Modulated Protein Binding Ability of Anti-Diabetic Drugs in Presence of Monodispersed Gold Nanoparticles and its Inhibitory Potential towards Advanced Glycated End (AGE) Product Formation

Journal of Fluorescence https://doi.org/10.1007/s10895-019-02485-y ORIGINAL ARTICLE

Modulated Protein Binding Ability of Anti-Diabetic Drugs in Presence of Monodispersed Gold Nanoparticles and its Inhibitory Potential towards Advanced Glycated End (AGE) Product Formation Imocha Rajkumar Singh 1 & Sivaprasad Mitra 1 Received: 19 October 2019 / Accepted: 26 December 2019 # Springer Science+Business Media, LLC, part of Springer Nature 2020

Abstract Binding strength of the anti-diabetic drugs chlorpropamide (CPM) and tolbutamide (TBM) with model protein bovine serum albumin (BSA) shows strong modulation in presence of colloidal gold nanoparticles (AuNP). Intrinsic tryptophan fluorescence of both the native BSA and BSA-AuNP conjugate quenched in presence of the drugs. Stern-Volmer quenching constant (KSV) of CPM binding to BSA-AuNP conjugate at different temperatures is almost twice (6.76~14.76 × 103 M−1) than the corresponding values in native BSA (3.21~5.72 × 103 M−1). However, the calculated KSV values with TBM show certain degree of reduction in presence of AuNP (6.46× 103 M−1), while comparing with native BSA (8.83 × 103 M−1). The binding mode of CPM towards BSAAuNP conjugate is mainly through hydrophobic forces; whereas, TBM binding is identified to be Van der Waal’s and hydrogen bonding type of interaction. Fluorescence lifetime analysis confirms static type of quenching for the intrinsic tryptophan fluorescence of BSA as well as BSA-AuNP conjugate with addition of CPM and TBM at different concentrations. The α-helical content in the secondary structure of BSA is decreased to 48.32% and 45. 28% in presence of AuNP, when the concentration of CPM is 0.08 mM and 0.16 mM in comparison with that of native protein (50.13%). On the other hand, the intensity of sugar induced advanced glycated end (AGE) product fluorescence is decreased by 55% and 80% at 0.13 nM and 0.68 nM AuNP, respectively. Change in the binding strength of the drugs with transport protein and reduced AGE product formation in presence of AuNP could lead to a major development in the field of nanomedicine and associated drug delivery techniques.

Keywords Serum albumin . Drug binding . AGE product . Nanomedicine . Fluorescence

Introduction Highlights • The formation of the BSA-AuNP conjugate is confirmed by the IR spectroscopy. • Binding efficiency of CPM and TBM to native BSA are strongly modulated in presence AuNP. • CPM binds to BSA-AuNP conjugate mainly through hydrophobic forces. • Binding of TBM to BSA-AuNP conjugate through van der Waals and HB interactions. • AuNP induces strong inhibition towards the glycation of serum proteins. Electronic supplementary material The online version of this article (https://doi.org/10.1007/s10895-019-02485-y) contains supplementary material, which is available to authorized users. * Sivaprasad Mitra smitra@nehu.ac.in; smitranehu@gmail.com 1

Centre for Advanced Studies, Department of Chemistry, North-Eastern Hill University, Shillong 793 022, India

Serum proteins present in blood serve diversified functions which include transport of drugs, lipids, hormones, vitamins, and minerals in the circulatory system and the regulation of a cellular activity and functioning of the immune system. The concentration of the albumin protein in the body fluid is around 35–50 mg/ml and has a half-life of ca. 21 days [1]. However, they are highly susceptible to glycation (non-enzymatic glycosylation) in presence of excess reducing sugars of the diabetic patients due to the presence of free amino functional groups. Normally, monitoring the amount of glycation is used as glycemic control index for a short-term measure because the amount of glycated albumin cannot be altered easily by any other biological metabolism, which is an essential tool to counter type II diabetes mellitus [2–4]. Also, the reducing sugars can interact with other amino acids, proteins and nucleic acids. The primary amine groups of lysine and arginine present in serum albumin interact with the carbonyl moiety of the reactive

reducing sugar molecules present in a diabetic species, thus forming a Schiff base, that undergoes subsequent reactions such as oxidation, dehydration and re-arrangement etc. leading to the formation of a heterogeneous groups of compounds like carboxymethyllysine (CML), carboxyethlylysine (CEL) and pyrraline; often collectively known as advanced glycated end (AGE) products [5, 6]. Prolonged accumulation of the AGE products causes many diabetic related complications, cataract formation, Alzheimer’s and Parkinson’s disease etc. [7–9]. With due advancement in the therapeutic science, many drugs are synthesized which can inhibit the glycation of proteins. Among these aminoguanidine, metformin etc. are strong inhibitors of glycation but has certain side effects [10, 11]. Bovine serum albumin (BSA) is a carrier protein which has a close homology with human serum albumin (HSA). Both these proteins are prone to undergo glycation. Recently a series of nanoparticles (NPs) have shown the ability to inhibit the glycation of albumins by D-ribose and glucose [12, 13]. Further, extremely small size and significant bio-compatibility of the NPs allow them to easily pass through the cell membrane, leading to the growth of the field called nanomedicine [14]. Also, biological macromolecular system like cells, proteins and membranes etc. can be easily adsorbed on the NP surface due to their high surface to volume ratio [15]. Similarly, stable colloidal gold nanoparticle (AuNP) solutions are used in the bioimaging, biosensors, drug delivery techniques etc. Various other conjugates of Au-DNA, Auantibiotics are also used in the detection and minimization of the toxicity of the drug [14, 16, 17]. As an accepted technique in nanomedicine, noble metal NPs are also used in the development of the prodrug by forming various complex of proteingold and protein-silver nanoconjugates. Since gold nanoparticles have characteristic localized surface plasmon resonance (SPR), it is often used in the targeted drug delivery technique because of their high selectivity towards the cancer cells, which is an emerging chemotherapy technique. There are also reports of AuNP’s forming conjugate with chitosan and other penetrating peptides to easily pass through the cell membrane and target the specific sites [18–20]. Functionalization of nanoparticle surface is an important tool in diagnostic and therapeutic applications in treating cancer [21]. Proteins play an essential role in natural systems as they are the mediators of all the biological activities. Any changes in the protein structure or conformation tend to have a loss in its activity. There are reports where thermo-sensitive polymers are coated with bioactive enzyme like trypsin to render the overall increase in performance of the enzyme [22–25]. However, the modification of the proteins by forming conjugates with other polymer molecules results a steric hindrance on the protein active site, which is responsible for the formation bio-inactive molecules [26]. The stability and performance of the proteins and enzymes also increases when they form conjugate with the nanoparticles [27]. Many complexes

of proteins and NPs are used in the field of therapeutic and analytical purposes. Adsorption of proteins forms a monolayer structure on the NP surface, commonly known as ‘protein corona’ [28]. The biological behavior of the NPs is entirely determined by the protein corona [29]. Due to the wide range of optical and electrical properties, varied sizes of metal nanoparticles (MNPs) are used to study the dynamics of protein unfolding and binding interactions. Adsorption on the AuNP surface can also change the α-helical content of the human serum albumin (HSA) [29, 30]. Also, citrate capped AuNPs can inhibit the fibrillogenesis of proteins in vitro [31, 32]. Chlorpropamide (CPM) and tolbutamide (TBM) are the sulfonylureas class II anti-diabetic drugs. CPM has an average half-life of 25–60 h and the glycemic effect is 24–72 h [33]. Both these drugs increase the production of the insulin by interacting with the pancreatic β-cell in type II diabetic mellitus [34, 35]. Sulfonylureas drugs are used for glycemic control; where, it also can modulate the hypoglycemic effect. In the present study, a model glycated system is constructed with BSA using L-arabinose, a five-carbon reducing sugar. It is well-known that arabinose is also formed by the autooxidation of glucose molecules during the glycation of proteins [36, 37]. It resembles D-ribose with the only difference at the anomeric carbon, due to the positioning of the hydroxyl group. The inhibition of glycation is studied using AuNPs by monitoring the AGE’s fluorescence intensity. Also, BSA adsorbed on the AuNP surface are used to study the binding interaction of the anti-diabetic drugs like CPM and TBM.

Experimental Details Reagents and chemicals. Bovine Serum Albumins (BSA), Tolbutamide (TBM) Chlorpropamide (CPM) and Trizma-HCl were bought from Sigma-Aldrich. Sodium hydroxide pellets were obtained from Merck. L-arabinose (ara) of highest grade was obtained from SRL, India. Trizma-HCl buffer solution (pH = 7.4) is prepared using highest quality water of analytical grade (type II) collected from Elix 10 water purification system (Millipore India Pvt. Ltd.). The pH of the solution was checked with a Systronics μ-pH system 361. The concentration of BSA was maintained at 5 μM throughout the experiment and the concentration of anti-diabetic drugs was varied from 0 to 0.7 mM.

Preparation of AuNPs and BSA-NP Conjugate Gold nanoparticles (AuNPs) were synthesized from the aqueous solution of HAuCl4 as precursor and tri-sodium citrate as reducing agent, where the later also acts as a capping agent [38]. Briefly, a solution of HAuCl4 is taken in a flask with a definite amount of water and heated to boil on a hot plate with

J Fluoresc continuous stirring. Then 1.6 mL of 1% (w/v) of tri-sodium citrate was added at the instant of boiling. After 20s of boiling, the solution turned faint blue which later changes readily to red indicating the formation of the AuNPs. The solution was boiled for another 30 min to ensure the complete reduction of the Au(III). The resulting red colored solution was cooled at room temperature. The diameter of the AuNPs was estimated to be around 14 ± 2 nm from TEM measurements (Fig. 1). The concentration of the stock colloidal AuNP solution was calculated using method developed by Haiss et al. [39] using Eq. (1), by substituting the value of absorbance at A450 (0.559) and Ɛ450 (1.76 × 108 M−1 cm−1) and found to be ca. 3.4 nM. A450 C¼ Ɛ450

A stock solution of 25 μM BSA was incubated with a fixed amount AuNP solution, making a final concentration of 5 μM BSA and 1.04 nM AuNP solution, to prepare BSA-AuNP conjugate. The incorporation of the BSA molecules on AuNP surface were also verified with IR spectroscopy and the aggregation of the AuNP in the protein matrix was checked with TEM measurements (Fig. 1).

Glycation of BSA (In-Vitro) A stock solution of 25 μM BSA was used to create a glycated system by reacting with 0.4 M of L-arabinose (ara). The concentration of the working protein solution was 5 μM. To check the inhibition of glycation by AuNP, a fixed amount of 0.13 nM and 0.68 nM solution of AuNP was also added

during the process of making a glycated system. The glycated BSA (gBSAara) and gBSAara containing AuNP were incubated for 80 h at 37 °C under the same condition. Meanwhile, the inhibition of the glycation by AuNP is checked by observing the characteristic AGE’s fluorescence intensity after taking an aliquot for each 10 h interval following the protocol described elsewhere [40]. The concentration of protein, sugar and AuNP were kept fixed during the glycation process.

Apparatus and Methods The absorption and fluorescence measurements were taken in a UV-visible PerkinElmer model Lambda25 and Quanta master (QM-40) from Photon Technology International (PTI), respectively. CPM, TBM and the sugar molecules do not show any significant absorption at the excitation wavelength (295 nm). Nevertheless, the emission spectra of all the solutions were subtracted from the blank mixture (without protein) in the same condition as a calibration measure. Consequently, the corresponding fluorescence emission is contributed only from the tryptophan (Trp) residue of the protein. Monitoring of the AGE’s fluorescence intensity was done at excitation wavelengths 335 nm and 370 nm. Quenching of the protein fluorescence data in presence of CPM and TBM (quencher, Q) of BSA-AuNP conjugate was analyzed using the Stern-Volmer relation (Eq. 2), where, F0 and F are the fluorescence generated by the protein in absence and presence of any quencher. τ0 represents the intrinsic Trp fluorescence lifetime and κq is the bimolecular quenching constant [41]. The analysis of fluorescence quenching data of the BSA-AuNP conjugate in presence of CPM and TBM shown negative deviation in all the systems. Therefore Lehrer equation (Eq. 3), a modified form of Eq. 2 is used to calculate the Stern-Volmer constant (KSV) values, where (fa) is the accessible fraction of the quencher Q [42]. F0 ¼ 1 þ KSV ½Q ¼ 1 þ k q τ 0 ½Q F F0 1 1 1 1 :þ ¼   f a KSV ½Q fa F0 − F

Fig. 1 TEM image of the synthesize AuNP’s in the size range 13 ± 2 nm (a) and (c). Agglomerated AuNP’s of BSA-AuNP conjugate (b) and (d) ð2Þ ð3Þ

Fluorescence decay time of the protein samples were measured by time-correlated single photon counting (TCSPC) technique in a pico-master (PM-3) instrument obtained from PTI with 295 nm LED excitation. Fluorescence decay spectra of the BSA-AuNP conjugate and BSA-AuNP conjugate with different concentrations of drug were collected at magic angles (54.7°). Instrument response function (IRF) were also collected in the same procedure by using the scatter solution of dried dairy coffee whitener. The traces of the fluorescence decay F(t) of all the protein samples were expressed as a sum of exponential (Eq. 4), where αi and τi are the amplitude and lifetime of the ith component, and analyzed by using non- J Fluoresc

linear least square iterative convolution method based on Lavenberg–Marquardt algorithm using FelixGX software (version 4.0.3).   −t FðtÞ ¼ ∑i αi exp ð4Þ τi τ av ¼ ∑i f i τ i ¼ ∑i αi τ 2i ∑ i αi τ i

ð5Þ

The statistical parameters like chi-squared (χ2) and the Durbin-Watson (DW) values were used for the validation of the fitting analysis, others observable like residual fitting and autocorrelation function were checked through visual inspection [43]. For multi exponential decay fitting, the average fluorescence lifetime was evaluated using (Eq. 5) [44].

moles and ‘n’ is the independent class of binding sites of the protein [46, 47]. The Ka values were calculated for different temperatures, and the thermodynamic parameters such as enthalpy (ΔH) and entropy (ΔS) change were evaluated using Van’t Hoff relation Eq. (9) for the interaction between CPM, TBM with the proteins [42, 48]. The total change in the free energy (ΔG) in each process was calculated by using Eq. (10). The nature of interaction between the drug and protein-NP conjugate was evaluated by monitoring the symbol of thermodynamic parameters. For example, the drug binds to the protein through hydrophobic mechanism when the values of ΔH and ΔS are positive; whereas, if the values of ΔH and ΔS are negative then the drug binds to the protein through hydrogen bonding and Van der Waals interaction [49]. r¼

Evaluating the Protein Structure Any noticeable change in the secondary structure of the protein were studied by measuring IR and Circular dichroism (CD) spectra. IR spectra are taken in a BRUKER model Eco-ATR (attenuated total reflectance) with a resolution of 2 cm−1 with 30 times scanning. About 20 μL of the solution sample is placed in the surface of the ZnSe crystal and the spectra are collected. All the spectra are taken in the spectral range from 1400 cm−1 to 1800 cm−1. Amide I region of the protein around 1600 cm−1 to 1700 cm−1 give the information related to the secondary structure of the protein; whereas, amide II peaks around 1500 cm−1 to 1600 cm−1 indicated adsorption of the protein due to surface energy [45]. The circular dichroism (CD) readings within the spectral range of 200– 250 nm were recorded on a JASCO J-1500 spectrometer using a quartz cuvette of 0.2 cm path length. The mean residue ellipticity (MRE) and the α-helix content of the protein were calculated using eqs. (6) and (7) by observing the CD values at 208 nm and 222 nm. For measuring IR spectra, the protein solutions were kept at 30 mg/mL; whereas, in CD experiment the concentrations of the protein was kept constant at 1 μM while varying the CPM concentration from 0.08 mM to 0.016 mM. MRE ¼

Binding of Drugs to BSA-AuNP Conjugate The binding of CPM and TBM to BSA-AuNP conjugate were evaluated using association constants (Ka) as the binding parameter calculated from the Klotz equation (Eq. 8). The parameter ‘r’ is the bound drug molecules per protein in terms of

n  Ka  ½drug 1 þ Ka  ½drug ΔH ΔS þ RT R ΔG ¼ ΔH−T:ΔS lnKa ¼ − ð8Þ ð9Þ ð10Þ

BSA Activity BSA undergoes non-enzymatic protein directed hydrolysis when it interacts with para-nitrophenyl acetate (PNA). The amino acid residues such as Try-411, Lys-412 and Lys- 413 present inside the subdomain III A show these catalytic properties. Amino acid residue Try-411 is easily acetylated by PNA [50–52]. The activity of BSA was checked by monitoring the release of p-nitrophenyl through single wavelength absorbance at 410 nm using UV-6300PC MAPADA spectrophotometer. A 5 μM BSA solution is incubated with 1 mM PNA at 370 C for 30 min. The release of para-nitrophenol was checked in presence and absence of colloidal AuNP solution.

Calculation of AGE’s Fluorescence Intensity Many of the AGE’s products shows fluorescence properties, argpyrimidine and pentosidine shows emission at 400 nm and 385 nm when excited at 335 nm. Similarly crossline and vesperlysine along with some other unidentified AGE compounds, when excited at 370 nm, show the fluorescence emission at 430–450 nm [53–55]. Generally, excitation at 335 nm is used to check the protein modification induced by the monosaccharide’s pentosidine type compound; whereas, excitation at 370 nm is usually used to monitor the alteration in the protein due to different AGE types of compounds [56–62]. The inhibition of the glycation was checked after monitoring the AGE’s fluorescence intensity. The relative decrease in the production of AGE’s fluorescence intensity of the glycated BSA in presence of different concentrations of AuNP’s was

studied within a time span of 80 h. The AGE’s fluorescence intensity was checked at each 10 h interval. Although, AuNP has no significant absorption at the excitation wavelengths (335 nm and 370 nm), AGE’s fluorescence spectra were corrected using Eq. (11).

F corr λE; λ F ¼ F λE; λ F  efAðλE ÞþAðλ F Þg=2 ð11Þ where, A represent the absorption of the sample at the excitation (λE) and emission (λF) wavelength, respectively. The final steady state data was taken as the average of three independent experiments and analyzed further.

Results and Discussion Interaction of CPM and TBM to BSA-AuNP Conjugate The fluorescence emission of BSA-AuNP conjugate are monitored at different concentrations of the anti-diabetic drugs (CPM and TBM). Fluorescence intensity of the BSA-AuNP conjugate is quenched regularly with an increase in the concentration of the drugs from 0.08 mM to 0.64 mM (Fig. 2) without any shift of the emission maxima. The fluorescence quenching is analyzed by using the Stern-Volmer relation (Eq. 2). A negative deviation from the linear straight-line fitting is observed from the onset of 0.16 mM concentration of CPM as shown in the inset of (Fig. 2). In one of our recent publications, the deviations from straight line plot was also observed for CPM binding to native BSA at the similar concentration range [40]. Normally, negative deviation in SV plot signifies that the drug is interacting with two distinct type of fluorophores (tryptophan residues) present in the protein. The tryptophan residues residing outside the protein surface is easily accessible by the drugs in comparison with the tryptophan residue, which is buried inside the hydrophobic domain [63]. The onset of negative deviation in similar concentration range both for the native protein and BSA-AuNP conjugate indicates almost equal accessibility of the tryptophan residues by the drugs in both the Fig. 2 Quenching of BSA-AuNP cojugate fluorescence by CPM at 298 K (a) and TBM (b) with increasing concentration of drugs from 0 to 0.68 mM. Inset is the onset of negative deviation from SV plot for CPM at 0.16 mM (a) and TBM at 0.24 mM (b) respectively. BSA concentration at remained constant at 5 μM

cases. On the other hand, the binding of TBM to BSA and BSA-AuNP conjugate results the onset of negative deviation at [TBM] ~ 0.24 mM. Similar types of tolbutamide binding to native BSA is reported by Szkudlarek et al. [64]. Fluorescence quenching spectra with CPM at higher temperatures and deviation from the linearity in the SV plot are shown in Fig. S1. The quenching constant KSV for the interaction between the drug and BSA-AuNP conjugate are calculated from the modified SV plot (Eq. 3) and given in Table 1. Some representative plots are presented in Fig. 3 and in the supplementary section (Fig. S2). Generally, for the interaction with BSA-AuNP conjugate, KSV values are found to be higher in the case of quenching by CPM as compared to TBM (for example, 6.76 × 103 M−1 and 6.46 × 103 M−1 for CPM and TBM, respectively at 298 K) and consistent with the requirement of lower CPM concentration (0.16 mM) in comparison with TBM (0.24 mM) to induce the deviation from linearity in SV plot discussed above. Interestingly, KSV values for the binding of CPM to BSAAuNP conjugate is higher than the corresponding data with native BSA [40]. However, the interaction of TBM with native BSA as well as BSA-AuNP conjugate shows the opposite trend and needs further discussion (see below). The magnitude of the accessible fraction ‘fa’ are also calculated using (Eq. 2) and given in Table 1. It may be noted here that the adsorption of the BSA on the surface of the AuNP reduces the number of accessible tryptophan residue present on the surface of the protein. Therefore, the accessible fraction ‘fa’ decreases during the interaction of CPM and/or TBM with BSA-AuNP conjugate, as compared to their interaction towards native BSA. The mechanism of fluorescence quenching can be classified into two types, i.e. dynamic and static. Dynamic quenching occurs when the quencher interacts with the fluorophore in the excited state; whereas, if the quencher interacts with the fluorophore in the ground state then it is known as static quenching. The fluorophore becomes less fluorescent in both static and dynamic quenching. The mode of the quenching can be identified by conducting the steady state fluorescence experiment at different temperatures. In case of dynamic quenching the magnitude of KSV increases

J Fluoresc Table 1 Stern-Volmer quenching constant (KSV) and fraction of accessible fluorophores (fa) for the interaction of BSA-AuNP conjugate with CPM and TBM at different temperatures Temp /K Modified Stern-Volmer analysis CPM

298 303 308 313 318 a TBM KSV /103 M−1 fa BSAa AuNP-BSA BSAa 3.21 5.64 4.88 5.20 5.72 6.76 7.78 9.95 7.50 14.26 0.50 0.44 0.48 0.48 0.48 fa AuNP-BSA BSA AuNP-BSA BSA AuNP-BSA 0.32 0.37 0.33 0.42 0.40 8.83 6.80 3.11 3.21 2.80

6.46 6.87 5.77 3.55 3.23 0.31 0.33 0.38 0.43 0.40 0.26 0.34 0.35 0.44 0.37 Data from ref. [40] with increase in temperature due to the higher amount of diffusion of the quencher molecule towards the fluorophore; whereas, in case of static quenching the magnitude of KSV is lowered at higher temperature due to the dissociation of the ground state complex. However, establishing the quenching mechanism based on fluorescence lifetime measurement is more accurate and reliable [44]. In our previous publication it was shown that for the interaction of BSA with CPM, static quenching mechanism operates as the lifetime remains constant (~3.63 ns) with varying concentration of the quencher [40]. Typically, the fluorescence lifetime (τ0) decreases with increase in the concentration of the drugs in the case of dynamic quenching. But in case of static quenching, the quencher (Q) molecule interacts with the fluorophore only in the Fig. 3 KSV values from modified SV plot at different temperature at 303 K (a) and 318 K (c) for CPM binding to BSA-AuNP conjuagte and at 303 K (b) and 318 K (d) for TBM binding to BSA-AuNP conjugate. R2 is the regression value

ground state and, therefore, the fluorescence lifetime remains constant [65]. The fluorescence decay traces of BSA-AuNP conjugate is analyzed by using two exponential fitting model to get acceptable statistical parameters with average lifetime ~5.3 ns (Fig. S3). Fluorescence decay traces of CPM (0.08 mM) interaction with BSA-AuNP conjugate is shown in (Fig. 4). The corresponding parameter for BSA-AuNP conjugate at varying concentrations of both CPM and TBM are given in Table S1. On the other hand, the lifetime values for the binding of TBM to native BSA are shown in Table S2. Interestingly, for both CPM and TBM binding to BSA-AuNP conjugate, there is no change in the lifetime of the protein with continuous addition of the drugs (Fig.S3) suggesting that quenching occurs via static mechanism.

Fig. 4 Fluorescence decay traces spectra of 0.08 mM CPM binding to BSA-AuNP conjugate showing a lifetime of 5.30 ns

To understand more on the extent of binding mechanism, association constants (Ka) are calculated using Klotz relation (Eq. 8). The calculated values are shown in Table 2 and some characteristic Klotz plots at different temperatures are shown in Fig. 5. The magnitude of the Ka is found to be higher in case of TBM binding to native BSA than CPM, which is consistent with the values of KSV given in Table 1. However, the binding of CPM to BSA-AuNP conjugate is greater than TBM binding to BSA-AuNP conjugate. This may be due to the accessibility of the fluorophore by the drug molecules towards it. The accessible fraction ‘fa’ of CPM in BSA-AuNP conjugate is more than for TBM in BSA-AuNP conjugate (Table 1). Ka values are also evaluated at different temperatures (Fig. S4) to calculate different thermodynamic parameters using van’t Hoff plot (Fig. 6) and listed in Table 3. The values of ΔH and ΔS are positive in case of CPM binding to BSA-AuNP conjugate signifying that CPM binds through the hydrophobic mode of interaction; which is indeed similar to the binding mode of CPM towards native BSA [40]. On the contrary, in case of TBM binding to BSA-AuNP conjugate the values of ΔH and Table 2 Association constant (Ka) and the number of binding sites (n) determined from Klotz plot for BSA-AuNP conjugate in binding of CPM and TBM

ΔS are found to be negative suggesting that TBM interacts through Van der Waals and hydrogen bonding interaction. Molecular docking reported earlier [40] shows that CPM does not bind to any specific drug binding site of native BSA [40]. The binding affinity of CPM is −27.61 kJmol−1 and marginally different to that calculated for TBM (−26.35 kJmol−1) towards native BSA (data not shown). Therefore, both CPM and TBM seem to bind in the same position of BSA (in between sites I and site II). However, these two drugs show significant differences both in affinity as well as their mode of interaction in case of binding with BSA-AuNP conjugate. It is to be noted here that the time-resolved fluorescence data unambiguously confirms the quenching mechanism to be static type in both the cases of CPM and TBM. However, contrary to the observation in case of TBM, the magnitude of KSV increases with increase in temperature for the quenching with CPM for both the native BSA and BSA-AuNP conjugate (Table 1). This apparent paradoxical observation for static quenching of BSA as well as BSA-AuNP conjugate fluorescence by CPM in varying temperature can be explained by closely looking at the thermodynamic parameters of the protein-drug complex formation process given in Table 3. Significant positive entropy change (ΔS) nullifies the unfavorable positive enthalpy change (ΔH) in case of the protein interaction with CPM and the association is clearly an “entropy-driven” process. Typically for the interaction of strong lipophilic drugs (with higher LogP values) with biological macromolecules, the entropy-driven pathway is primarily contributed by the entropy change due to solvent release upon binding and variation in KSV with temperature is often inadequate to judge the mechanism of fluorescence quenching (either static or dynamic) [66]. Essentially, timeresolved fluorescence measurement gives the final verdict about the quenching mechanism as indeed mentioned before and noted in the literature [44]. Interestingly, the interaction of TBM is markedly different than CPM having both negative ΔH and ΔS (Table 3). The magnitude of KSV also decreases with increase in temperature and follows the expected trend for static type interaction. The results confirm a more significant entropy-driven pathway for CPM interaction with BSA (or

J Fluoresc Fig. 5 Klotz plot for the interaction of TBM (a), (c) and CPM (b) and (d) towards BSAAuNP conjuagte at different temperature with increasing concentration of the drugs from 0 to 0.64 mM

BSA-AuNP conjugate) in comparison with TBM, which is consistent with higher lipophilicity of the former (LogP = 2.15 and 2.04, respectively).

Change in the Protein Conformations The change in the conformation of the protein structure due to glycation or adsorption on the NP surface are observed through monitoring the position of amide I peak in the IR spectra. In native BSA, the amide I peak is observed at ~1637 cm−1. There is a shift in this peak position by ~ 8 cm−1 in case of gBSAara and

appears at ~1645 cm−1 along with a small shoulder at ~1672 cm−1 (Fig. 7). The results indicated that there is a slight change in the protein structure of the gBSAara and consistent with other studies reported recently [67]. In case of the BSA-AuNP conjugate there is an increase in the amide II peak intensity around 1513 cm−1 which is due to the adsorption of the BSA molecules on the NP surface to form BSA-AuNP conjugate [45, 68]. The α-helix content of the BSA and BSA-AuNP conjugate with CPM and TBM are calculated using Eq. (6) and Eq. (7) by observing the value at 208 nm in CD spectra (Fig.S5). As reported in our earlier publication [40], the α-helical content of native BSA is found to be 50. 58% and glycation of BSA by arabinose decreases this quantity to 43.65%. Interaction of CPM to native BSA decrease the α-helical content by 11.30% and 16.11% when the concentration of the CPM is 0.08 mM and 0.16 mM, respectively. Similarly, a decrease in the α-helix content is observed upto 48% and 49%, respectively when 0.016 mM of CPM and TBM are added to the BSA-AuNP conjugate. There is no shift in the spectral position at 208 nm and 222 nm on addition of the drugs (CPM and TBM) to BSA-AuNP conjugate as observed in the CD spectra. The calculated values of MRE and the α-helical content (in percentage) are shown in Table S3.

Effect of AuNP in BSA Activity Fig. 6 Van′t Hoff plot of TPM (upper panel) and CPM (lower panel) binding to BSA-AuNP conjugate

The presence of different types of catalytic amino acid residues (Try-411, Lys-412, Lys- 413 and Try-411) in the protein J Fluoresc Table 3

Thermodynamic parameters for the interaction of CPM and TBM with native BSA and BSA-AuNP conjugate at different temperatures ΔG /kJ mol−1 at different temperature (K) CPMa TBMa TBMb CPMc a 298 303 308

313 318 323 −4.59 −4.98 −5.41 −3.67 −5.17 −4.51 −4.70 −3.86 −5.75 −4.03 −3.98 −4.04 −6.33 −3.56 −3.26 −4.23 −6.90 −3.08 −2.55 −4.42 −7.48 −2.60 −1.83 −4.60 ΔH/kJ mol−1 ΔS/J K−1 mol−1 29.84 −33.33 −48.05 7.45

115.56 −95.11 −143.08 37.35 Interaction with BSA-AuNP conjugate; b,c Interaction with native BSA. c Data from ref. [40]

structure of BSA enables the protein to undergo esterase like activity [52]. The esterase like activity of BSA is disturbed when there is any change in the conformation of the protein [69]. Adsorption of proteins on the NP surface changes the esterase like activity of the proteins. Recent results suggest that adsorption of BSA on the surface of the AuNP retain ~88% the its activity; whereas, the retention capacity is merely 5%c in the case of the gold nanorods [45]. Therefore, in addition to the modulation in binding ability, the retention of BSA activity in protein-Np conjugate is also a matter of great concern. The activity of BSA is checked in presence of pnitrophenyl acetate by monitoring the amount of release pnitrophenol at 410 nm [50]. After subtracting the inherent absorbance of the AuNP at this wavelength, the esterase like activity of BSA-AuNP conjugate is calculated with reference

Fig. 7 IR spectra of gBSAara (a) BSA-AuNP conjuagte (b) and native BSA (c) to the activity of free BSA. The esterase like activity of BSA is decreased by ~6% only when 0.13 nM of AuNP is used to form BSA-AuNP conjugate; whereas, ~ 11% decrease in activity is noted when the concentration of the AuNP is increased upto 0.68 nM (Fig. S6).

Inhibition of Glycation by AuNP The inhibition of glycation of BSA by L-arabinose is checked after monitoring the AGE’s fluorescence intensity with excitation at 335 nm and 370 nm. Glycation of the BSA with 0.4 M of L-arabinose solution is incubated with two different concentrations of AuNP for a total length of 80 h. The generated AGE’s intensity during this time is checked at each 10 h interval. The concentration of the L-arabinose sugar used in the model glycated system is more than the physiological system to create a glycated protein in a short period of time [70, 71]. It is interesting to note that there is a sharp decrease in the AGE’s intensity of the gBSAara incubated with AuNP solution when excited at 335 nm (Fig. 8). The decrease in the AGE’s fluorescence intensity suggested that there is an inhibition in the formation of the AGE’s related compounds. Further, the intensity is decreased with increasing the concentration of the AuNP solution in each interval of time. For example, a concentration of 0.13 nM AuNP solution decreases the AGE’s fluorescence intensity of gBSAara by 55%; whereas it decreases almost 80% when the concentration of AuNP is increased up to 0.68 nM (Fig. 8). There is a reduction about 16% and 40% in the AGE’s fluorescence intensity of the gBSAara, when the concentration of the AuNP are 0.13 nM and 0.68 nM, respectively (Fig. 8). Similarly, when gBSAara incubated with AuNP is excited at 370 nm, the AGE’s intensity of gBSAara decreases by 49% and 84%, at [AuNP] = 0.13 nM and 0.68 nM, respectively. A time dependent effect of AuNP’s in the inhibition of AGE’s fluorescence intensity is also observed when gBSAara is treated with AuNP for 80 h. The inhibitory strength of the AuNP decreases when the AGE’s intensity is checked by exciting at 335 nm after 34 h of incubation period as shown in Fig. 8. At 24 h, the production of AGE’s intensity is inhibited by 55%

J Fluoresc Fig. 8 AGE’s fluorescence spectra for excitation at λex = 335 nm after 24 h (a) and 34 h (b) where 1,2 and 3 represents AGE’s fluorescence spectra of gBSara in presence of AuNP at 0 nM, 0.13 nM and 0.68 nM respectively

when the concentration of the AuNP is 0.13 nM; whereas, the inhibition of AGE’s intensity of gBSAara is decreases to 16% after 34 h with the same concentration of the AuNP. Similar observations are also made when the AGE’s fluorescence intensity is monitored by exciting gBSAara incubated with AuNP’s at 370 nm. AGE’s fluorescence intensity of gBSAara is reduced to 49% at 24 h when gBSAara is incubated with 0.13 nM concentration of AuNP; whereas, there is only 16% inhibition of the AGE’s fluorescence intensity at the same concentration of the AuNP when it is checked after 34 h incubation period (Fig.S7). The AGE’s fluorescence intensity vs time plots for 335 nm and 370 nm excitation are shown in Fig. S8. The decrease in the generation of the AGE’s fluorescence intensity suggested that AuNP inhibited the formation of the AGE’s related products. Interestingly, AuNPs retain a promising inhibitory effect throughout the incubation period for a total length of 80 h.

Conclusion The binding of TBM is relatively stronger to native BSA than CPM. However, the presence of spherical colloidal NPs modulates the binding behavior in a significant way. For example, in case of drug binding to BSA-AuNP conjugate, the binding of CPM becomes stronger and the binding strength of TBM decreases. CPM binds to the BSA-AuNP conjugate through hydrophobic forces; whereas, TBM binding is best explained by a combination of van der Waals and hydrogen bonding interaction. Both the drugs bind to the BSA-AuNp conjugate by ground state complex formation through a static mechanism. AuNP also inhibit the production of the AGE’s compounds in a concentration dependent fashion. There is no major change in the conformation of the protein through glycation and in the formation of the BSA-AuNP conjugate. The inhibition of AGE product formation by AuNP reveals the anti-glycating ability and its modulatory effect in drug binding process could be a useful tool in the field of nanomedicine. Further the difference in the mode of

interaction of the drugs can help in developing and designing various type of anti-diabetes drug.