Mass Spectrometry of Non-Protein Amino Acids: BMAA and Neurodegenerative Diseases

Mass Spectrometry of Non-protein Amino Acids: BMAA and Neurodegenerative Diseases

Liying Jiang

Doctoral Thesis in Analytical Chemistry at Stockholm University, Sweden 2015

Mass Spectrometry of Non-protein Amino Acids: BMAA and Neurodegenerative Diseases

Liying Jiang

Doctoral Thesis in Analytical Chemistry Department of Environmental Science and Analytical Chemistry Stockholm University 2015

Academic dissertation for the Degree of Doctor of Philosophy in Analytical Chemistry at Stockholm University to be publicly defended at 10:00 am on Thursday 9th of April 2015 in Magnélisalen, Kemiska övningslaboratoriet, Svante Arrheniusväg 16 B, Stockholm, Sweden

Doctoral thesis in Analytical Chemistry Department of Environmental Science and Analytical Chemistry (ACES)

ISBN 978-91-7649-028-0 Printed by Holmbergs, Malmö, Sweden, 2015

Distributor: Department of Environmental Science and Analytical Chemistry, Stockholm University

To my parents, Qunfeng and Yitang

献给我的父母,群锋和忆唐 Abstract

Neurodegenerative diseases have been shown to correlate positively with an ageing population. The most common neurodegenerative diseases are amyotrophic lateral sclerosis (ALS), Parkinson’s disease and Alzheimer’s disease. The cause of these diseases is believed to be the interaction between genetic and environmental factors, synergistically acting with ageing. BMAA (β-methylamino-L-alanine) is one kind of toxin present in our environment and might play an important role in the development of those diseases.

BMAA was initially isolated from cycad seeds in Guam, where the incidence of ALS/Parkinsonism-dementia complex among the indigenous people was 50 – 100 times higher than the rest of the world in the 1950’s. BMAA can induce toxic effects on rodents and primates. Furthermore, it can potentiate neuronal injury on cell cultures at concentrations as low as 10 µM. BMAA was reported to be produced by cyanobacteria, and could bio-magnify through the food chain.

In this thesis, work was initially focused on the improvement of an existing analytical method for BMAA identification and quantification using liquid chromatography coupled with tandem mass spectrometry. Subsequently, the refined method was applied to environmental samples for probing alternative BMAA producer(s) in nature and to seafood samples for estimation of human exposure to this toxin.

In Paper I, a systematic screening of the isomers of BMAA in a database was performed and seven potential isomers were suggested. Three of them were detected or suspected in natural samples. In Paper II, a deuterated internal standard was synthesized and used for quantifying BMAA in cyanobacteria. In Paper III, Diatoms were discovered to be a BMAA producer in nature. In Paper IV, ten popular species of seafood sold in Swedish markets were screened for BMAA. Half of them were found to contain BMAA at a level of 0.01 – 0.90 µg/g wet weight. In Future perspectives, the remaining questions important in this field are raised.

Abstrakt

Neurogenerativa sjukdomar har visats korrelera positivt hos en åldrande population i Guam. De vanligaste neurogenerativa sjukdomarna är amyotrofisk lateral skleros (ALS), Parkinson’s och Alzheimer’s sjukdom. Orsaken till dessa sjukdomar tros ligga bakom interaktioner mellan genetiska och miljöfaktorer,som synergiskt agerar med åldrandet. BMAA (β-methylamino-L- alanine)är ett toxin som finns närvarande i våra miljöer och kan spela en viktig roll i utvecklandet av dessa sjukdomar.

BMAA var ursprungligen isolerad från cycad frön i Guam, där förekomsten av ALS/Parkinsonism-demenskomplex bland de infödda invånarna låg på en nivå 50 - 100 gånger högre än vid resten av världen under 1950-talet. BMAA kan inducera toxiska effekter hos gnagare och primater. Dessutom, kan det förstärka skador i cell kulturer vid koncentrationer så låg som 10 µM. BMAA som ska produceras från cyanobakterier kan biomagnifieras via mat kedjor.

I denna avhandling, var arbetet initialt fokuserad på förbättring av en existerande analytisk metod för BMAA identifiering och kvantifiering vid användning utav vätske kromatografi kopplad med en trippel quadropol masspektrometer. Därefter, appliceras den förbättrade metoden på miljöprover för screena alternativt BMAA producenter in natur och skaldjur för en uppskattning för en mänsklig exponering till detta toxin.

I Artikel I, ett systematiskt screening av isomerer från BMAA i databasen utfördes och sju potentiella isomerer föreslogs. Tre av dessa detekterades eller misstänktes i naturliga prov. I Artikel II, en denaturerad intern standard syntetiserades och användes för kvantifiering utav BMAA i cyanobakterier. I Artikel III, kiselalger upptäcktes vara en BMAA producent i naturen. I Artikel IV, tio populära fisk och skaldjur som säljs i den svenska marknaden undersöktes med avseende för BMAA. Hälften av dem påvisades innehålla BMAA vid en nivå av 0.01 – 0.90 µg/g i våtvikt. För Framtida perspektiv, den kvarvarande viktiga frågan i detta område är belysta.

List of Publications

I. Jiang L., Aigret B., De Borggraeve W. M., Spacil Z. & Ilag L. L. (2012) Selective LC- MS/MS method for the identification of BMAA from its isomers in biological samples. Analytical and Bioanalytical Chemistry 403: 1719-1730 DOI: 10.1007/s00216-012-5966-y The author was responsible for the design of the study, all the experimental work except chemical synthesis, data evaluation and a major part of writing the Paper.

II. Jiang L., Johnston E., Åberg K. M., Nilsson U. and Ilag L. L. (2013) Strategy for quantifying trace levels of BMAA in cyanobacteria by LC/MS/MS. Analytical and Bioanalytical Chemistry 405: 1283-1292 DOI: 10.1007/s00216-012-6550-1 The author was responsible for partially generating the idea, design of the study, all the experimental work except chemical synthesis, data evaluation and writing the Paper.

III. Jiang L., Eriksson J., Lage S., Jonasson S., Shams S., Mehine M., Ilag L. L. and Rasmussen U. (2014) Diatoms: A Novel Source for the Neurotoxin BMAA in Aquatic Environments. PLoS ONE 9(1): e84578 DOI:10.1371/journal.pone.0084578 The author was responsible for partially developing the sample preparation method, LC- MS/MS analysis, data evaluation and a major part of writing the Paper.

IV. Jiang L., Kiselova N., Rosén J. and Ilag L. L. (2014) Quantification of neurotoxin BMAA (β-N-methylamino-L-alanine) in seafood from Swedish markets. Scientific Reports 4: 6931 DOI: 10.1038/srep06931 The author was responsible for generating the idea, a part of design of the study, a part of the experimental work, data evaluation and writing the Paper.

Permissions to reproduce the publications were kindly obtained from the publishers. Publications/Manuscripts not included in the thesis:

V. Jiang L., Dziedzic P., Spacil Z., Zhao G. L., Nilsson L., Ilag L. L. and Cordova A. (2014) Abiotic synthesis of amino acids and self-crystallization under prebiotic conditions. Scientific Reports 4: 6769 DOI: 10.1038/srep06769 The author was responsible for a part of the experiments, data evaluation and a major part of writing the Paper.

VI. Jiang L. and Ilag L. L. (2014) Detection of endogenous BMAA in dinoflagellate (Heterocapsa triquetra) hints at evolutionary conservation and environmental concern. PubRaw Science 1 (2): 1-8 The author was responsible for all the experiments, data evaluation and a part of writing the Paper.

VII. Karlsson O., Jiang L., Andersson M., Ilag L. L. and Brittebo E. B. (2014) Protein association of the neurotoxin and non-protein amino acid BMAA (β-N-methylamino-L- alanine) in the liver and brain following neonatal administration in rats. Toxicology Letters 226.1: 1-5 DOI: 10.1016/j.toxlet.2014.01.027. The author was responsible for a part of the experiments, a part of data evaluation and a part of writing the Paper.

VIII. Banack S. A., Metcalf J. S., Jiang L., Craighead D., Ilag L. L. and Cox P. A. (2012) Cyanobacteria produce N-(2-aminoethyl)glycine, a backbone for Peptide nucleic acids which may have been the first genetic molecules for life on Earth. PLoS One 7:e49043 DOI: 10.1371/journal.pone.0049043 The author was responsible for a part of the experiments and a part of data evaluation.

IX. Xie X., Backman D., Lebedev T. A., Artaev B. V., Jiang L., Ilag L. L., and Zubarev A. R. (2015) Primordial soup was edible: abiotically produced Miller-Urey mixture supports bacterial growth. Nature Communications, submitted The author was responsible for a part of the experiments, a part of data evaluation and a part of writing the Paper.

X. Karlsson O., Jiang L., Ersson L., Malmström T., Ilag L. L. and Brittebo B. E. Protein association/binding of the neurotoxin BMAA in neonatal rat tissues. (manuscript) The author was responsible for a part of the experiments, a part of data evaluation and a part of writing the Paper.

Table of Contents

1. Introduction ...... 13

1.1 BMAA and neurodegenerative diseases ...... 13

1.2 BMAA toxicity...... 14 1.2.1 Toxicokinetics ...... 14

1.2.2 In vivo study ...... 15

1.2.3 In vitro study ...... 15

1.3 BMAA natural production by phytoplankton ...... 16 1.3.1 The debates about BMAA production by cyanobacteria...... 16

1.3.2 BMAA production by diatoms and dinoflagellates ...... 17

1.3.3 The natural role of BMAA ...... 17

1.4 Case reports of BMAA in food ...... 18

1.5 BMAA analysis ...... 19 1.5.1 BMAA and its isomers ...... 19

1.5.2 Chemical derivatization and instrumental analysis...... 20

2. Aims of the thesis...... 22

3. Methods ...... 23

3.1 Sample preparation ...... 23

3.2 LC-MS/MS analysis...... 24

3.3 Method validation ...... 26

4. Results ...... 28

4.1 BMAA identification and quantification by LC-MS/MS (Paper I, II and IV)...... 28

4.2 BMAA producers in nature (Paper III)...... 29

4.3 BMAA in seafood from Swedish markets (Paper IV) ...... 30

5. Conclusions and comments...... 31

6. Future perspectives ...... 32

Acknowledgements...... 34

References...... 37

Abbreviations

AD Alzheimer’s disease AEG N-(2-aminoethyl) glycine ALS Amyotrophic lateral sclerosis ALS-PDC Amyotrophic lateral sclerosis-Parkinson dementia complex AQC 6-Aminoquinolyl-N-hydroxysuccinimidyl carbamate b. w. Body weight BAMA β-amino-N-methylalanine BBB Blood-brain barrier BMAA β-Methylamino-L-alanine CNS Central nervous system CSF Cerebrospinal fluid d3-BMAA Deuterium-labeled BMAA d. w. Dry weight Da Dalton DAB α-γ-Diamino butyric acid DAD Diode-array detector DCM Dichloromethane DNS Dansyl chloride DOPA 3,4-Dihydroxyphenylalanine ECF Ethyl chloroformate EI Electron impact ionization ESI Electrospray ionization FLD Fluorescence detector FMOC 9-Fluorenylmethoxycarbonyl chloride GC-MS Gas chromatography coupled with mass spectrometry HD Huntington’s disease i. v. intravenous injection LC-MS/MS Liquid chromatography coupled with tandem mass spectrometry LLE Liquid-liquid extraction LOD Limit of detection LOQ Limit of quantification m/z Mass-to-charge ratio mGluR5 Metabotropic glutamate receptor 5 MAM Methylazoxymethanol NICI Negative ion chemical ionization NMDA N-methyl-D-aspartate ODAP β-N-oxalyl-α, β-diaminopropionic acid PD Parkinson’s disease PFPA Pentafluoropropionic anhydride pKa Negative logarithm of acid dissociation constant Ka PNA Peptide nucleic acid QC Quality control

RNA Riboneucleic acid SPE Solid phase extraction t1/2 Half-life TCA Trichloroacetic acid TFA Trifluoroacetic acid TFE Trifluoroethanol Th Thomson UHPLC Ultra-high performance liquid chromatography w. w. Wet weight

1. Introduction 1.1 BMAA and neurodegenerative diseases Neurodegenerative diseases are a group of hereditary or sporadic conditions characterized by progressive dysfunction, degeneration and death of specific populations of neurons which are often synergistically interconnected. The most common neurodegenerative diseases are amyotrophic lateral sclerosis (ALS), Parkinson’s disease (PD) and Alzheimer’s disease (AD).

In the 1950’s, the incidence of amyotrophic lateral sclerosis-Parkinson dementia complex (ALS-PDC) among the local people (Chamorro) in Guam, a territory of the United States in the western Pacific Ocean, was estimated to be 50-100 times higher than the one in the rest of the world1. The high incidence stimulated the U.S. National Institute of Neurological Disorders and Stroke to launch a registry for the cases of ALS-PDC in Guam in 1956. Epidemiological studies revealed the importance of inheritance2, and interaction with the environmental factors3 for ALS-PDC susceptibility among Chamorro people. In a cohort study that started in 1968 and lasted 15 years, 23 people developed PD among 965 Chamorro people involved. Traditional foods and ageing were the only two significant variables that correlated with PD incidence4.

In the traditional Chamorro diet, cycads have been long used as food and medicine for Chamorro. The toxic effects accompanying with the cycad’s consumption stimulated researchers to investigate the toxic components that might be related to ALS-PDC in Guam5. β-Methylamino-L-alanine (BMAA) was first isolated and identified from the seeds of cycad tree (Cycas circinalis) in 19676, when searching for the neurotoxin β-oxalyl-L-α,β- diaminopropionic acid (ODAP) or its analogs for lathyrism. The toxic effects of BMAA were demonstrated on chickens and rats7, indicating that it might be a causative agent for the ALS- PDC disease in Guam. However, the co-occurrence of other toxins in cycad seeds5, such as carcinogens cycasin, methylazoxymethanol (MAM), pakoein, macrozamin and sterol β- glucosides, makes the link between BMAA and neurodegenerative diseases ambiguous; the BMAA concentration in cycad seeds after a proper preparation decreased dramatically and was far to reach the level that can induce toxic effects observed in animal experiments; and the water-solubility of BMAA makes it unlikely to access the central neurological system through the blood-brain barrier (BBB).

The hypothesis that BMAA is a causative agent of neurodegenerative diseases was revitalized by two evidences discovered by Cox and coworkers: 1) Besides the free form, BMAA was also found to be present in a protein-bound form which may serve as an endogenous reservoir that slowly releases BMAA through protein metabolism8; and 2) There was a biomagnification of BMAA from cycads/cyanobacterial symbionts to flying foxes which can consume cycad seeds up to two and half times of their body weight per night9. Flying foxes were regarded as a desirable and luxury food material in Chamorros’ culture, and Chamorro people consumed flying foxes on ceremonial occasions. In the 1940’s, Americans built up their military bases in Guam and flying foxes became almost extinct by 1974 because of the

13 overhunting with the aid of firearms and the introduction of brown tree snakes10. Interestingly, the decrease of the incidence of ALS-PDC in Guam correlated with the decline of the flying fox population in Guam9, although the population of flying foxes was poorly estimated.

To investigate the hypothetical link between BMAA and neurodegenerative diseases, efforts have also been made to measure the presence of BMAA in the brain of patients who have died from neurodegenerative diseases. There were four reports about BMAA in human tissues so far. Among them, three BMAA-positive results were reported from Cox and his coworkers8, 11, 12 whereas another study reported negative results13. Both free and protein- bound BMAA were detected in brain tissues of Chamorro patients who died from ALS- PDC11; Canadian patients who died from AD8; and also North American patients who died from AD and ALS, but not those who suffered from Huntington’s disease (HD), another neurodegenerative genetic disorder12. The authors also commented that the observation of BMAA in Canadian and American patients suggested the possibility of human BMAA exposures through water supplies/marine food web that were contaminated by cyanobacteria, a phylum of bacteria that was shown by Cox’s group to be able to produce BMAA in nature14. However, another study by Montine and coworkers reported different results13. In this study, no free-BMAA was detected in brain tissues of eight Chamorro patients who died from PDC and five American patients who died from AD, although the limit of detection (LOD) in this study was orders of magnitude lower than the one reported by Cox’s and coworkers. The authors speculated that the contradictory data may be resulted from the different tissue fixation methods used, i.e. in their study, flash-frozen samples without fixation or preservation were analyzed, whereas in the studies of Cox and coworkers, samples were fixed in paraformaldehyde before the storage in a 15% sucrose maintenance solution for almost a decade which might produce artefacts. Since the four studies cited above, there have been no new reports about BMAA being detected in human tissues linked to neurodegenerative diseases.

1.2 BMAA toxicity

1.2.1 Toxicokinetics The oral bioavailability of BMAA in animals was estimated to be very high, i.e. more than 80 % in rats in 8 h after gastric intubation15; and 79 % at low dose (2 mg/kg) or 100 % at high dose (90 mg/kg) in monkeys in 48 h after gavage administration16. However, the blood-brain barrier permeability of BMAA was very limited, less than 0.08 % of the total dose in rat brains 2 h after intravenous injection (i. v.) with the BMAA concentration at 10 µg/g15, and less than 1 % of the total dose in mice17. The transport of BMAA into the brain was suggested to be through the large neutral amino acid carrier18 or cerebral capillary 17. The amount of BMAA reached a maximum in the brain of mouse at 1.5 h after bolus i. v. injection, and 80 % of total BMAA left in brain was present in protein-bound fraction after seven days17. However, the protein-bound BMAA detected in brains of neonatal rats was completely cleaned up from rat bodies in seven months19. BMAA was highly distributed across peripheral

14 tissues of rats, such as liver, kidney and muscle, and the distribution in central nervous system (CNS) was higher in pons, thalamus and cerebellum, but lower in spinal cord, cerebral cortex and cerebrospinal fluid (CSF) within 30 min15. The uptake of BMAA by liver or brain was a 15, 19 dose-dependent . The half-life (t1/2) of BMAA in those tissues was estimated to be around one day in plasma and half day in others, such as liver, skeletal muscle and brain15. After seven days, only less than 10 % of unchanged BMAA was excreted through urine and feces, thus the authors speculated that most of the BMAA was metabolized or bile was a significant pathway for BMAA elimination15, 16.

1.2.2 In vivo study Early studies showed that BMAA can induce convulsions, weakness and dragging gait in chickens, rats and mice after intraperitoneal injection of BMAA solution7, 20. However, the dosage that can induce these toxic effects was around 0.2 – 3.3 mg BMAA/g body weight (b.w.)/day which was higher than the BMAA concentration in cycad seeds. Later in another study21, neonatal rats (males and females) were administrated with 100 or 500 ng/kg (b.w.) of BMAA by subcutaneous injection. BMAA can produce permanent changes in adult (at 101 days of age) motor function and spinal cord neurochemistry, which may be related to the reorganizational effects induced by toxin-mediated neuroplasticity in developing neurons. The neurochemical changes were both treatment and sex dependent, but the pattern was not consistent with the changes reported in postmortem tissue from ALS patients.

There were three studies about BMAA toxicity on animals using oral administration route, which is relevant to human exposure to this toxin. In 1987, Spencer et al., published a landmark study on monkeys (Macaca fascicularis)22. The monkeys were administrated 100 – 315 mg/kg (b.w.)/day by gavage for 5 – 17 weeks. Animals developed corticospinal dysfunction, Parkinsonian features (such as expressionless face, blank stare and shuffling gait, etc.) and behavioral anomalies, with degenerative changes of motor neurons in cerebral cortex and spinal cord. Symptoms were selectively attenuated within 30 min by L-3, 4- dihydroxyphenylalanine (L-DOPA) an antiParkinsonian drug that acts as an antagonist to N- methyl-D-aspartate (NMDA) receptor, with a peripheral decarboxylase inhibitor. This study has been criticized for using unrealistic dosages. Two other studies using oral administration failed to induce behavior abnormalities, and neurochemical or neuropathological changes on mice23, 24. The doses used were 500 – 1000 mg/kg (b.w.)/day for 11 weeks23 and 28 mg/kg (b.w.)/day for 30 days24, respectively. The authors commented that the mouse might not be a suitable animal model for BMAA toxicological tests.

1.2.3 In vitro study There were a number of studies that investigated the mechanism of BMAA toxicity using brain slices, primary cells, spinal cord cultures and neuronal cultures. BMAA carbamate, a form that is structurally similar to glutamate, has been suggested to be transported into the brain and act as an excitotoxin25. BMAA was shown to be a weak glutamate receptor agonist that can cause neurotoxicity at high concentrations in the mM range, whereas most excitotoxins are toxic in the µM range. However, one study showed that BMAA at concentrations as low as 10 µM, together with other insults, can potentiate neuronal injury using mixed cortical cell cultures containing neurons and astrocytes, mechanistically as an

15 agonist for NMDA and metabotropic glutamate receptor 5 (mGluR5) and induction of oxidative stress26. 1.3 BMAA natural production by phytoplankton

1.3.1 The debates about BMAA production by cyanobacteria BMAA was thought to be endogenously formed in the cycad plants during the first half century of BMAA research. In 2003, Cox et al. found that BMAA was present in symbiotic cyanobacteria (Nostoc) that live in coralloid roots of the cycad tree27. Following this line, they further analyzed free-living cyanobacteria that were isolated worldwide and cultured in Stockholm, Manoa and Dundee. BMAA was detected in 95 % of the cyanobacterial genera; 97 % of the strains of free-living cyanobacteria that cover all major taxonomies; as well as 73 % of symbiotic cyanobacteria (Nostoc) in lichen, cycad and other plants, at µg – mg/g dry weight (d.w.) range in both free and protein-bound forms14.

These finding have been confirmed by other research groups thereafter: BMAA was detected in all 12 cyanobacterial samples from fresh/brackish water bodies in Britain used for drinking water/recreation at µg/g (d.w.) level28; 96 % of 27 cyanobacteria collected from fresh water in South Africa contain BMAA at µg – mg/g (d.w.) range29; 43 % of 21cyanobacterial scums from Dutch urban waters contain BMAA at low µg/g (d.w.) level30; all 21 cyanobacterial samples from the Baltic Sea contain BMAA at ng/g (d.w.) level31; and the latest study showed that, all 10 cyanobacterial samples from Tai lake in China contain BMAA at low µg/g (d.w.) level32. Most of these studies employed a liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) approach for analysis of BMAA28, 30-32, and only one study used gas chromatography coupled with mass spectrometry (GC-MS)29.

However, in other studies employing equally reliable analytical methodology, BMAA was absent in cyanobacterial samples tested, such as: 1) 30 laboratory cultured cyanobacterial strains, four cyanobacterial bloom samples from the Baltic Sea and two commercial Spirulina tablets33; 2) 30 free-living cyanobacteria mainly from China and Germany, and 32 Spirulina supplements34; 3) four cyanobacterial scums from Dutch urban water and four laboratory cyanobacterial strains35; and most recently, 4) eight commercial Spirulina products36.

The contradictory results prevalent in this field have to be resolved by further investigations. These discrepancies can be caused by the inherent properties of the analytical method used, such as insufficient sensitivity and/or selectivity; or by the differences in sample materials, such as laboratory culture conditions and sample impurity. Notably, most of the negative results were generated by laboratory raised cyanobacteria or nutrition supplements which contain only Spirulina, while most of the positive results were generated by field cyanobacteria. An explanation for this discrepancy has been attributed to the difficulty in mimicking the natural environmental conditions that allow cyanobacteria to produce BMAA under laboratory conditions. Researchers have put considerable effort to investigate the optimal culture conditions for cyanobacteria to produce BMAA in laboratory. Until now, no research group could succeed in consistently demonstrating that cyanobacteria can substantially produce BMAA under laboratory conditions.

1.3.2 BMAA production by diatoms and dinoflagellates Another possible reason for the disagreement regarding BMAA production between laboratory and field cyanobacterial samples could be that other organism(s) in field cyanobacterial samples might contain BMAA and give a positive signal. Indeed, in 2014, we demonstrated that diatoms are another BMAA producer in nature37; possibly dinoflagellates as well38, 39.

Interestingly, in some of the BMAA-positive reports mentioned above, Metcalf et.al. did mention that field samples also contained diatoms besides cyanobacteria28; Faassen et al. stated that “dominant species were qualitatively determined by light microscopy” which indicated there were other species in the samples as well30; and Jiao et al., described that “the collected cyanobacteria were separated from the zooplankton using a light trap” which did not mention if other phytoplankton (such as diatoms and dinoflagellates) than zooplankton were separated32. Field cyanobacterial bloom, scum or mat can contain many other organisms besides cyanobacteria. If those cyanobacteria were not carefully purified, BMAA might be generated by other organisms present in the samples.

1.3.3 The natural role of BMAA There is no solid evidence to support that cyanobacteria can produce substantial amounts of BMAA. Nevertheless, efforts have been made to investigate the role of BMAA in cyanobacteria. In cycads, BMAA was initially suspected to serve as an anti-herbivory compound and for nitrogen storage27. In cyanobacteria, it was observed that BMAA was synthesized or produced as a result of catabolism in response to nitrogen starvation40. It was reported in both non-nitrogen-fixing and nitrogen-fixing cyanobacteria, that exogenous BMAA can result in growth arrest and induction of chlorosis41, 42. With the observation of the accumulation of massive cellular glycogen, a carbon-storage molecule, BMAA was suggested to function as nitrogen fixation inhibitor through deactivating nitrogenase, rather than acting as a nitrogen source42. Instead, a mechanism related to impaired glutamine biosynthesis/degradation or oxidative stress was proposed42. BMAA seems detectable in the laboratory cultured cyanobacterial species used in these three studies. However, N-(2- aminoethyl) glycine (AEG), a structural isomer of BMAA that was commonly present in cyanobacteria43 was not included in their analytical methods. Thus it remains unclear if BMAA was distinguished from AEG. Regarding the fact that several studies reported absence of BMAA in many cyanobacterial species in laboratory cultures33, 35, there are concerns about the quality of the analytical methods used.

With the emergence of two other natural BMAA producers: diatoms37 and dinoflagellates38, 39, we suspected that BMAA may be a barometer for nitrogen levels which functions partly to be involved in nitrogen fixation signaling by cyanobacteria for itself and its symbionts, i.e. diatoms or dinoflagellates38. Interestingly, cyanobacterial symbionts can create anaerobic microenvironments to stabilize the nitrogenase for nitrogen fixation within the dinoflagellate host44. The authors in this study suspected that the natural role of BMAA might be related to nitrogen fixation signaling or biochemistry associated with plastids.

Besides the coexistence as symbionts, cyanobacteria, diatoms and dinoflagellates also exhibit close relationship in the natural environment. For instance, in the Baltic Sea, cyanobacteria bloom during the summer when nutrients are rich and temperature is high which favor their growth. Diatoms start blooming during the autumn when the temperature decreases, but the nutrient and light are still sufficient. During the spring, dinoflagellates become prevalent phytoplankton which coincide with the decline of diatoms45. The annual blooms of cyanobacteria, diatoms and dinoflagellates consequently make BMAA available in the food web of the ecosystem all year round.

Cyanobacteria, diatoms and dinoflagellates are the most common types of phytoplankton. They are ubiquitous organisms that can be found in almost all terrestrial and aquatic systems, from hot springs to deserts or even cold polar regions. Together, they play an important role in global carbon fixation and silicon regulation46. Together with plants, they are the primary producers that sustain directly or indirectly the entire food web. BMAA produced by these can potentially be transferred and bio-magnified up the trophic levels and eventually reach humans. Consequently, the health risk of BMAA is increased.

1.4 Case reports of BMAA in food Although the production of BMAA by cyanobacteria is still uncertain, the reports about BMAA-contaminated organisms from ecosystems or food markets have continuously emerged and have become more convincing over the time. This transition can be explained by improvements with time, of the analytical methods used for identification of BMAA and the discoveries of new BMAA producers, i.e. diatoms and dinoflagellates, in nature.

When BMAA was first discovered in cycad seeds in Guam, researchers limited their exploration in this Pacific island. Chamorro people used the flour that was made from the kernel of cycad seeds to make tortillas, dumpling, soups and cakes up to the 1950’s. BMAA was detected in a level up to g/kg in seed kernels47, 48. When preparing the cycad seed flour, Chamorros washed the seed kernels in water repeatedly for 1 – 3 weeks, since they were aware that the unwashed seed kernels are toxic49. The washed material contained much lower amounts of BMAA50. In a study by Banack et al., BMAA was detected at mg – g/kg range in flying foxes, and 3 mg/kg in deer hair. The authors commented that BMAA may also be present in pigs and land crabs that feed upon cycad seeds49.

After BMAA was reported to be produced by both symbiotic and free-living cyanobacteria worldwide, researchers have expanded their exploration beyond the island of Guam. BMAA has been detected in food made up of cyanobacteria. For instance, llullucha, a local food consumed by the indigenous people in the mountains of Peru, which are globular colonies of Nostoc commune (Nostocales) collected in highland lakes51; and fa cai, the cyanobacterium Nostoc flagelliforme that was a popular luxury food among the Chinese and fortunately has been banned by the Chinese government concerned with environmental protection52. BMAA has also been detected in various organisms from different aquatic ecosystems especially those with extensive algal blooms, for instance, in Florida Bay and Biscayne Bay in South

Florida which are eutrophic habitats caused by human activities53; in the Baltic Sea which is a temperate brackish water body with dramatically increased eutrophication over the past few decades31; and in Gonghu Bay of Tai lake in China which is another highly eutrophic fresh water body54. Those organisms analyzed cover a wide range of species that are used as human food, including fishes, crabs, blue mussels, oysters, shrimps, etc. However, it was not mentioned clearly in those studies how large the affected population is. Besides these studies with large scale samples, there were several small scale studies that reported BMAA in one or two organisms that are used as human food or nutritional supplements. Those foods include mussels and oysters from Thau lagoon in France where one significant ALS cluster was surrounding55; blue crabs that were frequently consumed by three sporadic ALS patients in Maryland, USA56; and commercial shark cartilage supplements in the USA57.

Although most of these studies were not designed to estimate the BMAA intake by the public, the wide spread/occurrence of BMAA in diverse organisms from those ecosystems or food markets raised public concern about human health risk posed by this toxin.

1.5 BMAA analysis

1.5.1 BMAA and its isomers BMAA, i.e. β-methylamino-L-alanine or L- α-amino-β-methylamino propionic acid was firstly isolated from Cycas circinalis seeds in 19676. It is a non-protein amino acid with a secondary amine on the side chain. Its monohydrochloride, a white solid salt, is soluble in water and has a melting point at 181 – 182 °C58. BMAA contains a carboxyl group, a primary amine attached to the α-carbon, and a secondary amine attached to the β-carbon. The corresponding pKa1, pKa2 and pKa3 were determined to be 2.1, 6.6 and 9.8 respectively59.

The big challenge in BMAA analysis using mass spectrometry is the presence of its structural isomers in a sample. Since the structural isomers share with BMAA the same molecular mass at 118 g/mol, fragmentation to generate structure-specific fragments, i.e. diagnostic product ions, is needed to improve method selectivity to ensure the reliable identification of BMAA. The first isomer that was discovered in samples during the BMAA analysis was α-γ-diamino butyric acid (DAB)33, a plant neurotoxin. The authors suggested the diagnostic product ions that can be used to distinguish BMAA and DAB in methods that employ either AQC- derivatization or a non-derivation approach, and use of the relative fragment ion ratios was also recommended. Later, two other structural isomers, i.e. β-amino-N-methylalanine (BAMA)60 and AEG43, 60 were discovered in natural samples. The former was detected in blue mussels and oysters collected from the west coast of Sweden, and the latter was detected in almost all cyanobacterial samples and was recognized as part of the backbone of the peptide nucleic acids (PNA), a putative information coding molecule that is suspected to be present on earth prior to RNA. The wide spread occurrence of AEG in cyanobacteria might shed some light on the controversy surrounding the BMAA production by cyanobacteria.

1.5.2 Chemical derivatization and instrumental analysis A number of reagents were used to derivatize BMAA before instrumental analysis to improve its separation, ionization and detection. In GC-methods, both amino and carboxyl groups of BMAA were modified to decrease the polarity so that the analyte is amenable for GC analysis. Duncan et al. developed a GC-MS method using pentafluoropropionic anhydride (PFPA) as derivatization reagent58. Firstly, the carboxyl group of BMAA was reacted with trifluoroethanol (TFE) to form an ester. Secondly, the sample was cleaned up and the analyte was extracted by dichloromethane (DCM). Finally, the analyte was derivatized with PFPA before GC-MS analysis. The qualification of BMAA was ensured by detection of molecular ion in negative ion chemical ionization (NICI) and fragments in electron impact ionization (EI), and the quantification was ensured by using deuterated-BMAA as internal standard. The method LOD was estimated to be 24 µg/g in this study and later 3 µg/g in another study61.

Pan et al. developed a GC-MS method using another reagent ethyl chloroformate (ECF). The method was only qualitative and both the intact molecular ion and reagent adducts were observed using CI. With similar derivatization principle, Tim Downing et al., developed both GC-MS and LC-MS/MS methods using EZ:faastTM amino acid analysis kit. In the GC-MS method29, positive EI was used and fragment 130 (m/z) was selected for BMAA identification. The LOD was estimated to be 84 pmol/injection in cyanobacterial matrix. In the LC-MS/MS method40, BMAA was identified by ultra-high performance liquid chromatography (UHPLC) retention time and precursor ion at 333 (m/z) and product ion at 245 (m/z) which unfortunately were shared by all isomers of BMAA. No LOD was reported in this study.

In LC-methods, BMAA was derivatized to improve LC separation, ionization when using electrospray ionization (ESI), or fluorescent detection when using fluorescence detector (FLD). BMAA was derivatized with fluorescent reagent, such as 9-fluorenylmethoxycarbonyl chloride (FMOC)62 or 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC) when using FLD8,14, 48, 63-65 or diode-array detector (DAD) detection27, 49. FLD offers very specific and selective detection of the compounds that usually contain conjugated double bonds on their structures thus largely reducing the background and ensures very low LOD. However, all primary and secondary amines/amino acids can be derivatized by FMOC and AQC, thus can be detected by FLD. Some of them might have similar chromatographic properties and co- elute with BMAA, thus generating false-positive qualitative and inaccurate quantitative results. Although in several studies mentioned above, the authors claimed that the identification of BMAA was validated by LC-MS8, 14, 27, 49 or GC-MS63 methods, the method selectivity is still not good enough to distinguish BMAA from its structural isomers.

Nowadays, LC-MS/MS analysis with the diagnostic fragment of BMAA has been commonly accepted as a reliable method for BMAA identification. Those studies include LC-MS/MS without derivatization30, 33, 34, 36, or with derivatization using AQC reagent19, 31, 32, 36, 37, 39, 39, 55, 57, 60, 66-69 or dansyl chloride (DNS) reagent70. Unlike BMAA isomers AEG and DAB, another BMAA isomer, i.e. BAMA can generate the same diagnostic product ion species as BMAA, but with different relative abundances. Thus, it is highly recommended to not only include

20 diagnostic product ion of BMAA, but also use product ion ratio between fragments to improve method selectivity for BMAA identification60.

2. Aims of the thesis There was an urgent need for the improvement of the analytical methodology, both qualitatively and quantitatively, to resolve the prevalent controversy about BMAA production by cyanobacteria. Furthermore, whether cyanobacteria produce BMAA exclusively in nature or other organism(s) can also be BMAA producer(s) remains an unresolved question. Finally, the BMAA dietary exposure of the public through the consumption of seafood from food markets has to be assessed. To answer these questions, the aims of this thesis were set as follows:

 To develop a reliable analytical method for the identification of BMAA in biological samples.

 To develop an accurate analytical method for the quantification of BMAA in biological samples.

 To explore new BMAA producer(s) in nature.

 To assess the exposure of the Swedish public to dietary BMAA through seafood consumption.

3. Methods 3.1 Sample preparation BMAA is expected to be present at low concentrations, between ng/g and µg/g (d.w.), in complex biological samples. Sample preparation is essential to sufficiently extract and enrich trace levels of the analyte, and also remove compounds from complex matrix prior to LC- MS/MS analysis. Therefore, a thorough sample pre-treatment method has been developed, including multiple steps, such as cell lysis, protein precipitation, hydrolysis, liquid-liquid extraction (LLE), solid phase extraction (SPE) and derivatization, as shown in Fig. 1. Since BMAA is known to be present in two forms in biological samples, i.e. one form as free amino acid and another form bound to proteins, two different workflows were developed, aimed at measuring the total amount of BMAA (Fig. 1 right panel, used in Papers I, II, IV) or free and protein-bound forms separately (Fig. 1 left panel, modified and used in Paper III).

water

Fig. 1: Sample preparation workflows for identifying and quantifying total BMAA (right panel) or free and protein-bound BMAA separately (left panel) in biological samples.

Compared to the sample preparation procedure that was developed previously in our group and others in the literature, there are several major modifications, namely: 1) water, instead of 80 % methanol, was used as extraction solvent to avoid the loss of proteins that can precipitate with organic solvent; 2) protein pellet was reconstituted and precipitated in 10 % TCA with two more cycles to remove possible free-BMAA residues in the pellet; 3) an LLE step was added prior to the SPE to remove lipophilic impurities, such as lipids and/or

23 chlorophyll in cyanobacterial and animal tissue samples; and 4) AQC derivatization protocol was modified to improve the derivatization efficiency. (See the protocol in details in Paper IV). 3.2 LC-MS/MS analysis Since Cox and coworkers reported BMAA production by cyanobacteria in 200327, a number of contradictory results have been generated regarding the detection of BMAA from these organisms. In 2008, Johan Rosén and Karl-Erik Hellenas reported the coexistence of BMAA isomer DAB in cyanobacterial samples and suggested to use diagnostic fragment ions and relative ion ratio to distinguish them33. This strategy has been demonstrated by Spacil et al., in 201066. However, the confusion surrounding the BMAA identification has not been completely resolved. In our early experience with cyanobacterial analysis, BMAA was detected at the correct retention time with a good signal from the product ion 119.08 (m/z), but the diagnostic product ion 258.09 (m/z) was either missing or at relatively low abundance compared to the standard (See Fig. 2 below). This observation intrigued us to investigate other potential BMAA isomers that could co-elute with BMAA and thus contribute to the signal of fragment 119 (m/z).

Fig. 2: LC-MS/MS chromatograms of BMAA standard (left panel) and a laboratory cultured cyanobacterial (Nostoc PCC8001) sample (right panel).

A systematic study of BMAA structural isomers has been performed in Paper I60. According to the sample pretreatment and LC-MS/MS methods we are using, the possible BMAA isomers were narrowed down to seven as listed in Fig. 3 (See more details in Paper I.). Among them, BAMA can produce the same fragment species but at different relative ratio compared to BMAA (See Fig. 4). Thus using chromatographic retention time and relative abundance of fragment 119.08 (m/z) and 258.09 (m/z) of BMAA and BAMA are crucial to distinguish them. Therefore, four criteria were used in our method to ensure reliable identification of BMAA, namely: 1) chromatographic retention time; 2) isolation of precursor

24 ions at 458.18 (m/z) for BMAA and its isomer derivates; 3) fragments of one general and one diagnostic ions for each isomer; and 4) peak area ratio between general and diagnostic fragments. (See Fig. 6 d).

To ensure an accurate quantification of BMAA in biological samples, deuterium-labeled

BMAA (d3-BMAA) was synthesized and spiked into the samples at the beginning of sample preparation procedure, compensating for the loss of BMAA during sample preparation, derivatization efficiency and matrix effects using ESI. Isotopic labeling results in an m/z shift of 3 thomson (Th) for both precursor ion and product ion at 119.08 (m/z) which were used for quantification (See Fig. 6 d). The peak area ratios of the transition 459.18 > 119.08 (m/z) from BMAA and the transition 462.20 > 122.10 (m/z) from d3-BMAA that were detected from a sample and from the standard solution were compared for quantification.

Fig. 3: The structures of BMAA and selected isomeric compounds (Reproduced from Paper I).

Fig. 4: The diagnostic fragments for the AQC double-labeled BMAA and its isomers.

To improve the detection limits, a post-column addition of organic solvent (acetonitrile) strategy was developed as shown in Fig. 5. The setting in detail can be found in Paper II.

Fig. 5: The sketch of LC-MS/MS instrument using post-column addition 3.3 Method validation The recovery of the whole sample preparation procedure until derivatization was estimated by spiking 2 mg of ground spirulina tablets with known amount of BMAA and d3-BMAA. The matrix effects were estimated by comparing the slopes of a matrix-adapted (using 2 mg of ground spirulina tablets) and matrix-free standard curve (See details in Paper II).

The linearity of the calibration curves for quantification was assessed using linear regression analysis. The calibration curve was evaluated by spiking 2 mg of ground spirulina tablet matrix with a dilution series of BMAA and a fixed amount of d3-BMAA internal standard before starting sample preparation procedure (Paper II and III). A similar calibration curve, but without matrix, was used for quantification in Paper IV.

The method LOD and LOQ were determined experimentally by spiking with different amounts of BMAA standard in 2 mg of ground spirulina tablet matrix (Paper II) or in 10 mg wet weight (w.w) of crayfish muscle matrix (Paper IV). LOD was estimated using the diagnostic product ion at 258.09 (m/z) and LOQ was estimated using the general product ion at 119.08 (m/z).

The accuracy and precision of the quantification method was evaluated by using the QC samples that were prepared in-house, spiked with low, medium and high levels of BMAA together with d3-BMAA in spirulina (Paper II) or crayfish muscle matrix (Paper IV) and determined on the same day or each of three consecutive days.

In all studies, 2 mg (d.w.) of broccoli was prepared in the same way as unknown samples and used as negative control sample to evaluate the possible contamination during sample pretreatment. AQC reagent blank was used across the whole sample batch analysis to evaluate the possible carryover in the LC-MS/MS system.

4. Results

4.1 BMAA identification and quantification by LC-MS/MS (Paper I, II and IV) In Paper I, surprisingly, BMAA and AEG co-eluted in our system when using a HPLC column with 0.1 % formic acid as mobile phase additive (Fig. 6 a). To improve LC resolution, TFA was added, but at a very low concentration at 0.005 % to minimize ion suppression (Fig. 6 b). In Paper II and III, a UHPLC column was used to improve the chromatographic separation and thus peak resolution without introducing TFA additive (Fig. 6 c). In Paper IV, a total LC run was shortened from 27 to 16 min using higher flow rate during washing and conditioning periods to improve sample throughput. Moreover, IS was added initially and identification and quantification were monitored simultaneously to decrease workload for large scale sample analysis (Fig. 6 d). Therefore, one HPLC-MS/MS and one UHPLC- MS/MS system were demonstrated here for versatile application. Each of them fulfills the four criteria mentioned above in the Methods section for unambiguous identification of BMAA, and also accurate quantification of BMAA in biological sample.

Fig. 6: LC-MS/MS chromatograms of BMAA and its isomer standards using HPLC column (Hypersil GOLDTM C18, 100 × 2.1 mm, 3 µm particle size): (a) with 0.1 % formic acid as mobile phase additive; (b) with 0.3 % acetic acid + 0.005 % TFA as mobile phase additive; or UHPLC column (ACCA-TAGTM ULTRA C18, 100 × 2.1 mm, 1.7 µm particle size) with 0.3 % acetic acid as mobile phase additive and total LC run for: (c) 27 min and (d) 16 min. (Reproduced from Paper I and IV)

The sample preparation recovery was estimated to be about 63.3 % at a concentration level of 1.25 µg BMAA/g spirulina tablet matrix (Paper I). The matrix effects were estimated to be 18 % of signal enhancement in 2 mg of spirulina matrix (Paper I). The method LOD and LOQ were estimated at 0.1 µg/g (d.w.) of spirulina matrix (Paper I) and 0.01 µg/g (w.w.) of crayfish muscle matrix (Paper IV). The calibration curve exhibited good linearity (R2 = 0.998) across the evaluated range. QC samples demonstrated good accuracy and intra- day/inter-day precision (See details in Paper II and IV).

4.2 BMAA producers in nature (Paper III) In Paper III, BMAA was detected in six axenic diatom species that were cultured in the laboratory. For field cyanobacterial/diatom samples, BMAA was not detectable in two field samples which were dominated by the benthic cyanobacteria without visible diatoms. This observation does not support the hypothesis that benthic cyanobacteria might produce more BMAA than do planktonic cyanobacteria. BMAA was also detected in one field sample which was mixed with cyanobacteria and diatoms. Further treatment of this field sample using germanium dioxide killed most of the diatoms, which was in line with the reduction of BMAA signal, thus indicating a correlation between the presence of diatoms and BMAA signal. Taken together, this study discovered that diatoms can produce BMAA in both axenic laboratory conditions and natural conditions.

This study also indicates that an axenic culture of dinoflagellate might also produce BMAA which was demonstrated by a follow-up study38. These studies expanded the BMAA natural producers to diatoms and dinoflagellates besides cyanobacteria.

4.3 BMAA in seafood from Swedish markets (Paper IV) Due to the ubiquity of phytoplankton (cyanobacteria, diatoms and dinoflagellates) and their central role in aquatic food web as primary producers, it is not difficult to infer that BMAA can be transferred and accumulated in aquatic food web and finally reach humans through seafood consumption.

In Paper IV, a systematic screening of BMAA exposure of a large population through consumption of seafood sold in metropolitan markets in Stockholm was conducted. BMAA was detected in all blue mussel, oyster, shrimp and plaice, and some of char and herring, but undetectable in salmon, cod, perch and crayfish. The content of BMAA in the BMAA- contaminated seafood was between 0.01 – 0.90 mg/kg. The average value of per capita consumption of seafood in Sweden during 2005 – 2007 was estimated to be 28.7 kg/year, i.e. 0.6 kg/week (according to the Food and Agriculture Organization of the United Nations). Therefore, the BMAA intake of Swedes through seafood consumption can be up to 540 µg per week.

5. Conclusions and comments Based on the four studies that this thesis covered, conclusions can be drawn as follows:

 A selective LC-MS/MS method has been developed to distinguish BMAA from its isomers AEG, DAB and BAMA for the reliable identification of BMAA in biological samples. For the analytical methods either with or without a derivatization step, it is crucial to include both general and diagnostic fragments and their relative abundance to improve the selectivity of the method. Since DAB and AEG were observed in most kinds of samples, it is highly recommended to include them into analytical method development, which some groups in BMAA research field are still not doing.

 An accurate quantitative method using deuterium-labeled BMAA as internal standard

has been developed and applied for the analysis of different sample matrices. d3- BMAA exhibits good stability in strong acid and high temperature conditions, and almost identical physicochemical properties, thus similar performance as BMAA during sample preparation and LC-MS/MS analysis. It is necessary to use isotopic- labeled internal standard or at least matrix-adapted calibration curve with external standard for accurate quantification of BMAA. However, we should remain in mind that isotopic labeled internal standards cannot eliminate error originating from analyte extraction efficiency during cell lysis and protein precipitation/hydrolysis. Furthermore, the possibility of unlabeled internal standard present in the synthesized isotopic-labeled internal standard has to be carefully examined to avoid false positive results.

 In addition to cyanobacteria, diatoms and probably dinoflagellates as well can produce BMAA in nature. This discovery suggests a new pathway of BMAA transfer and accumulation in the food web, and increases dramatically the human health concern posed by this toxin. It also highlights the importance to know the exact organism/species in field cyanobacterial samples so that the correct conclusion can be drawn regarding the origin of BMAA.

 BMAA was detected in blue mussel, oyster, shrimp, plaice, char and herring, but undetectable in salmon, cod, perch and crayfish, and quantified in a range of 0.01 – 0.90 µg/g (w.w.) of sample tissues. The BMAA intake for Swedes through seafood consumption can be up to 540 µg per week. According to the available data on BMAA toxicity currently, this intake is far to reach the level that can cause toxic effects on animal experiment. However, there are several perspectives that have to be kept in mind: 1) BMAA can potentially bio-accumulate in a protein-bound form and reach the level for the onset of disease in the long run; 2) Humans might have different sensitivity to this toxin compared to animals; 3) Besides seafood consumption, human might be exposed to this toxin by other routes, such as recreation in water bodies with algae blooming, nutritional supplements which contain algae, and even drinking water. Nevertheless, the BMAA intake estimated here can be used as a basis for refining the design of toxicological experiments.

6. Future perspectives There are still gaps in our knowledge before we can definitively link BMAA with neurodegenerative diseases or initiate a risk assessment of BMAA in the environment. To fill these gaps, the questions listed below have to be addressed in the near future:

 A fast and economically effective analytical method without compromising the method selectivity and detection limits has to be developed to improve method throughput for environmental monitoring, food screening or large scale clinical/animal sample analysis.

 It would be very valuable to establish a reference material, or perform proficiency tests in order to standardize and improve the analytical methods for BMAA identification and quantification used by different laboratories.

 The presence of BMAA in tissues from patient with neurodegenerative diseases has to be confirmed by using reliable analytical methodology and conducting large scale screening. It should be noted that, even if this connection is confirmed, BMAA still remains as an associated factor rather than the direct cause of these diseases.

 The evaluation of the toxicity of BMAA using standard protocols has to be performed, preferably in an oral administration route with a dose that is comparable to the degree of realistic human exposure over a long period of time, to investigate its neurotoxic potential, toxicokinetics and finally to estimate Tolerant Daily Intake.

 The interaction between BMAA and protein(s) has to be elucidated at the molecular level, and the BMAA-bound protein(s) have to be identified to understand the mechanism of toxicity of BMAA, and thus the potential pathway that may lead to neurodegenerative diseases.

 An analytical method to identify and quantify L- and D-form BMAA in biological samples has to be developed. Accordingly, animal studies on D-BMAA can be performed in order to investigate the possible therapy of using D-BMAA to inhibit the toxicity of L-BMAA.

 The production of BMAA by various species of cyanobacteria and dinoflagellates in laboratory and natural conditions has to be confirmed.

 Food screening covering more food types, larger numbers of samples and considering sampling places and seasonal variation can be performed in order to more accurately estimate the human exposure of BMAA through food consumption.

 Evaluation of the different food preparations and cooking strategies on the effect of removing or degrading BMAA can be investigated in order to calculate the human BMAA intake in daily life.

 The measurement of the presence of BMAA in water bodies, especially those used as drinking water sources, before, during and after algal blooms, i.e. the period of algae cell lysis, has to be performed.

 Laboratory studies on the efficiency of removing BMAA by different drinking water treatments. And the measurement of the BMAA concentration in tap water.

Acknowledgements It would not be possible to finish my Ph. D study and this thesis without the great support and contribution from a lot of people I met during this five-year study period. I might not be able to mention all of you, but you all own my gratitude. Here I would like to especially acknowledge some of you.

I would like to thank the Department of Analytical Chemistry (the Department of Environmental Science and Analytical Chemistry since 2015) to provide the fantastic education for my M. Sc. and Ph. D studies in Analytical Chemistry. It is an amazing scientific field and I will never regret choosing this discipline.

I would like to extend my deepest gratitude to my main supervisor Assoc. Prof. Leopold L. Ilag. Thank you for taking me as a Ph. D student and guiding me on the path of science. Your great passion for science always encourages me to focus on the research. Your intelligent scientific ideas always give me inspiration. I gratefully appreciate your endless patience, being always available for discussion and help, and giving prompt feedback even when you are busy. I also appreciate that you always trust me and give me freedom to try different tasks. You are a supervisor who doesn’t just let students work in the lab, but also help them to grow up as good scientists. I feel I am so lucky to have you as my first guide on the path of science. I have learnt so much from you on both science and diplomacy, as a researcher and a person.

I would also like to extend my deepest gratitude to my assistant supervisors, Prof. Ulrika Nilsson and Assoc. Prof. Gunnar Thorsén. Ulrika, you are the most popular teacher in our courses. (That is not only my own opinion.) When I encountered very challenging questions in mass spectrometry, you gave me very helpful discussions and expert suggestions. You always encourage me to learn new knowledge about mass spectrometry and learn instrument from solving problems. Your infinite curiosity about new analytical techniques and your passion for research gave me a lot of positive influence during my study. As the leader of our unit of the new Dept., you have done a great job! Gunnar, you are an icon of a good teacher because of your wide, but also deep knowledge in chemistry. And most improtantly, you are always willing to share your knowledge with others. I still remember clearly that one day in the lab, you explained for me the difference between instrument sensitivity and limits of detection, using an example of feeling pain on a finger. How lucid it was! I also respect you as one of the persons who is not afraid to show different opinions about things. It is my great pleasure to have both of you as teachers and supervisors!

I would like to send my sincerest thanks to Prof. Anders Colmsjö. Thank you for giving me the opportunity to take the M. Sc. and Ph. D studies in the department. As the boss of the Dept. of Analytical Chemistry, you try to create a pleasant and international working envorinment. As the mentor of my Ph.D study, you always try to follow up, guide and help solve different problems in my five-year study. I appreciate your support during the authorship conflict between us and our collaborators, and your tolerance for my sharpness sometimes . It was a great memory of our trip to China, especially the big table . I have to learn your diligence for learning Chinese to learn Swedish.

I would like to thank Assoc. Prof. Ingrid Granelli. It was a very pleasant and valuable experience to work with you in the Bioanalytical Chemistry course. Many thanks for your

34 comments on this thesis. Your insights as a senior scientist in pharmaceutical industry made me think about academic research in a different perspective. I would like to thank Assoc. Prof. Magnus K. Åberg. There is no doubt that you have excellent expertise in chemometrics. You are leading an essential part in the Analytical Chemistry discipline in our department. I appreciate what I have learnt from you during the Chemometrics course and our collaboration during the project. I am also grateful for your suggestions for my postdoc position. I would like to thank Prof. Roger Westerholm and Prof. Conny Östman. I respect your scientific passion in the same field persistently for decades, eventually becoming outstanding in your field. I appreciate what I learnt from you during the courses and my studies. I would like to thank Prof. Mohamed Abdel-Rehim for always being kind and your suggestion for my postdoc position. I would also like to thank Prof. Carlo Crescenzi for both scientific and nonscientific discussions, and delicious Italian desserts.

I would like to thank all my collaborators in my projects. Without you, my Ph.D study would not be diverse and fruitful. Thanks to all internal collaborators in our department: Ulrika Nilsson, Gunnar Thorsén, Magnus K. Åberg and Nadezda Kiselova. Thanks to all external collaborators: Prof. Wim M. De Borggraeve and his Ph.D student Benoit Aigret in KU Leuven in Belgium, Dr. Zdenek Spacil in University of Washington in USA, Dr. Eric Johnston in the Dept. of Organic Chemistry of Stockholm University, Assoc. Prof. Ulla Rasmussen and her coworkers Johan Eriksson, Sara Jonasson, Sandra Lage and Shiva Shams from the Dept. of Ecology, Environment and Plant Sciences in Stockholm University (It was an invaluable experience that makes me aware of the complexity of research in the real world.), Dr. Johan Rosén in National Food Administration in Uppsala (Without your valuable scientific input, our project would not be successful!), Professor Eva B. Brittebo and her coworkers Dr. Oskar Karlsson, Ph.D student Lisa Ersson and master student Tim Malmström in Uppsala University (It was very pleasant to work with all of you, good researchers and scientists, and I wish our collaboration will be continuously successful!), Prof. Paul Alan Cox and his group members Dr. Sandra Anne Banack and Dr. James S. Metcalf in the Institute for Ethnomedicine in Jackson, USA (Paul, without you, the BMAA research field would not be so exciting and successful!), Prof. Roman A. Zubarev and his group members Ph.D student Xie Xu and Dr. Susanna Lundström in Karolinska Institute (Roman, it was my great pleasure to work with you, one of the pioneers in mass spectrometry.), Prof. Armando Cordova and his group members Dr. Pawel Dziedzic and Dr. Gui-Ling Zhao (also as my close friend!) (Armando, thank you to let me join such an exciting project about the origin of life!), Prof. Heinrich Dircksen in the Dept. of Zoology of Stockholm University (Heinrich, thank you for your many long talks about research and your generosity to share daphnia and ultrasonication machine with us!), my host and supervisor during my visit in Hong Kong, Assoc. Prof. Zhong-Ping Yao and his group members Dr. Pui-Kin So, Dr. Bo Zheng, Ph.D student Bin Hu, Dr. Haixing Wang, and Mr. Tommy and HoYi in the Hong Kong Polytechnic University (Dr. Yao, thank you for giving me the chance to join your excellent research in Ambient mass spectrometry! It was invaluable to learn from you during projects and grant application, and also listen to your insights about research. Thanks to your group for their help and friendliness!), and Traditional Chinese Medicine collaborators: Lanzhou Foci pharmaceutical Co. Ltd (特别感谢贾总,李总,孙总,杨所长,黄维和田亚男), Tsinghua University (Prof. Luo and Prof. Wang), China Academy of Chinese Medical Sciences (Prof. Yang and Dr. Xu) and Wilkris & Co. AB (Ms. Christina Chuck and Mr. Jeremy Jia).

I would like to thank all my former and current colleagues in the Analytical Chemistry Dept. Thanks to the former Ph.D students: Gianluca Maddalo; Mohammad Reza Shariatgorji

(Shahram); Zdenek Spacil, also ex-members of Pol’s group, I believe all of you will become outstanding scientists. Helena Hansson; Nahid Amini (Nana); Leila Zamani; Yasar Thewalim; Caroline Bergh; Thuy Tran; Eric Alm; Aljona Saleh; Silvia Masala. Thanks to the current Ph.D students and colleagues: Trifa Mohammad Ahmed, for sharing all the happiness and frustration in our study and life, and great support always; Christoffer Bergvall, for many help and discussions about research and study; Ioannis Sadiktsis, for professional help on computer and thesis writing; Aziza EI Beqqali, for nice conversations and sharing your special tea and healthy sandwich; Hwamni Lim, for a lot of conversations about research and our lives, a great roommate and friend; Ahmed Ramzy, for our daily chair-fighting and being such “good” roommates; Alessandro Quaranta (for nice collaboration in the course and being a nice roommate; Javier Zurita and Nadia Kiselova (for joining the BMAA research. I am sure both of you will be successful; Giovanna Luongo, for bringing Italian coffee although I only smell and also the nice conversations; Petter Olsson, for helpful discussion about lipids and pieces of advice during the preparation of my thesis and dissertation; Erik Tengstrand, for sharing your genius on different things; Farshid Mashayekhy Rad, for sharing the instrument with me so nicely and being always helpful to solve the instrument problems; Francesco Iadaresta (for nice collaborations in the course and being a nice ex-roommate. We always have an extra chair waiting for you; Jonas Fyrestam, for impressive cream cakes; Rozanna Avagyan, Keep on going your good work; Anna Sroka-Bartnicka, for many nice conversations and being always helpful; Tuula Larsson, I am still using your gift during the winter; Jan Holmbäck, for sharing and great help on your microscope machine; Karin Lidén, for being always willing to help and share different things, like the tips for swimming, Swedish, computer and many others! The companionship of all of you during my Ph.D study will be a great memory in my life!

I would like to thank Pol, Magnus, Ninni, Gunnar, Ioannis and Rina Argelia Andersson for your valuable inputs in this thesis! I also would like to thank Petra Agirman for Swedish translation of the Abstract of my thesis.

I would like to thank all diploma students who worked with us in our projects: Mokhles Hafez, Anna Strzelczak, Martin Meine, Tim Forsgren Malmström, Rudolf Andrys and Petra Agirman. I wish all of you a sucessful future in your study and life!

I would like to thank all administration staff: Anita Engqvist, Anne-Marie Hagelroth, Jonas Rutberg and Lena Elfver. It is your good work that makes the department function!

I would like to thank all the foundations I received that support for my travel and research during my Ph.D study: Swedish Society for Medical Research, Stiftelsen Lars Hiertas minne, Henning och Johan Trone-Holsts Stiftelse, Research Student Attachment Programme of Hong Kong, K. A. Wallenbergs stiftelse and L&E Kinanders. To take this opportunity, I would like to thank Sweden and Swedes for trusting in scientific research and being passionate to donate funding for research work. It is your generosity that made my Ph.D study so colorful! I appreciate it very much!!

Finally, my special thanks to my parents, my brother and his family! 爸妈，谢谢你们毫无保留的无私的爱；谢谢你们总是给我自由，支持我做任何我想做的事情；谢谢你们对我工作，学习和生活一贯的支持和关心！哥嫂，谢谢你们在爸妈身边照顾他们, 让我可以心无旁骛的学习和工作！雪雪，你是我们家新的骄傲！加油！

My deepest thanks to my husband, Qunfeng Jiang (Lake), my lovely girl Tangtang and my coming angel! Lake, thank you for always being there when I needed you! Tangtang, thank you for understanding me when I have to work during your holidays, and thank you for being my sunshine! My coming angel, I will welcome you to this exciting world as a Dr. Mom.

References 1. Kurland, L. T. & Mulder, D. W. Epidemiologic investigations of amyotrophic lateral sclerosis. I. Preliminary report on geographic distribution, with special reference to the Mariana Islands, including clinical and pathologic observations. Neurology 4, 355-378 (1954).

2. Plato CC, Galasko D, Garruto RM, Plato M, Gamst A, Craig UK, Torres JM, Wiederholt W. ALS and PDC of Guam: forty-year follow-up. Neurology 58, 765-773 (2002).

3. Poorkaj P, Tsuang D, Wijsman E, Steinbart E, Garruto RM, Craig U, Chapman NH, Anderson L, Bird TD, Plato CC. TAU as a susceptibility gene for amyotropic lateral sclerosis–parkinsonism dementia complex of Guam. Arch. Neurol. 58, 1871-1878 (2001). 4. Reed, D., Labarthe, D., Chen, K. M. & Stallones, R. A cohort study of amyotrophic lateral sclerosis and parkinsonism-dementia on Guam and Rota. Am. J. Epidemiol. 125, 92-100 (1987).

5. Whiting, M. G. Toxicity of cycads. Econ. Bot. 17, 270-302 (1963). 6. Vega, A. & Bell, E. A. alpha -Amino-beta -methylaminopropionic acid, a new amino acid from seeds of Cycas circinalis. Phytochemistry 6, 759-762 (1967). 7. Vega, A., Bell, E. & Nunn, P. The preparation of l-and d-α-amino-β-methylaminopropionic acids and the identification of the compound isolated from Cycas circinalis as the l-isomer. Phytochemistry 7, 1885-1887 (1968).

8. Murch, S. J., Cox, P. A. & Banack, S. A. A mechanism for slow release of biomagnified cyanobacterial neurotoxins and neurodegenerative disease in Guam. Proc. Natl. Acad. Sci. USA 101, 12228-12231 (2004). 9. Cox, P. A. & Sacks, O. W. Cycad neurotoxins, consumption of flying foxes, and ALS-PDC disease in Guam. Neurology 58, 956-959 (2002). 10. Wiles, G. J. & Payne, N. H. The trade in fruit bats Pteropus spp. on Guam and other Pacific islands. Biol. Conserv. 38, 143-161 (1986). 11. Murch, S. J., Cox, P. A., Banack, S. A., Steele, J. C. & Sacks, O. W. Occurrence of β- methylamino-L-alanine (BMAA) in ALS/PDC patients from Guam. Acta Neurol. Scand. 110, 267-269 (2004).

12. Pablo J, Banack SA, Cox PA, Johnson TE, Papapetropoulos S, Bradley WG, Buck A, Mash DC. Cyanobacterial neurotoxin BMAA in ALS and Alzheimer's disease. Acta Neurol. Scand. 120, 216-225 (2009). 13. Montine, T. J., Li, K., Perl, D. P. & Galasko, D. Lack of beta-methylamino-l-alanine in brain from controls, AD, or Chamorros with PDC. Neurology 65, 768-769 (2005). 14. Cox PA, Banack SA, Murch SJ, Rasmussen U, Tien G, Bidigare RR, Metcalf JS, Morrison LF, Codd GA, Bergman B. Diverse taxa of cyanobacteria produce beta -N-

38 methylamino-L-alanine, a neurotoxic amino acid. Proc. Natl. Acad. Sci. USA 102, 5074-5078 (2005). 15. Duncan MW, Villacreses NE, Pearson PG, Wyatt L, Rapoport SI, Kopin IJ, Markey SP, Smith QR. 2-amino-3-(methylamino)-propanoic acid (BMAA) pharmacokinetics and blood- brain barrier permeability in the rat. J. Pharmacol. Exp. Ther. 258, 27-35 (1991). 16. Duncan MW, Markey SP, Weick BG, Pearson PG, Ziffer H, Hu Y, Kopin IJ. 2-Amino-3- (methylamino) propanoic acid (BMAA) bioavailability in the primate. Neurobiol. Aging 13, 333-337 (1992). 17. Xie, X., Basile, M. & Mash, D. C. Cerebral uptake and protein incorporation of cyanobacterial toxin β-N-methylamino-L-alanine. NeuroReport 24, 779-784 (2013). 18. Smith, Q. R., Nagura, H., Takada, Y. & Duncan, M. W. Facilitated Transport of the Neurotoxin, β –N-Methylamino‐l‐Alanine, Across the Blood‐Brain Barrier. J. Neurochem. 58, 1330-1337 (1992). 19. Karlsson, O., Jiang, L., Ilag, L. L., Andersson, M. & Brittebo, E. B. Protein association of the neurotoxin and non-protein amino acid BMAA (β-N-methylamino-l-alanine) in the liver and brain following neonatal administration in rats. Toxicol Lett (2014).

20. Polsky, F. I., Nunn, P. B. & Bell, E. A. Distribution and toxicity of β-amino-N- methylaminopropionic acid. Fed. Proc. , Fed. Amer. Soc. Exp. Biol. 31, 1473-1475 (1972). 21. Dawson Jr R, Marschall E, Chan K, Millard W, Eppler B, Patterson T. Neurochemical and Neurobehavioral Effects of Neonatal Administration of β-N-Methylamino-l-Alanine and 3, 3-Iminodipropionitrile. Neurotoxicol. Teratol. 20, 181-192 (1998). 22. Spencer PS, Nunn PB, Hugon J, Ludolph AC, Ross SM, Roy DN, Robertson RC. Guam amyotrophic lateral sclerosis-parkinsonism-dementia linked to a plant excitant neurotoxin. Science 237, 517-522 (1987). 23. Perry, T. L., Bergeron, C., Biro, A. J. & Hansen, S. β-N-Methylamino-L-alanine: chronic oral administration is not neurotoxic to mice. J. Neurol. Sci. 94, 173-180 (1989). 24. Cruz-Aguado, R., Winkler, D. & Shaw, C. A. Lack of behavioral and neuropathological effects of dietary β-methylamino-L-alanine (BMAA) in mice. Pharmacol. , Biochem. Behav. 84, 294-299 (2006). 25. Brownson, D. M., Mabry, T. J. & Leslie, S. W. The cycad neurotoxic amino acid, ß-N- methylamino-l-alanine (BMAA), elevates intracellular calcium levels in dissociated rat brain cells. J. Ethnopharmacol. 82, 159-167 (2002). 26. Lobner, D., Piana, P. M. T., Salous, A. K. & Peoples, R. W. β-N-methylamino-l-alanine enhances neurotoxicity through multiple mechanisms. Neurobiol. Dis. 25, 360-366 (2007).

27. Cox, P., Banack, S. & Murch, S. Biomagnification of cyanobacterial neurotoxins and neurodegenerative disease among the Chamorro people of Guam. Proc. Natl. Acad. Sci. USA 100, 13380-13383 (2003). 28. Metcalf JS, Banack SA, Lindsay J, Morrison LF, Cox PA, Codd GA. Co-occurrence of beta -N-methylamino-L-alanine, a neurotoxic amino acid with other cyanobacterial toxins in British waterbodies, 1990-2004. Environ. Microbiol. 10, 702-708 (2008).

29. Esterhuizen, M. & Downing, T. beta -N-methylamino-l-alanine (BMAA) in novel South African cyanobacterial isolates. Ecotoxicol. Environ. Saf. 71, 309-313 (2008). 30. Faassen, E. J., Gillissen, F., Zweers, H. A. J. & Luerling, M. Determination of the neurotoxins BMAA (beta -N-methylamino-L-alanine) and DAB (alpha -,gamma - diaminobutyric acid) by LC-MSMS in Dutch urban waters with cyanobacterial blooms. Amyotrophic Lateral Scler. 10, 79-84 (2009). 31. Jonasson S, Eriksson J, Berntzon L, Spacil Z, Ilag LL, Ronnevi L, Rasmussen U, Bergman B. Transfer of a cyanobacterial neurotoxin within a temperate aquatic ecosystem suggests pathways for human exposure. Proc. Natl. Acad. Sci. USA 107, 9252-9257 (2010). 32. Jiao Y, Chen Q, Chen X, Wang X, Liao X, Jiang L, Wu J, Yang L. Occurrence and transfer of a cyanobacterial neurotoxin β-methylamino-L-alanine within the aquatic food webs of Gonghu Bay (Lake Taihu, China) to evaluate the potential human health risk. Sci. Total Environ. 468-469, 457-463 (2014).

33. Rosen, J. & Hellenaes, K. Determination of the neurotoxin BMAA (β-N-methylamino-l- alanine) in cycad seed and cyanobacteria by LC-MS/MS (liquid chromatography tandem mass spectrometry). Analyst (Cambridge, U. K.) 133, 1785-1789 (2008). 34. Kruger, T., Monch, B., Oppenhauser, S. & Luckas, B. LC-MS/MS determination of the isomeric neurotoxins BMAA (beta-N-methylamino-L-alanine) and DAB (2,4-diaminobutyric acid) in cyanobacteria and seeds of Cycas revoluta and Lathyrus latifolius. Toxicon 55, 547- 557 (2010). 35. Faassen, E. J., Gillissen, F. & Luerling, M. A comparative study on three analytical methods for the determination of the neurotoxin BMAA in cyanobacteria. PLoS One 7, e36667 (2012).

36. McCarron, P., Logan, A. C., Giddings, S. D. & Quilliam, M. A. Analysis of beta-N- methylamino-L-alanine (BMAA) in spirulina-containing supplements by liquid chromatography-tandem mass spectrometry. Aquat. Biosyst 10, 5-9063-10-5. eCollection 2014 (2014).

37. Jiang L, Mehine M, Ilag LL, Eriksson J, Lage S, Jonasson S, Rasmussen U, Shams S. Diatoms: A Novel Source for the Neurotoxin BMAA in Aquatic Environments. PLoS One 9, e84578 (2014). 38. Jiang L. and Ilag L.L. Detection of endogenous BMAA in dinoflagellate (Heterocapsa triquetra) hints at evolutionary conservation and environmental concern. PubRaw Science 1 (2): 1-8 (2014).

39. Lage S, Costa PR, Moita T, Eriksson J, Rasmussen U, Rydberg SJ. BMAA in shellfish from two Portuguese transitional water bodies suggests the marine dinoflagellate Gymnodinium catenatum as a potential BMAA source. Aquatic Toxicology 152, 131-138 (2014).

40. Downing, S., Banack, S., Metcalf, J., Cox, P. & Downing, T. Nitrogen starvation of cyanobacteria results in the production of β-N-methylamino-L-alanine. Toxicon 58, 187-194 (2011).

41. Downing, S., van de Venter, M. & Downing, T. G. The effect of exogenous β-N- methylamino-L-alanine on the growth of Synechocystis PCC6803. Microb. Ecol. 63, 149-156 (2012). 42. Berntzon L, Erasmie S, Celepli N, Eriksson J, Rasmussen U, Bergman B. BMAA inhibits nitrogen fixation in the cyanobacterium Nostoc sp. PCC 7120. Marine drugs 11, 3091-3108 (2013). 43. Banack SA, Metcalf JS, Jiang L, Craighead D, Ilag LL, Cox PA. Cyanobacteria produce N-(2-aminoethyl)glycine, a backbone for Peptide nucleic acids which may have been the first genetic molecules for life on Earth. PLoS One 7, e49043 (2012). 44. Gordon, N., Angel, D., Neori, A., Kress, N. & Kimor, B. Heterotrophic dinoflagellates with symbiotic Cyanobacteria and nitrogen limitation in the Gulf-of-Aqaba. Mar. Ecol. Prog. Ser. 107, 83-88 (1994). 45. Wasmund, N. & Uhlig, S. Phytoplankton trends in the Baltic Sea. ICES Journal of Marine Science: Journal du Conseil 60, 177-186 (2003). 46. Bidle, K. D., Manganelli, M. & Azam, F. Regulation of oceanic silicon and carbon preservation by temperature control on bacteria. Science 298, 1980-1984 (2002). 47. Kisby, G. E., Ellison, M. & Spencer, P. S. Content of the neurotoxins cycasin (methylazoxymethanol beta-D-glucoside) and BMAA (beta-N-methylamino-L-alanine) in cycad flour prepared by Guam Chamorros. Neurology 42, 1336-1340 (1992).

48. Banack, S. A. & Cox, P. A. Distribution of the neurotoxic nonprotein amino acid BMAA in Cycas micronesica. Bot. J. Linn. Soc. 143, 165-168 (2003). 49. Banack, S. A., Murch, S. J. & Cox, P. A. Neurotoxic flying foxes as dietary items for the Chamorro people, Marianas Islands. J. Ethnopharmacol. 106, 97-104 (2006). 50. Duncan, M. W., Kopin, I. J., Garruto, R. M., Lavine, L. & Markey, S. P. 2-Amino-3 (methylamino)-propionic acid in cycad-derived foods is an unlikely cause of amyotrophic lateral sclerosis/parkinsonism. Lancet 2, 631-632 (1988). 51. Johnson HE, King SR, Banack SA, Webster C, Callanaupa WJ, Cox PA. Cyanobacteria (Nostoc commune) used as a dietary item in the Peruvian highlands produce the neurotoxic amino acid BMAA. J. Ethnopharmacol. 118, 159-165 (2008).

52. Roney BR, Li R, Banack SA, Murch S, Honegger R, Cox PA. Consumption of fa cai Nostoc soup: A Potential for BMAA exposure from Nostoc cyanobacteria in China? Amyotrophic Lateral Scler. 10, 44-49 (2009). 53. Brand, L. E., Pablo, J., Compton, A., Hammerschlag, N. & Mash, D. C. Cyanobacterial blooms and the occurrence of the neurotoxin, beta-N-methylamino-l-alanine (BMAA), in South Florida aquatic food webs. Harmful Algae 9, 620-635 (2010).

54. Liu Y, Chen W, Li D, Huang Z, Shen Y, Liu Y. Cyanobacteria-/cyanotoxin- contaminations and eutrophication status before Wuxi drinking water crisis in Lake Taihu, China. J Environ Sci (China) 23, 575-581 (2011).

55. Masseret E, Banack S, Boumédiène F, Abadie E, Brient L, Pernet F, Juntas-Morales R, Pageot N, Metcalf J, Cox P. Dietary BMAA Exposure in an Amyotrophic Lateral Sclerosis Cluster from Southern France. PloS one 8, e83406 (2013). 56. Field NC, Metcalf JS, Caller TA, Banack SA, Cox PA, Stommel EW. Linking β- methylamino-l-alanine exposure to sporadic amyotrophic lateral sclerosis in Annapolis, MD. Toxicon 70, 179-183 (2013). 57. Mondo K, Broc Glover W, Murch SJ, Liu G, Cai Y, Davis DA, Mash DC. Environmental neurotoxins β-N-methylamino-l-alanine (BMAA) and mercury in shark cartilage dietary supplements. Food and Chemical Toxicology 70, 26-32 (2014). 58. Duncan, M. W., Kopin, I. J., Crowley, J. S., Jones, S. M. & Markey, S. P. Quantification of the putative neurotoxin 2-amino-3-(methylamino) propanoic acid (BMAA) in cycadales: analysis of the seeds of some members of the family Cycadaceae. J. Anal. Toxicol. 13, 169- 175 (1989).

59. Nunn, P. B. & O'Brien, P. The interaction of β-N-methylamino-L-alanine with bicarbonate: a sup H-NMR study. FEBS Lett. 251, 31-35 (1989). 60. Jiang, L., Aigret, B., De Borggraeve, W. M., Spacil, Z. & Ilag, L. L. Selective LC-MS/MS method for the identification of BMAA from its isomers in biological samples. Anal. Bioanal. Chem. 403, 1719-1730 (2012). 61. Duncan, M. W., Steele, J. C., Kopin, I. J. & Markey, S. P. 2-Amino-3-(methylamino)- propanoic acid (BMAA) in cycad flour: an unlikely cause of amyotrophic lateral sclerosis and parkinsonism-dementia of Guam. Neurology 40, 767-772 (1990). 62. Kisby, G. E., Roy, D. N. & Spencer, P. S. Determination of β-N-methylamino-L-alanine (BMAA) in plant (Cycas circinalis L.) and animal tissue by precolumn derivatization with 9- fluorenylmethyl chloroformate (FMOC) and reversed-phase high-performance liquid chromatography. J. Neurosci. Methods 26, 45-54 (1988).

63. Banack, S. A. & Cox, P. A. Biomagnification of cycad neurotoxins in flying foxes. Neurology 61, 387-389 (2003). 64. Cervantes Cianca, R., Baptista, M., da Silva, L. P., Lopes, V. & Vasconcelos, V. Reversed-phase HPLC/FD method for the quantitative analysis of the neurotoxin BMAA (β- N-methylamino-l-alanine) in cyanobacteria. Toxicon 59, 379-384 (2012). 65. Cianca, R. C. C., Baptista, M. S., Lopes, V. R. & Vasconcelos, V. M. The non-protein amino acid β-N-methylamino-l-alanine in Portuguese cyanobacterial isolates. Amino Acids 42, 2473-2479 (2012). 66. Spacil Z, Eriksson J, Jonasson S, Rasmussen U, Ilag LL, Bergman B. Analytical protocol for identification of BMAA and DAB in biological samples. Analyst 135, 127-132 (2010).

67. Jiang, L., Johnston, E., Aaberg, K. M., Nilsson, U. & Ilag, L. L. Strategy for quantifying trace levels of BMAA in cyanobacteria by LC/MS/MS. Anal. Bioanal. Chem. 405, 1283-1292 (2013). 68. Jiang, L., Kiselova, N., Rosén, J. & Ilag, L. L. Quantification of neurotoxin BMAA ([bgr]-N-methylamino-L-alanine) in seafood from Swedish markets. Scientific reports 4 (2014).

69. Banack, S. A., Metcalf, J. S., Bradley, W. G. & Cox, P. A. Detection of cyanobacterial neurotoxin β-N-methylamino-l-alanine within shellfish in the diet of an ALS patient in Florida. Toxicon 90, 167-173 (2014). 70. Salomonsson, M. L., Hansson, A. & Bondesson, U. Development and in-house validation of a method for quantification of BMAA in mussels using dansyl chloride derivatization and ultra-performance liquid chromatography tandem mass spectrometry. Analytical Methods 5, 4865-4874 (2013).