Coupling of redox indicator dyes into an enzymatic reaction cycle

Journal of Biochemical and Biophysical Methods, 15 (1988) 241-248

241

Elsevier BBM 00638

Coupling of redox indicator dyes into an enzymatic reaction cycle George H. Czerlinski, Byron Anderson, June T o w and D e b o r a S. Reid Department of Molecular Biology, Northwestern University Medical School, Chicago, IL 60611, U.S.A. (Received 11 August 1987) (Accepted 3 November 1987)

Summary The spectral properties of ten redox indicator dyes were evaluated with the aim of finding the optimal choice for coupling to enzymatic reactions with high sensitivity for the production of the reduced form. Eight of the dyes were selected for coupling into a reaction cycle formed by yeast alcohol dehydrogenase with substrates ethanol and nicotinamide adenine dinucleotide (NAD + ) and diaphorase with substrates reduced nicotinamide adenine dinucleotide (NADH, produced by the prior reaction) and the oxidized form of the respective dye. Two of the dyes exhibited decreased absorption on reduction, whereas all (eight) tetrazolium dyes increased in their absorption substantially upon reduction. Bis-tetrazolium dyes had a significantly higher molar extinction coefficient (up to 23000 M - L c m -1) than mono-tetrazolium dyes (down to 8000 M - l . c m - 1 ) . Kinetic.ally, most dyes could be reduced with NADH (and diaphorase), but the rate of reduction varied considerably among the dyes with nitrobhie tetrazolium (NBT) and tetranitrobhie tetrazolium (TNBT) being the fastest. Therefore, NBT and TNBT seem to be the most suitable for fast response. Key words: Alcohol dehydrogenase; Diaphorase; Tetrazolium dye; Redox indication

Introduction There are various ways in which nature accomplishes the amplification of biological processes, such as in cascade organizations as summarized and reviewed by Goldbeter and Koshland [1]. Recently, Johannson et al. [2] reported on the coupling of the hydrolysis of NADP (to NAD) to an enzyme cycle involving the reduction of a dye. This latter method aimed at a simple colorimetric detection of metabolic processes. The effectiveness of any method using colorimetric dyes depends considerably on the dye used. We decided to analyze a series of dyes Correspondence address: G. Czerlinski, 301 E. Chicago Av. W4-067, Chicago, IL 60611, U.S.A. 0165-022X/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)

242 regarding their suitability as colorimetric redox indicators. We based our work in part on data reported by Prince et al. [3] in which a polarographic study of 19 redox indicators was conducted over a p H range of 3 to 12; however, the use of tetrazolium dyes was not reported. Tetrazolium dyes are of interest, as the reduced form (the formazan) absorbs strongly in the visible range, while the oxidized form does not. 2,3,5-Triphenyltetrazolium chloride (TTC), prepared by Pechman and Runge [4], seems to be the earliest known dye in the tetrazolium family. Jue and Lipke [5] determined reducing sugars in the nanomole range with tetrazolium blue in 0.05 N N a O H with 0.5 N tartrate at 90 to 100 o C (using a reaction volume of 4 ml and observing the reduced dye at 660 nm). They found that tartrate increased the sensitivity and decreased the reaction time for the determination of reducing sugars. Racek [6] determined the concentration of lactate with nitrotetrazolium blue (NTB), using ferricytochrome-c reductase. They acidified the solution at the end of the reaction with 0.1 N HC1 to stabilize the color and to prevent the spontaneous reduction of excess tetrazolium salts. Raap [7] later looked into an apparent reductive side reaction. He found that the frequently used electron transfer mediator phenazine methosulfate produced the superoxide anion ( 0 2 ) in a side reaction from molecular oxygen, the superoxide anions being the reducing agents. We did not use any electron transfer mediator in the studies reported here. Mii and Green [8] effectively used tetrazolium dyes in a sequence of enzyme reactions, terminating with diaphorase in an assay system to quantitate fatty acid oxidation. To further increase the sensitivity of redox systems, we decided to systematically investigate the spectral properties of ten dyes and then select the most appropriate ones for an enzymatic redox cycle involving alcohol dehydrogenase and diaphorase, using ethanol in considerable excess. We show here the spectral characteristics of ten dyes, as they are reduced. We then selected eight dyes for further kinetic studies which revealed that NBT and T N B T gave the best responses and that the addition of the enzyme alcohol dehydrogenase and the substrate ethanol did not effect significantly the dye reduction kinetics.

Materials and Methods

All enzymes, dyes and other chemicals were purchased from Sigma Chemical Co. (St. Louis, MO). Tris buffer (0.1 M, p H 7.2) was used throughout, prepared by mixing Tris base with Tris-HC1 (in equiriaolar amounts) and distilled water in 200 ml volumetric flasks and adjusting the p H with NaOH. All dyes 1 were weighed out in quantities of 5.0 mg, which was first dissolved in 0.8 ml methanol and then 1 The followingabbreviations (in parentheses) are used for the ten dyes: phenazine ethosulfate (PES), iodo nitro tetrazolium (INT), (3-[4,5-dimethylthiazol-3-yl]-2,5-diphenylenetetrazolium (MTT), nitroblue tetrazolium (NBT), safranin O (SAF), tetranitroblue tetrazolium (TNBT), tetrazolium violet (TV), neotetrazolium chloride (NTC), tetrazolium blue chloride (TBC), 2,3,5-triphenyltetrazoliumchloride (TTC).

243 diluted to 10 ml with buffer ('dye stock'). This stock was diluted further with Tris buffer to produce an easily measurable spectrum. Sodium dithionite was added directly in small portions for non-enzymatic stepwise reduction to 3 ml of diluted dye solution and the spectrum was taken after each addition on a Beckman model DU-6 spectrophotometer with Tris buffer as reference (glass cells with 1 cm light path, 23 o C). Alcohol dehydrogenase (from baker's yeast [9], 148 000 g/mol) and diaphorase (from beef heart [10], 98 000 g/mol) were only used for kinetic experiments, which were also conducted on the Beckman DU-6 spectrophotometer. In selecting the enzyme concentrations we initially stayed close to the assay instructions of the supplier, as these are the conditions of proven performance. Nicotinamide adenine dinucleotide (NAD ÷ [also for NADH]) was dissolved in water (14 mg in 1 ml). Ethanol (100%, 0.2 ml) was added directly to the reaction mixture of 3 ml final volume.

Results and Discussion

Fig. 1 shows the spectrum of TTC, as it is reduced by the addition in increments of dithionite thoroughly mixed in the absorption cell. Fig. 1 defines a wavelength of the 'early' absorption maximum and a wavelength of the 'later' absorption maximum of the reduced form. The spectra are typical of the tetrazolium dyes, which exhibited either two maxima in sequence or one maximum and later also a shoulder (at decreased maximum). Table 1 supplies a summary of the spectral investigations on ten dyes, listing extinctions (and extinction coefficients) only at the wavelength of maximum absorption by the reduced form of the dye (or oxidized form, if its maximum is larger than that of the reduced form). The dye concentrations were computed from the dilution of the dye stock solution with Tris buffer. Addition of dithionite in small increments led to a maximum absorption, which shifted with subsequent additions. Dithionite is a very powerful reducing agent, which is able to produce a higher state of reduction in the dye than common biological reducing agents. Of the ten compounds, two (NBT and TNBT) showed no shift in the wavelength of maximum absorption with successive additions of dithionite, five showed a shift of 5 to 10 nm and the remaining three (PES, MTT and SAF) showed a larger shift. On standing, the reduced compounds developed turbidity and eventually definite precipitation was noted. The dye MTT exhibited the least amount of turbidity and precipitate formation. When two of the dyes (INT and TTC) were reduced by the addition of NADH and diaphorase and spectra taken as a function of time, these spectra were practically identical with the early and late spectra obtained by reduction with increasing amounts of dithionite. In their spectral redox titrations Wagner and Grossmann [11] found differences between INT on the one hand and NBT (as well as TNBT) on the other. INT showed a spectral change linear with INT concentration at all wavelengths, NBT did not. They explained that NBT forms the monoformazan first, which is only

244

~1 400 100/"

450 '

~2

500 '

'

550 '1

600 '

o.oot ............................... i....................... °=°I

.i

.

...............................

650 '

,.oo

...........................0.90

..............................

.....................o.=o

o.0oj°'7° !//iiii iiii ~iiiiiii iiI.~iiiii....._o.=oI ......ii..................o.~o .... o.5o~-. . . . . . ./. . . . / ~ ...........!. . . . . . .X....-o.5o ...

=

.............................................................. i...........\-'

............................... 0.40

0 •30 1 - ' " V - ' ' " " ~

..........~ .................................... 0.30

o.2o1~ ..................... ~

................ \ ................................ 0.20

O.lO-t..................~

. . . . . . .~...-O.lO ...

o.oo] 400

tI 450

500

550

o.oo 600

650

wavelength (nm)

Fig. 1. Spectra of reduced TTC in 50 mM Tris, pH 7.0 (25 ° C). The spectrum with maximum at wavelength hi developed after initial additions of dithionite to the oxidized dye (which showed absorption in the ultraviolet region only). After further additions of dithionite a new extinction maximum developed at wavelength A2. With many of the other dyes used in this study, no distinct second maximum developed but a shoulder; A2 then refers to this shoulder. TABLE 1 SPECTRAL CHARACTERISTICS A N D MOLAR EXTINCTION COEFFICIENTS OF R E D U C E D A N D OXIDIZED DYES Dye

PES INT MTT NBT SAF TNBT TV NTC TBC TTC

Dye conc. (/~M)

~1

(nm)

~2 (nm)

~ * of oxidized form at ~1

c of reduced form at )k1 Jk2

30 49 60 31 29 28 65 37 34 75

387 480 570 550 520 565 540 535 580 570

460 490 595 550 < 350 565 550 540 585 580

22100 80 70 330 38 200 145 62 107 116 54

11000 16 200 7 950 20600 140 20000 10000 16 600 22 700 13 400

11000 15 600 7 950 16400 140 15 600 9 700 13 600 20100 12 700

The spectra of the dyes were examined after reduction with dithionite in "Iris buffer as described under Materials and Methods. The two wavelengths represent the early, ?h, and later, h2, absorption maxima, as defined in Fig. 1 for TTC. If an extinction coefficient of a dye is smaller than 2% of the highly absorbing form, this value represents an upper limit (given by the read-out limitation of the instrument). * c, molar extinction coefficient (in M - 1. c m - 1).

245

further reduced subsequently. Their change found with NBT corresponds to the change, which we found. We may therefore assume, that in the other instances a shift in the absorption maximum with continued dithionite addition involves continued reduction. Analysis is complicated by the fact that the formazan eventually precipitates under the conditions of the experiment (preceded by increasing turbidity). The maximal extinction coefficients (Table 1) varied from 7900 M -1- cm -1 to 23 000 M-1 cm-1 for the tetrazolium compounds, the larger extinction coefficients seen with the di-formazans. PES and SAF have a high absorption in the oxidized state, which is diminished on reduction. These two dyes were therefore not used for the kinetic experiments. MTT showed a definite (but small) absorption in the oxidized state. The remaining (all tetrazolium) dyes showed an absorption below the detection limit of 0.004 optical density units in the oxidized state (400 to 700 nm); the extinction coefficient increased upon reduction by at least two orders of magnitude. Table 2 provides a summary of the results of the kinetic investigations on eight tetrazolium dyes. Measurements were taken at the wavelength close to maximum absorption of the formazan, obtained in the 'early' reducing stages. Diaphorase and NADH were added initially to each of the dye solutions and the kinetic measurements taken. The amount of diaphorase used for four of the dyes (TV, NTC, TBC and TTC) was tenfold greater than the quantity used for the other dyes. Alcohol dehydrogenase and excess ethanol were added later, producing no change with six of the dyes (0.004 represents the measurement limit) and negative changes with two dyes (by 6% for NBT and 4% for TNBT). These negative changes are larger than expected by dilution and are the result of further reduction, leading to a decrease in the early maximum and the formation of a shoulder (or new shifted maximum). In the kinetic experiments (Table 2) diaphorase was used first. The concentration of NADH used remained fairly constant for all dyes. The diaphorase concentration .

TABLE 2 KINETIC RESULTS FROM THE REDUCTION OF DYES WITH SUBSTRATES NADH AND ETHANOL USING ALCOHOL DEHYDROGENASE AND DIAPHORASE AS CATALYSTS Dye

[Dye] (jaM)

[NADH] (~M)

[Diaphorase] (/~M)

Wavelength (nm)

Initial kinetic slope (OD/min)

INT MTT NBT TNBT TV NTC TBC TIC

106 215 22 39 232 131 121 267

143 143 143 143 143 141 141 133

0.18 0.18 0.18 0.18 1.8 1.8 1.8 1.8

480 565 550 565 520 535 580 520

0.08 0.038 0.083 0.20 0 0.023 0.033 0.025

The reduction of the dyes was monitored at the 'early' absorption maximum after the addition of the substrates (NADH at 143/~M and ethanol at 140 raM) and then the enzymes (alcohol dehydrogenase at 0.015/~M and diaphorase at 0.18 or 1.8/IM, as shown).

246

was increased about tenfold for the last four dyes (as lower concentrations produced negligible changes). TNBT clearly showed the largest kinetic change (0.20 O D / m i n ) and NBT exhibited a kinetic slope of 0.083 OD/min. INT exhibited a kinetic change close to that of NBT, whereas TV showed no kinetic change, even after a tenfold increase in diaphorase concentration. The data clearly show that TNBT and NBT, followed by INT, are the kinetically best responding, and together with their high molar extinction coefficients are therefore the dyes of choice for quantitating reactions generating reducing product. MTT was not found to be as desirable because both its kinetic rate of reduction and its molar extinction coefficient were about half that of INT.

Conclusion In conclusion, MTT was found to be the more stable compound in the reduction process with dithionite in terms of precipitation, whereas the kinetically best responding dyes were NBT and TNBT, followed by INT. MTT was reduced at a rate about half as fast as INT under the conditions of the experiment. However, MTT has a molar extinction coefficient, which is only half as large as that of INT. In these studies we avoided the addition of the electron transfer mediator PES because of the already mentioned production of superoxide anion [7], which acts as a reducing agent of tetrazolium compounds. Our addition of alcohol dehydrogenase showed that the reaction with diaphorase is not perturbed by this addition, except by a small amount in two instances (NBT and TNBT).

Simplified description of the method and its applications Antibody coated magnetic microspheres are used to separate cells either by gravity or with the use of an opposing magnetic field. Eight thousand cells per minute can be separated.

Acknowledgement The authors gratefully acknowledge the partial support of this work by the Microtech Medical Corporation of Arlington Heights, Illinois.

References 1 Goldbeter, A. and Koshland, D.E., Jr. (1982) Sensitivity amplification in biochemical systems. Q. Rev. Biophys. 15, 555-591. 2 Johannsson, J., Stanley, C.J. and Self, C.H. (1985) A fast highly sensitive colorimetric enzyme immunoassay system demonstrating benefits of enzyme amplification in clinical chemistry. Clin. Claim. Acta 148, 119-124.

247 3 Prince, R.C., Linkletter, S.J.G. and Dutton, P.L. (1981) The thermodynamic properties of some commonly used oxidation-reduction mediators, inhibitors and dyes, as determined by polarography. Biochim. Biophys. Acta 635, 132-148. 4 Von Pechmarm, H. and Runge, P. (1894) Oxydation der Formazylverbindungen. Ber. Deut. Chem. Ges. 27, 2920-2930. 5 Jue, C.K. and Lipke, P.N. (1985) Determination of reducing sugars in the nanomole range with tetrazolium blue. J. Biochem. Biophys. Methods 11,109-115. 6 Racek, J. (1985) Bestimmung der Lactatkonzentration im Plasma mit L-Lactat: Ferricytochrome c-Oxidoreductase und Tetrazoliumsalz. J. Clin. Chem. Clin. Biochem. 23, 883-885. 7 Raap, A.K. (1983) Studies on the phenazine methosulfate-tetrazolium salt capture reaction in NAD(P)÷-dependent dehydrogenase cytochemistry. III. The role of superoxide in tetrazolium reduction. Histochem. J. 15, 977-986. 8 Mii, S. and Green, D. (1954) Studies on the fatty acid oxidizing system of animal tissues. VIII. Reconstruction of fatty acid oxidizing system with triphenyltetrazolium as electron acceptor. Biochim. Biophys. Acta 13, 425-432. 9 Backlin, K.I. (1958) The equilibrium constant of the system ethanol, aldehyde, DPN ÷, DPNH and H ÷. Acta Chem. Scand. 12, 1279-1285. 10 Lusty, C.J. (1963) Lipoyl dehydrogenase from beef liver mitochondria. J. Biol. Chem. 238, 3443-3452. 11 Wagner, H. and Grossmann, H. (1975) Duennschichtchromatographische und spectralphotometrische Pruefung von Tetrazoliumsalzen und deren Formazanen fiir ihren Einsatz als quantitative Redoxindikatoren. Z. Med. Labortech. 16, 94-103.