Biochemical Pharmacology 92 (2014) 506–516

Contents lists available at ScienceDirect

Biochemical Pharmacology journal homepage: www.elsevier.com/locate/biochempharm

In vitro inhibition of lysine decarboxylase activity by organophosphate esters Sufang Wang, Bin Wan, Lianying Zhang, Yu Yang *, Liang-Hong Guo * State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, P.O. Box 2871, 18 Shuangqing Road, Beijing 100085, China

A R T I C L E I N F O

A B S T R A C T

Article history: Received 20 July 2014 Received in revised form 18 September 2014 Accepted 18 September 2014 Available online 28 September 2014

Organophosphate esters (OPEs), a major group of organophosphorus flame retardants, are regarded as emerging environmental contaminants of health concern. Amino acid decarboxylases catalyze the conversion of amino acids into polyamines that are essential for cell proliferation, hypertrophy and tissue growth. In this paper, inhibitory effect of twelve OPEs with aromatic, alkyl or chlorinated alkyl substituents on the activity of lysine decarboxylase (LDC) was assessed quantitatively with an economic and label-free fluorescence sensor and cell assay. The sensor comprises a macrocyclic host (cucurbit[7]uril) and a fluorescent dye (acridine orange) reporter. The twelve OPEs were found to vary in their capacity to inhibit LDC activity. Alkyl group substituted OPEs had no inhibitory effect. By contrast, six OPEs substituted with aromatic or chlorinated alkyl groups inhibited LDC activity significantly with IC50 ranging from 1.32 mM to 9.07 mM. Among them, the inhibitory effect of tri-mcresyl phosphate (TCrP) was even more effective as an inhibitor than guanosine 50 -diphosphate-30 diphosphate (ppGpp) (1.60 mM), an LDC natural inhibitor in vivo. Moreover, at non-cytotoxic concentrations, these six OPEs showed perceptible inhibitory effects on LDC activity in PC12 living cells, and led to a marked loss in the cadaverine content. Molecular docking analysis of the LDC/OPE complexes revealed that different binding modes contribute to the difference in their inhibitory effect. Our finding suggested that LDC, as a new potential biological target of OPEs, might be implicated in toxicological and pathogenic mechanism of OPEs. ß 2014 Elsevier Inc. All rights reserved.

Keywords: Organophosphate esters Toxicity Lysine decarboxylase Inhibition Molecular docking

1. Introduction Due to the ongoing worldwide phase-out and restriction of the brominated flame retardants (BFRs) since the early 2000s and the recent banning of polybrominated diphenyl ethers by the Stockholm Convention on persistent organic pollutants, the production and application of organophosphate esters (OPEs) as substitutes for BFRs has increased dramatically [1–3]. In 2004, OPEs accounted for 14% of the global production volume of flame retardants, compared to 21% for BFRs. OPEs are man-made industrial chemicals used as flame retardants, plasticizers, antifoaming agents and additives in various household and industrial products such as building and insulation materials, textiles, furniture, floor polishes, paints, lubricants, hydraulic fluids, cables and electronics [4–6]. OPEs are the derivatives of phosphoric acid with different substitutes including alkyl chains (e.g., tri-n-butyl phosphate (TnBP)), partly halogenated alkyl chains (e.g., tri(2-chloroethyl)

* Corresponding authors. E-mail addresses: [email protected] (Y. Yang), [email protected] (L.-H. Guo). http://dx.doi.org/10.1016/j.bcp.2014.09.011 0006-2952/ß 2014 Elsevier Inc. All rights reserved.

phosphate (TCEP)) as well as aromatic contents (e.g., triphenyl phosphate (TPhP)) [7–9]. Similar to BFRs, OPEs are not chemically bound in the fire-proofed material and can easily leach into the environment via volatilization, abrasion and dissolution. As a result, OPEs have been frequently detected in the environment, mainly in the aquatic environment (e.g., river, groundwater, surface water, drinking water, wastewater) (0.015 ng L1– 24 mg L1) [10,11], indoor air (0.05 ng m3–730 mg m3) [12,13] and dust (0.04–1800 mg g1) [6,14], sediment (0.05– 24,000 mg kg1) [12,15] and marine coastal biota (0.025– 810 mg kg1) [15,16]. Especially, the concentration of OPEs detected in house dust is comparable, or in some cases exceeds concentrations of BFRs. OPEs have been reported to induce various toxic effects including skin irritation, neurotoxicity, reproductive toxicity, carcinogenicity and developmental toxicity [4]. In some previous studies, esterases, such as the neuropathy target receptors, have been identified as important biological targets of OPEs. Animal data showed that some OPEs including tributoxyethyl phosphate (TBEP), TCEP, TPhP and tri-m-cresyl phosphate (TCrP) caused the red blood cell cholinesterase activity in rats to decrease

S. Wang et al. / Biochemical Pharmacology 92 (2014) 506–516

significantly after long-term exposure [17]. TCrP and TPhP remarkably inhibited acetyl cholinesterase (AChE) activity in cholinergic nerve synapses by phosphorylating a serine hydroxyl group at the active site of the enzyme, which resulted in the accumulation of neurotransmitter acetylcholine in nicotinic and muscarinic receptor and subsequently caused the impairment of neurological and neurobehavioral functions [18]. In vitro experiments revealed that five arylphosphates including diphenyl phosphate, tribenzyl phosphate, tricyclohexyl phosphate, diphenyl methyl phosphate and TPhP were relatively effective inhibitors of human monocyte carboxylesterase (CBE) activity, but the alkylphosphates (tributyl phosphate and triethyl phosphate) had no inhibitory activity [19]. In addition, TPhP has been shown to bind with androgen receptor (AR) by using a radiolabeled competitive binding assay [20]. Amino acid decarboxylases, which catalyze the conversion of amino acids (e.g. lysine, arginine, histidine, ornithine) into polyamines (e.g. cadaverine, agmatine, histamine, putrescine), play dual roles in acid resistance and the synthesis of polyamines [21]. Polyamines, as naturally occurring organic cations, are found in plants, animals and microbes. Early studies demonstrated that polyamines (putrescine, spermidine and cadaverine) are essential for optimal growth and viability [22,23]. Some polyamines such as spermidine and hypusine are requirement for cell proliferation in eukaryotic cells [24,25]. Recently, this paradigm has extended to mammals, with the observation that polyamine is essential for tissue development and growth. For example, polyamines are necessary for blood-vessel development occurring in response to damage to normal tissues or tumor growth [26]. In addition, Polyamine levels have also been associated with the normal growth and hypertrophy of several tissues-including skin, breast, kidney and heart-in rodents [27]. The role of polyamines in tissue repair might be to facilitate tissue remodelling, as has been reported for certain types of lung damage [28]. Polyamines have also been implicated in the development and function of both male and female reproductive organs [29]. Other studies indicated that polyamine-synthesis inhibitors disrupted intestinal development in mice. Inhibition of polyamine synthesis also suppressed wound healing and decreased hormone responsivity in rodents [30,31]. In addition, some diseases including cancers are closely associated with the abnormal polyamine level [32]. Given the important biological functions of polyamines, inhibition of the enzymes responsible for their in vivo synthesis, i.e., amino acid decarboxylases by exogenous chemicals is worth investigating. In this work, we investigated the inhibitory activity of twelve structurally diverse OPEs on lysine decarboxylase (LDC), an amino acid decarboxylase, using a fluorescence-based enzyme activity assay developed in our laboratory. These OPEs carry different substituents including alkyl, chlorinated alkyl and aromatic groups. By combining the fluorescence sensing assay, in vitro cell experiments and molecular docking, the structural requirements for the inhibition of LDC by OPEs, as well as LDC as a possible cellular target of OPEs, were investigated and assessed in detail.

507

2-ethylhexyl diphenyl phosphate (EHDPP) and tris(2-ethylhexyl) phosphate (TEHP) were purchased from Dr. Ehrenstorfer Gmbh (Germany) (Fig. 1). Benzoyl chloride was from TCI (Tokyo, Japan), and BCA protein assay kit was from ComWin Biotech (Beijing, China). HPLC grade acetonitrile was purchased from J.T. Baker (Phillipsburg, NJ, USA). Sodium dodecyl sulfate (SDS), glycerol and 2-mercaptoethanol were acquired from Amresco (Ohio, USA). NH4OAc, Tris–HCl, HCl and K2CO3 were all obtained from Sinopharm Chemical Reagent Beijing Co. Ltd (Beijing, China). De-ionized water (18.2 MV cm) obtained from a Millipore ultrapure water system (Millipore, Bedford, USA) was used throughout the experiment. 2.2. Fluorescence enzyme activity assay In our work, a self-assembled host-guest inclusion complex was employed as fluorescence sensor for the enzyme activity assay of LDC. The complex was composed of macrocyclic host CB7 and fluorescent dye AO. Competitive binding of enzyme product cadaverine with CB7 displaces AO from CB7, leading to reduced fluorescence signal. If LDC activity is inhibited, the fluorescence intensity would remain unchanged (Fig. 2). Steady-state fluorescence was measured on a Horiba Fluoromax-4 spectrofluorimeter (Edison, NJ, USA). The excitation and emission wavelength was 485 nm and 510 nm, respectively. Excitation and emission slits were all set at 3 nm. First of all, the binding interaction of AO with CB7 was investigated by successive addition of a known amount of CB7 into 0.5 mM AO solution. With the optimized CB7/AO concentration, fluorescence displacement measurements were carried out to examine the binding affinity of lysine (LDC substrate) and cadaverine (LDC product) with CB7. It was performed by successive addition of a known amount of lysine or cadaverine into a mixed solution of 0.5 mM AO and 5 mM CB7. In the LDC activity assay, an LDC stock solution was prepared by dissolving 500 mg solid in 1 mL HCl–NH4OAc buffer (10 mM, pH 6.0). Enzyme reaction was performed in an Eppendorf tube containing a 100 mL solution of 8.0 mg mL1 LDC, 50 mM lysine, 0.5 mM AO, and 5 mM CB7, and the temperature was thermostatically controlled at 37.0  0.1 8C. The reaction was monitored by recording AO fluorescence emission spectrum with time. For the LDC inhibition assay, 8.0 mg mL1 LDC and different concentrations of OPEs were mixed in HCl-NH4OAc buffer (pH 6.0) and incubated for 5 h at 37.0  0.1 8C. Then, 0.5 mM AO, 5 mM CB7 and 50 mM lysine were added. Fluorescence spectra were recorded after reacting for 1.5 h. The change of fluorescence intensity versus time was taken as a relative reaction rate, and was plotted as a function of inhibitor concentration. The dose-response curve was fitted with a sigmoidal model (Origin Lab 8.0, Northampton, MA, USA) and analyzed with the Hill equation to obtain IC50 value [33]. The IC50 can be readily converted into the inhibition constants Ki by considering the enzyme concentration [34]: IC 50 ¼ K i þ

1 ½E 2

2. Materials and methods 2.3. Cell culture, treatment and viability test 2.1. Chemicals and materials Lysine, cadaverine, acridine orange (AO), cucurbit[7]uril (CB7), lysine decarboxylase (LDC, from Bacterium cadaveris), 2,4,6trinitrebenzenesulfonic acid (TNBS), and 1,7-diaminoheptane (DAH) were purchased from Sigma–Aldrich (St. Louis MO, USA). Guanosine 50 -diphosphate, 30 -diphosphate (ppGpp) was obtained from Trilink BioTechnologies (San Diego, CA). Trimethyl phosphate (TMP), triethyl phosphate (TEP), TCEP, tri-n-propyl phosphate (TPrP), tris(2-chloroisopropyl)phosphate (TCPP), tri(2-chloro-1(chloromethyl)ethyl) phosphate (TDCP), TPhP, TnBP, TBEP, TCrP,

Pheochromocytoma PC12 cells, a cell line derived from rat adrenal medulla, were obtained from ATCC (Manassas, VA, USA), and maintained in Dulbecco’s modified Eagle’s medium supplemented with 6% fetal calf serum, 6% horse serum, 100 units/mL penicillin and 100 mg/mL streptomycin in a humidified CO2 incubator at 37 8C. All culture reagents were purchased from Invitrogen Corporation (Carlsbad, CA, USA). Approximated 0.5– 1  105 cells mL1 in 96-well plates was seeded for cell viability assay. The exposure was performed in DMEM with or without (DMSO control) OPEs (up to 250 mM), respectively. The change of

508

S. Wang et al. / Biochemical Pharmacology 92 (2014) 506–516

Fig. 1. Structures of the chemicals used in the present study.

cell viability after 24 h OPEs exposure was determined by cell proliferation reagent WST-1 (Roche Applied Science, Penzberg, Germany). In brief, after the exposure, WST-1 reagents (1:10 dilution) were added into each well of the 96-plate, and then incubated at 37 8C for 4 h. The absorbance were recorded on a Varioskan Flash microplate reader (Thermo Fisher Scientific, Waltham, MA, USA) at 440 nm, and the absorbance at 600 nm was used as reference. 2.4. LDC activity assay in cells To further verify the effects of OPEs on LDC enzymatic activity in living cells, changes in LDC activity in PC12 cell after OPEs exposure were investigated by spectrophotometric method [35]. The basic principle is that cadaverine, a LDC product, readily reacts with TNBS to give an intensely colored adduct with higher molar extinction coefficient and soluble in toluene whereas lysine does not. After 24 h OPEs exposure, the cells were lysed in a buffer

containing 0.125 M Tris–HCl (pH 6.8), 4% SDS, 20% glycerol, 2% 2mercaptoethanol, and centrifuged to obtain the supernatant. The protein content in the lysate was quantified by BCA protein assay kit. Appropriate amounts of cell extracts were incubated with lysine for 1 h in 500 mM acetate buffer (pH 6.0). To each sample, K2CO3 and TNBS were added and the samples were incubated for 5 min at 40 8C and thoroughly mixed with toluene and then separated. LDC activity was determined spectrophotometrically by measuring the concentration of N,N0 -bistrinitrophenylcadaverine (TNP-cadaverine) extracted into the organic phase. UV–visible absorption measurements at 340 nm were performed on an Agilent 8453 UV-vis spectrophotometer (Santa Clara, CA, USA). 2.5. Variations of cadaverine content in exposed cells We detected the cadaverine content in control and exposed cell by high-performance liquid chromatography (HPLC) described previously [36,37]. After 24 h exposure, cells were lysed in buffer

S. Wang et al. / Biochemical Pharmacology 92 (2014) 506–516

509

Fig. 2. Design of a label-free fluorescence displacement method for the assay of inhibitory effect of OPEs on LDC activity.

and separated by centrifuging the tubes at 8000  g for 20 min. The supernatant were derivatized with benzoyl chloride and then injected into an Eclipse plus C18 column (50 mm  3 mm; 20/80 water/acetonitrile; 0.1 mL/min, 25 8C) with UV detector (254 nm) (Agilent 1260, Hewlett Packard, Wilmington, NC, USA). The amount of cadaverine was measured in ng/mg protein using DAH as the internal standard. 2.6. Molecular docking The 3D crystal structure of LDC was provided by Protein Data Bank for constructing the receptor model [38]. All molecular simulation studies on the structural and conformational feature of OPEs/LDC complexes were performed using Lamarckian genetic algorithm provided by AutoDock 4.2 software [39,40]. Grid boxes were built around the protein with 60 points cube coverage. A spacing of 0.375 A˚ between the grid points was used. All other docking parameters were set to defaults, including a population size of 150, maximum number of 2.5 million evaluations, maximum of 2700 generations, gene mutation rate of 0.02, crossover rate of 0.8 and 10 GA runs. Ten docked conformations for each ligand were scored according to a free energy cost function (DG*), and the lowest-energy conformation was selected as the most likely bioactive conformation.

peak intensity increased by 3-fold and the maximum emission wavelength of AO shifted towards shorter wavelength 510 nm (Fig. 3). The fluorescence intensity at 525 nm relative to the one without the CB7 addition was plotted as a function of the total CB7 concentration. Using the modified Benesi-Hildebrand equation [41], the association constant of AO with CB7 was calculated to be 1.50  105 M1, which is agreeable with the reported value (2.90  105 M1) [42]. Based on the fluorescence titration curve, the concentrations of CB7 and AO were chosen to be 5 mM and 0.5 mM respectively. To be effective in LDC activity assay, its substrate and product must demonstrate differential capability in displacing AO from CB7 and thereby changing the fluorescence signal. To compare their capability, LDC substrate lysine and product cadaverine were titrated separately into a solution containing 5 mM CB7 and 0.5 mM AO. As depicted in Fig. 4, after the addition of cadaverine, the fluorescence intensity decreased significantly. However, no substantial change in the spectra was observed when lysine was

2.7. Statistical analysis We analyzed p values of the experimental data by a two-tailed paired T-test using the Microsoft Excel software. A p value of less than 0.05 was considered statistically significant. All the experiments were repeated for three times and data points represent the mean  S.D. of three individual experiments. 3. Results 3.1. Fluorescence enzyme activity assay Our method is based on the competitive encapsulation of a fluorescent dye and an enzymatic product by macrocyclic host to monitor enzymatic reaction. We first optimized the concentration of CB7/AO pair. When excited with 485 nm light, free AO displayed a fluorescence emission peak at 525 nm. After addition of CB7, the

Fig. 3. Fluorescence spectra of AO (0.5 mM) with increasing CB7 concentration at (1) 0, (2) 2.5, (3) 5, (4) 7.5, (5) 10, (6) 12.5, (7) 15, (8) 17.5, and (9) 20 mM in 10 mM NH4OAc buffer (pH 6.0) at 37 8C. Inset: Plot of the relative fluorescence against CB7 concentration. I and I0 are the fluorescence intensity values of AO with and without the addition of the CB7.

510

S. Wang et al. / Biochemical Pharmacology 92 (2014) 506–516

plateau when the time exceeded 5 h. Thus, we chose 5 h as the optimal incubation time. In the inhibition assay, the six OPEs substituted with alkyl chains did not show any effect on the activity of LDC. By comparison, remarkable inhibitory effect was observed for the other six OPEs substituted with chlorinated alkyl chains or aromatic groups. A dose-response curve similar to ppGpp was obtained with TCrP and the other five OPEs (Fig. 7). The IC50 values and inhibition constants Ki for LDC are summarized in Table 1. The inhibitory effect follows the order of TCrP > TPhP > TDCP > EHDPP > TCEP  TCPP. Among them, TCrP, an OPE substituted with cresyl, showed the strongest inhibitory effect on LDC with a Ki of 1.27 mM, which is even stronger than ppGpp. Besides, the significant decrease of LDC reaction rate was observed at 0.01 mM TCrP exposure (**p < 0.01) and the inhibition was reversible. 3.3. LDC activity assay in cells after OPEs exposure Fig. 4. Plots of the relative fluorescence intensity of AO against the concentration of lysine and cadaverine in fluorescence competitive displacement experiments with addition of lysine or cadaverine into a solution of 0.5 mM AO and 5 mM CB7 in 10 mM NH4OAc buffer, pH 6.0 at 37 8C. I and I0 are the fluorescence intensity values of CB7/AO with and without the addition of the cadaverine (or lysine).

added. The relative fluorescence intensity at 510 nm was plotted as a function of the competitor concentration. The association constants of CB7 with lysine and cadaverine were estimated to be 1.01  103 M1 and 4.94  106 M1, which were in good accordance with previous reports, 8.70  102 M1 for lysine and 1.40  107 M1 for cadaverine [43]. Obviously, cadaverine has much higher affinity with CB7 than lysine. Based on the above results, a CB7/AO system for real-time monitoring of LDC activity was established. Firstly, the concentrations of LDC and substrate lysine were optimized (Fig. 5). The optimal concentrations of LDC and lysine were determined to be 8.0 mg mL1 and 50 mM. The limit of detection for LDC is 0.25 mg mL1 (signal-to-noise ratio of 3), which is much lower than the values in the CB7/Dapoxyl and CX4/DBO systems [43]. After addition of 8.0 mg mL1 LDC and 50 mM lysine into a solution of 0.5 mM AO and 5 mM CB7, the fluorescence peak redshifted gradually to 525 nm, and the intensity decreased with time until it reached the background (Fig. 6). This reflects timedependent enzymatic conversion of lysine to cadaverine.

PC12 cells were selected to further investigate the effect of OPEs on LDC activity in living cells. In order to obtain non-cytotoxic concentrations, we first examined the effect of OPEs on cell viability. WST-1 assay was employed to detect the cell viability of PC12 cells after 24 h exposure to OPEs substituted with aromatic or chlorinated alkyl groups (0–250 mM). We found that cell viability significantly decreased about 160 mM when exposed to TCrP, TPhP, EHDPP or TDCP (Fig. 9), whereas it did not change appreciably after TCEP or TCPP exposure. The order of cytotoxicity is TDCP > EHDPP > TPhP > TCrP > TCEP  TCPP. Then, LDC activity was measured in cells under the exposure of non-cytotoxic concentrations of TCrP, TPhP, TDCP, TCEP, TCPP (0–100 mM), EHDPP (0–50 mM) or control medium (VC). This allowed us to ensure that changes in LDC activity were due to OPE treatment but not a secondary effect of cell death. An appropriate amount of cell lysate containing 30 mg protein was mixed with lysine at 37 8C for 1 h, LDC activity was determined by measuring the cadaverine amount produced enzymatically. As shown in Fig. 10a–c and g–i, six OPEs were all found to inhibit the LDC activity in a concentrationdependent manner. Among them, the significant decrease of LDC activity was observed at 50 mM TCrP exposure (**p < 0.01). And the LDC activity decreased from 1.37 to 0.88 (mM/mg protein). Using normalization against control, the relative LDC activity was a 35.4% decrease. It demonstrates that the strongest inhibitory effect of TCrP on LDC activity was observed in living cells, which further verify the results obtained from the fluorescence assay.

3.2. Inhibitory effect of OPEs on LDC activity 3.4. Cadaverine content in exposed cells Having established the LDC activity assay, inhibition of twelve OPEs on LDC activity was examined. The assay was first verified by reacting with ppGpp, a naturally occurring LDC inhibitor in vivo. The inhibitor was incubated with 8.0 mg mL1 LDC for 5 h, and then added into a solution of 50 mM lysine, 5 mM CB7, and 0.5 mM AO. As ppGpp concentration was increased, the fluorescence intensity increased gradually until it reached a plateau (Fig. 7). The IC50 value was 1.60 mM and Ki was 1.55 mM for ppGpp. The value is close to that reported in the literature (Ki = 0.68 mM) [38], which verifies the effectiveness of this method. In the following experiments, twelve OPEs with different substitution groups were tested for their LDC inhibitory effect, which include six alkyl chains (TMP, TEP, TPrP, TnBP, TBEP and TEHP), three chlorinated alkyl chains (TCEP, TCPP and TDCP), and three aromatic groups (TPhP, TCrP and EHDPP). Addition of an OPE alone into a CB7/AO solution did not induce any fluorescence change, suggesting the chemical does not interfere with the signaling system. Then, we investigated the time-dependent inhibition of LDC by OPEs. As illustrated in Fig. 8, the fluorescence intensity increased with the incubation time, and reached a

We made further investigation on the change of cadaverine content in the cells with or without OPEs exposure. A certain amount of cell lysate containing 2500 mg protein was derivatized with benzoyl chloride, the cadaverine content was then determined by HPLC. Compared with cells treated with control medium, OPEs treated cells had a significant dose-related reduction in the cadaverine content (Fig. 10d–f and j–l). The cadaverine content markedly decreased at 50 mM OPEs exposure (**p < 0.01) including TCrP, TPhP, TDCP, and TCEP (Fig. 10d, e, j and k). And it declined from 11.8 to 8.0 ng/mg protein (TCrP), 9.3 ng/mg protein (TDCP), 9.4 ng/ mg protein (TPhP) and 8.5 ng/mg protein (TCEP), and was lowered by 32.2%, 21.5%, 20.4%, and 27.6%. The change of cadaverine content was most evident in cells exposed to TCrP, which again provide a powerful support for the result from fluorescence assay. 3.5. Molecular docking To better understand the inhibitory effect of OPEs, the binding interaction of OPEs with LDC was simulated by molecular docking.

S. Wang et al. / Biochemical Pharmacology 92 (2014) 506–516

511

Fig. 5. Plots of the relative fluorescence intensity of AO against the concentration of LDC (a) and lysine (b) with addition of lysine or LDC into a solution of 0.5 mM AO and 5 mM CB7 in 10 mM NH4OAc buffer, pH 6.0 at 37 8C.

The X-ray crystal structure of LDC shows that the protein is a decamer composed of an oligomer of five dimers, and the active site is a narrow cleft of the dimer interface [38]. We firstly docked ppGpp to LDC using Autodock software. Results show that ppGpp locates at the interface of the cleft, and the binding pocket comprises three contiguous parts: a hydrophobic portion that interacts with the guanosine base of ppGpp, and two other hydrophobic portions that interact with 30 - and 50 -phosphate respectively (Fig. 11a). Three hydrogen bonds are formed between the phosphate group and the side chain of Arg206 and the backbone amide of Gly418. These results are consistent with the crystallographic structure (Fig. 11a) [38]. By using the same docking parameters, the twelve OPEs were docked with LDC. All the OPEs were found to bind to LDC, and do so at the active site of LDC. But the exact binding geometry is strikingly different. The six OPEs substituted with chlorinated alkyl chains or aromatic groups (TCEP, TCPP, TDCP, TCrP, TPhP and EHDPP) all reside at the upper right of the cleft (Fig. 11b–g). However, for the six OPEs substituted with alkyl chains (TMP, TEP, TPrP, TnBP, TBEP and TEHP), they all reside at the bottom of the cleft (Fig. 11h–m). The binding energy between OPEs and LDC was calculated by Autodock (Table 1), and is from 7.77 to 2.98 kcal/ M. Hydrogen bond interactions exist between the phosphate

Fig. 6. Change of the relative fluorescence intensity of AO with time in LDC activity assay in which 8.0 mg mL1 LDC and 50 mM lysine were added into a solution of 0.5 mM AO and 5 mM CB7 in 10 mM NH4OAc buffer, pH 6.0. Inset: Lineweaver–Burk plot obtained for varying lysine concentrations.

oxygen of OPEs and the side chain of Tyr203 of LDC or Lys417, Arg558 and Arg97, or Asn76 and Lys422, or Arg97 (Fig. 11b–m).

4. Discussion To investigate the inhibitory effect of OPEs on LDC, an enzyme activity assay is needed. Because both the substrate and product of LDC are optically inert, it is difficult to develop colorimetric or fluorescence assays based on conventional methods. In 2007, Henning et al. described a novel concept for the determination of enzyme activity of amino acid decarboxylases, using macrocyclic receptors and fluorescent dyes such as CB7/Dapoxyl and CX4/DBO [43]. This competitive displacement method is simple and convenient. But unfortunately, LDC consumption in these systems is quite high (40.0 mg mL1). Recently, the same group employed the CB7/AO pair for the monitoring of protease activity and inhibition [42]. We therefore tested the possibility of using this system to reduce the enzyme consumption in LDC activity assay and subsequently used it to screen a large number of inhibitors for LDC. Fortunately, this system could successfully reduce the enzyme consumption (8.0 mg mL1) and determine the OPE’s inhibitory effect on LDC activity.

Fig. 7. Dose-response inhibitory curves of ppGpp (black), TCrP (red), TPhP (blue), EHDPP (green), TDCP (magenta), TCEP (dark yellow) and TCPP (navy) in the presence of 8.0 mg mL1 LDC, 50 mM lysine, 0.5 mM AO and 5 mM CB7 in 10 mM NH4OAc buffer, pH 6.0 at 37 8C. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

512

S. Wang et al. / Biochemical Pharmacology 92 (2014) 506–516

Fig. 8. Dependence of the relative fluorescence intensity of AO on the LDC/OPE incubation time. 100 mM TCrP and 8.0 mg mL1 LDC were incubated for different time and then added into a solution of 50 mM lysine, 0.5 mM AO and 5 mM CB7 in 10 mM NH4OAc buffer, pH 6.0 at 37 8C.

For inhibitory assay, we found that OPEs inhibited LDC activity with Ki from ND to 1.55 mM. Based on the relatively large number of OPEs tested in our study, a clear trend of inhibition potency emerged. First, OPEs inhibition potency is clearly associated with substituent species. The distinctive difference in LDC inhibitory effect among OPEs underscores the importance of size matching in ligand/LDC interaction. This interesting structure-effect dependence was also observed in the inhibition of organophosphorus (OP) compounds on human monocyte carboxylesterase activity [19]. In addition, there exists a correlation between inhibition potency and hydrophobicity among aromatic groups and chlorinated alkyl chains substituted OPEs (R2 = 0.86). This suggests that the hydrophobic interactions between OPEs and amino acids residues in the binding pocket of LDC contribute to the stabilization of OPE/enzyme complex. Molecular docking analysis was performed between OPEs and LDC so as to provide further explanation about the structural characteristics of the binding. From the docking results, the exact binding geometry of OPEs is different. The six OPEs substituted with chlorinated alkyl chains or aromatic groups (LDC inhibitors) all reside at the upper right of the cleft (involving Thr127, Ile128, Leu129, Pro130, Pro121, Tyr203 and Lys417 of dimer C), extend into the inner part of the dimer C, and are in contact with the hydrophobic pocket formed by a4, a8, a15 and a16 helices. Especially, one of the hydrophobic groups of an OPE extends inside the hydrophobic space between a4 and a16 of LDC (Fig. 11b–g). In

Fig. 9. Cell viability assays of PC12 cells after 24 h exposure to TCrP (black), TPhP (blue), EHDPP (red) and TDCP (green). Data points represent the mean  S.D. of three individual experiments. Significance was set as *p < 0.05, **p < 0.01. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

addition, there is only one hydrogen bond between the phosphate oxygen of an OPE and the side chain of Tyr203 or Lys417 of LDC. However, for the six OPEs substituted with alkyl chains (TMP, TEP, TPrP, TnBP, TBEP and TEHP, none is LDC inhibitor), a completely different binding geometry was obtained. They all reside at the bottom of the cleft (involving Leu562, Leu564 of dimer A and Gly418, Asn419, Asp95, Arg97, Ser71, Glu75 and Asn76 of dimer C) and stay outside the hydrophobic pocket formed by a4, a8, a15 and a16 helices (Fig. 11h–m). Hydrogen bond interactions exist between the phosphate oxygen of OPEs and the side chain of Arg558 and Arg97 of LDC, or Asn76 and Lys422, or Arg97, depending on the OPE. The fundamental difference in binding geometry between the two groups of OPEs may provide a clue why the six alkyl-substituted OPEs did not inhibit LDC activity. For the six OPE inhibitors, we compared their inhibition strength with their binding potency to LDC. The order of binding energy is TCrP > TPhP > EHDPP > TDCP > TCEP  TCPP, with TCrP substantially larger than the other five OPEs. The inhibitory effect measured in our fluorescence experiments follows the order of TCrP > TPhP > TDCP > EHDPP > TCEP  TCPP. A high correlation exists between the calculated binding energy and experimentally measured inhibitory effect, with the exception of EHDPP. The good correlation suggests that, once an OPE binds to LDC in the correct geometry, its inhibition strength is dictated by its binding affinity with the enzyme. According to the results in Table 1, EHDPP forms a hydrogen bond with the side chain of Lys417 on LDC, whereas the

Table 1 IC50, Inhibition constant Ki and some molecular docking results of OPEs. Compound name

Abbreviation

IC50 (mM)

Ki (mM)

Binding energy (kcal/M)

Hydrogen-bonding interaction

Guanosine 50 -diphosphate, 30 -diphosphate Tri-m-cresyl phosphate Triphenyl phosphate Tri(2-chloro-1-(chloromethyl)ethyl) phosphate 2-Ethylhexyl diphenyl phosphate Tri(2-chloroethyl) phosphate Tris(2-chloroisopropyl)phosphate Tris(2-ethylhexyl) phosphate Tributoxyethyl phosphate Tri-n-butyl phosphate Tri-n-propyl phosphate Triethyl phosphate Trimethyl phosphate

ppGpp TCrP TPhP TDCP EHDPP TCEP TCPP TEHP TBEP TnBP TPrP TEP TMP

1.60 1.32 3.99 5.65 6.40 8.99 9.07 ND ND ND ND ND ND

1.55 1.27 3.94 5.60 6.35 8.94 9.02 ND ND ND ND ND ND

7.77 7.34 6.54 3.80 4.41 3.28 3.47 2.56 1.84 3.38 3.30 3.02 2.98

Arg206, Gly418 Tyr203 Tyr203 Tyr203 Lys417 Tyr203 Tyr203 Arg97 Asn76, Lys422 Arg558, Arg97 Arg558, Arg97 Arg558, Arg97 Arg558, Arg97

ND: not detected.

S. Wang et al. / Biochemical Pharmacology 92 (2014) 506–516

513

Fig. 10. LDC activity (black bar) and cadaverine content (red bar) in PC12 cells exposed to (a, d) TCrP, (b, e) TPhP, (c, f) DHDPP, (g, j) TDCP, (h, k) TCEP and (i, l) TCPP for 24 h. PC12 cells treated with ppGpp (75 mM, 24 h) were used as positive control. Values are expressed as the fold of change (Basal) against the control (without OPEs treatment; set as 1), and in mean  S.D. of three individual measurements. Significance was set as *p < 0.05, **p < 0.01. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

other five OPEs interact with the side chain of Tyr203 by hydrogen bonding. This might be the reason for the exception in the orders given above. WST-1 results revealed that the cytotoxicity of OPEs toward PC12 cells follow the order of TDCP > EHDPP > TPhP > TCrP > TCEP  TCPP. Among them, TDCP, which differs from TCEP and TCPP most notably by the number halogen substituent, showed strongest effect on cell viability. EHDPP, with relatively larger size than other five OPEs showed stronger effect on cell viability. The results of these studies suggest that the differences in the molecular backbone and halogen substitution patterns appear to

be important factor in determining the cyctotoxicity of OPEs. Since the catalytic residues are highly conserved among species, the effects of OPEs on standard LDC catalytic activity may reflect the influence of these chemicals on rat cells. Our results demonstrated that enzyme activity and cadaverine production in PC12 cells were all reduced after OPEs exposure, with TCrP showing the highest potency. TCrP, one of the three isomers (i.e., o-, m-, or p-cresyl) of tricresyl phosphate which have been widely used as plasticizers, plastic softeners, flam-retardants, and jet oil additive in industry [44,45], is able to induce chronic neurological disorders known as OP-induced delayed neurotoxicity (OPIDN) [46]. Inhibition and

514

S. Wang et al. / Biochemical Pharmacology 92 (2014) 506–516

Fig. 11. Graphs for the docked complexes between LDC and ppGpp, TCrP, TPhP, EHDPP, TDCP, TCEP, TCPP, TEHP, TBEP, TnBP, TPrP, TEP and TMP. The structure with the lowest binding energy for each ligand is shown. The OPEs are colored by atom type (carbon in gray, nitrogen in blue, oxygen in red, hydrogen in white and phosphor in orange). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

S. Wang et al. / Biochemical Pharmacology 92 (2014) 506–516

aging of neuropathy target esterase has been proposed as the initial effect of OPs that induce OPDIN [47–49]. Unfortunately, the mechanism responsible for neurotoxicity of OPs has not been elucidated. LDC inhibition may be a new clue for the pathogenic factors. We speculate that other enzymes might be involved in the mediation of such neuropathies. As far as we known, there was no previous study on the inhibitory effect of OPEs on LDC activity. We provide the first evidence showing that OPEs inhibited LDC activity remarkably in PC12 living cells, and subsequently led to a marked loss in cadaverine content. LDC, as we know, is a very important enzyme in most cells. Inhibition of LDC activity would suppress the cadaverine synthesis. Cadaverine is essential for optimal growth and viability of cells. Therefore, we deduce that at non-cytotoxic concentrations, OPEs might affect the function and physiological status of nerve cells and display potential neurotoxicity. But our study does not provide direct association between LDC inhibition and neurotoxicity of OPEs. In conclusion, inhibitory effect of twelve OPEs on LDC activity was investigated by a label-free competitive fluorescence displacement method, cell assay and molecular docking analysis. Six of the OPEs including three chlorinated alkyl chain (TCEP, TCPP and TDCP) and three aromatic groups (TPhP, TCrP and EHDPP) substituted ones exhibited significant inhibitory effect. Moreover, the results at cellular level further confirmed that the LDC activity was inhibited by these six OPEs, leading to significantly reduced cadaverine content in living cells. Molecular docking revealed that the inhibitory effect depended critically on the binding geometry of an OPE with LDC at its active site. Our study indicates that LDC might be a potential biological target for OPEs toxicity in living systems. And LDC inhibition by OPEs may be implicated in the pathogenic mechanisms of OPEs neurotoxicity. Conflict of interest None of the authors have any conflict of interest to declare. Acknowledgements This work was supported by the National Basic Research Program of China (2011CB936001), National Natural Science Foundation of China (21077124, 21377142, 21277158, 21321004), and Chinese Academy of Sciences (XDB14040100, YSW2013A01). References [1] Reemtsma T, Quintana JB, Rodil R, Garcia-Lopez M, Rodriguez I. Organophosphorus flame retardants and plasticizers in water and air I. Occurrence and fate. Trends Anal Chem 2008;27:727–37. [2] Quintana JB, Rodil R, Reemtsma T, Garcia-Lopez M, Rodriguez I. Organophosphorus flame retardants and plasticizers in water and air II. Analytical methodology. Trends Anal Chem 2008;27:904–15. [3] Brandsma SH, de Boer J, Cofino WP, Covaci A, Leonards PEG. Organophosphorus flame-retardant and plasticizer analysis, including recommendations from the first worldwide interlaboratory study. Trends Anal Chem 2013;43:217–28. [4] van der Veen I, de Boer J. Phosphorus flame retardants: properties, production, environmental occurrence, toxicity and analysis. Chemosphere 2012;88: 1119–53. [5] Bergman A, Ryden A, Law RJ, de Boer J, Covaci A, van der Veen I. A novel abbreviation standard for organobromine, organochlorine and organophosphorus flame retardants and some characteristics of the chemicals. Environ Int 2012;49:57–82. [6] Van den Eede N, Dirtu AC, Neels H, Covaci A. Analytical developments and preliminary assessment of human exposure to organophosphate flame retardants from indoor dust. Environ Int 2011;37:454–61. [7] Moller A, Sturm R, Xie ZY, Cai MH, He JF, Ebinghaus R. Organophosphorus flame retardants and plasticizers in airborne particles over the Northern Pacific and Indian Ocean toward the polar regions: evidence for global occurrence. Environ Sci Technol 2012;46:3127–34. [8] Bollmann UE, Moller A, Xie ZY, Ebinghaus R, Einax JW. Occurrence and fate of organophosphorus flame retardants and plasticizers in coastal and marine surface waters. Water Res 2012;46:531–8.

515

[9] Regnery J, Puttmann W. Occurrence and fate of organophosphorus flame retardants and plasticizers in urban and remote surface waters in Germany. Water Res 2010;44:4097–104. [10] Rodriguez I, Calvo F, Quintana JB, Rubi E, Rodil R, Cela R. Suitability of solidphase microextraction for the determination of organophosphate flame retardants and plasticizers in water samples. J Chromatogr A 2006;1108: 158–65. [11] Marklund A, Andersson B, Haglund P. Organophosphorus flame retardants and plasticizers in Swedish sewage treatment plants. Environ Sci Technol 2005;39:7423–9. [12] Norwegian Pollution Control Authority. Screening of selected metals and new organic contaminants 2007. Trondheim; 2008. [13] Marklund A, Andersson B, Haglund P. Organophosphorus flame retardants and plasticizers in air from various indoor environments. J Environ Monit 2005;7:814–9. [14] Stapleton HM, Klosterhaus S, Eagle S, Fuh J, Meeker JD, Blum A, et al. Detection of organophosphate flame retardants in furniture foam and US house dust. Environ Sci Technol 2009;43:7490–5. [15] Norwegian Enviroment Agency. Screening of organophosphor flame retardants 2010. Trondheim; 2011. [16] Sundkvist AM, Olofsson U, Haglund P. Organophosphorus flame retardants and plasticizers in marine and fresh water biota and in human milk. J Environ Monit 2010;12:943–51. [17] Agency for Toxic Substances and Disease Registry. Toxicological profile for phosphate ester flame retardants. Atlanta, Georgia; 2012. [18] Abou-Donia MB. Organophosphorus pesticides. In: Chang LW, Dyer RS, editors. Handbook of toxicology. New York: Marcel Dekker; 1995. p. 419–73. [19] Saboori AM, Lang DM, Newcombe DS. Structural requirements for the inhibition of human monocyte carboxylesterase by organophosphorus compounds. Chem-Bio Interact 1991;80:327–38. [20] Fang H, Tong W, Branham WS, Moland CL, Dial SL, Hong H, et al. Study of 202 natural, synthetic, and environmental chemicals for binding to the androgen receptor. Chem Res Toxicol 2003;16:1338–58. [21] Kanjee U, Gutsche I, Ramachandran S, Houry WA. The enzymatic activities of the Escherichia coli basic aliphatic amino acid decarboxylases exhibit a pH zone of inhibition. Biochemistry 2011;50:9388–98. [22] Tabor H, Hafner EW, Tabor CW. Construction of an Escherichia coli strain unable to synthesize putrescine, spermidine, or cadaverine: characterization of two genes controlling lysine decarboxylase. J Bacteriol 1980;144:952–6. [23] Tabor CW, Tabor H, Tyagi AK, Cohn MS. The biochemistry, genetics, and regulation of polyamine biosynthesis in Saccharomyces cerevisiae. Fed Proc 1982;41:3084–8. [24] Porter CW, Bergeron RJ. Spermidine requirement for cell proliferation in eukaryotic cells: structural specificity and quantitation. Science 1983;219: 1083–5. [25] Park MH, Lee YB, Joe YA. Hypusine is essential for eukaryotic cell proliferation. Biol Signals 1997;6:115–23. [26] Brzozowski T, Konturek SJ, Drozdowicz D, Dembinski A, Stachura J. Healing of chronic gastric ulcerations by L-arginine. Role of nitric oxide, prostaglandins, gastrin and polyamines. Digestion 1995;56:463–71. [27] Cohen SS. A guide to the polyamines. New York: Oxford University Press; 1998. [28] Babal P, Manuel SM, Olson JW, Gillespie MN. Cellular disposition of transported polyamines in hypoxic rat lung and pulmonary arteries. Am J Physiol Lung Cell Mol Physiol 2000;278:L610–7. [29] Ahuja V, Tantry U, Park J, Barbul A. Effect of difluoromethylornitine, a chemotherapeutic agent, on wound healing. J Surg Res 2003;114:308–9. [30] Calandra RS, Rulli SB, Frungieri MB, Suescun MO, Gonzalez-Calvar SI. Polyamines in the male reproductive system. Acta Physiol Pharmacol Ther Latinoam 1996;46:209–22. [31] Potter CL, Sipes IG, Russell DH. Inhibition of ornithine decarboxylase activity by 2,3,7,8-tetrachlorodibenzo-para-dioxin. Biochem Pharmacol 1982;31: 3367–71. [32] Gerner EW, Meyskens FL. Polyamines and cancer: old molecules, new understanding. Nat Rev Cancer 2004;4:781–92. [33] Copeland RA. Mechanistic considerations in high-throughput screening. Anal Biochem 2003;320:1–12. [34] Nau WM, Ghale G, Hennig A, Bakirci H, Bailey DM. Substrate-selective supramolecular tandem assays: monitoring enzyme inhibition of arginase and diamine oxidase by fluorescent dye displacement from calixarane and cucurbituril macrocycles. J Am Chem Soc 2009;131:11558–70. [35] Phan APH, Ngo TT, Lenhoff HM. Spectrophotometric assay for lysine decarboxylase. Anal Biochem 1982;120:193–7. [36] Flores HE, Galston AW. Analysis of polyamines in higher plants by high performance liquid chromatoguaphy. Plant Physiol 1982;69:701–6. [37] Liu XQ, He YM, Ren YS, Cai FQ, Wang YL. Determining the content of polyamine in cell lysis liquid by high performance liquid chromatography. Chin Hosp Pharm J 2010;30:1159–61. [38] Kanjee U, Gutsche I, Alexopoulos E, Zhao BY, EI Bakkouri M, Houry WA. Linkage between the bacterial acid stress and stringent responses: the structure of the inducible lysine decarboxylase. EMBO J 2011;30:931–44. [39] Morris GM, Goodsell DS, Halliday RS, Huey R, Hart WE, Belew RK, et al. Automated docking using a Lamarckian genetic algorithm and an empirical binding free energy function. J Comput Chem 1998;19:1639–62. [40] Morris GM, Huey R, Lindstrom W, Sanner MF, Belew RK, Goodsell DS, et al. AutoDock4 and AutoDockTools4. Automated docking with selective receptor flexibility. J Comput Chem 2009;30:2785–91.

516

S. Wang et al. / Biochemical Pharmacology 92 (2014) 506–516

[41] Catena GC, Bright FV. Thermodynamic study on the effects of beta-cyclodextrin inclusion with anilinonaphthalene sulfonates. Anal Chem 1989;61:905–9. [42] Ghale G, Ramalingam V, Urbach AR, Nau WM. Determining protease substrate selectivity and inhibition by label-free supramolecular tandem enzyme assays. J Am Chem Soc 2011;133:7528–35. [43] Henning A, Bakirci H, Nau WM. Label-free continuous enzyme assays with macrocycle-fluorescent dye complexes. Nat Methods 2007;4:629–32. [44] Craig PH, Barth ML. Evaluation of the hazards of industrial exposure to tricresyl phosphate: a review and interpretation of the literature. J Toxicol Environ Health B Crit Rev 1999;2:281–300.

[45] Winder C, Balouet JC. The toxicity of commercial jet oils. Environ Res 2002;89:146–64. [46] Veronesi B, Padilla SS, Newland D. Biochemical and neuropathological assessment of triphenyl phosphate in rats. Toxicol Appl Pharmacol 1986;87:203–10. [47] Glynn P. Neural development and neurodegeneration: two faces of neuropathy target esterase. Prog Neurobiol 2000;61:61–74. [48] Johnson MK. Organophosphates and delayed neuropathy – Is NTE alive and well? Toxicol Appl Pharmacol 1990;102:385–99. [49] Lotti M. The pathogenesis of organophosphate delayed polyneuropathy. Crit Rev Toxicol 1992;21:465–87.