Inhibition of choline oxidase by quinoid dyes

Abstract

Choline oxidase catalyzes the oxidation of choline to glycine-betaine, with betaine-aldehyde as intermediate and molecular oxygen as primary electron acceptor. This study reports on the inhibitory effects of triarylmethanes (cationic malachite green; neutral leukomalachite green), phenoxazines (cationic, meldola blue and nile blue; neutral nile red) and a structurally-related phenothiazine (methylene blue) on choline oxidase, assayed at 25°C in 50 mM MOPS buffer, pH 7, using choline as substrate. Methylene B acted as a competitive inhibitor with K_i = 74 ± 7.2 μM, pointing to the choline–binding site of the enzyme as a target site. Nile B caused noncompetitive inhibition of enzyme activity with K_i = 20 ± 4.5 μM. In contrast to methylene B and nile B, malachite G and meldola B caused complex, nonlinear inhibition of choline oxidase, with estimated K_i values in the micromolar range. The difference in kinetic pattern was ascribed to the differential ability of the dyes to interact (and interfere) with the flavin cofactor, generating different perturbations in the steady-state balance of the catalytic process.

Keywords::

• Choline oxidase
• enzyme inhibition
• malachite green
• meldola blue
• nile blue
• methylene blue

Introduction

Choline oxidase (CHO, E.C. 1.1.3.17) is a FAD-containing enzyme that catalyzes the oxidation of choline to glycine-betaine, with betaine-aldehyde as intermediate and molecular oxygen as primary electron acceptor (Scheme I). Being one of the few flavoproteins that carry out a four-electron oxidation of an alcohol to a carboxylic acid through an aldehyde intermediate, the enzyme has been subjected to detailed mechanistic studies [1,2]. In addition, the need to develop biosensors for the detection of choline and choline esters such as acetylcholine in serological samples and foods has focused clinical and industrial interest on the enzyme as an analytical tool [3,4]. Since many bacterial and plant species accumulate glycine-betaine in response to various conditions such as osmotic stress and extreme temperatures, CHO promises to be of further value in relation to the development of therapeutic agents and the design of genetically modified, drought-resistant crop plants 5-7.

Scheme 1 Oxidation of choline by choline oxidase.

CHO has been purified and characterized from Arthrobacter globiformis [8] and Alcaligenes sp. [9] and the fungus, Cylindrocarbon didymium [10]. The choline oxidase gene of A. globiformis has recently been cloned and sequenced [11]. The resulting protein is a homodimer of 120 kDa, with each subunit containing covalently bound FAD in 8α-N(1)-histidyl linkage [12,13]. The catalytic mechanism of the enzyme has been elucidated by using kinetic and spectroscopic methods 14-18.

Relatively little is known about the active site topology and the substrate and inhibitor specificity of CHO. The trimethylammonium group is a major structural determinant for substrate turnover. Competitive inhibition studies point to the involvement of alkyl-substituted amine headgroups in inhibitor binding (K_i = 2–26 mM) [19]. In the present study, aiming to identify further structural preferences in ligand binding, we have tested the effects of a selection of dyes/leukodyes on choline oxidase activity. The selection includes triarylmethanes (cationic malachite green; neutral leukomalachite green), phenoxazines (cationic, meldola blue and nile blue; neutral nile red) and a structurally-related phenothiazine (methylene blue) (Figure 1). Earlier work in our laboratory has shown that malachite G, meldola B and nile B can reversibly bind to (or form adducts with) a variety of proteins and protein-related nucleophiles 20-22. More specifically, they are highly effective as inhibitors of plasma cholinesterase, with K_i values in the μM range [23]. The phenothiazine ring system is a structural feature of a series of cholinesterase inhibitors which have been studied in detail [24,25]. The cationic dyes were therefore tested for their potential to inhibit choline oxidase as well. The neutral species (leucomalachite G and nile R) were included in the study for comparative purposes. The preliminary results reported here show that the charged dyes bind CHO with 2–3 orders of magnitude higher affinity than the choline analogues cited in reference [19] and might be helpful in the design of new CHO inhibitors.

Figure 1 The quinoid dyes studied.

Materials and methods

Horseradish peroxidase (HRP, E.C. 1.11.1.7, Type VI A), choline oxidase (E.C. 1.1.3.17, from Alcaligenes sp.), malachite G hydrochloride, leucomalachite G, meldola B, methylene B, nile B chloride and nile R were purchased from Sigma-Aldrich (USA). All remaining chemicals were obtained from Sigma-Aldrich or Merck (Germany). Stock solutions of the ligands (5 mM) were prepared in dimethyl sulfoxide.

Measurement of choline oxidase activity

Choline oxidase activity was assayed spectrophotometrically at 25°C, in the presence of HRP and o-dianisidine as the chromogenic component. The assay mixture (1.2 ml) contained 50 mM MOPS buffer (pH 7), 0.1–0.4 mM choline as substrate, 4.2 U HRP, 0.16 mM o-dianisidine. The reactions were initiated by adding ca. 0.3 U/ml CHO (20 μl), followed by 0–40 μM dye ( ≤ 20 μl). The progress of the o-dianisidine oxidation was monitored by the increase in absorbance at 500 nm, using a Shimadzu 1601 PC spectrophotometer equipped with a Peltier unit. CHO activity was calculated using the linear segments of the progress curves in the initial 60-s period (ε₅₀₀, o-dianisidine = 13.5 mM^{− 1} cm^{− 1}, determined in this study).

The potential of the dyes to affect HRP activity was tested by assaying the indicator enzyme in the presence and absence of 40 μM dye. The assays were carried out at 25°C, in 50 mM MOPS buffer (pH 7) containing 0.16 mM o-dianisidine, 2 mM H₂0₂. The reactions were started by the addition of HRP (3 mU/ml) and the oxidation of o-dianisidine was monitored as above. HRP activity was found to be inhibited by < 20%, i.e. not to an extent which would render the chromogenic system rate-limiting in the CHO assay. The stability of the dyes in the assay system was verified spectroscopically. No redox changes were detected.

Results and discussion

All ligands tested were found to act as reversible, linear inhibitors of choline oxidase. The initial rate data were analyzed using a simplified rapid equilibrium model for linear inhibition (Scheme II) and the corresponding rate equations (Equations 1 and 2) [26].

Scheme 2 Rapid equilibrium model for linear inhibition.

Relative inhibitory potencies

Inhibitory potencies were compared on the basis of intercept/slope (i.e. apparent K_i or I₅₀) values derived from Dixon plots of v^{− 1} versus [I] at 0.1 mM choline. The results are given in Table I. All 6 ligands were effective as choline oxidase inhibitors; the I₅₀ values for the cationic species were 100- to 800-fold lower than the K_i values of the aliphatic choline analogues previously studied [19]. The 3- to 6-fold difference between the affinities of the charged and neutral or leuko ligands is similar to the differential affinity of CHO for its substrates, choline (K_m = 0.43 mM) and 3,3-dimethyl-1-butanol (K_m = 1 mM) [19], suggesting that charge-charge interactions are not primary determinants in ligand selectivity and pointing to a major hydrophobic driving force in enzyme-inhibitor complex formation.

Table I.  Choline oxidase inhibitory potencies of triarylmethane, phenoxazine and phenothiazine dyes.

Compound	I50, μM^a,b
Nile blue	18 ± 2.8
Meldola blue	24 ± 4.3
Methylene blue	89 ± 3.8
Malachite green	29 ± 8.7
Nile red	120 ± 3.1
Leukomalachite green	140 ± 6.2

^a[choline]_o = 0.1 mM;

^bSelected K_i values (mM) for choline analogs [19]: trimethylamine, 2.4; diethylmethylamine, 5; trimethylethylamine, 13; allyltrimethylamine, 15.

Notes on the inhibitory mechanism

The effects of the cationic dyes were analyzed in greater detail at multiple [S]_o and [I] values. The resultant family of Dixon plots was utilized to construct diagnostic slope replots according to Equation (3): Scheme II requires that the slope replot intersect the abscissa at [I] = 0, − 1/K_s or − 1/αK_s, depending on whether the inhibitor has a competitive, noncompetitive or mixed effect on substrate binding and turnover [26].

Dixon plots and the slope replot of the inhibition of CHO by methylene B are given in Figure 2. The slope replot (Figure 2B) for the phenothiazine dye pointed to a competitive pattern of inhibition (α = ∞), with K_i = 74 ± 7.2 μM (based on Equation (3) and K_s/V_max = 32.3 ± 2.84 min as determined in independent experiments). The corresponding slope replot of the inhibition by nile B (Figure 3a) indicated noncompetitive inhibition with K_i = 20 ± 4.5 μM. In contrast to methylene B and nile B, the slope replots for inhibition by meldola B and malachite G were inconsistent with Scheme II operating under rapid equilibrium conditions. The plot for meldola B (Figure 3b), while seemingly linear (r² = 0.985), yielded a positive abscissa intercept if processed as a straight line (i.e. a negative value for K_s or α). The inconsistency was even more pronounced in the case of malachite G, where the slope replots showed pronounced curvature (Figure 3c).

Figure 2 The inhibition of choline oxidase by methylene blue. A. Dixon plots of inhibition at 0.1 (♦), 0.14 (▪), 0.2 (▴) and 0.4 (•) mM choline. (Average of 2 independent experiments). B. Slope replot (showing independent data sets).

Figure 3 Slope replots of the inhibition of choline oxidase by nile blue (A), meldola blue (B) and malachite green (C).

The irregularity of the inhibition data for meldola B and malachite G was taken to reflect the complex nature of the multi-step, steady-state CHO system: In steady-state systems, the initial rate equation for competitive inhibition (α = ∞) has the same form as the corresponding equation for a rapid equilibrium scheme. Steady-state initial rate equations for noncompetitive and linear mixed inhibition (1 ≤ α < ∞), on the other hand, contain second-order terms in [S] and [I] [27]. Thus plots based on Equations (2) and (3) are both potentially nonlinear. In the present case, [S]² was apparently the only higher-order term to make a significant contribution within the experimental ranges of [I] and [S], such that only the slope replots deviated from linearity. Proceeding with the assumption of steady-state, noncompetitive (or linear mixed) inhibition, the replots for meldola B and malachite G were fitted to an exponential trendline and αK_i was estimated from the ordinate intercepts with reference to Equation (3) (). The αK_i estimates for meldola B and malachite G were 33 ± 6.5 and 34 ± 3.3 μM, respectively, and in reasonable agreement with the I₅₀ values in Table I.

In conclusion, the results of this study show that choline oxidase can bind and is susceptible to inhibition by a variety of molecules with quinoid structural elements. The competitive nature of the inhibition by methylene B points to the choline-binding site of the enzyme as a target site. The inhibition by nile B, meldola B and malachite G is noncompetitive with respect to choline (but might further have a competitive component obscured by the complexity of the inhibition patterns). The predominantly noncompetitive nature of inhibition by these dyes implicates the flavin-binding site as a possible target. Why structurally similar species such as methylene blue, nile B, meldola B and malachite G exert distinct effects on choline oxidase is not immediately obvious. The difference may stem from the differential ability of the dyes to interact (and interfere) with the flavin cofactor, generating different perturbations in the steady-state balance of the catalytic process. A study on the impact of binding on the spectral properties of the dye and flavin moieties should provide information regarding the proximity of the inhibitor binding domain to the redox active site of the enzyme.

Abbreviations
CHO:	=	choline oxidase
HRP:	=	horseradish peroxidase
; leucomalachite G:	=	leucomalachite green
; malachite G:	=	malachite green
; meldola B:	=	meldola blue
; methylene B:	=	methylene blue
; MOPS:	=	3-(N-morpholino)propanesulfonic acid
; nile B:	=	nile blue
; nile R:	=	nile red

Acknowledgements

This study has been supported in part by a grant (SBAG-2912) from the Scientific and Technical Research Council of Turkey.