Liquid Chromatography with Amperometric Detection at a Silver

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES

Article

Liquid chromatography with amperometric detection at a silver based detector for the determination of thiocompounds: application to the assay of thiopurine antimetabolites in urine Nurgul Karadas-Bakirhan, Ahmad Sarakbi, Marie Vandeput, Sibel A. Ozkan, and Jean Michel Kauffmann Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.5b00879 • Publication Date (Web): 29 May 2015 Downloaded from http://pubs.acs.org on June 6, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Liquid chromatography with amperometric detection at a silver based detector for the determination of thiocompounds: application to the assay of thiopurine antimetabolites in urine Nurgul Karadas-Bakirhana, Ahmad Sarakbib, Marie Vandeputb, Sibel A. Ozkana, Jean-Michel Kauffmannb⃰ a

Ankara University, Faculty of Pharmacy, Department of Analytical Chemistry, 06100 Tandogan, Ankara, Turkey b

Université libre de Bruxelles (ULB), Faculty of Pharmacy, Campus Plaine, CP 205/6, 1050 Bruxelles, Belgium

A silver amperometric detector coupled to liquid chromatography (LC) was used for the determination of 6-thioguanine (6-TG) and two of its metabolites, thiouric acid (TU) and 2amino-6-mercaptopurine riboside (6-TGR). The silver detector coupled to LC operated at a low applied potential (0.08 V vs Ag/AgCl) and offered a chromatogram with peak responses corresponding to molecules interacting with silver namely, chloride ions and small soluble biothiols in addition to the organothiol drug compounds investigated. On-line electrochemical surface cleaning permitted to improve the repeatability and peak shape of the recorded signal compared to direct current amperometric detection (AD) when operating in chloride containing media. The studied molecules were eluted isocratically within 5 min on a reversedphase C18 column without interference from endogenous biothiols present in urine samples. Diluted urine samples (1:1) were directly injected in the LC setup, a linear calibration curve was obtained between peak area and analyte concentration between 0.1-10 µM for all the studied molecules. Limits of detection (LOD) were 0.03, 0.008 and 0.01 µM, and the limits of quantification (LOQ) were 0.1, 0.02 and 0.03 µM for TU, 6-TG and 6-TGR, respectively. Within-day RSDs were 2%, 0.8%, and 1 % and between-days RSDs were 2 %, 0.9 %, 2 % for TU, 6-TG, 6-TGR, respectively. Recoveries in spiked urine were 99.8%, 99.9 % and 99.0 % for TU, 6-TG and 6-TGR, respectively. ⃰ Corresponding

author: [email protected]

Fax: 00 32 26505225

Key words: amperometry, liquid chromatography, silver electrode, thioguanine, thiouric acid

1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Gold and platinum are suitable as electrode material for the determination of thiols and disulfides by liquid chromatography (LC).1 Due to rapid surface inactivation and analyte deposits, a pulsed electrochemical detection (LC-PED) is applied at these metallic based detectors. The signal corresponds to the analyte oxidation and metallic oxide formation.1-3 The selected PED waveform consists of alternated anodic and cathodic polarizations to clean the electrode through surface oxide formation and subsequent removal. This detection mode has been extensively investigated and is currently used in routine LC assays of sulfurcontaining compounds such as antibiotics and biothiols.3-8 Mercury amalgamated gold electrodes in a dual electrode configuration also allow the sensitive LC determination of both thiols and disulfides.9 Carbon based electrodes such as glassy carbon carbon paste or boron dopped diamond based detectors are suitable for thiols and disulfides but high positive potentials are required.10,11 The use of silver as working electrode in liquid chromatography (LC) is seldom considered except for the amperometric detection of halides, cyanide, sulfide, thiosulfate and thiocyanate at a potential close to 0.0 V vs Ag/AgCl.12,13 The signal corresponds to the facilitated oxidation of silver in the presence of the studied analytes ie. a surface redox phenomenon occurs where the electrode is consumed in the process. The LC eluent is strongly basic and application of an electrochemical waveform (PAD) permits online surface cleaning. The latter phenomenon corresponds to silver deposits removal via anodic formation of Ag(OH) and its subsequent electroreduction to reactivate the silver electrode surface.12,13 Recently, liquid chromatography with amperometric detection (LC-AD) at a silver electrode polarized at a potential close to 0.0 V vs Ag/AgCl was found to be suitable for the determination of biological thiols.14 The LC mobile phase was acidic which is of particular interest for the assay of biothiol compounds since it is well-known that they are unstable in alcaline media. The method was applied to the determination of glutathione, homocysteine, N-acetylcysteine and cysteine in white wine.15 The detector response mechanism was inferred to correspond to the reaction of the analyte, via its thiol functional group, both with the metallic silver surface and the silver ions present at the electrode-solution interface. A deposit of a biothiol salt and the formation of a self assembled monolayer (SAM) of the biothiol onto the silver electrode were inferred to occur during the detection. These surface phenomena

2


Page 2 of 15

Page 3 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


generated LC post-peak dipping and a surface pre-conditionning, by performing some biothiol injections, was required for subsequent signal repeatability.14 The present work aimed at showing the interest of applying a pulsed potential waveform at a silver based electrochemical detector coupled to liquid chromatography operating under acidic mobile phase conditions for the determination of organothiols in complex samples such as urine. This detection mode does not exploit the formation of metal oxide intermediates for surface cleaning. It can, however, be considered as a pulsed amperometric detection (PAD) mode since it imposes a repeated change of potentials and generates a signal by integrating the working electrode current over a very short period of time during which the applied potential remains unchanged.13 The preferential selection of a silver rather than a noble metal or a carbon based detector in LC for thiols was governed by the fact that the chromatogram is relatively “blind” with respect to numerous molecules which may interfere when applying high positive potentials.14 The studied drug compounds were 6thioguanine (6-TG) and two of its metabolites, thiouric acid (TU) and 2-amino-6mercaptopurine riboside (6-TGR). These molecules belong to the purine antimetabolites family (Figure 1). Thioguanine has been recognized as an important chemotherapeutic drug in the treatment of childhood acute lymphoblastic leukemia,16 inflammatory bowel disease, Crohn’s disease, AIDS, and some other pathologies.17,18

TU

6-TG

Figure 1. Schematic structures of TU, 6-TG and 6-TGR

3


6-TGR


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 15

LC methods for the determination of TG and its metabolites apply UV,19-24 fluorimetry,25-27 mass spectrometry 28 and amperometry

29

as detection mode. Amperometry at carbon based

electrodes is perturbated by surface fouling problems, yet better responses were observed at a cobalt phthalocyanine (CoPc) modified carbon paste electrode (CPE) thanks to the catalytic action of CoPc.11 MATERIALS AND METHODS All chemicals were of analytical reagent grade. 6-TG and 6-TGR were obtained from SigmaAldrich, respectively. TU was obtained from Toronto Research Chemicals Inc. (Toronto, Canada). The metabolite 2-amino-6-methylmercaptopurine was obtained from Acros (Geel, Belgium). Glutathione, homocysteine and N-acetylcysteine were purchased from SigmaAldrich (Stelnheim, Germany). Cysteine was purchased from Fluka (Steinheim, Germany). Acetic acid, formic acid, sodium acetate, sodium nitrate, disodium hydrogen phosphate were obtained

from

Merck

(Darmstadt,

Germany).

Methanol

and

disodium

ethylenediaminetetraacetic acid (EDTA) were purchased from VWR (Leuven, Belgium). The mobile phase (0.1 M formic acid, 0.5 mM EDTA, 0.025 M sodium nitrate, 10% methanol adjusted pH 6.0) was prepared by dissolving 4.250 g sodium nitrate, 0.375 g EDTA and 9.20 g formic acid (98%) in 1.5 L of water and the pH was adjusted to 5.5 (for LC-AD) or 6.0 (for LC-PAD) with a 2.0 M sodium hydroxide solution and the volume was filled up to 2 L with water after addition of 200.0 mL of methanol. Stock solutions of the TU (1×10−3 M) was prepared in the mobile phase at pH 6.0. The 6-TG and 6-TGR (1×10−3 M) solutions were prepared in 0.1 M NaOH and kept at 4°C in the fridge. Standard solutions were prepared twice a month using appropriate dilution of the stock solution with the mobile phase pH 6.0. The LC experiments were performed using a Dionex ICS-5000 LC system, equipped with an autosampler controlled by Chromeleon software. Separation was performed in isocratic mode on an Atlantis®dC18 column (3 µm, 4.6 × 100 mm) from Waters. The mobile phase was deoxygenated by nitrogen during the experiments. Experiments were performed at 25.0 ± 0.2◦C. A silver disk (1 mm diameter) working electrode housed in a flow-through electrochemical cell was coupled to the LC system. The auxiliary electrode was titanium and the reference electrode was made of Ag/AgCl, 3 M KCl. The cell volume of the detector was 0.2 µL. PAD was applied using the following parameters; applied potential sequence E1 = 0.08 V, E2 = 0.08 V, E3 = - 0.800 V with pulse duration time t1= 1,750 ms, t2 = 200 ms and t3= 10 ms (Figure 2). 4


Page 5 of 15

0.2

T1

T2

T3

0.0

Potential (V)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


-0.2 -0.4 -0.6 -0.8 0.0

0.5

1.0 Time (ms)*103

1.5

2.0

Figure 2. Shape of the pulsed potential waveform: t1 = stabilization period (1750 ms), t2 = measuring period (200 ms), t3= forced electroreduction period (10ms).

The flow rate of the mobile phase was 0.9 mL min−1. When not in use, the LC setup was maintained at a flow rate of 50 µL/min with the detector at open circuit. Manual cleaning of the working electrode was realized monthly by surface smoothing on a water wetted polishing cloth in the presence of alumina powder (P/N 036318 from Dionex, Aalst, Belgium). When urine samples were studied by LC-AD or LC-PAD, the electrode needed a manual polishing before starting the experiments on a daily basis or every 2 weeks, respectively. The autosampler and the entire LC setup were thermostatted at 25.0 ± 0.2◦C. HPLC grade water was obtained from a Milli-Q filtration station (Millipore Filter Corp., Bedford, MA, USA). The mobile phase was filtered through a 0.2 µm GH Polypro hydrophilic polypropylene membrane fixed in a SolVacTM filter holder (Pall, WVR Belgium). The pH of the mobile phase was determined with a Metrohm 744 pH Meter (Metrohm, Belgium). Prior to LC injection, standard solutions and urine samples diluted with the mobile phase (1:1) were filtrated through a 0.45 µm membrane (PAL Acrodisc® LCPVDF, VWR-Belgium). A loop of 20 µL served for sample injection. Cyclic voltammetry (CV) was performed using a potentiostat Epsilon (BASi – West Lafayette – USA) and a three electrode cell configuration with a homemade silver wire housed in solid paraffin as working electrode, a Ag/AgCl, 3 M KCl and a platinum wire served as counter 5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 15

electrodes. The silver electrode surface (exposed disk area diameter: 0.5 mm) was renewed, between each voltammogram, by cutting a thin slice of the tip of the electrode with a titanium based scissor. Dissolved oxygen of analyzed solutions was removed by deoxygenating with nitrogen. CVs were realized in acetate buffer 0.01 M + NaCl 0.01 + EDTA 0.5 mM + NaNO3 0.025 M + 10 % methanol at pH 5.5. Method validation. The developed method was validated according to USP and ICH guidelines.30,31 The linearity was evaluated by analyzing a series of varied concentrations for 6-TU, 6-TG and 6-TGR. The detection limit (LOD) and limit of quantification (LOQ) were calculated from the equation LOD =3.3×s/m and LOQ =10×s/m, where m is the slope of the calibration curve and s the standard deviation of the response.32 The ruggedness and precision of experiments were checked in the same day and between days, results were expressed as percentage of relative standard deviations (RSD %). All solutions were kept in the dark in a refrigerator at 4oC. Application to urine samples. Fresh thiopurine metabolites-free urine samples were obtained from a healthy volunteer and simply diluted with the mobile phase (1:1, v/v) and filtrated before LC analysis (see materials and method). A calibration curve was realized by spiking the urine sample with increasing amounts of the studied drug compounds. A linear regression of peak area versus concentration was applied to fit the calibration curve. The TU, 6-TG and 6-TGR prepared in dopped diluted urine samples were determined by the method of standard addition. Recovery was calculated by studying comparatively the peak area of TU, 6-TG and 6-TGR spiked at 2 µM in the diluted urine sample (1:1) and the peak area of

these

metabolites spiked at 2 µM in the mobile phase.

RESULTS AND DISCUSSION Cyclic voltammetry. Experiments were realized with the silver microelectrode in acetate buffer pH 5.5 in the presence of chloride ions. Acetate was required instead of formate for the CV experiments due to the buffering capacity of the former at pH 5.5 or 6.0. By starting at 0.4 V and scanning towards positive direction in the supporting electrolyte, the facilitated oxidation of silver at + 0.280 V and the corresponding silver chloride reduction peak at 0.020 V (Figure 3 A dotted lines) were noticed. In the presence of 6-TG (2 mM), a slight oxidation wave started at – 0.100 V followed by the facilitated anodic peak of silver (Ep = + 0.300 V). During the reverse scan, two reduction peaks close to 0.0 V (Figure 3 A solid line) and one 6


Page 7 of 15

cathodic peak (Ep = – 1.00 V) were noticed. The slight anodic wave starting at -0.100 mV was attributed to the facilitated oxidation of the silver electrode to give silver ions in the presence of 6-TG. This phenomenon was exploited in subsequent LC experiments with aperometric detection at the silver working electrode (see below). We had observed already a similar CV pattern for cysteine and thiocholine at a silver microelectrode.33 The two reduction peaks noticed at 0.0 V were attributed to the reduction of silver chloride and silver thiolate formed at the electrode surface as also reported in the literature for thiol based aminoacids reduction at a silver electrode.34 By starting the scanning potential at - 0.2 V towards negative direction, a broad reduction peak starting at approximately - 0.800 V was obtained (Ep = 1.00V). The origin of the cathodic peak at – 1.00 V noticed both in Figure 3 A and B was attributed to the reductive desorption of a 6-TG self assembled monolayer (SAM) ie., 6-TG covalently linked to metallic silver. Reductive desorption of a SAM of thiols has been described to occur on gold and silver electrodes at potentials close to -1.0 V vs Ag/AgCl.35 Taking into account the CV results and in order to avoid electrode surface fouling phenomena a PAD strategy was implemented in LC analysis of urine samples. The pulsed waveform consisted of a cathodic polarization in order both to desorb the thiolate SAM and to force the electroreduction of silver thiolate or silver chloride formed deposit. Special care was taken to apply positive potentials lower than + 0.1 V vs Ag/AgCl in order to avoid facilitated silver oxidation by chloride ions ie. minimize the formation of a silver chloride deposit.

8 A

2

4

1

0

0

I / µA

I / µA

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


-4

B

-1

-8 -2

-12 -1.2

-0.8

-0.4 E/ V

0.0

0.4

-1.2

-0.8 E/ V

-0.4

Figure 3. Cyclic voltammograms at the silver microelectrode as a function of investigation potential domain in acetate buffer 0.01 M + NaCl 0.01 + EDTA 0.5 mM + NaNO3 0.025 M + 10 % methanol at pH 5.5. A) Starting potential at - 400 mV towards positive direction, 7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 15

reversing at + 400 mV, scan rate 50 mV/s. (dotted line); in the presence of 2 mM 6-TG (solid line). B) Starting potential at -200 mV towards negative direction (dotted line); in the presence of 2 mM 6-TG (solid line). Scan rate 50 mV/s.

LC-AD The mobile phase was based on formate ions (instead of acetate, or phosphate) in order to minimize interaction of the mobile phase components with silver ions at the electrode solution interface.14 EDTA was added in order to complex possible ionic species such as iron, copper or nickel which might interfere at the electrode-solution interface and/or catalyze the oxidation of thiols.36 It was previously verified that EDTA gave no CV or amperometric signal at the silver electrode under the selected experimental conditions likely due its relatively weak complexation with silver ions (log Kf = 7.3) compared to species which react strongly with silver ions such as cysteine or glutathion (log Kf = approximately 12).

14

The

LC-AD method operated at 0.08 V using a pH 5.5 mobile phase with 10% methanol content at a flow rate of 0.7 mL/min. It was applied to the study of 6-TG and 6-TGR. The recorded signal was linearly related to the concentration of the analytes between 0.1 and 6 µM both in the mobile phase and in spiked diluted urine samples: Y (nA)= 743 X (mM) - 0.2765 (nA)

(r= 0.999) for 6-TG in mobile phase (pH 5.5)

Y (nA)= 477 X (mM) - 0.2657 (nA)

(r= 0.999) for 6-TGR in mobile phase (pH 5.5)

Y(nA)= 737 X (mM) - 0.0599 (nA)

(r= 0.997) for 6-TG in diluted urine

Y (nA)= 539 X (mM) - 0.0462 (nA)

(r= 0.999) for 6-TGR in diluted urine

The silver detector was suitable for one month without surface cleaning when studying buffer samples but it needed daily a manual cleaning when studying urine samples. The need for a manual cleaning was attributed to the high concentration of chloride ions and other interfering species present in urine which reacted with the silver ions at the electrode-solution interface and formed a deposit onto the electrode surface.

8


Page 9 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


LC-PAD The pulsed potential mode was applied in order to try to improve the silver electrode operational stability by on-line electrochemical cleaning and to eventually eliminate the post peak dropping observed in the AD mode. The strategy consisted to force the reduction of the Ag adducts back to Ag on the electrode. The applied potential waveform comprised a stabilization period (t1), a measuring period (t2) and a forced electroreduction period (t3). The t2 period was varied between 50 and 250 ms: the signal of the studied thiols increased and started to level off at 200 ms. By varying the t3 period it was found that duration times longer than 10 ms produced a decrease of the signal. The selected potential values were optimized by studying the negative potential value for deposit desorption between -0.6 V and -1.0 V and the detection and stabilization potentials were varied between 0.0 and + 0.1 V. Care was taken to avoid potentials higher than + 0.1 V in order to minimize the formation of a silver chloride deposit. The signal (peak height and area) increased by lowering the reduction potential value from - 0.6 V till - 0.8 V but, at values lower than - 0.8V the baseline and the peak magnitude decreased and a poor repeatability was noticed. The final selected pulsed potential sequence was : E1= 0.08 V, E2= 0.08 V, E3= - 0.800 V. The pH of the mobile phase influenced both the chromatographic pattern and the electrode response. This parameter was studied at pH 3.0, 4.0, 5.0 and 6.0. The baseline noise and the TU, 6-TG and 6-TGR post-peak drop became lower by raising the pH and peak resolution remained good. By increasing the methanol content from 5 till 10 % , the signal raised but the retention time (tR) decreased substantially. A good compromise in terms of peak shape, peak resolution, post-peak dropping and signal /noise was obtained at pH 6.0 and 10% methanol content. Under the selected conditions the compounds were eluted within less than 5 min. The flow rate effect was studied in the range 0.5–0.9 mL/min. An inverse correlation was observed between the flow rate and tR in agreement with the theory for isocratic separation. From 0.5 till 0.9 mL/min, the peak height raised by 50% while the peak area decreased by 50% then both trends leveled off. The peak current increase was attributed to tR shift to lower values, i.e., reduced band dispersion with the appearance of a more concentrated band at the detector, and to a thinner diffusion layer at the electrode-solution interface at high flow rates. The peak area decrease at high flow rates suggested that the overall process was controlled by a chemical reaction kinetic at the electrode surface.37 Based on these data a flow rate of 0.9 9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 15

mL min−1 was selected as it offered a good compromise between peak shape and magnitude and selectivity over possible interfering species. The response (peak area) was studied as a function of concentration in the range comprised between 0.1 and 10 µM. The between-day experiments permitted to achieve a RSD of the signal lower than 2% (peak area) during a two weeks period without the need for manual surface cleaning neither when studying diluted urine (1:1) nor buffer solutions. Calibration and linearity. Some matrix effect was noticed when comparing the slopes of the

calibration curves realized in buffer and in diluted urine (1:1); the latter gave higher responses (1.5 higher slopes in diluted urine compared to buffer). This matrix effect could not be explained yet. Further validation experiments were realized by studying diluted urine (1:1) spiked with the investigated drug compounds. The latter were eluted within 5 min; TU tR = 1.90 min, 6-TG tR = 3.75 min, 6-TGR tR = 4.77 min. The peak area was linearly related to the concentration of the studied molecules within the range 0.1 -10 µM (Figure 4). The following equations for the regression line were obtained:

Y (nC) = 48.4 X (mM) + 0.031 (nC)

% RSD: 2 (n=5)

r = 0.999

(TU)

Y (nC) = 89.3 X (mM) - 0.0014 (nC)

% RSD: 1 (n=5)

r = 0.999

(6-TG)

Y (nC) = 45.4 X (mM) + 0.0003 (nC)

% RSD: 0.8 (n=5)

r =0.999

(6-TGR)

Limits of detection (LOD) were 0.03, 0.008 and 0.01 µM, and the limits of quantification (LOQ) were 0.1, 0.02 and 0.03 µM for TU, 6-TG and 6-TGR, respectively, with RSD values equal to 2 %. All results were obtained from 5 measurements (n=5) for each concentration.

10


Page 11 of 15

35

PAD Response (nC)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TU

6-TG

6-TGR

28 21

10 µM

14 7

0.4 µM

0

Urine Blank 0

1

2

3

4

5

6

Retention Time (min)

Figure 4. LC-PAD chromatograms of a mixture of TU, 6-TG and 6-TGR at 0.4 µM; 0.6 µM; 0.8 µM; 1.0 µM; 2.0 µM; 4.0 µM; 6.0 µM; 8.0 µM; 10 µM in diluted urine samples (1:1).

The within-day and between day RSDs of the peak area (n=5) were 2 % for TU, 6-TG and 6TGR, respectively. The RSDs (n=5) of the retention time of TU, 6-TG and 6-TGR were 0.5 %, 0.1 % and 0.2 % for the within-day, and 0.4 %, 0.4 % and 0.2 % for the between-day, respectively. Recoveries in diluted urine were 99.8 %, 99.9 % and 99.0 % for TU, 6-TG and 6-TGR, respectively. Stability. The stability of TU, 6-TG and 6-TGR stock solutions were checked. After 7 days stored at 4 °C no statistically significant signal decrease was noticed (p > 0.05). At room temperature these solutions were stable only for one day, then the signal decreased. The stability of the detector was also checked in diluted urine samples. The LC-PAD method could be used for two weeks, then manual cleaning of the electrode surface was required due to a decrease of the signal. The developed method permitted the use of the LC column for more than 6 months of daily use.

11



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 15

Interference study and LC-PAD determination of TU, 6-TG and 6-TGR in spiked urine. By applying the selected conditions, the analysis of the three studied compounds

was

successfully achieved in diluted urine (Figure 4). Several peaks were noticed in the blank urine sample namely chloride (tR= 1.21) with a typical postpeak dropping followed by Lcysteine (tR = 1.35). Identification of the individual peaks from the urine samples was based on tR and by spiking the urine sample with increasing amounts of glutathione, L-cysteine, Nacetylcysteine and homocysteine reference standards. By spiking glutathione, L-cysteine, Nacetylcysteine, homocysteine in the diluted urine sample (1:1), the same retention time of glutathione and L-cysteine were noticed with tR= 1.37 and tR=1.35, respectively, homocysteine at tR= 1.52, N-acetylcysteine at tR= 1.67. Glutathione and L-cysteine gave nearly the same retention time, they could not be separated under these conditions. All the detected biothiols were estimated to be present in urine at a concentration below 10 µM. The presence of chloride ions and the biothiols, especially, glutathione and cysteine, did not affect TU, 6-TG and 6-TGR responses in the studied concentration range. Disulfides gave no signal at the detector under the selected PAD conditions. It is interesting to mention, however, that we noticed experimentally that by using an electrochemical reaction cell inserted on-line (LC post-column)

38

the disulfides were reduced and readily detected by PAD at the Ag detector

(results not shown).

CONCLUSION A silver based detector with pulsed electrochemical detection can be advantageously applied for the determination of organothiols under hydrodynamic conditions as illustrated by the LCPAD assay of the drug compound 6-TG and some of its metabolites namely TU and 6-TGR in urine. The acidity of the mobile phase and the short chromatographic run time were appropriate conditions for avoiding TU, 6-TG and 6-TGR degradation during the assay. The silver based detector offered a “clean” chromatogram blind to readily oxidized species present in biological samples. The isocratic LC-PAD could readily be implemented with minor instrumentation handling and it offered a long operational shelf life of the RP C18 column and the silver based detector. The method required no pretreatment of urine samples other than dilution with the mobile phase which is advantageous since sample processing may alter the redox status of the TU, 6-TG and 6-TGR. It is anticipated that the presently developed methodology could be implemented in different instrumental configurations exploiting 12


Page 13 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


hydrodynamic conditions and could be applied to the investigation of a variety of thio-based drug compounds. Acknowledgements Thanks are expressed to TUBITAK for grant support to Karadas-Bakirhan for a scientific stay in the Université libre de Bruxelles.

REFERENCES (1) Jonhson, D. C.; Polta, J.A.; Polta, T.Z.; Neuburger, G.G.; Jonhson J.; Tang, A P-C.; Yeo, I-H., Baur, J. J.Chem. Soc., Faraday Trans. 1986, 82, 1081-1098. (2) Jonhson, D.C.; LaCourse, W.R. Anal.Chem. 1990, 62,589A-597A. (3) Fedorowski, J.; LaCourse W.R. Anal.Chim.Acta, 2015, 861, 1-11. (4) LaCourse, W.R.; Owens, G.S. Anal.Chim.Acta 1995, 307, 301-319. (5) Dionex (Thermo Scientific) Application note 132. (6) Cataldi, T.R.I.; Nardiello, D. J. Chromatogr.A 2005, 1066, 133-142. (7) Modi , S.J.; LaCourse W.R.; Shansky, R.E. J.Pharm.Biomed. Anal. 2005, 37, 19-25. (8) Cheng, J.; Jandik, P.; Avdalovic N. Anal.Chem.2003, 75, 572-579. (9) Stenken, J.A.; Puckett, D.L.; Lunte, S.M.; Lunte C.E. J. Pharm. Biomed. Anal. 1990, 8, 85-89. (10)

Park, H.J., Mah, E.; Bruno, R.S. Anal. Biochem. 2010, 407,151-159.

(11)

Halbert, M.K.; Baldwin, R.P. Anal.Chim.Acta 1986, 187, 89-97.

(12)

Giuriati, C.; Cavalli, S.; Gorni, A.; Badocco, D.; Pastore, P. J. Chromatogr.A, 2004,

1023, 105-112. (13)

Cheng, J.; Jandik, P.; Avdalovic, N. Anal.Chim.Acta 2005, 536, 267-274.

(14)

Sarakbi, A.; Aydogmus, Z.; Dago, A.; Mertens, D.; Dewert J.Y.; Kauffmann J. M.

Anal. Chim. Acta 2013, 786, 22– 28. 13



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(15)

Page 14 of 15

Sarakbi, A.; Kauffmann, J-M. Food Chem. 2014, 153, 321-326.

(16) Tan, C. T.; Wollner, N.; Trippett, T.; Goker, E.; Tong, W. P.; Kheradpour, A.; Meyers, P. A.; Van Syckle, K. M.; Guarino, L.; Elisseyeff, Y. J. Clin. Oncol. 1994, 12, 1955 – 1962. (17) Spackova, N.; Cubero, E.; Sponer, J.; Orozco, M. J. Am. Chem. Soc. 2004, 126, 1464214650 (18) Bruunshuus, I.; Schmiegelow, K. Scand. J. Clin. Lab. Invest. 1989, 49, 779 – 784. (19) Boulieu, R.; Lenoir, A. J. Chromatogr. B 1995, 665, 213 – 216. (20) Su, Y.; Hon, Y. Y.; Chu, Y. Q.; van de Poll, M. E. C.; Relling, M. V. J. Chromatogr. B 1999, 732, 459 – 468. (21) Dervieux, T.; Chu, Y.; Su, Y.; Pui, C. H.; Evans,W. E.; Relling, M. V. Clin. Chem. 2002, 48, 61 – 68. (22) Stefan, C.; Walsh, W.; Banka, T.; Adeli, K.; Verjee, Z. Clin. Biochem. 2004, 37, 764 – 771. (23) Erb, N.; Haverland, U.; Harms, D. O.; Escherich, G.; Janka-Schaub, G. J. Chromatogr. B 2003, 796, 87 – 94. (24) Lennard, L.; Singleton, H. J. J. Chromatogr.,Biomed. Appl. 1992, 583, 83 – 90. (25) Dooley, T.; Maddocks, J. L. J. Chromatogr. 1982, 229, 121 – 127. (26) Erdmann, G. R.; France, L. A.; Bostrom, B. C. Biomed. Chromatogr. 1990, 4, 47 – 51. (27) Warren, D. J.; Slordal, L. A. Anal. Biochem. 1993, 215, 278 – 283. (28) Wang, H.;Wang,Y. Anal.Chem. 2010, 82, 5797-5803. (29) Imai, H.; Yoshida, H.; Masujima, T.; Morita, I.; Matsuura, K.; Nakamaru, A.; Katayama, K.; Matsuo, H. Anal.Letters, 1983, 16, 1109-1114. (30) United States Pharmacopeia 32 (USP 32) Revision; United States Pharmacopeia Convention: Rockville, MD, USA, 2009. (31) ICH, Topic Q2A validation of analytical procedures, methodology, PMP/ICH/281/95. 14


Page 15 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(32) Ozkan, S.A. Electroanalytical methods in pharmaceutical analysis and their validation, HNB Publishing, New York, 2012. (33) Vandeput,M.; Parsajoo, C.; Vanheuverzwijn, J.; Patris, S.; Yardim, Y.; le Jeune, A.; Sarakbi, A.; Mertens, D.; Kauffmann, J-M. J. Pharm. Biomed. Anal. 2015, 102, 267-275. (34)

Vidugiris,

G.-J.A.;

Razumas,

V.J.;

Drungiliene,

A.A.;

Kulys,

J.J.

Bioelectrochem.Bioenerg. 1988, 19, 513-520. (35) Widrig, C.A.; Chung, C.; Porter, M.D. J.Electroanal.Chem., 1991, 310, 335-359. (36) Hansen, R.E.; Winther, J.R. Anal. Biochem. 2009, 394, 147-158. (37) Stulik, K.; Pacakova, V.; Le, K.; Hennissen, B. Talanta 1988, 35, 455-460. (38) Kraj, A.; Brouwer, H-J.; Reinhoud, N.; Chervet, J-P. Anal.Bioanal.Chem. 2013, 405, 9311- 9320.

Graphical abstract

Organothiol (RSH) interaction with the silver electrode polarized at 0.08 V

15


Liquid Chromatography with Amperometric Detection at a Silver

Recommend Documents