2.1. Effect of Glycation on Serum Albumin Tertiary Structure—Fluorescence Characteristic
To prove the impact of glycation by fructose—“fructation”—on the formation of Advanced Glycation End-products (AGEs) in human serum albumin, emission and synchronous fluorescence spectra of AGEs coming from HSA and created in gHSA
FRC were recorded at λ
ex = 370 nm (
Figure 3a) and at the wavelengths range λ
ex = 320–440 nm (Δλ = λ
em − λ
ex = 40 nm) (
Figure 3b).
From the results presented in
Figure 3, the higher relative fluorescence intensity obtained for gHSA
FRC compared with HSA indicates that the glycation products in gHSA
FRC are formed. After excitation at λ
ex = 370 nm (
Figure 3a) the increase in the AGEs fluorescence intensity in gHSA
FRC compared to HSA was about 80% with accompanied by a blue-shift maximum fluorescence (Δλ = 5 nm). Specific emission fluorescence intensity (λ
em = 440 nm) observed for gHSA
FRC at 370 nm excitation wavelength indicates the formation of argpyrimidine (a typical fluorescent AGEs) [
19]. It can be seen in
Figure 3b, the marked increase in fluorescence intensity observed in glycated sample (F = 37) was about 80% compared to non-glycated HSA (F = 7.30). The main characteristic of synchronous fluorescence spectra of AGEs in gHSA
FRC is the red-shift maximum fluorescence (Δλ = 13 nm) from λ
em = 407 nm to λ
em = 420 nm. This miscellaneous shift of maximum fluorescence after albumin glycation (
Figure 3a,b) indicates that fluorescence AGEs are chemically heterogeneous compounds. Because in the circulation, the HSA becomes glycated by reducing sugars, and the reference range of a healthy person vary between 1% and 10% [
20], the blue-shift or red-shift of AGEs maximum fluorescence in gHSA
FRC compared with HSA makes that AGEs environment becomes more or less hydrophobic. In our previous study [
21] we also observed the blue-shift of bovine serum albumin maximum fluorescence under the glycation process by fructose indicating the reduction in the polarity of AGEs environment.
Synchronous (
Figure 4, main view) and emission (
Figure 4, insert) fluorescence spectra of HSA and gHSA
FRC were used to show the conformational changes in the environment of the tryptophanyl (Trp-214) and tyrosine residues (Tyr-30, -84, -138, -140, -148, -150, -161, -263, -319, -332, -334, -341, -353, -370, -401, -411, -497) of human serum albumin influenced by glycation process. As is known, a wavelength of 275 nm excites not only Trp-214 but also tyrosine residues and it is impossible to observe separately the fluorescence of these fluorophores. Synchronous fluorescence spectroscopy allows for separation of the emission spectra originating from the Trp-214 and Tyrs (as illustrated in
Figure 4a, main view), which results more specific informations about the structure of the protein. According to literature data [
22,
23], the synchronous fluorescence spectra were obtained considering the wavelength intervals Δλ = 60 nm and Δλ = 15 nm to evidence the Trp-214 and Tyrs, respectively (Δλ = λ
em − λ
ex).
Fluorescence of HSA fluorophores is sensitive to the changes of albumin tertiary structure and environmental properties. Albumin slight structural changes near the Trp-214 and Tyrs residues affect the fluorescence intensity and position of maximum fluorescence (λ
max) [
24]. The shift of λ
max position corresponds to the change in polarity around the chromophore of molecule. A blue-shift of λ
max indicates that the amino acid residues are located in more hydrophobic environment, while a red-shift of λ
max implies that the Trp-214 and Tyrs residues are in a polar environment and are more exposed to the solvent [
22]. Using Δλ = 60 nm (
Figure 4a, main view) and Δλ = 15 nm (
Figure 4b, main view), no changes in the maximum emission wavelength of HSA and gHSA
FRC Trp-214 and tyrosine residues were observed. It points to the stability of both bands in the synchronous spectra, irrespective of glycation process. No synchronous spectra shift caused by fructose glycation indicates no change in the polarity around Trp-214 and Tyr residues or/and a modification of the structure of human serum albumin in the environment of other residues, e.g., Cys-34. On the other hand, the main characteristic of gHSA
FRC emission fluorescence spectra excited at λ
ex = 275 nm (
Figure 4b, insert) is the blue-shift maximum fluorescence (Δλ = 9 nm) from λ
em = 331 nm to λ
em = 322 nm. This phenomenon suggests that Trp-214 and Tyr residues of glycated human serum albumin are less exposed to the solvent than non-glycated macromolecule. The fluorescence intensity of both types of fluorophores in the gHSA
FRC spectrum is lower than in the HSA. The reduction in Trp-214 and Tyrs fluorescence intensities at λ
max of gHSA
FRC relative to HSA 45.55% (
Figure 4a main view), 46.27% (
Figure 4a, insert) and 20.79% (
Figure 4b, main view) and 40.24% (
Figure 4b, insert) have been registered. These results indicate an alteration of the albumin tertiary structure by “fructation” in the region of tryptophanyl and tyrosyl residues, which can affect the binding of drugs in subdomain IIA (Trp-214, Tyr-263), IB (Tyr-138, Tyr-140, Tyr-148, Tyr-150, Tyr-161), IIB (Tyr-319, Tyr-332, Tyr-334, Tyr-341, Tyr-353, Tyr-370) and IIIA (Tyr-401, Tyr-411, Tyr-452, Tyr-497). The loss of fluorescence intensity of Trp-214 observed for HSA glycated by glucose compared with non-glycated albumin Sakurai et al. [
25] explained by energy transfer from the tryptophanyl residue to the newly chromophore formed in gHSA. Mendez et al. [
26] suggested that the different fluorescence of tryptophanyl residue of non-glycated and glycated albumin can be caused due to the different hydration of the whole protein induced by glycation. Nakajou et al. [
27] emphasized that an unlikely reason of the differences in fluorescence emission spectra of both albumins (non-glycated and glycated) can be directly modification of the tryptophanyl residue.
In order to study the structure-function relationship in proteins it is necessary to appreciate the environment and dynamics of albumin fluorophores. Red Edge Excitation Shift (REES) is an another method to directly monitor of the region surrounding the tryptophanyl residue of non-glycated and glycated human serum albumin [
28,
29]. In order to study REES effect, fluorescence spectra of HSA (
Figure 5a, insert) and gHSA
FRC (
Figure 5b, insert) excited at λ
ex = 290 nm, λ
ex = 295 nm and λ
ex = 300 nm wavelengths have been recorded. Emission fluorescence spectra of gHSA
FRC Trp-214 residue is different than for Trp-214 of HSA at all excitation wavelengths. A slight red-shift maximum emission of gHSA
FRC fluorescence (Δλ
em = 5 nm) relative to HSA (Δλ
em = 2 nm) has been obtained (
Figure 5b, insert). Higher shift for glycated albumin indicates that the “fructation” of HSA decreases mobility of Trp-214 inducing changes of albumin conformation. Similarly, larger REES in case of modified-oxidized (oHSA, Δλ
em = 39 nm) vs. non-modified (HSA, Δλ
em = 4 nm) human serum albumin Maciążek-Jurczyk et al. have observed [
30]. As the authors mentioned, it points to the structural modifications in the hydrophobic pocket containing the tryptophanyl residue due to the oxidation process, which contribute to stiffening of the Trp-214 environment or/and limited access to the polar solvent.
A more sensitive indicator of spectral shifts is the parameter A (
, which is less sensitive to experimental errors and as a consequence provides more accurate position of fluorescence spectra than in comparison with a position of the maximum fluorescence (λ
max) [
31]. In order to verify change in the fluorescence intensity of HSA and gHSA
FRC, spectral parameter A has been calculated (
Figure 5, main view). With the increase of excitation wavelength from 290 nm to 300 nm spectral parameter A decreases 1.3 times and 1.5 times for non-glycated and glycated albumin, respectively. It means that fluorescent spectra of tryptophanyl residue of HSA and gHSA
FRC shift towards long wavelengths (red-shift).
In this paper, we have also used another sensitive and useful method capable of identifying subtle structural changes in the tertiary conformation of albumin caused by glycation namely second derivative of fluorescence spectra. One advantage of using this method is based on the possibility of monitoring processes in albumins, which involves relatively small changes in the environment of aromatic amino acids residues not clearly visible in classical fluorescence spectra and even in fourth derivative absorption spectroscopy [
32].
Figure 6 presents the comparison of non-glycated (HSA) and glycated albumin emission spectrum normalized to non-glycated ((gHSA
FRC)
norm) and their second derivative fluorescence spectra (2nd HSA), (2nd (gHSA
FRC)
norm) for the excitation λ
ex = 275 nm (
Figure 6a) and λ
ex = 295 nm (
Figure 6b). The changes in the second derivative spectra in the wavelength range 370–400 nm and in the region below 320 nm point to the structure reorganization in albumin microenvironment, where tryptophan (Trp-214) and seventeen tyrosyl residues are located (in hydrophobic subdomain IIA and IB, IIB, IIA, IIIB), respectively [
32,
33]. It points the advantage of using the second derivative over emission fluorescence spectra, which it is more difficult to discriminate both tyrosine and tryptophan emissions [
32].
At λ
ex = 275 nm the second derivative fluorescence spectra of non-glycated albumin (2nd HSA) exhibits three peaks maximum at wavelengths 301 nm, 324 nm and 331 nm and marked one valley at 311 nm, while the second derivative spectra of normalized glycated albumin (2nd (gHSA
FRC)
norm) has peaks maximum at wavelengths 302 nm and 320 nm with marked valley at 314 nm and also one shoulder from the red side of the peak at wavelength 331 nm (
Figure 6a, main view). For λ
ex = 295 nm the second derivative fluorescence spectra of both albumins exhibit only one peak maximum at wavelengths 329 nm (for HSA) and 323 nm (for (gHSA
FRC)
norm) and shoulder at 333 nm (
Figure 6b, main view). Hypsochromic shift of second derivative spectra of tryptophanyl (Trp-214) residue of (gHSA
FRC)
norm in comparison with the second derivative fluorescence spectra of Trp-214 of HSA is observed at λ
ex = 275 nm (
Figure 6a, insert). The second derivative spectra of Trp-214 of non-glycated albumin has only one peak maximum appear at 382 nm with a slightly marked shoulder from the red side of the peak, while the second derivative fluorescence spectra of Trp-214 of glycated albumin ((gHSA
FRC)
norm) exhibits two peaks maximum at 378 nm and at 394 nm. Hypsochromic shift indicates that in glycation human albumin environment around tryptophanyl residue (surrounding of subdomain IIA) becomes more hydrophobic. It confirms our previous conclusion obtained from the analysis of HSA and gHSA
FRC emission fluorescence spectra excited at λ
ex = 275 nm (
Figure 4b, insert) that Trp-214 residue of glycated human serum albumin is less exposed to the solvent than tryptophanyl residue of non-glycated macromolecule. Two maxima in both second derivative fluorescence spectra of Trp-214 HSA and (gHSA
FRC)
norm are observed at excitation λ
ex = 295 nm (
Figure 6b, insert): for HSA at wavelengths 379 nm and 392 nm and for (gHSA
FRC)
norm at wavelengths 378 nm and 388 nm. As illustrated in
Figure 6b only second peak of Trp-214 of 2nd (gHSA
FRC)
norm in comparison with Trp-214 of 2nd HSA is blue-shifted. On the contrary to the results observed for albumins excited at λ
ex = 275 nm (
Figure 6a, insert), the more changes in the fluorescence intensities of mentioned peaks is observed for albumins excited at λ
ex = 295 nm (
Figure 6b, insert). The use of the second derivative of the fluorescence spectra has shown that not only primarily Trp-214 located in subdomain IIA, but also Tyr residues located in subdomain IB, IIB, IIA and IIIB human serum albumin participate in the process of glycation.
In our studies we used a sensitive indicator for monitoring changes in the degree of polarity in the environment of Trp-214 and Tyr residues in both albumin (HSA and gHSA
FRC)—empirical parameter
H (relative peak composition) [
32]. The values of parameter
H calculated for tryptophanyl and tyrosyl residues of non-glycated (HSA) and normalized glycated (gHSA
FRC)
norm albumin for λ
ex = 275 nm and λ
ex = 295 nm are collected in
Table 1.
Glycation of serum albumin causes the increase in polarity around the tyrosyl (Tyr) residues (λ
ex = 275 nm) and the decrease in polarity around Trp-214 (λ
ex = 295 nm) that was shown as increase (
H275nm) and decrease (
H295nm) in the value of parameter
H in (gHSA
FRC)
norm, respectively (
Table 1). The decrease in polarity around Trp-214 in glycated human serum albumin results in blue-shift of second derivative fluorescence spectra at excitation λ
ex = 275 nm (
Figure 6a, insert). Qualitative analysis of the second derivative spectra indicate that glycation of human serum albumin reorganizes the structure of macromolecule around Trp-214 and Tyr residues, reflecting the subtle changes in the HSA tertiary structure.
Subdomains of human serum albumin (A and B in domain I, II and III) with separate helical structures mediate albumin binding with various endogenous and exogenous ligands. There are two main binding sites for drugs in the albumin structure. According to Sudlow’s nomenclature–sites I (located in subdomain IIA) has binding affinity for heterocyclic compounds such as phentylbutazone and warfarin and site II (hydrophobic pocket in subdomain IIIA) binds to aromatic compounds such as ibuprofen [
4]. Because based on the emission and synchronous fluorescence spectroscopy the changes in glycated HSA structure in the region of tryptophanyl (subdomain IIA) and tyrosyl (subdomain IB, IIB, IIA and IIIA) residues have been obtained, in order to determine the influence of glycation on hydrophobic nature of the specific binding sites, the method of fluorescent probes: 5-dimethylaminonaphthalene-1-sulfonamide (DNSA), warfarin (WAR), dansyl-
l-glutamine (dGln) and
N-dansyl-
l-proline (dPro) was used. These probes do not fluoresce or exhibit weak fluorescence in the polar environment while strongly fluoresce in organic solvents (non-polar environment) or when combined with hydrophobic protein structures [
34]. As seen in
Figure 7 fluorescence intensity of DNSA, WAR, dGln and dPro in the complex with HSA and gHSA
FRC increases.
By titrating the protein solution by DNSA at increasing concentration no significant difference in the fluorescence intensity of the probe in the presence of HSA and gHSA
FRC has been observed (
Figure 7a). In turn, by titrating the HSA and gHSA
FRC by warfarin solution it has been suggested that the binding of WAR with glycated HSA is stronger than with non-glycated albumin (
Figure 7b). DNSA, similarly as warfarin, locates in hydrophobic regions of Sudlow’s site I in subdomain IIA [
27,
35]. This place consists of three subregions: Ia, Ib and Ic [
36]. Registered difference in warfarin and DNSA fluorescence may result from the binding of probes in the different subregions of albumin. Based on the conducted experiment it can be concluded that glycation of HSA changes its conformation in the environment of macromolecule subdomain IIA however the magnitude of change is different for each of the Sudlow’s site I subregions. Dansyl amino acids bind to hydrophobic sites of serum albumin and dansyl-
l-glutamine (dGln) was used as a marker for Sudlow’s binding site I, while
N-dansyl-
l-proline (dPro) for Sudlow’s binding site II in the HSA molecule [
37]. The experiment with the dGln and dPro probes conducted a gradual increase in dGln-HSA, dGln-gHSA
FRC (
Figure 7c) and dPro-HSA, dPro-gHSA
FRC (
Figure 7d) fluorescence with the increase of the probe concentration. An increase in probe fluorescence, stronger for glycated than non-modified albumin, is a proof of the influence of glycation on conformation changes, both in the region of subdomain IIA, and IIIIA. Fluorescence analysis enabled the conclusion that environment of both binding site I and II is modified by fructose glycation.
2.5. Fluorescence Quenching of Non-Glycated and Glycated Human Serum Albumin Induced by Tolbutamide and Losartan in the Binary and Ternary Complex
Fluorescence quenching of proteins can be used to obtain detailed ligand-albumin binding informations. The fluorescence quenching of non-glycated (HSA) and glycated (gHSA
FRC) human serum albumin excited at λ
ex = 275 nm and λ
ex = 295 nm in the binary (TB-HSA, TB-gHSA
FRC (
Figure 10a) and LOS-HSA, LOS-gHSA
FRC (
Figure 11a)) complexes was conducted to determine interaction of tolbutamide (TB) and losartan (LOS) in binding hydrophobic pockets with both albumins HSA and gHSA
FRC in the high-affinity binding sites. Glycation altered the microenvironment around tyrosyl (Tyr) and tryptophanyl (Trp-214) residues and as was described in
Section 2.1 and
Section 2.4, quenching curves of HSA and gHSA
FRC excited at λ
ex = 275 nm and λ
ex = 295 nm were compared (
Figure 10b and
Figure 11b). The use of the excitation wavelength of 275 nm allows to observe one tryptophanyl residue (Trp-214) and seventeen tyrosyl residues (Tyr-30, -84, -138, -140, -148, -150, -161, -263, -319, -332, -334, -341, -353, -370, -401, -411, -497) of non-glycated and glycated albumin, whereas 295 nm wavelength excites only Trp-214 of the albumins. The quenching protein fluorescence takes place when the distance between the chromophores of aromatic rings in ligand chemical structure and the fluorophores (tryptophanyl or/and tyrosyl residues) of albumin is smaller than 10 nm, typically 1–10 nm. Then, the Fluorescence Resonance Energy Transfer (FRET) donor (fluorophore)—acceptor (chromophore) is possible [
43].
The quenching curves of HSA and gHSA
FRC in the presence of tolbutamide at increasing concentration 1 × 10
−5 mol∙L
−1 to 1 × 10
−4 mol∙L
−1 (molar ratio TB:HSA and TB:gHSA
FRC 2:1 to 20:1) (
Figure 10a) and losartan at increasing concentration 5 × 10
−6 mol∙L
−1–5 × 10
−5 mol∙L
−1 (molar ratio LOS:HSA and LOS:gHSA
FRC 1:1 to 10:1) (
Figure 11a) show the decrease in both non-glycated and glycated albumin fluorescence for λ
ex = 275 nm and λ
ex = 295 nm. Correction for inner filter effect has been applied (Equation (5)), so the quenching of HSA and gHSA
FRC fluorescence could be considered as a result of the formation of TB-HSA, TB-gHSA
FRC and LOS-HSA, LOS-gHSA
FRC complex. Tolbutamide quenches fluorescence HSA by 23% and by 33% for excitation 275 nm and 295 nm (
Figure 10a, main view), respectively. Fluorescence of gHSA
FRC excited at λ
ex = 275 nm and λ
ex = 295 nm decreases by 18% and 35%, respectively, for the same molar ratio ligand:albumin (
Figure 10a, insert). Losartan quenches fluorescence HSA by 36% and by 47%, respectively, for λ
ex = 275 nm and λ
ex = 295 nm, at molar ratio LOS:HSA 10:1 (
Figure 11a, main view). Fluorescence of gHSA
FRC excited wavelength at 275 nm and 295 nm decreases at the same LOS:gHSA
FRC molar ratio by 33% and 50%, respectively (
Figure 11a, insert). The quenching curves of HSA and gHSA
FRC excited at λ
ex = 275 nm and λ
ex = 295 nm in the presence of TB (
Figure 10a) and LOS (
Figure 11a) at increasing drug concentration do not overlap. This phenomenon probably means that in the interaction of tolbutamide and losartan with both serum albumins the tryptophanyl residue of subdomain IIA (Trp-214) and tyrosyl residues located in the hydrophobic subdomains i.e., IB, IIB, IIIA and IIIB take part. As seen in
Figure 10a and
Figure 11a, fluorescence quenching of the proteins by TB and LOS is more extended excited at 295 nm than that excited at 275 nm. The changes observed in the run of the quenching curves at λ
ex = 275 nm and λ
ex = 295 nm probably indicate significant participation of Trp-214 in the interaction between ligands (TB, LOS) and albumins (HSA, gHSA
FRC).
The quenching curves of non-glycated and glycated albumin excited at 275 nm and 295 nm in the presence of both TB (
Figure 10b) and LOS (
Figure 11b) shows slight differences which originate from lower by 5% and 3% quenching of gHSA
FRC by TB (
Figure 10b, main view) and LOS (
Figure 11b, main view) and greater by 2% and 3% quenching of gHSA
FRC by TB (
Figure 10b, insert) and LOS (
Figure 11b, insert). This may be explained by structural changes around Trp-214 and Tyr microenvironment as a result of glycation.
The influence of TB on the LOS and LOS on the TB affinity towards HSA and gHSA
FRC was studied by the comparison of the quenching curves of albumins (HSA, gHSA
FRC) in the presence of TB in the binary TB-HSA, TB-gHSA
FRC and ternary TB-LOS
(const)-HSA, TB-LOS
(const)-gHSA
FRC (
Figure 12) and in the binary LOS-HSA, LOS-gHSA
FRC and ternary LOS-TB
(const)-HSA, LOS-TB
(const)-gHSA
FRC complexes (
Figure 13).
The quenching of fluorescence of TB-LOS
(const)-HSA (
Figure 12a,c), TB-LOS
(const)-gHSA
FRC (
Figure 12b,d) and LOS-TB
(const)-HSA (
Figure 13a,c), LOS-TB
(const)-gHSA
FRC (
Figure 13b,d) systems slightly differs from the binary TB-HSA, TB-gHSA
FRC and LOS-HSA, LOS-gHSA
FRC systems, respectively. The quenching of HSA and gHSA
FRC fluorescence by TB and LOS at the constant concentration is lower by 4%, 1% (for λ
ex = 275 nm) and 2%, 5% (for λ
ex = 295 nm) than in the system without additional ligand added to the binary system. The presence of LOS probably makes the interaction TB-HSA and TB-gHSA
FRC more difficult or hinders the formation of the TB-HSA and TB-gHSA
FRC complex. Losartan may probably cause displacement of TB from its complex with non-glycated and glycated serum albumin. The quenching of HSA and gHSA
FRC fluorescence by LOS and TB at the constant concentration is lower by 7%, 3% (λ
ex = 275 nm) and 5%, 4% (λ
ex = 295 nm) than in the binary LOS-HSA and LOS-gHSA
FRC systems, respectively. It shows that transfer of energy from excited fluorophores of HSA and gHSA
FRC to losartan is easier without tolbutamide in the system. Tolbutamide can reduce the affinity of LOS to HSA and gHSA
FRC because of the hydrophobic interactions stabilizing the complex TB-HSA and TB-gHSA
FRC. It can be concluded that additional ligand changes structure of HSA and gHSA
FRC or/and character of binding and this may suggests that the presence of the second drug (LOS or TB) causes drug-albumin complex more stable.
Based on the Stern-Volmer curves (Equation (6)) the mode of the interaction between tolbutamide or losartan and both serum albumin (non-glycated and glycated) in binary TB-HSA, TB-gHSA
FRC, LOS-HSA, LOS-gHSA
FRC and ternary TB-LOS
(const)-HSA, TB-LOS
(const)-gHSA
FRC, LOS-TB
(const)-HSA, LOS-TB
(const)-gHSA
FRC systems were analyzed (data not shown). The dependence
on TB or LOS concentration when tryptophanyl (Trp-214) and 17 tyrosyl residues of the albumin have been excited displays negative deviation from the linearity. Two of the reasons is the presence of more than one fluorophore with different accessibility to the quencher (TB or LOS) and different value of the Stern-Volmer constant
(moL
−1∙L) or the system contains a fluorophore in different environments [
44,
45]. The negative deviation observed in the Stern-Volmer plots is also explained in terms of intramolecular and intermolecular hydrogen bond complex formation with the fluorophore [
46]. Analysis of the modified by Lehrer Stern-Volmer plots (Equation (7),
Figure 14a–d) allows to determine the Stern-Volmer constant
(mol
−1∙L) and the fractional maximum protein fluorescence accessible for the quencher
for the binary (
Table 3) and ternary systems (
Table 4). The
is a mean value of the Stern-Volmer constants characterizing all binding sites of human serum albumin.
The Stern-Volmer constant is used to assess the availability of the quencher to the excited fluorophore. The growth of
value is associated with the increase of ligand molecule availability to the macromolecule and the formation of the complex in an excited state. The
determined on the basis of Stern-Volmer equation modified by Lehrer (Equation (7)) for binary system with non-glycated albumin (TB-HSA, LOS-HSA) is higher than that for the ternary system (TB-LOS
(const)-HSA, LOS-TB
(const)-HSA) while for binary system with glycated albumin (TB-gHSA
FRC, LOS-gHSA
FRC)
is lower than that for the ternary system (TB-LOS
(const)-gHSA
FRC, LOS-TB
(const)-gHSA
FRC) (
Table 3 and
Table 4). The presence of TB and LOS probably makes formation of LOS-HSA and TB-HSA complex difficult while the presence of TB and LOS in the system makes formation of LOS-gHSA
FRC and TB-gHSA
FRC complex easier, respectively. The rate of biomolecular quenching constants
(10
12) for the binary (
Table 3) and ternary system (
Table 4) points to the static mechanism of fluorescence quenching. By Lakowicz, the maximum value of the
constant for collision fluorescence quenching in the aqueous solution equals to 2 × 10
10 (mol
−1∙L∙s
−1) [
47]. Static quenching leads to a decrease in the intensity of emitted fluorescence when the ligand binds to a fluorophore molecule in its basic state (non-excited) and reduces the population of fluorescents capable of excitation [
48]. The Stern-Volmer values and quenching rate constants obtained for glycated albumin are higher than in comparison with
and
values for non-glycated macromolecule. These results indicate that TB and LOS molecules locate closer to fluorophores of gHSA
FRC than HSA, both in the binary and ternary complexes. This phenomenon may probably suggests that tolbutamide (TB-HSA, TB-LOS
(const)-HSA) and losartan (LOS-HSA, LOS-TB
(const)-HSA) bind to non-modified albumin at such a distance that makes transfer the donor-acceptor energy difficult.
A model of drug-binding to non-glycated and glycated human serum albumin in binary and ternary system has been obtained on the basis of the binding isotherms plotted based on the
dependence and representative data for λ
ex = 275 nm are presented in
Figure 15. The nonlinear shape of the isotherms for TB-HSA, TB-gHSA
FRC (
Figure 15a), LOS-HSA, LOS-gHSA
FRC (
Figure 15b), TB-LOS
(const)-HSA, TB-LOS
(const)-gHSA
FRC (
Figure 15c) and LOS-TB
(const)-HSA, LOS-TB
(const)-gHSA
FRC (
Figure 15d) complexes indicates a mixed (specific and non-specific) nature of drugs interaction with both albumins in the binary and ternary system, respectively.
It means that non-specific binding sites on a HSA and gHSAFRC surface, in the neighborhood of excited tyrosyl residues or/and formation of TB-HSA, TB-gHSAFRC, LOS-HSA, LOS-gHSAFRC and TB-LOS(const)-HSA, TB-LOS(const)-gHSAFRC, LOS-TB(const)-HSA, LOS-TB(const)-gHSAFRC complexes in hydrophobic pocket of albumin takes place.
The Scatchard curves (the dependence of
on
,
Figure 16) and the Klotz curves (the dependence of
on
, data not shown) allowed to determine association constants
(mol
−1∙L) and the number of binding sites
for the independent class of drug binding sites in albumin. The changes in high affinity binding of TB and LOS to non-glycated and glycated serum albumin in the binary and ternary systems on the basis of association constants
, the number of TB and LOS molecules bound with 1 mole of HSA and gHSA
FRC in a particular binding site
and also Hill’s coefficient
cooperative obtained for the binary (TB-HSA, TB-gHSA
FRC, LOS-HSA, LOS-gHSA
FRC) and ternary systems (TB-LOS
(const)-HSA, TB-LOS
(const)-gHSA
FRC, LOS-TB
(const)-HSA, LOS-TB
(const)-gHSA
FRC) have been summarized in
Table 5 and
Table 6, respectively (λ
ex = 275 nm and λ
ex = 295 nm).
The Scatchard plot determined for TB-HSA, TB-gHSA
FRC (
Figure 16a) and LOS-HSA, LOS-gHSA
FRC (
Figure 16b) complex shows a linear dependence. This results indicates the existence of one class of specific and also non-specific binding sites on the surface of albumin for TB and LOS in HSA and gHSA
FRC. The association constant
for TB-HSA and TB-gHSA
FRC complex is equal to (2.84 ± 0.07) × 10
4 mol
−1∙L and (5.17 ± 0.23) × 10
4 mol
−1∙L for λ
ex = 275 nm and (1.94 ± 0.09) × 10
4 mol
−1∙L and (2.61 ± 0.19) × 10
4 mol
−1∙L for λ
ex = 295 nm, respectively (
Table 5). The value of
for LOS-HSA and LOS-gHSA
FRC complex equals to (8.13 ± 0.41) × 10
4 mol
−1∙L and (9.21 ± 0.40) × 10
4 mol
−1∙L for λ
ex = 275 nm and (4.62 ± 0.07) × 10
4 mol
−1∙L and (5.31 ± 0.11) × 10
4 mol
−1∙L for λ
ex = 295 nm, respectively (
Table 6). Both TB and LOS has a high affinity towards hydrophobic subdomains IB, IIB, IIIA and IIIB of HSA due to the glycation (λ
ex = 275 nm) and lower, when only tryptophanyl residue was excited (λ
ex = 295 nm). The growth of
in TB-gHSA
FRC and LOS-gHSA
FRC versus TB-HSA and LOS-HSA complexes means that in vitro glycation of albumin with lower SH-content (
Section 2.2 and
Section 3.3) has a higher affinity for TB and LOS compared with non-glycated albumin. In addition, losartan at 10:1 LOS:HSA (LOS:gHSA
FRC) molar ratio has a higher affinity for HSA and gHSA
FRC than tolbutamide at 20:1 TB:HSA (TB:gHSA
FRC) molar ratio. This effect shows that the transfer of energy from albumin fluorophores (Trp-214 and Tyrs) to LOS is easier than to TB. Using high-performance affinity chromatography and frontal analysis Joseph et al. studied the binding of tolbutamide to non-glycated and glycated human serum albumin at different levels of glycation [
49]. The authors observed that
for tolbutamide increased by 1.2- to 1.3-fold and by 1.1- to 1.4-fold in going from normal HSA to all glycated HSA at Sudlow’s site I and II, respectively. They emphasized that glycation of albumin may affect the rate of metabolism or extraction and the overall half-life of TB in the circulation. Similarly as in the previous analyzed binary systems TB-HSA, TB-gHSA
FRC and LOS-HSA, LOS-gHSA
FRC, linear run of Scatchard plot for λ
ex = 275 nm (main view) and λ
ex = 295 nm (insert) indicates the existence of one class of specific and mixed, i.e., specific and non-specific binding sites for ternary system TB-LOS
(const)-HSA, TB-LOS
(const)-gHSA
FRC (
Figure 16c) and LOS-TB
(const)-HSA, LOS-TB
(const)-gHSA
FRC (
Figure 16d).
In order to investigate the effect of losartan and tolbutamide interactions with non-glycated and glycated albumin, the association constant
for TB-HSA and TB-gHSA
FRC complex in the presence of LOS was carried out (
Table 5). The association constants
obtained for TB-HSA are slightly higher than that obtained for TB-LOS
(const)-HSA (
= (2.61 ± 0.10) × 10
4 mol
−1∙L,
= (1.90 ± 0.14) × 10
4 mol
−1∙L), while
obtained for TB-gHSA
FRC are slightly lower than obtained for TB-LOS
(const)-gHSA
FRC (
= (6.03 ± 0.17) × 10
4 mol
−1∙L,
= (3.86 ± 0.20) × 10
4 mol
−1∙L) complex, at both λ
ex = 275 nm and λ
ex = 295 nm. Tolbutamide has almost the same affinity to HSA than LOS
(const)-HSA complex and to gHSA
FRC than LOS
(const)-gHSA
FRC complex. Losartan added to the TB-HSA and TB-gHSA
FRC complex at 1:1 LOS:HSA and at 1:1 LOS:gHSA
FRC molar ratio has the abilities to form a complex with non-glycated and glycated serum albumin, where tyrosyl residues or/and tryptophanyl residue take places (in subdomain IB, IIB, IIA, IIIB or/and IIA). Because TB and LOS have albumin common binding sites, the possibility of competitive interaction between analyzed drugs should be taken into account. The number of TB molecules bound to one molecule of both HSA and gHSA
FRC was observed and do not change in the presence of LOS. This is similar to the results obtained for binary and ternary complexes (
Table 5). Comparison of association constants calculated for binary LOS-HSA, LOS-gHSA
FRC and ternary complexes (LOS-TB
(const)-HSA:
= (2.10 ± 0.18) × 10
4 mol
−1∙L,
= (2.64 ± 0.09) × 10
4 mol
−1∙L and LOS-TB
(const)-gHSA
FRC:
= (10.37 ± 0.70) × 10
4 mol
−1∙L,
= (4.26 ± 0.07) × 10
4 mol
−1∙L for λ
ex = 275 nm and λ
ex = 295 nm, respectively) confirms the existence of the competition between losartan and tolbutamide and the displacement of losartan from the binding site especially in non-glycated human serum albumin. It was found that the presence of TB changes the affinity of non-glycated albumin towards losartan binding site. This phenomenon is associated with the fact, that tolbutamide at 1:1 TB:HSA molar ratio interacts with tryptophanyl residue or/and tyrosyl residues located in the subdomain IB, IIB, IIA, IIIB or/and IIA, which are probably the common binding sites for both TB and LOS. Tolbutamide displaces losartan from the complex or makes the formation of LOS–HSA more difficult. The decrease in
values in LOS-HSA complex due to the presence of TB means a reduction of studied system stability. The mean number of LOS molecules bound to one molecule of HSA in the given class of binding sites
is about 1 and decreased to about 0.5 for ternary LOS-TB
(const)-HSA complex. On the other hand, glycation of albumin changes tyrosyl or/and tryptophanyl residues environment, that in the presence of TB the association constant obtained for LOS-TB
(const)-gHSA
FRC is slightly higher than that obtained for LOS-gHSA
FRC complex at λ
ex = 275 nm and slightly lower at λ
ex = 295 nm. The mean number of LOS molecules bound to one molecule of glycated albumin in the given class of binding sites
is about 1 and does not change under the influence of the additional TB (
Table 6).
The association constants
(mol
−1∙L) and the number of binding sites
(number of ligand molecules bound per protein) for the binary and ternary systems were also determined using Klotz method (the dependence of
on
). The obtained results are comparable to the
and
values determined using Scatchard method (
Table 5 and
Table 6).
In order to measure of cooperativity in a binding process, the values of the Hill coefficient has been used (Equation (5)). For TB-HSA, TB-gHSAFRC and LOS-HSA, LOS-gHSAFRC complex is equal to 1 or slightly more than 1. It indicates independent TB and LOS binding sites in both, non-modified and modified albumin or/and shows positive cooperativity—binding of one ligand facilitates binding of subsequent ligand at the sites on the ligand-protein complex. For ternary complexes (TB-LOS(const)-HSA, TB-LOS(const)-gHSAFRC, LOS-TB(const)-HSA, LOS-TB(const)-gHSAFRC) the Hill’s coefficients are equal to 1 and slightly less than 1 especially for LOS-TB-HSA ( = 0.80 ± 0.03) and TB-LOS-gHSAFRC ( = 0.88 ± 0.02). This indicates negative cooperativity—binding of one ligand hinders binding of subsequent ligands at the sites on the complex.