Comparative studies on biophysical interactions between gambogic acid and serum albumin via multispectroscopic approaches and molecular docking

Author’s Accepted Manuscript Comparative Studies on Biophysical Interactions between Gambogic Acid and Serum Albumin via Multispectroscopic Approaches and Molecular Docking Yi Wang, Lijun Wang, Meiqing Zhu, Jiaying Xue, Rimao Hua, Qing X. Li www.elsevier.com/locate/jlumin

PII: DOI: Reference:

S0022-2313(18)31013-5 https://doi.org/10.1016/j.jlumin.2018.09.005 LUMIN15888

To appear in: Journal of Luminescence Received date: 8 June 2018 Revised date: 15 August 2018 Accepted date: 3 September 2018 Cite this article as: Yi Wang, Lijun Wang, Meiqing Zhu, Jiaying Xue, Rimao Hua and Qing X. Li, Comparative Studies on Biophysical Interactions between Gambogic Acid and Serum Albumin via Multispectroscopic Approaches and Molecular Docking, Journal of Luminescence, https://doi.org/10.1016/j.jlumin.2018.09.005 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Comparative Studies on Biophysical Interactions between Gambogic Acid and Serum Albumin via Multispectroscopic Approaches and Molecular Docking Yi Wang*1, Lijun Wang1, Meiqing Zhu1, Jiaying Xue1, Rimao Hua*1, Qing X. Li2 (1. Key Laboratory of Agri-food Safety of Anhui Province, College of Resources and Environment, Anhui Agricultural University, Hefei 230036, China. 2. Department of Molecular Biosciences and Bioengineering, University of Hawaii at Manoa, Honolulu, HI 96822, USA.)

* Corresponding Author Email address: [email protected]; [email protected]

Abstract Gambogic acid (GA) is insecticidal and cytotoxic to various cancer cells. This study focused on mechanisms of interactions between GA and human serum albumin (HSA) and bovine serum albumin (BSA). Spectra of steady-state fluorescence, UV-Vis, and time-resolved fluorescence indicated that GA binding to HSA/BSA is a static process. The site maker experiments suggest the binding for HSA-GA and BSA-GA systems both occurs at site II (subdomain IIIA). The complex of GA with HSA/BSA can distribute efficiently in vivo under the value of binding constants in the intermediate range. Thermodynamic parameters illustrate the binding of GA to HSA/BSA is collaboratively driven by van der Waals force and

a hydrogen bond in a spontaneous process. Molecular modeling studies revealed the binding of GA to the site II via hydrogen bond and π-cation interactions. Such interactions between GA and HSA/BSA are informational for use of gambogic acid as a lead compound for insecticide development.

Graphical abstract:

Key words: Gambogic acid, Serum albumin, Spectroscopy, π-cation interaction

1. Introduction Gambogic acid (GA) (Fig. 1) is a xanthonoid extracted from the resin of Garcinia hanburyi. GA exhibits good insecticidal activities, detoxification effects and low toxicity in normal tissues and can prevent from bleeding [1-6]. Recently, some pharmacological research

revealed that GA could be a promising chemotherapeutic agent for the treatment of various cancers [7], such as lung cancer [8], hepatoma [9], leukemia [10], gastric carcinoma [11], and breast carcinoma [12] both in vitro and in vivo. In addition, GA bound with other chemotherapeutics or nanoparticles can significantly improve antitumor effects against various cancers with reduced side effects, overcoming the limitation of poor water solubility and low chemical stability [13, 14]. The widespread use of anticancer drugs in human treatments and veterinary applications has attracted research attention in recent years [15, 16]. Both human serum albumin (HSA) and bovine serum albumin (BSA) are commonly used as representative model proteins in chemistry, clinical medicine and life sciences. HSA and BSA have been used to serve as the possible targets to investigate the interaction of small molecules with biomacroprotein. HSA and BSA share a similar conformation presenting a high homology about 76% of sequences [17, 18]. They both consist of three domains I, II and III with two sub-domains (A and B) being a hydrophobic cavity. The primary structure of HSA and BSA contains 585 and 583 amino acid residues, respectively [19-21]. However, HSA contains only a single tryptophan (Trp-214) in subdomain IIA, while BSA owns two tryptophan’s (Trp-134 in subdomain IB and Trp-212 in subdomain IIA). Trp-134 is located on the BSA surface, while Trp-212 is wrapped in a hydrophobic environment inside a protein pocket [22-24]. Up to now, many researchers have done research on the binding of different drug ligands with serum albumins. Tee and coworkers investigated the interaction between gefitinib and HSA through static quenching mechanism [25]. Ding’s group reported several pesticides upon binding to HSA, such as the fungicide metalaxyl, insecticide imidacloprid and

sulfonylurea herbicides [26-28]. Sharifi and colleagues provided the information of the organophosphorus insecticide chlorfenvinphos interacting with both serum albumins [29]. He et al. [30] reported static quenching mechanism of acotiamide hydrochloride upon HSA binding. Handing et al. [31] found a novel drug binding (CBS1) when cetirizine connects with serum albumins. Zhang et al. [31] prepared the HSA-GA nanoparticles as a new delivery system for lung cancer therapy. Moreover, the binding of drugs to protein led a direct effect on drug transport in blood, drug elimination rate and binding site in vivo [33, 34]. In the present study, we studied quenching mechanism, thermodynamic parameters, conformation changes and high-affinity binding site of HSA/BSA in the presence of GA (Fig. 1). We also compared HSA-GA and BSA-GA systems via multiple spectroscopic techniques and molecular docking. The study acquired information valuable for use of GA as a lead compound for insecticide development and potential effect to human.

2. Materials and Methods 2.1 Reagents Both stock solutions of HSA (J&K, Beijing, China) and BSA (Sigma-Aldrich, St. Louis, MO, USA) were prepared in Tris-HCl buffer (10 mM, pH = 7.4). GA (purity ≥ 98%) was obtained from Huzhou Zhanshu Bio-technology Co., Ltd (China). GA was dissolved in 99.9% ethanol and finally diluted in Tris-HCl buffer to appropriate concentrations. The values of pH were measured with a REX PHS-25 digital pH meter (Shanghai, China). Other reagents were analytical grade without further purification. Ultrapure water was obtained from a Milli-Q Plus System (Billerica, MA).

2.2 Absorbance measurements Absorption spectra of GA binding to HSA/BSA were recorded on a UV-1800 UV spectrophotometer (Shimadzu, Japan) with 1-cm quartz cell, scanning from 200 to 500 nm. The background of the corresponding buffer solution was subtracted to correct the signal caused by the GA standard in the UV-vis spectra. 2.3 Fluorescence spectroscopy measurements Fluorescence spectra were recorded on an Inesa 970CRT fluorescence spectrophotometer (Shanghai, China) under the condition of excitation wavelength at 295 nm (only exciting Trp residues) and the emission wavelength in the range of 300-450 nm. In addition, the excitation and emission slit widths were set at 5.0 nm and the scanning speed was 240 nm / min. The final concentrations of GA ranged from 0.2 × 10−6 M to 1.6 × 10−6 M in titration experiments at three different temperatures 298, 307 and 316 K controlled by water bath. In the site marker experiments, the reaction of GA and HSA/BSA was in a 3-mL volumetric cuvette and the site markers were in equimolar concentrations. The quenching experiments were triplicate. Time-resolved fluorescence was measured on a FluoroLog-3 fluorescence lifetime spectrometer (Horiba, France) at room temperature, through time-resolved intensity decay using the time-correlated single photon counting (TCSPC). The fluorescence decays were deconvoluted. The average lifetime was calculated via the DataStation software to confirm the binding mechanism of GA to HSA/BSA. 2.4 Circular Dichroism (CD) spectrum measurements CD spectra of HSA and BSA with and without GA were measured on a Jasco 810

spectropolarimeter (Tokyo, Japan) from 200 to 240 nm in a 1-cm length quartz cuvette. All CD spectra were recorded with a 1-nm step resolution and the scan speed of 50 nm min-1, and then final data was averaged after five continuous scans. The buffer blank background was subtracted from all CD spectra. The HSA/BSA secondary structures were calculated with CDSSTR software supported by CDPro software package [34]. 2.5 FT-IR spectrum measurements FT-IR spectra were recorded on a Thermo Scientific Nicolet iS50 FTIR spectrometer (MA, USA) equipped with attenuated total reflection (ATR). All FT-IR measurements between HSA/BSA and GA were collected from 500 to 4000 cm-1 at ambient temperature and taken by 32 scans at 4 cm-1 resolution. All original data were minus the corresponding absorbance contributions of buffer under the same conditions, and analyzed with the OMNIC software. 2.6 Synchronous fluorescence spectrum measurements Synchronous fluorescence spectra were recorded with a Cary Eclipse fluorescence spectrophotometer (Agilent Technologies) using with a 1-cm path length and 3-mL volume in a quartz cuvette. The scan range of wavelength was set from 260 to 320 nm when Δλ = 15 nm, and from 240 to 320 nm when Δλ = 60 nm. 2.7 Three dimension (3D) fluorescence spectrum measurements 3D fluorescence spectra were obtained on a Cary Eclipse fluorescence spectrophotometer (Agilent) with the scanning rate of 9600 nm min-1, excitation from 200 to 350 nm and emission wavelength from 220 to 500 nm. 2.8 Molecular modelling

All molecular dockings experiments were accomplished with Docking Software Schröodinger 2009. The crystal structures of HSA-CMPF (PDB ID: 2BXA, resolution: 2.35 Å) and BSA-naproxen (PDB ID: 4OR0, resolution: 2.58 Å) were applied for molecular docking of GA. The protein in the crystal structures was prepared using the Protein Preparation Wizard workflow. GA was prepared with LigPrep. The binding site was defined with the crystal ligand in each complex structure and GA was docked into the binding site using the InducedFit docking mode. The binding free energy was calculated with MM/GBSA method. 3. Results and Discussion 3.1 Fluorescence quenching of GA to HSA/BSA To date, quenching of intrinsic Trp fluorescence in existence of drugs is the most convenient method, benefitting from its sensitivity towards the disturbance of the local microenvironment [36]. The fluorescence quenching of both HSA and BSA accompanied by a gradient of concentration of GA was described in Fig. 1. The addition of GA caused the declining fluorescence intensity and slight changes in the maximum peak, indicating that the microenvironment of Try residue changed in both HSA-GA and BSA-GA systems [37]. The maximum emission wavelength of peak in HSA-GA and BSA-GA systems was 345 nm and 347 nm, respectively. [Q]90 concentration refers to quenching 90% of protein fluorescence and smaller [Q]90 value represents stronger quenching ability in the literature [38]. In Fig. 1D, [Q]90 value in HSA-GA system was lower than that in BSA-GA system by nonlinear curve fit, suggesting the quenching ability of GA towards HSA is stronger than that in BSA solution. The result is in accordance with the previous report that the binding affinity of drug ligand

binding with HSA is more powerful than that with BSA [39]. 3.2 Fluorescence quenching mechanisms of GA to HSA/BSA Quenching mechanisms are often classified into dynamic or static, depending on their different responses upon temperature and viscosity. Fluorescence quenching, whether dynamic or static, both can be analyzed with the Stern-Volmer equation [40]: F0 ⁄F =1+Kq τ0 [Q]=1+KSV [Q] (1) where F0 and F are the fluorescence intensities in the absence and presence of the quencher, respectively. Kq and KSV are the bimolecular quenching constant and the Stern-Volmer quenching constant, respectively. [Q] is the quencher concentration.𝜏0 is the average lifetime of fluorophore in lack of quencher. In Fig. 2, plots of F/F0 versus [Q] for interaction of GA with HSA/BSA show a single class of quenching mechanism from the linear plots of GA-HSA/BSA. The values of KSV increased gradually both in HSA-GA and BSA-GA systems by increasing the temperatures from 295 to 316 K. Moreover, the values of Kq calculated in Table 1 are as large as 1000-fold than the limiting diffusion rate constant of the biomolecule (Kq is near 1 × 1010 M-1 s-1), indicating that the quenching mechanism of both HSA-GA and BSA-GA systems are probably static mechanism [40]. The absorption spectra of fluorophore can also confirm static or dynamic quenching [40]. The absorption spectra of dynamic quenching are that the excited states of fluorophore with no changes. On the contrary, a static progress often leads to disturbance of fluorophore absorption spectra [41]. As shown in Fig. S2, the absorption intensities of HSA/BSA were on the rise with the GA concentration increasing, indicating unity of HSA-GA and BSA-GA complexes [23]. These results can verify the quenching mechanism of GA binding to

HSA/BSA is static quenching. Importantly, the result of lifetime measurements on the serum albumins is the most reliable evidence to differentiate whether static or dynamic quenching [40]. When the dynamic quenching occurs, the fluorophore lifetime reduces constantly, while there is no variation in presence of quencher in static quenching [42]. Based on the data obtained from lifetime experiments in Fig. S1, the concentration radio of HSA and GA was set at 1:0, 1:1, 1:2, thus the corresponding 𝜏0 was 5.30, 5.21, 5.29 ns in HSA-GA system, respectively. For the same radio of BSA and GA, the corresponding 𝜏0 was 5.81, 5.72, 5.73 ns in BSA-GA system, respectively. From the above results of 𝜏0 , the lifetime of HSA/BSA in the absence or presence of GA did not change obviously. Therefore, the quenching of HSA/BSA-GA is indeed static quenching, which is also supported by the absorption spectra between HSA/BSA and GA. 3.3 Binding constants and location To recognize the binding sites of HSA/BSA with GA, the fluorescent data can be expressed mathematically by double-logarithmic equation (2) [43]: log[ (F0 -F)⁄F]=log Ka + n log[Qt -(F0 -F)Pt /F0 ] (2) where the definition of F0, F are the same as that of eq. (1), [Qt] and [Pt] are the total quencher concentration and the total protein concentration, respectively. n is the average number of binding sites, and Ka is the binding affinity constant between quencher and biomolecule at temperature. Fig. 2 shows the plots of log[ (F0 -F)⁄F] versus log[Qt -(F0 -F)Pt /F0 ] at 298 K. The value of n between HSA/BSA and GA can be obtained from the slope of a linear line by the

plots and listed in Table 2, which is approximately 1 in both HSA-GA and BSA-GA systems. It shows the interaction of GA with HSA/BSA at one binding site in the experimental concentration range. The competitive binding experiments were performed via the fluorescence titration method using three site markers phenylbutazon (PB), flufenamic acid (FA) and digitoxin (Dig) for sites I (subdomain IIA), II (subdomain IIIA) and III, respectively [44, 45]. The binding affinity constants in the presence of site markers were listed in Table S1 according to eq. (2). The values of binding affinity constants reduced distinctly upon the addition of FA, demonstrating that site II (subdomain IIIA) might be the primary binding site both in HSA-GA and BSA-GA systems. Binding constants of some tight protein-ligand compounds range from 107 to 109 L mol-1. In the contrast, the value of Ka of GA binding to HSA/BSA in the range of 105 - 106 is below this range (Table 3), indicating that the affinity binding in HSA/BSA for GA was relatively moderate [46]. Thus, when GA gets into circulatory system of the human or animal body, it would be transported and reach its target sites easily. 3.4 Thermodynamic parameters and interaction modes Noncovalent bonds define the strength between a drug molecule and its target. Noncovalent bonds include hydrogen bonds, salt bridges, hydrophobic and van der Waals interactions, π effects and steric contacts within the binding site [47, 48]. The analysis of thermodynamic parameters including enthalpy change ΔH, and entropy change ΔS, and free energy change ΔG can account for which force is involved and further clarify interaction modes. The calculation of these parameters is according to Van’t Hoff equation (3 and 4) [49].

ln Ka =- ∆H⁄RT +∆S/R (3) ∆G=∆H-T∆S (4) where Ka is the same as that of eq. (3) and universal gas constant R is about 8.314 J mol-1 K-1. The data of thermodynamic parameters shown in Table 3 were acquired at 298, 307 and 316 K. On the basis of Ross’s theory, the negative values of ΔH and ΔS suggest that van der Waals interactions and a hydrogen bond play a predominant role in the interaction of GA binding to HSA/BSA with exothermic performance [50]. Besides, the negative ΔG, the combined performance of enthalpy and entropy changes, reveals that the GA interacts with HSA/BSA spontaneously [39]. 3.5 CD response to HSA/BSA The protein secondary structure can be described by α-helices, β-sheets, β-turns and random coils. These contents of HSA/BSA with and without GA were calculated and shown in Table S2 via CD spectroscopy. The CD spectra of both HSA and BSA with GA in different concentration ratios depicted two negative minima at 208 and 222 nm in Fig. 3, which is typical evaluation of the α-helix structure of class proteins [51]. In different molar ratios of HSA-GA and GA-BSA complexes (1:0, 1:1, and 1:2), the α-helix structure is increased in comparison to free HSA/BSA (Table S2) and the intensity of the negative peaks enhanced without any detectable alter in the peak positions. These results indicate the secondary structure changed because of the GA binding to HSA/BSA [52, 53]. 3.6 FT-IR spectroscopy Information of amide bands in proteins can be obtained by FT-IR spectra, which contributes to realizing the frequencies of the peptide moiety and analyzing the influence of

GA on HSA/BSA secondary structures. Among the amide bands, the amide I (1700 - 1600 cm-1) is mainly associated with the C=O stretch band and the amide II (1600-1500 cm-1) is the C-N stretching coupled with N-H bending, which are the most widely-used vibrational bands in study [54]. The FT-IR spectra of GA combined with HSA/BSA were displayed in Fig. 3. The amide I band peak position in HSA-GA and BSA-GA systems shift from 1657.9 cm-1 to 1651.2 cm-1, 1645.5 cm-1 to 1654.6 cm-1, respectively. The peak position of amide II band moved from 1543.7 cm-1 to 1552.9 cm-1 (in HSA-GA system) and from 1543.7 cm-1 to 1544.9 cm-1 (in BSA-GA system), respectively. The changes of positions and shapes of peaks showed that the GA bound to HSA/BSA induces the polypeptide carbonyl hydrogen-bonding network achieved rearrangement. In addition, the intensity variation of the amide I band in the presence of GA is attributed to the content of protein α-helical structure [39]. Changes of protein secondary structure with the addition of GA were verified. 3.7 Synchronous fluorescence spectra Synchronous fluorescence spectra include the simultaneously scanned excitation and the emission fluorescence spectra, which provide feature information about the tryptophan (Trp) and tyrosine residues of HSA/BSA upon GA binding when Δλ = 15 and 60 nm [55, 56]. In Fig. 4 (A and B), fluorescence intensity quenching of Trp was stronger in compassion with that of Tyr, which implied that Trp contribute more to quenching the intrinsic fluorescence of HSA/BSA. No changes in the maximum emission were obtained when Δλ = 60 nm, and there appears a red shift when Δλ = 15 nm with GA concentration increasing both in HSA-GA and BSA-GA system, indicating the conformations and polarity of HSA and BSA upon binding with GA changed.

3.8 3D fluorescence spectra 3D fluorescence spectrum can provide some detailed information about the conformational and micro-environmental changes of proteins in comparison to their spectral variation. It is therefore used widely in fluorescence analysis in recent years. The peak position (λex/λem) and intensity representing characteristic information were listed in Table S3 and the corresponding contour maps data were shown in Fig. 5. One of two peaks is the Rayleigh scattering peak a (λex = λem) and the other is the second-order scattering peak b (2λex = λem) [57,58]. In addition, the strong peaks 1 and 2 illustrated the spectral features of the polypeptide backbone structure and Trp and Tyr residues on proteins, respectively. Upon the addition of GA, the strong peaks 1 and 2 generated a slight blue shift. Their fluorescence intensities reduced in HSA-GA system, while a slight redshift occurred and the changed trend in terms of fluorescence intensity also decreased in the BSA-GA system (Table S3). Hence, these results revealed that some conformational and micro-environmental changes appeared upon binding of GA to HSA/BSA, which agreed with the synchronous fluorescence spectra. 3.9 Resonance energy transfer from HSA/BSA to GA Förster resonance energy transfer (FRET) occurred between a donor molecule in the excited state and an acceptor molecule in the ground state, which can be used to determine the drug binding site distance between amino acid residues and a drug in vitro and in vivo [59, 60]. This parameter is expressed by the HSA/BSA fluorescence emission spectrum and the overlapping of the drug absorption spectrum. Förster’s nonradiation energy transfer theory suggests that the energy transfer efficiency (E) can be calculated by the equation (5). F

E=1- F = 0

R0 6 R0 6 +r6

(5)

where F and F0 are the HSA/BSA fluorescence intensity with and without GA. The binding distance r and the critical distance R0 with 50% transfer efficiency between the acceptor (GA) and donor (HSA/BSA) are acquired by the equation (6). R0 6 =8.8×10-25 K2 N-4 ϕ J (6) where N is the refractive index of a medium, K2 is the spatial orientation factor of the dipole,

ϕ is the donor fluorescence quantum yield. The degree of overlap integral between the acceptor absorption spectrum and donor fluorescence emission spectrum called J was shown in Fig. S2, and with its value computed by equation (7).8.8 J=

∑ F(λ)ε(λ)λ4∆λ ∑ F(λ)∆λ

(7)

where F(λ) and ε(λ) are the fluorescence intensity of the donor and donor at wavelength λ， respectively. In case of the interaction of HSA/BSA with drugs, N, K2 and ϕ are 1.336, 2/3 and 0.14, respectively. Table S4 displays the binding distance r is 4.51 nm (HSA-GA system) and 4.79 nm (BSA-GA system), both are below 8 nm as well as 0.5R0 < r < 1.5R0, suggesting energy transfer from the Trp residue in HSA/BSA to GA and the existence of static quenching interaction between HSA/BSA and GA with high probability [48]. Besides, the different values of r imply that the conformational changes of HSA-GA system differ from that in BSA-GA system, although GA binds the similar site on the HSA and BSA [61]. 3.10 Molecular modeling study Molecular modes of GA in the binding sites of HSA and BSA predicted by molecular docking and attested to the discussion of site competition experiments. In Fig. 6, the results of molecular docking for interaction of GA and HSA/BSA revealed that GA located in in site II (subdomain IIIA) in consistent with previous results. Fig. 6 shows the proteins in cartoon, the

key residues in stick, GA in cyan stick and hydrogen bonds in magenta dashes with the label of length. GA binding to HSA formed six hydrogen bond forces surrounded by Lys414 (one H-bond), Arg410 (three H-bonds), Gln390 (one H-bond), Ser489 (one H-bond). GA binding to BSA formed seven hydrogen bond forces surrounded by Lys412 (two H-bonds), Arg409 (two H-bonds), Gln393 (two H-bonds) and Tyr410 (one H-bond). Numerous hydrogen bonds can strengthen the interactions between GA and HSA/BSA and stabilize GA within the binding pocket at site II (subdomain IIIA) of HSA and BSA [50]. In addition, π-cation interaction was formed between aromatic moieties of GA and Arg410 of HSA or Arg409 of BSA at the binding site (Fig. S3). The binding free energy ΔG by molecular docking is -28.00 kcal/mol (HSA-GA system) and -25.80 kcal/mol (BSA-GA system). Therefore, molecular modeling studies indicated the binding of GA to site II of HSA/BSA by formation of hydrogen bond and π-cation interactions. These results can verify the validity of fluorescence quenching experiments. In conclusion, interactions of GA binding to HSA/BSA were described with multi-spectroscopic techniques and molecular docking. GA binds efficiently with HSA/BSA protein in site II (subdomain IIIA) and furthermore quenched HSA/BSA intrinsic fluorescence, accompanying their secondary structure changes. In addition, [Q]90 suggests that the quenching ability of GA in HSA solution is more efficient than that in BSA solution. The binding affinity of HSA-GA system is greater than that of BSA-GA system, which are of great benefit to comprehend the transport mechanism of GA in vivo and provide valuable information for developing its derivatives with valid pharmacology properties.

Acknowledgements This work was supported in part by the Natural Science Foundation of China (NSFC) (No. 31601657), Open Funding of Key Laboratory of Agri-food Safety of Anhui Province and Open Funding of State Key Laboratory of Elemento-organic Chemistry.

References

[1] H. Auterhoff, H. Frauendorf, W. Liesenklas, C. Schwandt, The chief constituents of gamboge resins. 1. Chemistry of gamboge, Arch. Pharm. Ber. Dtsch. Pharm. Ges., 295/67 (1962) 833-846. [2] D.L. Wang, G.Y. Cao, P. Fang, L. Xia, B.J. Cheng, Comparative transcription analysis of different Antirrhinum phyllotaxy nodes identifies major signal networks involved in vegetative-reproductive transition, PLoS One, 12 (2017) e0178424. [3] X. Wu, L. Long, J. Liu, J. Zhang, T. Wu, X. Chen, B. Zhou, T.-Z. Lv, Gambogic acid suppresses inflammation in rheumatoid arthritis rats via PI3K/Akt/mTOR signaling pathway, Mol. Med. Report., 16 (2017) 7112-7118. [4] Q. Guo, Q. Qi, Q. You, H. Gu, L. Zhao, Z. Wu, Toxicological studies of gambogic acid and its potential targets in experimental animals, Basic Clin. Pharmacol. Toxicol., 99 (2006) 178-184. [5] S. G. Li, B.G. Zhou, M.Y. Li, S. Liu, R.M. Hua, H.-F. Lin, Chemical composition of Illicium verum fruit extract and its bioactivity against the peach-potato aphid, Myzus persicae (Sulzer), Arthropod-Plant Inte, 11 (2017) 203-212. [6] J. Liang, H.Yan, P. Puligundla, X. L. Gao, Y. B. Zhou, X. C. Wan. Applications of

chitosan nanoparticles to enhance absorption and bioavailability of tea polyphenols: A review. Food Hydrocolloids, 69 (2017) 286-292. [7] K. Hao, X.Q. Liu, G.J. Wang, X.P. Zhao, Pharmacokinetics, tissue distribution and excretion of gambogic acid in rats, Eur. J. Drug Metabol. Pharmacokinet., 32 (2007) 63-68. [8] X.L. Zhu, H.M. Zhang, Y. Lin, P.S. Chen, J. Min, Z.Z. Wang, W. Xiao, B.A. Chen, Mechanisms of gambogic acid-Induced apoptosis in non-small cell lung cancer cells in relation to transferrin receptors, J. Chemother., 21 (2009) 666-672. [9] Y. Yang, L. Yang, Q.-D. You, F.-F. Nie, H.-Y. Gu, L. Zhao, X.-T. Wang, Q.-L. Guo, Differential apoptotic induction of gambogic acid, a novel anticancer natural product, on hepatoma cells and normal hepatocytes, Cancer Lett., 256 (2007) 259-266. [10] M.K. Pandey, B. Sung, K.S. Ahn, A.B. Kunnumakkara, M.M. Chaturvedi, B.B. Aggarwal, Gambogic acid, a novel ligand for transferrin receptor, potentiates TNF-induced apoptosis through modulation of the nuclear factor-kappa B signaling pathway, Blood, 110 (2007) 3517-3525. [11] L. Zhao, Q.L. Guo, Q.D. You, Z.Q. Wu, H.Y. Gu, Gambogic acid induces apoptosis and regulates expressions of Bax and Bcl-2 protein in human gastric carcinoma MGC-803 cells, Biol. Pharm. Bull., 27 (2004) 998-1003. [12] J. Chen, H.-Y. Gu, N. Lu, Y. Yang, W. Liu, Q. Qi, J.-J. Rong, X.-T. Wang, Q.-D. You, Q.-L. Guo, Microtubule depolymerization and phosphorylation of c-Jun N-terminal kinase-1 and p38 were involved in gambogic acid induced cell cycle arrest and apoptosis in human breast carcinoma MCF-7 cells, Life Sci., 83 (2008) 103-109.

[13] T.T. Wang, J. Wei, X.P. Qlan, Y.T. Ding, L.X. Yu, B.R. Liu, Gambogic acid, a potent inhibitor of survivin, reverses docetaxel resistance in gastric cancer cells, Cancer Lett., 262 (2008) 214-222. [14] L.H. Fang, B.A. Chen, S.L. Liu, R.P. Wang, S.Y. Hu, G.H. Xia, Y.L. Tian, X.H. Cai, Synergistic effect of a combination of nanoparticulate Fe3O4 and gambogic acid on phosphatidylinositol 3-kinase/Akt/Bad pathway of LOVO cells, Int. J. Nanomed., 7 (2012) 4109-4118. [15] Z. Zhang, L. Yang, H. Ye, X. F. Du, Z. M. Gao, Z. L. Zhang. Effects of pigment extract from black glutinous corncob in a high-fat-fed mouse model of hyperlipidemia. Eur. Food Res. Technol., 230 (2010) 943-946. [16] H. Huang, X. Xu, J. Lv, G. Li, E. Wang, Y. Gao, Impact of proteins and saccharides on mass production of Tyrophagus putrescentiae (Acari: Acaridae) and its predator Neoseiulus barkeri (Acari: Phytoseiidae), Biocontrol Sci. Technol., 23 (2013) 1231-1244. [17] L. Chen, H. Guo, Y. Lin, Y. Wu, H. Cheng. Molecular cloning and characterization of the cinnamate 4-hydroxylase gene from eupatorium adenophorum. Weed Biol. Manage., 14 (2015) 167-177. [18] Y. Wang, Y. Yang, C. Wei, X. Wan, H. J. Thompson. Principles of Biomedical Agriculture Applied to the Plant Family Theaceae to Identify Novel Interventions for Cancer Prevention and Control. J. Agric. Food Chem., 64 (2016) 2809-2814. [19] P. Sengupta, P.S. Sardar, P. Roy, S. Dasgupta, A. Bose, Investigation on the interaction of Rutin with serum albumins: Insights from spectroscopic and molecular docking

techniques, J. Photochem. Photobiol. B: Biol., 183 (2018) 101-110. [20] Y. Wu, X. Du, H. Ge, Z. Lv,. Preparation of microporous starch by glucoamylase and ultrasound. Starch - Starke, 63 (2011), 217–225. [21] B.J. Zhu, C.L. Liu, Q.N. Liu, L.S. Dai, Characterization of a female-specific protein from the wild silkworm Actias selene, Bull. Insectology, 65 (2012) 107-112. [22] A. Papadopoulou, R.J. Green, R.A. Frazier, Interaction of flavonoids with Bovine serum albumin:  A fluorescence quenching study, J. Agric. Food. Chem., 53 (2005) 158-163. [23] Z.J. Cheng, Studies on the interaction between scopoletin and two serum albumins by spectroscopic methods, J. Lumin., 132 (2012) 2719-2729. [24] B.J. Zhu, Y.Y. Yu, J. Gao, Y.Y. Feng, L. Tang, Y.X. Sun, L.L. Yang, Characterization and function of a novel calmodulin-like protein from crayfish Procambarus clarkii, Fish Shellfish Immunol., 67 (2017) 518-522. [25] M.Z. Kabir, W.-V. Tee, S.B. Mohamad, Z. Alias, S. Tayyab, Interaction of an anticancer drug, gefitinib with human serum albumin: insights from fluorescence spectroscopy and computational modeling analysis, RSC Adv., 6 (2016) 91756-91767. [26] F. Ding, W. Liu, X. Zhang, L.J. Wu, L. Zhang, Y. Sun, Identification of pyrazosulfuron-ethyl binding affinity and binding site subdomain IIA in human serum albumin by spectroscopic methods, Spectrochim. Acta, Pt. A: Mol. Spectrosc., 75 (2010) 1088-1094. [27] F. Ding, B.Y. Han, W. Liu, L. Zhang, Y. Sun, Interaction of imidacloprid with hemoglobin by fluorescence and circular dichroism, J Fluoresc., 20 (2010) 753-762. [28] F. Ding, X.N. Li, J.X. Diao, Y. Sun, L. Zhang, Y. Sun, Chiral recognition of metalaxyl

enantiomers by human serum albumin: evidence from molecular modeling and photophysical approach, Chirality, 24 (2012) 471-480. [29] T. Sharifi, Y. Ghayeb, T. Mohammadi, Study of conformational changes in serum albumin by binding of chlorfenvinphos using multispectroscopic techniques and molecular dynamic simulation, Monatsh. Chem., 148 (2017) 781-791. [30] J. He, H. Yang, S. Li, K. Xu, Q. Wang, Y. Huang, H. Li, Characterization of the interaction between acotiamide hydrochloride and human serum albumin: 1H STD NMR spectroscopy, electrochemical measurement, and docking investigations, RSC Adv., 6 (2016) 61119-61128. [31] K.B. Handing, I.G. Shabalin, K. Szlachta, K.A. Majorek, W. Minor, Crystal structure of equine serum albumin in complex with cetirizine reveals a novel drug binding site, Mol. Immunol., 71 (2016) 143-151. [32] Y. Zhang, Z. Yang, X. Tan, X. Tang, Z. Yang, Development of a more efficient albumin-based delivery system for gambogic acid with low toxicity for lung cancer therapy, AAPS PharmSciTech, 18 (2016) 1987-1997. [33] B. Sandhya, A.H. Hegde, S.S. Kalanur, U. Katrahalli, J. Seetharamappa, Interaction of triprolidine hydrochloride with serum albumins: Thermodynamic and binding characteristics, and influence of site probes, J. Pharm. Biomed. Anal., 54 (2011) 1180-1186. [34] J.N.A. Tian, Y.C. Zhao, X.H. Liu, S.L. Zhao, A steady-state and time-resolved fluorescence, circular dichroism study on the binding of myricetin to bovine serum albumin, Lumin., 24 (2009) 386-393.

[35] F. Ding, J.X. Diao, Y. Sun, Y. Sun, Bioevaluation of human serum albumin-hesperidin bioconjugate: insight into protein vector function and conformation, J. Agric. Food. Chem., 60 (2012) 7218-7228. [36] V.K. Sonu, M.M. Islam, A.B. Gurung, A. Bhattacharjee, S. Mitra, Serum albumin interaction with xanthine drugs at nano-bio interfaces: A combined multi-spectroscopic and molecular modelling approach, J. Mol. Liq., 242 (2017) 919-927. [37] T.H. Wang, Z.M. Zhao, B.Z. Wei, L. Zhang, L. Ji, Spectroscopic investigations on the binding of dibazol to bovine serum albumin, J. Mol. Struct., 970 (2010) 128-133. [38] C.Y. Tan, E. Alas, J.G. Muller, M.R. Pinto, V.D. Kleiman, K.S. Schanze, Amplified quenching of a conjugated polyelectrolyte by cyanine dyes, J. Am. Chem. Soc., 126 (2004) 13685-13694. [39] X. Li, G. Wang, D. Chen, Y. Lu, β-Carotene and astaxanthin with human and bovine serum albumins, Food Chem., 179 (2015) 213-221. [40] J.R. Lakowicz, Principles of fluorescence spectroscopy, 3rd Springer Science+Business Media, New York, 2006. [41] N. Cacita, S. Nikolaou, Studying the interaction between trinuclear ruthenium complexes and human serum albumin by means of fluorescence quenching, J. Lumin., 169 (2016) 115-120. [42] M. Mondal, P. Lakshmi, R. Krishna, N. Sakthivel, Molecular interaction between human serum

albumin

(HSA)

and

phloroglucinol

derivative

that

shows

selective

anti-proliferative potential, J. Lumin., 192 (2017) 990-998. [43] P. Mandal, T. Ganguly, Fluorescence spectroscopic characterization of the interaction of

human adult hemoglobin and two isatins, 1-methylisatin and 1-phenylisatin: A comparative study, J. Phys. Chem. B, 113 (2009) 14904-14913. [44] G. Sudlow, D.J. Birkett, D.N. Wade, Further characterization of specific drug binding sites on human serum albumin, Mol. Pharmacol., 12 (1976) 1052-1061. [45] I. Sjoholm, B. Ekman, A. Kober, I. Ljungstedt-Pahlman, B. Seiving, T. Sjodin, Binding of drugs to human serum albumin:XI. The specificity of three binding sites as studied with albumin immobilized in microparticles, Mol. Pharmacol., 16 (1979) 767-777. [46] U. Kragh-Hansen, Structure and ligand binding properties of human serum albumin, Dan. Med. Bull., 37 (1990) 57-84. [47] Y. Wang, J. Liu, M. Zhu, L. Wang, X. Zen, S. Fan, Z. Wang, H. Li, R. Na, X. Zhao, Q.X. Li, Biophysical characterization of interactions between falcarinol-type polyacetylenes and human serum albumin via multispectroscopy and molecular docking techniques, J. Lumin., 200 (2018) 111-119. [48] Z. Liang, Q.X. Li, π-Cation Interactions in molecular recognition: Perspectives on pharmaceuticals and pesticides, J. Agric. Food. Chem., 66 (2018) 3315-3323. [49] X. Peng, X. Wang, W. Qi, R. Su, Z. He, Affinity of rosmarinic acid to human serum albumin and its effect on protein conformation stability, Food Chem., 192 (2016) 178-187. [50] P.D. Ross, S. Subramanian, Thermodynamics of protein association reactions: forces contributing to stability, Biochemistry, 20 (1981) 3096-3102. [51] J. Tian, J. Liu, Z. Hu, X. Chen, Interaction of wogonin with bovine serum albumin, Biorg. Med. Chem., 13 (2005) 4124-4129.

[52] M.B. Bolattin, S.T. Nandibewoor, S.D. Joshi, S.R. Dixit, S.A. Chimatadar, Interaction of hydralazine with human saerum albumin and effect of β-Cyclodextrin on binding: Insights from spectroscopic and molecular docking techniques, Ind. Eng. Chem. Res., 55 (2016) 5454-5464. [53] M.A. Rub, J.M. Khan, N. Azum, A.M. Asiri, Influence of antidepressant clomipramine hydrochloride drug on human serum albumin: Spectroscopic study, J. Mol. Liq., 241 (2017) 91-98. [54] J. Tian, J. Liu, W. He, Z. Hu, X. Yao, X. Chen, Probing the binding of scutellarin to human serum albumin by circular dichroism, fluorescence spectroscopy, FTIR, and molecular modeling method, Biomacromolecules, 5 (2004) 1956-1961. [55] S.R. Feroz, S.B. Mohamad, N. Bujang, S.N.A. Malek, S. Tayyab, Multispectroscopic and molecular modeling approach to investigate the interaction of flavokawain B with human serum albumin, J. Agric. Food. Chem., 60 (2012) 5899-5908. [56] N. Ibrahim, H. Ibrahim, S. Kim, J.P. Nallet, F. Nepveu, Interactions between antimalarial indolone-N-oxide derivatives and human serum albumin, Biomacromolecules, 11 (2010) 3341-3351. [57] M.S. Zaroog, S. Tayyab, Formation of molten globule-like state during acid denaturation of Aspergillus niger glucoamylase, Process Biochem., 47 (2012) 775-784. [58] Q. L. Fu, J. Z. He, L, Blaney, D. M. Zhou. Roxarsone binding to soil-derived dissolved organic matter: Insights from multi-spectroscopic techniques. J. Agric. Food. Chem., 155 (2016) 225-233. [59] T. Förster, O. Sinanoglu Modern quantum chemistry, in, Academic Press, New York,

1996. [60] X. Guo, L. Xu, L. Zhang, H. Wang, X. Wang, X. Liu, J. Yao, A. Hao, One-pot solid phase pyrolysis synthesis of highly fluorescent nitrogen-doped carbon dots and the interaction with human serum albumin, J. Lumin., 196 (2018) 100-110. [61] J. Zhang, X.J. Wang, Y.J. Yan, W.S. Xiang, Comparative studies on the interaction of genistein, 8-chlorogenistein, and 3’,8-dichlorogenistein with bovine serum albumin, J. Agric. Food. Chem., 59 (2011) 7506-7513.

Figure Captions Figure 1. The molecular structure of gambogic acid (A); Fluorescence quenching spectra of HSA/BSA in the presence of different concentrations of GA (B: HSA-GA, C: BSA-GA); (D) The plots as extinction of HSA/BSA fluorophore (F/F0) against GA concentration at 298 K, λex = 295 nm. CHSA = CBSA= 2 × 10-6 M, x: only GA; (a-i) GA concentrations were 0, 0.2, 0.4, 0.6, 0.8, 1.0, 1.2, 1.4 and 1.6 × 10−6 M, respectively. Figure 2. The plots of Stern-Volmer (A: HSA-GA; B: BSA-GA) at 298, 307, 316 K and log[ (F0 -F)⁄F] versus log[Qt -(F0 -F)Pt /F0 ] of the GA-HSA/BSA systems (C: HSA-GA; D: BSA-GA) at 298 K. Data are mean ± SE (bars) (n=3). Figure 3. Far-UV CD spectra (A: HSA-GA; B: BSA-GA) and FTIR spectra (C: HSA-GA; D: BSA-GA): CHSA = CBSA = 1.0 × 10−6 M. The molar ratios of HSA-GA and BSA-GA systems were 1:0, 1:1, and 1:2. Figure 4. Synchronous fluorescence spectra of HSA and BSA in the presence of different concentrations of GA at 298 K: Δλ = 60 nm (A: HSA-GA, B: BSA-GA) and Δλ = 15 nm (C: HSA-GA, D: BSA-GA). CHSA = CBSA = 2×10-6 M. (a-i) GA concentrations were 0, 0.2, 0.4, 0.6, 0.8, 1.0, 1.2, 1.4 and 1.6 ×10−6 M, respectively.

Figure 5. Three dimensional fluorescence spectra of HSA/BSA in the absence and presence of GA at 298 K CHSA = CBSA= 2×10-6 M; A: HSA-GA, 1:0; B: HSA-GA, 1:1; C: BSA-GA, 1:0; D: BSA-GA, 1:1). Figure 6. Molecular modeling of HSA and BSA in presence of GA (A: HSA-GA; B: BSA-GA). The proteins are shown in cartoon while the key residues are shown in stick. GA is shown in cyan stick. Hydrogen bonds are shown in magenta dashes with the label of length.

Figure 1. The molecular structure of gambogic acid (A); Fluorescence quenching spectra of HSA/BSA in the presence of different concentrations of GA (B: HSA-GA, C: BSA-GA); (D) The plots as extinction of HSA/BSA fluorophore (F/F0) against GA concentration at 298K, λex = 295 nm. CHSA = CBSA = 2×10-6 M, x: only GA; (a-i) GA concentrations were 0, 0.2, 0.4, 0.6, 0.8, 1.0, 1.2, 1.4 and 1.6 ×10−6 M, respectively.

Figure 2. The plots of Stern-Volmer (A: HSA-GA; B: BSA-GA) at 298, 307, 316 K and log[ (F0 -F)⁄F] versus log[Qt -(F0 -F)Pt /F0 ] of the GA-HSA/BSA systems (C: HSA-GA; D: BSA-GA) at 298 K. Data are mean ± SE (bars) (n=3).

Figure 3. Far-UV CD spectra (A: HSA-GA; B: BSA-GA) and FTIR spectra (C: HSA-GA; D: BSA-GA). CHSA = CBSA = 1.0 × 10−6 M. The molar ratios of HSA-GA and BSA-GA systems were 1:0, 1:1, and 1:2.

Figure 4. Synchronous fluorescence spectra of HSA and BSA in the presence of different concentrations of GA at 298 K: Δλ = 60 nm (A: HSA-GA, B: BSA-GA) and Δλ = 15 nm (C: HSA-GA, D: BSA-GA). CHSA = CBSA = 2×10-6 M. (a-i) GA concentrations were 0, 0.2, 0.4, 0.6, 0.8, 1.0, 1.2, 1.4 and 1.6 ×10−6 M, respectively.

Figure 5. 3D fluorescence spectra of HSA/BSA in the absence and presence of GA at 298 K. CHSA = CBSA = 2×10-6 M; A: HSA-GA, 1:0; B: HSA-GA, 1:1; C: BSA-GA, 1:0; D: BSA-GA, 1:1.

Figure 6. Molecular modeling of HSA and BSA in presence of GA (A: HSA-GA; B: BSA-GA).The proteins are shown in cartoon while the key residues are shown in stick. GA is shown in cyan stick. Hydrogen bonds are shown in magenta dashes with the label of length.

Table 1. Stern-Volmer quenching constants for the interactions of GA with HSA/BSA at different temperatures. KSV ± SDb Kq R2 (×105 M−1) (×1013 M−1 s−1) 298 4.61 ± 0.02 8.30 0.994 HSA-GA 307 6.51 ± 0.03 11.73 0.989 316 9.96 ± 0.04 17.95 0.989 298 4.54 ± 0.02 7.57 0.989 BSA-GA 307 4.98 ± 0.01 8.98 0.997 316 7.18 ± 0.02 11.98 0.993 2 b R is the correlation coefficient; SD is the standard deviation for the KSV values. System

T (K)

Table 2. Binding numbers of GA to HSA/BSA. T (K) n ± SDb R2 298 0.79 ± 0.03 0.988 HSA-GA 307 0.99 ± 0.03 0.984 316 1.00 ± 0.04 0.997 298 0.89 ± 0.03 0.991 BSA-GA 307 0.96 ± 0.02 0.990 316 0.99 ± 0.05 0.995 2 b R is the correlation coefficient; SD is the standard deviation for the n values. System

Table 3. Thermodynamic parameters of HSA-GA and BSA-GA binding systems at different temperatures. System HSA-GA

BSA-GA

T (K)

R2

298 307 316 298 307 316

0.988 0.994 0.991 0.990 0.998 0.984

Ka ΔH ΔG (×105 M−1) (kJ mol-1) (kJ mol-1) 7.80 -33.24 1.34 -59.21 -32.46 2.02 -31.67 5.59 -32.36 1.04 -59.88 -31.52 1.43 -30.69 R2 is the correlation coefficient.

ΔS (J mol-1 K-1) -87.14

-92.36

Highlights



GA-HSA/BSA interactions in vitro were studied by multispectroscopy techniques.



GA induces the conformational changes of HSA/BSA.



GA binds to the site II of HSA/BSA via hydrogen bond and π-cation interactions.