Detection, transfer and role of an environmentally spread neurotoxin (BMAA) with focus on cyanobacteria and the Baltic Sea region

(1)

D e t e c t i o n , t r a n s f e r a n d r o l e o f a n e n v i r o n m e n t a l l y s p r e a d n e u r o t o x i n ( B M A A ) w i t h f o c u s o n c y a n o b a c t e r i a a n d t h e B a l t i c S e a r e g i o n

(2)

(3)

Detection, transfer and role of an

environmentally spread neurotoxin

(BMAA) with focus on cyanobacteria and

the Baltic Sea region

(4)

©Lotta Berntzon, Stockholm University 2015

Cover image: Baltic Sea cyanobacteria collected and photographed outside of Askö Laboratory, 8 August 2013. Photograph taken by Lotta Berntzon.

ISBN 879-91-7649-142-3 Printed: Holmbergs, Malmö 2015

Distributor: Department of Ecology, Environment and Plant Sciences, Stockholm University, Stockholm, Sweden

(5)

(6)

Abstract

β-N-methylamino-L-alanine (BMAA) is one of the more recently discovered bioactive compounds produced by cyanobacteria. BMAA is a non-protein amino acid reported present in human brain tissues of patients deceased from a neurodegenerative disease, such as Alzheimer´s disease or amyotrophic lateral sclerosis (ALS). This observation in combination with its neurotoxic effects in eukaryotes (in vivo and in vitro) and its potential to incorporate into (human) proteins, causing protein aggregation, suggests BMAA as a possible causative environmental agent for neurodegenerative diseases. Due to the ubiquitous nature of cyanobacteria with a wide occurrence in both aquatic and terrestrial environments, BMAA could be globally spread. Hence, investigations of a possible coupling between BMAA and neurodegeneration are urgently needed as well as to identify sources of BMAA in Nature.

The aim of this thesis was to examine the potential occurrence of BMAA in bloom forming cyanobacteria of the Baltic Sea and its possible transfer to other organisms of this ecosystem. Of importance was also to reveal any likely routes for human BMAA exposure in the Baltic Sea region and to further investigate BMAA as a triggering agent for neurodegenerative diseases. Acknowledged difficulties of analysing BMAA in biological samples also inferred method development as part of the experimental studies. Investigating the role of BMAA in its producers was another purpose of the thesis, which may be crucial for future management of BMAA-producing cyanobacteria.

By screening natural populations of the major filamentous bloom forming cyanobacteria of the Baltic Sea, we discovered the presence of BMAA throughout the entire summer season of two consecutive years, using a highly specific analytical method (liquid chromatography-tandem mass spectrometry; LC-MS/MS). BMAA was found to bioaccumulate in zooplankton and fish, as well as in mussels and oysters from the Swedish west coast. To improve the understanding of BMAA analyses in natural samples, the formation of carbamate adducts in the presence of bicarbonate was examined. Using two derivatization techniques in combination with LC-MS/MS, we could show that BMAA detection was not hindered by carbamate formation. Exogenously added BMAA inhibited nitrogen fixation in the model cyanobacterium Nostoc sp. PCC 7120, which was also hampered in growth and displayed signs of nitrogen starvation. Finally, BMAA was detected in cerebrospinal fluid in three of 25 Swedish test individuals, and represents the first confirmation of BMAA in the human central nervous system using LC-MS/MS as the primary analytical method. However, the association of BMAA to neurodegenerative

(7)

diseases could not be verified as BMAA was present in both control individuals (two) and in one ALS-patient. Nevertheless, the finding of a known neurotoxic compound in the human central nervous system is alarming and potential consequences should be investigated.

The discovery of the neurotoxic compound BMAA in Baltic Sea organisms, and in the central nervous system of humans potentially consuming fish from this ecosystem is concerning and warrants continued investigations of BMAA occurrence and human exposure. Further knowledge on the function and regulation of BMAA may help in developing strategies aiming to minimise human exposure.

(8)

Sammanfattning

Cyanobakterier producerar många bioaktiva ämnen, varav β-N-metylamino-L-alanin (BMAA) är ett av de mer nyligen upptäckta. BMAA är en ickeprotein-aminosyra som har återfunnits i hjärnvävnad hos människor som avlidit på grund av en neurodegenerativ sjukdom, så som Alzheimers sjukdom eller amyotrofisk lateral skleros (ALS). Att BMAA därtill uppvisar neurotoxiska effekter hos eukaryoter (in

vivo och in vitro), och potentiellt kan infogas i (mänskliga) proteiner och därmed

orsaka proteinaggregering, talar för att BMAA skulle kunna vara en potentiell miljöfaktor inblandad i uppkomsten av neurodegenerativa sjukdomar. Då cyanobakterier är vanligt förekommande i de flesta akvatiska och terrestra miljöer är BMAA troligtvis globalt spritt. Det är därför av vikt att undersöka den möjliga kopplingen mellan BMAA och neurodegeneration, och att även identifiera BMAA-källor i naturen.

Syftet med denna avhandling var att undersöka förekomsten av BMAA i cyanobakterier i Östersjön och om BMAA kan överföras till andra organismer i detta ekosystem. Av intresse var även att utröna ifall BMAA på något vis kan överföras till människor i Östersjöregionen och att vidare undersöka om BMAA kan vara associerat till neurodegenerativa sjukdomar. Kunskap om BMAA:s funktion i cyanobakterier eftersöktes därtill, vilken bland annat kan vara till hjälp för framtida hantering av dessa BMAA-producerande organismer. Metodutveckling var också en given del av det experimentella arbetet, då BMAA är erkänt svåranalyserat i biologiska prover.

Specifika analyser (vätskekromatografi-tandem masspektrometri; LC-MS/MS) av naturliga populationer av de filamentösa cyanobakterier som dominerar blommningarna i Östersjön, visade på närvaro av BMAA hela sommaren under två på varandra följande år. Vi upptäckte att BMAA bioackumulerades i zooplankton och fisk (Östersjön) och i musslor och ostron från den svenska västkusten. Tillsatt BMAA verkade starkt inhiberande på kvävefixeringen hos modellcyanobakterien

Nostoc sp. PCC 7120, som även uppvisade tecken på kvävesvält och avstannad

tillväxt. För att öka kunskapen om BMAA-analyser i naturliga prover undersöktes bildandet av BMAA-karbamater som sker i närvaro av bikarbonat. Genom att använda två olika derivatiseringsmetoder i kombination med LC-MS/MS, kunde vi visa att BMAA-detektionen inte påverkades negativt av karbamatbildningen. Slutligen påvisades BMAA i cerebrospinalvätska hos tre av 25 svenska testpersoner, vilket är den första bekräftelsen på BMAA:s närvaro i det mänskliga centrala nervsystemet som gjorts med LC-MS/MS som primär analysmetod. Någon koppling mellan BMAA och neurodegeneration kunde dock inte göras då BMAA

(9)

detekterades i både kontrollindivider (två stycken) och i en ALS-patient. Trots detta är närvaron av ett känt neurotoxiskt ämne i det centrala nervsystemet hos människa oroande och eventuella konsekvenser bör undersökas vidare.

Upptäckten av BMAA i organismer från Östersjön och i det centrala nervsystemet hos människor som möjligtvis konsumerar fisk från detta ekosystem är oroande och yrkar på fortsatta undersökningar av BMAA:s förekomst och den mänskliga exponeringen. Utökad kunskap om BMAA:s funktion och dess reglering hos cyanobakterier kan därtill underlätta utvecklandet av strategier för att minimera människors exponering för BMAA.

(10)

List of Papers

This thesis is based on the following publications and manuscript, referred to in roman numerals.

I. Jonasson S, Eriksson J, Berntzon L, Spáčil Z, Ilag LL, Ronnevi L-O, Rasmussen U, Bergman, B (2010) Transfer of a cyanobacterial neurotoxin within a temperate aquatic ecosystem suggests pathways for human exposure. Proc Natl Acad Sci USA 107:9252-9257.

II. Berntzon L, Erasmie S, Celepli N, Eriksson J, Rasmussen U, Bergman

B (2013) BMAA inhibits nitrogen fixation in the cyanobacterium

Nostoc sp. PCC 7120. Mar Drugs 11:3091-3108.

III. Berntzon L, Bergman B, Eriksson J (2015) Effects of bicarbonate on

LC-MS/MS analysis of BMAA using AQC or EZ:Faast™ pre-column derivatization. (manuscript)

IV. Berntzon L, Ronnevi L-O, Bergman B, Eriksson J (2015) Detection of

BMAA in the human central nervous system. Neuroscience 292:137-147.

My contributions to the papers were:

Paper I: Collaborated in field sampling, extraction and sample preparation of cyanobacteria and zooplankton. Commented on the manuscript.

Paper II: Participated in planning, performed the major part of the experiments and data analysis. Main writer of the paper.

Paper III: Participated in planning, performed experiments and analysis of data. Main writer of the paper.

Paper IV: Participated in planning, performed the major part of the experiments and analysis of data. Main writer of the paper.

(11)

Reprints of Paper I, II and IV were made with permission from the respective publisher.

Additional paper:

Jonasson S, Eriksson J, Berntzon L, Rasmussen U, Bergman, B (2008) A novel cyanobacterial toxin (BMAA) with potential neurodegenerative effects. Plant Biotechnol 25:227-232. (review)

(12)

Abbreviations

ADP adenosine diphosphate

AEG N-(2-aminoethyl) glycine

ALS amyotrophic lateral sclerosis

ALS-PDC amyotrophic lateral sclerosis-parkinsonism dementia complex

AMPA α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid

ATP adenosine triphosphate

AQC 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate

BAMA β-amino-N-methylalanine

BG-11 blue green medium

BMAA β-N-methylamino-L-alanine

BMAA-α-NCO2 α-N-carboxy-β-N-methylamino-L-alanine

BMAA-β-NCO2 β-(N-carboxy-N-methyl)-amino-L-alanine

BOAA β-oxalylaminopropionic acid

C carbon

°C degree celsius

14

C carbon-14 isotope

Ca2+ divalent calcium ion

CID collision induced dissociation

CNS central nervous system

CO2 carbon dioxide

CSF cerebrospinal fluid

Cu2+ divalent copper ion

D3 deuterated

DAB 2-4-diaminobutyric acid

DDT dichlorodiphenyltrichloroethane,

DW dry weight

e- electron

EAATs excitatory amino acid transporters

ER endoplasmic reticulum

ESI electrospray ionisation

Fe iron

FNR ferredoxin:NADP+ oxidoreductase

GC gas chromatography

GC-ARA gas chromatography-acetylene reduction assay

Gln glutamine

(15)

GluR0/SGluR0 cyanobacterial glutamate receptor GR glutathione reductase h hour H+ hydrogen ion H2 dihydrogen HCl hydrochloric acid HCO3- bicarbonate

HILIC hydrophilic interaction liquid chromatography

H2O2 hydrogen peroxide

HPLC high performance liquid chromatography

HPLC-FLD high performance liquid chromatography-fluorescent detection

ICBN the International Code of Botanical Nomenclature

ICNP the International Code of Nomenclature of Bacteria

K+ potassium ion

LC liquid chromatography

LC-FLD liquid chromatography-fluorescent detection

LC-MS liquid chromatography-mass spectrometry

LC-MS/MS liquid chromatography-tandem mass spectrometry

LOD limit of detection

LOQ limit of quantification

LPS lipopolysaccharide

M Molar (mol/L)

mg milligram

mGluRs metabotropic glutamate receptors

mM millimolar

Mn manganese

mRNA messenger ribonucleic acid

MRM multiple reaction monitoring

MS mass spectrometry

MS/MS tandem mass spectrometry

m/z mass-to-charge ratio N nitrogen 13 N nitrogen-13 isotope 15 N nitrogen-15 isotope N2 dinitrogen

Na+ divalent sodium ion

NAPD+ _{reduced nicotinamide adenine dinucleotide phosphate}

NH3 ammonia

NH4+ ammonium

Ni2+ divalent nickel ion

NMDA N-methyl-D-aspartate

NMR nuclear magnetic resonance

(16)

O2 oxygen

O2- singlet oxygen

P phosphorus

Pi inorganic phosphorus

PCBs polychlorinated biphenyls

PCC Pasteur Culture Collection of Cyanobacteria

PDC programmed cell death

psu practical salinity units

Q1 first quadrupole

Q2 second quadrupole/collision cell

Q3 third quadrupole

Rbr rubrerythrin

ROS reactive oxygen species

RRHT Rapid Resolution High Throughput

RT retention time

S sulphur

SIMS secondary ion mass spectrometry

S/N signal-to-noise ratio

SOD superoxide dismutase

sp. species (singular)

SPE solid phase extraction

spp. species (plural)

SRM selective reaction monitoring

TCA trichloroacetic acid

TEM transmission electron microscopy

TFA trifluoroacetic acid

TIC total ion count

µ micro

µg/g microgram per gram

UHPLC ultra high performance liquid chromatography

UPLC ultra performance liquid chromatography

Xc- cystine/glutamate antiporter system

(17)

17

Introduction

The discovery of BMAA – a historical overview

BMAA (β-N-methylamino-L-alanine) is a non-protein amino acid first isolated from the seeds of the gymnosperm Cycas circinalis (now Cycas micronesica Hill) (Vega and Bell, 1967), a palm-like tree widespread on the tropical island of Guam (part of the Marianas islands) in the Pacific Ocean. The cycad seeds were initially analysed in the search for neurotoxic compounds that could cause the high incidence of the neurodegenerative disorder amyotrophic lateral sclerosis-Parkinsonism dementia complex (ALS-PDC) on the island. In the mid-20th century, the incidence rate of this disease, also known as Lytico-Bodig, was found to be 50-100 times higher amongst the native population (Chamorros) of Guam, compared to the occurrence of neurodegenerative diseases, such as amyotrophic lateral sclerosis (ALS), Alzheimer´s and Parkinson´s disease, in the rest of the world (Zimmerman, 1945; Kurland and Mulder, 1954). The high incidence rates on Guam aroused tremendous interest among neurologist trying to finally find a causative agent of human neurodegeneration. For instance, the anthropologist Dr. Marjorie Whiting found that cycad seeds, known to be toxic, were used to produce tortilla flour among the Chamorros (Whiting, 1964; Whiting et al., 1966). Inspired by the finding of a neurotoxic amino acid, β-oxalylaminopropionic acid (BOAA), in Lathyrus latifolius seeds causing neurolathyrism (Murti et al., 1964; Rao et al., 1964), a search for BOAA also in the cycad seeds was initiated. However, instead of BOAA, Vega and Bell (1967) discovered the previously unknown amino acid BMAA in the cycad seeds.

Subsequent animal studies showed that BMAA possessed neurotoxic properties (Vega et al., 1968; Polsky et al., 1972; Spencer et al., 1986; 1987) which suggested that BMAA could be involved in eliciting ALS-PDC. These animal studies were however criticised (Duncan et al., 1988) as the Chamorros, being aware of cycad toxicity, washed the seeds carefully before use, lowering the BMAA concentrations to almost undetectable levels. In addition, Garruto and colleagues (1988) postulated that if the sensitivity of humans equalled that of primates, about 1500 kg unwashed cycad seeds would need to be consumed to reach similar concentrations as that administrated during 12 weeks to the macaques by Spencer and colleagues (1987). After these findings, interest in BMAA as a causative agent for ALS-PDC or other neurodegenerative diseases waned.

(18)

18

Rejuvenation of the BMAA-hypothesis

Interest in BMAA was renewed at the beginning of the 21th century due to several studies published by the ethnobotanist Dr. Paul Cox and co-workers. They noticed another food item of traditional importance for the Chamorros, namely flying foxes feeding on cycad seeds (Cox and Sacks, 2002; Monson et al., 2003). The Chamorro consumption of flying foxes was considerably elevated after the Japanese occupation of Guam during World War II. The increased access of firearms eased the hunt for flying foxes to the extent that one of the two endemic species (Pteropus

tokudae) became extinct, while the other (P. mariannus mariannus) was listed as an

endangered species (Cox and Sacks, 2002; Monson et al., 2003). Hence, during the 1940s the Chamorros consumed flying foxes in unpreceded quantities, a consumption that declined in parallel with decreasing numbers of flying foxes on Guam (Cox and Sacks, 2002). Cox and Sacks (2002) noticed that the incidence rates of ALS-PDC in the Chamorro population peaked at a period just after the peak in consumption of Guamanian flying foxes (1960s). They could further show a correlation between the decline of the flying fox population and the subsequent decrease in PDC incidence rates. This gave rise to the hypothesis that ALS-PDC was related to the consumption of flying foxes due to the biomagnification of a toxic compound derived from cycad seeds in these animals (Cox and Sacks, 2002). The hypothesis was soon supported by the discovery of high concentrations of BMAA in tissues of flying foxes (museum specimens) (Banack and Cox, 2003a; Banack et al., 2006).

By examining distribution patterns of BMAA in the cycad Cycas micronesica Hill, it was soon discovered that BMAA was present in the specialised ‘coralloid’ roots but not in other cycad roots (Banack and Cox, 2003b). The coralloid roots typically grow apogeotropically and often reach the soil surface. These are roots specialised to harbour symbiotic nitrogen-fixing cyanobacteria of the genus Nostoc which support the cycad plant with its total need of combined nitrogen (Lindblad and Costa, 2002). This opened for the possibility that the prokaryotic cyanobacteria, rather than the plants, were the true producers of BMAA in cycads. This was further substantiated by the finding of BMAA in a cultured Nostoc symbiont isolated from coralloid roots of C. micronesica Hill (Cox et al., 2003). This discovery highly influenced the continued BMAA research of the 21st century. Notably, BMAA was furthermore detected in post mortem brain tissues of six Chamorro ALS-PDC patients as well as in two Alzheimer´s patients living geographically far from Guam and in a temperate ecosystem devoid of cycads in its flora, namely in Canada. In contrast, BMAA was not detected in the 13 tested Canadian ‘control’ post mortem brains (Cox et al., 2003). Biomagnification of BMAA in the Guam ecosystem was also observed as the BMAA concentrations increased at higher trophic levels, from isolated cyanobacteria, to cycad tissues and flying foxes (Cox et al., 2003) (Figure

1, free BMAA). This scenario was postulated to illustrate a likely route for human

exposure to high BMAA concentrations and that BMAA may be the causative agent for human neurodegeneration in general (Cox et al., 2003). Analyses of BMAA in

(19)

19

biological material were up to this point only concerned with the free BMAA (dissolved BMAA in the soluble cell fraction), an approach that was reversed when high BMAA concentrations were found after hydrolysing (6 M HCl) non-soluble protein fractions of the biological material tested, as in axenic Nostoc isolates, cycad seeds and in human brain tissues (Murch et al., 2004a; Murch et al., 2004b) (Figure

1, protein associated BMAA). The finding further substantiated the neurodegenerating role of BMAA in ALS-PDC, as it revealed that the BMAA concentrations might have been seriously underestimated, and the Chamorros exposed to considerably higher doses of BMAA than previously known. In addition, other Chamorro food items such as meat from pigs and deer, also feeding on cycad seeds, were identified (Banack et al., 2006). An investigation of the protein fraction of previously analysed material (free BMAA only) (Banack and Cox, 2003a), verified the pattern of BMAA biomagnification in cycad seeds, flying foxes and humans of the Guam ecosystem but also revealed the presence of even higher BMAA concentrations in some matrices (Murch et al., 2004a; Murch et al., 2004b) (Figure 1). Due to these discoveries it was suggested that BMAA may accumulate in endogenous, protein associated reservoirs within the body (e.g. the brain), and that BMAA is slowly released during protein metabolism (Murch et al., 2004a). Hence, BMAA would act as a ‘slow’ toxin with a long latency period, as early suggested by Spencer and colleagues (1991). In a subsequent study, BMAA was detected in a majority (90%) of the strains in a large sample set of cyanobacteria (Cox et al., 2005), with representatives from all five morphological sections of cyanobacteria (Rippka et al., 1979), including symbiotic and free-living species of both terrestrial and aquatic origin (Cox et al., 2005).

The discovery of BMAA in post mortem brain tissues of Canadian Alzheimer´s patients (Cox et al., 2003; Murch et al., 2004b) together with the finding that BMAA may be produced by a large number of cyanobacteria of worldwide distribution

Figure 1. Bioaccumulation and biomagnification of free and protein-associated BMAA

in the Guam ecosystem (from Murch et al., 2004b, copyright (2004) National Academy of Sciences).

(20)

20

(Cox et al., 2005) implied that the occurrence of BMAA was of global concern. This led to a strong renewed interest in BMAA research, still continuing to this date. Notably however, a number of questions remain and importantly, a direct link between BMAA and human neurodegenerative diseases still remains to be proven.

BMAA – the molecule

BMAA is a non-protein diamino acid, hence not one of the 20 amino acids that represents the building blocks of proteins. There are hundreds of different non-protein amino acids, produced foremost by plants (Bell, 2003; Vranova et al., 2011), but also by microorganisms such as cyanobacteria (Harada, 2004).

Today, β-N-methylamino-L-alanine (abbreviated as BMAA) is the commonly used chemical name, but was earlier referred to as: L

-α-amino-β-methylaminopropionic acid, 2-amino-3-(methylamino)-propanoic acid or β-methyl-α, β-diaminopropionic acid and often abbreviated MeDAP. BMAA is a small hydrophilic molecule (MW: 118.13) with the molecular formula C4H10N2O2 (CAS:

16012-55-8 for L-BMAA hydrochloride) (Figure 2). It has three titratable groups;

the carboxyl group (pKa1 2.1), the primary amino group (pKa2 6.5) and the

secondary amino group (pKa3 9.8) (Nunn and O'Brien, 1989). The pKa values for

BMAA infers that it is positively charged at low (acidic) pH, negatively charged at high (basic) pH and a neutral ion (zwitter ion; the net charge is 0) at pH 7 (Nunn and Ponnusamy, 2009).

Even though BMAA appears to be stable under certain external conditions (Banack and Cox, 2003a; Banack et al., 2006; Lürling et al., 2011), several reports describe the reactive nature of BMAA (Hashmi and Anders, 1991; Nunn and Ponnusamy, 2009; Glover et al., 2012). A most important reaction contributing to the toxicity of BMAA in eukaryotes is the reversible formation of stable BMAA-carbamates in the presence of CO2/HCO3- (Nunn and O'Brien, 1989; Weiss et al.,

1989; Myers and Nelson, 1990). The addition of CO2 to the amino groups of BMAA

creates β-(N-carboxy-N-methyl)-amino-L-alanine (BMAA-β-NCO2), which is

structurally similar to glutamate and α-N-carboxy-β-N-methylamino-L-alanine (BMAA-α-NCO2), resembling N-methyl-D-aspartate (NMDA), an agonist of the

glutamate receptor with the same name (N-methyl-D-aspartate; NMDA) (see Figure 1 in Paper III). Under physiological conditions both forms of BMAA carbamates exist.

(21)

21

Yet another important feature of BMAA is its chelating capacity. BMAA forms stable complexes with Cu2+, Zn2+ and Ni2+ (Nunn et al., 1989b).

Neurodegeneration and signalling

Neurodegenerative diseases

The most common neurodegenerative diseases worldwide are Alzheimer´s disease, Parkinson´s disease and ALS. These diseases have some common neuropathological features, such as a progressive degradation of neurons. In addition, they all feature different forms of protein aggregation (Tanner et al., 2014). Alzheimer´s disease primarily affects neurons in the cerebral cortex of the brain and causes dementia (Murray et al., 2011), while the neurons in substantia nigra are primary targets in Parkinson´s disease leading to movement dysfunctions (Fearnley and Lees, 1991). ALS primarily affects the upper (cortical) and lower (brainstem and spinal cord) motor neurons (Burvill, 2009; Trojsi et al., 2013). Alzheimer´s disease is globally the most common of the three syndroms, with a prevalence rate in Europe of 1200 cases per 100 000 individuals, followed by Parkinson´s disease (120-170 per 100 000) and ALS (~5.4 per 100 000) (Chiò et al., 2013). Together these diseases give rise to tremendous human suffering and large expenses for the society (Olesen et al., 2012). ALS is least common, but causes the most severe effects with a life expectancy of only about 2-4 years after diagnosis (Chiò et al., 2013). Symptoms are paralysis, muscle weakness and ultimately death from respiratory failure (Burvill, 2009; Trojsi et al., 2013). As for Alzheimer´s and Parkinson´s disease, only about 10% of the ALS cases are hereditary (familial ALS) while the rest (sporadic ALS) still have an unknown background. Besides being more common in men and at higher ages, no other risk factors for sporadic ALS have been ascertained (Ingre et al., 2015). Over the years however, several behavioural and environmental factors have been suggested and investigated, such as exposure to pesticides, metals (e.g. aluminium, lead, manganese, iron and copper), cigarette smoking, physical activity, virus infection and head injuries (Ingre et al., 2015).

Notably, the incidence rates of ALS have more recently been observed to be increasing in some European countries (Sweden, Norway, Finland, France), raising the concern for an even wider spread of ALS (Fang et al., 2009; Chiò et al., 2013; Tanner et al., 2014). This evokes questions of potential human lifestyle patterns or environmental triggers in these regions, even though improved ALS diagnosis may to some extent explain the increase. To verify and eventually prevent this negative trend, additional studies on potential causative agents of ALS are urgently warranted (Ingre et al., 2015).

(22)

22

Neuron signalling and glutamate receptors

Communication and signal transduction between neurons occurs at synapses between the transmitting (presynaptic) and the receiving (postsynaptic) neuron (Figure 3). In chemical synapses, the presynaptic neuron releases a neurotransmitter into the synaptic cleft (31-50 nm wide) which activates receptors in the postsynaptic neuron, triggering signal transduction mechanisms. In mammals, glutamate (Glu) is the major excitatory neurotransmitter and is present at high concentrations (4-15 mmol/kg wet weight intracellularly and 0.2–5 µM extracellularly) in the brain (Nakanishi and Masu, 1994; Smith, 2000; Niciu et al., 2012).

Glutamate receptors are generally grouped in two major categories: ionotropic and metabotropic receptors (Figure 4). The ionotropic receptors (NMDA and α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid [AMPA]/kainate) are ion channels allowing the influx of Ca2+ and Na+ and efflux of K+ upon binding of glutamate. The chemical signal is thereby transduced to an electrical impulse. The transmembrane metabotropic receptors (mGluRs) are coupled to G-proteins that initiate downstream second messenger systems when activated by glutamate (Niciu et al., 2012).

Glutamate receptors was first thought to exist only in humans and animals, but the discovery of glutamate receptors in plants (Lam et al., 1998) soon lead to the finding of a glutamate receptor also in prokaryotic cyanobacteria (Chen et al., 1999), the progenitors of plant chloroplasts. The cyanobacterial GluR0 (or SGluR0)

Figure 3. Typical glutamatergic synapse signalling in mammals (glutamate is

represented by a red dot). Displayed is also the glutamate recycling in glial cells and the presynaptic neuron (from Niciu et al., 2012).

(23)

23

receptor is a K+ selective ion channel activated by L-α-amino acids such as most proteinogenic amino acids, for instance glutamate (highest affinity), serine and glutamine (Chen et al., 1999). The amino acid sequence of GluR0 resembles both glutamate receptors and K+ ion channels in eukaryotes, and may be considered as the ‘missing link’ between these two structures (Chen 1999; Kuner et al., 2003). Database searches in cyanobacterial genomes have revealed that this glutamate receptor sequence exists in many cyanobacteria (Kuner et al., 2003), and that it seems specifically common in cyanobacteria compared to in other prokaryotes (Ger et al., 2010).

Excitotoxicity (Olney, 1969) is a term that describes the neuronal damage caused by high concentrations of extracellular glutamate, in particular of extrasynaptic glutamate. This situation may arise during disturbances of the tightly regulated excitatory neurotransmission. Normally, glutamate is rapidly removed from the synaptic cleft by excitatory amino acid transporters (EAATs) primarily localised at adjacent glial cells and to some extent also in postsynaptic neurons (Niciu et al., 2012) (Figure 3). Disturbances in this system cause receptor overstimulation, which can lead to neurodegeneration (Choi, 1988; Weiss and Sensi, 2000) and potentially play a role in ALS (Rothstein et al., 1992; Gredal and Møller, 1995; Trotti et al., 1999).

Figure 4. The different types of glutamate receptors in mammals, showing subgroups of

the two major categories, ionotropic (ion channels) and metabotropic (G-protein coupled) receptors (modified from Niciu et al., 2012).

(24)

24 Ta bl e 1 . Su mm ar y o f BM A A ( β-N -me th y la m in o -L -a la n in e) an a ly se s p erfo rmed i n p o st m o rt em h u m an b rai n t is su es . G iv en ar e th e n u m b er a n d t y p es o f sam p le s a n al y sed , a n al y ti cal m et h o d an d BMA A co n ce n trat io n s (T ab le m o d if ie d fro m P a per IV ). Da ta se t no . H um an sa m pl es a nd th eir geo gra ph ic al o ri gi n T iss ue/ flu id F ra ct io n an al yse d M et ho d o f a na ly sis N um ber o f BM A A po sit iv e sa m pl es / to ta l s am pl es BM A A co ncen tra ti o ns Re fe ren ce 1 G ua m a. 6 A LS -P DC b. 2 c tr l Ca na da c. 2 A D d. 13 ct rl br ai n (α ) f re e (β ) f re e and p rot ei n (α , β ) H PL C -F LD * MS v er ifi cat ion 1 A QC -d er iv at iz at ion (α ) a. n ot g iv en 3 b. no re sul ts g iv en c. not g iv en 3 d. 0/ 13 (β ) a. 6 /6 b. 1/ 2 c. 2/ 2 d. 0/ 13 (α ) a. 6 µg/ g (av er ag e) b. not g iv en c. 6. 6 µg /g (av er ag e) d. – (β ) a. 3 .3 -1190 µ g/ g b. 48 -82 µg /g c. 3. 4 and 2 64 µg /g d. -C ox et a l. 2003 ( α) M ur ch et a l. 2004a 4(β ) 2 Ca na da a. 7 A D b. 1 c tr l br ai n p rot ei n H PL C -F LD * M S v er ifi cat ion 1 A QC -d er iv at iz at ion a. 6/ 7 b. 0/ 1 a. 25. 9-235. 6 µg /g b. -M ur ch et a l. 2 004b 5 3 G ua m a. 8 P DC b. 2 c tr l U SA _c.5 A D d. 5 c tr l br ai n (γ ) f re e (δ , ε ) fr ee and p rot ei n (ζ ) f re e (γ ) H PL C -F LD * F M OC -de ri vat iz at ion (δ , ε ) G C -MS * , I S, E C F-de ri vat iz at ion (ζ ) G C x G C -T OF M S * , T M S-de ri vat iz at ion (γ ,δ ,ε ,ζ ) 0/ 20 (γ ,δ ,ε ,ζ ) - M ont ine e t a l. 2005 ( γ) Sny de r et a l. 2009a 6(δ ) Sny de r et a l. 20 09b (ε ) Sny de r et a l. 2010 ( ζ) 4 U SA _a.13 A LS b. 8 H D c 1 2 A D d. 12 ct rl br ai n pr ot ei n H PL C -F LD * M S/ M S v er ifi cat ion 1 A QC -d er iv at iz at ion a. 13/ 13 b. 1/ 8 c. 12/ 12 d. 2/ 12 a. 31 -256 µg /g b. 11 µg /g c. 10 -228 µg /g d. 39 and 4 5 µg /g Pabl o et a l. 2009 5 Fr an ce 2 A LS br ai n fr ee LC -M S/ M S * und er iv at iz ed 0/ 2 - C ombes e t a l. 2 014

(25)

25 A LS -P DC : amy ot rophi c lat er al s cl er os is P arki ns oni sm d em ent ia compl ex; c tr l: cont rol ; A D: A lz he im er ´s d is eas e; H PL C -F LD: hi gh pe rform anc e liqui d chr om at og raphy -fluores ce nc e de te ct ion; M S: m as s spe ct romet ry ; A QC : 6-ami noqui nol yl -N -hy dr oxy suc ci ni m id yl carbam at e; FM O C : 9-fluoreny lm et hy lc hl oroform at e; G C -M S: ga s chr omat og raphy -m as s spe ct romet ry ; IS : int er nal s tand ard ; EC F: e thy l c hl oroform at e; G C x G C T OF M S: t w o -d im ens iona l ga s chr omat og raphy c oupl ed t o tim e-of -fl ight m as s spe ct romet ry ; T M S: tr im et hy ls ilat ion re ag ent ; A LS : amy ot rophi c lat er al sc le ros is ; H D: H unt ingt on´ s di se as e; M S/ M S: tand em m as s spe ct romet ry ; LC -M S/ M S: liqui d chr omat og raphy -t and em mas s s pe ct romet ry . * BM A A q uant ifi cat ions w er e b as ed o n t hi s me th od . 1N um be r o f v er ifi ed s ampl es n ot c le arl y s tat ed . 2R eport s L OD/ LOQ = 0. 0001/ 0. 013 µmol , mat ri x/ buffe r n ot g iv en. 3Onl y me an val ue o f t he B M A A c on ce nt rat ions g iv en. 4R eport s me tho ds a nd re sul t d et ai ls of the fr ee B M A A fr ac tion o f C ox et a l. 2003. 5Inc lud es re sul ts o f C ox et a l. 2003 a nd M ur ch et a l. 2004a and a d is cus si on th er eof. 6Short n ot es o n re sul ts o f S ny de r et a l. 2009b. 7 T he fr ee fr ac tion w as h yd rol ys ed . 8IS n ot u se d f or al l s ampl es . 9C onc ent rat ions n ot c orr ec te d w ith IS .

(26)

26

Detection of BMAA in humans and its correlation to neurodegeneration

The discovery of BMAA in the brains of deceased patients with ALS-PDC in Guam and in Canadian Alzheimer´s patients, while not in control individuals, (Cox et al., 2003; Murch et al., 2004b) was a breakthrough finding, supporting the hypothesis of BMAA being causative of neurodegenerative diseases. As seen in Table 1, several subsequent investigations corroborated these findings, including when examining

post mortem brain samples from ALS-patients, while brain tissues from patients

with the genetically underpinned Huntington´s disease were (primarily) devoid of BMAA (Murch et al., 2004a; 2004b, Pablo et al., 2009).

However, due to difficulties in replicating these data (Montine et al., 2005; Snyder et al., 2009a; 2009b; 2010), and the use of less specific methods for the analyses (high performance liquid chromatography-fluorescent detection; HPLC-FLD) (Table 1), a connection between BMAA and human neurodegeneration has occasionally been questioned. Although differences in analytical methods and in interpretation of data may have caused discrepancies in reported data (Table 1), natural variations in the material analysed could also have influenced the results. For instance it is possible that BMAA could still be a causative agent although not present at the time of analysis. Expanded investigations with regard to BMAA and neurodegeneration are therefore needed, including screening of cerebrospinal fluid (CSF) for BMAA during the entire course of the disease while employing the most appropriate methods available.

More indirect attempts to investigate the relevance of BMAA in neurodegeneration have been made through epidemiological studies. In some examinations, areas of high ALS incidence rates were identified near lakes prone to support cyanobacterial blooms (Bradley et al., 2013). The increased ALS occurrence was proposed to be related with the exposure to a cyanobacterial toxin(s) such as BMAA through for instance inhalation of aerosolised cyanobacterial cells or consumption of toxin contaminated lake organisms (Caller et al., 2009; Masseret et al., 2013). Case studies revealing BMAA in the diet of ALS-patients lend support to these suggestions (Field et al., 2013; Banack et al., 2014).

BMAA toxicity – effects and mechanisms

It is by now well established that BMAA negatively affects neurons and neurotransmission at the cellular level and some mechanisms of actions have been revealed (Chiu et al., 2011). The oral bioavailability of BMAA in mammals is high (~80%) (Duncan et al., 1991; 1992), and although the passage of the blood brain barrier appears quite restricted, low concentrations of BMAA can pass and enter the central nervous system (CNS) (Duncan et al., 1991; Smith et al., 1992; Xie et al., 2013). Transport of BMAA to the CNS may occur via the large neutral amino acid carrier (Duncan et al., 1991; Smith et al., 1992) or as recently suggested, through cerebral capillary transfer (Xie et al., 2013).

(27)

27

BMAA neurotoxicity in eukaryotic cells proceeds via over-activation of the glutamate receptors (excitotoxicity) NMDA at mM concentrations (Lindstrom et al., 1990; Copani et al., 1991; Rakonczay et al., 1991; Matsuoka et al., 1993; Brownson et al., 2002), AMPA/kainate at µM concentrations (Rakonczay et al., 1991; Matsuoka et al., 1993; Rao et al., 2006; Cucchiaroni et al., 2010) and mGluR1 (Copani et al., 1990; 1991; Lobner et al., 2007; Cucchiaroni et al., 2010) and mGluR5 (Lobner et al., 2007; Liu et al., 2009) at low µM concentrations. The activation causes increased Ca2+ levels resulting in the release of cytochrome-c from mitochondria in eukaryotic cells (Brownson et al., 2002; Rao et al., 2006; Cucchiaroni et al., 2010). Oxidative stress is caused by inhibition of the cystine/glutamate antiporter system Xc- (Lobner et al., 2007; Liu et al., 2009) and release of reactive oxygen species (ROS) from mitochondria (Rao et al., 2006; Cucchiaroni et al., 2010). Excitotoxicity is further enhanced by glutamate escaping through system Xc- (Lobner et al., 2007; Liu et al., 2009). The prolonged depolarization of the cell also leads to increased membrane permeability and release of noradrenalin (Lindstrom et al., 1990) (Figure 5). Disturbance of protein homeostasis displayed by endoplasmic reticulum (ER) stress, caspase 12 activity and increased protein ubiquitination has also been observed (Okle et al., 2012). The majority of the in vitro experiments employ cortical neurons from mice or rats, but in recent years complemented with investigations using human cell lines. Although one of the human cell studies reports only a weak toxicity of BMAA (Lee and McGeer, 2012), others corroborate many of the effects found in rodent cells (Chiu et al., 2012; Okle et al., 2012; Muñoz-Saez et al., 2013; 2015).

In addition, mis-incorporation of BMAA into proteins instead of serine was recently demonstrated in human cells (Dunlop et al., 2013), as well as in an artificial

Figure 5. Summary of some of the toxic effects excerted by the β-carbamate of BMAA

(blue dot) in neurons. The neuron is represented by the large circle, and its dotted line illustrates the permeability of the membrane caused by neuron depolarization (from Chiu et al., 2011).

(28)

28

protein system (Glover et al., 2014). This process leads to misfolded, dysfunctional proteins and protein aggregation, similar to the different protein aggregates typical of neurodegenerative diseases (Ross and Poirier, 2004; Dunlop et al., 2013). Moreover, protein incorporation of BMAA could perhaps result in a chronic toxicity with slow continuous loss of motor neurons resembling that of neurodegeneration, as previously suggested (Murch et al., 2004a).

A more unexplored mechanism suggested for BMAA neurotoxicity is that BMAA, due to its chelating capacities may disturb metal homeostasis in the CNS (Zn2+ and Cu2+) (Choi and Koh, 1998; Nunn et al., 1989b; Nunn and Ponnusamy, 2009), or generate oxidative stress by forming new oxidative centres (Bradley and Mash, 2009).

Further, still other yet unknown mechanisms might be induced by BMAA in eukaryotic cells, which could potentially be the cause of neurodegenerative diseases. Besides the wealth of BMAA studies in mammals, there are also some investigations in other organisms. Aquatic organisms can be negatively affected upon BMAA exposure as for instance observed in the increased mortality rate and developmental disturbances in zebrafish (Danio rerio), brine shrimp (Artemia

salina), the protozoan Nassula sorex (Purdie et al., 2009a; 2009b) as well as in

reduced mobility of Daphnia magna (Lürling et al., 2011). Moreover, BMAA inhibits enzymes protecting against oxidative stress in Daphnia magna and in the aquatic plants Ceratophyllum demersum, Lomariopsis lineata, Fontinalis

antipyretica, Riccia fluitans and Taxiphyllum barberi (Esterhuizen-Londt et al.,

2011, Contardo-Jara et al., 2013, Esterhuizen-Londt et al., 2015). Excitotoxic effects of BMAA have been further found in leech (Haemopsis sanuisuga) (Nedeljkov et al., 2005; Lopicic et al., 2009), honeybee (Apis mellifera) (Okle et al., 2013) and in

Drosophila (Zhou et al., 2009; 2010; Goto et al., 2012; Koenig et al., 2015). In the

latter, delayed chronic and progressive neurological effects was additionally observed (Zhou et al., 2009; 2010).

The recorded uptake of BMAA in plants potentially constitutes a route for human BMAA exposure. For instance, BMAA has been shown to be taken up in wheat (Triticum aestivum) upon irrigation with BMAA containing water (Contardo-Jara et al., 2014). On the other hand, the BMAA uptake capacity in aquatic plants such as Fontinalis antipyretica, Riccia fluitans and Taxiphyllum barberi may be employed to remediate BMAA-contaminated waters (Contardo-Jara et al., 2013).

Cyanobacteria

Cyanobacteria are the only oxygenic photosynthetic prokaryotes. They evolved some 3.5 billion years ago (Schopf, 1993), and according to the endosymbiont theory they are the progenitors of the chloroplast (Deusch et al., 2008), thereby giving rise to all plants on Earth today. In time, their oxygen (O2) evolving lifestyle

led to a marked increase in atmospheric O2 concentration (the Great Oxygenation

(29)

29

Schopf, 2014; Swanner et al., 2015). Today cyanobacteria continue to be indispensable for life on Earth, being fundamental primary producers in globally spread terrestrial, limnic and oceanic ecosystems (Dodds et al., 1995; Bergman et al., 2012). Moreover, cyanobacteria are particularly crucial agents for the glo bal carbon (C) fixation and some fix vast amounts of atmospheric dinitrogen (N2) (Karl

et al., 2002), estimated to reach ~250 Tg of ‘new’ nitrogen (N) per year (Gruber et al., 2008), thereby fuelling the biosphere with two highly sought after key-nutrients, C and N.

The phylum Cyanobacteria comprise organisms capable of surviving in nearly all illuminated environments on Earth, with the exception of low pH habitats (< ~4) (Whitton, 2000). This is likely due to their diverse morphology and often complex physiology and metabolism. They thrive in terrestrial and aquatic (marine, limnic, brackish) habitats, the majority employing a free-living lifestyle while a few genera are able to enter symbiosis, primarily with a range of plants (Bergman et al., 2008) and with a few other eukaryotes (Usher et al., 2007).

Initially cyanobacteria were termed ‘blue-green algae’ and regarded as plants due to their rich pigmentation (chlorophyll and phycobiliproteins) and oxygenic photosynthesis. Hence, until the end of the 20th century they were classified according to the Botanical Code (ICBN, the International Code of Botanical Nomenclature). Discovery of their prokaryotic cellular characters such as lack of membrane bound organelles, revealed by electron microscopy, led Stanier and co-workers (1978) to introduce these apparently prokaryotic ‘blue-green algae’ into the ICNP (International Code of Nomenclature of Bacteria), and the name was subsequently changed to cyanobacteria. Next, Rippka and co-workers (1979) divided the cyanobacterial strains into 5 sections (Section I-V) based on morphological and developmental characters, a grouping still widely used (Table 2). With the introduction of the large scale gene and genome sequencing methodologies, the cyanobacterial radiation has been expanded and diversified, and it is by now apparent that the true phylogeny of cyanobacteria is still not elucidated (Palinska and Surosz, 2014; Singh et al., 2015).

In addition to organized colony formation of unicellular cyanobacteria, three Sections (III-V) exhibit multicellularity with cells attached to each other to form filaments (strings of cells). Some of these filamentous cyanobacteria (group IV and V; Table 2) also employ cell differentiation, giving rise to motile filaments (hormogonia), sporelike cells (akinetes) or cells specialised for nitrogen fixation (heterocysts) (Adams and Duggan, 1999; Haselkorn, 2007). A notable feature of aquatic cyanobacteria is their occasional mass appearance forming widespread surface ‘blooms’ (Capone et al., 1998; Stal et al., 2003). The bloom-forming cyanobacteria are characterised by gas vesicles which render buoyancy, and these forms are often associated with both nitrogen-fixation and the production of potent toxic compounds (e.g. Nodularia, Microcystis).

(30)

30

Table 2. Sections of the cyanobacterial phylum according to Rippka and co-workers (1979).

Section Morphotype Reproduction Cell

differentiation

I Chroococcales unicellular binary fission no

II Pleurocapsales unicellular multiple or binary fission no III Oscillatoriales filamentous divide in one plane no* IV Nostocales filamentous divide in one plane (false

branching)

heterocysts, akinetes, hormogonia

V Stigonematales filamentous divide in more than one plane (true branching)

heterocysts, akinetes, hormogonia

*) A notable exception is the marine genus Trichodesmium capable of differentiating diazocytes, specialised cells used for nitrogen fixation (see Bergman et al., 2012).

Nitrogen fixation

All organisms require N for protein and nucleic acid biosynthesis, and despite the abundance of N2 in our atmosphere (78%), N is generally a limiting factor for

growth on Earth. The strong triple bond of N2 renders this gas unreactive and

unavailable for any organisms other than diazotrophs; nitrogen-fixing organisms. These are represented by a handful of microbes possessing the ability to reduce the triple bond of the N2 gas, and produce ammonia (NH3), which in turn is rapidly

converted into N-containing amino acids (glutamine and glutamate) (Stal, 2012). A range of cyanobacterial genera are able to carry out nitrogen fixation, a trait which they share with a few heterotrophic bacteria, such as some free-living bacteria and symbiotic rhizobia (symbionts of legumes) and together they contribute substantially to the global N budget (Gruber et al., 2008, Karl et al., 2002, Berman-Frank et al., 2003).

Nitrogen fixation is catalysed by the nitrogenase enzyme complex, consisting of the MoFe-protein (dinitrogenase) encoded by nifDK and the Fe-protein (dinitrogenase reductase) encoded by nifH. As seen in the simplified reaction below, the energy and reducing power required for the reduction of one N2 molecule is high

(16 ATP; adenosine triphosphate), demanding that the process is strictly regulated (Kumar et al., 2010).

8H+ + 8 e- + N2 + 16 ATP → 2 NH3 + H2 + 16 ADP + 16 Pi

Nitrogenase is an extremely O2-sensitive enzyme that probably evolved before the

oxygenation of the atmosphere (Stueken et al., 2015; Berman-Frank et al., 2003). The Fe4S4 cluster of the Fe-protein is the most sensitive part of nitrogenase and is

irreversibly damaged by O2 (Gallon, 1992), accompanied by a decrease of

(31)

31

organisation levels (gene transcription, translation and activity), although mechanisms in cyanobacteria remain to be fully elucidated (Thiel and Pratte, 2014). To protect nitrogenase from atmospheric and intracellularly produced O2 (the latter

from their own photosynthesis), cyanobacteria have evolved different strategies. Photosynthesis and N2 fixation may be temporarily (photosynthesis at day and N2

fixation at night) or spatially separated (vegetative cells and heterocysts or diazocytes) (Stal and Zehr, 2008; Bergman et al., 2012). In addition, increased respiration and additional cell layers in heterocysts further create the micro-oxic environment suitable for sustaining nitrogenase activity (Haselkorn, 2007; Bothe et al., 2010). Heterocysts and vegetative cells are mutually dependent, with vegetative cells receiving glutamine as an N source from the heterocysts in exchange for carbon/carbon skeletons in the form of sucrose and glutamate from the vegetative cells (Pernil et al., 2015).

BMAA in cyanobacteria

Since the first report on BMAA in a majority of tested cyanobacterial taxa (Cox et al., 2005), the presence of BMAA has been substantiated in a variety of laboratory strains and field collected cyanobacteria from various ecosystems worldwide, Table

3 summarises 27 publications reporting presence of BMAA in cyanobacteria.

According to these (Table 3), BMAA production cannot be coupled to any specific taxonomic group or morphology type among the cyanobacteria, nor to their lifestyle, geographic origin or to the sample fraction analysed (free, protein associated or total). It is neither clear whether the BMAA production is coupled to specific growth stages or the environmental nutritional status. This suggests that the BMAA production in cyanobacteria is constitutive. Although the majority of published data supports a cyanobacterial BMAA production (27 of totally 39 publications; ~69%; summary of Table 3 and 4), still some do not (12 of the reports summarised in

Table 4). Since there are no obvious differences between the biological material in

the investigations detecting BMAA, compared to those that do not, the varying results may rather be due to the methods employed. A clear majority of the publications reporting BMAA production by cyanobacteria (Table 3) detect BMAA in almost all samples, whereas other investigators (Table 4) are unable to find BMAA in any of a variety of tested samples. For instance, some of the largest data sets, examining up to 62 samples, are part of this latter group (Krüger et al., 2010). The most prominent methodological difference between the investigations detecting BMAA (Table 3) and those reporting its absence (Table 4), is that hydrophilic interaction liquid chromatography (HILIC) analysis of underivatized BMAA using liquid chromatography-tandem mass spectrometry (LC-MS/MS) is employed in eight of the 12 latter studies (Table 4). To date, only one study reports the identification of underivatized BMAA in cyanobacteria using LC-MS/MS (Faassen et al., 2009).

(32)

32 Ta bl e 3 . Pre se n ce o f BM A A i n cy a n o b a ct er ia. S u mm ar y o f 2 7 p u b li cat io n s d et ec ti n g BM A A , i. e. i n ap p ro x im at el y 2 2 6 /3 0 6 t es te d s a m p le s (~ 7 4 % ). N o . o f sa m pl es co nt ai ni ng BM A A /t o ta l sa m pl es N o . o f fiel d sa m pl es 1 M ai n m et ho d (d eri va ti za ti o n) F ree BM A A (µ g/ g D W ) Bo un d BM A A (µ g/ g D W ) T o ta l BM A A (µ g/ g D W ) Re fe ren ce 1/ 1 1 H PL C -F LD (A QC ) MS -v er ifi cat ion 2 0. 3 n. a. n. a. C ox et a l. 2 003 37/ 41 0 3 H PL C -F LD (A QC ) MS -v er ifi cat ion 2 3-6478 4-5415 n. a. C ox et a l. 2 005 1/ 1 0 5 m et hod s 4 7 5 10 5 25 6 Bana ck e t a l. 2 007 26/ 27 0 7 GC -M S (EZ: Faa st ™) tr ac e a m ount s 8 and u p to147 0. 1-2756 n. a. Es te rhui ze n and D ow ni ng 2008 12/ 12 12 H PL C -F LD (A QC ) M S/ M S v er ifi cat ion 2 2-276 6-48 n. a. M et cal f e t a l. 2 008 21/ 21 21 4 m et hod s 9 n. a. n. a. 2-22 Johns on et a l. 2 008 BM A A found 10 onl y f ie ld sampl es 4 m et hod s 11 not s ta te d not s ta te d not s ta te d C ox et a l. 2 009 2/ 2 0 H PL C -F LD (A Q C) M S v er ifi cat ion (al l) n. a. n. a. 23 -30 Bi di ga re e t a l. 2 009 5/ 8 12 8 LC -M S/ M S (A QC ) n. a. n. a. not q uant ifi ed C rai ghe ad e t a l. 2 009 9/ 21 13 21 LC -M S/ M S (und er iv at iz ed ) 4-42 0 n. a. Faa ss en et a l. 2 009 4/ 7 14 0 LC -M S (A QC ) n. a. n. a. 0. 027 -0. 659 R one y e t a l. 2 009 2/ 2 1 LC -M S/ M S (A QC ) n. a. n. a. 1. 9-109. 4 µg /L Spáč il et a l. 2 010 1/ 3 0 LC -M S a nd L C -M S/ M S (und er iv at iz ed a nd de ri vat iz ed ; A QC ) n. a. n. a. T rac e a m ount s 15 Li e t a l 2 010 19/ 20 16 0 LC -M S (EZ: Faa st ™) M S/ M S v er ifi cat ion 2 0. 05 -0. 976 0. 0518 -10. 616 n. a. Es te rhui ze n-Lond t e t a l. 2 011

(33)

33 DW : d ry w ei ght ; H PL C -F D: hi gh pe rform anc e liqui d chr omat og raphy -fl uores ce nt d et ec tion; A QC : 6 -ami noqui nol yl -N -h yd roxy suc ci ni m id yl ; M S: m as s spe ct romet ry ; n. a: not ana ly se d; G C -M S: ga s chr omat og raphy -m as s spe ct romet ry ; EZ :F aa st ™; LC -M S; li qui d chr omat og raphy -m as s spe ct romet ry ; LC -M S/ M S: li qui d chr omat og raphy - tan de m m as s spe ct romet ry ; C E: c api llary e le ct ropho re si s; D N S: d ans yl c hl ori de . 1 Fi el d s ampl es a re h er e de fine d as s ampl es c ol le ct ed in fie ld w ithout s ub se que nt long -t er m lab cul tiv at ion. 2 N ot s tat ed h ow many s ampl es v er ifi ed . 3 M ent ions p re lim inary d at a fr om bl oo m s ampl es o f T ric hod es m iu m s p. c ont ai ni ng ~0. 007 µg B M A A /g w et w ei ght . 4 H PL C -F D, U PL C -U V , a m ino ac id ana ly se r, L C -M S, L C -M S/ M S. 5 H PL C -F D. 6 U PL C -U V . 7 N ot s tat ed for how long t he c ol le ct ed s ampl es w er e h el d i n lab cul tur e b efore BM A A -ana ly si s. 8 De te ct ed b ut n ot q uan tifi abl e. 2/ 2 0 LC -M S (EZ: Faa st ™) M S/ M S v er ifi cat ion not re port ed 18 not re port ed 18 n. a. Dow ni ng et a l. 2 011 2/ 2 19 2 LC -M S (EZ: Faa st ™) n. a. n. a. 25 µ g/ L 20 Es te rhui ze n-Lond t a nd Dow ni ng et a l. 2 011 12/ 12 0 C E (und er iv at iz ed ) n. a. n. a. 170 -810 Bapt is ta et a l. 2 011 18/ 18 0 H PL C -F LD (A QC ) n. a. n. a. 0. 04 -63 C er vant es -C ianc a et a l. 2 012a 1/ 1 0 H PL C -F LD (A QC ) 3. 4 n. a. 5. 7 C er vant es -C ianc a et a l. 2 012b 1/ 1 0 LC -M S/ M S (A QC ) n. a. n. a. 0. 73 Jiang e t a l. 2 012b 10/ 10 10 LC -M S/ M S (A QC ) not re port ed not re port ed 2. 01 -7. 23 Jiao e t a l. 2 014 7/ 10 10 LC -M S (EZ: Faa st ™) m ax conc . 0 .252 21 m ax conc . 0 .252 21 n. a Sc ot t e t a l 2 014 25/ 67 67 H PL C -F LD (A QC ) M S/ M S v er ifi cat ion n. a. n. a. 1. 8-25. 3 µ g/ L Al -S amm ak e t a l. 2 014 not re port ed 23 onl y f ie ld sampl es LC -M S/ M S (A QC ) n. a. n. a. no quant ifi cat ions m ad e Me tc al f e t a l. 2 015 3/ 4 4 LC -M S/ M S (A QC ) n. a. n. a. 0. 002 -0. 006 Lag e e t a l. 2 015 1/ 1 0 LC -M S/ M S (EZ: Faa st ™) 0. 571 2. 511 n. a. Es te rhui ze n-Lond t e t a l. 2 015 4/ 12 12 LC -M S/ M S 24 (DN S) 0. 01 -0. 3 µ g/ L n. a. n. a. R oy -L ac hape lle e t a l 2 015

(34)

34 9 H PL C -F D, a m ino ac id ana ly se r, U PL C -M S, L C -M S/ M S. 10 BM A A w as found in the s ampl es , b ut n o de tai ls g iv en. 11 H PL C -F D, a m ino ac id ana ly se r, U PL C -M S, L C -M S/ M S. 12 Sampl es o f c yano bac te ri a m ixe d w ith ot he r p hy topl ank ton, m ai nl y g re en al ga e a nd d iat oms . B M A A p re se nt in sampl es t ha t w er e d omi nat ed b y c yano bac te ri a. 13 Sc um mat er ial fr om c yano bac te ri al b loo m s. F ilament ous c ya noba ct er ia w er e d omi nant o r s ubd omi nant in 8 of the 9 B M A A p os iti ve sa m pl es . 14 18 sampl es o f fai c ai n oo dl es w er e t es te d, b ut o nl y 7 o f t he se a ct ual ly c ons is te d o f c yano bac te ri a. 15 BM A A -d et ec tion w ith pr e-col um n de ri vat iz at ion (A QC ). 16 T hr ee o f t he se B M A A -pos iti ve s ampl es w er e re port ed a s “d et ec te d b ut n ot q uant ifi abl e”. 17 N ot s tat ed h ow many c yano bac te ri al s ampl es v er ifi ed . 18 St ud y f oc us ed o n pr ov ing the p rod uc tion of BM A A w ithi n cy ano bac te ri a, n o conc ent rat ions g iv en. 19 Fi ltr at e o f w at er s ampl es . T hr ee w er e t es te d b ut o nl y t w o had v is ibl e b loo m mat er ial . 20 T he s ec ond s ampl e h ad e ve n hi ghe r c onc ent rat ions , a lthoug h abo ve t he q uant ifi cat ion rang e. 21 N o ind iv id ual v al ue s f or t he fr ee o r p rot ei n fr ac tions g iv en, o r f or e ac h sampl e. 23 BM A A w as found in de se rt c rus t s ampl es . 24 H igh re sol ut ion M S (H R M S) w ith Q -Exac tiv e M S in a fr ag m ent at ion m od e (t -MS 2).

(35)

35 Ta bl e 4 . A b se n ce o f BM A A i n cy a n o b a ct er ial s a m p le s (~ 1 6 3 t es te d s am p le s in t o tal ). S u mm ar y o f 1 2 p u b li sh ed s tu d ie s. LC -M S: li qui d chr omat og raphy -m as s spe ct romet ry ; L C -M S/ M S: li qui d chr omat og raphy -t and em m as s spe ct romet ry ; 1H N M R : pr ot on nuc le ar m ag ne tic re sona nc e ; F M OC : 9 -fl uoreny lm et hy lc hl or ofo rm at e; A QC : 6 -ami noqui nol yl -N -h yd roxy suc ci ni m id yl ; E LIS A : e nz ym e l ink ed im m unos orbent a ss ay . 1 T es ts a lga e b loo m s ampl es b ut d oe s n ot s tat e h ow many , n or t he c ompos iti on of the o rga ni sm s i n the b loo m . 2 H PL C -F LD w ith de ri vat iz ed s ampl es (A QC ), L C -M S/ M S w ith de ri vat iz ed s ampl es (A QC ) and L C -M S/ M S w ith und er iv at iz ed s am pl es . N o BM A A de te ct ed w ith the t w o M S/ M S me tho ds . B M A A d et ec te d i n the fr ee fr ac tion of 3 cy ano bac te ri al s ampl es w ith H PL C -F LD, a lthoug h sugg es te d to be caus ed b y l ow s el ec tiv ity o f t hi s me thod , h enc e p os si bl e mi si de nt ifi cat ion. 3 M ixe d w at er s ampl es w ith bl oo m s domi nat ed b y c yano bac te ri a. A ll s ampl es t es te d p os iti ve for B M A A u si ng the A br axi s E LIS A k it, w he re as no BM A A w as d et ec te d w ith th e L C -M S/ M S me tho d. C ros s re ac tiv ity o f t he E LIS A -k it w as s ugg es te d to expl ai n the fal se p os iti ve re sul ts o f t he E LIS A -k it. 4 10 st rai ns t es te d, b ut many more sampl es t es te d o f e ac h cy ano bac te ri um , e ve n at d iffe re nt g row th st ag es . 5 4 squares o f s oi l c rus ts c ont ai ni ng cy ano bac te ri a, t hr ee s ampl es fr om e ac h square . N o . s am pl es tes ted N o . o f fiel d s am pl es M ai n m et ho d S am pl e fra ct io n tes ted Re fe ren ce 2 0 LC -M S u nd er iv at iz ed fr ee K ubo et a l. 2 008 36 4 LC -M S/ M S u nd er iv at iz ed fr ee a nd p rot ei n R os én and H el le näs 2 008 not g iv en 1 onl y f ie ld s ampl es 1H N M R fr ee M oura e t a l. 2 009 11 0 H PL C -F LD de ri vat iz ed (F M OC ) fr ee a nd p rot ei n Sc ot t e t a l. 2 009 62 0 LC -M S/ M S u nd er iv at iz ed tot al fr ac tion K rüge r e t a l. 2 010 3 0 LC -M S/ M S u nd er iv at iz ed fr ee a nd p rot ei n Li e t a l. 2 012 2 1 LC -M S/ M S de ri vat iz ed (A QC ) tot al fr ac tion Jiang e t a l. 2 012a 8 4 3 m et hod s 2 fr ee a nd t ot al Faa ss en et a l. 2 012 7 3 7 EL IS A LC -M S/ M S u nd er iv at iz ed fr ee a nd t ot al Faa ss en et a l. 2 013 10 0 LC -M S/ M S de ri vat iz ed (A QC ) and u nd er iv at iz ed tot al fr ac tion M cC arr on et a l. 2 014 10 4 0 LC -M S/ M S u nd er iv at iz ed fr ee a nd t ot al R év ei llon et a l. 2 014 12 5 12 LC -M S/ M S de ri vat iz ed (A QC ) tot al fr ac tion R ic he r e t a l. 2 015

(36)

36

The BMAA concentrations detected in cyanobacteria varies considerably, sometimes up to several orders of magnitude (Table 3). A few studies report on BMAA concentrations up to several thousand µg/g dry weight (DW) (Cox et al., 2005; Esterhuizen and Downing, 2008), some in the hundreds of µg/g DW range (Metcalf et al., 2008; Baptista et al., 2011) and yet others in a few µg/g DW range up to 20-40 µg/g DW (Banack et al., 2007; Johnson et al., 2008; Bidigare et al., 2009; Faassen et al., 2009; Cervantes Cianca et al., 2012a; Al-Sammak et al., 2014; Jiao et al., 2014). Even lower concentrations of < 1 µg/g DW have been found in cyanobacteria (Roney et al., 2009; Jiang et al., 2012b; Scott et al., 2014; Esterhuizen-Londt et al., 2015; Lage et al., 2015). The reasons for the varying BMAA concentrations in cyanobacteria may be a reflection of i) choice of method for sample preparation and potential sample loss during the procedure, ii) low selectivity of the methodology used (e.g. HPLC-FLD), and iii) insufficiently stringent quantifications (e.g. without internal standard). These factors can contribute to overestimations as well as underestimations. The fact that the concentrations of BMAA reported within each individual study are more or less consistent (within the same range), supports these suggestions and points to that differences in sample handling and analytical methodology might explain discrepancies found. Even if data obtained so far suggest that BMAA could be produced constitutively, variations in the biosynthesis may indeed exist, depending on the geno- and phenotypes examined as well as on external conditions imposed at the time of sampling, to mention a few factors. In spite of the contradictions in the published reports of BMAA in cyanobacteria, both regarding the actual presence and in reported concentrations, the bulk of collected data, in total 226 BMAA positive samples out of 469 cyanobacterial samples tested (~48%; Table 3 and 4) implies that the biosynthesis of BMAA in cyanobacteria is a common trait. However, the results collectively stress two aspects: i) the need of considerably expanded research efforts with regards to the biosynthesis of BMAA in cyanobacteria and its regulation; as well as ii) a wider consensus in the methodological approach, from sampling to extraction and the analysis of BMAA, to be able to introduce needed standardised analytical methods.

Cyanotoxins

Cyanobacteria produce an astounding variety of bioactive compounds. Properties ascribed include antibacterial, antialgal, antiviral antimycotic and even antitumor capacities, hence some are of great biomedical interest (Singh et al., 2005; Jones et al., 2009; Leão et al., 2012). Others, commonly associated with cyanobacterial blooms, are referred to as cyanotoxins (such as BMAA) as they are toxic to humans and other eukaryotic organisms. The production of cyanotoxins is spread within several cyanobacterial genera. The most common and well-studied are the hepatotoxins (microcystins, nodularin(s), cylindrospermopsis) and some neurotoxins (anatoxin-a, anatoxin-a(S), saxitoxins) (Sivonen and Jones, 1999; Metcalf and Codd,