Nonenzymatic decarboxylation of pyruvate

ANALYTICAL

BIOCHEMISTRY

139, 353-358 (1984)

Nonenzymatic

Decarboxylation

GEORGE CONSTANTOPOULOS Developmental

of Pyruvate

AND JOHN A. BARRANGER

and Metabolic Neurology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland

and Communicative 20205

Received November 7, 1983 Triton X- 100, retinal, retinoic acid, retinal, hexane, dithiothreitol, mercaptoethanol, and some other commercially available chemicals caused nonenzymatic decarboxylation of pyruvate and cY-ketoglutarate. “Lipids” obtained from human or pigeon liver homogenates using isopropanol/ hexane also had very high nonenzymatic decarboxylating activity on these two a-ketoacids; most of this activity could be traced to the hexane (Eastman) used in the extraction. Optimum pH of the reaction with dithiothreitol and mercaptoethanol was 7-8 and with the other chemicals around 10, but considerable activity was present at pH 7-8. Liver homogenates had a scavenger effect on the decarboxylating activity of Triton X-100 and of dithiothreitol. Dithiothreitol and mercaptoethanol at high concentrations (11 mM) also had a scavengereffecton the decarboxylating activity of the “lipids.” Pretreatment of Triton X-100, dithiothreitol, retinol, and the “lipids” with catalase markedly decreased the decarboxylating activity, while treatment with boiled catalase failed to do so. The results suggest that these compounds contain oxidizing contaminants, perhaps peroxide derivatives. Powerful oxidizing impurities have been reported in Triton X-100 from various sources by Y. Ashani and G. N. Catravas (1980, Anal. Biochem 109, 55-62). Such peroxide derivatives may cause nonenzymatic decarboxylation of pyruvate and cu-ketoglutarate, presumably by a mechanism similar to the well-known nonenzymatic decarboxylation of a-ketoacids by hydrogen peroxide. In the absence of catalase and/or other protective agents against reactive oxygen derivatives, these chemicals would interfere in the assays of pyruvate dehydrogenase, pyruvate dehydrogenase complex, and cY-ketoglutarate dehydrogenase complex which depend on the release of %OZ from a-[ 1-“‘C]ketoacids. KEY WORDS: pyruvate; ot-ketoacids: decarboxylation; nonenzymatic; Triton X- 100; dithiothreitol; peroxide derivatives.

The most usual assay of pyruvate dehydrogenase complex (PDHC),’ in crude homogenates, is based on the cofactor-dependent production of 14C02 from [l-‘4C]pyruvate at pH 7.4-8.0. A similar method, with a-[l“C]ketoglutarate as substrate, is used for the measurement of the activity of a-ketoglutarate dehydrogenase complex (a-KGDHC). Also, one of the assays of pyruvate dehydrogenase (PDH, EC 1.2.4.1) is based on the determination of the thiamine pyrophosphate (TPP)dependent 14C02 release from [ 1-‘4C]pyruvate ’ Abbreviations used: PDHC, pyruvate dehydrogenase complex; a-KGDHC, a-ketoglutarate dehydrogenase complex; PDH, pyruvate dehydrogenase; TPP, thiamine pyrophosphate.

in the absence of NAD+ and CoA. These methods, in the presence of commercial preparations of Triton X- 100, dithiothreitol, and 2-mercaptoethanol, gave high substrate blanks. Further investigation showed that these agents plus some other chemicals and a lipid fraction obtained from pigeon liver homogenate with isopropanol/hexane caused nonenzymatic decarboxylation of pyruvate and cY-ketoglutarate, presumably because of the presence of oxidizing contaminants. Conditions for the nonenzymatic reaction and the nature of the decarboxylating activity and its possible interference with the enzymatic assays are also the subject of this investigation. An abstract of this work has appeared (1). 353

0003-2697184 $3.00 Copyright 0 1984 by Academic Press, Inc. All rights of reproductmn in any fwm resewed.

354 MATERIALS

AND

CONSTANTOPOULOS

AND

METHODS

stoppered again and complete CO2 collection was achieved by shaking the tubes for 90 min at 37°C. “Substrate” blanks at the pH of the reaction and also after the addition of 50 ~1 of 3 N HCl at zero reaction time were run in parallel. The various agents, depending on their solubility, were added in the reaction mixture in lo-20 ~1 aqueous or methanolic solution. Enzyme assays. The enzymes PDHC and a-KGDHC were assayed by the method of Blass et al. (3). The activity of PDH was measured by the method of Kresze (4). The procedure for the collection of CO2 was the same used in the nonenzymatic decarboxylation reaction.

[ I-‘4C]Pyruvic acid, sodium salt, 5-20 mCi/mmol, and cr-[ 1-14C]ketoglutaric acid, sodium salt, 40-60 mCi/mmol, were purchased from New England Nuclear (Boston, Mass.). Nonradioactive sodium pyruvate, sodium cu-ketoglutarate, CoA, NAD+, TPP, perox&se, superoxide dismutase, Da-tocopherol acetate, D-a-tocopherol succinate, Coenzyme QIO, and Dr.-cr-phosphatidylcholine (dipalmitoyl) were purchased from Sigma (St. Louis, MO.). Catalase was a product of Worthington Biochemical Corporation (Freehold, N. J.). Deionized water was used in this study. The peroxide test, using 10 ml water, was negative, less than 0.1 ppm (sensitivity of the assay). The source of the chemicals which caused nonenzymatic decarboxylation of pyruvate is shown in Table 1. Extraction of lipids. Pigeon or human liver was homogenized with 4 vol of 50 mM potassium phosphate buffer, pH 7.8, in a Sorvall homogenizer. The lipids were extracted from this homogenate with isopropanol/hexane by the method of Kawamura et al. (2). After removal of the solvents the lipids were dissolved in methanol (“lipids” from 2 g tissue/ml) and an aliquot was added in the reaction mixture. Assay of the nonenzymatic decarboxylation reaction. The decarboxylation reaction was studied by measuring the 14C02 produced from [ 1-r4C]pyruvate or a-ketoglutarate in 50 mM potassium phosphate buffer, pH 7.8, by different concentrations of the various agents. Reaction mixtures were prepared in 12 X 75mm disposable glass test tubes in an ice bath. The tubes were stoppered with corks from which were suspended plastic wells (cellulose propionate tubes size 4.8 X 19.9 mm, Beckman Cat. No. 341288). The wells contained paper wicks impregnated with 50 ~1 2.5 M NaOH in order to absorb evolved 14C02. Reactions were conducted at 37°C for 30 min in a total reaction volume of 200 ~1. At the end of the incubation the tubes were placed in an ice bath and the reaction mixture was acidified with 50 ~1 3 N HCl. The tubes were

BARRANGER

RESULTS

The chemicals which caused nonenzymatic decarboxylation of pyruvate, their source and grade, and also the optimum pH of the reaction, concentration of the chemicals, and CO1 produced are shown in Table 1. All chemicals were commercially available and of high grade. However, Triton X- 100 from New England Nuclear caused much more nonenzymatic decarboxylation of pyruvate than the Research Products International product. Similarly, the residue of hexane from Eastman-despite the fact that it was “Reagent grade”-had very large decarboxylating activity, while hexane from Burdick-Jackson was practically devoid of such activity. The latter was uv or HPLC grade. Isopropanol from either source was also practically free of decarboxylating activity. Retinol (vitamin A) had very high decarboxylating activity. Retinoic acid, retinal, and retinol acetate also caused nonenzymatic decarboxylation of pyruvate but to a lesser extent. Optimum pH of the reaction with all these chemicals but dithiothreitol and mercaptoethanol was around 10. However, considerable decarboxylating activity was present at pH 7.0-8.0, which is the optimum pH of PDH, PDHC, and cyKGDHC. Both dithiothreitol and mercaptoethanol exhibited their highest decarboxylating

NONENZYMATIC

DECARBOXYLATION TABLE

355

OF PYRUVATE

I

NONENZYMATIC DECARBOXYLATION OF PYRUVATE

‘TO* Chemical

Optimum PH

Source

Concentration of chemical (mM)

produced (pm01 min-‘)

Triton X-100

New England Nuclear (scintillation grade)

10.0

1.0 (0.065%) 3.0 (0.2%)

Triton X-100

Research Products International Corp. (scintillation grade)

10.0

1.0

0.0

10.0

12.0

Retinol (Vit. A)

Sigma (Type X)

10.0

0.1 1.0

58.0 229.0

Retinal

Sigma (Type XVI)

10.0

0.1 I.0

4.8 28.0

Retinoic acid

Sigma (Type XX)

10.0

I.0

47.0

Retinol acetate

Sigma (Type I)

10.0

1.0

20.0

Oleic acid

Sigma

10.0

2.5

17.0

Isopropanol (residue)

Fisher Scientific (certified reagent)

10.0

Residue of 0.5 ml

1.0

Isopropanol (residue)

Burdick-Jackson (HPLC grade)

10.0

Residue of 0.5 ml

1.0

Hexanes (residue)

Eastman (ACS grade)

10.0

Residue of 0.5 ml

575.0

Hexanes (residue)

Burdick-Jackson (UV, HPLC Grade)

10.0

Residue of 0.5 ml

1.0

“Lipids”

Human or pigeon liver extracted with Fisher and Eastman solvents

10.0

Eq to 2 mg tissue

335.0

“Lipids”

Human or pigeon liver extracted with Burdick-Jackson solvents

10.0

Eq to 2 mg tissue

5.8

Dithiothreitol

Nutritional

Dithiothreitol

California B&hem.

Dithiothreitol

Biochem. Corp.

32.0 64.0

7.0-8.0

0.01

17.2

7.0-8.0

0.01

17.0

Bethesda Research Laboratories

7.0-8.0

0.01

14.3

2-Mercaptoethanol

Sigma

7.0-8.0

0.01

19.0

L-Ascorbic acid

Fisher Scientific (reagent grade)

> 10.0

0.01

16.5

Corp.

activity at 0. l- 1.O I’nM concentration and pH 7.0-8.0. Most of the nonenzymatic decarboxylating activity found in the “lipids” could be traced to the hexane (Eastman) used in the extraction. A few experiments demonstrated that the chemicals which caused nonenzymatic decarboxylation of pyruvate also had the same effect on a-ketoglutarate. The fol-

lowing compounds, not shown in Table 1, have also been studied: DL-a-phosphatidylcholine (dipalmitoyl), Coenzyme Qlo, ~atocopherol acetate, and D-a-tocopherol acid succinate; none of them caused appreciable nonenzymatic decarboxylation of pyruvate. The conditions for the nonenzymatic decarboxylation of pyruvate have been studied

CONSTANTOPOULOS

356

PH

1. The pH optimum for the nonenzymatic decarboxylation of pyruvate. 0, With 5 mM Triton X-100 (New England Nuclear); A, with 0.1 mM retinal; 0, with “lipids” from pigeon liver obtained with isopropanol/hexane (Eastman); Cl, with 0.05 mM hydrogen peroxide. Reaction mixture in 200 ~1 was 50 mM in potassium phosphate buffer and contained 1.5 X IO5 cpm (20 nmol) [ I-r4C]pyruvate. Incubation for 30 min at 37°C. FIG.

in more detail and the results are shown in the next two figures. Optimum pH of the reaction with 5 mM Triton X- 100, 0.1 mM retinol, “lipids,” and 0.05 mM hydrogen peroxide is shown in Fig. 1. With Triton X- 100, retinol, and the “lipids,” optimum pH was around 10 but considerable activity was present at pH 7.0-8.0. Very little CO2 was released between pH 1.0 and 6.0. The reaction with hydrogen

AND BARRANGER

peroxide also had optimum pH around 10, but high activity was present at pH 5.0-7.0. The time course and the effects of concentration and temperature on the reaction with “lipids” are shown in Figs. 2a, b, and c, respectively. With a large excess of pyruvate it was found that the decarboxylating activity was limited (in Triton X-100 less than 0.5%). Liver homogenates had a scavenger effect on the decarboxylating activity of Triton X100 and dithiothreitol (Table 2). Bovine serum albumin had no appreciable effect. Surprisingly, although dithiothreitol and mercaptoethanol at low concentrations caused nonenzymatic decarboxylation of pyruvate, at higher concentrations they acted as scavengers on the nonenzymatic decarboxylation of pyruvate by other chemicals (Table 3). Pretreatment of the various chemicals and of the “lipids” with catalase at pH 7.8 markedly reduced their decarboxylating activity. Treatment with boiled catalase failed to do so. The results of such an experiment are shown in Table 4. Pretreatment with superoxide dismutase (and catalase) did not change the effect of catalase one way or another. DISCUSSION

Dithiothreitol and/or 2-mercaptoethanol are invariably included in the radiochemical assay of PDHC and a-KGDHC. Triton X100 is also used in certain assays of these en-

Ial

0 30so

90 TIME, MINUTES

120

lb)

0

0.5 1.0 “LIPID EXTRACT

2.5 1.11

ICI

0 25 50 TEMPERATURE

75 Oc

FIG. 2. (a) Time course of nonenzymatic decarboxylation of pyruvate by “lipids”; (b) effect of the concentration of the “lipids”; (c) effect of temperature. “Lipids” were extracted from pigeon liver with isopropanol/hexane (Eastman). 1 ~1 of “lipids” in methanol is equivalent to 0.4 mg tissue. The reaction mixture, in 200 pl, was 50 mM in potassium phosphate buffer and contained 1.2 X lo5 cpm (I 5 nmol) [ I-r4C]pyruvate. If not otherwise indicated, incubation time, 30 min; temperature, 37°C.

NONENZYMATIC

DECARBOXYLATION

357

OF PYRUVATE

TABLE 2

through the action of catalase which is almost always present. Bovine serum albumin had SCAVENGER EF’FE~~ OF PIGEON LIVER HOMOGENATE ON NONENZYMATIC DECARBOXYLATION OF PYRUVATE no appreciable effect. The effect of Triton BY TRITON X-100 AND DITHIOTHREITOL X-100 was more evident in the assay of purified pyruvate decarboxylase (EC 4.1.1.1) ‘%Or released from brewer’s yeast which could not act as a Reaction mixtures (cpm) scavenger. Homogenates may contain other protective agents as well. Buffer” 521 Buffer + 10e2 M Triton X-100 7,843 These findings and the effect of pretreatBuffer + 10m2M Triton X-100 ment of various chemicals with catalase (Table + homogenate’ 720 4) suggested that Triton X-100 and the other Buffer + lo-* M Triton X-100 chemicals contain oxidizing contaminants, + 200 pg BSA 6,737 perhaps peroxide derivatives. The presence of Buffer + lo-* M dithiothreitol 10,132 Buffer + lo-’ M dithiothreitol Buffer + lo-* M dithiothreitol Buffer + lo-’ M dithiothreitol + homogenate” Buffer + IO-’ M dithiothreitol + 200 pg BSA Buffer + homogenate

16,741 10,194 386 8,144 437

a 50 mM potassium phosphate buffer, pH 7.8. *Pigeon liver homogenate in 50 mM potassium phosphate buffer, pH 7.8; protein content = 250 pg. Reaction mixtures in duplicate contain 2 X 10’ cpm [ l-14C]pyruvate (20 nmol). Incubation at 37°C for 30 min.

TABLE 3 SCAVENGER EFFECTOF HIGH DITHIOTHREITOL NONENZYMATIC

Reaction mixtures

Buffer Buffer + 10m5M dithiothreitol Buffer + 10m4M dithiothreitol Buffer + 10e3 M dithiothreitol Buffer + 10-r M dithiothreitol Buffer + 10 ~1 “lipids” zymes. Moreover, these chemicals are utilized Buffer + IO-’ M dithiothreitol extensively in the biochemical laboratory. + 10 ~1 “lipids” Therefore, we were surprised to find that com- Buffer + 10m4M dithiothreitol + 10 ~1 “lipids” mercial preparations of these three compounds Buffer + lo-’ M dithiothreitol and the other chemicals listed in Table 1 cause + 10 ~1 “lipids” nonenzymatic decarboxylation of pyruvate, Buffer + 10m2M dithiothreitol cy-ketoglutarate, and presumably of other LY+ 10 ~1 “lipids” ketoacids. The reaction is base-catalyzed; op- Buffer + 10m5M mercaptoethanol Buffer + 10m4M mercaptoethanol timum pH with dithiothreitol and mercapBuffer + lo-’ M mercaptoethanol toethanol was between 7.0 and 8.0 and with Buffer + lo-* M mercaptoethanol all other chemicals around 10.0. Buffer + 10m5M mercaptoethanol While measuring the activity of PDHC in + 10 ~1 “lipids” tissue homogenates, we found that, in the Buffer + 10e4 M mercaptoethanol + 10 ~1 “lipids” presence of Triton X- 100 and/or mercaptoBuffer + lo-’ M mercaptoethanol ethanol, substrate blanks ( 14C02 released from + 10 ~1 “lipids” [ l-14C]pyruvate) obtained when only the ho- Buffer + lo-* M mercaptoethanol mogenate was omitted were much higher than + 10 ~1 “lipids”

blanks from which CoA was omitted. The results suggested that Triton X- 100 and/or mercaptoethanol caused nonenzymatic decarboxylation of pyruvate and that the homogenate acted as a scavenger of this activity, perhaps

CONCENTRATIONS

OF

AND MERCAPTOETHANOL ON THE DECARBOXYLATION OF FVRUVATE

14C02 released @pm) 356 2,103 3,219 2,895 1,385 20,357 22,420 20,396 17,551 4,027 548 908 2,077 1,572 19,131 19,040 11,030

Note. To the reaction mixtures, in 50 mM potassium phosphate buffer, pH 7.8, were added 6.7 X 10“ cpm (20 nmol) [ l-‘4C]pyruvate to a total volume of 200 ~1. Incubation at 37°C for 30 min. 10 pl “lipids” is equivalent to 4 mg tissue.

CONSTANTOPOULOS

358 TABLE 4

EFFECT OF CATALASE ON THE DECARLKIXYLATING ACXVIN OF TRITON X-100, DITHIOTHREITOL, VITAMIN A, AND “LIPIDS”

“‘CO2 released Reaction

mixtures

Buffer (50 mM potassium phosphate, pH 7.8) Buffer + lo-* M T&on X-100 Buffer + lo-* M Triton X-100 + 10 Pg catal= Buffer + lo-* M T&on X-100 + 10 pg boiled catalase Buffer + IO-’ M dithiothreitol Buffer + lo-’ M dithiothreitol + 10 PI2 catalase Buffer + lo-’ M dithiothreitol + 10 pg boiled catalase Buffer + IO-’ M vitamin A Buffer + 10m3M vitamin A + 10 pg catalase Buffer + 10 pl “lipids” Buffer + 10 pl “lipids” + 10 pg catalase Buffer + 10 Irp catalase

@pm)

950 9,222 2,858 8,760 22,969 871 8,728 72,886 43,775 95,417 66,104 786

Note. Preincubation with 10 pg (87 units) catalase (Worthington Biochemical Corp.) for 5 min at 37°C. Then 2 X 10’ cpm (20 nmol) [ 1-‘%Z] pyruvate were added and incubation continued for 30 min at 37°C. Boiled catalase: heated at 100°C for 5 min.

peroxides in ethyl acetate has been reported (5). Recently, powerful oxidizing impurities have been reported in Triton X- 100 from various sources by Ashani and Catravas (6). Some peroxide formed by air oxidation of 2-amino4- hydroxy-6,7 dimethyltetrahydropteridine, cofactor of phenylalanine hydroxylase, inactivated the enzyme and also attacked dithiothreitol and the cofactor (7). Such peroxide derivatives may cause nonenzymatic decarboxylation of pyruvate and cu-ketoglutarate by a mechanism similar to the well-known nonenzymatic decarboxylation of a-ketoacids by hydrogen peroxide (8). Other effects, such as oxidation of SH groups of enzymes or other proteins and cofactors, are discussed in (6).

AND BARRANGER

Dithiothreitol, mercaptoethanol, and ascorbic acid, at low concentrations, i.e., 1 mM or less, caused significant decarboxylation of pyruvate. However, at higher concentrations they had minimal decarboxylating activity toward pyruvate and acted as scavengers on the decarboxylating activity of other chemicals such as Triton X-100. It is important to mention that the decarboxylating activity of hexane is extracted with the lipids. It is well known that lipids under various conditions form peroxides. The mechanism of lipid peroxidation is under intensive investigation (9). In addition to their interference with various measurements (6), these chemicals, in the absence of catalase and/or other protective agents against reactive oxygen derivatives, would interfere in the assays of PDH, PDHC, and CZ-KGDHC. A spectrophotometric assay of PDHC which depends on the generation of acetyl-CoA and utilizes purified arylamine-Nacetyltransferase may be a better alternative to the radiochemical method for the measurement of the activity of PDHC in tissue homogenates ( 10). REFERENCES 1. Constantopoulos, G., and Barranger, J. A. (1983) Fed. Proc. 42, 1845. 2. Kawamura, N., Kishimoto, Y., Moser, H. W., Singer, H., and Baker, H. J. (198 1) Trans. Amer. Sot. Neurochem. 12, 235. 3. Blass, J. P., Cederbaum, S. D., and Kark, R. A. P. (1977) Clin. Chim. Acta 75, 21-30. 4. Kresze, G. B. (1979) Anal. Biochem. 98, 85-88. 5. Burstein, S., and Kimball, H. L. (1963) Steroids 2, 209-211. 6. Ashani, Y., and Catravas, G. N. (1980) Anal. B&hem. 109.55-62. 7. Jakubovic, A., Woolf, L. I., and &m-Henry, E. (1971) Biochem. J. 125, 563-568. 8. Metzler, D. E. (1977) Biochemistry. The Chemical Reactions of Living Cells, p. 49 1, Academic Press, New York. 9. Bucher, J. R., Tien, M., and Aust, S. D. (1983) Biochem. Biophys. Res. Commun. 111, 774-784. 10. Ksiezak-Reding, H., Blass, J. P., and Gibson, G. E. (1982) J. Neurochem. 38, 1627-1636.