Anal. Chem. 1994.66, 2240-2244
.
Biodegradation of Soluble Redox Polymers. 1 (0.01 7Ferrocene)amylose
Boris B. Gnedenkot and Alexander D. Ryabov’** Department of Chemistry, Moscow State Universi@, 119899 Moscow, Russia, and Division of Chemistry, G. V. Plekhanov Russian Economic Academy, Stremyanny per. 28, 1 13054 Moscow, Russia
The reaction of amylose with NaH and FcCHZNMeJ+I-in dimethylsulfoxidebrought about a redox-labeledpolymer with a low degree of modification, viz. 1 ferrocene residue per 60 glucose units. The preparation, (0.017ferrocene)amylose, displays one-electron irreversible behavior at a pyrographite electrode in terms of the Delahay formalism, the formal redox potential Eo’ being equal to 0.38 V at pH 6 and 40 O C versus SCE. Specific to amylose enzymeseodedepolymerases,which carry out random hydrolysis, accept the labeled amylose providing a significant increase in the peak current on cyclic voltammograms. The absence of potential drifts suggests that the effect is due an increase in the diffusion coefficients of the amylose fragments in the course of enzymatic digestion of (0.017ferrocene)amylose. This proposal was confirmed by the simulationof experimental cyclic voltammograms. Several practical applications of the results of this study for electrochemical assaying the amylolytic activity and evaluation of the mechanismsof the enzymatic catalysis by amylases have been demonstrated. Chemical and biochemical degradation of polymeric materials,1-3 often self-assembled via the secondary and tertiary structures, is an urgent challenge to the scientific community encouraged by growing requirements of environmental ~ o n t r o l . ~Naturally, J various approaches, techniques, etc., will contribute to this interdisciplinary area including easy and fast analytical methods controlling the progress of depolymerization reactions. Such features are typical of amperometric biosensors which have recently proved to be extremely competitive as compared with other traditional analytical techniquesU6 Our interests are associated with organometallic electron-transfer carriers,’J and in this paper we present a general approach to follow by cyclicvoltammetry a random biodepolymerization of soluble polymer modified with ferrocene residues. Ferrocenes are now widely used as
redox modifiers of various polymersg-ll mostly for electrode coating, but biodegradation of insoluble ferrocene-modified amylose, which was digested by amylases, was only reported by McNeil et al.12 The enzymatic cleavage provided electrochemically active species responsible for broad traces on cyclic voltammograms. These were much improved for using as an analytical tool by combination with the glucose oxidase amplification. l 3 To obtain water-soluble ferrocene-modified amylose with a low degree of substitution, we developed a new procedure which is shown in Scheme 1. Amylose is a convenient and promising molecule, since, in addition to its numerous hydroxy groups, it is a polymer with a secondary helical structureI4and, hence, there was an expectation that, by analogy with the complexation of ferrocenes with ~ y c l o d e x t r i n s ~and ~ - ~organic ~ molecules with amylose itself,19,20 bound ferrocene residues could be incorporated into the six to eight-membered inner cavities of amylose and this will decrease their ability to exchange electrons with an electrode. Finally, there is a big family of specific to amylose enzymes which differ in the mechanism of cleavage of the 0-1 ,Ca-glycosidic bonds. One can thus select the pathway of depolymerization, i.e. either endo, when all 1,rl-bondshave an equal probability to be cleaved, or exo, when dissociation of terminal residues is highly preferred. We will also present (i) a quantitative rationalization of experimental cyclic voltammograms, (ii) an approach for rapid assay of the amylolytic activity of enzymes and (iii) an easy procedure for preevaluation of either exoor endo mechanism of the enzymatic hydrolysis. EXPERIMENTAL SECTION Apparatus. Electrochemical voltammetric measurements were made in a three-electrode thermostated 10-mL cell with pyrographite as a working electrode (surface A = 1.2 f 0.2 cm2),a saturated calomel electrode (SCE) (Radiometer), and ~~~
~~
~~~~
(9) Inagaki, T.; Lcc,H. S.;Skotheim, T. A.; Okamoto, Y . J. Chem. Soc.. Chem. + Moscow State University. t G. V. Plekhanov Russian Economic Academy.
Commun. 1989, 1181-83.
(10) Creasy, K. E.; Shaw, E. R. AMI. Chem. 1989.61, 1460-6s. (1) Klyosov, A. A. Biochemistry 1990, 29, 10577-8s (cellulose.). (1 1) Saji, T.; Hoshino, K.;Ishii, Y.;Goto, M. 1.Am. Chem. Soc. 1991,113,450(2) Schoemaker, H. E. Recl. Trau. Chim. Pays-Bas 1990,109, 255-72 (lignin). 56. (3) Evans, J. D.; Sikdar, S . K. CHEMTECH 1990, 3 8 4 2 (plastics). (12) McNeil, C. J.; Grcen, M. J.; Hill, H. A. 0. Patent GB, Appl. 85/4, 522.21 (4) Manahan,% E. Environmental Chemistry. 5thed.;LewisPublihers: Chelsea, Feb 1985; Chem. Abstr. 1988, 108, 34501~. MI, 1991. ( I 3) Review: Heller, A. Acc. Chem. Res. 1990, 23, 128-34. ( 5 ) Jones, M. M.; Johnston, D. 0.;Netterville, J. T.; Wood, J. L.; Joesten, M. D. (14) Kubik, S.;Wullf. G. Srorch/Sf&ke 1993, IS, 220-25. Chemistry and Society, 5th ed.;Saunder College Publishing: Philadelphia, (15) Matsue, T.; Evans, D. H.; Osa, T.; Kohyashi. N. J. Am. Chem. Soc. 1985, 107, 3411-17. 1987. (6) A series of extensive and update reviews demonstrating the recent advantagca (16) Ryabov, A. D.; Tyapochkin, E. M.; Varfolomecv, S. D.; Karyakin, A. A. Bioelectrochem. Biocncrg. 1990, 24, 25762. in this area is summarized in: Advances in Biosensors, Turner, A. P. F., Ed.; JAI Press, Inc.: Greenwich, CT, 1992; Vol. 2. (17) Strelets, V. V.; Mamdjarova, I. A.; Nefdova, M.N.; Pysnograeva, N. I.; (7) Review: Ryabov, A. D. Angew. Chem., Int. Ed. Engl. 1991, 30, 93141. Pospiiil, L.; Hauzlik, J. J. Electroam/. Chem. 1991, 310, 179-86. (8) Ryabov,A. D.;Trushkin,A.M.;Bakshccva,L.I.;Gorbatova,R.K.;Kubrakova, (18) Isnin, R.; Salam, C.; Kaifer, A. K. J. Org. Chem. 1991, 56, 3541. I. V.;Mozhaev, V. V.;Gnedenko, E. B.; Levashov, A. V. Angew. Chem.. Int. (19) Wulff, G.; Kubik S.Mukromol. Chem. 1992, 193. 1071-80. Ed. Engl. 1992, 31, 789-90. (20) Wulff, G.; Kubik S . Carbohydr. Res. 1992. 237, 1-10.
2240
Analytikal Chemistry, Vol. 66, No. 14, July 15, 1994
QQQ927QQl94fQ366224Q$Q4.5Q/Q 0 1994 Amerlcan Chemical Society
Scheme 1 0
0
1.
II
H3C-S-CH3 + NaH ==== H3C-S-CH*-Na+ + H2
'CH2NMe3*
0
reagents which were additionally purified in some cases according to the standard procedures.26 Preparationof Ferrocene-ModifiedAmylose. Amylose (2 g) and trimethyl(ferrocenylmethy1)ammonium iodide (0.77 g, 2 mmol) were dissolved in 20 mL of DMSO. On intense stirring (magnetic bar) sodium hydride as a 50% suspension in Nujol(0.24 g, 5.0 mmol) was added to this yellow solution. After violent evolution of dihydrogen the resulting mixture was heated with continued mixing at 100-120 OC for 1 h. The brown-black solution was then cooled to room temperature and the modified amylose was precipitated by addition of a 3-4-fold excess of ethanol (60-80 mL) and washed with 50 mL of ethanol. The slightly colored solid was purified 3 times by dissolving it in 20 mL of DMSO and precipitating with 60-80 mL of ethanol. This was necessary for removal absorbed ferrocene derivatives from amylose. Yield was 1.7 g (80% with respect to amylose).
RESULTS AND DISCUSSION Synthesis of Soluble (0.017Ferrocene)amylose. The procedure is presented in Scheme 1. Sodium hydride was added to a solution of trimethyl(ferrocenylmethy1)ammonium iodide and amylose in dimethyl sulfoxide (DMSO) to generate a strong base able to deprotonate hydroxy groups of amylose. Subsequent nucleophilic substitution gave rise to a modified polymer which was precipitated by addition of ethanol. The a platinum wire as ancillary electrode. Scans of potential material is poorly soluble in pure water but much better in were made by using a potentiostat P-5827M and voltammowater-DMSO mixtures. It is slightly colored as a solid, but grams were registered by PDP4-002 and H307 XY recorders. the solutions in DMSO look yellow-brown and the presence The temperature of the cell was kept constant by circulating of ferrocene residues can be estimated by conventional UVwater pumped in by a 2219 MULTITEMP I1 thermostat vis spectrophotometry at 425 nm where the material has a (LKB). Electronic spectra were recorded on a Beckmann 25 maximal absorption in 0.1 M acetate buffer, pH 6, in the spectrophotometer. The iron content in the labeled amylose presence of 10 vol % DMSO at 40 OC (t 0.0175 dm3g-I cm-I was measured by atomic emission with an inductively coupled or 170 dm3 mol-' cm-I with respect to the concentration of plasma ICAP-61 spectrometer (TJA Corp.). 'H and I3C ferrocene residues, vide infra). Our goal was a polymer with NMR spectra were measured on a CXP 200 Bruker speca low degree of modification and this was confirmed by the trometer in DMSO-da as solvent. Simulations of the cyclic atomic absorption spectrometry. The content of iron attached voltammograms were made by a SigmaPlot (Version 4.0) to amylose was 0.103 mol g-I or, alternatively, 1 ferrocene computer program. residue per 60 glucose units (the degree of modification 0.017). Reagents. a-Amylases from porcine pancreas and from Interestingly, a very close value was obtained from the UVhuman saliva (EC 3.2.1.1) and glucoamylase (exo-l,4-aviz measurements using the extinction coefficient for ferglucosidase) from Aspergillus niger (EC 3.2.1-3)were Sigma rocenecarboxylic acid of 153 dm3 mol-' cm-' at 446 nm.27 preparations. Amylose from potato (Mn 150 000 g This limited amount of the conjugated ferrocenes does not was purchased from Serva. Samples of resting saliva were change the structure of the polymer in solution as comes from used in analyses after 10-fold dilution with 0.1 M acetate the comparison of the IH and I3C NMR spectra of the buffer (pH 6) containing 15 mM CaC12. Aliquots (100 pL) native28*29 and modified amylose. Moreover, no ferrocene were then added into the electrochemical cell. Dimethylsignals are observed in the proton spectrum giving additional sulfoxide, a Merck reagent, was dried over molecular sieves evidence for the low degree of modification. For the same 3A (Ferak). Trimethyl(ferrocenylmethy1)ammoniumiodide reason it was also impossible to indicate precisely to which of was prepared by alkylation of ferrocene with (N,N,N',N'the three hydroxy groups of a glucose unit the methylferrocene tetramethyldiamino)methaneZ2 followed by quartenization of moiety was bound. Thus, the binding mode shown in Scheme [(dimethy1amino)methyllferrocene formed with i~domethane.~~ 1 is one of the possibilities, although the hydroxy group at C6 Other inorganic salts, mineral acids, phenol, 3,5-dinitrosalycilic is usually preferred.lZ The helical structure of (0.017feracid, and other chemicals used for preparation of the reagents rocene)amylose manifests itself in complexation with iodine for determination of reducing saccharide^^^**^ were Reakhim l
(21) Shevel'kova, A. N.; Ryabov, A. D.; Sinitsyn, A. P. Biokhimia 1993,58,92837. (22) Osgerby, J. M.; Pauson, P. L. J. Chem. SOC.1958, 656-60. (23) Lindsay, J. K.; Hauser, C.R. J. Org. Chem. 1957, 22, 355-62. (24) Nelson, N. J . Biol. Chem. 1944, 153, 375-80. (25) Somogyi, M. J. Biol. Chem. 1952, 195, 19-23.
(26) Perrin, D. D.; Armarcgo, W.L. F.; Perrin, D. R.Purification of Luboratory Chemicals; Pergamon Press: Oxford, New York, Toronto, Sydney, Paris, Frankfurt, 1980. (27) Reinehart, Jr., K. L.; Motz, K. L.; Moon, S.J. Am. Chem. Soc. 1957, 79, 2749-2754. (28) Dorman, D.E.;Roberts, J. D. J. Am. Chem. Soc. 1971.93, 4463-72. (29) Dais, P.; Perlin, A. S . Carbohydr. Res. 1982, 100, 103-16.
Analytical Chemistry, Vol. 66, No. 14, July 15, 1994
2241
0.0
0.0 0.0
0.8
E@ SCE)/V
F l g m 1. Repetitive cyctlc voltammograms of the labeled amylose (10 g dmJ) obtained in the absence (A) and in the presence (B)of a-amylase from porcine pancreas (30 mg dmJ min-l) at a scan rate 0.01 V s-l, 40 %, pH 6 (0.1 M acetate buffer), and in the presence of 10% DMSO. The time interval between the approaching of anodic peaks was 2 min.
resulting in development of an intense blue color. This observation suggests the principal possibility of incorporation of hydrophobic ferrocene residues into the helical fragments of amylose. A limited amount of ferrocene moieties brought about only one disadvantage to amylose: common oxidants which are used for assayingthe concentration of the reducing saccharides and, thus, the activity of amylases24125were unapplicable for precise quantitative measurements due to oxidation of ferrocene residues. Nevertheless, qualitative experiments using the Somogyi reagent2sshowed that the polymer is still a good substrate for pancreas and salivary a-amylases and glucoamylase. Electrochemical Properties of (0.017Ferrocene)amylose. Repetitive cyclic voltammograms of the modified amylose dissolved in 0.1 M acetate buffer, pH 6, are presented in Figure 1A. The peak current has only a small tendency to increase on increasing the number of scans. The peak separation was equal to 0.100 V and did not change with time. Such an electrochemical behavior of the redox polymer at the pyrographite electrode can be considered as one-electron and irreversible in terms of the Delahay formalism.30 Another reason that forces the using of this approach is the fact that the functioningof redox-modifiedpolymers at electrodesmight not be necessarily Nernstian. As we will see below, the irreversiblebehavior of the ferrocene label is advantageous in this case, since, if such, a quantitative theoretical treatment of the data described below simplifies ~onsiderably.~~ The one-electron mode implies that only one ferrocene residue is involved in the electron exchange with the electrode provided by the length of the polymer and its low degree of substitution. The formal redox potential (vide infra) of the irreversible oxidation Eo’ was equal to 0.38 V at pH 6 and 40 OC versus SCE. The peak current ip (AI is proportional to the concentration of the conjugate c in the range 1.67-10 g dm-3 and the corresponding equation is 109, = (2.17 f 0.09)~ (0.8 f 0.5) without correction for background signal. The correlation coefficient r is 0.995. This observation rules out meaningful change in viscosity in the concentration range used.
+
(30) Dclahay, P. J. Am. Chcm. Soc. 1953, 75, 1190-96. (31) Galus, 2.Theoretical Basics of Electrochemical Analysis; Mir Publishers: Moscow, 1974; p 265.
2242 AMlyticelChemr(sby, Vol. 66, No. 14, Ju& 15, 1994
The electrochemicalbehavior of (0.017ferrocene)amylose changes significantly in the presence of pancreas a-amylase which hydrolyzes 0-1,Ca-glycosidic bonds, Figure 1B. An increase in the peak current is bigger, the higher the activity of a-amylase. Other enzymes tested, i.e. salivary a-amylase and glucoamylase, produce similar effects. It should be emphasized that the rates of increase in current were different for the endo- and exo-enzymes of the equal activity. Remarkably, the position of anodic and cathodic maxima, and, hence, the peak separation, does not change with increasing the peak current in all cases. This suggests that the redox label does not change its environment as the depolymerization proceeds, and the increase in current in the presence of depolymerases is not due to the release of ferrocene from the amylose helix. If the increase in current were due to the decomplexation, a drift of the peak potentials should be observed (cf. with the complexation of ferrocenes with cyclodextrinslS1*). Therefore, the reason for the current increase is likely associated with a change of the molecular weight of the labeled amylose and, hence, of its diffusion coefficient. In the next section, we will show that our electrochemical data can be rationalized in terms of this concept. Quantitative Treatment of Cyclic Voltammograms. The profile of the anodic wave of a voltammetric curve for an irreversibleredox process is given by eq 1 where the current ,31932
i = nFAcrd(rbDd)1’2X(bt)
(1)
-
is expressed in PA,^^ b = [( 1 a)nFu/(RT)] {s-l], Drd (cm2 s-l) and c d (mMJare the diffusion coefficient and concentration of reduced species, respectively, n is the number of electrons transferred, a is the transfer coefficientfor a cathodic reaction, u {V s-l) is the scan rate, F is the Faraday constant, A {cm2) is the electrode surface, and x(bt) is thedimensionlesscurrent function. If the change in current is only due to depolymerization of (O.O17ferrocene)amylose,the growth should arise from an increase in the diffusion coefficient D d which occurs due to random depolymerizationof the polymer. If all other characteristicsof the electrochemicalprocess are kept constant, it is possible to simulate experimental cyclic voltammograms. The simulation is justified for polymers with a low degree of labeling, since the probability of cleavage of the glycosidic bond in the vicinity of the redox label, which may change the ferrocene environment and, hence, alter the electrochemical properties, is insignificant. This might not be true for a polymer with a bigger amount of redox mediators attached to the polymeric molecule.12 The diffusion coefficient D d in eq 1 is related to a number-averaged molecular weight of a polymer through the EinsteinStokes relation 234where rrd Drd = kT/iiaqr,
(2)
(cm) is radius of a spherical particle and 7 (g cm-1 s-1) is the (32) Bard, A. J.; Faulkner, L.R. Electrochemical Methods. Fundamentals and Applications; J. Wiley & Sons: New York, Chichestcr, Brisbane, Toronto, 1980; pp 222-224. (33) Yoshida, K. Electrooxidation in Organic Chemistry. The Role of Cation RadicalsasSynthetic Intermdiateq J. Wiley & Sons: N e w York, Chichester, Brisbane, Toronto, 1984; Chapter 1. (34) Berezin, I. V.; Klyosw, A. A. Practical Course of Chemical and Enzymatic Kinetics; MSU Publishers: Moscow, 1976; p 262.
viscosity of the medium. On assumption that an amylose molecule can be approximated by a sphere of radius r, the latter can be derived from the number-averaged molecular weight of amylose (Mn). The volume (V) of such a molecule is equal to (4/3)7rr3and alternatively V = M n / ( p N A ) , where p (g cm-3} is the density of amylose and N A (mol-’) is the Avogadro number. A combination of the two expressions leads to eq 3. Substitution of eq 3 into eq 2 and the resulting r3 = 3 M n / 4 * p N A
(3) 0.0
one into eq 1 will give a final eq 4 that binds the current with
0.6
0.0
0.6
E(vs SCE)/ V
Flgure 2. (A) Anodic waves of the cyclic voltammograms obtained during random hydrolysis of (0.017ferrocene)amylose by a-amylase. i = 0 . 2 9 n F A ~ , , ( ~ ~ ~ ~ ’ p ~ / ~ N , ~ / ~ (M4 )~ ~ ~ Conditions ~ ( * ) ~ /are ~ ~as( in b tFigure ) 1. (B) Simulated curves for the same experimental conditions calculated on the basis of eq 7. For details, see text.
the number-averaged molecular weight of the labeled amylose. All other parameters in eq 4 can be estimated. In particular, the transfer coefficient for a cathodic reaction a,calculated from the known value of the peak separation E , - E , = 1.857RT/(1 c ~ ) n Fis, ~equal ~ to 0.5. We assume that the density of the polymer ( p ) and the viscosity of the solution ( q ) do not change in the course of hydrolysis and are equal to 1.O g ~ m and - ~0.01 kg m-1 s-1, respectively. The absolute value of the density does not play an important role in this case. It is obvious that it cannot change dramatically on hydrolysis of labeled amylose and the power (1 / 6 ) in eq 4 is a guarantee that this term willnot affect thecurrent much. Theassumption for q is justified because the peak current does not deviate from linearity at the concentrations of the labeled amylose used, vide supra. The numerical values of the function (.n)ll2 x(bt) for the given values of a! and the peak separation were obtained as described by Nicholson.36 The dimensionless rate parameter I)equal to 0.6 was found from the tabulated data. The relation I) = k0/(7rD,&)1/2 ( a = nFv/RT) is used for evaluation the electron-transfer rate constant ko .32 The value of Drd for (0.017ferrocene)amylose was estimated from eq 2 (5.9 X lo-’ cm2s-l) and ko was then calculated ( 5 X lo4 cm s-’). The rate constant is needed for calculation of the formal redox potential Eo’which was equal to 0.382 V versus SCE. When Eo’is known (7r)’/2 x(bt) can be obtained. Therefore, if the variation of the number-averaged molecular weight of (0.017ferrocene)amyloseis found, a simulation of experimental cyclic voltammograms becomes possible. An agreement between the simulated and the experimental voltammograms would support our mechanistic concept. The kinetics of random enzymatic hydrolysis of polymeric substrates has been analyzed in the l i t e r a t ~ r e . ~ ’ Equation -~~ 5 gives a time dependence of the number-averaged molecular weight of a polymer in the presence of a depolymera~e?~ where
-
(5) (35) Nicholson, R. S.; Chain, 1. AM^. Chem. 1964, 36, 706-23. (36) Nicholson, R. S. Anal. Chem. 1965, 37, 1351-55. (37) Almin, K. E.; Eriksson, K.-E. Biochim. Biophys. Acfa 1967, 139, 238-47. (38) Almin, K. E.; Eriksson, K.-E.; Peterson, B. Eur. J . Biochem. 1975,51,20711.
(39) Rabinovich, M. L.; Klyosov, A. A,. Berezin, I. V. Bioorg. Khim. 1977, 3, 405-14.
rate of enzymatic hydrolysis (activity of enzyme) is expressed in (M min-’}, c {gdm-3)is the weight concentration of a polymer which does not change during hydrolysis, Mn (g mol-’} is the number-averaged molecular weight of a polymer, and c JMn (M}is a current concentration of terminal reducing saccharides (RS). One can integrate eq 5 in the case of the steady-state hydrolysis, i.e. when the rate of enzymatic reaction does not change with time. In fact, we have shown that the rate of enzymatic hydrolysis of amylose (the rate of RS formation {mg dm-3)) is constant over the time of voltammetric experiment (25 min), i.e. [RS] = [(2.99 f 0.05) X 1k2]t ( r = 0.997) under the conditions given in the caption to Figure 1B. Integration of eq 5 gives
where Mnoand M , are the initial and current number-averaged molecular weight of a hydrolyzable polymer, respectively.The expression for Mn should be substituted into eq 4, and the current is now a function of time, the activity of enzyme, and the number-averaged molecular weight of the initial amylose preparation, eq 7. The enzymatic activity was determined
spectrophotometrically by measuring the rate of appearance of the reducing saccharidesZ4Js and the resulting value was used in further calculations. The anodic waves of the cyclic voltammograms simulated on the basis of eq 7 are presented in Figure 2B and compared with experimental voltammograms obtained in the presence of the enzyme preparation of equal activity, Figure 2A. The similarity between the two is surprisingly good and, in our opinion, the coincidence is the best proof for our model of the electrochemical response of a limitedly redox-modified polymer that undergoes random hydrolysis. AnalyticalChemistry, Vol. 66, No. 14, July 75, 1994
2243
o
o
0
o
2
T 4
1 6
t
8
l
10
'
12
' 14
'
I 16
' 18
20
~
T
Timed mln Flgwo 3. Solid lines are theoretical dependencies of the numberaveraged molecular welght of amylose 4 versus time calculated according to eq 6. Numbers above the curves indicate endoamylolyHc actMty (in mg dma min-1). Data were calculated from experimental cydicvoltammogfamobtalnedin the presenceof enzyme preparalbns of different parHal activity shown in parentheses and expressed in mg dm4 min-': V,rr-amylase from human saliva (60); 0 ,aamylase from porcine pancreas (30); m, human saliva of one of us (20); V, glucoemylase from Asperg&s niger (50);0, a-amylase from human saliva (4.5).
Applications. Equation 7 shows that the current is a function of the activity of amylase provided other variables are kept constant. Therefore, the "random" activity of amylose depolymerase can easily be assayed as described here. Figure 3 shows how the key parameter Mn-1/6of eq 4 varies with time in the presence of enzyme preparations of different activity (eq 6). Unfortunately, we could not obtain straightforward quantitativeinformationon variation of Mn-'/6 with time under exactly the same conditions using, for example, gel chromatography or fluorescence depolarization. However, our data are in agreement with the recent qualitative study of amylasecatalyzed digestion of fluorescein-labeled amylose by fluorescence depolarization.21 The values of Mn-l/6 can alternatively be derived from the experimental cyclic voltammograms using eq 4. A comparison between the theoretical and experimental curves is demonstrated in Figure 3. The proportionality between the current and Mn-lI6might not be that strong in some cases. However, it seems to be quite acceptable when the degree of depolymerization of high molecular weight compound is large provided the digestion is enzymatic, Figure 3. Several enzyme preparations, which carry out random hydrolysis of amylose, were tested and good agreement was observed in all cases. Prior to addition into (40) Bixler, J.; Baker, G.; McLcndon. G. J. Am. Chem. Soc. 1992,114,6938-39. (41) Bourdillon, C.; Demaille, C.;Moiroux, J.; SavCant, J.-M. J. Am. Chem. Soc. 1993, 115,2-10. (42) Badia, A.; Carlini, R.;Fernandez, A.; Battaglini, F.; Mikkelsen, S.R.;English, A. M.J. Am. Chem. Soc. 1993, 115, 705360.
2244
Analyticel Ch~dstry,Vol. 66, No. 14, Jurv 15, 1994
the electrochemical cell, the enzyme sample was assayed by measuring the rate of formation of reducing saccharides and the value obtained was then compared with that expected from the electrochemical data. Salivary human a-amylase, pancreas a-amylase, as well as a sample of saliva of one of us (B.B.G.) gave a very good coincidence. At the same time glucoamylase from Aspergillus niger, which is exo-depolymerase, did show remarkably different behavior, Figure 3. The activity of the sample according to the "reducing saccharides" was equal to 50 mg dm-3 m i d , while the value of only 6 mg dm-3 min-l was obtained by cyclic voltammetry. In our opinion, this big difference is indicative of the fact that this enzyme cleaves terminal glucose residues, rather than carries out random hydrolysis. Obviously, if a polymer with a low degree of redox modification is depolymerized by exoenzymes, its bulk electrochemicalproperties should not change much after dissociation of a few redox-silent terminal units. This is an illustration of advantages of dealing with redox polymers of the low degree of modification. The amylolytic activity can thus be determined electrochemically using the plots such as Figure 3 in a wide range of enzyme concentrations. The data obtained are in good agreement with the results evaluated by traditional methods. The log-log plot of the enzymatic activity of a-amylase from porcine pancreas determined from the electrochemical measurements AE (M m i d } versus its concentration in the electrochemical cell (M) is a straight line the analytical form of which is log AE = (1.01 f 0.03) log E - (4.9 f 0.1)in the range of E (0.02-8.0) X IO+? M with r = 0.997.
CONCLUSION This study has shown a general approach for the investigation of chemical or biochemical random depolymerizationof high molecular weight molecules by electrochemical means, cyclic voltammetry, in particular, which is now finding new applications for mechanistic studies of biochemical proc e ~ s e sA . ~polymeric ~ substrate must meet several requirements. The degree of its modification by a redox moiety must be low, it should be soluble, and, if a biodegradation is under investigation, the modification should not affect its ability to be accepted by corresponding enzymes-depolymerases. ACKNOWLEDGMENT This work was partly supported by Grant 7-192of the Russian state scientificand technologicalprogram "Advanced Methods of Bioengineering". The authors aregrateful to Drs. A. N. Shevel'kova and A. P. Sinitsyn for advice concerning catalysis by amylases, Dr. I. E. Nifant'ev for the help in preparation of the labeled amylose, Dr. V. A, Polyakov for recording the NMR spectra, Dr. V. V. Mozhaev for his interest in this work and helpful comments. Received for review November 2, 1993. Accepted February 18, 1994." *Abstract published in Aduance ACS Absfmcfs,April 1, 1994.
