2.1. Absorption and Fluorescence Characteristic of Aged Serum Albumin—Second Derivative
The normalized absorbance spectra of aged human (HSA) and bovine (BSA) serum albumin in phenylalanine (Phe), tyrosine (Tyr), and tryptophans (Trp) regions are shown in Figure 1
a,c. Figure 1
b,d presents their second derivatives.
A contribution of the protein aromatic residues, phenylalanine (Phe), tyrosyl (Tyr), and tryptophanyl (Trp), has been determined using derivative spectroscopy. It is noteworthy that the contribution of (Phe), (Tyr), and (Trp) residues in the total protein absorbance cannot be determine using the method of zero-order spectrum because molar absorption coefficients of Phe, Tyr, and Trp differ significantly. These values were obtained by Mihályi and are equal to:
= 110.6 dm3
= 1330 dm3
= 5620 dm3
]. The first derivative spectra dA/dλ is a rate of the absorbance (A) which changes with the change of the wavelength (λ), while the second derivative dA2
is the rate at which this change occurs. As it was reported by Kuś [8
] and Karpińska [11
], subtle structures (arms, inflection, size of descent angle of spectrum), faintly visible in the zero-order absorption spectrum become more pronounced in the second derivative spectrum. Figure 1
a,c present the zero-order spectra of HSA and BSA, arithmetically normalized to the maximum absorbance of the aged HSA (AHSA) and ABSA. In order to show subtle changes of albumin’s tertiary structure due to the ageing process, the second derivative spectra have been shown in Figure 1
b,d. The second derivative absorption spectra of HSA, AHSA, BSA, and ABSA allow us to distinguish albumin chromophores. The spectral wavelength range from 250 nm to 270 nm relates to phenylalanine (Phe) residues and the participation of other aromatic chromophores in this wavelength range is negligible. A wavelength range from 293 nm to 305 nm relates only to the tryptophanyl (Trp) residues and the wavelength range between 270 nm and 293 nm relates to tyrosyl (Tyr) and tryptophanyl residues (Trp214 in HSA and BSA and Trp135 in BSA) [12
From the second derivative absorption spectra, the d2
values of AHSA, normalized HSA, BSA, and normalized BSA were determined and collected in Table 1
Qualitative analysis (Figure 1
) and the second derivative values of the analyzed spectra (Table 1
) indicate that the ageing of human and bovine serum albumin leads to the alteration of the tertiary structure around Phe, Tyr, Trp214 (HSA), and Trp135 (BSA) residues. Human serum albumin (HSA) contains 31 phenyloalanine (Phe) residues located in each of six albumin subdomains and bovine serum albumin (BSA) contains 27. Eighteen Tyr residues of HSA and 19 of BSA are deployed in five subdomains of albumin (except subdomain IIIB) [1
]. Therefore, it can be supposed that the structural changes caused by the ageing process refer not only to molecule fragments but also to whole albumin. The second derivative values obtained for aromatic amino acids residues (Trp, Tyr, Phe) are lower for both AHSA and ABSA structures and have been collected in Table 1
As it was described by Kagan et al. [14
] and Ankaru et al. [15
], during the ageing process, the aromatic (tryptophan, tyrosine, phenylalanine) and basic (cysteine, methionine, lysine, arginine, histidine, proline) amino acids are oxidized since they are more sensitive to proteolysis than other amino acids. The formation of oxidized products as a result of ageing causes the changes in the function of albumin, for example, the changes of metal ions binding (Cu2+
), the reduction of antioxidant properties, and the reduction of thioesterase activity (Trp residue is oxidized). Furthermore, the ageing process causes the tyrosine oxidation and the generation of tyrosine radicals, which form the crosslinks and aggregates with the same or different molecules. Varshavsky [16
], in his paper, described the loss of serum albumin chaperone-like activity that cause the accumulation of modified protein leading to apoptosis or necrosis. Eldeeb and Fahlman [17
], in their scientific works, demonstrated anti-apoptopic roles for the N-end rule pathway. They also presented that degradation can be modified by changing the identity of the N-terminal amino acids. They noticed this phenomenon during the investigation the proteolytical degradation of Lyn tyrosine and bone marrow kinases by the N-rule machinery [17
]. Emission fluorescence spectra were obtained by the use of two excitation-wavelengths: λex
275 nm and 295 nm. When serum albumin is excited at λex
295 nm, tryptophanyl residues are responsible for all albumin fluorescence, and tyrosyl residue fluorescence is insignificant. When the protein is excited at λex
275 nm, the fluorescence from both tryptophanyl and tyrosyl residues of albumin is observed [20
]. Figure 2
illustrates emission fluorescence spectra of non-modified (HSA, BSA) and aged (AHSA, ABSA) human and bovine serum albumin at excitation wavelengths of 275 nm (a), (c) and 295 nm (b), (d). In the inserts of Figure 2
, the spectra normalized to their respective maxima around 340 nm are shown.
For aged human serum albumin (AHSA) the increase in fluorescence intensity is observed in comparison to non-modified HSA, while the fluorescence intensity obtained for non-modified bovine serum albumin (BSA) is higher in comparison to aged BSA. Before the determination of the second derivative spectra, the fluorescence spectra of non-modified HSA were normalized to AHSA fluorescence spectra and ABSA fluorescence spectra were normalized to non-modified BSA fluorescence spectra (Figure 2
, in the insert). Second derivative spectra of non-modified (HSA, BSA) and aged (AHSA, ABSA) human and bovine serum albumin normalized to their respective maxima are presented in Figure 3
Batochromic shift of second derivative fluorescence spectra of tryptophanyl residue (Trp214) of non-modified (HSA) and aged (AHSA) human serum albumin is observed at λex
275 nm (Figure 3
a, in the insert). The peaks maxima of the second derivative fluorescence spectra of HSA tryptophanyl residue (Trp214) appear at 398 nm and 415 nm and for AHSA at 403 nm and 414 nm. The valley in HSA and AHSA second derivative fluorescence spectra occurs at 408 nm and 411 nm, respectively (Figure 3
a, in the insert). Because bathochromic shift was observed in both maxima, valley, and shoulders it, means that the aged human albumin environment around Trp214 becomes more polar. It confirmed our previous conclusions obtained from the analysis of the fluorescence emission spectra that the ageing process of HSA causes an increase in polarity in subdomain IIA (Figure 2
a, in the insert).
Second derivative fluorescence spectra of HSA tryptophanyl residue (Trp214) excited at λex
295 nm exhibit two peaks maxima (383 nm and 395 nm) and mark valley at 389 nm. While second derivative spectra of AHSA tryptophanyl residue (Trp214) have only one peak maximum (395 nm) and small shoulder from the blue side of the peak. The ageing of HSA does not cause the shift of the second derivative fluorescence spectrum (Figure 3
a, in the insert).
In comparison to non-modified BSA (Figure 3
b, in the insert) and in opposition to HSA (Figure 3
a, in the insert), the second derivative fluorescence spectra of aged BSA (λex
275 nm) is blue shifted. This phenomenon is caused by the Trp135 located not only in the BSA subdomain IB but also in the environment containing charged residues. At the excitation wavelength λex
275 nm, second derivative fluorescence spectra of non-modified bovine serum albumin (BSA) tryptophanyl residues (Trp214, Trp135) exhibit two maxima (at wavelengths 402 nm and 413 nm) and three shoulders from the blue side of the peak. Aged BSA (ABSA) has one maximum at 398 nm and a shoulder from the red side of the peak instead of the second peak observed for non-modified BSA. In the ABSA second derivative fluorescence spectra, two shoulders from the blue side are observed (Figure 3
b, in the insert).
Two maxima of both second derivative fluorescence spectra, BSA and ABSA, are observed at excitation wavelength λex
295 nm: for BSA at wavelengths 384 nm and 395 nm and for ABSA at wavelengths 383 nm and 393 nm. It means that the second derivative fluorescence spectra of aged bovine serum albumin (ABSA) is blue-shifted. It is noteworthy that the second derivative fluorescence spectra values of HSA and BSA tryptophanyl residues are higher in comparison to that of AHSA and ABSA (Figure 3
, in the inserts). The same effect was obtained using UV technique (Table 1
The values of the second derivative fluorescence spectra were determined by peak to peak
method, where the distance from the positive peak maximum to the negative peak minimum is measured [7
]. Mozo-Vilları́as [22
] defines the peak to peak
method of the derivative value as an empirical parameter H
(relative peak composition
). For the excitation λex
295 nm, parameter H
is the difference between minimum and maximum in the region between 370 nm and 400 nm of the second derivative spectra. Whereas for the excitation λex
275 nm, parameter H
is the difference between the minimum and maximum below the wavelength 320 nm of the second derivative spectra. The empirical parameter H
is an indicator of the polarity in the tryptophanyl (Trp214 in HSA, Trp214, and Trp135 in BSA) and tyrosyl (Tyr) residues environment [21
]. The changes in the second derivative value in the wavelength range between 370 nm and 400 nm (the red-side of the absorption spectrum) point to the structure reorganization around tryptophanyl residues in HSA (Trp214) and BSA (Trp214, Trp135). The changes of the second derivative value in the wavelength range below 320 nm indicate the rearrangement of the HSA and BSA structures mainly around tyrosyl residues (Tyr). The values of parameter H
are collected in Table 2
The decrease in polarity around tyrosyl residues of aged human (AHSA) and bovine (ABSA) serum albumin results in the blue-shift of second derivative fluorescence spectra at excitation λex
275 nm (the value of parameter H
increases) (Table 2
, Figure 3
). The differences in polarity occurring at the ageing process are greater for HSA than BSA.
295 nm, the increase in polarity around aged human serum albumin tryptophanyl residue (Trp-214) was observed (the value of parameter H
decreases). Mozo-Vilarias [22
] determined the value of the HSA relative peak composition value (λex
295 nm) and demonstrated that the environment of tryptophanyl residue (Trp214) is rather hydrophobic. It was confirmed by Kosa et al. [23
]. The structural changes of human albumin N-B isomerization observed by the spectroscopic techniques (circular dichroism, fluorescence, 1
H-NMR) and differential scanning calorimetry (DSC) indicated that Trp residue is buried in a hydrophobic environment and fluctuations of the residue increased with the increase of pH (Trp residue is exposed to polar environment). The increase in polarity around the Trp214 of aged HSA can reduce the effective solubility of those hydrophobic drugs bound in the subdomain IIA of albumin and influence its physiological activity. Oettl and Stauber [24
] demonstrated that both ageing and in the presence of diseases influence the redox state of albumin which results in changes of ligand binding (i.e., some non-steroidal anti-inflammatory drugs, calcium channel blocker, benzodiazepine group) to serum albumin.
The ageing of BSA causes the decrease in polarity around the tryptophanyl residues (Trp-214 and Trp-135) (λex 295 nm), but the differences are negligible in comparison to that of BSA.
The modification of basic amino acids such as lysine and arginine influenced significantly on the hydrophobicity and net charge of albumin [25
]. Moreover, Varshavsky [16
] observed shortened HSA and BSA half-life due to the alterations of the N-terminal lysine and histidine, which are modified during ageing. It is possible that HSA and BSA are degraded by the N-end rule pathway [26
2.3. Effect of N-A Isomerization on Metoprolol (MET) Binding to Serum Albumin
There are two high affinity binding sites in the HSA and BSA tertiary structure for exogenous and endogenous ligands. According to Sudlow’s nomenclature, site I includes six helices of subdomains IIA and the loop associated with subdomain IB. Site II is much smaller than site I and includes six helices of subdomain IIIA [33
]. As a confirmation of MET binding to I and II, the binding sites of HSA and BSA crystal structures have been created using computational simulation. Figure 5
and Figure 6
presents the mechanism of MET binding to the fluorophores of the analyzed proteins.
In order to determine the binding affinity of metoprolol (MET) to specific binding sites located in subdomains IIA and IIIA of the HSA and BSA tertiary structure, the method of fluorescent probes (dansyl glicyne (DG) and dansyl-l
-proline (DP)) was used. Dansyl glicyne (DG) was used as a marker for Sudlow’s binding site I (subdomain IIA), while dansyl-l
-proline (DP) was used for Sudlow’s binding site II (subdomain IIIA). Fluorescence of DG and DP in aqueous solutions is only slight, while in the complex with albumin increases [35
]. A displacement of DG and DP from their binding site in non-modified human and bovine serum albumin by MET was studied by the determination of displacement percentage of bound DG and DP. Moreover, the influence of the ageing process on displacement percentage of bound DG and DP from Sudlow’s site I and II by MET was conducted. In this way, it was possible to determine the specificity and differences in MET binding affinity to the specific sites of non-modified HSA and BSA and aged AHSA and ABSA serum albumins. Figure 7
presents the percentage of DG and DP displacement from their binding site in HSA, AHSA, BSA, and ABSA by MET at the albumin:MET molar ratio of 1:10, and the values of the percentage of displacement are collected in Table 4
In the presence of metoprolol (MET), the fluorescence intensity of all DG-albumin (DG-HSA, DG-AHSA, DG-BSA, DG-ABSA) and DP-albumin complexes (DP-HSA, DP-AHSA, DP-BSA, DP-ABSA) was decreased and a displacement of bound to albumin dansyl glicyne (DG) and dansyl-l
-proline (DP) was registered. The displacement percentage of probes (DG, DP) from non-modified HSA binding sites (DG from site I and DP from site II) by metoprolol (MET) was similar and equals 47.12% and 39.16% at HSA:DG and HSA:DP, respectively, with a molar ratio of 1:0.5. At the molar ratio of HSA to both DG and DP at 1:1, the displacement percentage of DG and DP was reduced to 27.9% and 33.29%, respectively (Table 4
). It can be concluded that the binding site of MET is located in the subdomains IIA and IIIA of the non-modified HSA structure and these results from in vitro experiments are consistent with the data obtained based on the computational simulations with the use of the Piecewise Linear Potential Protein-Ligand ANT System (PLANTSPLP
The percentage of DG displacement from non-modified BSA by MET is low and did not exceed 13.0% at both molar ratios of BSA:DG at 1:0.5 and 1:1, while the calculated percentage of DP displacement from non-modified BSA equals 41.20% and 46.05% at BSA:DP molar ratios 1:0.5 and 1:1, respectively (Table 4
). It indicates that MET also binds to subdomain IIIA of non-modified BSA.
The binding affinity of HSA site I has been changed by the alkaline ageing. At the molar ratio AHSA:DG of 1:0.5, the percentage of DG displacement decreased two times (23.60%), while at the molar ratio AHSA:DG of 1:1, it increased two times (Table 4
The serum albumin structure binding site I is large enough to be divided into three regions: region Ia, Ib, and Ic corresponding to warfarin, azapropazone, and n-butyl p-ABE, respectively [36
]. Moreover, site I has been described as an “elongated sock shaped pocket” with hydrophobic foot and hydrophilic leg regions [36
]. In our manuscript, we demonstrated that the alkaline ageing causes the increase in a polarity in subdomain II that results in DG binding to another region (Ia, Ib, Ic) of aged albumin in comparison to non-modified HSA.
At the molar ratio albumin:DP of 1:0.5, MET displaces DP from site II of aged HSA (33.28%) and non-modified HSA (39.16%). At the molar ratio albumin:DP of 1:1, the displacement of DP by MET from site II of AHSA is more effective (47.59%) in comparison to that of HSA (33.29%). This phenomenon indicates that the binding affinity of subdomain IIIA increases during HSA alkaline ageing. Moreover, the molar ratio AHSA:DG and DP of 1:1 should be used because site I involves three regions: Ia, Ib, Ic. Concluding, MET has greater binding affinity to aged than to non-modified HSA. The similar phenomenon was described by Chudzik et al. [37
] for doxazosin binding affinity.
On the contrary to AHSA, at both albumin:probe molar ratios of 1:0.5 and 1:1, MET displaces DP from site II of aged BSA (ABSA) less effectively than from non-modified BSA (Table 4
). The binding of MET to site II involves helices: h1, h2, h3, h4 and h5. Because the amino acid sequence of these HSA and BSA helices are different in at least 18 positions, therefore, the alkaline ageing of HSA and BSA influences the metoprolol binding affinity in site II differently.
Based on the conducted experiment, we observed that the increase of HSA elimination is caused by not only the basic amino acid modification but also the ageing influence on the specific albumin binding sites.