FEMS Microbiology Letters 119 (1994) 283-288 © 1994 Federation of European Microbiological Societies 0378-1097/94/$07.00 Published by Elsevier

283

FEMSLE 06003

Quasi-irreversible inhibition of enolase of Streptococcus mutans by fluoride

T i m o t h y M. Curran *'a, Daniel H. Buckley a and R o b e r t E. Marqu i s a,b

Departments of a Microbiology and Immunology and of h Dental Research, University of Rochester School of Medicine and Dentistry, Rochester, IVY 14642-8672, USA

(Received 28 March 1994; revision received and accepted 4 April 1994)

Abstract: Fluoride at concentrat ions greater than 0.01 mM was found to be a quasi-irreversible inhibitor of enolase of permeabilized cells of Streptococcus mutans GS-5 and also of isolated yeast enolase. The inhibition appeared to be of the type that has been described for P-ATPases, but was not dependent on added AI 3+ or Be z+ ions. Fluoride inhibition of enolase was not reversed by repeatedly washing the permeabilized cells in chilled fluoride-free medium but could be reversed by the product, phosphoenolpyruvate, or by very high levels of the substrate, 2-phosphoglycerate. Irreversible inhibition of glycolysis was not evident after fluoride t reatment of intact cells, washing to remove unbound or loosely bound fluoride and addition of glucose, presumably because intracellular levels of phosphoenolpyruvate were sufficiently high to preclude irreversible fluoride inhibition of enolase.

Key words: Enolase; Fluoride; Streptococcus mutans

Introduction

Fluoride inhibition of enolase has been stud- ied extensively starting with the investigations of Warburg and Christian [1], who first showed that fluoride inhibits the isolated enzyme and that the degree of inhibition for yeast enolase could be predicted from a knowledge of F - , Mg 2÷, and phosphate levels. Their findings have been aug- mented and refined over the years with use of enzymes from a variety of organisms, with more precise kinetic analyses [2] and with detailed study

* Corresponding author. Tel: (0716) 275-1674; Fax: (716) 473- 9573.

of enzyme-metal-fluoride-phosphate complexes [3].

Fluoride is well known as an anticaries agent, and at least part of its effectiveness is thought to be due to inhibition of glycolytic acid production by cariogenic bacteria such as Streptococcus mu-

tans. It serves also to enhance remineralization of enamel and so appears to have dual action in reducing caries [4]. Fluoride inhibition of glycoly- sis in oral streptococci is thought to be due pri- marily to inhibition of enolase. Changes in cyto- plasmic levels of glycolytic intermediates after exposure of cells to fluoride suggest that such inhibition occurs in intact ceils and that sec- ondary inhibition of sugar uptake by the phos- photransferase system further reduces glycolysis

SSDI 0 3 7 8 - 1 0 9 7 ( 9 4 ) 0 0 1 4 9 - L

284

[5,6]. However, fluoride can act also as a weak acid to enhance t ransmembrane proton conduc- tivity with resultant acidification of the cytoplasm of bacteria in acid environments, and this acidifi- cation can adversely affect glycolytic enzymes [7].

Recent studies of the inhibitory effects of fluo- ride on enolase isolated from mutans streptococci [8,9] have led to the conclusion that inhibition is complex and has both competitive and noncom- petitive components. These findings of complex inhibition led us to reinvestigate fluoride inhibi- tion of enolase but with use of permeabil ized cells that allow for easy washing of the inhibited enzyme to remove unbound or loosely bound fluoride. Our initial view was that aluminofluo- ride could be involved in inhibition, as it is for the F-ATPase of S. mutans [10]. However, we found that inhibition of enolase more closely resembled fluoride inhibition of P-ATPases of mammalian cells, and that aluminofluoride is not essential for inhibition.

Materials and Methods

Bacteria Streptococcus mutans GS-5 was grown rou-

tinely in brain-heart infusion broth (Difco Labo- ratories, Detroit, MI) with 2% (wt/vol) added glucose in static culture at 37°C. Cells were har- vested during the early stationary phase of growth by means of centrifugation in the cold. Stock cultures of the organism were stored at - 70°C in 50% (w/v) glycerol solution.

Enolase assays Permeabil ized cells were prepared from cul-

tures in the early stationary phase of growth. 25-ml culture samples were centrifuged in the cold, and each pellet was resuspended in 2.5 ml 75 mM Tris-HCl buffer, pH 7.0, with 10 mM MgSO 4. Toluene (250 ~1) was added prior to vigorous vortex mixing and incubation for 5 min at 37°C. The cells were then subjected to two cycles of freezing in a dry-ice-ethanol bath with thawing at 37°C. The cells were harvested by centrifugation, resuspended in 1.0 ml of Tris buffer with magnesium, frozen rapidly and stored

at -70°C. The suspensions were thawed at 37°C before use.

1 ml samples of permeabil ized cell prepara- tions were suspended in 1.0 ml of 50 mM Tris- maleate, pH 7, containing 20 mM MgC12 and 1.0 mM K / N a P O 4. The enolase reaction was started by addition of 2-phosphoglycerate (2PG) to yield a final concentration of 22.7 mM. 100 ~1 samples were removed at intervals and centrifuged for 1 min in a microcentrifuge. The phosphoenolpyru- vate (PEP) formed was assayed using pyruvate kinase and lactic dehydrogenase (Sigma Chemical Co., St Louis, MO). A 50 /xl sample of super- natant fluid was mixed with 950 ml of a solution containing 0.2 M KPO 4, pH 7.4, 1.0 mM MgC1 z, 0.3 mg N A D H per ml and 10 mM ADP. The initial absorbance of the mixture for 340 nm light was determined. Then 20 jxl of a solution con- taining 700 units pyruvatc kinase per ml and 1,000 units lactate dehydrogenase per ml was mixed into the cuvette. The final absorbance was determined when the change in absorbance had become less than 0.001 optical density uni ts /min. Enolase activity was calculated in terms of ixmol PEP f o r m e d / m i n / m g cells.

Yeast enolase (Sigma Chemical Co., St Louis, MO) was assayed similarly only the samples were heated at 80°C for 5 min to inactivate the enzyme prior to enzymatic assay of PEP.

Results

Inhibition of enolase in permeabilized cells In our initial experiments, we found that pro-

duction of PEP by permeabilized cells of S. mu- tans GS-5 was inhibited by fluoride in a way suggestive of irreversible inhibition. Rates of PEP production in fluoride-treated samples were re- duced initially in relation to the fluoride level but then, after a time, production stopped, even though the PEP level was well below the final value for untreated permeabilized cells and the substrate 2PG was still present. The data in Fig. 1 show the course of PEP production by permeabi- lized cells that had been initially incubated with 0.05 mM NaF, 20 mM Mg 2÷ and 1 mM potas- sium phosphate in 50 mM Tris maleate, pH 7.0,

6

5

4 " o

t.u CL

2

z:l. 1

0 [ ~ 10 20

Minutes

Fig. 1. Quasi-irreversible inhibition of enolase of permeabi- lized cells of S. m u t a n s GS-5 by 0.05 mM NaF. Permeabilized cells were incubated at room temperature in 50 mM Tris maleate buffer, pH 7.0, containing 20 mM MgCI2, 1 mM potassium phosphate and 0.05 mM NaF. Samples were re- moved at 0 ( n ) , 5 ( i ) , 10 (A), 15 (× ) , 20 ( 0 ) and 30 (El) min, the cells were washed and assayed for production of PEP

from added 22.7 mM 2PG.

for various periods of time before assay of eno- lase activity. In other words, the cells were first t reated with fluoride, then washed twice with water to remove fluoride, and the enolase activity of the washed cells assayed. After only 5 min of exposure to fluoride, the subsequent initial rate of PEP production was reduced to only about 25% of that of the control rate for cells not exposed to fluoride. (In these experiments, con- trol preparat ions always showed a maximum PEP production at about 5 min followed by a slow decline, probably due to slow metabolism of PEP by the permeabil ized cells.) As shown, incubation of the permeabil ized ceils in the fluoride-contain- ing solution for 30 min resulted in essentially complete inactivation of the enzyme. When Mg 2+ and phosphate were not included in the inactiva- tion mixtures, there was little or no detectable change in the course of inactivation. Also, Tris maleate buffer appeared to play no essential role in the inactivation, and inactivation occurred even in suspensions without buffer in which the pH value was maintained due to buffering by the cells.

The loss of activity followed first-order kinet- ics, and the plot in Fig. 2 is of average k values

285

versus fluoride concentration. The k values were obtained from plots of log % activity versus time of exposure in minutes, k then is simply the slope of the line. It appears that k increases in an approximately linear fashion with increase in flu- oride concentration, although there is a slight upward inflection in the plot of Fig. 2.

Addition of 10 to 50 izmol of A1C13 along with the fluoride to the inactivation mixture had, on average, only a slight enhancing effect on inacti- vation. A13 + itself at these concentrations was not inhibitory for enolase. Also, the aluminum chela- tor deferoxamine at a concentration of 0.5 mM did not protect the enzyme against fluoride inac- tivation, as we had found previously for fluoride inhibition of F-ATPase [10]. In addition, BeSO 4 also did not significantly enhance inhibition. In- activation of the enzyme was not enhanced by addition of 2PG during the initial incubation with fluoride, and it appeared that the enzyme did not have to be catalytic ally active for inhibition to occur.

Inhibition was optimal at a pH value of about 6 but was not highly pH-dependen t in the 5 to 7 range. For example, average % inactivation val- ues caused by 0.05 mM NaF without added Mg were found to be 97%, 79% and 71% after 5-min exposure at pH values of, respectively, 6, 5 and 7.

0.3

~" 0.2

E

-~ 0.1

0 .0 I I I I I I i I I I I

0.00 O.Ot 0.02 0.03 0.04 0.05 0.06

mM F

Fig. 2. Effect of fluoride concentration on the first-order rate constant for inhibition of enolase. Cells were incubated with fluoride in the presence of 20 mM MgCI 2 and 1 mM potas- sium phosphate in 50 mM Tris maleate buffer at pH 7. Samples were removed, washed twice and assayed for produc- tion of PEP from 22.7 mM 2PO. The k values were assessed

from plots of log % activity versus time.

286

These values are reasonably within the range of expected cytoplasmic pH values for glycolysing cells of S. mutans GS-5.

Re~,ersibility of inhibition In experiments in which the enzyme was ini-

tially inactivated by incubation with fluoride, and then any free or loosely bound fluoride was washed away before the enzyme was incubated overnight at 4°C, there was no recovery of activ- ity. Enzyme not exposed to fluoride remained active during the manipulations. However, when the same procedure was followed only now with addition of 22.7 mM PEP to the wash solutions, the level of enzyme activity increased to be nearly equal to that of permeabil ized cells put through the same manipulations only without fluoride (Fig. 3). The substrate 2PG was much less effective in reversing fluoride inhibition, and levels of some 227 raM, or ten times those of the standard assay for enolase activity, were required for reversal of fluoride inhibition. When inactivated enzyme was incubated overnight at room temperature , activity was slowly regained.

8

"¢ 6

£2.

2

0 110 20

Minutes

3 0

Fig. 3. Reversal by PEP or 2PG of quasi-irreversible fluoride inhibition of enolase of permeabilized cells of S. m u t a n s

GS-5. The enzyme was inactivated with 0.1 mM NaF, then the cells were washed and incubated at 4°C overnight with buffer only (~), with 22.7 mM PEP ( I ) , with 22.7 mM 2PG ( 0 ) or with 227 mM 2PG (×) . After incubation overnight, the cells were washed twice and assayed for production of PEP from 22.7 mM 2PG. Data are shown also for permeabilized cells not exposed to fluoride but carried through the washing and

incubation procedures ( [] ).

5

4

1

M

20 40

M i n u t e s

Fig. 4. Pattern of inhibition of yeast enolase by NaF. The purified enzyme (0.05 units) was incubated with 22.7 mM 2PG and 0 ([]) , 0.01 ( 0 ) , 0.10 ( I ) or 1.00 (zx) mM NaF. At intervals, samples were taken, heated to inactivate the enzyme

and assayed for PEP.

Effects of fluoride on yeast enolase As shown by the data presented in Fig. 4,

isolated enolase from yeast was inactivated by fluoride in much the same manner as was the enzyme of permeabilized cells of S. rnutans. Thus, 0.1 mM NaF not only slowed the initial rate of PEP production but also stopped production of PEP after 20 minutes of incubation, even though the substrate 2PG still was present.

Effects of fluoride on glycolysis by intact cells In an at tempt to determine if irreversible inac-

tivation of enolase occurred within glycolysing ceils exposed to fluoride, we suspended cells of S. mutans GS-5 in a solution containing 50 mM KC1 and 1 mM MgCI 2. Companion suspensions con- tained also 10 mM NaF. Samples of the suspen- sions were withdrawn at 0, 0.5, 1, 2, 3 and 4 h, washed to remove fluoride, and the cells were resuspended in fresh KC1-MgC12 solution now with 1% (wt/vol) glucose but no fluoride. The pH drop in these suspensions was then followed. Average final pH values after 105 min in the suspensions were 4.1 with a range of values only from 4.08 to 4.13 in the various suspensions. Clearly the, incubation of intact cells in the pres- ence of 10 mM NaF for as long as 4 h had no effect on the glycolytic capacities of the bacteria. Fluoride itself was highly inhibitory for glycolysis, and if 10 mM NaF was added to cell suspensions

with excess glucose, the pH value fell to an aver- age of 6.1, compared with the value in these experiments of 4.1 for cells not exposed to fluo- ride. Other variants of the experiment involved lowering the pH value of the suspensions exposed to fluoride, adding 1% glucose during the initial incubation with fluoride to allow the cells to develop ApH across the cell membrane and vary- ing the fluoride concentration from 1 to 50 mM. However, there was still no indication of irre- versible damage to the glycolytic system.

Discussion

The data presented show clearly that fluoride can irreversibly inhibit enolase of permeabilized cells of S. m u t a n s GS-5, and this type of inhibi- tion may be related to the non-competitive inhibi- tion reported by others [8,9]. The term 'irreversi- ble' has to be qualified because of the finding that inhibition could be reversed by phospho- enolpyruvate or by very high concentrations of 2-phosphoglycerate. Presumably, fluoride binds tightly to the catalytic site, and this binding is not readily reversed during washing in the absence of product. Tight binding of fluoride did not require substrate, nor did it require magnesium or phos- phate, which are thought to bind separately from substrate [11].

The type of fluoride inhibition described here is similar in many ways to fluoride inhibition of P-ATPases, for example, the calcium ATPase of sarcoplasmic reticulum [12] or the Na /K-ATPase of mammalian tissue [13]. However, there are differences. For example, inhibition of the Ca- ATPase is very dependent on Mg 2+ and phos- phate and is enhanced by beryllium ion. How- ever, fluoride inhibition of enolase was not greatly affected by not adding Mg 2+ or phosphate to the inactivation mixture, and aluminum or beryllium ions did not enhance inactivation. Again, the inactivation of enolase appeared to be due to tight binding of fluoride to the catalytic site. Knowledge of this reaction is useful for studies of catalytic mechanism and of reversible inhibition by fluoride.

That irreversible inhibition of glycolysis by in-

287

tact cells did not occur is not surprising in view of the finding that PEP can reverse fluoride inhibi- tion of enolase. Bacteria are known to pool PEP even during times of starvation, presumably to allow for rapid uptake via the phosphotransferase system of sugars that may become available. S. m u t a n s cells poisoned with fluoride have in- creased cytoplasmic levels of 2PG and also have little if any reduction in levels of PEP [6]. Thus, the product, but also possibly the substrate, would act to reduce the quasi-irreversible type of fluo- ride inhibition occurring in permeabilized cells.

Acknowledgements

This research was supported by grant 5R01 DE06127 from the National Institute of Dental Research of the US Public Health Service.

References

1 Warburg, O. and Christian, W. (1942) Isolierung und kristallisation des g~irungsferment enolase. Biochem. Z. 310, 384-421.

2 Wang, T. and Himoe, A. (1974) Kinetics of the rabbit muscle enolase-catalyzed dehydration of 2-phospho- glycerate. J. Biol. Chem. 249, 3895-3902.

3 Mauer, P.J. and Nowak, T. (1981) Fluoride inhibition of yeast enolase. 1. Formation of the ligand complexes. Bio- chemistry 20, 6894-6900.

4 Murray, J.J., Rugg-Gunn, A.J. and Jenkins, G.N. (1982) FIuorides in caries prevention, 3rd Edition, Butterworth- Heinemann, Ltd., Oxford.

5 Hamilton, I.R. (1990) Biochemical effects of fluoride on oral bacteria. J. Dent. Res. 69 (spec. issue) 660-667.

6 Bender, G.R., Thibodeau, E.A. and Marquis, R.E. (1985) Reduction of acidurance of streptococcal growth and gly- colysis by fluoride and gramicidin. J. Dent. Res. 64, 90-95.

7 Marquis, R.E. (1990) Diminished acid tolerance of plaque bacteria caused by fluoride. J. Dent. Res. 69 (spec. issue) 672-675.

8 Hilther, F-J., Psarros, N. and Duschner, H. (1990) Isola- tion, characterization, and inhibition kinetics of enolase from Streptococcus rattus FA-1. Infect. Immun. 58, 1043- 1047.

9 Kaufmann, M. and Bartholmes, P. (1992) Purification, characterization and inhibition by fluoride of enolase from Streptococcus mutans DSM320523. Caries Res. 26, 110- 116.

10 Sturr, M.G. and Marquis, R.E. (1990) Inhibition of pro- tontranslocating ATPases of Streptococcus mutans and

288

Lactobacillus casei by fluoride and aluminum. Arch. Mi- crobiol. 155, 22-27.

11 Brewer, J.M. (1981) Yeast enolase: Mechanism of activa- tion by metal ions. CRC Crit. Rev. Bioehem. 11,209-254.

12 Murphy, A.J. and Coll, R.J. (1992) Fluoride is a slow,

tight-binding inhibitor if the calcium ATPase of sarcoplas- mic reticulum. J. Biol. Chem. 267, 5229-5235.

13 Murphy, A.J. and Hoover, J.C. (1992) Inhibition of the Na,K-ATPase by fluoride. J. Biol. Chem. 267, 16995- 17000.