AMI. Ctwm. 1993, 65,3076-3080
9076
Fluorometric-Enzymatic Lactate Determination Based on Enzyme Cytochrome b2 Fluorescence Javier GalbPn,' Susana de Marcos, and Juan R. Castillo Department of Analytical Chemistry, Faculty of Sciences, University of Zaragoza, Pza. San Francisco sln, Zaragoza-50009, Spain
This paper presents a procedure for fluorometricenzymatic lactate determination based on the modification of the fluorometric properties of the enzyme L-lactic dehydrogenase (cytochrome h), during the enzymatic oxidationof the analyte with ferricyanide. During the reactionone can observe an irreversiblefall in the intensity of the enzyme's fluorescence, the rate of which is proportional to the concentration of the lactate. The source of this signal has been investigated and it has been shown that, besides the formation of a complex between the enzyme and the ferricyanide (the constant of which can be determined),this signal loss can be explained by simultaneous inner filter effects caused by the ferricyanide and the ferrocyanide (generated in the enzymatic reaction). A mathematical model has been developed which makes it possible to establish a linear response between the enzyme's analytical signal of fluorescence and the concentrations of the lactate, the cytochrome,and the ferricyanide. The procedure makes it possible to determine the lactate in concentrations ranging from 0.2 to 45 mg/L. Determination of the analyte has been carried out in milk samples with great precision and accuracy.
INTRODUCTION Lactic acid (or the lactic ion) is a parameter of great interest in different kinds of samples. In milk derivatives, it is an indicator of the quantity of the product's fatty acids and therefore of its quality. In clinical analysis, lactate is essential for the diagnosis of lactate acidosis, which may be a result of respiratory, hemodynamic, or metabolic disturbances, and it is very useful in sports medicine. Lactate has poor intrinsic spectroscopic properties, and so one usually has to resort to indirect procedures of analysis for ita determination. One of the more interesting alternatives is the method based on the enzymatic oxidation of lactate. For this, essentially two enzyme/coenzyme pairs have been employed nicotine adenine dinucleotide/lactate dehydrogenase (NAD/LDH) and ferricyanide/cytochrome b2 (Fed Cyt). The mechanism by which the oxidation of the lactate occurs is a three-stage reaction,' the final products of which are the pyruvate and the reduced form of the coenzyme [NADH and ferrocyanide(Foc), respectively]. In the case of the Fec/Cyt, the stages could be (where L = lactate, P = pyruvate, and Foc = ferrocyanide) ki
Fec + Cyt a Fec-Cyt
K, = K,/k-,
(1)
k-i
(1)Dixon, M.; Webb, E. C. Enzymes; Elsevier: London, 1979;Chapter IV, pp 47-137. 0003-2700/93/03653076$04.0010
ka
Fec-Cyt
+ nL + Foc-Cyt + nP
K, = k,/k,
(2)
k-2
(This stage (2) is subdivided into three others.)
ks
F o d y t Foc + Cyt (control stage) (3) For other enzymatic reactions that obey this mechanism, it is possible to find some of the values for the equilibrium constants,' although no data have been found in the bibliography for the reactions corresponding to Fec:Cyt. From an analytical point of view, the reaction of Fec/Cyt for the determination of lactate has been used essentially as a basis for enzymaticelectrodes,2giventhe poor spectroscopic properties that the Fec and the F d exhibit, although these can be improved through indirect procedures of analysis.' Recently this reaction has been used as a basis for a surface plasmon resonance sensor for lactate.6 Enzymes tend to exhibit fluorescence due to the presence of tyrosine and tryptophan in their molecules. The tyrosine's fluorescence often appears masked by that of the tryptophan, while the latter's fluorescence is often much affected by the surroundings of the molecule. Thus, starting from the quenching effect of fluorescence that certain species produce on the tryptophan, it is possible to evaluate some thermodynamic data both for the enzymatic species and for the species that produces the quenching.6 In spite of this, the enzymatic species' fluorescencehas rarely been used as a basis for carrying out the analytical determination of the substrates. Regarding this, one can only cite the work of Trettnak and Wolfbeis? who designed a sensor for lactate determination based on the intrinsic fluorescence of the lactate monooxygenase. These researchers found a very short linear response range (50-100 mg/L) and suggested no possible application of the method. In this paper a study of the use of the cytochrome's intrinsic fluorescence is carried out, in order to determine lactate, with regard to the previously mentioned enzymaticreaction. This study allows us to obtain information on the mechanism by which the analytical signal appears, to obtain some other of the thermodynamicconstants of the previously cited process, and to establish a model which enables us to evaluate the influence of the Cyt and the Fec on the method's sensitivity. Finally, the determination of lactate has been carried out in milk samples. EXPERIMENTAL SECTION Apparatus. All of the measurements were made using a Perkin-Elmer LS-50luminometer, equipped with a measuring ~~
~
~~~
~
~~~
~
(2)Kulis, J.; Svirmickas, G. J. Anal. Chim. Acta 1980,117,116-120. (3)Racek, J. J. Clin. Chem. Clin. Biochem. 1985,23,883-886. (4)Durliat, H.;Comtat, M. Anal. Chem. 1980,52,21m2112. ( 5 ) Castillo, J. R.; CepriA, C.; de Marc-, S.; Galbb, J.; Mateo, J.; Garcfa N ~ Z , E. Sens. Actuators, in prese. (6) Lakowicz, J. R. Principles of Fluorescence Spectroscopy; Plenum Press: New York, 1983;Chapter 9,pp 268-297. (7)Trettnak, W.; Wolfbeis, 0. S. Fresenius Z.Anal. Chem. 1989,334, 427-430. 0 1993 American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 65, NO. 21, NOVEMBER 1, 1993
3077
800
600
-
.-
h VI
Y
I
Flgurr 1. Excitatlon (X,= 335 nm) and emlssion (k,,,, = 225 nm) spectra of cytochrome (0.4 IU/mL) in phosphate buffer (pH = 8.0).
cell with a magnetic stirrer. Quartz cells with an optic path length of 1 cm were used. Reactants. The following were used: phosphate buffer, HzNaPOI/HNa$Od, 0.1 M and pH = 8.0; 15.6 IU/mL Baker’s yeast L-lactic dehydrogenase (cytochrome bz) (EC 1.1.2.3) (Sigma L-4506 type IV-SS), solutions of lower concentrations were prepared by dissolvingthe yeast in phosphate buffer; potassium ferricyanide 109 M; lactate stock solution, 1000 mg/L, in water prepared from lithium lactate (puriss. Sigma). Procedure. To the luminometer’s cell were added 1 mL of the phosphate buffer, 100 pL of a 2.25 IU/mL enzyme solution, and 50 pL of a 109 M solution of ferricyanide. The stirrer was started and a fluorescence intensity value was obtained at a wavelength of 335 nm (excitation at 225 nm). A quantity of the lactate solution (maximum 1 mL), was added and following a 30-5 wait the intensity value under these conditionswas obtained. To determine the lactate in milk, to a sample volume of 50 mL was added 0.5 mL of concentrated HClOd and the resultant mixture was centrifuged (at 4000 rpm for 10 min). An aliquot of the supernatant solution was taken and subjected to the previously indicated procedure.
RESULTS AND DISCUSSION Origin of the Analytical Signal. The Cyt gives an excitation spectrum with peaks at 225 and 275 nm and just one emission peak at 335 nm (Figure 1). The intensity of the fluorescence varies with the enzyme’s concentration in the way indicated in Figure 2; the deviation in the linearity, which can be observed for high concentrations, can be satisfactorily explained by the primary inner filter effect.8 The addition of Fec to a solution of the enzyme produces an instantaneous decrease in its fluorescence for either of the two wavelengths of excitation (Figure 3); whereas subsequently adding lactate gives rise to a fall in the intensity, in this case gradually with time, when the excitation occurs at 225 nm. The emission remaining is practically invariable if the molecule is excited to 275 nm. For either of the two cases, the variations in intensity occur without displacements in the spectrum’s wavelengths. The possible mechanism responsible for these signal variations has been studied. At 275 nm as excitation wavelength, the intensity of the Cyt’s fluorescence (lo) varies linearly with the concentration of Fec (in the absence of lactate). The mathematical calculations indicate that the results obtained under this wavelength of excitation agree with the conventional Stern-Volmer expression: r = 1.0000 K~75[Cytl,-JZo = 0.97 + 5817[Feclo whereas the results observed at 225 nm (Figure 4) obey the (8) Henderson, G.J. Chem. Educ. 1977,54,57-69.
400
200
0
[Cytl, i.u./mL
Fbure 2. Variation of fluorescenceintensity at 335 nm with cytochrome hXc = 275 nm. concentration: A, Lo= 225 nm;
*,
expression Kf225[Cytl,,/Zo= (1.00 + 6019[Feclo) exp(1415[Feclo) (4) (where Kf275 and Kf225 are the proportionality constants of fluorescence for the enzyme and [Cytlo and [Feclo are the concentrations of the Cyt and Fec initially added). These results are consistent with the mechanism for this previously indicated enzymatic reaction, the Stern-Volmer (&VI constant being equivalent to that of the formation of the complex Cyt-Fec. Regarding the exponential part of (4), the expression complies well with the quenching of the “sphere of action” 6 represented by the expression Kf[CytldZo = exp(uN[Fecl/1000) where u is the volume of the sphere of action (volume around the molecule for which the probability of quenching for the molecule is its unity). However, the volume of the sphere of action obtained by the results of the experiment is not consistent with the size of the enzyme molecule. As alternative possibility that enables us to explain this expression is to consider that the Fec causes, simultaneously with the formation of the compound Fec-Cyt, an inner filter effect on the enzyme’s fluorescence, such that expression 4 would obey the following: Kf225[CytldZo= (1+ KSv[Feclo)e~p(2~~~,,1[FecI,)(5) (where z225F,represents the molar absorptivity of the Fec at 225 nm and 1 the length of the optical path). When the excitation spectrum of the fec is given, one observes that the value of Z2%,% is consistent with the value obtained from fluorescence measurements (the observed difference can be explained by considering that during fluorescence the length of the cell does not coincide with the optical path length of radiations), while at 275 nm the value
ANALYTICAL CHEMISTRY, VOL. 65, NO. 21, NOVEMBER 1, 1993
5078
,
4
Cyi. addilion
3 0
’ a z
3
0,
2
1
-
.
.
,
0
5
100
50 t,
(6)
-
Figure 5. Fluorescence intensity variation during the enzymatic reaction: - -, A=. = 275 nm; -, X, = 225 nm; [Fec] = 0.25 mM, [Cyt] = 0.4 IU/mL, [Lac] = 2 X lod M, and phosphate buffer pH
10 Cancantrmtian
16 .
M
K,27s[Cyt]o/Io variation with (m) ferricyanide or (*) ferrocyanide concentration: X, = 335 nm and = 225 nm; [Cit] = 0.4 IU/mL. Figure 4.
= 8.0.
of Z is much smaller and practically negligible, which enables UB alsoto explain the reason why this type of quenchingeffect is not observed at 275 nm. To complete this study it is interesting to observe how the enzyme’s fluorescence varies in the presence of Foc which, from what is indicated, is one of the products of the enzymatic reaction. Once again, the mathematical calculations applied to the experimental results obtained allow us to establish that, at 275 nm, the results are very consistent with the conventional Stern-Volmer equation: K:76[Cytl,,/lo = 1.02 + 6095[Focl
. .
r = 0.999
whereas the results obtained at 225 nm (Figure 4) obey an expression very similar to the one at ( 5 ) X:26[Cytl,,/Io = (0.99 + 5915[Feclo) exp(7300[Fecl0) (6) It is very interesting to confiim how, once again, for both wavelength, the Stern-Volmer constants are very similar to each other and, in turn, also very similar to the one observed in the case of the Fec, which indicates that the enzyme and the Foc also form a complex, with a constant similar to that of the complex Cyt-Fec, and this is also consistent with the mechanism of the enzymatic reaction. The constant of the exponential term observed for the Foc is greater than the one observed for the Fec, which is also consistent with the inner filter effect previously mentioned. Moreover, it is also interesting to confiim how the quotient of the molar absorptivities observed at 225 nm (2226F&226F, = 4.8)is very similar to the quotient of the exponential terms obtained in expressions 5 and 6 (K’F~JK’F,= 5.2), which enables us to establish that the quenching process produced by the Fec and the Foc on the Cyt obeys the established mechanism. In the sameway, one can also observe that the molar absorptivity of both species at 275 nm is very similar.
IFncl
*lo*
M
Figure 1. AUI0 variation with ferricyanide concentration: X, = 335 nm,,.A = 225 nm; [Cit] = 0.4 IU/mL, [lactate] = 2 X M, pH
= 8.0, phosphate buffer.
All of these considerations lead us to establish that, as the Fec is transformed into Foc, the inner filter effect on the enzyme’s fluorescence increases, causing a gradual decrease
ANALYTICAL CHEMISTRY, VOL. 65, NO. 21,NOVEMBER 1, 1993
in the enzyme's signal during the reaction; the decrease in the signal occurs at the same rate as the transformation of the Fec into Foc. The values of the equilibrium constanta for the formation of the complexes Cyt-Fec and Cyt-Foc can be considered equal (5961 f 121 IU/mL). Optimizationof Parameters. The enzyme's fluorescence signal during the enzymatic reaction decreases with time, although the most striking variations can be observed during the first 30 s of the reaction, for which, to start with, we chose the parameter of quantification to be IO- I2, (where IO is the intensity value observed on adding the Fec to a solution of the enzyme and I2 is the value observed 30 s after adding the lactate). However, given that the value of IOcould at times be considered slightly imprecise, we decided to choose Alllo = (IO-Iz)/Io,which had already been used on other occasions9 as a quantification parameter. Using this parameter, a model has been established which enables us to evaluate what effect the concentration of the enzyme, the ferricyanide, and the lactate has on their values. According to previous definitions, the value of IOis the one that is obtained in the absence of an enzymaticreaction, and therefore, it can be deduced from expression 5:
When the lactate is added the reaction will begin to take place. For the first 30 sone can assume that the concentration of the complex cyt-Foc that will have been formed will be very small in comparison with the total concentration of the enzyme added (i.e., [CytIt=, = [Cytlo - [Cyt-FecIt=ao) and [Fecit,, = [Feclo and therefore
I2 = [Kf226[Cytld(l+ KS~[FeclO)l X eXp(22u~,l[FeC]~+ Z2u~,1[FOC] t=m) If we calculate the parameter AI/&, it can be observed that it equates to
M/Io= 1- eXp(-222s~,~[FOC]t=90)
(7)
From reaction 3, it follows that d[Focl/dt = k,[Cyt-Foc] under stationary-state conditions (d[Cyt-Foclldt = 0); then [FOCI,,, = 30k3[Cyt-Focl,=, In order to express [Cyt-Foclt=m as a function of [Llo, [Feclo, and [Cytlo, we can assume for this enzymaticreaction conditions similar to those applied to other reactions of the same type,' i.e., a stationary state is achieved: d[Cyt-Focl/dt = d[Cyt-Fecltdt = 0 The concentrations of lactate and pyruvate, during these 30 s, can be considered to be equal to the initial one ([LIt=30= [L]o and [PIt,, = 0) and that k-l>> k2 (Le., k - l + kz[L]n = k-1). This gives us [FOCI,,, = (30)5961k2[LIon[Feclo[Cytl,,l(l+ 5961[Fecl,) which, by substituting into (71, gives AUI0 = 1- exp(-AILlon[Feclo[Cytl~[l + 5961[Feclol) (8) (where A = (30)5961222S~,lk2). (9) Galbb, J.; de Marcoe, S.;Vidal, J. C.; Dim, C.; Aznarez, J. A w l . Sci. 1990,6,187-190.
SO78
.
I 0
Flgure6. AZ/Zo varlation with cytochrome concentration: X, = 335 nm, bXc = 225 nm; [Fec] = 0.25 mM, [lactate] = 2 X lo4 M, pH = 8.0, phosphate buffer.
The previous exponential function can be developed like the mathematical series 1- lo-" = 2.303X - x2/(2.3O3l2+ ~~l(2.303)' - x'I(2.303)'
For small index values, the terms of a higher order (a power of 2 or more), can be discarded: 1- lo-" = 2.303X hence expression 8 can be simplified to
(9)
Al/Io = 2.303A[Llon[Feclo[Cytl~(l + 5961[Feclo) (10) This expression enables us to evaluate how the analytical signal obtained varies as a function of the concentrations of the added Cytand Fec. For low enzyme concentrations (with [Feclo and [Llo constant), M/Iovaries linearly with the Cyt's concentration; when the Cyt's concentration is high, the simplification (8) cannot be applied and the linearity is lost. In the same way, it can also be seen that, for low concentrations of Fec (i.e., 1+ 5961[FecIo = 1)and maintaining the [Lloand the [Cyt]~constant, there is a section of linear response, whereas for high concentrations (Le., 1+ 5961[Fec]o = 5961[Feclo), the value of AI/&, ceases to depend on the concentration of the added Fec. The results obtained (Figures 5 and 6) agree with these hypotheses. The effect of the pH of the medium on the value of AI/Io has also been studied. Testa were carried out with various pH values and with various buffer solutions (glycine/NaOH, HP042-/H2P04-). In all cases, the optimum pH for the determination was the same (pH = 8), with slightly higher values when HP04Z/H2P04- was used. It is important to indicate that the enzyme's fluorescence spectra present the same form for the pH values indicated in the study, independent of the type of buffer solution used. Analytical Parameters. Figure 7 shows the variation of All& as a function of the lactate concentration (for constant
3080
ANALYTICAL CHEMISTRY, VOL. 65, NO. 21, NOVEMBER 1, 1993
Table I. Results Obtained from Milk Sampler sample this method NAD/LDH method 1 2 3 4 5 6 7
1
2
3
4
[ L a ~ l a t e ] * * l O + ~M,
Fbun 7. AUXo variatton with lactate concentratlon: X, = 335 nm, Lc= 225 nm; [Cit] = 0.4 IU/ml, [Fec] = 0.25 mM, pH = 8.0,
log (AI/Io)= 1.38 + 0.44log [LI,
[lactatel,,,
(10)Bichsel, S.Pharm. Acta Helv. 1984,59,62-64. (11)Gaffney, M.; Morrice, N.; Cooke, M. Anal. R o c . 1984,21,434436. (12)Owens, J. A.; Robinson, J. S. J. Chromatogr. Biomed. Appl. 1984, 32,380-386. (13)Figenschou, D.L.;Marais, J. P. Anal. Biochem. 1991,195,308312. (14)Marce, R. M.; Calull, M.; Olucha, J. C.; Borull, F.; Rim, F. V. J . Chromatogr. 1991,542,273-293. (16)Hikima, S.;Haeebe, K.; Kakizaki, T. Anal. Sci. 1992,8,165-168. (16)Renneberg, R.; Trott-Kriegeskorte, G.; Lietz, M.; Jarger, V.; Pawlowa, M.; Kaiser, G.; Wollenberger, V.; Schubert, F.; Wagner, R.; Schmid, R. D.; Scheller, F. W. J. Biotechnol. 1991,21,173-176. (17)Benthin, S.;Nielsen, J.; Villadeen, J. A d . Chim. Acta 1992,261, 145-153. (18)Yao, T.;Kobayashi, N.; Wara, T. Electroanalysis 1991,3,493491.
(19)Simonidea,W. S.;Zaremba, R.; Van Hardeveld,C.; Van der Laarse, W. J. Anal. Biochem. 1988,169,268-273.
= 1.007 [lactate],,,
method
method
- 0.752
r = 0.98
From what one can observe, despite the nature of the analytical signal, no interference seem to be caused by species which could be present in the solution.
r = 0.998
which is according to the stoichiometry proposed for this reaction in others works. As can be seen, the interval of concentrations for which the response is linear between log (Alllo) and log [Llo spans at least 0.2-45 mg/L lactate. The relative standard deviation obtained for 10 lactate determinations (2mg/L) was 2.5%. The method's sensitivity is better than those obtained by liquid chromatography (direct1@12or derivatization13J4) and
88f2 118 f 3 80f3 112 f 4 124 f 4 81 f 3 102 f 3
amperometric methods15Je and similar to that obtained by chemiluminescence (with lactate oxidase and luminal)." Better results have been obtained by using amplification reaction'* and special derivatization liquid chromatographic methods.19 RSD values are better than or similar to those obtained using the cited methods. Application. This method has been applied to lactate determination in milk samples, for which a deproteinization of the sample was necessary. Before this, we studied the effect that could be caused by the HClO4 that was used for this. One could observe that for concentrations below 0.8 N HC104there are no alterations in the lactate's analyticalsignal, which is appropriate when considering that the total deproteinization of the sample occurs with a concentration of 0.4 N HC104. The results obtained from the indicated method using seven samples of commercial milk correlated well with those observed by the standard method based on lactate determination by its enzymatic reaction with NAD/LDH (molecular absorption of the NADH), as indicated in Table I, with a correlation line as
phosphate buffer.
concentrations of Fec and Cyt). By using the least squares regression method, we obtained
90f3 115 f 4 80 f 3 109 f 3 128 f 4 83 f 3 100i3
CONCLUSIONS The results obtained in this paper indicate that the study of enzyme fluorescence in the course of an enzymaticreaction can be a very useful and simple procedure for determining thermodynamicconstants associated with the process, as well as for studying the mechanism of the reaction. Although the wavelength of excitation used can be very prone to interference effects, these can be avoided in the majority of caaes by increasing the concentration of the enzyme or the coenzyme. Finally, the reaction studied in this paper, accordingtowhat can be gathered from the results, could be applied to the determination of ferricyanide and of cytochrome bz.
ACKNOWLEDGMENT With thanks to Cables de Comunicaciones S.A. (General Cable Group) for their financial support (Project CEDETIOTRI 90/0076). RECEIVED for review April 7, 1993. Accepted July 29, 1993." 0
Abstract published in Advance ACS Abstracts, September 1,1993.
