May 10, 1972 - Guanethidine was the gift of Dr. Albert J. Plummer of Ciba Pharmaceuticals Co.; 1-methyl4 (3- methyl-5-isoxazolyl) pyridinium chloride was ...
Proc. Nat. Acad. Sci. USA Vol. 69, No. 7, pp. 1957-1961, July 1972
Calcium-like Action of Phenethylbiguanide and Related Compounds: Inhibition of Pyruvate Kinase (enzyme kinetics/metal-bridge enzymes/substrate-bridge enzymes)
FRANK DAVIDOFF AND STEPHEN CARR Department of Medicine, Beth Israel Hospital and Harvard Medical School, 330 Brookline Ave., Boston, Massachusetts 02215
Communicated by Carl F. Cori, May 10, 1972
ABSTRACT Pyruvate kinase (EC 2.7.1.40) is inhibited by phenethylbiguanide. The kinetics of inhibition are competitive between biguanide and divalent, but not monovalent, metal cation activators of the enzyme; biguanide inhibition thus resembles inhibition by Ca++. Alteration of either the polar or nonpolar portion of the phenethylbiguanide molecule quantitatively reduces its effectiveness as an inhibitor of pyruvate kinase, but the kinetics of inhibition remain qualitatively unchanged. Measurements of ['HIphenethylbiguanide binding to the enzyme indicate the presence of a single class of about 12 binding sites per enzyme molecule; binding characteristics are not significantly different in the presence of either monovalent or divalent metal cations. Studies with *Ca++ and 54Mn++ demonstrate about 4 metal binding sites per enzyme molecule; phenethylbiguanide displaces these metal cations from the enzyme. Studies with several enzymes, dependent upon divalent metal cations, of both metal-bridge and substrate-bridge classes fail to show significant inhibition except at much higher phenethylbiguanide concentrations.
Biguanides affect the activities of soluble and particulate enzymes, including hexokinase (1, 2), phosphatase (1), and phosphorylase (2), as well as enzymes in the biosynthesis of cholesterol from acetate (3) and the activity of proteolytic clotting factors and trypsin (4). Although no common mechanism underlying these effects has emerged, Hollunger (5) demonstrated that Mg++ reactivated yeast hexokinase inhibited by guanidine, suggesting an interaction with divalent metal ions. We now report that phenethylbiguanide is an effective inhibitor of rabbit muscle pyruvate kinase (ATP: pyruvate phosphotransferase, EC 2.7.1.40). This enzyme is absolutely dependent on both divalent (Mg++, Mn++, or Co++) and monovalent cations (K+, Rb+, or NH4+) for activity (6), and Ca++ competitively inhibits binding of and activation by divalent metal ions (7). Inhibition by phenethylbiguanide is kinetically identical to that induced by Ca++, and the mechanism is shown to be due to the displacement of divalent metal from the enzyme. Structure-function correlations using related guanidine derivatives and organic amines indicate further that binding of phenethylbiguanide to pyruvate kinase and inhibition of enzyme activity depend on both polar and nonpolar interactions. A portion of this work has appeared in abstract form (8). Abbreviations: [Mg++]f and [Mn ++]f, concentrations of free magnesium and manganese. 1957
MATERIALS AND METHODS Rabbit muscle pyruvate kinase was obtained from Boehringer
Mannheim Corp., New York, and freed from NH4+ by passage over Sephadex G-25 in 0.05 M Tris HCl buffer (pH 7.4). Enzyme activity was assayed spectrophotometrically (7) with a Gilford automatic recording spectrophotometer. Concentrations of free Mn++ and Mg++, [Mn++]f and [Mg++]f, in the assay mixture were calculated by the method of Mildvan and Cohn (7). Phosphoglucomutase (a-D-glucose-
1,6-diphosphate: a-D-glucose-i-phosphate phosphotransferase, EC 2.7.5.1), enolase (2-phospho-D-glycerate hydro-lyase, EC 4.2.1.11), hexokinase (ATP: D-glucose-6-phosphotransferase, EC 2.7.1.1), creatine kinase (ATP: creatine phosphotransferase, EC 2.7.3.2), and phosphoglycerate kinase (ATP: 3-phospho-D-glycerate 1-phosphotransferase, EC 2.7.2.3) were obtained from Boehringer Mannheim Corp. and assayed spectrophotometrically by coupling with NAD or NADPlinked reactions. In each assay, the ratio of the indicator enzyme activity was kept very high relative to the primary enzyme being studied (9), and in each instance the presence of the inhibitor did not restrict the activity of the indicator reaction significantly. Concentration of inhibitor required for 50% decrease in enzyme activity was determined graphically from a semi-logarithmic plot of inhibitor concentration against percent.of control activity. Monoguanidine derivatives were synthesized by the method of Phillips and Clarke (10). The products were recrystallized several times and characterized by melting points of the recrystallized sulfate and picrate derivatives. Amines were converted to the hydrochlorides and recrystallized before use. Guanethidine was the gift of Dr. Albert J. Plummer of Ciba Pharmaceuticals Co.; 1-methyl4 (3methyl-5-isoxazolyl) pyridinium chloride was obtained from Dr. A. D. Blickens of Lederle Laboratories; butacaine and isopentenyl guanidine (galegine) sulfate were obtained from K and K Laboratories. Isopentenyl guanidine sulfate was recrystallized twice from ethanol-water. [3H]Phenethylbiguanide was prepared by catalytic exchange and purified by preparative paper chromatography as described (11). Binding of tritiated biguanide and of labeled Ca++ and Mn++ to pyruvate kinase was measured by ultrafiltration (12), with UM-10 filters; 45Ca++ was counted in a liquid scintillation spectrometer and 54Mn++ in a crystal scintillation counter. Binding data are plotted where indicated by the method of Scatchard (13). Pyruvate
Biochemistry: Davidoff and Carr
1958
Proc. Nat. Acad. Sci. USA 69
(1972)
Under standard conditions of 40 mM K+ plus 0.17 mM was achieved at 88 mM Li+; this inhibition, therefore, does not appear to be due to the relatively nonspecific inhibition associated with high ionic strengths, as is seen with K+ concentrations over 150 mM (6). The interaction of Li+ with Mg++ was noncompetitive, while Li+ was uncompetitive with K+. The "nonspecific" monovalent cation, tetramethylammonium chloride, was less effective as an inhibitor, since 130 mM tetramethylammonium chloride was required to achieve 50% inhibition; the kinetic behavior of tetramethylammonium chloride appeared to be similar to that of Li+, but because of its weak inhibitory effect was more difficult to define clearly. The kinetics of phenethylbiguanide inhibition thus bear no resemblance to the kinetics with other monovalent cation inhibitors. Direct competition between phenethylbiguanide and divalent metal cation is further supported by the observation that the apparent Km for ADP increased roughly 2-fold, from 0.16 to 0.28 mM, in the presence of phenethylbiguanide, an effect identical to that observed with calcium, and which is the behavior expected when substrate binds mainly to the enzyme-metal complex (17).
Mg++, 50% inhibition
-10
0
10
20
[hz y
-50
0
50
100
~~~~~~~~//Mn+tlf
FIG. 1. Kinetics of inhibition of pyruvate kinase by phenethylbiguanide and Ca++. Enzyme activity was assayed spectrophotometrically in a volume of 1.5 ml; V is expressed in nmol of NADH formed per min per cuvette. [Mn++]f was calculated as described in Methods. Lines were fitted by least squares.
kinase was quantitated by measurement of absorbance at 280 nm, and converted to milligrams by the correction factor of Bucher and Pfleiderer (14); the molecular weight of pyruvate kinase was assumed to be 237,000 (7). RESULTS Inhibition of pyruvate kinase by phenethylbiguanide, calcium, lithium, and tetramethylammonium chloride Under standard assay conditions, which included 40 mM K+ and 0.16 mM Mn++, phenethylbiguanide was an effective
inhibitor of pyruvate kinase, 50% inhibition occurring at about 2.5 mM phenethylbiguanide. The degree of inhibition depended critically on the cation composition of the medium. Thus, kinetic studies indicated purely competitive interaction between Mn++ and phenethylbiguanide, while the interaction between K+ and phenethylbiguanide was mixed in type (Fig. 1). Identical kinetic observations were made with magnesium over a range of free Mg++ concentrations from 0.15-7 mM. As demonstrated by others (6), Ca++ also inhibited pyruvate kinase activity in this assay, and the kinetic interactions of Ca++ with Mn++ and K+ were identical to those of phenethylbiguanide (Fig. 1). When measured in the same experiment, the Lineweaver-Burke plots for 1/ K+ with Ca++ and with phenethylbiguanide both intersected the control plot at precisely the same point. By the graphic methods of Dixon (15), the Ks for phenethylbiguanide in the presence of 150 mM K+ was calculated to be about 8 X 10-3 M; in contrast, the Ki for Ca++, in terms of total rather than free Ca++, has been determined as 1.4 X 10-4 M (7). The great bulk (i.e., 99.99%) of biguanide in solution at pH 7.4 exists as the monovalent cation since the second pKa of biguanides is about 3.5 (16); the kinetic behavior of biguanides might therefore have been expected to resemble that of other inhibitory monovalent ions such as Na+ or Li+ (6). However, kinetic data for these two metal ions as inhibitors of pyruvate kinase were not available. The kinetics of Li+ as an inhibitor of pyruvate kinase were therefore examined.
Structural requirements for inhibition by phenethylbiguanide Alteration of the structure of the nitrogenous portion of the phenethylbiguanide molecule to the corresponding guanidine and amine progressively decreased the inhibitory potency for pyruvate kinase (Table 1), but both derivatives showed competitive kinetics with Mn++ and mixed kinetics with K+. Changes in the hydrophobic portion of the molecule also
TABLE 1. Inhibition of pyruvate kinase by structural analogues of phenethylbiguanide and related guanidine and amino derivatives
50% inhibitory Compound Phenethylbiguanide HC1 Phenethylguanidine S04 Phenethylamine HCI Biguanide HC1 Biguanide SO4 Dimethylbiguanide HCl Butylbiguanide HCl Octylguanidine S04 Octylamine HCl Isopentenyl guanidine SO4 Guanethidine SO4
concentration
Apo
(mM)*
(Cal)t
2.5 13 25 150 130 275 78
-35$ 75 37 26
-3.51
-2.54 -2.16 -1.15 -1.15 -0 . 76 -1.50
'-1.97 -1.52 -1.93 -2.14
* Enzyme activity was assayed spectrophotometrically at standard conditions of 40 mM K+ and 0.16 mM Mn++. t Free energy of binding calculated from the equation AF0 = - 1fTlnKO, where R = gas constant, T = absolute temperature, and Ko = association constant that was determined as the reciprocal of inhibitor concentration required for 50% inhibition, on the assumption that at this concentration 50% of the binding sites for the inhibitor were occupied. $ Limited solubility of octylguanidine required the use of ethanol to attain significant concentrations; ethanol was also added to controls; 50% inhibitory level of octylguanidine was determined by extrapolation.
Proc. Nat. Acad. Sci. USA 69
Calcium-like Action of Biguanides
(1972)
affected inhibitory capacity. Thus, biguanide inhibited pyruvate kinase no more effectively than tetramethylammonium chloride, while dimethylbiguanide was actually less effective than biguanide itself; again, the kinetic interactions of biguanide and dimethylbiguanide qualitatively resembled those of phenethylbiguanide. Butylbiguanide, which is equipotent with phenethylbiguanide as an inhibitor of mitochondrial respiration in vitro and as a hypoglycemic agent, was markedly less effective than phenethylbiguanide as an inhibitor of pyruvate kinase. Octylguanidine and octylamine, the n-alkyl compounds containing the same number of side-chain carbon atoms as phenethylbiguanide, were about one-third as potent inhibitors of the enzyme as the corresponding phenethyl derivatives (Table 1). Isopentenyl guanidine, which also demonstrates hypoglycemic activity, and is an effective inhibitor of respiration of isolated mitochondria, inhibited pyruvate kinase at relatively high concentrations; its kinetic interactions with metal cations, although resembling those for the other monoguanidine derivatives, were somewhat more complex. Guanethidine also inhibited pyruvate kinase, with its kinetic behavior resembling the phenethyl compounds. 1-methyl-4 (3-methyl-5-isoxazolyl) pyridinium chloride, a compound with hypoglycemic properties somewhat similar to those of biguanides, inhibited the enzyme no more effectively than tetramethylammonium chloride (i.e., as determined by extrapolation, 50% inhibition at about 150 mM), while butacaine, a local anesthetic structurally very similar to tetracaine, and which, like biguanides, is an inhibitor of hepatic gluconeogenesis, was totally without effect on pyruvate kinase at concentrations up to 20 mM. Binding of [3H]phenethylbiguanide to pyruvate kinase Direct measurement of the amount of [3H]phenethylbiguanide bound to pyruvate kinase indicated that binding paral-
1959
I/
.11 20
-
45CaOBinding
'_1,i
2
IF
3.
5
\
l1 lo
-
0. ~~~~~~~~~~~~~~~~~~2.( .1.0
5
0
2
4
6 0
0.5
Nfi-If -,A
1.0
0
50
100
fA(Sfh7ZfRMI#1/A-,wAf
FIG. 3. Binding of 45Ca++ to pyruvate kinase. Binding was determined by the ultrafiltration technique in a medium containing 0.05 mM Tris HOl (pH 7.4) but no nucleotide or substrate; where indicated (0) 75 mM KCl was added (0, no K+). Each point in A represents the mean of 5-8 separate experiments for the data without KCl; all other data represent means of 1-4 experiments. Symbols are as in Fig. 2. In B and C, 45Ca++ was present at a concentration of 0.25 mM.
lelled inhibition, with saturation at higher biguanide concentrations (Fig. 2, left). A Scatchard plot of the same data (Fig. 2, right) suggested that the enzyme contained about 12 binding sites per enzyme molecule; the stability constant for binding was about 0.24 liters mmolh'. The presence of monovalent cation (K+ or tetramethylammonium chloride) exerted no significant effect on the binding characteristics of phenethylbiguanide (Fig. 2). When added in the presence of 75 mM K+, manganese at lower concentrations (up to 0.13 mM of the free ion), produced no significant displacement of bound biguanide (present at a concentration of 4.4 mM). It should be noted that when pyruvate kinase activity was assayed spectrophotometrically in the presence of similar concentrations of phenethylbiguanide, at least 60% of the inhibition observed at a free manganese concentration of 0.012 mM was reversed by raising the free manganese concentration to 0.25 mM. At free manganese concentrations of 1.6 and 6.2 mM, the mean quantity of bound biguanide was reduced by 27 and 38%, respectively, over the control; however, because of large standard error, these latter results failed to reach the level of statistical significance (results of 6-10 separate experiments). Binding of divalent metal cations to pyruvate kinase
4
8
12
[PHENEMYLB /GUAN/DE
16 -mM
FIG. 2. Binding of ['H]phenethylbiguanide to pyruvate kinase. Binding studies were done with an ultrafiltration cell (see Methods), in the presence of 0.05 M Tris HCl buffer (pH 7.4)-3.3 mM ADP-0.8 mM phosphoenolpyruvate, and, where indicated, 0.1 M KCl or tetramethylammonium chloride. About 1 nmol (0.24 mg) of pyruvate kinase was present in each filtration cell in a total volume of 0.15-0.3 ml; duplicate or triplicate cells were run in each experiment, and controls containing no enzyme were done simultaneously for each experimental condition. Each point represents the mean of 1-7 separate experiments. v mol of ligand bound per mol of enzyme; A medium concentration of ligand in mM. *, tetramethylammonium chloride; A, K +; *, no cation. =
=
At Ca++ concentrations lower than 0.1 mM, the Scatchard plot of calcium binding to pyruvate kinase deviated somewhat from linearity (Fig. 3A), possibly because of nonspecific adsorption of Ca++ to the glass and plastic surfaces, although cooperativity of binding could not be excluded. In the absence of K+, Ca++ was probably bound to a single class of about 4 binding sites per pyruvate kinase molecule with a stability constant of about 5.1 liters mmolh', although the line is drawn in Fig. 3A somewhat arbitrarily. The relative affinities of Ca++ and phenethylbiguanide for pyruvate kinase as measured by direct binding thus compare quite well with the relative magnitude of the Ki values (see above). In the absence of K+, Mn++ diminished Ca++ binding to the enzyme, 50% of the bound Ca++ being displaced at a Mn++ concentration of 0.08 mM (Fig. 3B). Phenethylbiguanide also diminished Ca++ binding to the enzyme, 50% displacement of Ca++ being achieved at a concentration of about 15 mM biguanide (Fig. 3S). This displacement could not be attributed
Biochemistry: Davidoff and Carr
1960
A
54Mn+
Binding
Proc. Nat. Acad. Sci. USA 69 B
200
150
'0 .
100
\0
50F
.
0
k, 0
1
2
3
4
5
Z)
6
JL
0 10 20 30 40 50 60
[PHENETIMLB4/GUAN/uE-m,#
FIG. 4. Binding of 14Mn++ to pyruvate kinase. Conditions 54Mn++ was present at 0.05 mM in B.
were described in Fig. 3; *, no K +; O, 75 mM K+.
to ionic strength effects, since 50 mM tetramethylammonium chloride was found to displace only 12% of the bound Ca++. Inclusion of 75 mM K+ in the incubation mixture consistently reduced Ca++ binding by about 17% over a wide range of Ca++ concentrations, with a resultant shift in stability constant to 3.6 (Fig. 3A). In the presence of K+, 50% displacement of bound Ca++ was achieved by Mn++ at about the same concentration as in the absence of K+ (Fig. 3B). Measurement of 54Mn++ binding to pyruvate kinase also revealed a single class of 4 binding sites with a stability constant of about 24 liters mmol-h (Fig. 4A), which is in agreement with the somewhat greater affinity of the enzyme for Mn++ than for Ca++ as determined by other methods (7); deviation of the Scatchard plot from linearity was not observed. In the presence of 75 mM K+, the stability constant for Mn++ binding doubled, to 53 liters mmolh'; phenethylbiguanide, added in the presence of K+, displaced Mn++ from the enzyme, 50% displacement occurring at a biguanide concentration of about 12 mM (Fig. 4B). This displacement was not an effect of ionic strength, since 50 mM tetramethylammonium chloride actually increased Mn++ binding by 20%. When phenethylbiguanide was added under the standard conditions described above for kinetic measurements of inhibition (40 mM K+ and 0.16 mM Mn++), 2.5 mM phenethylbiguanide displaced 23% of the bound Mn++. Inhibition of other divalent metal cation-dependent enzymes by phenethylbiguanide On the basis of measurements of Mn++ binding, Mildvan and Cohn have classified divalent metal cation-dependent enzymes into metal-, substrate-, and enzyme-bridge groups
(18). Inhibition of pyruvate kinase, a metal-bridge enzyme, by phenethylbiguanide, as well as the above kinetic and binding data indicating direct biguanide binding to a divalent metal site on the enzyme, suggested that if biguanide inhibition occurred with other enzymes it might be limited to members of the metal-bridge group. Several metal-bridge and several substrate-bridge enzymes were, therefore, screened
(1972)
for susceptibility to inhibition by phenethylbiguanide. Phenethylbiguanide did not affect the activity of the two other metal-bridge enzymes or any of the three substratebridge enzymes until concentrations of the biguanide were nearly an order of magnitude greater than that which inhibited pyruvate kinase. Phosphoglucomutase, a metalbridge enzyme, was inhibited 50% by 75 mM phenethylbiguanide, which may be a specific, calcium-like effect, since under the same conditions 10 mM calcium inhibited phosphoglucomutase by 85% (19). In the case of the substratebridge enzymes, hexokinase, creatine phosphokinase, and 3-phosphoglycerate kinase, however, 50% inhibition was achieved only at phenethylbiguanide concentrations of 170 mM, 83 mM, and 47 mM, respectively; this inhibition was probably nonspecific, since Ca++, which is normally an activator of the latter two enzymes, was also inhibitory in these concentration ranges. Phenethylbiguanide had no effect on enolase, a metal-bridge enzyme, in concentrations up to 50 mM. DISCUSSION
Inhibition of pyruvate kinase by phenethylbiguanide and certain related derivatives appears to result from displacement of divalent metal ions from the enzyme; this displacement is highly specific, as evidenced by (i) close resemblance between Ca++ and guanidine-derivative inhibition kinetics, (ii) lack of kinetic similarity between lithium and guanidinederivative inhibition, and (iii) binding data with labeled biguanide, Ca++, and Mn++. The finding that about 12 biguanide molecules are bound per molecule of enzyme is of some interest in view of the previous suggestion that 4 divalent metal-ion-binding sites are present per enzyme molecule (20), a conclusion supported by our own binding data (Figs. 3 and 4). The monotonic slope of the Scatchard plot indicates a single class of phenethylbiguanide binding sites (Fig. 2), presumably the same sites that bind divalent metal ion. At saturation, then, each divalent metal-ion-binding site may possibly accommodate up to 3 biguanide molecules, perhaps one biguanide associated with each of the three coordination groups estimated to be present at the divalent cation site (20). Triethyltin, an alkyl-substituted cation whose inhibition of mitochondrial respiration and phosphorylation closely resembles biguanide inhibition, binds to paired histidine residues both in purified proteins and in model systems (21), suggesting the possibility that histidines may also be involved in the coordination of divalent metals and of biguanides to pyruvate kinase*. Although biguanides chelate certain divalent metal cations, such as Co++, Ni++, and Cu++, chelation is effectively ruled out as the mechanism of pyruvate kinase inhibition, since biguanides do not chelate either Mg++ or Mn++ (22). The small amount of doubly protonated phenethylbiguanide present at pH 7.4 cannot be the molecular species responsible for inhibition, since the monoguanidine and amino derivatives, which do not form such divalent species, are also inhibitory. As estimated from the 50% inhibitory concentrations of phenethylamine, -monoguanide and -biguanide, the polar group in the phenethylbiguanide molecule is clearly of some importance in determining binding affinity to pyruvate Preliminary studies in this laboratory [Davidoff, F. & Carr, S. (1972) Fed. Proc. 31, 853 Abstr.] indicate that triethyltin is also an effective inhibitor of pyruvate kinase. *
Proc. Nat. Acad. Sci. USA 69
(1972)
kinase (Table 1). However, the change in structure from amine to biguanide only changes the binding energy about 30%, while introduction of the phenethyl group contributes 70% (Table 1). The ineffectiveness of monovalent or divalent metal cations in displacing bound phenethylbiguanide from the enzyme (see Fig. 2 and text) is consistent with a major dependence of binding on hydrophobic interactions; restoration of enzyme activity with increasing Mn++ concentrations despite continued binding of biguanide suggests that, while the nitrogenous base is displaced from the cation-binding site itself, the hydrophobic portion of the molecule remains bound to adjacent regions of the enzyme. Steric factors also appear to influence binding affinity, since the addition of two methyl groups to the N1 nitrogen of biguanide actually decreases biguanide inhibitory effectiveness below that of the "nonspecific" cation, tetramethylammonium. This steric effect is strongly reminiscent of the diminished effectiveness of N1, N1-dimethylguanidine compared with guanidine itself as an inhibitor of mitochondrial respiration and ATPase activity (5). Hydrophobic interactions not only contribute importantly to the binding affinity of organic bases for pyruvate kinase but also alter the specificity of action, as evidenced by the conversion of an activating monovalent cation (NH,+) to an inhibitory one (phenethylamine) upon addition of a phenethyl or acyl group, presumably by altering the locus on the enzyme to which the cation is bound. Differences in inhibitory potency (23) as well as inhibitory specificity (24) in mitochondrial systems have previously been observed with changes in the hydrophobic N-substituent of mono- and biguanide derivatives. The biological implications of "calcium-like" behavior of biguanides may be considerable. However, prediction of guanidine derivative effects on any particular biological activity is difficult, since the outcome depends on whether Ca++ itself or some other divalent metal cation is displaced, and whether the bound biguanide resembles Ca++ sufficiently in all critical physical chemical parameters to substitute for it completely. Nevertheless, many actions of guanidine derivatives on subcellular fractions, whole tissues, or intact animals may be reinterpreted in terms of calcium-like behavior. Thus, the inhibition of respiration and stimulation of inorganic phosphate uptake by mitochondria that accompanies the uptake of guanidine derivatives (23, 25) may result from interaction of these compounds with the same binding and transport system that handles Ca++ (26). In intact animals, intravenous guanethidine and phenethylguanidine bring about acute release of catecholamines (27), which may in part be due to adrenal medullary catecholamine release, since Ca++ is a potent secretagogue for the adrenal medulla (28). In view of the Ca++ requirement for activation of target organs dependent on catecholamines and cyclic-3': 5'-AMP, it is also of interest that the response of rabbit uterus and ileum to epinephrine can be specifically blocked by phenethylamine, phenethylguanidine, and phenethylbiguanide, as well as by several other biguanide derivatives
Calcium-like Action of Biguanides
1961
(27). While interference with catecholamine storage, release, and action may underlie both the antihypertensive actions of guanethidine and its effects on carbohydrate metabolism (29), the hypoglycemic activity of the biguanides may be mediated directly by "calcium-mimetic" activity at several other tissue sites. Because the binding of guanidine derivatives to lipid-containing membrane structures appears to be stronger than binding to soluble proteins (25), guanidine derivative activity in vivo is probably exerted largely at membrane loci, rather than on soluble enzymes. This investigation was supported in part by Research Grants AM 10541 and FR 05479 from the National Institutes of Health, U.S. Public Health Service, and a Research Grant from the American Diabetes Association. F.D. is the recipient of Career Development Award 1 KO 4 AM 34961 from the National Institutes of Health. 1. Dobson, H. L. & Allen, M. B. (1967) Excerpta Med. 140, 139. 2. Hildmann, W. & Zillman, R. (1965) Acta Biol. Med. Ger. 15, 375-380. 3. Dempsey, M. E. (1968) Ann. N.Y. Acad. Sci. 148, 631-646. 4. Roberts, P. S. & Fleming, P. B. (1972) Fed. Proc. 31, 218 Abstr. 5. Hollunger, G. (1955) Acta. Pharmacol. Toxicol. 11, 14-24 (Suppl.). 6. Kachmar, J. F. & Boyer, P. D. (1953) J. Biol. Chem. 200, 669-682. 7. Mildvan, A. S. & Cohn, M. (1965) J. Biol. Chem. 240, 238246. 8. Davidoff, F. (1971) Diabetes 20, 325. 9. Bergmeyer, H.-U. (1955) in Methods of Enzymatic Analysis, ed. Bergmeyer, H.-U. (Academic Press, New York), pp. 10-13. 10. Phillips, R. & Clarke, H. T. (1923) J. Amer. Chem. Soc. 45, 1755-1757. 11. Davidoff, F. (1968) J. Clin. Invest. 47, 2344-2358. 12. Paulus, H. (1969) Anal. Biochem. 32, 91-100. 13. Scatchard, G., Coleman, J. S. & Shen, A. L. (1957) J. Amer. Chem. Soc. 79, 12-20. 14. Bucher, T. & Pfleiderer, G. (1955) in Methods in Enzymology, eds. Colowick, S. P. & Kaplan, N. 0. (Academic Press, New York), Vol. I, pp. 435-440. 15. Dixon, M. & Webb, E. C. (1964) Enzymes (Academic Press, New York), p. 329. 16. Doornbos, D. A. (1967) Pharm. Weekbl. 102, 269-287. 17. Dixon, M. & Webb, E. C. (1964) Enzymes (Academic Press, New York), p. 432. 18. Mildvan, A. S. & Cohn, M. (1970) Advan. Enzymol. 33, 1-70. 19. Cori, G. T., Colowick, S. P. & Cori, C. F. (1938) J. Biol. Chem. 124, 543-555. 20. Reuben, J. & Cohn, M. (1970) J. Biol. Chem. 245, 65396546. 21. Rose, M. S. (1969) Biochem. J. 111, 129-137. 22. Doornbos, D. A. (1968) Pharm. Weekbi. 103, 1213-1227. 23. Pressman, B. C. (1963) J. Biol. Chem. 238, 401-409. 24. Schafer, G. (1969) Biochim. Biophys. Acta. 172, 334-337. 25. Davidoff, F. (1971) J. Biol. Chem. 246, 4017-4027. 26. Carafoli, E. & Lehninger, A. L. (1971) Biochem. J. 122, 681-
690. 27. Kroneberg, G. & Stoepel, K. (1958) Arzneim. Forsch. 8, 470475. 28. Rubin, R. P. (1970) Pharmacol. Rev. 22, 389-428. 29. Kansal, P. C., Buse, J., Durling, F. C. & Buse, M. G. (1971) Curr. Ther. Res. CGin. Exp. 13, 517-522.