ seem that acetate thiokinase is largely or exclusively in duca sexta) and 0.012% of the administered dose was glia, an assumption that remains to be rigorously proven. converted to (2R,3R)-2-fluoro-[2-3H]citrate. This low Other corollaries of the metabolic pools hypothesis have conversion is consistent with the fact that previous at­ received support from immunocytochemical studies viz. tempts using 14C labeled precursors all failed to result that glutamine synthesase (16) and glutamate dehydro­ in a definitive chemical identification of 2-fluorocitrate. genase (1) are enriched in glial cells. This idea that fluo­ This also makes the original suggestion of Peters that rocitrate is having its effect in glial cells was further fluorocitrate is inhibiting aconitase competitively rather developed in Fonnum's laboratory (17). These authors implausible. found that injection of 1 nanomole of fluorocitrate in­ Over the years many investigators have failed to trastriatally induces an in vivo condition in which glial find large accumulations of citrate in fluoroacetate poi­ metabolism is selectively impaired. They used this con­ soned animals, especially in brain tissue, and thus have dition as a model for the study of metabolic relationships questioned the validity of Peters' explanation. In fact between neurons and glial cells and to assess the im­ Peters and Shorthouse (19) and later Morselli et al. (14) portance of glial cells for brain function in vivo. suggested that "lethal synthesis" occurred in other or­ By analogy with the Sokoloff deoxyglucose method gans and fluorocitrate was then carried to brain tissue to for studying brain metabolism and relating it to function produce its toxic effect. Kidney, heart and liver are the we considered that labeled fluoroacetate might be a pos­ tissues in which most investigators have been able to sible radiochemical marker for studying glial metabo­ demonstrate significant elevations of citrate in fluoroac­ lism. Sokoloff had similar ideas (private communication). etate poisoned animals. However, [14C]fluoroacetate was the only material com­ When tracer studies were brought to bear on the mercially available and it would have required toxic doses subject Quastel's group (12) showed that the most not­ of that material to get sufficient radioactivity into brain able change observed in the brains of fluoroacetate poi­ tissue for such experiments. We were able to convince soned rats was decreased labeling of glutamine from Dr. Richard Young of the New England Nuclear Corp. glucose or other precursors and these effects could be to prepare highly labeled [2-3H]fluoroacetate for us and demonstrated in brain slices. Hence they suggested that this material allowed us to do tracer studies to test this glutamine synthetase may be inhibited. We were able to hypothesis, viz. that [3H]fluoroacetate would be acti­ show that fluorocitrate did not inhibit purified glutamine vated in glia and be converted to [3H]fluorocitrate which synthetase from sheep brain (7). Rather we advanced the would accumulate there and be detectable by autora­ theory that aconitase was being inhibited in a pool of diography. Approx. 1% of a dose of labeled fluoroace­ citric acid cycle intermediates that was much smaller tate administered to rats i. p. enters brain cortex and of than the total pool in brain cells (7). Later work sug­ this only < 0.02% was found in the fraction containing gested that this pool is probably in glial cells while the fluorocitrate. The fluorocitrate fraction in heart ac­ larger pools are in neurons. The experiments on brain counted for 0.25% of the tracer and in kidney for 0.35% slices also indicated that fluoroacetate must be converted (6). This agrees with previous reports in the literature. to fluorocitrate in brain tissue. In that work we attempted The analogy with the Sokoloff deoxyglucose method to isolate labeled fluorocitrate when using [U-14C]-as­ was flawed because no accumulation of labeled fluoro­ partic acid as the tracer. The precursor produces labeled citrate was observed to occur over time. Rather increased oxaloacetate which in turn should produce labeled fluo­ release of tritium into water was found. However, when rocitrate. We failed to detect it (10). However, as men­ autoradiographs were made of different brain areas dif­ tioned above the degree of conversion of fluoroacetate ferential labeling was observed consistent with the pos­ to fluorocitrate is quite low and we would have had to tulate that glia were much more heavily labeled than use almost 100% 14C labeled substrates to have found neurons (15). Tritiated acetate, which is much more readily it. This alerted us to the need to use tritium labeled available than is tritiated fluoroacetate, also yielded sim­ material for tracer experiments in search of labeled fluo­ ilar autoradiographs in experiments of short duration when rocitrate. little of the tritium had been released into water. The metabolic compartmentation hypothesis pro­ When toxic doses of labeled fluoroacetate are used posed by Berl and Clarke (3) was the basis for the above it is possible to get sufficient quantities of labeled ma­ explanation of "lethal synthesis" of fluorocitrate occur­ terials for chemical isolation and study. This approach ring in a small compartment characterized by rapid Ia- has been used recently (22) in conjunction with 19F hnd Fluoroacetate and Fluorocitrate: Mechanism of Action 1057

13C NMR. Fluoroacetate was observed to be extensively concentration of this substance at the site of inhibition conjugated with glutathione in agreement with previous will be of the order of 1 nanomolar, a concentration reports in the literature ..The most significant observation which can be expected to be an effective inhibitor only by these authors was that aconitase defluorinates (-)­ if it is inactivating an enzyme or similar type of protein erythro-fluorocitrate but doesnt defluorinate the ( +) en­ such as a carrier system. antiomer significantly. This suggests that the ( - ) or toxic In a more recent paper (9) it is shown that direct isomer is a substrate for aconitase and the cis-fluoroa­ intra-thecal injection of fluorocitrate produces seizures conitate formed immediately reacts with a nucleophile in mice with a latency of the order of 15 seconds as near the active center of the enzyme with the liberation compared to 30 mins. after intracerebroventricular in­ of fluoride ion. This possibility was suggested by Vil­ jection. Citrate and other Ca2 + chelators produce similar lafranca and Platus (24) but seems to have been dis­ seizures and co-administration of Ca2 + with the chelator missed by those who have studied this topic extensively attenuated the neurotoxicity. In contrast Ca2 + did not (25). prevent the seizures elicited by strychnine. Hence Hom­ Fluoroaconitate is a relatively unstable molecule and feldt and Larson concluded that the proximate cause of was successfully synthesized for the first time by Can­ seizures is the complexation of Ca2 + in the spinal cord tacuzene et al. (4). As an allylic halide it can be expected and the long delay commonly observed in seizure pro­ to be a very reactive substance. Thus while the triethyl duction after i. p. or central injection of fluoroacetate ester was synthesized as early as 1960 (2) all attempts and fluorocitrate is caused by the time needed for dif­ to obtain the free acid by hydrolysis of this ester failed. fusion down the spinal column. In the successful synthesis the tri-t-butyl ester was pre­ We therefore seem to have come almost full circle pared and the ester cleaved under mild conditions using back to the original hypothesis of Peters regarding lethal trimethylsilyl iodide. The trans isomer of 4-fluoroacon­ synthesis of fluorocitrate. The original postulate of com­ itic acid was obtained in pure form but readily isomer­ petitive inhibition of aconitase is untenable. However ised to one with fluorine bound to a vinylic carbon (2- the concept that fluorocitrate was a "suicide" substrate fluoroaconitate). The cis- isomer was not obtained and for aconitase (or for a citrate carrier protein) was not it is one of this enantiomeric pair which would be ex­ possible in the 1950's when Peters offered his theory for pected to be formed from (-)-erythro-2-fluorocitrate. the mechanism of fluoroacetate poisoning. Neither was This is an indication of its great reactivity. the low conversion of fluoroacetate to fluorocitrate ap­ 4-Fluoroaconitate has two stereogenic centers and preciated. Inactivation of an enzyme or carrier also fits so, as with 2-fluorocitrate, there are four stereoisomers. well with the observation that if an animal survives be­ (2R,3R)-2-fluorocitrate, often referred to as (-)-ery­ yond two to three hours, or is given a subtoxic dose, no thro-2-fluorocitrate, should yield (2E,4R)-4-fluoroacon­ long term effects are observed. If the dose administered itate. This stereoisomer might then be expected to be is insufficient to overwhelm the CNS and be lethal, re­ positioned correctly for nucleophilic attack in the 4-po­ covery is total after the synthesis of a new supply of sition by a group near the active center of aconitase to protein or reversal of its inactivation. The action in the cause inactivation of the enzyme. The results of Villa­ spinal cord would explain the long latency period to franca and Platus (24) confirm this. The fact that this produce toxic effects (30--90 mins) after i.p. or similar inactivation is readily reversed when attempts were made injections even though biochemical effects of added to isolate the inactivated enzyme do not detract from this fluoroacetate and fluorocitrate in tissues slices and in the observation. brains of intact animals can be detected within a few Kun has reported a covalently bound "fluorocitry­ minutes after administration of these substances (7). lated'' membrane protein (11) which was believed to be the citrate carrier protein. The exact nature of this protein does not seem to have been established. Whether acon­ ACKNOWLEDGMENT

itase or a citrate carrier protein inactivation explains the The writing of this review was made possible in part by support pile up of citrate in fluoroacetate or fluorocitrate poi­ from a grant from the NIH (51'34-GM-08055) for the training of MARC soned animals, (2R,3R)-2-fluorocitrate seems to be a scholars. mechanism based or "suicide" substrate. This is much more consistent with the data at hand than the originally assumed competitive inhibition. Toxic doses of fluo­ REFERENCES roacetate are of the order of 1 mg/Kg or approx. 10 1. Aoki C., Milner T. A., Sheu K -F. R., Blass J. P., and Pickel J.Lmolar. If only 0.01% is converted to fluorocitrate the V. M. 1987. Regional Distribution of Astrocytes with Intense ..

... 1058 Clarke

Immunoreactivity for Glutamate Dehydrogenase in Rat Brain: Im­ W., Valzelli L., and Peters R. A. 1968. The Effect of Injection plications for Neuron-Glia Interactions in Glutamate Transmis­ of Fluorocitrate into the Brains of Rats. Biochem. Pharmacal. sion. J. Neurosci. 7:2214-2231. 17:195-202. 2. Bergmann, E. D., and Shahak, I. 1960. Triethyl a-fluoroaconi­ 15. Muir D., Berl S., and Clarke D. D. 1986. Acetate and fluoroac­ tate. Nature 185:529-530. etate as possible markers for glial metabolism in vivo. Brain. Res. 3. Berl S., and Clarke D. D., 1969. Compartmentation of Amino 380:336-340. Acid Metabolism. In A. Lajtha (ed), Handbook of Neurochem­ 16. Norenberg M. D. 1983. Immunohistochemistry of Glutamine istry, Vol II, Plenum Press, NY pp. 447-472. Synthetase. pp. 95-111, in L. Hertz, E. Kvamme, E. G. McGeer 4. Catacuzene D., Wakselman C. and Massoudi H. 1984. Fluoroa­ and A. Schousboe (eds. ), Glutamine, Glutamate, and GABA in conitates: Wittig Reaction of Oxalofluoroacetic Acid Esters. Syn. the Central Nervous System. Alan R. Liss, Inc. Comm. 14:1067-1072. 17. Paulsen R. E., Contestabile A., Villoni L., and Fonnum F. 1987. 5. Clarke D. D., Nicklas W. J., and Palumbo J. 1968. Fluorofu­ An in Vivo Model for Studying Function of Brain Tissue Tem­ marate-A Substrate for Fumarate Hydratase. Arch. Biochem. Bio­ porarily Devoid of Gial Cell Metabolism. The Use of Fluoroci­ phys. 123:205-206. trate. J. Neurochem. 48:1377-1385. 6. Clarke D. D., and Berl S. 1983. Fluoroacetate as a Possible Marker 18. Peters R. A. 1972. Some Metabolic Aspects of fluoroacetate es­ for Glial Cell Metabolism in Vivo. Pages 233-240, in Hertz, L., pecially related to fluorocitrate. in Carbon-Fluorine Compounds: Kvamme, E., McGeer, E. and Schousboe, A. (eds.), Glutamine, Chemistry, Biochemistry & Biological Activities (A Ciba Foun­ Glutamate, and GABA in the Central Nervous System. Alan R. dation Symposium) 55-76. Liss, Inc. 19. Peters R. A., and Shorthouse M. 1966. Behavior of Rat Brain 7. Clarke D. D., Nicklas W. J., and Berl S. 1970. Tricarboxylic tissue Treated with Fluoroacetate in Vitro. Biochem. Pharmacol. Acid Cycle Metabolism in Brain: Effect of Fluoroacetate and 15:2130-2131. Fluorocitrate on the Labeling of Glutamate, Asparate, Glutamine 20. Prestwich G. D., Yamaoka R., Phirwa S., and DePalma A. 1984. and -y-Aminobutyrate. Biochem. J. 120:345-354. Isolation of 2-Fluorocitrate Produced by in Vivo Dealkylation of 8. Dumel R. J., and Kun E. 1969. Studies with Specific Enzyme 29-Fluorostigmasterol in an Insect. J. Bioi. Chern. 259:11022- Inhibitors XII. Resolution of DL-erythro-Fluorocitric Acid into 11026. Optically Active Isomers. J. Bioi. Chern. 244:2966-2969. 9. Hornfeldt C. S., and Larson A. A. 1990. Seizures induced by 21. Rivett D. E. A. 1953. The Synthesis of Monofluorocitric Acid. J. Chern. Soc. 3710-3711. fluoroacetic acid and fluorocitric acid may involve chelation of divalent cations in the spinal cord. Eur. J. Pharmacal. 179:307- 22. Teele B., and Casida J. E. 1989. Enzymatic Defluorination and Metabolism of Fluoroacetate, Fluoroacetamide, Fluoroethanol, and 313. 19 10. Rudick J. 1972. Studies of the Toxic Isomer of Fluorocitrate. Ph. (-)-erythro-Fluorocitrate in Rats and Mice Examined by F and 13 D. thesis Submitted to Fordham University. C NMR. Chern. Res. in Toxicol. 2:429-435. 11. Kun E., Kirsten E., and Sharma M. L. 1974. Enzymatic For­ 23. Stallings W. C., Monti C. T., Belvedere J. F., Preston R. K., mation of Glutathione-citryl thioester by a Mitochondrial System and Glusker J. P. 1980. Absolute Configuration of the Isomer of and Its Inhibition by (-)erythrofluorocitrate. Proc. Natl. Acad. Fluorocitrate That Inhibits Aconitase. Arch. Biochem. Biophys. Sci. 77:4942-4946. 203:65-72. 12. Lahiri S., and Quastel J. H. 1963. Fluoroacetate and the Metab­ 24. Villafranca J. J., and Platus E. 1973. Fluorocitrate Inhibition of olism of ammonia in the Brain. Biochem. J. 89:157-163. Aconitase. Reversibility of the Interaction. Biochem. Biophys. 13. Marietta M., Cheung Y., and Walsh C. 1982. Stereochemical Res. Comm. 55:1197-1207. Studies on the Hydration of Monofluorofumarate and 2,3-Difluo­ 25. Walsh C. 1983. Fluorinated Substrate Analogs: Routes of Metab­ rofumarate by Fumarase. Biochemistry. 21:2637-2643. olism and Selective Toxicity. in Adv. in Enzymol. Meister, A. 14. Morselli P. L., Garattini S., Marcucci F., Mussini E., Rewersky (ed.), 55, 197-289.