/ . Biochem., 80, 455-461 (1976)
Interaction of Diiodo-L-tyrosine and Triiodophenol with Bovine Serum Albumin Circular Dichroism and Fluorescence Studies Nobuo OKABE, Noriko MANABE,1 Ryozi TOKUOKA, and Ken-ichi TOMITA Faculty of Pharmaceutical Sciences, Osaka University, Yamada-kami, Suita, Osaka 565 Received for publication, February 2, 1976
As a model study to investigate the binding mechanism between thyroid hormones and carrier protein, the interaction of diiodo-L-tyrosine (DIT) and triiodophenol (h) with bovine serum albumin (BSA) was investigated by circular dichroism (CD) and fluorescence methods. In both the DIT-BSA system and the Is0-BSA system, induced Cotton effect was observed in the wavelength region near 320 nm. This induced Cotton effect was measured at various molar ratios of ligands to BSA (L/P). The value of the ellipticity at 319 nm, [6]M, in the I8$6-BSA system was remarkably large compared with that of the DIT-BSA system, and [0]s« at an L/P ratio of one was -1.96xlO 4 (degree cma decimole"1) for the 1,0-BSA system and - 0 . 1 xlO 4 for the DIT-BSA system. The binding constants for the combination of BSA with a single molecule of ligand, calculated by measuring the quenching of the fluorescence of the protein, were 1.33x10* M"1 at 15° for the DIT-BSA system and 1.6x10° M"1 at 28° for the I80-BSA system. These results suggest that the binding of lt to BSA is stronger than that of DIT and a cleft may exist more congruent with the molecular dimensions of 1$^ than with those of DIT.
Thyroid hormones, L-thyroxine (T4), and triiodo-L-thyronine (T,) t bound to thyroxine-binding globulin, prealbumin or serum albumin in the blood are transferred to their receptor sites
from the thyroid gland (1,2). In the previous paper (3), we investigated the interaction between bovine serum albumin (BSA) and thyroid hormones (T4 and T s ) by circular di-
1
Present address: CIBA-Geigy (Japan) Limited, 10-66 Miyuki-cho, Takarazuka-shi, Hyogo Prefecture. Abbreviations: BSA, bovine serum albumin ; HSA, human serum albumin ; T DlIODO-L-TYROSINE
(DIT) TRIIODOPHENOL
Fig. 1. Thyroid hormones and their analogs.
Circular dichroism (CD) measurements were carried out using a Jasco ORD/UV-5 spectropolarimeter equipped with a CD attachment. Samples at various molar ratios of ligands to protein (L/P) were obtained by adding various amounts of DIT or \t at concentrations of about 3 x 10"' M with a micropipette to 5 ml of albumin solution (about 3.6x10"'M in 0.1 M phosphate buffer, pH 7.4, containing 3xlO" 4 M EDTA) with continuous stirring. BSA concentration was determined spectrophotometrically, using £}* m =6.54 at 280 nm (6) and a molecular weight of 64,300 ( 7 ) . Molar ellipticity, [d], (based on the concentration of BSA) was calculated from the equation, [0]=3,3OOx(£,-er)
chroism (CD) measurements, and induced Cotton effect {4, 5) was observed at around 320 run, the absorption wavelength region of T 4 and T 8 . It was concluded that the induced Cotton effect was due to the binding of these Hgands to protein. However, it was not obvious whether the observed Cotton effect was induced by the interaction between an asymmetric region of the protein and the chromophoric environment of the a-ring or between the protein and the £-ring of T 4 and T,. (The a- and /3-rings of these ligands are illustrated in Fig. 1). In order to investigate more precisely the induced Cotton effect in relation to the structure of the ligand or the protein, we used diiodo-L-tyrosine (DIT) and triiodophenol (I30), as model compounds for the a- and /)ring moieties of T 4 (see Fig. 1), respectively. The binding characteristics of DIT and IS0 to BSA were investigated by circular dichroism and fluorescence methods. MATERIALS AND METHODS
where (EI — £r) is the difference between the molar extinction coefficients for left- and rightcircularly polarized light. The ultraviolet fluorescence was measured with a Hitachi MPF-2A spectrophotofluorometer. The excitation and emission wavelengths of protein fluorescence were 290 nm and 350 nm, respectively. The molar ratio of ligands to protein was varied in the same way as for CD measurements. Several measurements were made and the emitted intensities were averaged for the DIT-BSA and I3?J-BSA systems. All fluorescence measurements were made in 0.1 M phosphate buffer, pH 7.4, containing 3xlO" 4 M EDTA at the temperatures indicated in each figure. The primary binding constant between BSA and a single molecule of DIT or Is0, KU was calculated from the fluorescence quenching data using a procedure entirely analogous to that employed by Steiner et al. (8). Their procedure for the determination of the primary binding constants for the T 4 (or Ti)-BSA and T 4 (or T,)-HSA systems is as follows: The observed relative intensity (I) is given by
Crystallized bovine serum albumin (BSA) and diiodo-L-tyrosine (DIT) were purchased from Y,Iy (1) Sigma Chemical Co., and triiodophenol (Ji) was from Nakarai Chemicals, Ltd. These reagents were used without further purification. where Yo is the mole fraction of albumin which DIT and Is$6 were dissolved in distilled water has no bound T 4 (or T 8 ); Io is the relative and 30% ethanol, respectively. fluorescence intensity corresponding to such / . Biochem.
SMALL MOLECULE-MACROMOLECULE INTERACTIONS
uncombined albumin; Yi and I t are the corresponding quantities for the species with 1 mole of bound T 4 (or T 3 ); and Yj and Iy are the corresponding parameters for the species with j moles of bound T 4 (or Tj), where j is equal to or greater than 2. Introducing the reduced quantities n and ii, which are denned by
457
where [P], [L], [PL], efc., are the molar concentrations of free albumin, free T4 (or T8), and the 1 : 1 complex, etc., then (11) • •}
and
(2)
Equation 1 may be rewritten «=Yo+Y1«1+ £ Yjij
(3)
If i is defined as the average value of i for the higher species ( » 1 ) , i=ZYjijlEYj= 2
Thus Ki is obtainable from the initial slope of the plot of 1/Yo with respect to [L]. The quantity «i may be determined from the initial slope of the plot of i as a function of r. From Eqs. 3, 5, 10, and 11,
(4)
(12) then
r= Y —Y Yi
^ c: \
Defining r as the average number of T4 (or Ti) molecules bound per molecule of albumin, r=Y 1 + f ( l - Y , - Y 1 )
•Ki\LVI(l+
The quantity » may be expanded as a MacLaurin series in terms of r: (13)
(6)
The limiting slope of i as a function of r is where r is the average value of r for the given by higher species and is defined by -(di/dr),.0=l-ii (14) f = E rjY,l E Y, For the determination of Ku as in the y>i i>i case of the T< (or T,>BSA (or HSA) system (7) = E ryYj/a-Yo (
Interaction of Diiodo-L-tyrosine and Triiodophenol with Bovine Serum Albumin Circular Dichroism and Fluorescence Studies Nobuo OKABE, Noriko MANABE,1 Ryozi TOKUOKA, and Ken-ichi TOMITA Faculty of Pharmaceutical Sciences, Osaka University, Yamada-kami, Suita, Osaka 565 Received for publication, February 2, 1976
As a model study to investigate the binding mechanism between thyroid hormones and carrier protein, the interaction of diiodo-L-tyrosine (DIT) and triiodophenol (h) with bovine serum albumin (BSA) was investigated by circular dichroism (CD) and fluorescence methods. In both the DIT-BSA system and the Is0-BSA system, induced Cotton effect was observed in the wavelength region near 320 nm. This induced Cotton effect was measured at various molar ratios of ligands to BSA (L/P). The value of the ellipticity at 319 nm, [6]M, in the I8$6-BSA system was remarkably large compared with that of the DIT-BSA system, and [0]s« at an L/P ratio of one was -1.96xlO 4 (degree cma decimole"1) for the 1,0-BSA system and - 0 . 1 xlO 4 for the DIT-BSA system. The binding constants for the combination of BSA with a single molecule of ligand, calculated by measuring the quenching of the fluorescence of the protein, were 1.33x10* M"1 at 15° for the DIT-BSA system and 1.6x10° M"1 at 28° for the I80-BSA system. These results suggest that the binding of lt to BSA is stronger than that of DIT and a cleft may exist more congruent with the molecular dimensions of 1$^ than with those of DIT.
Thyroid hormones, L-thyroxine (T4), and triiodo-L-thyronine (T,) t bound to thyroxine-binding globulin, prealbumin or serum albumin in the blood are transferred to their receptor sites
from the thyroid gland (1,2). In the previous paper (3), we investigated the interaction between bovine serum albumin (BSA) and thyroid hormones (T4 and T s ) by circular di-
1
Present address: CIBA-Geigy (Japan) Limited, 10-66 Miyuki-cho, Takarazuka-shi, Hyogo Prefecture. Abbreviations: BSA, bovine serum albumin ; HSA, human serum albumin ; T DlIODO-L-TYROSINE
(DIT) TRIIODOPHENOL
Fig. 1. Thyroid hormones and their analogs.
Circular dichroism (CD) measurements were carried out using a Jasco ORD/UV-5 spectropolarimeter equipped with a CD attachment. Samples at various molar ratios of ligands to protein (L/P) were obtained by adding various amounts of DIT or \t at concentrations of about 3 x 10"' M with a micropipette to 5 ml of albumin solution (about 3.6x10"'M in 0.1 M phosphate buffer, pH 7.4, containing 3xlO" 4 M EDTA) with continuous stirring. BSA concentration was determined spectrophotometrically, using £}* m =6.54 at 280 nm (6) and a molecular weight of 64,300 ( 7 ) . Molar ellipticity, [d], (based on the concentration of BSA) was calculated from the equation, [0]=3,3OOx(£,-er)
chroism (CD) measurements, and induced Cotton effect {4, 5) was observed at around 320 run, the absorption wavelength region of T 4 and T 8 . It was concluded that the induced Cotton effect was due to the binding of these Hgands to protein. However, it was not obvious whether the observed Cotton effect was induced by the interaction between an asymmetric region of the protein and the chromophoric environment of the a-ring or between the protein and the £-ring of T 4 and T,. (The a- and /3-rings of these ligands are illustrated in Fig. 1). In order to investigate more precisely the induced Cotton effect in relation to the structure of the ligand or the protein, we used diiodo-L-tyrosine (DIT) and triiodophenol (I30), as model compounds for the a- and /)ring moieties of T 4 (see Fig. 1), respectively. The binding characteristics of DIT and IS0 to BSA were investigated by circular dichroism and fluorescence methods. MATERIALS AND METHODS
where (EI — £r) is the difference between the molar extinction coefficients for left- and rightcircularly polarized light. The ultraviolet fluorescence was measured with a Hitachi MPF-2A spectrophotofluorometer. The excitation and emission wavelengths of protein fluorescence were 290 nm and 350 nm, respectively. The molar ratio of ligands to protein was varied in the same way as for CD measurements. Several measurements were made and the emitted intensities were averaged for the DIT-BSA and I3?J-BSA systems. All fluorescence measurements were made in 0.1 M phosphate buffer, pH 7.4, containing 3xlO" 4 M EDTA at the temperatures indicated in each figure. The primary binding constant between BSA and a single molecule of DIT or Is0, KU was calculated from the fluorescence quenching data using a procedure entirely analogous to that employed by Steiner et al. (8). Their procedure for the determination of the primary binding constants for the T 4 (or Ti)-BSA and T 4 (or T,)-HSA systems is as follows: The observed relative intensity (I) is given by
Crystallized bovine serum albumin (BSA) and diiodo-L-tyrosine (DIT) were purchased from Y,Iy (1) Sigma Chemical Co., and triiodophenol (Ji) was from Nakarai Chemicals, Ltd. These reagents were used without further purification. where Yo is the mole fraction of albumin which DIT and Is$6 were dissolved in distilled water has no bound T 4 (or T 8 ); Io is the relative and 30% ethanol, respectively. fluorescence intensity corresponding to such / . Biochem.
SMALL MOLECULE-MACROMOLECULE INTERACTIONS
uncombined albumin; Yi and I t are the corresponding quantities for the species with 1 mole of bound T 4 (or T 3 ); and Yj and Iy are the corresponding parameters for the species with j moles of bound T 4 (or Tj), where j is equal to or greater than 2. Introducing the reduced quantities n and ii, which are denned by
457
where [P], [L], [PL], efc., are the molar concentrations of free albumin, free T4 (or T8), and the 1 : 1 complex, etc., then (11) • •}
and
(2)
Equation 1 may be rewritten «=Yo+Y1«1+ £ Yjij
(3)
If i is defined as the average value of i for the higher species ( » 1 ) , i=ZYjijlEYj= 2
Thus Ki is obtainable from the initial slope of the plot of 1/Yo with respect to [L]. The quantity «i may be determined from the initial slope of the plot of i as a function of r. From Eqs. 3, 5, 10, and 11,
(4)
(12) then
r= Y —Y Yi
^ c: \
Defining r as the average number of T4 (or Ti) molecules bound per molecule of albumin, r=Y 1 + f ( l - Y , - Y 1 )
•Ki\LVI(l+
The quantity » may be expanded as a MacLaurin series in terms of r: (13)
(6)
The limiting slope of i as a function of r is where r is the average value of r for the given by higher species and is defined by -(di/dr),.0=l-ii (14) f = E rjY,l E Y, For the determination of Ku as in the y>i i>i case of the T< (or T,>BSA (or HSA) system (7) = E ryYj/a-Yo (
