InhIbItory potentIal of polyhydroxylated fullerenes agaInst proteIn tyrosIne phosphatase 1 b

Inhibition of PTP1B by polyhydroxylated fullerenes was studied in silico and in vitro. The enzyme kinetics in the presence of polyhydroxy small gap fullerenes showed that reciprocal value of maximum velocity non-linearly increases with increasing the inhibitor concentration. analysis of the dose-dependent curve of PTP1B inhibition suggests an apparent positive cooperativity with involvement of at least two binding sites for the hydroxylated fullerene cages. Molecular docking calculations indicated that highly hydroxylated fullerene c60 may occupy the active site and additional allosteric binding site with similar affinity. In silico analysis of a number of fullerenols with 6, 12, 18, 24, 30, and 36 hydroxyl groups showed that the inhibitory activity may depend on the degree of hydroxylation of the nanoparticles surface. These data provide some understanding of the mechanisms of inhibitory action of fullerenols on activity of protein tyrosine phosphatases.

Inhibition of PTP1B by polyhydroxylated fullerenes was studied in silico and in vitro.The enzyme kinetics in the presence of polyhydroxy small gap fullerenes showed that reciprocal value of maximum velocity non-linearly increases with increasing the inhibitor concentration.analysis of the dose-dependent curve of PTP1B inhibition suggests an apparent positive cooperativity with involvement of at least two binding sites for the hydroxylated fullerene cages.Molecular docking calculations indicated that highly hydroxylated fullerene c 60 may occupy the active site and additional allosteric binding site with similar affinity.In silico analysis of a number of fullerenols with 6, 12, 18, 24, 30, and 36 hydroxyl groups showed that the inhibitory activity may depend on the degree of hydroxylation of the nanoparticles surface.These data provide some understanding of the mechanisms of inhibitory action of fullerenols on activity of protein tyrosine phosphatases.K e y w o r d s: protein tyrosine phosphatase, PTP1B, fullerenes, fullerenols.P olyhydroxylated fullerene derivatives (fullere nols) are characterized by promising biologi cal properties including immunomodulating, anticancer [13] and antimetastatic activities [4].The ability of fullerenols to neutralize free radicals is as sociated with their possible use as antioxidants and radioprotective agents in living cells [5,6].In addi tion, fullerenols are able to inhibit the activity of enzymes such as P450dependent monooxygenase, Mg 2+ -ATPase [7], М-MuLV reverse transcriptase [8], lysozyme [9], and monoamine oxidase B [10].
We have previously reported that fullerenol and some other fullerene derivatives can also be effective as inhibitors of human protein tyrosine phosphatases (EC 3.1.3.48)[11][12][13].These regulatory enzymes catalyze dephosphorylation of tyrosine residues in signaling proteins, and abnormal changes in their ac tivities may lead to development of serious diseases [14].Some receptor and nonreceptor protein tyrosi ne phosphatases are involved in the regulation of in sulin signaling.Among them, protein tyrosine phos phatase 1B (PTP1B) provides dephosphorylation of the insulin receptor tyrosine residues contribu ting to type 2 diabetes risk [15,16].Negative role of PTP1B in obesity is associated with the blocking of transduction signal from the leptin hormone via dephosphorylation of Jak2 [17].Furthermore, the en zyme may act as a promoter or suppressor of tumor develop ment [18].
In this context, search for effective PTP1B inhibitors is essential for drug design [19], and the polyhydroxylated fullerenes can be considered as potential therapeutic agents exhibiting significantly lower toxicity in comparison with the unmodified ones [2022].The aim of the present report was to elucidate some characteristics of polyhydroxy small gap fullerenes as inhibitors of PTP1B in vitro.In ad dition, molecular docking was performed to predict the inhibitory potential of the nanoparticles with dif ferent number of hydroxyl groups against PTP1B as a favorable target.

Materials and Methods
PTP1B, pnitrophenyl phosphate ( pNPP), 4methylumbelliferyl phosphate disodium salt (4MUP), ethylenediamine tetraacetate disodium salt (EDTA) and dithiotreitol (DTT) were pur chased from SigmaAldrich.Preparation of poly hydroxy small gap fullerenes (SigmaAldrich) was used without additional purification as a representa tive fullerenol with about thirty oxy functionali ties per one fullerene core of the common formula C 120 O 30 (OH) 30 For studying the inhibition of PTP1B in vitro, a typical reaction mixture consisted of 0.05 M bis Tris buffer (рн 7.2), 100 mM NaCl, 3 mM EDTA, 1 mM DTT, 1% dimethyl sulphoxide, recombinant PTP1B 1436 (1436 amino acid residues) and pNPP as an artificial substrate.In some of kinetic experi ments, fluorogenic substrate 4-MUP was employed to analyze the inhibition of PTP1B 1322 .The mixture was incubated during 5 min at 37 °с.The reac tion was started by adding the enzyme (in the case of pNPP) or substrate (in the case of 4MUP).The enzymatic hydrolysis of pNPP was monitored spectrophotometrically at 410 nm by watching the accumulation of pnitrophenol.The rate of 4MUP hydrolysis was observed due to increasing concen trations of fluorescent product, 4-methylumbelliferone (excitation wavelength and emission wavelength were 360 and 450 nm, respectively).Hill slope was calculated using fourparameter equation from the dosedependent inhibition curve and shown as the mean value ± standart error.Apparent IC 50 value was the inhibitor concentration that inhibits the en zyme activity by 50% and is given as the mean value ± standart deviation.Kinetic data of pNPP and 4MUP enzymatic transformations were analyzed using Lineweaver-Burk double reciprocal plots.The kinetic and statistical calculations were carried out using MS Excel software.
Computersimulation studies of PTP1B (struc tures with PDB identification codes 1NL9 and 2CM8) were performed by Autodock Vina [23].Ad ditional calculations with Autodock 4.2 showed the similar positions of the hydroxylated fullerenes on the surface of PTP1B.Energetically favorable struc tures [24,25] of the fullerene ligands were optimized by Avogadro software [26] using the MMFF94s force field.

results and discussion
Polyhydroxy small gap fullerenes with OH and ONa groups covalently attached to the nanoscale car bon core were used as a watersoluble polyhydroxy fullerene example [27] to inhibit PTP1B in vitro.The previous studies [12] showed that the polyhydroxy lated fullerenes can display strong inhibitory effects towards this enzyme.In this paper, the two artificial substrates, pNPP and 4MUP, were applied to eluci date the kinetic characteristics of the inhibition.
Progress curves of pNPP hydrolysis cata lyzed by PTP1B in the presence of polyhydroxylated fullerenes (Fig. 1) showed a timedependent decrease in the reaction rate.Reaching a steadystate veloci ty was observed after about the first five minutes of pnitrophenol accumulation.Taking into account the slowbinding inhibition, the steadystate kinetics was used to analyze the enzyme inhibition.
The dosedependent curve of PTP1B suppres sion by the polyhydroxylated fullerenes is presented in Fig. 2.Under assay conditions, the fullere nols have a high inhibitory potential with activity at concentration from 0.01 to 1 µg/ml (based on rep resentative fullerene derivative C 60 (OH) 30 , molar concentrations of the inhibitor may be given in range from 5 to 500 nM).The apparent IC 50 value was 0.114 ± 0.033 µg/ml that is in agreement with •min B previously reported data [12].The hill coefficient value was found to be 2.20 ± 0.51.This suggests an apparent positive cooperative effect with involve ment of at least two binding sites of PTP1B for the hydroxylated fullerene cages.
The plots of the steadystate velocity against the concentration of the substrate (1/V versus 1/[S]) were linear (Fig. 3, a, the values of r 2 are 0.98-0.99),however, the graphical analysis does not allow iden tifying the type of inhibition.
Similar results were obtained with 4MUP as a substrate (Fig. 3, B, the values of r 2 are 0.99).The observed kinetic effects of the polyhydroxylated fullere nes at concentration 0.1 µg/ml may be cor related with selfassembly of the nanoparticles in the aqueous phase [28,29].At the same time, analysis of

Scheme. Possible kinetic scheme of PTP1B inhibition by polyhydroxylated fullerenes
the secondary replots of slope and intercept versus the inhibitor concentration (Fig. 4) shows that value of the intercept (1/V max ) increases nonlinearly with increasing the inhibitor concentration but dependen ce of the slope on the inhibitor concentration is lin ear, which can be characteristic of Slinear Ipara bolic mechanism of inhibition [30,31].The parabolic character of the dependence of 1/V max on the inhibi tor concentration is in accordance with interaction of fullerenol molecules with two binding sites of PTP1B .According to this suggestion, polyhydroxy lated fullerenes interact with the enzyme to form EI, IEI, and IES inactive complexes (Scheme).Optimization of ligand structures shows that at tachment of hydroxyl groups to C 60 leads to increase of molecular volume of the inhibitor, reduces aroma ticity and consequently results in a distorted fullere ne cage [24,25].Docking modeling of enzymein hibitor complexes (PDB code 1NL9) indicated that two wide cavities on PTP1B surface that are compa rable to the size of the molecule of C 60 (OH) 30 can be formed by the active site region and allosteric inhibi tor binding site (Fig. 5).Docking binding energies were calculated to be -8.2 kcal/mol and -7.7 kcal/ експериМентальні роботи mol, respectively.Location of the inhibitor molecule in the active site is characterized by the presence of several hydrogen bonds and electrostatic interac tions with Asp48, Glu115, Lys116, Lys120, Asp181, Ala217, ser216, Gly220, Arg221, and Gln266 (Fig. 5,  a).Additional forces that stabilize the enzymein hibitor complex are π-π and π-cationic interaction with Tyr46 and Lys120, respectively.Distance to catalytically important Cys215 is approximately of 3.72 Å.However, the molecule of the hydroxylated fullerene can not occupy the active site of PTP1B with closed WPD-loop (PDB code 2CM8).In this case (as well as in case of PTP1B with open WPD loop), the fullerenol may be located at allosteric site at the distance of about 20 Å from the catalytically important Cys215 [32].The molecule of C 60 (OH) 30 occupies the region formed by helices α3, α4, α6, α7, and Sloop (Fig. 5, B) that are involved in conforma tional changes of PTP1B [3234].The inhibitor forms Hbonds and electrostatic interaction with Arg79, Arg199, Glu200, Gly202, ser205, Glu207, Asp236, and Gln288.
The results of this docking revealed molecular fragments of representative hydroxylated fullerene C 60 that are responsible for the interaction with the PTP1B.It should be noted that polyhydroxy small gap fullerene product is heterogeneous and rich in giant fullerenes larger than C 60 .Hence their com plexes with the enzyme also may be suggested in cluding, for example, the interaction with dimeric fullerene derivatives.
Molecular docking simulations of polyhydro xylated fullerene derivatives C 60 (OH) 6 , C 60 (OH) 12 , C 60 (OH) 18 , C 60 (OH) 24 , C 60 (OH) 30 , C 60 (OH) 36 as inhibit ors of PTP1B (Fig. 6) showed that the inhibi tory potential of these compounds depends on de gree of hydroxylation of the fullerene.The efficiency of the inhibitors decreases with increasing the num ber of hydroxyl groups and remains almost stable in case of the fullerenols containing more than 18 OH groups.More hydrophilic nanoparticles bind to active and allosteric sites with approximately equal binding energies.In contrast, the fullerene deriva tives with a lower degree of hydroxylation form more stable enzymeinhibitor complexes interacting with hydrophobic residues at the active site of PTP1B .This is consistent with the fact that pyr rolidino[60]fullere ne triscarboxylic acid inhibits the activity of PTP1B in a competitive manner [12], whereas the polyhydroxy lated fullerenes show the other inhibition types.It should also be noted that some of low functionalized fullerene derivatives ex hibited selective inhibition of receptortype protein tyrosine phosphatase CD45 (in contrary to PTP1B) which may be explained by significant hydrophobic ity of CD45 surface [12,13].
In conclusion, the results of kinetic studies showed that polyhydroxylated fullerenes may inhibit activity of PTP1B with positive cooperativity.Com puter docking simulation indicated that C 60 (OH) 30 particles occupy the substrate binding site and al losteric inhibitor binding site of the enzyme with similar docking energies.The inhibition of PTP1B may depend on the degree of fullerene surface hy droxylation.These data provide some basis for un derstanding the mechanisms of inhibitory action of fullerenols on activity of PTP1B and other protein tyrosine phosphatases.

Fig. 4 .
Fig. 4. The dependence of the slope and intercept (data from Fig. 3, B) on the inhibitor concentration
Fig. 6.Molecular docking energies of various polyhydroxylated fullerene derivatives bound to the PTP1B active site (PDB code 1NL9) and allosteric site (PDB codes 1NL9 and 2cM8)к