Optimization of the Determination Method for Dissolved

Sep 17, 2017 - There is a serious dispute on the existence of β-N-methylamino-l-alanine (BMAA) in water, which is a neurotoxin that may cause amyotro...
1 downloads 0 Views 1MB Size
Subscriber access provided by UNIV OF CAMBRIDGE

Article

Optimization of the determination method for dissolved cyanobacterial toxin BMAA in natural water Boyin Yan, Zhiquan Liu, Rui Huang, Yongpeng Xu, Dongmei Liu, Tsair-Fuh Lin, and Fuyi Cui Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b02867 • Publication Date (Web): 17 Sep 2017 Downloaded from http://pubs.acs.org on September 19, 2017

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 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

Analytical Chemistry

Optimization of the determination method for dissolved cyanobacterial toxin BMAA in natural water Boyin Yan,1, †Zhiquan Liu,1,†,*Rui Huang,†YongpengXu,†Dongmei Liu,†Tsair-Fuh Lin,‡Fuyi Cui†, * †

State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology, Harbin 150090, China Department of Environmental Engineering, National Cheng Kung University, Taiwan 1 These authors contributed equally to this work. ‡

ABSTRACT: There is a serious dispute on the existence of β-N-methylamino-L-alanine (BMAA) in water, which is a neurotoxin that may cause Amyotrophic lateral sclerosis/Parkinson's disease (ALS/PDC) and Alzheimer’ disease. It is believed that a reliable and sensitive analytical method for the determination of BMAA is urgently required to resolve this dispute. In the present study, the solid phase extraction (SPE) procedure and the analytical method for dissolved BMAA in water were investigated and optimized. The results showed both derivatized and underivatized methods were qualified for the measurement of BMAA and its isomer in natural water, and the limit of detection and the precision of the two methods were comparable. Cartridge characteristics and SPE conditions could greatly affect the SPE performance, and the competition of natural organic matter is the primary factor causing the low recovery of BMAA, which reduced from approximately 90% in pure water to 38.11% in natural water. The optimized SPE method for BMAA was a combination of rinsed SPE cartridges, controlled loading and elution rates and elution solution, evaporation at 55oC, reconstitution of a solution mixture and filtration by polyvinylidene fluoride membrane. This optimized method achieved> 88% recovery of BMAA in both algal solution and river water. The developed method can provide an efficient way to evaluate the actual concentration levels of BMAA in actual water environments and drinking water systems.

INTRODUCTION β-N-methylamino-L-alanine (BMAA) is a neurotoxin found from the seeds of Cycasmicronesicain 19671. This compound can potentiate neuronal injury caused by amyloid-β or 1-methyl-4-phenylpyridinium ion at concentrations as low as 10µM2. BMAA is further hypothesized as an important factor in the induction of some neurodegenerative disease such as Amyotrophic lateral sclerosis/Parkinson's disease (ALS/PDC) and Alzheimer's disease3-4. Karlsson et al.5 reported the transplacental transfer of BMAA and uptake in fetal mouse brain. They also elucidated the presence of protein-bound BMAA in neonatal rat brain6-7and proposed the urgency of evaluating actual human BMAA exposure levels8. Moreover, a variety of freshwater algae species such as cyanobacteria9, diatoms10 and flagellates11 are considered to be the original producers of BMAA. It has been reported that BMAA has accumulated and transferred through food chains in the ecosystems of Guam12, south-eastern United States13, Baltic Sea14 and Tai Lake15. BMAA has even been found under harsh conditions16-17. Furthermore, it has been found that diverse aquatic products contain BMAA at different concentration levels18-21. In addition, Chen et al.22reported the probable presence and fate of BMAA and its chlorinated byproducts in drinking water systems. All of these studies suggest that it is easy to exposure to BMAA through one’s daily diet and BMAA has become a worldwide problem, which may seriously threaten the human health. It is important and urgent to understand the concentration level of BMAA in aquatic environments.

However, there are still serious disputes on the presence of BMAA in cyanobacteria laden water among different research groups23. The supporters believe that there are two primary forms, free fraction and protein-bound fraction, in cyanobacteria cells, and they report BMAA is detectable in diverse strains of cyanobacteria grown under laboratory culture and samples collected from actual waters9, 24. Others argue that many previous positive studies adopt a liquid chromatography fluorescence detector (LC-FLD) method, by which BMAA may be overestimated because identifying an analyte though retention time and fluorescence rather than the analyte itself may result in false positive results25. They also believe isomers of BMAA, such as 2,4-diaminobutyric acid (DAB) which is considered to be a component of cell wall peptidoglycan of 65 species of actinomycetes26, may contribute to the false positive results25. Kubo et al.27 and Rosén and Hellenäs28 directly analyzed the cyanobacteria samples with a liquid chromatography-tandem quadrupole mass spectrometry (LC-MS/MS) coupled with a HILIC column and the results showed no BMAA exist in algal samples. Nevertheless, these results are contradicted by the supporters, who believe the existence of BMAA, because of the low performance of HILIC column and the low ionization efficiency in the electrospray ion source (ESI) for the direct measurement of BMAA. Both of the two sides, at least, endorse that a reliable, sensitive and validated method is essential to find out the actual truth about BMAA in aquatic environments. Faassenet al25 compared LC-FLD, derivatized and underivatized LC-MS/MS methods for the determination of BMAA in eight species of cyanobacteria, and found the limits of detection (LOD) of the three methods are comparable

ACS Paragon Plus Environment

Analytical Chemistry

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

but LC-FLD method may overestimate BMAA concentrations in some samples. They indicate BMAA is undetectable in the eight species by either derivatized or underivatized LC-MS/MS method. It is interesting that in an early study the same group claimed BMAA was detected by LC-MS/MS method in the cyanobacteria scum samples of urban drinking water sources contaminated by cyanobacteria blooms with a positive ratio of 42.85%29.The researchers do not get an agreement on a validated BMAA determination method as well as the real status of BMAA in aquatic environments up to date. It is difficult to measure BMAA accurately because of some of its characteristics. First of all, BMAA is hydrophilic and is more difficult to concentrate than hydrophobic compounds30. Secondly, there is no fluorescent or ultraviolet characteristic of BMAA as well as many other amino acids31, which implies traditional detectors such as UV detector, diode-array detector and FLD cannot be directly used for the measurement of BMAA. Thirdly, the isomers of BMAA, some of which has been found in natural water samples14, may cause false positive results during the determination. These isomers are difficult to separate from BMAA during the LC separation process and can provide similar transitions as BMAA during the MS/MS analytical process. Apart from BMAA in cyanobacteria cells, including both free fraction and protein-bound fraction, the dissolved BMAA in water is rarely reported. As a hydrophilic compound with a small molecular weight of 118, BMAA may penetrate through cyanobacteria cells during the process of biological metabolism and exist in natural water in dissolved form32, which may serve as a potential threat to the safety of urban drinking water11, 33. However, the available research papers about dissolved BMAA are still limited. BMAA may exist in water at an extremely low concentration, and a solid phase extraction (SPE) method is commonly employed as a concentration process for the measurement of BMAA. Considering the hydrophilicity of BMAA, SPE for a large volume of BMAA sample may cause a low efficiency of concentration, and the current determination method for free and protein-bound BMAA may not be qualified for the measurement of dissolved BMAA in natural water. The purpose of this study was to develop a highly sensitive and accurate analytical method for dissolved BMAA in aquatic environmental samples, which may be helpful for researchers to understand the real concentration level of BMAA in water. In order to improve the recoveries of BMAA, the pretreatment procedure was optimized step by step. Derivatized and underivatized LC-MS/MS analytical methods were also compared in terms of LOD and spike recovery in natural water. The results may also be useful to improve the determination method for free and protein-bound BMAA in algal cells.

Page 2 of 9

D-2,4-diaminobutyric acid-2,3,3,4,4-2D5 dihydrochloride (D5DAB) was provided by C/D/N Isotopes Inc. (Canada). The AccQ-Tag Ultra Derivatization Kit was obtained from the Waters Corporation (USA). The acetonitrile, methanol and formic acid obtained from Sigma-Aldrich Co. LLC (USA) were high-performance liquid chromatography (HPLC) grade. Solution of ammonium hydroxide and humic acid (HA) were purchased from Aladdin Industrial Corporation (Shanghai, China) and 2M hydrochloric acid was provided by TCI Development Co. Ltd (Shanghai, China). Pure water was supplied by a MilliQ water purification system (Millipore Ltd., Bedford, MA, USA) to 18MΩ quality or better. Polytetrafluoroethylene (PTFE), polyvinylidene fluoride (PVDF), polyether sulfone (PES), nylon (NY) and glass fiber (GF) were purchased from JinTeng Experimental Equipment Co. Ltd (Tianjin, China). Stock solutions (100mg/L) of BMAA, DAB and D5DAB were prepared in MilliQ water and stored at 0°C. These stock solutions could be stable for more than 6 months under this storage condition 34. The prepared method of the stock solution of HA was prescribed by Gaoet al. in 2015 35. Method optimization SPE process. The optimization of SPE process included the selection of SPE cartridges (Oasis-MCX (6cc, 500mg, Waters, USA), Strata-X-C (6cc, 500mg, Phenomenex, USA), Bond ElutPlexa PCX (6cc, 500mg, Agilent, USA), DSC-MCAX (3cc, 500mg, Supelco, USA) and LC-SCX (3cc, 300mg, Supelco, USA)), the optimization of sample loading (1 mL/min, 1.85 mL/min and 3 mL/min) and eluting rate (1 mL/min and 3 mL/min), as well as the adjustment of eluent constitute (2%, 3% and 5% of NH3·H2O in methanol) and volumes (1mL, 3mL, 5mL, 7.5mL and 10mL). The evaporation (room temperature, 55oC, 80oC and vacuum) and filtration (PTFE, PVDF, PES, NY and GF) procedures after SPE were also studied. The details of the optimization can be found in Supporting Information. Improvement of SPE performance in nature water. The interferences of ions and natural organic matter (NOM) were investigated by adding cations (Ca2+, Mg2+, Na+ and NH4+ at a constant concentration of 3 mM, 5mM or 10mM) or humic acid (2 mg/L, 5 mg/L or 10 mg/L) into samples before the SPE treatment. The performance of adjusting pH (acidized to pH 2, 3 or 4 with 2M hydrochloric acid), combining a traditional C18 (Oasis-HLB, 6cc, 500mg, Waters, USA) SPE cartridge and changing sample volume (50mL, 60mL, 70mL, 80mL and 90mL) on the improvement of BMAA recoveries were investigated. The activation and sample loading conditions of C18 cartridge were conducted according to the study by Nicolas Mazzella et al.36. The C18 cartridges were discarded after the sample loading procedure, and the targets on the Oasis-MCX cartridges were washed, eluted and analyzed under the optimized conditions of SPE.

MATERIALS AND METHODS Analysis by LC/MS-MS Chemicals and reagents BMAA hydrochloride and DAB dihydrochloride were purchased from Sigma-Aldrich Co. LLC. (USA);

LC-MS/MS analysis was performed using an Ultra Performance Liquid Chromatography (UPLC, Acquity UPLC, equipped with 2777 injector) coupled with a triple

2

ACS Paragon Plus Environment

Page 3 of 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

Analytical Chemistry

quadrupoledetector (Xevo TQ-S, equipped with an electrospray interface), both of which were purchased from Waters Corporation (Milford, MA, USA). In order to optimize the MS tuning parameters, a standard underivatized or derivatized solution (1mg/L) was injected into the mass spectrometer directly. For the underivatized analytical method, the mass spectrometer was used in positive mode with the capillary set at 4.0 kV, cone at 15 V, desolvation temperature at 650oC, desolvation gas at 1000L/Hr and source temperature at 150oC. Six transitions including 119>102, 119>101, 119>88, 119>76, 119>44 and 124>107 were monitored at respective collisions energies (CE) of 14, 14, 16, 16, 18 and 14V using multiple reaction monitoring (MRM) scan mode. Transition 119>102 was used to quantify BMAA and DAB, and transition 125>107 was used to quantify D5DAB. Transition 119>88 and 119>101 were used as qualitative fragment ions for BMAA and DAB, respectively. When the samples were analyzed by derivatized method, the mass spectrometer was also used in positive mode with the capillary set at 4.0 kV, cone at 20 V, desolvation temperature at 600oC, desolvation gas at 1000L/Hr and source temperature at 150oC. Five transitions, including 459>119, 459>171, 459>188, 459>258 and 464>171, were monitored at respective collisions energies (CE) of 18, 14 16, 16 and 14V using MRM scan mode. Transition 459>171 was used to quantify AQC-BMAA and AQC-DAB, and transition 464>171 was used to quantify AQC-D5DAB. Transition 459>258 and 459>188 were used as qualitative fragment ions for AQC-BMAA and AQC-DAB, respectively. Nitrogen and argon were used as drying gas and collusion gas of both analytical methods, respectively. The underivatized samples were analyzed through a hydrophilic interaction chromatography column (Acquity UPLC BEH, 2.1mm 150 mm, with 1.7 mm particle size), provided by Waters as well. The temperature of column was 30oC. Mobile phases were 0.1% (v/v) formic acid in acetonitrile (A) and 0.1% (v/v) formic acid in MilliQ water (B). An isocratic gradient (30%B) was used at a flow rate of 0.4mL/min. The derivatized samples were analyzed through a C18 column (Acquity UPLC BEH, 2.1mm 100 mm, with 1.7 mm particle size) and the temperature of the column was 50oC. A binary mobile phase system comprised of acetonitrile containing 0.1% formic acid (A) and water containing 0.1% formic acid and 0.05% NH3.H2O (B) was adopted. An isocratic gradient (90%B) was used at a flow rate of 0.4mL/min. A series of low concentrations of standard solutions was prepared and then analyzed by both underivatized and derivatized method. The LODs were ascertained by 3 times of signal to noise ratio of chromatograms. RESULTS AND DISCUSSION LC-MS/MS analysis and comparison A sensitive and validated instrumental analytical method is the fundamental requirement to obtain reliable results. The optimization of instrumental analysis for BMAA included mass spectrum optimization and LC optimization. MS/MS detector can provide detailed chemical structure information,

which is necessary to distinguish BMAA from its isomers. The running conditions were tuned and optimized by the analysis of BMAA, DAB and D5DAB standards or their derivative with AQC at a concentration of 1mg/L. The optimized results are described in the section of Analysis by LC/MS-MS. MS/MS detector can distinguish BMAA and its isomers by their different transitions, but well separated components before the entrance of MS/MS can improve the sensitivity and the reliability of this analysis method. Five different isocratic elution programs were compared for the separation of underivatized BMAA and DAB on a HILIC column: 65%A, 70%A, 75%A, 80%A and 85%A (Fig. S1). Resolution was improved with the increase of the proportion of organic mobile phase, since HILIC columns were designed to retain polar compounds and the force of elution decreased with the increased proportion of organic mobile phase. However, a high proportion of organic mobile phase made BMAA difficult to ionize at ESI and lowered the peaks of targets down. Considering the separation performance and the signal strength, 70%A was the best condition for the measurements of BMAA and DAB. The retention times of BMAA and DAB were 2.60 min and 2.96 min, respectively, both of which were earlier than those in previous papers28, 37. A similar procedure was adopted to optimize the LC elution program for derivatized BMAA and DAB on a C18 column, and 10%A was chosen as the proportion of organic mobile phase (Fig. S2). Linear analytical curves for both BMAA (r2=0.9924, Y=62.9554X-40.1376) and DAB (r2=0.9951, Y=37.6817X-2.4855) with the underivatized method were 0.8-20µg/L (Fig S3A), while the curves for BMAA (r2=0.9982, Y=23.0698X-2.4855) and DAB (r2=0.9993, Y=14.2897X-5.1027) with derivatized method (Fig S3B) were 0.2-5µg/L and 0.5-10µg/L, respectively. The detection limit of underivatizd and derivatized BMAA were 500 ng/L (equivalent to 42.37 fmol/injection at an injection volume of 10 µL) and 100 ng/L (equivalent to 8.47 fmol/injection at an injection volume of 10 µL), respectively. The result in the present study is in accordance with the study by Faassenet al.25., who indicate the LOD of BMAA analyzed by underivatized and derivatized LC-MS/MS method are 106 fmol/injection and 85 fmol/injection. It is reported that UPLC has significant advantages in speed and resolution compared with HPLC38, which not only reduced the retention time but also increased the sensitivity of the measurement in the present study. The LC-MS/MS chromatograms in Daughter scan mode for underivatized or derivatized BMAA and DAB could be found in Fig. S4 and S5. Comparing the LOD of the two detection methods, it can be found that the derivatized method performed a higher sensitivity. The reason is that the molecular weight is increased by the derivatization of BMAA and DAB with AQC, which facilitates the identification of the parent ion and fragment ions by the MS/MS detector31. Accuracy and precision of both methods were tested by spiking standards at a concentration of 20µg/L into the supernatant of Microcystisaeruginosa (a lab-culture of cyanobacteria), waters from Songhua River (Harbin, China) and waters from Kongmu Lake (Nanchang, China),

3

ACS Paragon Plus Environment

Analytical Chemistry

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

respectively. The spiked samples were analyzed by both derivatized and underivatized LC-MS/MS methods (without SPE treatment), and the results were summarized in Table 1. Both of the two methods can separate and measure BMAA and DAB correctly. The relative standard deviation (RSD) of the two methods, approximately 3~5%, were also comparable. This implies the disputes on the existence of BMAA in aquatic environment in previous studies are not caused by the difference between underivatized and derivatized instrumental methods. Thus, pretreatment methods may result in the loss of targets and may be the primary reason causing the disputes. Table1 Recoveries (n=4) of BMAA and DAB spiked at the concentration of 20µg/L in blank supernatants of Microcystis aeruginosa, in waters from Songhua River (Harbin, China) and waters from Kongmu Lake (Nanchang, China). The spiked samples were analyzed by both derivatized and underivatized LC-MS/MS methods.

sample

supernatants Songhua River Kongmu Lake

Recoveries of underivatization (mean%±RSD)

Recoveries of derivatization (mean%±RSD)

BMAA

DAB

BMAA

DAB

95.22 ±2.91 93.46 ±4.91 95.50 ±4.78

87.14 ±0.78 90.35 ±4.72 97.25 ±7.15

115.23± 2.81 113.17± 5.78 109.25± 1.71

122.62± 6.46 128.96± 2.65 114.38± 5.48

Comparison and optimization of SPE procedure SPE cartridge comparison. The SPE cleanup step for the freshwater samples is necessary to purify and concentrate BMAA and consequently increase the sensitivity of the analytical technique, because dissolved BMAA may be at a trace level in aquatic environment. Cation-exchange SPE cartridges are widely-used to purified for low molecular mass basic compounds like BMAA and DAB. Five kinds of cartridges (Oasis-MCX, Strata-X-C, Plexa PCX, DSC-MCAX, and LC-SCX) were tested for the treatment of 100 mL BMAA and DAB standard solutions (1µg/L), and the recoveries and precisions of BMAA and DAB are shown in Fig.1. The results indicated Oasis-MCX was the best cartridge for BMAA, which is also widely used for the purification of free and protein-bound BMAA in previous studies21, 27, 32.

Page 4 of 9

The low recovery may be caused by either the loss of BMAA during the sample loading and the washing process or the insufficient elution caused by the excessive adsorption of BMAA on the cartridges. In order to study this question, the variations of BMAA and DAB in the sample loading and washing filtrate were monitored and the results were shown in Fig. 2. At the beginning, most of BMAA and DAB were retained in the cartridges. Nevertheless, as the volume of loading samples increased, the capacity of retention tended to be saturated and the leakage increased. At the end of sample loading process, some cartridges, e.g. Strata-X-C, DSC-MCAX and LC-SCX, not only lost their ability to retain BMAA, but also released a part of BMAA which had been retained on the cartridges. During the washing process, the leakages of targets were observed among the five kinds of cartridges, and 2% formic acid, which was a strong polar rinse agent, washed out more BMAA and DAB than methanol (Fig. 2B and D, table S1). Obviously, the leakage of targets during the sample loading and washing processes reduced the recovery of BMAA and DAB, which was caused by the insufficient retention of BMAA by cartridges. Because Oasis-MCX performed the best capacity of retain BMAA and DAB, it provided the highest recovery of BMAA among the five cartridges. Oasis-MCX is selected as the SPE cartridge in the following studies. The different performances of SPE cartridges may be relevant to the structure of skeleton and the sulfonation degree in the fillers of cartridges. The skeletons of Oasis-MCX, Strata-X-C, Plexa PCX cartridges were polymer, while the skeletons of DSC-MCAX, LC-SCX cartridges were silica gel that bind alkane groups. According to the experimental results, the performances of cartridges with different structures of skeleton were comparable except MCX, which implied skeleton structure may not be the primary factor affecting the SPE performance. Thus, the amounts of functional group (sulfonation degree) may play an important role, which should be further studied. Table 2Recoveries (n=3) of BMAA and DAB obtained from different flow rates of loading and eluting. Flow rate of loading (mL/min) 1

Flow rate of eluting (mL/min) 1

1.85 3

Fig.1 Recoveries of BMAA and DAB at a spiked concentration of 1µg/L in pure water treated with different kinds of SPE cartridges. (n=3)

Recoveries of SPE cleanup (%) BMAA 61.62±1.02

DAB 81.52±0.83

3

89.45±3.14

87.20±2.15

1

56.05±8.96

54.41±0.47

3

78.52±0.42

67.01±0.71

1

34.68±2.20

17.34±3.69

3

55.76±2.81

37.49±1.59

Optimization of sample loading and eluting rate. The rates of sample loading and eluting could affect the performance of SPE. The combinations of sample loading and eluting rates were compared and the results were shown in Table 2. High recoveries of BMAA and DAB were obtained under slow sample loading rate and fast eluting rate. The SPE process can be regarded as a combination of adsorption and desorption. Thus, the transport of the targeted chemicals between the

4

ACS Paragon Plus Environment

Page 5 of 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

Analytical Chemistry

Fig.2 The concentration of BMAA (A) and DAB (C) in the sample loading filtrate passed though the cartridges and the concentration of BMAA (B) and DAB (D) in the washing filtrate (2% FA in water). The dotted line represents the spiked concentration of BMAA and DAB in pure water.

Fig.3 Impacts of eluent constitution and volumes on the recoveries of BMAA (A) and DAB (B).

liquid phase and cartridge materials may involve a few processes such as film transport, intraparticle diffusion, and adsorption onto (desorption from) the sites39-40. During the loading process, a low loading rate reduced the flow velocity, but extended the contact time, which enhanced the mass transfer from the liquid phase to cartridge. The result, high recoveries obtained at low sample loading rates, implied that intraparticle diffusion may be the rate limiting factor in the sample loading process. Classical adsorption studies41 indicate a high flow velocity can reduce external surface film resistance and promote intraparticle diffusion. During the elution process, a high elution rate may also increase the film transport rate and then accelerate the desorption of BMAA from cartridges. According to the results, the sample loading rate and the eluting rate were fixed at 1 mL/min and 3 mL/min, respectively.

Optimization of elution conditions. The performance of elution with different volumes and ratios of NH3·H2O in methanol were investigated and the results were shown in Fig. 3. Obviously, increasing the volume of eluents and the ratio of NH3·H2O was beneficial to the elution. It is worth noting that the eluent of 2% NH3·H2O in methanol, widely used in previous studies42-44, is insufficient to elute BMAA from the cartridges. This result implies that an insufficient elution may exist in some previous studies which may cause a failed measurement when the actual concentration of BMAA is just above the LOD. According to the results, 7.5 mL 5% NH3·H2O in methanol was adopted as the eluent. Optimization of evaporation and filtration. Evaporation is a necessary procedure after the elution step, which can eliminate the eluent and maintain the target compound. The evaporation was carried out under four different conditions (room

5

ACS Paragon Plus Environment

Analytical Chemistry

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

temperature, 55oC, 80oC and vacuum at room temperature) and the results were shown in Fig. 4A. The losses of BMAA and DAB during evaporation were negligible except that at the temperature of 80oC. The loss may be caused by either the thermal decomposition of BMAA or the volatilization effect accompanying with the evaporation of eluent. Considering the fact that BMAA is stable under 110°C for as long as 24h during the proteolysis process45, the latter may be the primary reason for the loss of BMAA. The evaporation rate decreased as the following sequence: 80°C>55°C≈vacuum>room temperature. Considering the time and energy cost, evaporation at 55°C was adopted in this study, which is also widely used for the measurements of free and protein-bound BMAA in previous studies20, 46-47.

Page 6 of 9

rather than organic solvent. It is reported that PES membrane can remove a part of organic micro-pollutants during the ultra-filtration process and hydrophobic organic is easier to be adsorbed on the membrane48. It was possible that a small amount of acetonitrile, which is more hydrophobic than BMAA, was retained during the filtration, causing the higher observed result. PVDF was deemed the best membrane material in this study.

Fig.4 Impacts of evaporation (A) and filtration (B) on the recoveries of BMAA and DAB. The volume of eluent was 10mL and the volume of reconstituted solution for filtration was 1mL.

Filtration, a necessary step to remove particles with the size larger than 0.22µm before the analysis with LC-MS/MS, may also cause the loss of BMAA. In order to reduce undesirable interactions and/or interception between BMAA and the filtermembrane material, five different membranes, PVDF, PTFE, PES, NY, and GF, were tested by filtrating standard reconstituted solutions at a concentration of 100µg/L, and the results were presented in Fig. 4B. PVDF filter membranes provided the best recovery value as high as 97%. The recoveries of BMAA treated by PTFE, NY and GF membranes were lower than 90%, which implied there may be chemical reaction and/or physical interception of BMAA with the membranes. It is interesting that the recoveries of BMAA treated by PES filter membranes were always higher than 100%. PES filters are generally used for aqueous solutions

Fig.5 Recoveries of BMAA and DAB at a spiked concentration of 1µg/L in cation solutions (A: BMAA; B: DAB) and in HA solutions (C) at different concentrations.

Improvement of SPE performance for BMAA in natural water Interference from natural water. Although the recoveries of BMAA and DAB with MCX cartridges were approximately 90% in pure water, the value decreased significantly to 38.11% and 26.44% in actual water samples (water from Songhua River with spiked BMAA of 1 µg/L). Obviously, the impurities in natural water reduced the recovery. It is reported that the retention mechanism of chemical organics by MCX

6

ACS Paragon Plus Environment

Page 7 of 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

Analytical Chemistry

included both ion exchange and electrostatic interaction34. Targets were ionized under acidic condition and then were retained on the cartridges. The retained targets were converted back to neutral under basic condition and were desorbed from the cartridges by elution. Thus, the competition of cations and NOM may result in the decreased SPE performance. The impacts of cations and NOM (Fig. 5) were investigated by adding Ca2+, Mg2+, Na+, NH4+ or humic acid in BMAA solution. Cations are the competitors for the active ion-exchange sites, while NOM may cause pore blocking to hinder the transfer of targets from the liquid phase to active sites. The results that the competition of cations was negligible but NOM did reduce the SPE performance indicated that there may be sufficient active sites for ion exchange, but limited channels for the transfer of targets. Thus, the pore blocking was identified as the primary mechanism, resulting in significantly reduced recoveries in natural waters. Efficient methods should be conducted to eliminate the interference from NOM. Adjusting pH. pH adjusting is a necessary step before the sample loading process, which can ionize BMAA and improve the retention capacity of the cartridges. The water sample collected from Songhua River and spiked with 1µg/L of BMAA were acidified with HCl to 242,334 or 449, as adopted in previous studies. Because the pKa1 of BMAA and DAB are 6.48 and 8.4137, ionizations of the two targets are complete under all of the three pH conditions. However, the recoveries were different and the best recoveries of BMAA and DAB were obtained with pH 3 (Fig 6). The results indicated a proper acidification could reduce the competition from the impurities in water and improve the performance of SPE. Under pH 2, the electrostatic competition may be enhanced because excessive organics were ionized, while under pH 4, numerous impurities may block the channels by electrostatic force, also resulting in a decrease of BMAA and DAB recoveries.

BMAA) and improve the performance of SPE. The results showed that the combination of HLB and MCX improved the recoveries of BMAA and DAB from 54.36% and 36.64% to 63.76% and 47.76%, respectively. The improvement was barely satisfactory because HLB was designed for the separation of hydrophobic compounds, which may not be the direct competitors of BMAA and DAB. Optimization of sample loading volume. The insufficient retention is the primary reason which causes the loss of BMAA and DAB (Section SPE cartridge comparison), and the competition of NOM even enhanced the leakage of targets and lowered the recovery (Section Interference from natural water). The recoveries of BMAA and DAB with different sample loading volumes in NOM solution of 5 mg/L were shown in table 3. The recoveries decreased sharply, as well as the adsorbed ratios of targets by MCX in the last 10 mL samples, with the increased sample loading volume. The competition of NOM greatly reduced the adsorption capacity of cartridges and accelerated the leakage of BMAA and DAB. Thus, the sample loading volume should be no more than 50 mL for the measurement of BMAA in actual water samples. Table3 Recoveries (n=3) of BMAA and DAB with different sample loading volumes. The spiked concentration was 1µg/L and the HA concentrations was 5mg/L. Recoveries of SPE cleanup Volume (mL)

(mean%±RSD)

Ratios of adsorbed BMAA in the last 10mL samples (mean%)*

BMAA

DAB

BMAA

DAB

50

94.28±2.50

62.36±3.72

-

-

60

85.52±1.30

56.22±3.57

40.78

25.52

70

77.09±2.67

55.09±3.24

30.16

48.36

80

72.67±3.76

50.50±4.38

36.15

18.47

90

67.50±1.69

44.20±3.16

26.12

-6.2

*Calculated by (initial concentration-filtrate concentration)/initial concentration

Fig. 6 Impacts of pH adjustment on the recoveries of BMAA and DAB in Songhua River samples

Removing NOM. A traditional C18 SPE cartridge (Oasis-HLB, Waters, USA), which was tested by pre-experiments that this cartridge cannot retained BMAA or DAB (less than 3%), was coupled with MCX to remove NOM (pure water spiked with 5 mg/L humic acid and 1µg/L

Validation of analytical procedures for BMAA in natural water samples. The analytical method developed in this study included two parts: SPE concentration and LC-MS/MS analysis. The SPE is performed in the form of tandem C18 and MCX cartridges. 50mL of samples are acidified with HCl to pH 3, loaded on the column at rate of 1 mL/min, washed by 10 mL 2% FA water and methanol sequentially, and eluted by 7.5 mL 5% NH3·H2O in methanol at a rate of 3 mL/min. Eluents are dried up under N2 blow at 55oC and then reconstituted as the method described in the section of SPE process. PVDF membranes are used to filter samples before the injection. BMAA and DAB were analyzed by underivatized LC-MS/MS method as described in the section of Analysis by LC/MS-MS. BMAA in the lab culture of M. aeruginosa, Songhua River and Kongmu Lake were analyzed in December, 2016 by the optimized method. No BMAA was found neither in the supernatant of M. aeruginosa nor in Songhua River samples, but there is a very weak signal of BMAA was observed (lower than LOD) in the Kongmu Lake samples. Since the samples

7

ACS Paragon Plus Environment

Analytical Chemistry

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

were collected in winter, the current results cannot represent the actual concentrations of BMAA in summer when the algae are flourishing and further analytical work was required. To verify the method, BMAA standards were spiked into the supernatant of M. aeruginosa and waters from Songhua River at a concentration of 1µg/L. The recoveries of BMAA in algal solution and in river water were 88.36±0.99% and 89.56±2.47%, respectively. The result indicated the developed method in the present study was qualified for the measurement of dissolved BMAA in actual water samples.

Page 8 of 9

The details of the optimization of the SPE process; pictures of LC-MS/MS chromatograms for BMAA, DAB, AQC-BMAA and AQC-DAB, standard working curves, and spectrogram and daughter ions of BMAA and DAB (PDF) AUTHOR INFORMATION Corresponding Author Zhiquan Liu, [email protected]; Fuyi Cui, [email protected], [email protected]

Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. ‡These authors contributed equally. (match statement to author names with a symbol)

CONCLUSION A highly selective determination method for dissolved BMAA, including pretreatment procedure and analytical method, was investigated in the present study, and the key findings were listed below. 

The sensitivity and precision of underivatized and derivatized LC-MS/MS methods were comparable and both methods were qualified for the measurement of BMAA and DAB in natural water.



The SPE procedure was optimized and the best conditions were described below: (1) Using tandem C18 and MCX cartridges; (2) loading acidified pH 3 samples of 50 mL at a loading rate of 1min/L; (3) washing with 10mL of 2% formic acid and 10mL of methanol; (4) eluting with 7.5mL of 5% NH3·H2O at a rate of 3mL/min; (5) drying the eluted samples at 55oC and (6) filtering by PVDF membrane.



The competition of NOM in natural water greatly reduced the recovery of BMAA, and the SPE performance could be improved by adjusting pH to 3, using a C18 cartridge for pretreatment and reducing the volume of samples to 50 mL.



A high recovery of dissolved BMAA, more than 88% in natural water and algal solution, could be obtained by using the optimized method in the present study.

ACKNOWLEDGEMENTS

We would like to show our highly appreciations to Mr. Casey Finnerty, a Ph.D candidate in Dept. of Civil & Environmental Engineering at the University of California, Berkeley, for his kindly linguistic modification. This work was supported by National Natural Science Foundation of China (Grant No. 51508130), Open Project of State Key Laboratory of Urban Water Resource and Environment (Grant No. ES201511), Fundamental Research Funds for the Central Universities (Grant No. HIT.NSRIF.201673) and HIT Environment and Ecology Innovation Special Funds (No. HSCJ201606). ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website.

Notes Any additional relevant notes should be placed here.

REFERENCES (1) Vega, A.; Bell, E. A. Phytochemistry1967, 6, 759-763. (2) Lobner, D.; Piana, P. M. T.; Salous, A. K.; Peoples, R. W. Neurobiol Dis2007, 25, 360-366. (3) Spencer, S. P.; Nunn, P. B.; Hugon, J.; Ludolph, A. C.; Ross, S. M.; Roy, D. N.; Robertson, R. C. Science 1987, 237 (4814), 517-522. (4) Murch, S.; Cox, P. A.; Banack, S. A.; al., e. ActaNeurologicaScandinavica2004, 110 (4), 267-269. (5) Karlsson, O.; Lindquist, N. G.; Brittebo, E. B.; Roman, E. ToxicolSci2009, 109 (2), 286-95. (6) Karlsson, O.; Jiang, L.; Ersson, L.; Malmstrom, T.; Ilag, L. L.; Brittebo, E. B. Sci Rep2015, 5, 15570. (7) Karlsson, O.; Jiang, L.; Andersson, M.; Ilag, L. L.; Brittebo, E. B. ToxicolLett2014, 226 (1), 1-5. (8) Karlsson, O.; Berg, A.-L.; Hanrieder, J.; Arnerup, G.; Lindström, A.-K.; Brittebo, E. B. Arch Toxicol2015, 89, 423-436. (9) Cox, P. A.; Banack, S. A.; Murch, S.; Rasmussen, U.; Tien, G.; Bidigare, R. R.; Metcalf, J. S.; Morrison, L. F.; Codd, G. A.; Bergman, B. ProcNatlAcadSci U S A2005, 102 (14), 5074–5078. (10) Jiang, L.; Eriksson, J.; Lage, S.; Jonasson, S.; Shams, S.; Mehine, M.; Ilag, L. L.; Rasmussen, U. PLoS One2014, 9 (1), e84578. (11) Lage, S.; Costa, P. R.; Moita, T.; Eriksson, J.; Rasmussen, U.; Rydberg, S. J. AquatToxicol2014, 152, 131-8. (12) Cox, P. A.; Banack, S. A.; Murch, S. J. ProcNatlAcadSci U S A2003, 100 (23), 13380-3. (13) Bidigare, R. R.; Christensen, S. J.; Wilde, S. B.; Banack, S. A. AmyotrophLatSclFr2009, 10 (s2), 71–73. (14) Jonasson, S.; Eriksson, J.; Berntzon, L.; Spacil, Z.; Ilag, L. L.; Ronnevi, L. O.; Rasmussen, U.; Bergman, B. ProcNatlAcadSci U S A2010, 107 (20), 9252–9257. (15) Jiao, Y.; Chen, Q.; Chen, X.; Wang, X.; Liao, X.; Jiang, L.; Wu, J.; Yang, L. Sci Total Environ 2014, 468-469, 457-63. (16) Chatziefthimiou, A. D.; Metcalf, J. S.; Glover, W. B.; Banack, S. A.; Dargham, S. R.; Richer, R. A. Toxicon2016, 114, 75-84. (17) Metcalf, J. S.; Banack, S. A.; Richer, R.; Cox, P. A. J Arid Environ2015, 112, 140-144. (18) Rosen, J.; Westerberg, E.; Schmiedt, S.; Hellenas, K. E. Toxicon2016, 109, 45-50. (19) Reveillon, D.; Sechet, V.; Hess, P.; Amzil, Z. Toxicon2016, 110, 35-46. (20) Beach, D. G.; Kerrin, E. S.; Quilliam, M. A. Anal BioanalChem2015, 407 (28), 8397-409. (21) Reveillon, D.; Abadie, E.; Sechet, V.; Brient, L.; Savar, V.; Bardouil, M.; Hess, P.; Amzil, Z. Mar Drugs2014, 12 (11), 5441-67.

8

ACS Paragon Plus Environment

Page 9 of 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

Analytical Chemistry

(22) Chen, Y. T.; Chen, W. R.; Liu, Z. Q.; Lin, T. F. Environ SciTechnol2017, 51 (3), 1303-1311. (23) Faassen, E. J. Toxins2014, 6 (3), 1109-38. (24) Metcalf, J. S.; Banack, S. A.; Lindsay, J.; Morrison, L. F.; Cox, P. A.; Codd, G. A Environ Microbiol2008, 10 (3), 702-8. (25) Faassen, E. J.; Gillissen, F.; Lurling, M. PLoS ONE2012, 7 (5), 1-8. (26) Groth, I.; Schumann, P.; Weiss, N.; Martin, K.; Rainey, F. A., Agrococcusjenensis gen. nov., sp. nov.Int J SystBacteriol1996, 46, 234-239. (27) Kubo, T.; Kato, N.; Hosoya, K.; Kaya, K. Toxicon2008, 51 (7), 1264-8. (28) Rosen, J.; Hellenas, K. E. Analyst2008, 133 (12), 1785-9. (29) Faassen, E. J.; Gillissen, F.; Zweers, H. A.; Lurling, M. Amyotroph Lateral Scler2009, 10 Suppl 2, 79-84. (30) Murch, S. J.; Cox, P. A.; Banack, S. A. ProcNatlAcadSci U S A 2004, 101 (33), 12228-31. (31) Cohen, S. A. Analyst2012, 137 (9), 1991-2005. (32) Elisabeth, J. F.; Antoniou, M. G.; Beekman-Lukassen, W.; Blahova, L.; Chernova, E.; Christophoridis, C.; Combes, A.; Edwards, C.; Fastner, J.; Harmsen, J.; Hiskia, A.; Ilag, L. L.; Kaloudis, T.; Lopicic, S.; Lurling, M.; Mazur-Marzec, H.; Meriluoto, J.; Porojan, C.; Viner-Mozzini, Y.; Zguna, N. Mar Drugs2016, 14 (3). (33) Banack, S. A.; Caller, T.; Henegan, P.; Haney, J.; Murby, A.; Metcalf, J. S.; Powell, J.; Cox, P. A.; Stommel, E. Toxins2015, 7 (2), 322-36. (34) Combes, A.; ElAbdellaoui, S.; Sarazin, C.; Vial, J.; Mejean, A.; Ploux, O.; Pichon, V.; group, B. Anal ChimActa2013, 771, 42-9. (35) Guo, B.; Yu, H.; Gao, B. Y.; Rong, H. Y.; Dong, H. Y.; Ma, D. F.; Li, R. H.; Zhao, S. Colloid Surface. A2015, 481, 476-484.

(36) Nicolas, M.; Sophie, L.; Sylvia, M.; Francois, D.; Patrick, M.; James, N. H. Environ SciTechnol2010, 44, 1713-1719. (37) Li, A.; Fan, H.; Ma, F.; McCarron, P.; Thomas, K.; Tang, X.; Quilliam, M. A. Analyst2012, 137 (5), 1210-9. (38) Wu, T.; Wang, C.; Wang, X.; Xiao, H. Q.; Ma, Q.; Zhang, Q. Chromatographia2008, 68 (9-10), 803-806. (39) Mckay, G.; Oppenhauser, M. S.; Sweeney, A. G. Water Res1980, 14, 15-20. (40) Badruzzaman, M.; Westerhoff, P.; Knappe, D. R. U. Water Res 2004, 38 (18), 4002-12. (41) Allen, S. J.; McKay, G.; Khader, K. Y. H. Environ Pollut1989, 56, 39-50. (42) Jiang, L. Y.; Aigret, B.; De Borggraeve, W. M.; Spacil, Z.; Ilag, L. L. Anal BioanalChem2012, 403 (6), 1719-30. (43) Spacil, Z.; Eriksson, J.; Jonasson, S.; Rasmussen, U.; Ilag, L. L.; Bergman, B. Analyst2010, 135 (1), 127-32. (44) Jiang, L. Y.; Johnston, E.; Aberg, K. M.; Nilsson, U.; Ilag, L. L. Anal BioanalChem2013, 405 (4), 1283-92. (45) Banack, S. A.; Murch, S. J.; Cox, P. A. J Ethnopharmacol2006, 106 (1), 97-104. (46) Fan, H.; Qiu, J.; Fan, L.; Li, A. Environ SciPollut Res2015, 22 (8), 5943-51. (47) Pearse, M. C.; Alan, C. L.; Sabrina, D. G.; Michael, A. Q. Aquatic Biosystems2014, 10 (1), 5. (48) Kaminska, G.; Bohdziewicz, J.; Calvo, J. I.; Prádanos, P.; Palacio, L.; Hernández, A. J Membrane Sci2015, 493, 66-79. (49) Roy-Lachapelle, A.; Solliec, M.; Sauve, S. Anal BioanalChem2015, 407 (18), 5487-501.

For TOC only

9

ACS Paragon Plus Environment