- Email: [email protected]
HPLC-HG-AFS determination of arsenic species in acute promyelocytic leukemia (APL) plasma and blood cells
Accepted Manuscript Title: HPLC-HG-AFS determination of arsenic species in acute promyelocytic leukemia (APL) plasma and blood cells Authors: Meihua Guo, Wenjing Wang, Xin Hai, Jin Zhou PII: DOI: Reference:
S0731-7085(17)31282-7 http://dx.doi.org/doi:10.1016/j.jpba.2017.07.001 PBA 11378
To appear in:
Journal of Pharmaceutical and Biomedical Analysis
Received date: Revised date: Accepted date:
23-5-2017 29-6-2017 1-7-2017
Please cite this article as: Meihua Guo, Wenjing Wang, Xin Hai, Jin Zhou, HPLC-HG-AFS determination of arsenic species in acute promyelocytic leukemia (APL) plasma and blood cells, Journal of Pharmaceutical and Biomedical Analysishttp://dx.doi.org/10.1016/j.jpba.2017.07.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
SHPLC-HG-AFS determination of arsenic species in acute promyelocytic leukemia (APL) plasma and blood cells Meihua Guoa1, Wenjing Wanga1, Xin Haia*, Jin Zhoub,c,* aDepartment
of Pharmacy, the First Affiliated Hospital, Harbin Medical University, 23 Youzheng Street, Nangang
District, Harbin, 150001, China bDepartment
of Hematology, the First Affiliated Hospital, Harbin Medical University, 23 Youzheng Street,
Nangang District, Harbin, 150001, China cHeilongjiang
1These
Institute of Hematology & Oncology, 23 Youzheng Street, Nangang District, Harbin,150001, China
authors contributed equally to this work.
*Corresponding
author
Jin Zhou, Department of Hematology, the First Hospital, Harbin Medical University, 23 Youzheng Street, Nangang District, Harbin, 150001, China E-mail address: [email protected] Xin Hai, Department of Pharmacy, the First Hospital, Harbin Medical University, 23 Youzheng Street, Nangang District, Harbin, 150001, China E-mail address: [email protected]
Highlights
The methods of HG-AFS and HPLC-HG-AFS were established and validated for quantitative analysis of total arsenic in whole blood and arsenic species of As(III), DMA, MMA and As(V) in plasma, respectively.
Sample preparation for speciation involved only a simple deproteinization procedure with HClO4 and the four arsenic species was separated using HILIC column within 6 min.
The methods have been successfully applied to the assay of arsenic compounds in plasma and blood cells from patients with acute promyelocytic leukemia, which were found to be simple, low cost as well as could be easily implemented in clinical laboratory.
Abstract Arsenic trioxide (ATO) has been successfully used in the treatment of acute promyelocytic leukemia
(APL). To clarify the arsenic species in APL patients, high performance liquid chromatography-hydride generation-atomic fluorescence spectrometry (HPLC-HG-AFS) and HG-AFS methods were developed and validated to quantify the plasma concentrations of inorganic arsenic (As(III) and As(V)) and methylated metabolites (MMA and DMA), and the total amounts of arsenic in blood cells and plasma. Blood cells and plasma were digested with mixtures of HNO3-H2O2 and analyzed by HG-AFS. For arsenic speciation, plasma samples were prepared with perchloric acid to precipitate protein. The supernatant was separated on an anion-exchange column within 6 min with isocratic elution using 13 mM CH3COONa, 3 mM NaH2PO4, 4 mM KNO3 and 0.2 mM EDTA-2Na. The methods provided linearity range of 0.2-20 ng/mL for total arsenic and 2.0-50 ng/mL for four arsenic species. The developed methods for total arsenic and arsenic species determination were precise and accurate. The spiked recoveries ranged from 81.2%-108.6% and the coefficients of variation for intra- and inter-batch precision were less than 9.3% and 12.5%, respectively. The developed methods were applied successfully for the assay of total arsenic and arsenic species in 5 APL patients. The HPLC-HG-AFS may be a good alternative for arsenic species determination in APL patients with its simplicity and low-cost in comparison with HPLC-ICP-MS.
Abbreviations: ATO, arsenic trioxide; APL, acute promyelocytic leukemia; HPLC-HG-AFS, high performance liquid chromatography-hydride generation-atomic fluorescence spectrometry; As(III), arsenite; DMA, dimethylarsinic acid; MMA, monomethylarsonic acid; As(V), arsenate.
Keywords High performance liquid chromatography-hydride generation-atomic fluorescence spectrometry Arsenic speciation Acute promyelocytic leukemia Plasma Blood cells
1. Introduction Arsenic trioxide (ATO, arsenite (As(III)) in solution) has been successfully used in the treatment of
patients with acute promyelocytic leukemia (APL) [1]. As firmly established, As(III) can be transformed into the major metabolites monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA) by the metabolism that involves oxidation and oxidative methylation by a series of enzymes [2, 3]. It is essential to determine inorganic arsenic as well as arsenic metabolites, namely DMA and MMA in patients with APL, revealing the relationship between arsenic speciation and clinical response. Because of the different toxic effects induced by different As species [3-5], speciation analysis is critical when evaluating the therapeutic efficiency of arsenic t7rioxide for treating individual with APL. Although a few studies concerning the arsenic speciation in urine and plasma of APL were reported [6-10], the detailed systematic analysis of the arsenic species in blood cells of APL patients has not yet been performed. Inductively coupled plasma mass spectrometry (ICP-MS) has been chosen by many authors as a chromatographic detector system in the measurements of total arsenic or arsenic species, because of the rapidity and very high sensitivity of measurements [8-12]. Considering the samples to be analyzed in most hospitals, a fast, simple and low cost analytical method is more feasible to make it perform easily. However, both the cost of ICP-MS instrumentation as well as its running costs are quite high that limit its applications for routine analysis in many clinical laboratories. In addition, the determination of arsenic at m/z of 75 by ICP-MS could be compromised by
40
Ar35Cl+ and
38
Ar37Cl+ [8, 13-15]. It is
therefore necessary to overcome the polyatomic interference by taking some measurements, which increase complexity. As an alternative, hydride generation-atomic fluorescence spectrometry (HG-AFS) has been described to be similar to ICP-MS for arsenic speciation analysis in linear range and sensitivity [6, 12, 16]. The cost of the HG-AFS instrumentation is about 1/10 of that for most ICP-MS instruments and it has lower running costs [12]. AFS detects arsenic by a high-intensitive arsenic lamp, avoiding the interferences from ArCl. The application of HG is beneficial for quantifying arsenic, since it separates the analytes from the sample matrix, minimizes or reduces chemical and spectral interferences [15]. Recently, HG-AFS has been applied for determination of arsenic compounds in urine samples of APL patients [6]. Few analytical methods have been performed to determine total arsenic and arsenic species in whole blood of APL by HG-AFS. An HPLC-HG-AFS method has been used by Šlejkovec et al. [7] for the determination of arsenic species in blood samples with a complex sample treatment, yet no validation data have been provided to prove its sensitivity and specificity. The aim of the present
study was to develop a fast, accurate and low cost analytical method by HPLC-HG-AFS for total arsenic and arsenic speciation in APL whole blood. Special attention was focused on all aspects of the analytical procedure to simplify sample preparation and retrieve as high quality data as possible. The methods were fully validated and applied in APL patients following ATO treatment. HPLC-HG-AFS is approved to be a feasible alternative for arsenic speciation study which would be helpful in understanding the relationship between arsenic concentrations and clinical response. 2. Materials and methods 2.1. Chemicals and Reagents The chemicals used in this work were of guarantee grade and were all purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). The deionized water was obtained from a water purification system (Thermo Genpure, USA) and always freshly prepared. Stock solutions were prepared using the following standard As and arsenic compounds: As (GBW(E)080117), arsenite (As(III), GBW08666), arsenate (As(V), GBW08667), monomethylarsonic acid (MMA, GBW08668) and dimethylarsinic acid (DMA, GBW08669) purchased from National Institute of Metrology (Beijing, China). Stock solutions were prepared with fresh deionized water and stored in the dark at 4oC. Working solutions were prepared daily by serial dilutions of the stock solutions in deionized water. 2.2 Chromatographic and AFS conditions Arsenic species in plasma were determined by coupling an HPLC system with HG-AFS as detector (LC-AFS 6500, Beijing Haiguang Instrument Co., Ltd, Beijing, China). The separation of the arsenic species was performed on an anion-exchange column (Hamilton PRP-X100, 250 mm×4.1 mm, 10 µm particle size) at 30oC. The mobile phase was a mixed solution consisting of 13 mM CH3COONa, 3 mM NaH2PO4, 4 mM KNO3 and 0.2 mM EDTA-2Na, at an isocratic elution flow rate of 1.0 mL/min. The column effluent was mixed at a three-way cock with the carrier liquid (10% HCl) and then met with the reductant (3% KBH4 in 0.5% KOH) in another three-way cock, delivered by the peristaltic pump at the same flow rates of 4.2 mL/min, where hydride generation took place. The mixture was separated by the first gas-liquid separator and the arsenic hydrides were carried by argon carrier gas (purity: 99.999%). After being separated by the next gas-liquid separator, the analytes were carried by the argon shielding gas and detected by the AFS detector as shown in Fig.1. The total arsenic concentration in blood was determined by HG-AFS. The HG-AFS system consists of a continuous flow hydride generator coupled with a high-intensity As hollow cathode lamp. The
mixture pre-treated for total arsenic performed the same procedure mentioned above with different parameters as shown in Fig.1. Details of the HG-AFS conditions used in the work are summarized in Table 1. 2.3. Sample preparation 2.3.1. Total arsenic 0.3 g plasma or 0.2 g blood cells (wet weight) were digested by microwave irradiation using MARS6 One-Touch microwave digestion system (CEM Crporation, NC, USA). The samples were accurately weighed into PTFE vessels and added a mixture of 6 mL HNO3 (65%) and 0.5 mL H2O2 (30%). The vessels were sealed and then subjected to a pre-selected program of digestion (raised the temperature to 190oC, then held 2 h). The vessels were opened after cooling, then removed into a dry bath which was heated to 150oC to evaporate the acid until about 0.5 mL solution remained. The residual solution was diluted to 25 mL with 2.5 mL pre-reducing agent (10% ascorbic acid-10% thiourea) and appropriate carrier liquid (10% HCl) in volumetric flasks, then directly analyzed by HG-AFS. 2.3.2. Arsenic speciation 450 μL Plasma samples were accurately weighed and treated with 50 μL of HClO4 (20%) for deproteinization, followed by vortexing for 120 s. The mixture was then centrifuged at 15000 rpm for 15 min at 4oC . 100 µL of the supernatant was injected into the HPLC-HG-AFS system. 50 mg blood cells (wet weight) was accurately weighed and lysed in 490 µL of mobile phase plus 10 µL of ammonia (5%). The mixture was thoroughly vortex-mixed. 500 µL of mobile phase was added to the mixed solution. An aliquot of 450 µL was mixed with 50 µL of HClO4 (20%) for deproteinization. The solution was then centrifuged at 15000 rpm for 15 min at 4 oC. 100 µL aliquots of the supernatant were injected into HPLC-HG-AFS system. 2.4. Calibration standards 2.4.1. Total arsenic The stock solutions of total arsenic were prepared from the standard solution of As by making appropriate dilutions with deionied water at 1000 ng/mL. The series of calibration standards of total arsenic were prepared in 25 mL volumetric flasks with 2.5 mL pre-reducing agent (10% ascorbic acid-10% thiourea) and appropriate carrier liquid (10% HCl). The range was from 0.2 to 20 ng/mL. The different concentration levels samples were prepared in the same way at 3 different concentration
levels of As: 0.5 ng/mL, low (L); 8.0 ng/mL, medium (M); 16 ng/mL, high (H). 2.4.2. Arsenic species The individual stock solutions of As(III), DMA, MMA and As(V) were prepared in water at a concentration of 1000 ng/mL, respectively. The series of mixed working standard solutions were prepared from the stock solutions by making appropriate dilutions with deionied water. The calibration standards were prepared daily by spiking human drug-free plasma with the above working standard solutions covering the ranges from 2.0 to 50 ng/mL for As(III), DMA, MMA and As(V), respectively. Samples were prepared in the same way for As(III), DMA, MMA and As(V) : 2.0 ng/mL, lower limit of quantification (LLOQ); 4.0 ng/mL, L; 20 ng/mL, M; 40 ng/mL, H. The concentrations of arsenic compounds were expressed as the concentrations of arsenic element (As), unless otherwise indicated. 2.5. Method validation 2.5.1. Specificity for arsenic species The specificity of arsenic species was assessed by comparing the chromatograms of drug-free plasma samples from 6 individual donors with those of spiked with arsenic standards. 2.5.2. Linearity and sensitivity Calibration curves of total arsenic were assessed by analyzing a series of standard solutions at concentrations from 0.2 to 20 ng/mL. Calibration curves were plotted as the fluorescence intensity against the corresponding concentration of As. Calibration curves of arsenic species were assessed by analyzing a series of standard samples at concentrations from 2.0 to 50 ng/mL for As(III), DMA, MMA and As(V). Calibration curves were plotted as the peak area of As(III), DMA, MMA and As(V) against the corresponding concentration of the four arsenic species by the analysis of calibration standards, respectively. Correlation coefficient (r) greater than 0.99 demonstrates an adequate linearity. The limits of detection (LOD) of the arsenic compounds were based on the signal-to-noise ratio (S/N) of at least 3. The LLOQ was determined as the lowest concentration of the corresponding calibration curve. 2.5.3. Precision and recovery The intra-batch precision was verified by analyzing five replicates at the L, M and H concentration levels of the same batch. The inter-batch precision was determined by analyzing five replicates at the respective L, M and H concentration levels in three different batches. The spiked
recovery of assay was assessed to investigate the accuracy and reliability of the method. The spiked recovery was carried out by analyzing spiking patient samples at 3 different concentration levels. 2.5.4. Stability of arsenic species The stability of individual stock solution of As(III), DMA, MMA and As(V) was evaluated at 4 oC for 120 days. The autosampler stabilities of the four arsenic species were evaluated at L and H concentration levels in triplicate by reinjecting the samples after storage in the autosampler for 6 h. The room temperature stability of the arsenic compounds was determined at L and H levels concentration in triplicate by analyzing the spiked plasma samples stored at ambient temperature for 6 h. Long-term stability in plasma was evaluated at L and H levels in triplicate on samples stored at -80oC for 60 days. Freeze–thaw stability in plasma was determined at L and H levels concentration in triplicate by conducting three freeze–thaw cycles. The span of time is considered long enough to reveal any possible problems of samples stability during processing procedures and the sample storing. If the stability was within ±15% (bias), the assay was considered stable. 2.6. Application The methods were applied to analyze the plasma and blood cells concentrations of total arsenic and arsenic species in patients who were diagnosed with APL. The study was approved by the Ethics Committee of Harbin Medical University. Informed consent was obtained from each subject. 5 patients with newly diagnosed APL who received ATO (arsenic trioxide injection, Harbin Yida Pharmaceutical Company, Harbin, China) treatment at our hospital were observed. Seafood consumption was prohibited during the therapy. ATO was administered intravenously at a dose of 0.16 mg/kg/day in 0.9% normal saline. ATO was given at the usual rate for 40 min only on the first day of the whole course. Then ATO was given continuously at a very slow rate over 18 h daily. Blood samples were collected in EDTA-containing vacuum collection tubes on the 7th (patient a), 8th (patient b), 11th (patient c), 12th (patient d) and 23th day (patient e) of the slow-rate administration, respectively. 3-5 mL venous blood was collected through a peripheral IV at the time just before the next dose. The blood samples were separated into plasma and blood cells by centrifugation at 4000 rpm for 10 min at 4oC. The samples were immediately frozen at -80oC until analysis. 3. Results and discussion 3.1. Optimization of the chromatographic conditions Because the four arsenic species can exist as anionic form in water, the anion-exchange column
(PRP-X100) was selected, which is commonly employed to separate arsenic speciation. Considering the HG-AFS detection was used in the study, the mobile phase should not only be suitable for the separation of arsenic species on the anion-exchange column but also compatible with the HG-AFS system. In the previous reports [6, 7, 16, 17], phosphate buffer solutions such as KH2PO4, (NH4)2HPO4 and NH4H2PO4 were employed as mobile phase for the ion-exchange separation of arsenic species in HG-AFS. In our study, these phosphate solutions were investigated by varying their concentrations and pHs. However, the separation and retention time of arsenic species were not ideal. To achieve good separation, retention time and LOD, the solution of CH3COONa-NaH2PO4-KNO3-EDTA-2Na was selected and optimized for further experiments, which showed full compatibility with the HPLC-HG-AFS system. A certain amount of arsenic in blood cells and plasma can get lost during the separation procedure [7]. In order to prevent the losses, EDTA-2Na was added into mobile phase. We found the concentration of EDTA-2Na have influence on the retention times of DMA, MMA and As(V). Therefore, the concentration ranged from 0.1-0.8 mM was investigated in our study. 0.2 mM EDTA-2Na was selected, which allowed a satisfying separation of arsenic species in a reasonable time. The influence of mobile phase pH and concentration on separation of arsenic species was investigated. Good separation was obtained for the four arsenic species using the solution of 13 mM CH3COONa - 3 mM NaH2PO4 - 4 mM KNO3 - 0.2 mM EDTA 2Na (pH 6.0). The analysis time was 6 min which is greatly shortened compared with previous reports [6, 7, 17]. The retention times were 2.2, 2.8, 3.5 and 4.3 min for As(III), DMA, MMA and As(V), respectively. Typical chromatograms obtained from drug-free plasma samples, plasma spiked with standards of As(III), DMA, MMA and As(V), plasma sample from the patient with APL are shown in Fig 2. 3.2. Optimization of the HG-AFS system In the HG-AFS system, hydride reaction depended greatly upon the acidity of the reaction. KBH4 was used not only as the reductant but also as the hydrogen supply to maintain the argon-hydrogen flame. It is well established that hydride generation sensitivity is strongly affected by HCl and KBH 4 solutions concentrations, which were optimized in our study. The KBH4 solution concentration was investigated in the range from 1% to 5% (m/v), and the HCl solution concentration was determined from 5% to 15% (v/v) at a pump speed of 60 rpm/min. Lower concentrations of KBH 4 might lead to low fluorescence intensity. Higher concentrations of KBH 4 generated some additional drawbacks, such
as high background signals. Considering the accuracy and sensitivity, a KBH 4 solution of 3.0% (m/v) was selected. The results of test showed that the fluorescence intensity kept stable when the concentrations of HCl solutions were of 10% (v/v). 3.3. Sample preparation Sample pre-treatment is crucial for blood samples due to its complex matrix and low concentration level of arsenic species. As the most frequently used protein precipitant, organic solvents such as acetonitrile and methanol were not suitable for determination of arsenic species due to the interference on As signals with ion-exchange chromatography HG-AFS system as well as the dilution caused by tripling the volume of samples. To solve this problem, Šlejkovec et al. [7] and Zhang et al. [17] conducted a process of evaporation followed by reconstitution to solve the problems. However, stable and reliable recoveries for the arsenic species were not obtained by investigating the method in the present study. Besides, the sample preparations steps were time-consuming and complicated, which lead to poor recoveries [7, 17]. In our study, moderate HClO4 solution (20%, 0.1 volume) was used as protein precipitant for sample preparation, which was demonstrated to be fast, simple and robust in comparison with the previous studies. 3.4. Method validation results People are usually exposed to arsenic through consumption of foods. The arsenic is absorbed and distributed throughout the body once ingested by foods. Therefore, it is difficult to find blank blood samples for method validation. Considering the consumption of seafood is the major source of dietary arsenic, seafood consumption was prohibited before 1 month when drug-free blood samples were collected from donors. Calibration curves of arsenic species were compared between spiked drug-free plasma and aqueous standards. There was an obvious difference in the curve slopes that varied from day-to-day, indicating an apparent matrix effect. Thus, matrix matched calibration was necessary for the assay of arsenic species in plasma. The typical equations are displayed in Table 2. The ranges were 0.2-20 ng/mL for total arsenic in solution and 2.0-50 ng/mL for four arsenic compounds in plasma (r >0.995). The ranges are suitable for the determination of the arsenic compounds in patients with APL. The reported therapeutic and toxic concentration of arsenic in human blood-plasma/serum is 2-70 and 50-250 ng/mL, respectively [18]. In a clinical pharmacokinetic study by Fukai et al. [19], ATO was administered intravenously at a dose of 0.08 mg/kg/day in a patient, the serum concentration ranges of 18.2-40.7 for
total arsenic, 5.8-13.1 ng/mL for As(III), 5.1-14.4 ng/mL for DMA and 5.5-13.2 ng/mL for MMA. In another clinical pharmacokinetic study by Fox et al. [20], ATO was administered intravenously at a recommended dose of 0.15 mg/kg/day, the patients plasma concentration of As(III) ranged from 5 to 28 ng/mL. Therefore, it is concluded that HPLC-HG-AFS method might be sensitive enough for monitoring the speciation of ATO metabolites in blood samples in pharmacokinetic, therapeutic and toxicological studies. Five replicates of LLOQ samples of arsenic species were prepared and analyzed, respectively. The relative standard deviation (RSD%) data for precision was within ±20%. The corresponding limits of detection estimated in water were very low, which were 0.12, 0.19, 0.20 and 0.27 ng/mL for As(III), DMA, MMA and As(V), respectively. The precision of total arsenic can be accepted. The coefficients of variation for intra- and inter-batch precision of total arsenic ranged from 0.89% to 4.54% and 0.62% to 7.32%, respectively. The results demonstrated the coefficients of variation for intra- and inter-batch precision of arsenic species in plasma ranged from 1.82% to 9.33% and 3.08% to 12.49%, respectively. A critical problem of the present assay is that there is no standard reference material for plasma and blood cells. Therefore, it is impossible to investigate the accuracy by checking the certified reference materials. To solve the problem, the spiked arsenicals recovery technique was used for quality assurance. Standards were spiked to the samples from the ATO-treated patient with APL. The samples were prepared in the same method of corresponding protocols above-mentioned. Spiked recovery (%) = [(Measured value-Background value)/Spiked value]×100%. The results of spiked recovery were summarized in Table 3. The RSD% of each concentration levels was less than 11.90%. Good accuracy and repeatability were obtained for all the analytes at each concentration levels with the spiked recoveries. Potential inter-conversion among different arsenic species is a problem during the storage. Therefore, the stability of As(III), DMA, MMA and As(V) was investigated in the present study. The stability test results are presented in Table 4. The results demonstrated that the individual arsenic species in the stock solutions were stable at 4oC for at least 120 days. The average bias of accuracy ranged from -5.35% to 6.80% for spiked plasma samples placed at room temperature for 6 h. The average bias of accuracy ranged from -3.69% to -4.37% for extracted plasma samples at ambient temperature for 6 h. The freeze-thaw stability bias for As(III), DMA, MMA and As(V) in plasma was from -9.05% to -3.88% after three cycles. The long-term stability in plasma was from -4.11% to 7.52%. The results demonstrated the arsenic species were stable during the sample storing and processing
procedures. 3.5. Application Arsenic concentrations in blood samples from 5 patients with APL are shown in Table 5. The total arsenic ranged from 114.3 ng/g to 328.9 ng/g in blood cells and from 28.3 ng/g to 59.2 ng/g in plasma, respectively, indicating remarkable individual variation. The results demonstrate that the total arsenic concentrations in the blood cells were over 4 times higher than those in the plasma, which is consistent with the previous reports by Yuta Yoshino [8]. The results also suggest that most of arsenic is present in blood cells of the APL patients, indicating that the profiles of arsenic species in blood cells should be noticed. The ratio of the sum concentration of arsenic species to the total arsenic concentration in plasma was more than 80%. The results suggest that the extraction recovery of the method for arsenic species in plasma was higher than 80%, which was higher than in other reports [8, 9]. Thus, the sample preparation in our study is suitable for the determination of arsenic species in plasma. In agreement with previous reports [7-9], the sum of methylated metabolites (DMA+MMA) was higher than the sum of inorganic arsenic (As(III)+As(V)) during the therapy period, suggesting that methylated metabolites (DMA and MMA) were major metabolites in plasma. As(III) can be determined in our study while no As(III) can be seen in a previous report [7] as shown in Fig 3. The reasons for these results may be the prevention of arsenic losses from separation procedure or the individual variation. In
our
study,
the
As(III)>DMA>MMA>As(V)
concentrations for
patient
of
arsenic
compounds
ranked
a,
MMA>As(III)>DMA>As(V)
in for
the
order
patient
b,
DMA>MMA>As(III)>As(V) for patient c and patient e, MMA>DMA>As(III)>As(V) for patient d as shown in Fig 3. These results indicate large variations among individual patients in metabolic efficiency and methylation levels. These inter-individual differences in total arsenic and arsenic speciation were probably due to different age, treatment duration or pathological state. But we should notice that genetic polymorphisms in arsenic metabolism genes, such as arsenic (+III oxidation state) methyltransferase genes (AS3MT), is considered to be related to individual differences in the arsenic metabolism [21]. These results also suggest that continuous efforts to understand the interindividual variations in arsenic metabolism and the relationship between factors and arsenic metabolism are greatly useful for proposing effective therapy protocols for individual patients with APL.
In the present study, we attempted to apply the method to determine the arsenic species in blood cells. The results show that the sum of four arsenic species was significantly lower than that the total arsenic after digestion in blood cells from the same patient. The problem may be explained by the presumption that the major arsenic in blood cells is likely bound to hemoglobin [14, 22]. The analytical method of arsenic species in blood cells was problematic. The protein-bound arsenicals were not released from hemoglobin during the sample preparation procedure and produced precipitation losses. As far as the analytical problems are concerned, we would focused on accurately determining the arsenic species in blood cells of patients with APL by cleaving the bonds between As and hemoglobin in the following study. 4. Conclusion Despite the methods of total arsenic and arsenic species being widely developed, arsenic is not routinely afforded as a determination in clinical laboratory. The methods of HG-AFS and HPLC-HG-AFS were established and validated for the determination of total arsenic in whole blood and arsenic species of As(III), DMA, MMA and As(V) in plasma, respectively. The methods were applied to the assay of arsenic compounds in plasma and blood cells from patients with APL. The methods were found to be simple, low cost and suitable for the determination of total arsenic and arsenic species as well as could be easily implemented in clinical laboratory. This approach provided the basis for further clinical research.
Acknowledgements
This project was supported by Heilongjiang Province Science Foundation for Youths (No. QC2015119) and Foundation of The First Affiliated Hospital of Harbin Medical University (No. 2012lx003).
References
[1] Z. Wang, Z. Chen,;1; Acute promyelocytic leukemia: from highly fatal to highly curable, Blood. 111 (2008) 2505-2515. [2] W.R. Cullen,;1; Chemical Mechanism of Arsenic Biomethylation, Chem Res Toxicol. 27 (2014) 457-461. [3] K. Rehman, H. Naranmandura,;1; Arsenic metabolism and thioarsenicals, Metallomics. 4 (2012) 881-892. [4] B. Yuan, Y. Yoshino, T. Kaise,;1; H. Toyoda, Application of arsenic trioxide therapy for patients with leukaemia, Biological Chemistry of Arsenic, Antimony and Bismuth. (2011) 263-292. [5] H. Naranmandura, M. W. Carew, S. Xu, J. Lee, E.M. Leslie, M.Weinfeld, X.C. Le,;1; Comparative toxicity of arsenic metabolites in human bladder cancer EJ-1 cells, Chem Res Toxicol. 24 (2011) 1586-1596. [6] Z.;1; Šlejkovec, H. Podgornik, P. Černelč, I. Falnoga, Exceptions in patterns of arsenic compounds in urine of acute promyelocytic leukaemia patients treated with As2O3, Biometals. 29 (2016) 107-118. [7] Z.;1; Šlejkovec, I. Falnoga, W. Goessler, J.T. van Elteren, R. Raml , H. Podgornik, P. Černelč, Analytical artefacts in the speciation of arsenic in clinical samples, Analytica Chimica Acta. 607 (2008) 83-91. [8] Y. Yoshino, B. Yuan, S. I. Miyashita, N. Iriyama, A. Horikoshi, O. Shikino, H. Toyoda, T. Kaise,;1; Speciation of arsenic trioxide metabolites in blood cells and plasma of a patient with acute promyelocytic leukemia, Anal. Bioanal. Chem. 393 (2009) 689-697. [9] N. Iriyama1, Y. Yoshino, B. Yuan, A. Horikoshi, Y. Hirabayashi, Y. Hatta1,H. Toyoda, J. Takeuchi1.;1; Speciation of arsenic trioxide metabolites in peripheral blood and bone marrow from an acute promyelocytic leukemia patient, J. Hematol. Oncol. 5 (2012) 1-11. [10] S. Fujisawa, R. Ohno, K. Shigeno, et al.;1; Pharmacokinetics of arsenic species in Japanese patients with relapsed or refractory acute promyelocytic leukemia treated with arsenic trioxide, Cancer Chemoth. Pharmacol. 59 (2007) 485-493. [11] K.C. Hsu,;1; C.C. Sun, Y.L. Huang, Arsenic speciation in biomedical sciences: Recent advances and applications, Kaohsiung J. Med. Sci. 7 (2011) 82-389. [12] A.L.;1; Lindberg , W. Goessler, M. Grandér, B. Nermell, M. Vahter, Evaluation of the three most commonly used analytical methods for determination of inorganic arsenic and its metabolites in urine,
Toxicol Lett. 168 (2007) 310-318. [13] K. Ito,;1; C.D. Palmer, A.J. Steuerwald, P.J. Parsons, Determination of five arsenic species in whole blood by liquid chromatography coupled with inductively coupled plasma mass spectrometry, J. Anal. At. Spectrom. 25 (2010) 1334-1342. [14] B. Chen, X. Lu, S. Shen, L.L. Arnold, S.M. Cohen, X. C. Le,;1; Arsenic speciation in the blood of arsenite-treated F344 rats, Chem. Res. Toxicol. 26 (2013) 952-962. [15] M. Welna, A. Szymczycha-Madeja, P. Pohl,;1; Comparison of strategies for sample preparation prior to spectrometric measurements for determination and speciation of arsenic in rice, Trends Anal. Chem. 65 (2015) 122-136. [16] C. Wei, J. Liu,;1; A new hydride generation system applied in determination of arsenic species with ion chromatography-hydride generation-atomic fluorescence spectrometry (IC-HG-AFS), Talanta. 73 (2007) 540-545. [17] Y.J. Zhang, S.P. Qiang, J. Sun, M. Song, T.J. Hang,;1; Liquid chromatography-hydride generation-atomic fluorescence spectrometry determination of arsenic species in dog plasma and its application to a pharmacokinetic study after oral administration of Realgar and Niu Huang Jie Du Pian, J Chromatogr B, 917-918 (2013) 93-99. [18] M. Schulz, A. Schmoldt,;1; Therapeutic and toxic blood concentrations of more than 800 drugs and other xenobiotics, Pharmazie. 58 (2003) 447-474. [19] Y. Fukai，M. Hirata，M. Ueno，N. Ichikawa，H. Kobayashi，H. Saitoh, T. Sakurai;1; , K. Kinoshita, T. Kaise , S. Ohta, Clinical pharmacokinetic study of arsenic trioxide in an acute promyelocytic leukemia (APL) patient: speciation of arsenic metabolites in serum and urine, Biol. Pharm. Bull. 29 (2006) 1022-1027. [20] E. Fox,;1; B.I. Razzouk, B.C. Widemann, S. Xiao, M. O'Brien, W. Goodspeed, G.H. Reaman, S.M. Blaney, A.J. Murgo, F.M. Balis, P.C. Adamson, Phase 1 trial and pharmacokinetic study of arsenic trioxide in children and adolescents with refractory or relapsed acute leukemia, including acute promyelocytic leukemia or lymphoma, Blood. 111 (2008) 566-573. [21] K. Engström, M.Vahter, S.J. Mlakar, G.Concha, B. Nermell, R. Raqib, A.Cardozo, K. Broberg,;1; Polymorphisms in arsenic (+III oxidation state) methyltransferase (AS 3MT) predict gene expression of AS3MT as well as arsenic metabolism, Environ. Health Perspect. 119 (2011) 182-188. [22] S. Shen, M. Lu, X. F. Li, W. R. Cullen, M. Weinfeld, X. C. Le,;1; Arsenic binding to proteins, Chem. Rev. 113 (2013) 7769-7792.
Figure Legends
Fig. 1. Schematic diagram of HPLC-HG-AFS system used for the determination of the total arsenic and arsenic species, respectively.
Fig. 2. Representative HPLC-HG-AFS chromatograms: drug-free plasma (A); drug-free plasma spiked with standards of the four arsenic species at LLOQ (B1) and 10.0 ng/mL (B2) (As(III): tR=2.2 min; DMA: tR=2.8 min; MMA: tR=3.5 min; As(V): tR=4.3 min); real subject plasma sample (patient b) (C). The four arsenic species were separated on an anion exchange column (Hamilton PRP-X100) with
13 mM CH3COONa - 3 mM NaH2PO4 - 4 mM KNO3 - 0.2 mM EDTA-2Na (pH 6.0) as a mobile phase.
Fig. 3. Arsenic species concentration in plasma of five APL patients (a, b, c, d and e) at different stages of the therapy. The length of the bar represents the total arsenic concentration in plasma with the sub-divisions for the individual arsenic species quantified; * indicates the gap between sum of species and total arsenic concentration.
Table 1 Details of the HG-AFS conditions Total arsenic Pre-reducing agent (freshly prepared) Reductant (freshly prepared)
Arsenic speciation
10% ascorbic acid (m/v) - 10% thiourea (m/v) 3% KBH4 (m/v) in 0.5% KOH (m/v)
3% KBH4 (m/v) in 0.5% KOH (m/v)
10% HCl (v/v)
10% HCl (v/v)
Primary current (mA)
70
80
Boost current (mA)
35
40
Carrier liquid
Pump speed (rpm/min)
80
60
Shielding gas flow rate (mL/min)
800
900
Carrier gas flow rate (mL/min)
300
300
Negative high voltage (V)
280
300
Table 2 Calibration curves and limits of detection
Total As As(III) DMA MMA As(V)
Linearity range (ng/mL)
Typical equations
r
Limits of detection (ng/mL)
0.2-20
A= 84.2 + 145.9C
0.9997
0.08
2.0-50
Y=1.178×105+
2.184×105X
0.9996
0.12
2.0-50
Y=3.714×103+
1.778×105X
0.9965
0.19
2.0-50
Y=4.527×104+
2.287×105X
0.9989
0.20
2.0-50
Y=1.259×105+
1.822×105X
0.9986
0.27
Table 3 Spiked recovery of total As in pasma and blood cells and arsenic species in plasama Background value (ng/g)
Spiked value
Spike recovery (%)
50 ng/g
89.4
100 ng/g
106.3
200 ng/g
94.3
20 ng/g
87.6
50 ng/g
92.6
100 ng/g
100.2
5 ng/mL
100.4
10 ng/mL
92.4
20 ng/mL
101.5
5 ng/mL
85.6
Blood cells
Total As
138.8
Plasma
Total As
As(III)
DMA
MMA
As(V)
36.7
7.39
8.79
8.96
5.03
10 ng/mL
96.3
20 ng/mL
104.3
5 ng/mL
108.6
10 ng/mL
99.6
20 ng/mL
98.7
5 ng/mL
103.9
10 ng/mL
81.2
20 ng/mL
95.3
Total As = concentration of total arsenic after digestion
Table 4 Stability data of As(III), DMA, MMA and As(V) in plasma
Ambient
Stock solutions Arsenic
(4oC,
120 d)
species RSD
Accuracy
(%)
Bias (%)
As(III)
0.59
0.25
DMA
1.08
0.15
MMA
1.46
-1.05
As(V)
2.10
-1.15
Initial
(room temperature,
concentration
6 h)
(ng/mL)
Ambient
Freeze-thaw
(autosamper, 6 h)
(-80°C, 3 cycles)
60 d (-80°C)
RSD
Accuracy
RSD
Accuracy
RSD
Accuracy
RSD
Accuracy
(%)
Bias (%)
(%)
Bias (%)
(%)
Bias (%)
(%)
Bias (%)
4.05
6.04
-5.35
10.73
-2.88
5.90
-9.05
8.28
-4.11
39.8
3.04
-3.60
7.55
1.26
4.05
-3.01
3.06
0.75
4.46
2.22
1.04
9.20
0.37
7.30
-1.93
10.64
1.35
41.5
2.38
-2.81
0.52
-3.69
2.27
0.80
1.23
-3.69
3.95
5.35
2.53
5.78
0.93
4.71
-3.12
0.86
3.37
39.7
4.09
-1.87
7.57
4.37
5.78
1.85
3.50
0.92
4.12
7.75
6.80
12.03
2.59
10.33
3.88
6.24
7.52
40.8
1.93
2.78
0.51
0.82
3.28
1.88
6.11
2.45
Table 5 The concentrations of arsenic in blood samples collected from patients with APL Concentrations of arsenic (ng/g) Patients
Patient a
Patient b
Patient c
Patient d
Patient e
Age, sex
9, female
55, female
46, female
30, female
31, male
Time
Day-7
Day-8
Day-11
Day-12
Day-23
114.3
172.3
328.9
138.8
186.9
Total As
28.3
38.1
59.2
36.7
39.0
As(III)
7.75
7.98
10.71
7.39
6.59
DMA
7.14
7.47
20.21
8.79
13.22
MMA
5.46
13.05
16.87
8.96
9.44
As(V)
3.58
4.61
6.51
5.03
3.24
tAs
23.93
33.11
54.3
30.17
32.49
Blood cells Total As Plasma
Total As = concentration of total arsenic after digestion; tAs = the sum of arsenic species
S0731-7085(17)31282-7 http://dx.doi.org/doi:10.1016/j.jpba.2017.07.001 PBA 11378
To appear in:
Journal of Pharmaceutical and Biomedical Analysis
Received date: Revised date: Accepted date:
23-5-2017 29-6-2017 1-7-2017
Please cite this article as: Meihua Guo, Wenjing Wang, Xin Hai, Jin Zhou, HPLC-HG-AFS determination of arsenic species in acute promyelocytic leukemia (APL) plasma and blood cells, Journal of Pharmaceutical and Biomedical Analysishttp://dx.doi.org/10.1016/j.jpba.2017.07.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
SHPLC-HG-AFS determination of arsenic species in acute promyelocytic leukemia (APL) plasma and blood cells Meihua Guoa1, Wenjing Wanga1, Xin Haia*, Jin Zhoub,c,* aDepartment
of Pharmacy, the First Affiliated Hospital, Harbin Medical University, 23 Youzheng Street, Nangang
District, Harbin, 150001, China bDepartment
of Hematology, the First Affiliated Hospital, Harbin Medical University, 23 Youzheng Street,
Nangang District, Harbin, 150001, China cHeilongjiang
1These
Institute of Hematology & Oncology, 23 Youzheng Street, Nangang District, Harbin,150001, China
authors contributed equally to this work.
*Corresponding
author
Jin Zhou, Department of Hematology, the First Hospital, Harbin Medical University, 23 Youzheng Street, Nangang District, Harbin, 150001, China E-mail address: [email protected] Xin Hai, Department of Pharmacy, the First Hospital, Harbin Medical University, 23 Youzheng Street, Nangang District, Harbin, 150001, China E-mail address: [email protected]
Highlights
The methods of HG-AFS and HPLC-HG-AFS were established and validated for quantitative analysis of total arsenic in whole blood and arsenic species of As(III), DMA, MMA and As(V) in plasma, respectively.
Sample preparation for speciation involved only a simple deproteinization procedure with HClO4 and the four arsenic species was separated using HILIC column within 6 min.
The methods have been successfully applied to the assay of arsenic compounds in plasma and blood cells from patients with acute promyelocytic leukemia, which were found to be simple, low cost as well as could be easily implemented in clinical laboratory.
Abstract Arsenic trioxide (ATO) has been successfully used in the treatment of acute promyelocytic leukemia
(APL). To clarify the arsenic species in APL patients, high performance liquid chromatography-hydride generation-atomic fluorescence spectrometry (HPLC-HG-AFS) and HG-AFS methods were developed and validated to quantify the plasma concentrations of inorganic arsenic (As(III) and As(V)) and methylated metabolites (MMA and DMA), and the total amounts of arsenic in blood cells and plasma. Blood cells and plasma were digested with mixtures of HNO3-H2O2 and analyzed by HG-AFS. For arsenic speciation, plasma samples were prepared with perchloric acid to precipitate protein. The supernatant was separated on an anion-exchange column within 6 min with isocratic elution using 13 mM CH3COONa, 3 mM NaH2PO4, 4 mM KNO3 and 0.2 mM EDTA-2Na. The methods provided linearity range of 0.2-20 ng/mL for total arsenic and 2.0-50 ng/mL for four arsenic species. The developed methods for total arsenic and arsenic species determination were precise and accurate. The spiked recoveries ranged from 81.2%-108.6% and the coefficients of variation for intra- and inter-batch precision were less than 9.3% and 12.5%, respectively. The developed methods were applied successfully for the assay of total arsenic and arsenic species in 5 APL patients. The HPLC-HG-AFS may be a good alternative for arsenic species determination in APL patients with its simplicity and low-cost in comparison with HPLC-ICP-MS.
Abbreviations: ATO, arsenic trioxide; APL, acute promyelocytic leukemia; HPLC-HG-AFS, high performance liquid chromatography-hydride generation-atomic fluorescence spectrometry; As(III), arsenite; DMA, dimethylarsinic acid; MMA, monomethylarsonic acid; As(V), arsenate.
Keywords High performance liquid chromatography-hydride generation-atomic fluorescence spectrometry Arsenic speciation Acute promyelocytic leukemia Plasma Blood cells
1. Introduction Arsenic trioxide (ATO, arsenite (As(III)) in solution) has been successfully used in the treatment of
patients with acute promyelocytic leukemia (APL) [1]. As firmly established, As(III) can be transformed into the major metabolites monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA) by the metabolism that involves oxidation and oxidative methylation by a series of enzymes [2, 3]. It is essential to determine inorganic arsenic as well as arsenic metabolites, namely DMA and MMA in patients with APL, revealing the relationship between arsenic speciation and clinical response. Because of the different toxic effects induced by different As species [3-5], speciation analysis is critical when evaluating the therapeutic efficiency of arsenic t7rioxide for treating individual with APL. Although a few studies concerning the arsenic speciation in urine and plasma of APL were reported [6-10], the detailed systematic analysis of the arsenic species in blood cells of APL patients has not yet been performed. Inductively coupled plasma mass spectrometry (ICP-MS) has been chosen by many authors as a chromatographic detector system in the measurements of total arsenic or arsenic species, because of the rapidity and very high sensitivity of measurements [8-12]. Considering the samples to be analyzed in most hospitals, a fast, simple and low cost analytical method is more feasible to make it perform easily. However, both the cost of ICP-MS instrumentation as well as its running costs are quite high that limit its applications for routine analysis in many clinical laboratories. In addition, the determination of arsenic at m/z of 75 by ICP-MS could be compromised by
40
Ar35Cl+ and
38
Ar37Cl+ [8, 13-15]. It is
therefore necessary to overcome the polyatomic interference by taking some measurements, which increase complexity. As an alternative, hydride generation-atomic fluorescence spectrometry (HG-AFS) has been described to be similar to ICP-MS for arsenic speciation analysis in linear range and sensitivity [6, 12, 16]. The cost of the HG-AFS instrumentation is about 1/10 of that for most ICP-MS instruments and it has lower running costs [12]. AFS detects arsenic by a high-intensitive arsenic lamp, avoiding the interferences from ArCl. The application of HG is beneficial for quantifying arsenic, since it separates the analytes from the sample matrix, minimizes or reduces chemical and spectral interferences [15]. Recently, HG-AFS has been applied for determination of arsenic compounds in urine samples of APL patients [6]. Few analytical methods have been performed to determine total arsenic and arsenic species in whole blood of APL by HG-AFS. An HPLC-HG-AFS method has been used by Šlejkovec et al. [7] for the determination of arsenic species in blood samples with a complex sample treatment, yet no validation data have been provided to prove its sensitivity and specificity. The aim of the present
study was to develop a fast, accurate and low cost analytical method by HPLC-HG-AFS for total arsenic and arsenic speciation in APL whole blood. Special attention was focused on all aspects of the analytical procedure to simplify sample preparation and retrieve as high quality data as possible. The methods were fully validated and applied in APL patients following ATO treatment. HPLC-HG-AFS is approved to be a feasible alternative for arsenic speciation study which would be helpful in understanding the relationship between arsenic concentrations and clinical response. 2. Materials and methods 2.1. Chemicals and Reagents The chemicals used in this work were of guarantee grade and were all purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). The deionized water was obtained from a water purification system (Thermo Genpure, USA) and always freshly prepared. Stock solutions were prepared using the following standard As and arsenic compounds: As (GBW(E)080117), arsenite (As(III), GBW08666), arsenate (As(V), GBW08667), monomethylarsonic acid (MMA, GBW08668) and dimethylarsinic acid (DMA, GBW08669) purchased from National Institute of Metrology (Beijing, China). Stock solutions were prepared with fresh deionized water and stored in the dark at 4oC. Working solutions were prepared daily by serial dilutions of the stock solutions in deionized water. 2.2 Chromatographic and AFS conditions Arsenic species in plasma were determined by coupling an HPLC system with HG-AFS as detector (LC-AFS 6500, Beijing Haiguang Instrument Co., Ltd, Beijing, China). The separation of the arsenic species was performed on an anion-exchange column (Hamilton PRP-X100, 250 mm×4.1 mm, 10 µm particle size) at 30oC. The mobile phase was a mixed solution consisting of 13 mM CH3COONa, 3 mM NaH2PO4, 4 mM KNO3 and 0.2 mM EDTA-2Na, at an isocratic elution flow rate of 1.0 mL/min. The column effluent was mixed at a three-way cock with the carrier liquid (10% HCl) and then met with the reductant (3% KBH4 in 0.5% KOH) in another three-way cock, delivered by the peristaltic pump at the same flow rates of 4.2 mL/min, where hydride generation took place. The mixture was separated by the first gas-liquid separator and the arsenic hydrides were carried by argon carrier gas (purity: 99.999%). After being separated by the next gas-liquid separator, the analytes were carried by the argon shielding gas and detected by the AFS detector as shown in Fig.1. The total arsenic concentration in blood was determined by HG-AFS. The HG-AFS system consists of a continuous flow hydride generator coupled with a high-intensity As hollow cathode lamp. The
mixture pre-treated for total arsenic performed the same procedure mentioned above with different parameters as shown in Fig.1. Details of the HG-AFS conditions used in the work are summarized in Table 1. 2.3. Sample preparation 2.3.1. Total arsenic 0.3 g plasma or 0.2 g blood cells (wet weight) were digested by microwave irradiation using MARS6 One-Touch microwave digestion system (CEM Crporation, NC, USA). The samples were accurately weighed into PTFE vessels and added a mixture of 6 mL HNO3 (65%) and 0.5 mL H2O2 (30%). The vessels were sealed and then subjected to a pre-selected program of digestion (raised the temperature to 190oC, then held 2 h). The vessels were opened after cooling, then removed into a dry bath which was heated to 150oC to evaporate the acid until about 0.5 mL solution remained. The residual solution was diluted to 25 mL with 2.5 mL pre-reducing agent (10% ascorbic acid-10% thiourea) and appropriate carrier liquid (10% HCl) in volumetric flasks, then directly analyzed by HG-AFS. 2.3.2. Arsenic speciation 450 μL Plasma samples were accurately weighed and treated with 50 μL of HClO4 (20%) for deproteinization, followed by vortexing for 120 s. The mixture was then centrifuged at 15000 rpm for 15 min at 4oC . 100 µL of the supernatant was injected into the HPLC-HG-AFS system. 50 mg blood cells (wet weight) was accurately weighed and lysed in 490 µL of mobile phase plus 10 µL of ammonia (5%). The mixture was thoroughly vortex-mixed. 500 µL of mobile phase was added to the mixed solution. An aliquot of 450 µL was mixed with 50 µL of HClO4 (20%) for deproteinization. The solution was then centrifuged at 15000 rpm for 15 min at 4 oC. 100 µL aliquots of the supernatant were injected into HPLC-HG-AFS system. 2.4. Calibration standards 2.4.1. Total arsenic The stock solutions of total arsenic were prepared from the standard solution of As by making appropriate dilutions with deionied water at 1000 ng/mL. The series of calibration standards of total arsenic were prepared in 25 mL volumetric flasks with 2.5 mL pre-reducing agent (10% ascorbic acid-10% thiourea) and appropriate carrier liquid (10% HCl). The range was from 0.2 to 20 ng/mL. The different concentration levels samples were prepared in the same way at 3 different concentration
levels of As: 0.5 ng/mL, low (L); 8.0 ng/mL, medium (M); 16 ng/mL, high (H). 2.4.2. Arsenic species The individual stock solutions of As(III), DMA, MMA and As(V) were prepared in water at a concentration of 1000 ng/mL, respectively. The series of mixed working standard solutions were prepared from the stock solutions by making appropriate dilutions with deionied water. The calibration standards were prepared daily by spiking human drug-free plasma with the above working standard solutions covering the ranges from 2.0 to 50 ng/mL for As(III), DMA, MMA and As(V), respectively. Samples were prepared in the same way for As(III), DMA, MMA and As(V) : 2.0 ng/mL, lower limit of quantification (LLOQ); 4.0 ng/mL, L; 20 ng/mL, M; 40 ng/mL, H. The concentrations of arsenic compounds were expressed as the concentrations of arsenic element (As), unless otherwise indicated. 2.5. Method validation 2.5.1. Specificity for arsenic species The specificity of arsenic species was assessed by comparing the chromatograms of drug-free plasma samples from 6 individual donors with those of spiked with arsenic standards. 2.5.2. Linearity and sensitivity Calibration curves of total arsenic were assessed by analyzing a series of standard solutions at concentrations from 0.2 to 20 ng/mL. Calibration curves were plotted as the fluorescence intensity against the corresponding concentration of As. Calibration curves of arsenic species were assessed by analyzing a series of standard samples at concentrations from 2.0 to 50 ng/mL for As(III), DMA, MMA and As(V). Calibration curves were plotted as the peak area of As(III), DMA, MMA and As(V) against the corresponding concentration of the four arsenic species by the analysis of calibration standards, respectively. Correlation coefficient (r) greater than 0.99 demonstrates an adequate linearity. The limits of detection (LOD) of the arsenic compounds were based on the signal-to-noise ratio (S/N) of at least 3. The LLOQ was determined as the lowest concentration of the corresponding calibration curve. 2.5.3. Precision and recovery The intra-batch precision was verified by analyzing five replicates at the L, M and H concentration levels of the same batch. The inter-batch precision was determined by analyzing five replicates at the respective L, M and H concentration levels in three different batches. The spiked
recovery of assay was assessed to investigate the accuracy and reliability of the method. The spiked recovery was carried out by analyzing spiking patient samples at 3 different concentration levels. 2.5.4. Stability of arsenic species The stability of individual stock solution of As(III), DMA, MMA and As(V) was evaluated at 4 oC for 120 days. The autosampler stabilities of the four arsenic species were evaluated at L and H concentration levels in triplicate by reinjecting the samples after storage in the autosampler for 6 h. The room temperature stability of the arsenic compounds was determined at L and H levels concentration in triplicate by analyzing the spiked plasma samples stored at ambient temperature for 6 h. Long-term stability in plasma was evaluated at L and H levels in triplicate on samples stored at -80oC for 60 days. Freeze–thaw stability in plasma was determined at L and H levels concentration in triplicate by conducting three freeze–thaw cycles. The span of time is considered long enough to reveal any possible problems of samples stability during processing procedures and the sample storing. If the stability was within ±15% (bias), the assay was considered stable. 2.6. Application The methods were applied to analyze the plasma and blood cells concentrations of total arsenic and arsenic species in patients who were diagnosed with APL. The study was approved by the Ethics Committee of Harbin Medical University. Informed consent was obtained from each subject. 5 patients with newly diagnosed APL who received ATO (arsenic trioxide injection, Harbin Yida Pharmaceutical Company, Harbin, China) treatment at our hospital were observed. Seafood consumption was prohibited during the therapy. ATO was administered intravenously at a dose of 0.16 mg/kg/day in 0.9% normal saline. ATO was given at the usual rate for 40 min only on the first day of the whole course. Then ATO was given continuously at a very slow rate over 18 h daily. Blood samples were collected in EDTA-containing vacuum collection tubes on the 7th (patient a), 8th (patient b), 11th (patient c), 12th (patient d) and 23th day (patient e) of the slow-rate administration, respectively. 3-5 mL venous blood was collected through a peripheral IV at the time just before the next dose. The blood samples were separated into plasma and blood cells by centrifugation at 4000 rpm for 10 min at 4oC. The samples were immediately frozen at -80oC until analysis. 3. Results and discussion 3.1. Optimization of the chromatographic conditions Because the four arsenic species can exist as anionic form in water, the anion-exchange column
(PRP-X100) was selected, which is commonly employed to separate arsenic speciation. Considering the HG-AFS detection was used in the study, the mobile phase should not only be suitable for the separation of arsenic species on the anion-exchange column but also compatible with the HG-AFS system. In the previous reports [6, 7, 16, 17], phosphate buffer solutions such as KH2PO4, (NH4)2HPO4 and NH4H2PO4 were employed as mobile phase for the ion-exchange separation of arsenic species in HG-AFS. In our study, these phosphate solutions were investigated by varying their concentrations and pHs. However, the separation and retention time of arsenic species were not ideal. To achieve good separation, retention time and LOD, the solution of CH3COONa-NaH2PO4-KNO3-EDTA-2Na was selected and optimized for further experiments, which showed full compatibility with the HPLC-HG-AFS system. A certain amount of arsenic in blood cells and plasma can get lost during the separation procedure [7]. In order to prevent the losses, EDTA-2Na was added into mobile phase. We found the concentration of EDTA-2Na have influence on the retention times of DMA, MMA and As(V). Therefore, the concentration ranged from 0.1-0.8 mM was investigated in our study. 0.2 mM EDTA-2Na was selected, which allowed a satisfying separation of arsenic species in a reasonable time. The influence of mobile phase pH and concentration on separation of arsenic species was investigated. Good separation was obtained for the four arsenic species using the solution of 13 mM CH3COONa - 3 mM NaH2PO4 - 4 mM KNO3 - 0.2 mM EDTA 2Na (pH 6.0). The analysis time was 6 min which is greatly shortened compared with previous reports [6, 7, 17]. The retention times were 2.2, 2.8, 3.5 and 4.3 min for As(III), DMA, MMA and As(V), respectively. Typical chromatograms obtained from drug-free plasma samples, plasma spiked with standards of As(III), DMA, MMA and As(V), plasma sample from the patient with APL are shown in Fig 2. 3.2. Optimization of the HG-AFS system In the HG-AFS system, hydride reaction depended greatly upon the acidity of the reaction. KBH4 was used not only as the reductant but also as the hydrogen supply to maintain the argon-hydrogen flame. It is well established that hydride generation sensitivity is strongly affected by HCl and KBH 4 solutions concentrations, which were optimized in our study. The KBH4 solution concentration was investigated in the range from 1% to 5% (m/v), and the HCl solution concentration was determined from 5% to 15% (v/v) at a pump speed of 60 rpm/min. Lower concentrations of KBH 4 might lead to low fluorescence intensity. Higher concentrations of KBH 4 generated some additional drawbacks, such
as high background signals. Considering the accuracy and sensitivity, a KBH 4 solution of 3.0% (m/v) was selected. The results of test showed that the fluorescence intensity kept stable when the concentrations of HCl solutions were of 10% (v/v). 3.3. Sample preparation Sample pre-treatment is crucial for blood samples due to its complex matrix and low concentration level of arsenic species. As the most frequently used protein precipitant, organic solvents such as acetonitrile and methanol were not suitable for determination of arsenic species due to the interference on As signals with ion-exchange chromatography HG-AFS system as well as the dilution caused by tripling the volume of samples. To solve this problem, Šlejkovec et al. [7] and Zhang et al. [17] conducted a process of evaporation followed by reconstitution to solve the problems. However, stable and reliable recoveries for the arsenic species were not obtained by investigating the method in the present study. Besides, the sample preparations steps were time-consuming and complicated, which lead to poor recoveries [7, 17]. In our study, moderate HClO4 solution (20%, 0.1 volume) was used as protein precipitant for sample preparation, which was demonstrated to be fast, simple and robust in comparison with the previous studies. 3.4. Method validation results People are usually exposed to arsenic through consumption of foods. The arsenic is absorbed and distributed throughout the body once ingested by foods. Therefore, it is difficult to find blank blood samples for method validation. Considering the consumption of seafood is the major source of dietary arsenic, seafood consumption was prohibited before 1 month when drug-free blood samples were collected from donors. Calibration curves of arsenic species were compared between spiked drug-free plasma and aqueous standards. There was an obvious difference in the curve slopes that varied from day-to-day, indicating an apparent matrix effect. Thus, matrix matched calibration was necessary for the assay of arsenic species in plasma. The typical equations are displayed in Table 2. The ranges were 0.2-20 ng/mL for total arsenic in solution and 2.0-50 ng/mL for four arsenic compounds in plasma (r >0.995). The ranges are suitable for the determination of the arsenic compounds in patients with APL. The reported therapeutic and toxic concentration of arsenic in human blood-plasma/serum is 2-70 and 50-250 ng/mL, respectively [18]. In a clinical pharmacokinetic study by Fukai et al. [19], ATO was administered intravenously at a dose of 0.08 mg/kg/day in a patient, the serum concentration ranges of 18.2-40.7 for
total arsenic, 5.8-13.1 ng/mL for As(III), 5.1-14.4 ng/mL for DMA and 5.5-13.2 ng/mL for MMA. In another clinical pharmacokinetic study by Fox et al. [20], ATO was administered intravenously at a recommended dose of 0.15 mg/kg/day, the patients plasma concentration of As(III) ranged from 5 to 28 ng/mL. Therefore, it is concluded that HPLC-HG-AFS method might be sensitive enough for monitoring the speciation of ATO metabolites in blood samples in pharmacokinetic, therapeutic and toxicological studies. Five replicates of LLOQ samples of arsenic species were prepared and analyzed, respectively. The relative standard deviation (RSD%) data for precision was within ±20%. The corresponding limits of detection estimated in water were very low, which were 0.12, 0.19, 0.20 and 0.27 ng/mL for As(III), DMA, MMA and As(V), respectively. The precision of total arsenic can be accepted. The coefficients of variation for intra- and inter-batch precision of total arsenic ranged from 0.89% to 4.54% and 0.62% to 7.32%, respectively. The results demonstrated the coefficients of variation for intra- and inter-batch precision of arsenic species in plasma ranged from 1.82% to 9.33% and 3.08% to 12.49%, respectively. A critical problem of the present assay is that there is no standard reference material for plasma and blood cells. Therefore, it is impossible to investigate the accuracy by checking the certified reference materials. To solve the problem, the spiked arsenicals recovery technique was used for quality assurance. Standards were spiked to the samples from the ATO-treated patient with APL. The samples were prepared in the same method of corresponding protocols above-mentioned. Spiked recovery (%) = [(Measured value-Background value)/Spiked value]×100%. The results of spiked recovery were summarized in Table 3. The RSD% of each concentration levels was less than 11.90%. Good accuracy and repeatability were obtained for all the analytes at each concentration levels with the spiked recoveries. Potential inter-conversion among different arsenic species is a problem during the storage. Therefore, the stability of As(III), DMA, MMA and As(V) was investigated in the present study. The stability test results are presented in Table 4. The results demonstrated that the individual arsenic species in the stock solutions were stable at 4oC for at least 120 days. The average bias of accuracy ranged from -5.35% to 6.80% for spiked plasma samples placed at room temperature for 6 h. The average bias of accuracy ranged from -3.69% to -4.37% for extracted plasma samples at ambient temperature for 6 h. The freeze-thaw stability bias for As(III), DMA, MMA and As(V) in plasma was from -9.05% to -3.88% after three cycles. The long-term stability in plasma was from -4.11% to 7.52%. The results demonstrated the arsenic species were stable during the sample storing and processing
procedures. 3.5. Application Arsenic concentrations in blood samples from 5 patients with APL are shown in Table 5. The total arsenic ranged from 114.3 ng/g to 328.9 ng/g in blood cells and from 28.3 ng/g to 59.2 ng/g in plasma, respectively, indicating remarkable individual variation. The results demonstrate that the total arsenic concentrations in the blood cells were over 4 times higher than those in the plasma, which is consistent with the previous reports by Yuta Yoshino [8]. The results also suggest that most of arsenic is present in blood cells of the APL patients, indicating that the profiles of arsenic species in blood cells should be noticed. The ratio of the sum concentration of arsenic species to the total arsenic concentration in plasma was more than 80%. The results suggest that the extraction recovery of the method for arsenic species in plasma was higher than 80%, which was higher than in other reports [8, 9]. Thus, the sample preparation in our study is suitable for the determination of arsenic species in plasma. In agreement with previous reports [7-9], the sum of methylated metabolites (DMA+MMA) was higher than the sum of inorganic arsenic (As(III)+As(V)) during the therapy period, suggesting that methylated metabolites (DMA and MMA) were major metabolites in plasma. As(III) can be determined in our study while no As(III) can be seen in a previous report [7] as shown in Fig 3. The reasons for these results may be the prevention of arsenic losses from separation procedure or the individual variation. In
our
study,
the
As(III)>DMA>MMA>As(V)
concentrations for
patient
of
arsenic
compounds
ranked
a,
MMA>As(III)>DMA>As(V)
in for
the
order
patient
b,
DMA>MMA>As(III)>As(V) for patient c and patient e, MMA>DMA>As(III)>As(V) for patient d as shown in Fig 3. These results indicate large variations among individual patients in metabolic efficiency and methylation levels. These inter-individual differences in total arsenic and arsenic speciation were probably due to different age, treatment duration or pathological state. But we should notice that genetic polymorphisms in arsenic metabolism genes, such as arsenic (+III oxidation state) methyltransferase genes (AS3MT), is considered to be related to individual differences in the arsenic metabolism [21]. These results also suggest that continuous efforts to understand the interindividual variations in arsenic metabolism and the relationship between factors and arsenic metabolism are greatly useful for proposing effective therapy protocols for individual patients with APL.
In the present study, we attempted to apply the method to determine the arsenic species in blood cells. The results show that the sum of four arsenic species was significantly lower than that the total arsenic after digestion in blood cells from the same patient. The problem may be explained by the presumption that the major arsenic in blood cells is likely bound to hemoglobin [14, 22]. The analytical method of arsenic species in blood cells was problematic. The protein-bound arsenicals were not released from hemoglobin during the sample preparation procedure and produced precipitation losses. As far as the analytical problems are concerned, we would focused on accurately determining the arsenic species in blood cells of patients with APL by cleaving the bonds between As and hemoglobin in the following study. 4. Conclusion Despite the methods of total arsenic and arsenic species being widely developed, arsenic is not routinely afforded as a determination in clinical laboratory. The methods of HG-AFS and HPLC-HG-AFS were established and validated for the determination of total arsenic in whole blood and arsenic species of As(III), DMA, MMA and As(V) in plasma, respectively. The methods were applied to the assay of arsenic compounds in plasma and blood cells from patients with APL. The methods were found to be simple, low cost and suitable for the determination of total arsenic and arsenic species as well as could be easily implemented in clinical laboratory. This approach provided the basis for further clinical research.
Acknowledgements
This project was supported by Heilongjiang Province Science Foundation for Youths (No. QC2015119) and Foundation of The First Affiliated Hospital of Harbin Medical University (No. 2012lx003).
References
Toxicol Lett. 168 (2007) 310-318. [13] K. Ito,;1; C.D. Palmer, A.J. Steuerwald, P.J. Parsons, Determination of five arsenic species in whole blood by liquid chromatography coupled with inductively coupled plasma mass spectrometry, J. Anal. At. Spectrom. 25 (2010) 1334-1342. [14] B. Chen, X. Lu, S. Shen, L.L. Arnold, S.M. Cohen, X. C. Le,;1; Arsenic speciation in the blood of arsenite-treated F344 rats, Chem. Res. Toxicol. 26 (2013) 952-962. [15] M. Welna, A. Szymczycha-Madeja, P. Pohl,;1; Comparison of strategies for sample preparation prior to spectrometric measurements for determination and speciation of arsenic in rice, Trends Anal. Chem. 65 (2015) 122-136. [16] C. Wei, J. Liu,;1; A new hydride generation system applied in determination of arsenic species with ion chromatography-hydride generation-atomic fluorescence spectrometry (IC-HG-AFS), Talanta. 73 (2007) 540-545. [17] Y.J. Zhang, S.P. Qiang, J. Sun, M. Song, T.J. Hang,;1; Liquid chromatography-hydride generation-atomic fluorescence spectrometry determination of arsenic species in dog plasma and its application to a pharmacokinetic study after oral administration of Realgar and Niu Huang Jie Du Pian, J Chromatogr B, 917-918 (2013) 93-99. [18] M. Schulz, A. Schmoldt,;1; Therapeutic and toxic blood concentrations of more than 800 drugs and other xenobiotics, Pharmazie. 58 (2003) 447-474. [19] Y. Fukai，M. Hirata，M. Ueno，N. Ichikawa，H. Kobayashi，H. Saitoh, T. Sakurai;1; , K. Kinoshita, T. Kaise , S. Ohta, Clinical pharmacokinetic study of arsenic trioxide in an acute promyelocytic leukemia (APL) patient: speciation of arsenic metabolites in serum and urine, Biol. Pharm. Bull. 29 (2006) 1022-1027. [20] E. Fox,;1; B.I. Razzouk, B.C. Widemann, S. Xiao, M. O'Brien, W. Goodspeed, G.H. Reaman, S.M. Blaney, A.J. Murgo, F.M. Balis, P.C. Adamson, Phase 1 trial and pharmacokinetic study of arsenic trioxide in children and adolescents with refractory or relapsed acute leukemia, including acute promyelocytic leukemia or lymphoma, Blood. 111 (2008) 566-573. [21] K. Engström, M.Vahter, S.J. Mlakar, G.Concha, B. Nermell, R. Raqib, A.Cardozo, K. Broberg,;1; Polymorphisms in arsenic (+III oxidation state) methyltransferase (AS 3MT) predict gene expression of AS3MT as well as arsenic metabolism, Environ. Health Perspect. 119 (2011) 182-188. [22] S. Shen, M. Lu, X. F. Li, W. R. Cullen, M. Weinfeld, X. C. Le,;1; Arsenic binding to proteins, Chem. Rev. 113 (2013) 7769-7792.
Figure Legends
Fig. 1. Schematic diagram of HPLC-HG-AFS system used for the determination of the total arsenic and arsenic species, respectively.
Fig. 2. Representative HPLC-HG-AFS chromatograms: drug-free plasma (A); drug-free plasma spiked with standards of the four arsenic species at LLOQ (B1) and 10.0 ng/mL (B2) (As(III): tR=2.2 min; DMA: tR=2.8 min; MMA: tR=3.5 min; As(V): tR=4.3 min); real subject plasma sample (patient b) (C). The four arsenic species were separated on an anion exchange column (Hamilton PRP-X100) with
13 mM CH3COONa - 3 mM NaH2PO4 - 4 mM KNO3 - 0.2 mM EDTA-2Na (pH 6.0) as a mobile phase.
Fig. 3. Arsenic species concentration in plasma of five APL patients (a, b, c, d and e) at different stages of the therapy. The length of the bar represents the total arsenic concentration in plasma with the sub-divisions for the individual arsenic species quantified; * indicates the gap between sum of species and total arsenic concentration.
Table 1 Details of the HG-AFS conditions Total arsenic Pre-reducing agent (freshly prepared) Reductant (freshly prepared)
Arsenic speciation
10% ascorbic acid (m/v) - 10% thiourea (m/v) 3% KBH4 (m/v) in 0.5% KOH (m/v)
3% KBH4 (m/v) in 0.5% KOH (m/v)
10% HCl (v/v)
10% HCl (v/v)
Primary current (mA)
70
80
Boost current (mA)
35
40
Carrier liquid
Pump speed (rpm/min)
80
60
Shielding gas flow rate (mL/min)
800
900
Carrier gas flow rate (mL/min)
300
300
Negative high voltage (V)
280
300
Table 2 Calibration curves and limits of detection
Total As As(III) DMA MMA As(V)
Linearity range (ng/mL)
Typical equations
r
Limits of detection (ng/mL)
0.2-20
A= 84.2 + 145.9C
0.9997
0.08
2.0-50
Y=1.178×105+
2.184×105X
0.9996
0.12
2.0-50
Y=3.714×103+
1.778×105X
0.9965
0.19
2.0-50
Y=4.527×104+
2.287×105X
0.9989
0.20
2.0-50
Y=1.259×105+
1.822×105X
0.9986
0.27
Table 3 Spiked recovery of total As in pasma and blood cells and arsenic species in plasama Background value (ng/g)
Spiked value
Spike recovery (%)
50 ng/g
89.4
100 ng/g
106.3
200 ng/g
94.3
20 ng/g
87.6
50 ng/g
92.6
100 ng/g
100.2
5 ng/mL
100.4
10 ng/mL
92.4
20 ng/mL
101.5
5 ng/mL
85.6
Blood cells
Total As
138.8
Plasma
Total As
As(III)
DMA
MMA
As(V)
36.7
7.39
8.79
8.96
5.03
10 ng/mL
96.3
20 ng/mL
104.3
5 ng/mL
108.6
10 ng/mL
99.6
20 ng/mL
98.7
5 ng/mL
103.9
10 ng/mL
81.2
20 ng/mL
95.3
Total As = concentration of total arsenic after digestion
Table 4 Stability data of As(III), DMA, MMA and As(V) in plasma
Ambient
Stock solutions Arsenic
(4oC,
120 d)
species RSD
Accuracy
(%)
Bias (%)
As(III)
0.59
0.25
DMA
1.08
0.15
MMA
1.46
-1.05
As(V)
2.10
-1.15
Initial
(room temperature,
concentration
6 h)
(ng/mL)
Ambient
Freeze-thaw
(autosamper, 6 h)
(-80°C, 3 cycles)
60 d (-80°C)
RSD
Accuracy
RSD
Accuracy
RSD
Accuracy
RSD
Accuracy
(%)
Bias (%)
(%)
Bias (%)
(%)
Bias (%)
(%)
Bias (%)
4.05
6.04
-5.35
10.73
-2.88
5.90
-9.05
8.28
-4.11
39.8
3.04
-3.60
7.55
1.26
4.05
-3.01
3.06
0.75
4.46
2.22
1.04
9.20
0.37
7.30
-1.93
10.64
1.35
41.5
2.38
-2.81
0.52
-3.69
2.27
0.80
1.23
-3.69
3.95
5.35
2.53
5.78
0.93
4.71
-3.12
0.86
3.37
39.7
4.09
-1.87
7.57
4.37
5.78
1.85
3.50
0.92
4.12
7.75
6.80
12.03
2.59
10.33
3.88
6.24
7.52
40.8
1.93
2.78
0.51
0.82
3.28
1.88
6.11
2.45
Table 5 The concentrations of arsenic in blood samples collected from patients with APL Concentrations of arsenic (ng/g) Patients
Patient a
Patient b
Patient c
Patient d
Patient e
Age, sex
9, female
55, female
46, female
30, female
31, male
Time
Day-7
Day-8
Day-11
Day-12
Day-23
114.3
172.3
328.9
138.8
186.9
Total As
28.3
38.1
59.2
36.7
39.0
As(III)
7.75
7.98
10.71
7.39
6.59
DMA
7.14
7.47
20.21
8.79
13.22
MMA
5.46
13.05
16.87
8.96
9.44
As(V)
3.58
4.61
6.51
5.03
3.24
tAs
23.93
33.11
54.3
30.17
32.49
Blood cells Total As Plasma
Total As = concentration of total arsenic after digestion; tAs = the sum of arsenic species