Glycation of human serum albumin alters its binding efficacy towards the dietary polyphenols: a comparative approach

Diabetes is a major problem in the world. The proteins became modified during glycation after reacting with the reducing sugars (e.g. D-glucose) via non-enzymatic pathways. The glycated analogue of human serum albumin (HSA) has been characterized with the help of multi-spectroscopic methods. It has been observed that six glucose molecules can bind covalently to HSA under experimental condition. The binding affinity of the modified HSA towards the dietary polyphenols has been estimated using UV–vis and fluorescence spectroscopic techniques. The binding constant values of the ligands were found to decrease after the modification of HSA. Graphical abstract The binding affinities (Kb) of the polyphenols decreased towards human serum albumin after its structural modification with D-glucose. Highest percentage decrease in the binding is observed for quercetin among all the polyphenols.


Introduction
Serum albumins (SAs) maintain the osmotic pressure and are indispensable for proper delivery of body fluids between intravascular compartments and body tissues (Kragh-Hansen, 1981). They are also accountable for the maintenance of blood pH level (Peters, 1975). HSA contains three structurally equivalent α-helical domains (I-III); each of them consist of two subdomains, A and B (Carter & Ho, 1994;Petitpas, Grüne, Bhattacharya, & Curry, 2001) with only tryptophan residue (Trp 214) placed in subdomain IIA of site 1. SAs are efficient in enhancing the solubility of ligands in blood plasma and are able to improve their delivery to cells effectively. The following factors are considerable for the albumin binding efficacy of ligand molecules: lipophilicity, acidity, H-bonding potential and shape factors. The drug molecules are released at target locations by binding (binding is generally reversible in nature, Ishtikhar et al., 2015;Singh & Mehta, 2006) with SAs in blood plasma. Information concerning the binding affinities of the ligands with SAs is enviable to develop the therapeutic effectiveness of these compounds at the molecular level.
The reactions of SAs with reducing sugars via nonenzymatic pathways cause the structural alteration of protein. Many people are affected by diabetes worldwide and it is actually associated with insulin deficiency, insulin ineffectiveness or both causing hyperglycemia (Unwin, Whiting, Gan, Jacqmain, & Ghyoot, 2009). Such structural adaptation of the proteins by reducing sugars occurs naturally in the body and has been allied to many chronic complications such as renal failure, blindness, nervous breakdown and cardiovascular disease. The major glycation position of HSA is Lysine 525, but other Lysine (Lys) residues such as 199, 281 and 439 are also susceptible to glycation. It affects the structural and functional aspects of the protein due to the confinement of Lys residues (Hudson et al., 2002;Rondeau & Bourdon, 2011;Taghavi et al., 2014;Trynda-Lemiesz & Wiglusz, 2011;Vetter & Indurthi, 2011). Literatures have already established that glycation of HSA alters its binding efficacy towards the ligands (Joseph, Anguizola, Jackson, & Hage, 2010;Joseph, Anguizola, & Hage, 2011;Trynda-Lemiesz & Wiglusz, 2011), e.g. the binding affinity of bilirubin reduced to 50% after glycation of HSA (Bourdon, Loreau, & Blache, 1999). Binding constant of HSA with warfarin and L-tryptophan is changed as the glycation level is augmented, which occurs in diabetes . Hence the basic understanding of binding of HSA with different drugs under diabetic atmosphere has significant clinical importance.
Flavonoids are classified into (contain three rings A, B and C; Figure 1) three main categories e.g. flavone, flavanones, isoflavones. This complex framework enables the compound to bind with proteins which on the other hand has a distinct effect on their structural and functional assets (Matei & Hillebrand, 2010;Matei, Ionescu, & Hillebrand, 2011;Pal & Saha, 2014;Singha Roy, Ghosh, & Dasgupta, 2013;Singha Roy, Tripathy, Chatterjee, & Dasgupta, 2010;Wang & Ho, 2009). They are generally present in various dietary sources such as onion, tomato, apple, grape, tea, red wine etc. (Wang & Ho, 2009). Flavonoids are beneficial due to their extensive uses such as anti-oxidant, anti-inflammatory, anti-diabetic, anti-allergic, anti-bacterial and anti-carcinogenic agents (Khan, Afaq, Syed, & Mukhtar, 2008;Mabry, Markham, & Thomas, 1970;Wang & Ho, 2009;Wang, Noh, & Koo, 2006). The anti-oxidant activities of the polyphenols is primarily due to their free radical scavenging capabilities and the B-ring of the flavonoid category is anticipated to be the major site of anti-oxidant reactions (Khan & Mukhtar, 2007). Hence it is essential to study the binding of these ligands with the carrier proteins which plays an important role in the field of pharmacology and pharmacodynamics. The interaction of flavonoids with glycated HSA is significant as we are looking for the changes in binding affinities of these compounds towards the modified protein. In our recent work we have discussed a comparative synopsis of the binding affinities of these naturally occurring polyphenols (genistein: Bian, Liu, Tian, & Hu, 2004;morin: Xie, Long, Liu, Qin, & Wang, 2006;rutin: Pastukhov, Levchenko, & Sadkov, 2007;quercetin: Singha Roy, Tripathy, Ghosh, & Dasgupta, 2012;fisetin: Singha Roy, Dinda, & Dasgupta, 2012) with the native HSA and the glycated analogue of HSA.
In this proposed manuscript, we have prepared the glycated analogue of HSA from the native state of HSA using the procedure described by Joseph et al. (2011). The detail characterizations of the glycated HSA have been incorporated in the supplementary section of this manuscript. In this report we have investigated the binding of dietary polyphenols with the glycated analogue of HSA using UV-vis, steady state fluorescence and circular dichroism (CD) spectroscopic methods as reported earlier by several research groups for protein-ligand binding studies (Chandrasekaran, (Pace, Vajdos, Fee, Grimsley, & Gray, 1995). The stock solutions of the polyphenols were prepared by dissolving the solids in the HPLC grade ethanol. The experiments were performed keeping less than 5% ethanol in the resultant mixtures.

UV-vis study
UV-vis experiments were performed on UVPC 2450 spectrophotometer at room temperature in the range of 200-500 nm with the help of a 1 cm quartz cuvette in 0.02 M phosphate buffer of pH 7.0. A 1.0 mL of 20 μM polyphenol in the buffer was titrated with consecutive addition of the protein sample (0 to 25 μM).
where L T indicates total ligand concentration, ΔA represents the alteration in absorbance at a fixed wavelength of our interest. ε is known as the extinction coefficient and M is the concentration of the glycated protein used in the binding. On the other hand the subscripts b, f and T designate bound, free and total ligand concentrations under the experimental environment. Hence the plot of 1/ΔA vs 1/[glycated HSA], will give the K a value.

Steady-state fluorescence spectroscopy
Fluorescence studies have been executed on a Horiba Jobin Yvon spectrofluorometer (Fluoromax 4) with the help of a 1 cm quartz cell. The glycated protein sample (4 μM) was titrated with successive addition of the polyphenols (0 to 14 μM). The protein sample was excited at 295 nm (at 26°C) and the fluorescence emission spectra were obtained in the range of 305-500 nm.
During the experiment the excitation and emission slit widths were kept at 2 nm each.

Result and discussion
The interaction of dietary polyphenols (quercetin, rutin, morin, fisetin and genistein) with glycated HSA has been explored using UV-vis and steady state fluorescence techniques. The study has been performed to determine the binding affinities (K b ) of these glycated HSApolyphenol complexes and to compare the values with native HSA-polyphenol complexes as reported earlier. Figure S5 exhibits the UV-vis spectra of different polyphenols in presence of increasing concentrations of the glycated HSA. The UV-vis spectra presented are those obtained by subtracting the glycated albumin spectrum from the corresponding glycated HSA-flavonoid complex spectrum. In each case a red shift was observed after the addition of the glycated protein indicating the involvement of specific binding occurs between the protein and the ligands. The extent of red shift was observed to be the highest in case of quercetin (25 nm) and the same for other ligands are listed in Table S2. The red shifts are originated from the deprotonation of the hydroxyl groups that lead to the extended π conjugation (Zsila, Bikádi, & Simonyi, 2003). The absorbance at the wavelength maxima were found to decrease in presence of glycated protein. It may be due to the altered planarity of the ligands after protein binding that is required for their accommodation in the binding site of the protein (Zsila et al., 2003;Kanakis et al., 2006). The absence of hydroxyl group at the C3 position in genistein and rutin results in a lower extent of red shift compared to the other systems. The absorption maximum of genistein shifted from 262 to 268 nm and that of rutin occurred from 360 to 364 nm. The ground state association constants (K a ) for the individual systems were estimated according to the Equation 1 and are summarized in the Table S2. Quercetin-glycated HSA complex showed the highest stability over the other complexes.
Rutin exhibited the lower affinity towards the glycated analogue which is also reflected in the UV-vis spectra of rutin after the protein addition. Small decrease in absorbance value was found in case of rutin suggesting the lower affinity of the ligand towards the protein. Similar kinds of observations were found in case of the interactions of quercetin, kaempferol and delphinidin with HSA using UV-vis study as reported earlier by Kanakis et al. (2006). The bathochromic shift may be arises due to the presence of polar interactions between the protein and the polyphenols. The red shift in the UV-vis spectra established the formation of ground state complexes between the polyphenols and the protein. Interactions of rutin with the native state of HSA showed the similar type of results as found in the present study (Pastukhov et al., 2007).
The fluorescence emission spectra of glycated HSA in absence and presence of the polyphenols are provided in Figure 2. Glycated HSA exhibits fluorescence (emission at 347 nm) but the emission intensity is less than native HSA ( Figure S1b). The polyphenols were also capable to quench the fluorescence intensity of the glycated HSA under experimental conditions. The quenching indicated the possibility of interactions taking place between the glycated protein and the polyphenols. The blue shifts observed suggest that the fluorophore of glycated HSA (Trp 214) is located in a more nonpolar microenvironment as a result of hydrophobic interactions with the polyphenols (Mandeville et al., 2009). This study also supported the fact that the polyphenols bind near Trp 214 (site 1, subdomain IIA) of the protein.
Fluorescence quenching can occur due to excited state reactions, ground state complexation or collisional quenching between a ligand and a protein (Lakowicz, 2006). To determine the quenching type involved, we have estimated the Stern-Volmer constant (K SV , calculated from equation 2) for the individual systems that are provided in Table 1. The K SV values are found in the order of 10 4 M −1 and the bimolecular quenching constant (K q ) of different protein-ligand complexes are estimated to be in the order of 10 13 M −1 s −1 . The bimolecular quenching constant (K q ) for the binding processes showed that it had crossed the maximum limit (K diffusion = 2 × 10 10 M −1 s −1 ) of the highest probable value for a dynamic collision in aqueous solution (Lakowicz, 2006) and the results ruled out the incidence of dynamical quenching mode in the binding.
The linear nature of the Stern-Volmer plots (Figure 3(a)) in correlation with the ground state complex formation (confirmed from the UV-vis studies) also indicate the presence of static quenching mechanism. Figure S6a and S6b showed that the subtracted UV-vis spectra of the ligand (or protein) from protein-ligand complex were different in nature suggesting the formation of ground state complexes. The ground state complex formation between protein and ligand is likely to affect the absorption spectra of the protein or ligand and that will confirm the absence of dynamical quenching. The binding parameters for the interactions of the polyphenols were calculated according to equation (3) and the related plots are presented in Figure 3(b). The binding parameters are summarized in Table 2 and the binding constants are found in the order of 10 4 to 10 5 M −1 . These results indicate the presence of moderate and reversible kind of binding between glycated HSA and polyphenols. Similar kind of results is also obtained in case of binding of fisetin and morin with hen egg lysozyme (Utreja, Badhei, & Singha Roy, 2015). The binding of polyphenols are with the native HSA are all non-covalent in nature and the binding affinities are listed in Table 2 (Bian et al., 2004;Pastukhov et al., Xie et al., 2006). It was found that quercetin has the highest affinity (11.89 ± .11 × 10 4 M −1 ) towards the glycated protein. Rutin showed the lowest binding affinity (2.78 ± .71 × 10 4 M −1 ) towards the modified protein. One interesting observation is that K b values for individual systems were found to be lower than their individual K b values with native HSA ( Table 2). The number of binding sites (n) for the polyphenols remained very close to unity in of both cases (Table 2).
where ΔF = F 0 −F; F 0 and F are the fluorescence emission intensities of glycated HSA in the absence and presence of the polyphenols respectively. The number of binding sites and binding constant are designated by n and K b respectively. The tryptic digestion analysis established that either Lys 199 or Lys 205 was modified during glycation time.
Moreover, the hydrophobic pocket (near Trp 214) is unfolded to some extent during the same time period of glycation (Coussons et al., 1997;Chamani et al., 2006). A steric hindrance near site 1 (subdomain IIA) of the glycated sample is created and the partially unfolded hydrophobic pocket prevent the binding of the polyphenols in this site near to Trp 214. The glycosylation of the flavonoid at the C3 position lowers the planarity of rutin and steric hindrance takes place that also reduced the binding affinity of this compound towards the proteins (Xiao, Cao, Wang, Zhao, & Wei, 2009;Xiao et al., 2011). The binding affinities of quercetin, morin and fisetin were reduced by 64, 35 and 56% respectively after binding with glycated HSA ( Table 2). The K b values obtained from the UV-vis studies are in good correlation with the K b values from fluorescence study and a correlation coefficient of .74 was obtained (Figure 4(a)). The stability constant for fisetin with bovine milk proteins was found to be lower than quercetin due to the absence of a hydroxyl group at the C5 position of ring A (Xiao et al., 2011). The lack of a 3′-OH group in   Bian et al. (2004) kaempferol reduced the binding affinity towards the bovine milk proteins which got significantly enhanced when the 3′-OH group was present (quercetin) (Xiao et al., 2011). Quercetin contains 3′-OH, 3-OH and 5-OH groups in the chemical structure and it does not contain any glycoside moiety at C3 or C7 position. The molecule also contains a double bond between C2 and C3 atoms. Hence due to its perfect structural planarity (Xiao et al., 2011), quercetin binds strongly with the native HSA and its glycated analogue in compare to other polyphenols. The isoflavone genistein is structurally different from these molecules and that is why it has not been considered in the above comparison. The binding constant values of genistein and rutin with HSA were taken from the works of Bian et al. (2004) and Pastukhov et al. (2007) respectively (Bian et al., 2004;Pastukhov et al., 2007). Binding affinities in the absence and presence of D-glucose (that is in case of glycated HSA) are plotted (Figure 4(b)) simultaneously and a correlation coefficient of .76 was obtained.

Conclusion
The characterization of the glycated HSA has been carried out by UV-vis, fluorescence, MALDI-TOF and CD spectroscopic techniques. The MALDI-TOF analysis established that six glucose molecules were attached to HSA. The probable glycation sites were identified from the tryptic digestion analyses and the results were compared with the literatures available. The binding of the polyphenols with the glycated HSA were investigated using UV-vis and steady-state fluorescence techniques. The binding constants are found to be reduced in compare to the native HSA as determined from steady state fluorescence spectroscopy. Rutin exhibits the lowest binding affinity towards the glycated HSA due to its bulkier size and less planar geometry. The decrease in the binding affinities of the flavonoids occurred either due to glycation of Lys 199 or Lys 205 which results in a steric barrier to the ligands or due to partial unfolding of the hydrophobic pocket during glycation.

Supplementary material
The supplementary material for this paper is available online at http://dx.doi.org/10.1080/07391102.2015. 1094749.