- Email: [email protected]
Biotin ligands labeled with daunomycin as an electrochemical probe for avidin and biotin interaction
Talanta ELSEVIER
Talanta 44 (1997) 357-363
Biotin ligands labeled with daunomycin as an electrochemical probe for avidin and biotin interaction Shunitz Tanaka *, Fumie Yamamoto, Kazuharu Sugarwara 1, Hiroshi Nakumura Division of Material Science, Graduate School of Environmental Earth Science, Hokkaido University, Sapporo. 060. Japan
Received 2 April 1996; received in revised form 6 August 1996; accepted 9 August 1996
Abstract Electroactive biotin ligands were prepared by the reaction of daunomycin with biotinylating reagents with a different spacer. These biotin ligands exhibited similar electrochemical properties to those of daunomycin, but the adsorptivity of the ligands on the electrode increased with increasing length of the spacer. The electrode response of these ligands decreased when specifically bound with avidin. This made it possible to detect electroinactive avidin indirectly. Biotin was detected by observing the competitive reaction between biotin and the ligands for the limited binding sites of avidin. The binding strength of the labeled biotins with avidin was compared with that of unlabeled biotin by using an enzyme assay. © 1997 Elsevier Science B.V. Keywords: Avidin-biotin interaction; Daunomycin-labeled biotin ligands; Electrochemical detection
I. Introduction Avidin-biotin" binding is one of the strongest bindings between proteins and ligands. Therefore, the strong interaction has been applied in various fields and is referred to as avidin-biotin technology [1-5]. An avidin-biotin binding assay is one of the typical avidin-biotin technologies using this strong interaction. In these assays, it is necessary to detect the information about the binding by using various methods. Generally, the a v i d i n biotin interaction has been investigated by a pho* Corresponding author. Tel.: + 181 11 7062219; fax: + 181 11 7166101. Present address: Faculty of Engineering, Kigami Institute of Technology, Kitami 090, Japan.
tometric procedure which uses fluorescent ligands and enzyme conjugates [6,7]. However, these methods usually require the use of a separation procedure such as filtration or centrifugation before measuring in order to separate a free labeled biotin from a bound biotin. Radioisotopic ligands are very useful but the handling and disposal of the ligands is problematic. An avidin-biotin assay using an electrochemical procedure is difficult because both avidin and biotin are electroinactive. If it were possible to evaluate the avidin-biotin interaction electrochemically, a new electrochemical assay without a separation procedure could be developed. Therefore, we attempted to detect avidin-biotin interaction electrochemically by preparing biotins labeled with an electroactive compounds (LBs). When the LB and avidin are
0039-9140/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S0039-9140(96)02062-0
358
s. Tanaka et al./Talanta 44 (1997) 357 363
mixed, the LB becomes electroinactive through binding with avidin. Therefore, the electroinactive avidin can be detected indirectly from the change of the electrode response of the LB. The biotin can also be detected by observing the competitive reaction biotin and the LB for the limited binding sites of avidin. Previously, we prepared a labeled biotin ligand with daunomcyin as an electroactive compound. We then evaluated the interaction between avidin and biotin using the biotin ligand electrochemically [8]. It has also been reported that the affinity for labeled ligand-protein binding depends on the length of the spacer. The role of a spacer is to remove steric hindrance between two functional moieties and to facilitate the specific binding. If the length of the spacer is too short, the binding will be affected by steric hindrance. In this study, we prepared three LBs having spacers of different lengths as electrochemical probes for avidin-biotin binding assay. Subsequently, we observed the electrochemical behaviors of these LBs in the absence and presence of avidin. We also attempted to detect biotin using the competitive reaction of biotin and the LBs. Finally, the binding strength of these LBs with avidin was compared with that of biotin by an enzyme assay using a microtiter plate.
2. Experimental 2.1. Materials
Sulfosuccimidyl D-Biotin (Biotin-Sulfo-Osu), sulfosuccinimidyl N-(D-biotinyl)-6-aminohexanoate (Biotin-ACs-Sulfo-Osu) and sulfosuccinimidyl N-[N'-D-biotinyl-6-aminohexanoyl]-6'aminohexamoate (Biotin-(ACs)2-Sulfo-Osu) were purchased from Pierce Rockford (IL, USA). Daunomycin was supplied by Sigma, St. Louis, (MO, USA) and avidin and biotin by Wako Pure Chemicals (Osaka, Japan). Biotinamidocaproyl bovine serum albumin (Biotin-BSA; ca. 9.6 mol of biotin per mole of BSA), bovine serum albumin fraction V(BSA) and 3',3,5,5'-tetramethylbenzidine (TMB) were supplied by Sigma.
Horseradish peroxidase avidin D conjugate (AvP, 1.5 M peroxidase-1 M avidin) was purchased from Vector Labs Burlingame (CA, USA). Phosphate buffer was prepared with 0.1 MKH2PO4 and the pH was adjustable to pH 7 with the addition of 0. l M NaOH. High quality nitrogen gas was used for deaeration of solutions for electrochemical measurements. All other reagents were of analytical-reagent grade. 2.2. Apparatus
All electrochemical measurements were carried out using a Yanaco P-1100 polarographic analyzer (Yanagimoto Scientific, Kyoto, Japan) with a Rikadenki Electronic Model RY-101AT recorder (Rikadenki Kogyo, Tokyo, Japan). Visible spectra of daunomycin and several biotins modified with daunomycin were measured with a Shimazu UV-2200 UV-VIS recording spectrophotometer (Shimazu Tokyo, Japan). 2.3. Electrode
A glassy carbon electrode (Model No. 11-2012, 3.0 mm diameter; Bioanalytical Systems (BAS, West Lafayette IN, USA)) was used as the working electrode. Before a measurement was carried out, the electrode was polished with 1.0, 0.3 and 0.05 gm alumina (Baikoski International, Charlotte, NC, USA) followed by 10 min of ultrasonication. An Ag/AgC1 electrode (Model No. 11-2020; BAS) was used as the reference electrode and a platinum wire as the counter electrode. All potentials were measured against the Ag/AgC1 electrode. Strong adsorption of the LBs on the electrode required us to reproduce a new electrode surface for all subsequent attempts of the procedure described above. 2.4. Preparation o f L B
LBs were prepared as follows: 2 mM biotinylation reagent in DMF (500 lal) and 1 mM daunomycin in 50 mM phosphate buffer (pH 7.0) (250 lal) were mixed and incubated for 24 h at 4°C. The reaction product was separated from the biotiny-
359
S. Tanaka et al. " Talanta 44 (1997) 357 363
H3CO
O
OH
0
OH 0
0 ii
RM.LB. D
H
o II
--(CH2)B--N~C-(CH2) 4 H
RS'LB-D ~(CH2) 4
H S -~',,y,. N,..pO I/ I~I--H
S-A~N-.,pO I/ ~-H
RL.LB. D
HI
H 0 - - (CH2)5 - - N - C-- (CH2),~.--N - I~-- (CH2)4• 0 H
S . , " ' ~ N.,,.pO
t f
'I~-H
Fig. 1. Structure of labeled biotins prepared in this study.
lation reagent by T L C (Developing solvent chlor o f o r m - m e t h a n o l - f o r m i c acid (80:20:2)). The concentration of LBs was determined from the absorbance at 475 nm because the spectra of the LBs were identical with that of daunomycin in visible region. The proposed structures of the LBs as shown in Fig. 1, and were supported by N M R data. The LBs labeled with Biotin-Sulfo-Osu, Biotin-ACs-Sulfo-Osu and Biotin-(ACs)z-Sulfo-Osu are abbreviated to S-LB-D, M - L B - D and L-LBD, respectively. To examine the stability of the reagent, the spectra of LBs were measured. After 2 months, the spectra were the same as those at the time of preparation. Only one spot was observed on T L C even it was developed 2 months later. Therefore, it was concluded that the LBs were stable for at least 2 months.
2.5. Procedure .for electrochemical avidin-biotin assay
Avidin and the LBs were mixed in l0 ml of 0.1 M phosphate buffer solution (pH 7.0) and the solution was incubated for 1 h at r o o m temperature while being stirred. To investigate the competitive reaction between the LBs and biotin, biotin was added to a solution containing LB and the avidin. The solution was incubated under same conditions as described above. After the incubation, the solution (10 ml) was subjected to a voltammetric measurement. Deaeration of the so-
lution was carried out for 15 min under nitrogen, then a polished electrode was immersed in the solution and a potential of - 1.00 V was applied to the electrode for 5 min to reduce and accumulate the LB with stirring. After a rest time of 15 s, voltammograms of the LB were recorded by scanning the potential from - 1.00 to - 0.30 V using differential-pulse polarography (scan rate 5 mV S
]).
2.6. Procedure Jor enzyme assay After microtiter plates (Wako Pure Chemicals) had been washed with a detergent solution and water, 0.002°/,, biotin BSA (200 ~tl) in phosphate buffer solution was added to the wells of the plate, which was left to stand at 4°C for 16 h. The plate was then washed three times with 10 m M phosphate buffer containing 0.1 M sodium chloride (pH 7.4) and 0.1% BSA (PBS-B) solution. Furthermore, 1% BAS solution (200 ~tl) was added to each well and kept for 1 h at 4°C to eliminate all remaining hydrophobic binding sites. The solution was discarded and the plate rinsed three times with PBS-B and subsequently once with PBS-B before being used. Sample (100 ~tl), PBS-B (50 ~tl) and Av (0.12 nM, 50 ~tl) were mixed in the wells. The mixture was incubated overnight at 4°C with continuous shaking at 80 oscillations rain ~ on a Thermo Shaker NTS-1300 (Eyela, Tokyo, Japan). After incubation, the wells of the plate were washed five
360
s. Tanaka et al./ Talanta 44 (1997) 357-363
times with a 10 m M phosphate buffer containing 0.1 M sodium chloride (pH 7.4) (PBS), then TMB solution (100 ~tl) was added to each well. The enzyme reaction began with the addition of 0.01% hydrogen peroxide (50 ~tl). This mixture was shaken for 30 min. The reaction was stopped by adding 0.1 M sulfuric acid (50 ~tl). The absorbance of this solution was measured at 450 nm. The absorbance of a blank solution was taken as the unit (1.0) and the absorbance with 1 × 10 5 M biotin was taken as zero, and the ratios of the absorbances at various concentrations of biotin and LBs were plotted against concentration.
3. Results and discussion 3. I. Adsorptive behavior
of L B
on an electrode
Daunomycin has two pairs of redox peaks. The oxidation peak at - 0 . 6 5 V and the reduction peak at - 0 . 7 0 V were based on the redox reaction of the quinone part of the reagent molecule. The oxidation peak at + 0.5 V and the reduction peak at + 0.30 V were due to the redox reaction of the hydroquinone part [9]. The biotins labeled with daunomycin (LBs) also have the same pairs of redox peaks at the same potentials as for daunomycin itself. This suggests that labeling with biotin did not affect the electrochemical behavior of the daunomycin moiety. We investigated the binding behavior by using the oxidation peak at - 0 . 6 5 V vs. (Ag/AgC1) by differential-pulse voltammetry because this peak was the sharpest among the four peaks. Fig. 2 shows the relationship between the peak current and the accumulation time for 4 x 10 7 M daunomycin and LBs on the electrode. Daunomycin was adsorbed on the glassy carbon electrode and the peak current increased linearly until 15 min. The peak currents of M-LB-D and S-LBD increased with increasing accumulation time up to 10 min and L-LB-D up to 5 min. When sufficient accumulation time had elapsed, the peak currents remained constant. These values are probably due to the adsorptive equilibrium of the LBs on the electrode surface. The peak current of
M-LB-D is similar to that of L-LB-D, but is larger than those of S-LB-D and daunomycin. It was found that the adsorption of M-LB-D and L-LB-D on the electrode is stronger than that of S-LB-D and daunomycin. This enhancement of the adsorbing properties of the ligands may be attributed to the increase in hydrophobicity with increasing length of the spacer and by converting the ionic amino group of daunomycin into a more hydrophobic biotin moiety. 3.2. Change o f the L B electrode response based on the interaction with avidin
The different pulse voltammograms of M-LB-D and daunomycin with and without avidin are shown in Fig. 3. The peak shape of the LB in the solution containing avidin was similar to that without avidin, but the peak current of the LB decreased considerably. When avidin was added to the solution containing daunomycin, the peak current of daunomycin did not decrease even after 2 h. Accordingly, it was concluded that the decrease in the peak current of LB was not due either to hindrance of the electrode reaction by the adsorption of protein on the electrode or to non-specific binding of ligand with protein. Therefore, the change in the peak current of the LB was considered to be due to the specific binding between avidin and biotin.
4
*. ..........v (c) O~-~O • (d) 3 ~. .,-" • ...A (b) o/~.."~" • .."; V ~ ~- ' "• .... "" "- 2 / / ~ (a) 1 O
0
r
i
I
5
10
15
20
Accumulation time/min
Fig. 2. Dependence of peak current on accumulation time 4 × 10 -7 M S-LB-D (a), daunomycin (b), M-LB-D (c) L-LBD (d). Results obtained by differential-pulsevoltammetry(0.01 M phosphate buffer, E, = - 1.0 V, t, = 5 min, scan rate = 5 mVs ').
S. Tanaka et al. /Talanta 44 (1997) 357 363
[ 0.SPA \
Ca)
I
-,.o
i
-o18
-o:4
I
-,.o
I
-o18 -o18 -o.,
Potential(V) vs. Ag/AgCl Fig. 3. Voltammograms of daunomycin and M-LB-D with and without avidin. (a) 4×10 7 M M-LB-D; as (a) + 6 x 10 Mavidin;(c) 4x 10 7Mdaunomycin;(d) as(c) +1.2x10 7 M avidin. The dependence of the peak current of the LBs on the concentration of avidin is shown in Fig. 4. The peak current of the LBs decreased with increasing concentration of avidin, which made the detection of the electroinactive avidin possible. The detection limits for avidin calculated from the relative standard deviation (10%) of the peak current without avidin were 5 x 1 0 _ 9 M using S-LB D, and 2 x 10 9 M using M - L B - D and L-LB-D. When a sufficient amount of avidin was
2
,b,
0I 0
iii
. . . . . . . r ~ . . . . . . . v - - j : . . . . . r x'" . . . . . . . . . . . . " ' ~
1
2
3
4
Concentration of avidin/10 Fig. 4. Effect of avidin on peak current. 4 (a) M-LB-D (b) and S-LB-D (c).
x 10-
5 -7 M
7 M L-LB-D
361
added to bind all of the LBs, the peak current became close to zero. It can be suggested that the LB lost its electroactivity because the electroactive part of the daunomycin was covered with a large volume of avidin. If so, the length of the spacer should affect the electrode response of the LBavidin conjugates. More detailed observations showed that when a high concentration of avidin was added, the peak current of S-LB-D was almost zero but M - L B - D and L-LB-D showed a small peak current even when a high concentration of avidin was added. This suggests that when avidin binds with M - L B - D and L-LB-D, the electroactive part is not completely covered with avidin because the electroactive part is located far from the binding site of avidin owing to the presence of a longer spacer. Therefore, the electroactivity of the ligands can be maintained but the diffusion constant decreases considerably. In contrast, the electroactive part of S-LB-D is covered completely with avidin because of the short spacer length. 3.3. Competitiz~e assay f o r biot#7
The competitive reaction of the LB ad biotin for the limited binding sites of avidin makes the assay of biotin possible. When biotin, at various concentrations, was incubated in a solution containing a constant concentration of avidin and the LB, the peak current of the LB increased with increasing the concentration of avidin and the LB, the peak current of the LB increased with increasing the concentration of biotin, as shown in Fig. 5. This is because biotin occupies the binding sites of avidin and, as a result, the amount of free LB increases with increase in biotin content. The response curves of three types of LBs were not as varied. The best response curve was obtained by using L- and M-LB-D, as shown in Fig. 5. The relative standard deviation of the peak current at 1 × 10 -v M was 12% using M-LB-D. 3.4, E n z y m e assay using microtiter plate
To investigate the strength of the binding of these ligands with avidin, an enzyme assay [10] was carried out and the results are shown in Fig.
362
S. Tanaka et al. / Talanta 44 (1997) 357-363
2.5
2
~
(a)
y."" (b) A ......"~ ..." ..--m (c)
< 1.5 _~.
The increase in the hydrophobicity of the ligand due to labeling caused it to bind non-specifically with avidin, having hydrophobic properties.
4. Conclusion
1 ....
0
-10
a
i
-8 -6 -4 Iog[Concn.ofBiotin/M]
Fig. 5. Competitive binding assay for biotin using LBs by the electrochemical procedure. 4 × 10-v M L-LB-D (a), M-LB-D (b), S-LB-D (c). Concentration of avidin 1.2 z 10-7 M.
6. The order of the binding strength was M-LBD > L-LB-D > biotin > S-LB-D. This is due to the difference in the length of spacer between biotin and the daunomycin part. The binding constant of M-LB-D is the largest of the LBs and that of S-LB-D is the smallest. Since the distance between the biotin part and the electroactive compound is short in S-LB-D, it is thought that steric hindrance weakens the binding between the LB and avidin. The distance of the spacer does not affect the binding. The binding strength of M-LBD and L-LB-D seemed to be slightly larger than that of biotin. If the distance of the spacer is too long, the compounds have high hydrophobicity.
Three electroactive ligands for avidin-biotin assay were prepared. The binding behaviors of these LBs with avidin were investigated by using electrochemical and enzyme assay methods. These ligands, adsorbed strongly on the electrode, were detected sensitively by accumulation voltammetry. It was found that the adsorption of the ligands becomes stronger as the length of the spacer increases because of an increase in hydrophobicity. The electrode response of these LBs decreased drastically by binding with avidin. Therefore, the detection of 10 -9 M avidin became possible. Biotin was also detected electrochemically using the competition reaction with these LBs. It is concluded that these LBs represent a new electrochemical probe for the measurement of the avidin-biotin interaction.
Acknowledgements The authors thank the Ministry of Education, Science and Culture of Japan for support of this work under a Grant-in-Aid for Scientific Research (No. 07640795).
References
_ •0.5
0
, -10
, ( a ) ~ -8 -6
-4
Log[Labeled biotin/M]
Fig. 6. Enzymeassay using microtiterplate for LBs and biotin (a) M-LB-D; (b) L-LB-D; (c) biotin; (d) S-LB-D.
[1] E.A. Bayer, M. Safars and M. Wilchek, Anal. Biochem., 161 (1987) 262. [2] F.R. Harmon, M. Berger, H. Beegen and H.G. Wood, J. Biol. Chem., 255 (1980) 9458. [3] Y. Germani, H. deRocqunigny and J.L. Guesdon, J. lmmunol. Methods, 146 (1992) 25. [4] M. Wilchek and E. Bayer, Anal. Biochem., 171 (1988) 1. [5] D.M. Mock, G. Langford, D. Dubois, N. Criscimagna and P. Horowitz, Anal. Biochem., 151 (1985) 178. [6] K.L. Brillhart and T.T. Ngo, Anal. Lett., 24 (1991) 2157. [71 S. Daunert and L.G. Bachas, Anal. Chim. Acta, 208 (1988) 43. [8] K. Sugawara, S. Tanaka and H. Nakamura, Anal. Chem., 67 (1995) 209.
S. Tanaka et a l . / T a l a n t a 44 (1997) 357 363
[9] E.N. Chaney and R,P. Baldwin, Anal, Chem., 54 (1982) 2556.
363
[10] S. Tanaka, T. Takeuchi and G.A. Rechnitz. J. Chromatogr., 597 (1992) 443.
Talanta 44 (1997) 357-363
Biotin ligands labeled with daunomycin as an electrochemical probe for avidin and biotin interaction Shunitz Tanaka *, Fumie Yamamoto, Kazuharu Sugarwara 1, Hiroshi Nakumura Division of Material Science, Graduate School of Environmental Earth Science, Hokkaido University, Sapporo. 060. Japan
Received 2 April 1996; received in revised form 6 August 1996; accepted 9 August 1996
Abstract Electroactive biotin ligands were prepared by the reaction of daunomycin with biotinylating reagents with a different spacer. These biotin ligands exhibited similar electrochemical properties to those of daunomycin, but the adsorptivity of the ligands on the electrode increased with increasing length of the spacer. The electrode response of these ligands decreased when specifically bound with avidin. This made it possible to detect electroinactive avidin indirectly. Biotin was detected by observing the competitive reaction between biotin and the ligands for the limited binding sites of avidin. The binding strength of the labeled biotins with avidin was compared with that of unlabeled biotin by using an enzyme assay. © 1997 Elsevier Science B.V. Keywords: Avidin-biotin interaction; Daunomycin-labeled biotin ligands; Electrochemical detection
I. Introduction Avidin-biotin" binding is one of the strongest bindings between proteins and ligands. Therefore, the strong interaction has been applied in various fields and is referred to as avidin-biotin technology [1-5]. An avidin-biotin binding assay is one of the typical avidin-biotin technologies using this strong interaction. In these assays, it is necessary to detect the information about the binding by using various methods. Generally, the a v i d i n biotin interaction has been investigated by a pho* Corresponding author. Tel.: + 181 11 7062219; fax: + 181 11 7166101. Present address: Faculty of Engineering, Kigami Institute of Technology, Kitami 090, Japan.
tometric procedure which uses fluorescent ligands and enzyme conjugates [6,7]. However, these methods usually require the use of a separation procedure such as filtration or centrifugation before measuring in order to separate a free labeled biotin from a bound biotin. Radioisotopic ligands are very useful but the handling and disposal of the ligands is problematic. An avidin-biotin assay using an electrochemical procedure is difficult because both avidin and biotin are electroinactive. If it were possible to evaluate the avidin-biotin interaction electrochemically, a new electrochemical assay without a separation procedure could be developed. Therefore, we attempted to detect avidin-biotin interaction electrochemically by preparing biotins labeled with an electroactive compounds (LBs). When the LB and avidin are
0039-9140/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S0039-9140(96)02062-0
358
s. Tanaka et al./Talanta 44 (1997) 357 363
mixed, the LB becomes electroinactive through binding with avidin. Therefore, the electroinactive avidin can be detected indirectly from the change of the electrode response of the LB. The biotin can also be detected by observing the competitive reaction biotin and the LB for the limited binding sites of avidin. Previously, we prepared a labeled biotin ligand with daunomcyin as an electroactive compound. We then evaluated the interaction between avidin and biotin using the biotin ligand electrochemically [8]. It has also been reported that the affinity for labeled ligand-protein binding depends on the length of the spacer. The role of a spacer is to remove steric hindrance between two functional moieties and to facilitate the specific binding. If the length of the spacer is too short, the binding will be affected by steric hindrance. In this study, we prepared three LBs having spacers of different lengths as electrochemical probes for avidin-biotin binding assay. Subsequently, we observed the electrochemical behaviors of these LBs in the absence and presence of avidin. We also attempted to detect biotin using the competitive reaction of biotin and the LBs. Finally, the binding strength of these LBs with avidin was compared with that of biotin by an enzyme assay using a microtiter plate.
2. Experimental 2.1. Materials
Sulfosuccimidyl D-Biotin (Biotin-Sulfo-Osu), sulfosuccinimidyl N-(D-biotinyl)-6-aminohexanoate (Biotin-ACs-Sulfo-Osu) and sulfosuccinimidyl N-[N'-D-biotinyl-6-aminohexanoyl]-6'aminohexamoate (Biotin-(ACs)2-Sulfo-Osu) were purchased from Pierce Rockford (IL, USA). Daunomycin was supplied by Sigma, St. Louis, (MO, USA) and avidin and biotin by Wako Pure Chemicals (Osaka, Japan). Biotinamidocaproyl bovine serum albumin (Biotin-BSA; ca. 9.6 mol of biotin per mole of BSA), bovine serum albumin fraction V(BSA) and 3',3,5,5'-tetramethylbenzidine (TMB) were supplied by Sigma.
Horseradish peroxidase avidin D conjugate (AvP, 1.5 M peroxidase-1 M avidin) was purchased from Vector Labs Burlingame (CA, USA). Phosphate buffer was prepared with 0.1 MKH2PO4 and the pH was adjustable to pH 7 with the addition of 0. l M NaOH. High quality nitrogen gas was used for deaeration of solutions for electrochemical measurements. All other reagents were of analytical-reagent grade. 2.2. Apparatus
All electrochemical measurements were carried out using a Yanaco P-1100 polarographic analyzer (Yanagimoto Scientific, Kyoto, Japan) with a Rikadenki Electronic Model RY-101AT recorder (Rikadenki Kogyo, Tokyo, Japan). Visible spectra of daunomycin and several biotins modified with daunomycin were measured with a Shimazu UV-2200 UV-VIS recording spectrophotometer (Shimazu Tokyo, Japan). 2.3. Electrode
A glassy carbon electrode (Model No. 11-2012, 3.0 mm diameter; Bioanalytical Systems (BAS, West Lafayette IN, USA)) was used as the working electrode. Before a measurement was carried out, the electrode was polished with 1.0, 0.3 and 0.05 gm alumina (Baikoski International, Charlotte, NC, USA) followed by 10 min of ultrasonication. An Ag/AgC1 electrode (Model No. 11-2020; BAS) was used as the reference electrode and a platinum wire as the counter electrode. All potentials were measured against the Ag/AgC1 electrode. Strong adsorption of the LBs on the electrode required us to reproduce a new electrode surface for all subsequent attempts of the procedure described above. 2.4. Preparation o f L B
LBs were prepared as follows: 2 mM biotinylation reagent in DMF (500 lal) and 1 mM daunomycin in 50 mM phosphate buffer (pH 7.0) (250 lal) were mixed and incubated for 24 h at 4°C. The reaction product was separated from the biotiny-
359
S. Tanaka et al. " Talanta 44 (1997) 357 363
H3CO
O
OH
0
OH 0
0 ii
RM.LB. D
H
o II
--(CH2)B--N~C-(CH2) 4 H
RS'LB-D ~(CH2) 4
H S -~',,y,. N,..pO I/ I~I--H
S-A~N-.,pO I/ ~-H
RL.LB. D
HI
H 0 - - (CH2)5 - - N - C-- (CH2),~.--N - I~-- (CH2)4• 0 H
S . , " ' ~ N.,,.pO
t f
'I~-H
Fig. 1. Structure of labeled biotins prepared in this study.
lation reagent by T L C (Developing solvent chlor o f o r m - m e t h a n o l - f o r m i c acid (80:20:2)). The concentration of LBs was determined from the absorbance at 475 nm because the spectra of the LBs were identical with that of daunomycin in visible region. The proposed structures of the LBs as shown in Fig. 1, and were supported by N M R data. The LBs labeled with Biotin-Sulfo-Osu, Biotin-ACs-Sulfo-Osu and Biotin-(ACs)z-Sulfo-Osu are abbreviated to S-LB-D, M - L B - D and L-LBD, respectively. To examine the stability of the reagent, the spectra of LBs were measured. After 2 months, the spectra were the same as those at the time of preparation. Only one spot was observed on T L C even it was developed 2 months later. Therefore, it was concluded that the LBs were stable for at least 2 months.
2.5. Procedure .for electrochemical avidin-biotin assay
Avidin and the LBs were mixed in l0 ml of 0.1 M phosphate buffer solution (pH 7.0) and the solution was incubated for 1 h at r o o m temperature while being stirred. To investigate the competitive reaction between the LBs and biotin, biotin was added to a solution containing LB and the avidin. The solution was incubated under same conditions as described above. After the incubation, the solution (10 ml) was subjected to a voltammetric measurement. Deaeration of the so-
lution was carried out for 15 min under nitrogen, then a polished electrode was immersed in the solution and a potential of - 1.00 V was applied to the electrode for 5 min to reduce and accumulate the LB with stirring. After a rest time of 15 s, voltammograms of the LB were recorded by scanning the potential from - 1.00 to - 0.30 V using differential-pulse polarography (scan rate 5 mV S
]).
2.6. Procedure Jor enzyme assay After microtiter plates (Wako Pure Chemicals) had been washed with a detergent solution and water, 0.002°/,, biotin BSA (200 ~tl) in phosphate buffer solution was added to the wells of the plate, which was left to stand at 4°C for 16 h. The plate was then washed three times with 10 m M phosphate buffer containing 0.1 M sodium chloride (pH 7.4) and 0.1% BSA (PBS-B) solution. Furthermore, 1% BAS solution (200 ~tl) was added to each well and kept for 1 h at 4°C to eliminate all remaining hydrophobic binding sites. The solution was discarded and the plate rinsed three times with PBS-B and subsequently once with PBS-B before being used. Sample (100 ~tl), PBS-B (50 ~tl) and Av (0.12 nM, 50 ~tl) were mixed in the wells. The mixture was incubated overnight at 4°C with continuous shaking at 80 oscillations rain ~ on a Thermo Shaker NTS-1300 (Eyela, Tokyo, Japan). After incubation, the wells of the plate were washed five
360
s. Tanaka et al./ Talanta 44 (1997) 357-363
times with a 10 m M phosphate buffer containing 0.1 M sodium chloride (pH 7.4) (PBS), then TMB solution (100 ~tl) was added to each well. The enzyme reaction began with the addition of 0.01% hydrogen peroxide (50 ~tl). This mixture was shaken for 30 min. The reaction was stopped by adding 0.1 M sulfuric acid (50 ~tl). The absorbance of this solution was measured at 450 nm. The absorbance of a blank solution was taken as the unit (1.0) and the absorbance with 1 × 10 5 M biotin was taken as zero, and the ratios of the absorbances at various concentrations of biotin and LBs were plotted against concentration.
3. Results and discussion 3. I. Adsorptive behavior
of L B
on an electrode
Daunomycin has two pairs of redox peaks. The oxidation peak at - 0 . 6 5 V and the reduction peak at - 0 . 7 0 V were based on the redox reaction of the quinone part of the reagent molecule. The oxidation peak at + 0.5 V and the reduction peak at + 0.30 V were due to the redox reaction of the hydroquinone part [9]. The biotins labeled with daunomycin (LBs) also have the same pairs of redox peaks at the same potentials as for daunomycin itself. This suggests that labeling with biotin did not affect the electrochemical behavior of the daunomycin moiety. We investigated the binding behavior by using the oxidation peak at - 0 . 6 5 V vs. (Ag/AgC1) by differential-pulse voltammetry because this peak was the sharpest among the four peaks. Fig. 2 shows the relationship between the peak current and the accumulation time for 4 x 10 7 M daunomycin and LBs on the electrode. Daunomycin was adsorbed on the glassy carbon electrode and the peak current increased linearly until 15 min. The peak currents of M-LB-D and S-LBD increased with increasing accumulation time up to 10 min and L-LB-D up to 5 min. When sufficient accumulation time had elapsed, the peak currents remained constant. These values are probably due to the adsorptive equilibrium of the LBs on the electrode surface. The peak current of
M-LB-D is similar to that of L-LB-D, but is larger than those of S-LB-D and daunomycin. It was found that the adsorption of M-LB-D and L-LB-D on the electrode is stronger than that of S-LB-D and daunomycin. This enhancement of the adsorbing properties of the ligands may be attributed to the increase in hydrophobicity with increasing length of the spacer and by converting the ionic amino group of daunomycin into a more hydrophobic biotin moiety. 3.2. Change o f the L B electrode response based on the interaction with avidin
The different pulse voltammograms of M-LB-D and daunomycin with and without avidin are shown in Fig. 3. The peak shape of the LB in the solution containing avidin was similar to that without avidin, but the peak current of the LB decreased considerably. When avidin was added to the solution containing daunomycin, the peak current of daunomycin did not decrease even after 2 h. Accordingly, it was concluded that the decrease in the peak current of LB was not due either to hindrance of the electrode reaction by the adsorption of protein on the electrode or to non-specific binding of ligand with protein. Therefore, the change in the peak current of the LB was considered to be due to the specific binding between avidin and biotin.
4
*. ..........v (c) O~-~O • (d) 3 ~. .,-" • ...A (b) o/~.."~" • .."; V ~ ~- ' "• .... "" "- 2 / / ~ (a) 1 O
0
r
i
I
5
10
15
20
Accumulation time/min
Fig. 2. Dependence of peak current on accumulation time 4 × 10 -7 M S-LB-D (a), daunomycin (b), M-LB-D (c) L-LBD (d). Results obtained by differential-pulsevoltammetry(0.01 M phosphate buffer, E, = - 1.0 V, t, = 5 min, scan rate = 5 mVs ').
S. Tanaka et al. /Talanta 44 (1997) 357 363
[ 0.SPA \
Ca)
I
-,.o
i
-o18
-o:4
I
-,.o
I
-o18 -o18 -o.,
Potential(V) vs. Ag/AgCl Fig. 3. Voltammograms of daunomycin and M-LB-D with and without avidin. (a) 4×10 7 M M-LB-D; as (a) + 6 x 10 Mavidin;(c) 4x 10 7Mdaunomycin;(d) as(c) +1.2x10 7 M avidin. The dependence of the peak current of the LBs on the concentration of avidin is shown in Fig. 4. The peak current of the LBs decreased with increasing concentration of avidin, which made the detection of the electroinactive avidin possible. The detection limits for avidin calculated from the relative standard deviation (10%) of the peak current without avidin were 5 x 1 0 _ 9 M using S-LB D, and 2 x 10 9 M using M - L B - D and L-LB-D. When a sufficient amount of avidin was
2
,b,
0I 0
iii
. . . . . . . r ~ . . . . . . . v - - j : . . . . . r x'" . . . . . . . . . . . . " ' ~
1
2
3
4
Concentration of avidin/10 Fig. 4. Effect of avidin on peak current. 4 (a) M-LB-D (b) and S-LB-D (c).
x 10-
5 -7 M
7 M L-LB-D
361
added to bind all of the LBs, the peak current became close to zero. It can be suggested that the LB lost its electroactivity because the electroactive part of the daunomycin was covered with a large volume of avidin. If so, the length of the spacer should affect the electrode response of the LBavidin conjugates. More detailed observations showed that when a high concentration of avidin was added, the peak current of S-LB-D was almost zero but M - L B - D and L-LB-D showed a small peak current even when a high concentration of avidin was added. This suggests that when avidin binds with M - L B - D and L-LB-D, the electroactive part is not completely covered with avidin because the electroactive part is located far from the binding site of avidin owing to the presence of a longer spacer. Therefore, the electroactivity of the ligands can be maintained but the diffusion constant decreases considerably. In contrast, the electroactive part of S-LB-D is covered completely with avidin because of the short spacer length. 3.3. Competitiz~e assay f o r biot#7
The competitive reaction of the LB ad biotin for the limited binding sites of avidin makes the assay of biotin possible. When biotin, at various concentrations, was incubated in a solution containing a constant concentration of avidin and the LB, the peak current of the LB increased with increasing the concentration of avidin and the LB, the peak current of the LB increased with increasing the concentration of biotin, as shown in Fig. 5. This is because biotin occupies the binding sites of avidin and, as a result, the amount of free LB increases with increase in biotin content. The response curves of three types of LBs were not as varied. The best response curve was obtained by using L- and M-LB-D, as shown in Fig. 5. The relative standard deviation of the peak current at 1 × 10 -v M was 12% using M-LB-D. 3.4, E n z y m e assay using microtiter plate
To investigate the strength of the binding of these ligands with avidin, an enzyme assay [10] was carried out and the results are shown in Fig.
362
S. Tanaka et al. / Talanta 44 (1997) 357-363
2.5
2
~
(a)
y."" (b) A ......"~ ..." ..--m (c)
< 1.5 _~.
The increase in the hydrophobicity of the ligand due to labeling caused it to bind non-specifically with avidin, having hydrophobic properties.
4. Conclusion
1 ....
0
-10
a
i
-8 -6 -4 Iog[Concn.ofBiotin/M]
Fig. 5. Competitive binding assay for biotin using LBs by the electrochemical procedure. 4 × 10-v M L-LB-D (a), M-LB-D (b), S-LB-D (c). Concentration of avidin 1.2 z 10-7 M.
6. The order of the binding strength was M-LBD > L-LB-D > biotin > S-LB-D. This is due to the difference in the length of spacer between biotin and the daunomycin part. The binding constant of M-LB-D is the largest of the LBs and that of S-LB-D is the smallest. Since the distance between the biotin part and the electroactive compound is short in S-LB-D, it is thought that steric hindrance weakens the binding between the LB and avidin. The distance of the spacer does not affect the binding. The binding strength of M-LBD and L-LB-D seemed to be slightly larger than that of biotin. If the distance of the spacer is too long, the compounds have high hydrophobicity.
Three electroactive ligands for avidin-biotin assay were prepared. The binding behaviors of these LBs with avidin were investigated by using electrochemical and enzyme assay methods. These ligands, adsorbed strongly on the electrode, were detected sensitively by accumulation voltammetry. It was found that the adsorption of the ligands becomes stronger as the length of the spacer increases because of an increase in hydrophobicity. The electrode response of these LBs decreased drastically by binding with avidin. Therefore, the detection of 10 -9 M avidin became possible. Biotin was also detected electrochemically using the competition reaction with these LBs. It is concluded that these LBs represent a new electrochemical probe for the measurement of the avidin-biotin interaction.
Acknowledgements The authors thank the Ministry of Education, Science and Culture of Japan for support of this work under a Grant-in-Aid for Scientific Research (No. 07640795).
References
_ •0.5
0
, -10
, ( a ) ~ -8 -6
-4
Log[Labeled biotin/M]
Fig. 6. Enzymeassay using microtiterplate for LBs and biotin (a) M-LB-D; (b) L-LB-D; (c) biotin; (d) S-LB-D.
[1] E.A. Bayer, M. Safars and M. Wilchek, Anal. Biochem., 161 (1987) 262. [2] F.R. Harmon, M. Berger, H. Beegen and H.G. Wood, J. Biol. Chem., 255 (1980) 9458. [3] Y. Germani, H. deRocqunigny and J.L. Guesdon, J. lmmunol. Methods, 146 (1992) 25. [4] M. Wilchek and E. Bayer, Anal. Biochem., 171 (1988) 1. [5] D.M. Mock, G. Langford, D. Dubois, N. Criscimagna and P. Horowitz, Anal. Biochem., 151 (1985) 178. [6] K.L. Brillhart and T.T. Ngo, Anal. Lett., 24 (1991) 2157. [71 S. Daunert and L.G. Bachas, Anal. Chim. Acta, 208 (1988) 43. [8] K. Sugawara, S. Tanaka and H. Nakamura, Anal. Chem., 67 (1995) 209.
S. Tanaka et a l . / T a l a n t a 44 (1997) 357 363
[9] E.N. Chaney and R,P. Baldwin, Anal, Chem., 54 (1982) 2556.
363
[10] S. Tanaka, T. Takeuchi and G.A. Rechnitz. J. Chromatogr., 597 (1992) 443.