29/2007 29/2007

AN

NE R

AN

TALA

Evolution and Detection of C

yanobacterial Hepatotoxin Synthetase G

enes

Evolution and Detection of Cyanobacterial Hepatotoxin Synthetase Genes

Dissertationes bioscientiarum molecularium Universitatis Helsingiensis in Viikki

ANNE RANTALA

Department of Applied Chemistry and MicrobiologyDivision of Microbiology

Faculty of Agriculture and Forestry and Viikki Graduate School in Biosciences

University of Helsinki

Recent Publications in this Series:

7/2007 Chun-Mei LiType III Secretion System of Phytopathogenic Bacterium Pseudomonas syringae: From Gene to Function8/2007 Saara NuutinenThe Effects of Nicotine on the Regulation of Neuronal alpha7 Nicotinic Acetylcholine Receptors and Intracellular Signalling Pathways9/2007 Jaakko AaltonenFrom Polymorph Screening to Dissolution Testing: Solid Phase Analysis during Pharmaceutical Development and Manufacturing10/2007 Jenni AntikainenSurface Proteins of Lactobacillus crispatus: Adhesive Properties and Cell Wall Anchoring11/2007 Jing LiNovel Molecular Mechanisms of Arabidopsis Disease Resistance12/2007 Piia SaloThin-Layer Chromatography with Ultraviolet and Mass Spectrometric Detection: From Preparative-Layer to Miniaturized Ultra-Thin-Layer Technique13/2007 Mikko SairanenNeurotrophins and Neuronal Plasticity in the Action of Antidepressants and Morphine14/2007 Camilla RibackaRedox-linked Proton Transfer by Cytochrome C Oxidase15/2007 Päivi RamuOuter Membrane Protease/adhesin PgtE of S. enterica: Role in Salmonella-Host Interaction16/2007 Joni AlvesaloDrug Discovery Screening and the Application of Genomics and Proteomics in the Drug Development Process for Chlamydia pneumoniae17/2007 Tomi JukkolaFGFR1 Regulated Gene-Expression, Cell Proliferation and Differentiation in the Developing Midbrain and Hindbrain18/2007 Maarit HellmanStructural Biology of the ADF-H Domains19/2007 Rasa Gabrenaite-VerkhovskayaMovement-Associated Proteins of Potato Virus A: Attachment to Virus Particles and Phosphorylation20/2007 Henna ViholaStudies on Thermosensitive Poly(N-Vinylcaprolactam) Based Polymers for Pharmaceutical Applications21/2007 Terhi HakalaCharacterization of the Lignin-Modifying Enzymes of the Selective White-Rot Fungus Physisporinus Rivulosus22/2007 Ilya BelevichProton Translocation Coupled to Electron Transfer Reactions in Terminal Oxidases23/2007 Johan PahlbergSpectral Tuning and Adaptation to Different Light Environments of Mysid Visual Pigments24/2007 Beata Kluczek-TurpeinenLignocellulose Degradation and Humus Modifi cation by the Fungus Paecilomyces infl atus25/2007 Sabiruddin MirzaCrystallization as a Tool for Controlling and Designing Properties of Pharmaceutical Solids26/2007 Kaisa MarjamaaPeroxidases in Lignifying Xylem of Norway Spruce, Scots Pine and Silver Birch27/2007 Pekka NieminenMolecular Genetics of Tooth Agenesis28/2007 Sanna KoutaniemiLignin Biosynthesis in Norway Spruce: from a Model System to the Tree

Helsinki 2007 ISSN 1795-7079 ISBN 978-952-10-4369-7

Evolution and Detection of Cyanobacterial Hepatotoxin Synthetase Genes

Anne Rantala

Department of Applied Chemistry and MicrobiologyDivision of Microbiology

Faculty of Agriculture and Forestry and

Viikki Graduate School in BiosciencesUniversity of Helsinki

Academic Dissertation in Microbiology

To be presented, with the permission of the Faculty of Agriculture and Forestry of the University of Helsinki,

for public criticism in the auditorium 1041 at Viikki Biocenter(Viikinkaari 5, Helsinki) on November 27th at 12 o’clock noon.

Helsinki 2007

Supervisor:Academy Professor Kaarina SivonenDepartment of Applied Chemistry and MicrobiologyUniversity of Helsinki, Finland

Reviewers:Dr., Docent Petri AuvinenInstitute of BiotechnologyUniversity of Helsinki, Finland

Dr. Rainer KurmayerInstitute for LimnologyAustrian Academy of Sciences, Austria

Opponent: Professor James M. TiedjeCenter for Microbial EcologyMichigan State University, USA

Printed Edita Prima OyLayout Timo Päivärinta

ISSN 1795-7079ISBN 978-952-10-4369-7 (paperback)ISBN 978-952-10-4370-3 (PDF)

e-mail anne.rantala@helsinki.fi

Front cover pictureLake Joutikas (photo by Eija Kolmonen) overlaid with an image of a DNA-chip hybridization

CONTENTS

LIST OF ORIGINAL PUBLICATIONS

THE AUTHOR’S CONTRIBUTION

ABBREVIATIONS

ABSTRACT

TIIVISTELMÄ (ABSTRACT IN FINNISH)

1. INTRODUCTION .....................................................................................................1 1.1 Cyanobacteria .....................................................................................................1 1.2 Mass occurrences of cyanobacteria ...................................................................1 1.3 Cyanobacterial toxins ........................................................................................3 1.3.1 Microcystins and nodularins ....................................................................3 1.3.2 Other toxins ...............................................................................................4 1.3.3 Detection of microcystins and nodularins...............................................5 1.4 Microcystin and nodularin biosynthesis ..........................................................5 1.5 Evolution of hepatotoxin synthetase genes .......................................................7 1.5.1 Phylogenetic sequence analysis ................................................................8 1.6 Microarrays in cyanobacterial research ...........................................................9 1.7 Detection of hepatotoxin producers by molecular methods .........................11 1.7.1 Indirect detection ....................................................................................11 1.7.2 Biosynthetic genes ...................................................................................11 1.7.2.1 Hybridization ................................................................................12 1.7.2.2 Conventional PCR ......................................................................12 1.7.2.3 Quantitative real-time PCR ........................................................14

2. AIMS OF THE STUDY ..........................................................................................22

3. MATERIALS AND METHODS ............................................................................23 3.1 Strains and environmental samples ................................................................23 3.2 Methods ............................................................................................................23 3.2.1 Methods described in publications ........................................................23 3.2.2 Additional optimization of the DNA-chip assay ...................................24

4. RESULTS .................................................................................................................25 4.1 Design of general and genus-specifi c primers for microcystin synthetase genes ................................................................................................25 4.1.1 General primers (I) .................................................................................25 4.1.2 Genus-specifi c primers (II, III) ..............................................................26 4.2. Evolution of microcystin synthetase genes in cyanobacteria (I) ..................26 4.3 Frequency and composition of microcystin producers in lakes in Finland (II) ..........................................................................................27 4.4 Association of environmental variables with the occurrence of microcystin producers and microcystin concentration (II) ..........................28 4.5 Development of the DNA-chip to detect hepatotoxin producers (III) .........30 4.5.1 Specifi city of the probes ..........................................................................30 4.5.2 Sensitivity of the probes ..........................................................................30 4.5.3 Validation of DNA-chip results with environmental samples .............30 4.6 Additional optimization of the DNA-chip assay ............................................32

5. DISCUSSION ..........................................................................................................34 5.1 Evolution of microcystin synthetase genes .....................................................34 5.2 mcyE as a molecular marker for hepatotoxin-producing cyanobacteria ....36 5.3 Potential microcystin producers in lakes in Finland ....................................38 5.4 DNA-chip as a detection method for hepatotoxin producers .......................39

6. CONCLUSIONS ......................................................................................................42

7. ACKNOWLEDGEMENTS .....................................................................................44

8. REFERENCES .........................................................................................................46

LIST OF ORIGINAL PUBLICATIONS

I Rantala A., D.P. Fewer, M. Hisbergues, L. Rouhiainen, J. Vaitomaa, T. Börner, and K. Sivonen. 2004. Phylogenetic evidence for the early evolution of microcystin synthesis. Proc. Nat. Acad. Sci. USA. 101:568-573.

II Rantala A., P. Rajaniemi-Wacklin, C. Lyra, L. Lepistö, J. Rintala, J. Mankiewicz-Boczek, and K. Sivonen. 2006. Detection of microcystin-producing cyanobacteria in Finnish lakes with genus-specifi c microcystin synthetase gene E (mcyE) PCR and associations with environmental factors. Appl. Environ. Microbiol. 72: 6101-6110.

III Rantala A., E. Rizzi, B. Castiglioni, G. de Bellis, and K. Sivonen. Identifi cation of hepatotoxin-producing cyanobacteria by DNA-chip. Accepted for publication in Environmental Microbiology

The original articles were reprinted with kind permission from the copyright holders: the National Academy of Sciences (USA), American Society for Microbiology, the Society for Applied Microbiology, and Blackwell Publishing.

THE AUTHOR’S CONTRIBUTION

I Anne Rantala contributed to designing of the study, extraction of DNA from strains, PCR, cloning, sequencing, and analysis of the mcyE gene. She participated in writing of the article.

II Anne Rantala contributed to designing of the study, fi ltration, DNA extraction from lake samples, and statistical analysis of the results. She designed Planktothrix-specifi c primers for the mcyE gene, performed all the genus-specifi c PCR reactions, ELISA analyses of unconcentrated water samples, interpreted the results, and wrote the article.

III Anne Rantala contributed to the mcyE gene PCR, some of the LDR/hybridization experiments, analysis of the hybridization data, and interpretation of the results. She carried out all the quantitative real-time PCR experiments and wrote the article.

ABBREVIATIONS

Adda Constituent C20β-amino acid in microcystins and nodularinsChl-a Chlorophyll-aCt Threshold cycle in qPCRDGGE Denaturing gradient gel electrophoresisDIP Dissolved inorganic phosphorusELISA Enzyme-linked immunosorbent assayHGT Horizontal gene transferITS Internal transcribed spacerLDR Ligation detection reactionmcyA-J, T Genes encoding microcystin synthetase subunitsMcyA-J, T Proteins encoded by the mcy genesMdha N-methyldehydroalanine, a constituent amino acid in microcystinsMdhb N-methyldehydrobutyrine, a constituent amino acid in nodularinsME Minimum evolutionML Maximum likelihoodMP Maximum parsimonyndaA-I Genes encoding nodularin synthetase subunitsnifH Gene encoding a component of nitrogenase enzyme PC-IGS Phycocyanin operon intergenic spacerPCA Principal component analysisPCR Polymerase chain reactionPPIA Protein phosphatase inhibition assayqPCR Quantitative real-time PCRRFLP Restriction fragment length polymorphismrpoC1 Gene encoding RNA polymerase subunit CRT-PCR Reverse transcriptase PCRsp./spp. SpeciesTN Total nitrogenTP Total phosphorus

ABSTRACT

Mass occurrences (blooms) of cyanobacteria are common in aquatic environments worldwide. These blooms are often toxic, due to the presence of hepatotoxins or neurotoxins. The most common cyanobacterial toxins are hepatotoxins: microcystins and nodularins. In freshwaters, the main producers of microcystins are Microcystis, Anabaena, and Planktothrix. Nodularins are produced by strains of Nodularia spumigena in brackish waters. Toxic and nontoxic strains of cyanobacteria co-occur and cannot be differentiated by conventional microscopy. Molecular biological methods based on microcystin and nodularin synthetase genes enable detection of potentially hepatotoxic cyanobacteria.

In the present study, molecular detection methods for hepatotoxin-producing cyanobacteria were developed, based on microcystin synthetase gene E (mcyE) and the orthologous nodularin synthetase gene F (ndaF) sequences. General primers were designed to amplify the mcyE/ndaF gene region from microcystin-producing Anabaena, Microcystis, Planktothrix, and Nostoc, and nodularin-producing Nodularia strains. The sequences were used for phylogenetic analyses to study how cyanobacterial mcy genes have evolved. The results showed that mcy genes and microcystin are very old and were already present in the ancestor of many modern cyanobacterial genera. The results also suggested that the sporadic distribution of biosynthetic genes in modern cyanobacteria is caused by repeated gene losses in the more derived lineages of cyanobacteria and not by horizontal gene transfer. Phylogenetic analysis also proposed that nda genes evolved from mcy genes.

The frequency and composition of the microcystin producers in 70 lakes in Finland were studied by conventional polymerase chain reaction (PCR). Potential microcystin producers were detected in 84% of the lakes, using general mcyE primers, and in 91% of the lakes with the three genus-specifi c mcyE primers. Potential microcystin-producing Microcystis were detected in 70%, Planktothrix in 63%, and Anabaena in 37% of the lakes. The presence and co-occurrence of potential microcystin producers were more frequent in eutrophic lakes, where the total phosphorus concentration was high. The PCR results could also be associated with various environmental factors by correlation and regression analyses. In these analyses, the total nitrogen concentration and pH were both associated with the presence of multiple microcystin-producing genera and partly explained the probability of occurrence of mcyE genes. In general, the results showed that higher nutrient concentrations increased the occurrence of potential microcystin producers and the risk for toxic bloom formation.

Genus-specific probe pairs for microcystin-producing Anabaena, Microcystis, Planktothrix, and Nostoc, and nodularin-producing Nodularia were designed to be used in a DNA-chip assay. The DNA-chip can be used to simultaneously detect all these potential microcystin/nodularin producers in environmental water samples. The probe pairs detected the mcyE/ndaF genes specifically and sensitively when tested with cyanobacterial strains. In addition, potential microcystin/nodularin producers were identifi ed in lake and Baltic Sea samples by the DNA-chip almost as sensitively as by quantitative real-time PCR (qPCR), which was used to validate the DNA-chip results. Further improvement of the DNA-chip assay was achieved by optimization of the PCR, the fi rst step in the assay.

Analysis of the mcy and nda gene clusters from various hepatotoxin-producing cyanobacteria was rewarding; it revealed that the genes were ancient. In addition, new methods detecting all the main producers of hepatotoxins could be developed. Interestingly, potential microcystin-producing cyanobacterial strains of Microcystis, Planktothrix, and Anabaena, co-occurred especially in eutrophic and hypertrophic lakes. Protecting waters from eutrophication and restoration of lakes may thus decrease the prevalence of toxic cyanobacteria and the frequency of toxic blooms.

TIIVISTELMÄ (ABSTRACT IN FINNISH)

Syanobakteerien massaesiintymät (kukinnat) vesistöissä ovat yleisiä kaikkialla maailmassa. Kukinnat ovat usein myrkyllisiä johtuen syanobakteerien tuottamista maksa- ja hermomyrkyistä. Yleisimmät syanobakteerien tuottamat maksamyrkyt ovat mikrokystiini ja nodulariini. Makeissa vesissä mikrokystiinejä tuottavat tavallisimmin Anabaena, Microcystis ja Planktothrix. Sen sijaan nodulariineja tuottavat vain Nodularia spumigena -syanobakteerit, joita esiintyy murtovesissä. Ongelmallista on, ettei myrkkyä tuottavia syanobakteerisoluja voida erottaa myrkkyä tuottamattomista soluista mikroskoopilla katsomalla. Sen sijaan mikrokystiini- ja nodulariinisyntetaasigeeneihin perustuvat molekyylibiologiset menetelmät mahdollistavat myrkyllisten syanobakteerien tunnistamisen.

Tässä työssä kehitettiin menetelmiä, jotka tunnistavat näiden maksamyrkyllisten syanobakteereiden DNA:sta yhden maksamyrkkyjen valmistukseen tarvittavan geenin osan. Aluksi suunniteltiin ”yleiset” alukkeet, jotka tunnistivat ja monistivat tämän geenin sekä mikrokystiinejä tuottavista Anabaena-, Microcystis-, Planktothrix- ja Nostoc-sukujen syanobakteereista (mcyE-geeni) että nodulariineja tuottavista Nodularia-suvun syanobakteereista (ndaF-geeni). Monistettujen geenien avulla tutkittiin miten nämä geenit ovat kehittyneet syanobakteerien evoluution aikana. Tutkimus paljasti, että mikrokystiini- ja nodulariinisyntetaasigeenit ovat erittäin vanhoja. Tulokset tukivat myös oletusta, että mikrokystiinisyntetaasigeenit ovat periytyneet sukupolvelta toiselle eivätkä ole siirtyneet satunnaisesti solujen välillä nk. lateraalisessa geenisiirrossa. Evoluution aikana mikrokystiinisyntetaasigeenit ovat todennäköisimmin hävinneet joiltakin syanobakteereilta ja vastaavasti säilyneet muilla, mikä on johtanut siihen että näitä geenejä on vain osalla nykyisistä syanobakteereista. Sukupuuanalyysin mukaan nodulariinisyntetaasigeenit kehittyivät mikrokystiinisyntetaasigeeneistä.

Tutkimme kuinka yleisiä eri sukuihin kuuluvat mikrokystiinin tuottajat olivat 70 suomalaisessa järvessä. Mahdollisia mikrokystiinin tuottajia havaittiin 84% järvistä käytettäessä ”yleisiä” alukkeita ja 91% järvistä, kun käytettiin eri sukuja tunnistavia alukkeita. Microcystis-sukuun kuuluvia tuottajia havaittiin 70% järvistä, Planktothrix- ja Anabaena-sukuihin kuuluvia tuottajia taas 63% ja 37% järvistä. Myrkyllisten syanobakteerisukujen esiintyminen ja erityisesti usean suvun esiintyminen samanaikaisesti oli yleisempää rehevöityneimmissä järvissä. Tutkimuksessa geenimonistustuloksia myös verrattiin tilastollisesti useisiin ympäristötekijöihin. Korrelaatioanalyysi osoitti usean suvun rinnakkaisesiintyminen liittyvän mm. veden korkeaan typpipitoisuuteen. Lisäksi regressioanalyysissä typpipitoisuuden osoitettiin osaltaan selittävän mcyE-geenien ja siten myös mikrokystiinejä tuottavien syanobakteereiden esiintymistä. Tuloksista voidaan päätellä järvien korkeiden ravinnepitoisuuksien nostavan usean mikrokystiinin tuottajasuvun esiintymistodennäköisyyttä, jolloin myös myrkyllisten massaesiintymisten kehittymisen riski on suurempi rehevissä vesissä.

Tutkimuksessa kehitettiin myös menetelmä, jolla kaikki vesinäytteessä olevat mikrokystiinin tai nodulariinin mahdolliset tuottajat voidaan tunnistaa yhdellä kertaa. Tätä nk. DNA-sirumenetelmää varten suunniteltiin Anabaena-, Microcystis-, Planktothrix-, Nostoc-, ja Nodularia-suvuille koetinparit, joilla niiden mcyE- tai ndaF-geenit voidaan tunnistaa. Koetinparit osoittautuivat spesifi siksi ja herkiksi, kun niitä

arvioitiin syanobakteerikannoilla sekä järvi- ja Itämeri-näytteillä. DNA-siruilla pystyttiin tunnistamaan ympäristönäytteissä olevat mahdolliset maksamyrkkyjen tuottajat lähes yhtä herkästi kuin kvantitatiivisella geenimonistusanalyysillä, jota käytettiin varmentamaan DNA-sirutuloksia. Optimoimalla menetelmän geenimonistusvaihetta DNA-sirumenetelmää voitiin parantaa.

Useiden eri sukuihin kuuluvien maksamyrkyllisten syanobakteerien mikrokystiini- ja nodulariinisyntetaasigeenien tutkiminen oli palkitsevaa. Geenien havaittiin olevan hyvin vanhoja. Sen lisäksi kehitettiin uusia menetelmiä, joilla voidaan tunnistaa kaikki tärkeimmät maksamyrkkyjen tuottajat. Mikrokystiiniä tuottavien Microcystis-, Planktothrix-, ja Anabaena-syanobakteerien havaittiin esiintyvän samanaikaisesti erityisesti runsasravinteisissa järvissä. Maksamyrkkyä tuottavien syanobakteerien ja siten myös myrkyllisten kukintojen esiintymistä voitaisiin mahdollisesti ehkäistä suojelemalla järviä rehevöitymiseltä ja niiden kunnostamisella.

1

Introduction

1. INTRODUCTION

1.1 Cyanobacteria

Cyanobacteria are oxygenic photosynthetic bacteria. Cyanobacteria have been and still commonly are referred to as blue-green algae, due to their capability for plantlike photosynthesis and their role as primary producers among phytoplankton species. However, the intracellular organization (no membrane-bound organelles or nucleus), cell wall structures (peptidoglycan layer), and protein synthesis machinery of cyanobacteria are similar to those of other prokaryotes, and hence they are now classifi ed as bacteria (Mur et al., 1999; Castenholz, 2001). The oldest cyanobacterial fossils have been found from approximately 3500 million-year-old rocks. Since those times, their oxygen-producing photosynthesis is believed to have developed the oxygen-rich atmosphere (Schopf, 2000). Oxygenic photosynthesis by modern green and red algae and plants is also of cyanobacterial origin. Endosymbiosis that occurred between a protist and a cyanobacterium led to formation of a membrane-bound photosynthetic organelle, the chloroplast, from the engulfed cyanobacterium (Bhattacharya et al., 2003).

Cyanobacteria are everywhere. They are ubiquitously found in marine, brackish, and freshwater environments, in sediments, attached to surfaces, or as planktonic species. They can grow even at the extreme temperatures of snow, ice, and hot springs. Cyanobacteria are also common in terrestrial environments, e.g. in hot and polar deserts, sand dunes, and rocks. In addition, cyanobacteria can form symbioses with various organisms including plants, fungi, and animals (Whitton and Potts, 2000).

In addition to their modest basic demands for life: water, light, carbon dioxide, and inorganic nutrients, certain specific features help cyanobacteria occur and predominate in such a vast variety of environments. They may tolerate desiccation, high ultraviolet radiation levels, high salt concentration, and a wide range of temperatures. They efficiently perform photosynthesis in a broad region of the light spectrum, and store essential nutrients, when plentiful in the environment. The ability of some cyanobacteria to fi x atmospheric nitrogen gives them the advantage, especially under nitrogen-limited conditions. Resting cell (akinete) formation enhances survival during adverse conditions. Gas vesicles give planktic cyanobacterial species the ability to regulate their vertical position in the water column to fi nd optimal light and nutrient concentrations (Mur et al., 1999; Castenholz, 2001)

1.2 Mass occurrences of cyanobacteria

Mass occurrences (blooms) of cyano-bacteria are common in aquatic environ-ments worldwide and are often toxic. Both hepatotoxic and neurotoxic blooms are widespread (Sivonen and Jones, 1999). These mass occurrences are usually formed by planktonic, gas-vacuolated cyanobacterial species, including Anabaena , Planktothrix (formerly called Oscillatoria), Microcystis, and Nodularia, which are also potential hepatotoxin producers. Although in the tropics cyanobacterial blooms can form throughout the year, in temperate and boreal waters they usually arise during the warm summer and autumn months when high temperatures allow for maximum

2

growth rate of cyanobacteria (Mur et al., 1999; Oliver and Ganf, 2000).

Blooms occur frequently in eutrophic waters, where cyanobacteria as well as other phytoplankton species fl ourish. Due to high biomass amounts, eutrophic waters are often turbid and light intensities are lowered. At low light intensities, however, cyanobacteria grow more efficiently than eukaryotic algae. They may also have better storage capability and higher affi nity for nitrogen and phosphorus and hence can outcompete other phytoplankton species when the amount of nutrients decreases (Mur et al., 1999; Isvánovics et al., 2000). Under the more nutrient-depleted conditions of marine waters, mass occurrences are formed mainly by cyanobacteria able to fi x nitrogen (Paerl, 2000). Mass occurrences can develop either at the water surface or the metalimnion. Surface blooms are commonly formed by Anabaena and Microcystis in freshwaters (Oliver and Ganf, 2000), and Nodularia and Aphanizomenon in brackish waters e.g. the Baltic Sea (Finni et al., 2001). Metalimnetic populations are often formed by low-light-adapted Planktothrix species (Oliver and Ganf, 2000).

A l t h o u g h e u t r o p h i c a t i o n i s acknowledged as a modern problem, cyanobacterial blooms have occurred for centuries or even millennia (Ressom et al., 1994; Bianchi et al., 2000). The fi rst scientifi c article on the toxic effects of a bloom was already published in the year 1878 (Francis, 1878). Surveys of blooms have indicated that 10-92% (mean 59%) are toxic, hepatotoxic being more common than neurotoxic blooms throughout the world (Ressom et al., 1994; Sivonen and Jones, 1999) as well as in lakes in Finland and the Baltic Sea (Sivonen et al., 1989,

1990). Hepatotoxic blooms have caused several incidents of animal poisonings and death including cattle, sheep, horses, birds, and dogs (Sivonen et al., 1990; Ressom et al., 1994). Exposure to toxins through recreational usage or drinking of contaminated water also poses a threat to human health (Kuiper-Goodman et al., 1999).

A bloom can comprise one or several species, each of which can include both toxic and nontoxic strains. Hence, the toxicity of a bloom is dependent both on the relative proportions of toxic and nontoxic strains present and the toxin production rates of the toxic strains (Sivonen and Jones, 1999). With the exception of Nodularia spumigena, strains of which are practically always toxic (Laamanen et al., 2001), toxic strains of the main microcystin producers, Microcystis, Anabaena, and Planktothrix, cannot be differentiated by conventional microscopy (Sivonen and Jones, 1999).

The same environmental variables e.g. nutrients, temperature, and light, that allow for effi cient growth of cyanobacteria and formation of blooms (Ressom et al., 1994; Mur et al, 1999) have also been associated with higher microcystin concentrations in the water (e.g. Wicks and Thiel 1990; Kotak et al, 1995, 2000; Lahti et al., 1997; Rolland et al., 2005). However, it has not been easy to identify the actual toxin producers present in the samples. Previously, this would have required isolation of several strains complemented with toxin analyses. For this reason, the statistical associations of environmental factors with different toxin producers have not been reported until introduction of molecular methods able to identify different toxin-producing genera.

Introduction

3

1.3 Cyanobacterial toxins

1.3.1 Microcystins and nodularinsThe most common cyanobacterial toxins are the cyclic peptide hepatotoxins: microcystins and nodularins. Microcystins are produced by freshwater cyanobacteria throughout the world. In addition to the main producers Microcystis, Anabaena, and Planktothrix (Oscillatoria), a few strains of planktonic Nostoc, Anabaenopsis, and Radiocystis also contain microcystins (Sivonen and Jones, 1999; Vieira et al., 2003). Furthermore, microcystin production has been linked to terrestrial/benthic strains of Anabaena, Hapalosiphon, Nostoc, Phormidium, and Plectonema, and a lichen-associated Nostoc strain (Sivonen and Jones, 1999; Oksanen et al., 2004; Mohamed et al., 2006; Izaguirre et al., 2007). Nodularins are produced solely by planktonic Nodularia spumigena strains in brackish waters (Sivonen and Jones, 1999). Recently, microcystin-producing Synechococcus was identifi ed from a marine environment (Carmichael and Li, 2006).

The general s tructure of the heptapeptide microcystins is cyclo-(D-alanine1-X2-D-MeAsp3-Z4-Adda5-D-glutamate6-Mdha7). X and Z designate variable L-amino acids, D-MeAsp is D-erythro-β-methylaspartic acid, Mdha is N-methyldehydroalanine, and Adda is (2S,3S,8S,9S)-3-amino-9-methoxy-2,6,8-trimethyl-10-phenyldeca-4,6-dienoic acid. The nodularins are pentapeptides with a general structure of cyclo-(D-MeAsp1-L-arginine2-Adda3-D-glutamate4-Mdhb5), where Mdhb stands for N-methyldehydrobutyrine (Sivonen and Jones, 1999). The main differences between the two compounds are the lack of amino acids D-Ala1 and X2, and the substitution of Mdha with Mdhb in the

nodularins (Fig. 1). Over 60 structural variants of microcystins have been identifi ed. Most commonly they consist of L-amino acid substitutions at positions 2 and 4, and demethylations at positions 3 and 7. In contrast to microcystins, only a few variants, usually with demethylations at positions 1 and 3, of nodularin have been reported. Most of these variants are highly toxic, while the rare modifi cations in the Adda-D-glutamate part of microcystins and nodularins, e.g. isomerization of Adda or acylation of D-glutamate, result in a loss of toxicity. Toxicity is also signifi cantly reduced in linear toxin molecules, which are most probably either synthesis or breakdown intermediates (Sivonen and Jones, 1999).

Microcystins and nodularins inhibit protein phosphatases types 1 and 2A of mammals and higher plants (MacKintosh et al., 1990; Honkanen et al., 1991). In mammals, the toxic effect is restricted mainly to liver cells, into which these hepatotoxins can enter via the bile salt transport system. In liver cells, inhibition of protein phosphatases leads to overphosphorylation, which in turn disturbs the delicate regulation cycles of the cells. As a result, the cytoskeleton of the liver cells shrinks, leading to collapse and separation of the hepatocytes from one another and leaving the sinusoidal capi l lary ce l ls wi thout suppor t . Consequently, the cells of capillaries also separate and blood is accumulated in the liver. With high doses this may lead ultimately to death by internal hemorrhage (Carmichael, 1994). In humans, exposure to microcystins was associated with acute liver damage, failure, and even death of hemodialysis patients after treatment with microcystin-contaminated water in Caruaru, Brazil (Jochimsen et al., 1998). Chronic low-level exposure, e.g. through

Introduction

4

drinking water, could also be harmful, since it is suspected of promoting liver cancer development (Kuiper-Goodman et al., 1999).

The purpose of microcystin and nodularin production for the organism is still not resolved. The most common role suggested is that of a deterrent against grazing by zooplankton (e.g. DeMott et al., 1991). Although microcystins were toxic against Daphnia spp., they were not responsible for actual feeding inhibition in studies with microcystin-producing and -nonproducing mutant strains that were both ingested at the same rate (Rohrlack et al., 1999, 2001). It was also suggested that microcystins act as intracellular chelators of iron, facilitating its uptake from the environment (Utkilen and Gjølme, 1995). Furthermore, sequences for ferric uptake regulator binding were found in promoter regions of microcystin synthetase (mcy) genes, indicating that mcy expression could be regulated by iron availability (Martin-Luna et al., 2006). Its role as a signaling molecule has been proposed, based on observations of microcystin regulating

Figure 1. The general structure of microcystins and nodularin. Microcystins contain seven amino acids: D-Ala1-X2-D-MeAsp3-Z4-Adda5-D-Glu6-Mdha7, where X2 and Z4 represent variable L-amino acids, e.g. microcystin-LR contains leucine (L) at position 2 and arginine (R) at position 4. Nodularin differs from microcystins in lacking the amino acids D-Ala1 and X2 (indicated by the dashed line) and having Mdhb in place of Mdha.

its own biosynthesis (Kaebernick et al., 2000, Schatz et al., 2007) or that of other proteins (Dittmann et al., 2001).

1.3.2 Other toxinsIn addition to cyclic peptide hepatotoxins, other toxins are produced by many cyanobacterial genera. A prominent group of toxins are the neurotoxic alkaloids, which include anatoxin-a, homoanatoxin-a, anatoxin-a(S), and the saxitoxins (Sivonen and Jones, 1999). Although these toxins have different modes of action, all impair the normal signaling between nerves and muscles, paralysis of respiratory muscles being the fi nal cause of death (Carmichael, 1994). Cylindrospermopsins represent alkaloids with cytotoxic effects mainly on the liver, but also on the kidney, spleen, thymus, and heart. Dermatotoxic alkaloids, aplysiatoxins and lyngbyatoxins, can cause severe dermatitis and oral and gastrointestinal infl ammation. In addition, irritant or allergenic responses are caused by cyanobacterial lipopolysaccharide endotoxins (Sivonen and Jones, 1999).

Introduction

5

1.3.3 Detection of microcystins and nodularinsMany methods can be used for detection of microcystins and nodularins and are broadly classified as biological and physicochemical methods. Bioassays using invertebrates such as brine shrimp and Daphnia sp., or cell cultures have been developed to replace the traditional mouse bioassay, which was previously used almost solely for toxicity determination (McElhiney and Lawton, 2005) . However, the more sensitive and reliable methods, protein phosphatase inhibition assays (PPIAs) and the enzyme-linked immunosorbent assays (ELISAs), are now widely used to measure these toxins. They are especially suitable for determining total toxin concentration and screening of samples (Harada et al., 1999; McElhiney and Lawton, 2005; Spoof, 2005).

PPIAs quantify the amount of toxins by measuring the inhibitory effect of microcystins and nodularins on protein phosphatases. However, PPIAs do not detect different toxin variants with the same sensitivity and they cannot differentiate microcystins from unrelated protein phosphatase inhibitors that may be present in the samples. ELISAs detect microcystins and nodularins through the use of antibodies. Although antibodies offer sensitive and specifi c detection and quantifi cation of the toxin molecule they were raised against, some of the other variants may be more poorly detected or even go undetected (McElhiney and Lawton, 2005).

Analyt ica l (physicochemical ) methods such as high performance liquid chromatography followed by detection with photodiode array or mass spectrometry provide tools for identification of microcystin variants present in the samples (McElhiney

and Lawton, 2005; Spoof, 2005). Lack of commercially available standards can, however, prevent unambiguous identifi cation of different variants (Spoof, 2005).

1.4 Microcystin and nodularin biosynthesis

Microcyst ins and nodularins are synthesized nonribosomally by large multienzyme complexes that include both peptide synthetases and polyketide synthases. These systems are widely used both by bacteria and lower eukaryotes to produce secondary metabolites, e.g. antibiotics (Hopwood and Sherman,1990; Kleinkauf and von Döhren, 1995; Marahiel et al., 1997). Cyanobacteria also produce a wide variety of secondary metabolite peptides other than microcystins/nodularins, most of which are synthesized nonribosomally (Welker and von Döhren, 2006). The sequence of the peptide produced is ‘written’ in the modular structure of the peptide synthetases. Each module activates and incorporates one amino acid (substrate) of the fi nal peptide by the constituent adenylation, thiolation, and condensation domains (von Döhren et al., 1999). Further diversity in peptides is introduced by tailoring enzymes, e.g. methyltransferases and epimerases (Marahiel et al., 1997). The substrate specificity and hence the amino acid activated by an adenylation domain can be predicted from its signature sequence, i.e. the amino acid residues lining the substrate-binding pocket (Stachelhaus et al., 1999). The role of peptide synthetases in microcystin biosynthesis was first implicated when certain peptide synthetase adenylation domain gene sequences were discovered only in microcystin-producing Microcystis strains either by hybridization

Introduction

6

(Meißner et al., 1996; Neilan et al., 1999; Nishizawa et al., 1999) or polymerase chain reaction (PCR) amplifi cation (Dittmann et al., 1996; Neilan et al., 1999). This soon led to characterization of the entire gene cluster in Microcystis (Nishizawa et al., 2000; Tillett et al., 2000). Involvement of the mcy genes was further confi rmed by insertional mutagenesis experiments that abolished microcystin production (Dittmann et al., 1997; Nishizawa et al., 1999, 2000; Tillett et al., 2000). Peptide synthetase and polyketide synthase gene sequences were also detected in nodularin-producing Nodularia strains by PCR (Moffi tt and Neilan, 2001).

The enzymes needed for microcystin and nodularin synthesis are encoded by microcystin (mcy) and nodularin synthetase (nda) genes. These large, approximately 48-55 kb, synthetase gene clusters have been sequenced from the main producer organisms: Microcystis (Nishizawa et al., 1999, 2000; Tillett et al., 2000), Planktothrix (Christiansen et al., 2003), Anabaena (Rouhiainen et al., 2004), and Nodularia (Moffitt and Neilan, 2004). There are differences in the number and order of the mcy genes of Microcystis, Anabaena, and Planktothrix, but they all contain mcyA-E and mcyG (Fig. 2). The peptide synthetases encoded by mcyA, mcyB, and mcyC activate and incorporate the precursors for Mdha7, D-Ala1 (mcyA), X2, D-MeAsp3 (mcyB), and Z4

(mcyC) amino acids found in microcystins. Peptide synthetases encoded by the mixed peptide synthetase/polyketide synthase genes mcyE and mcyG incorporate the precursors for D-Glu6 and Adda5, respectively. The polyketide synthases encoded by mcyD, mcyE, and mcyG carry out the remaining Adda5 synthesis (Tillett et al., 2000; Christiansen et al., 2003; Rouhiainen et al., 2004). Almost all genes

with equivalent functions can be found in the nda cluster of Nodularia. The main difference is due to the lack of amino acids D-Ala and X and the corresponding gene regions (Moffi tt and Neilan, 2004). Of the tailoring genes, the ones encoding a putative ABC transporter, and an O-methyltransferase are found in each of the mcy/nda gene clusters. However, genes coding for an aspartate racemase and a dehydrogenase are not found in the mcy gene cluster of Planktothrix. In addition, the mcyT that encodes a putative thioesterase is only found in the mcy gene cluster of Planktothrix (Christiansen et al., 2003) (Fig. 2).

Microcystin production varies between individual strains and the microcystin content is, although to a lesser extent, also affected by environmental factors such as temperature, light, pH, and nutrients (Sivonen and Jones, 1999). Data on the effects of environmental factors are contradictory, but cyanobacteria are commonly believed to produce most toxins under the most favorable growth conditions (Orr and Jones, 1998; Sivonen and Jones, 1999). Accordingly, molecular studies have shown an increase in the transcription levels of mcy genes when transferred to light and under nitrogen-replete conditions (Nishizawa et al., 1999; Gobler et al., 2007). The transcription levels of mcyB and mcyD remained constant in low and medium light in a Microcystis strain, but increased under high and red light conditions and decreased when exposed to darkness or oxidative or salt stress. The light intensities did not, however, affect the toxin content of the cells, which was suggested to be due to active transport of microcystins outside the cells by an ABC transporter (McyH) under high-light conditions (Kaebernick et al., 2000). In Planktothrix, the amounts

Introduction

7

of both mcyA transcripts and microcystins as well as the growth rate increased when light intensity was increased to 60 μmol m-2 s-1, but decreased at an intensity of 100 μmol m-2 s-1 (Tonk et al., 2005).

1.5 Evolution of hepatotoxin synthetase genes

Strains of cyanobacteria that produce micro cystins or nodularins are sporadi-cally distributed among modern cyano-bacterial genera. In addition, all the main hepatotoxin-producing genera (Anabaena, Microcystis, Planktothrix, Nostoc, and Nodularia) include both toxic and nontoxic strains (Sivonen and Jones, 1999). This phenomenon could have been caused by two mechanisms: vertical inheritance of mcy genes from common ancestors coupled with gene loss in some of the more derived lineages or horizontal

Figure 2. Schematic representation of microcystin synthetase gene (mcy) clusters of Microcystis, Planktothrix, and Anabaena, and nodularin synthetase gene (nda) cluster of Nodularia. The light-green bars represent peptide synthetase genes and orange bars the polyketide synthase genes. Genes encoding tailoring enzymes are indicated with the other colors: the black bars denote aspartate racemase genes, light-blue bars ABC transporter genes, dark-green bars dehydrogenase genes, red bars O-methyltransferase genes, and yellow bar a thioesterase gene.

Microcystis PCC 7806 mcy gene cluster (Tillett et al., 2000)

A B CH G E DJ FI

Anabaena sp. 90 mcy gene cluster (Rouhiainen et al., 2004)

A B CH GE DJFI

G

Planktothrix CYA 126/8 mcy gene cluster (Christiansen et al., 2003)

A B C JHEDT

Nodularia NSOR10 nda gene cluster (Moffitt and Neilan, 2004)

A BI CF DEH G

gene transfer (HGT) transmitting the gene cluster randomly among strains. Both gene loss and HGT are now acknowledged to have an important role in shaping prokaryotic genomes (Ochman et al., 2000; Mira et al., 2001; Ragan, 2001; Berg and Kurland, 2002; Beiko et al., 2005).

Gene loss and HGT have also been suggested to explain the sporadic distribution of mcy genes. Constant connection of the mcyABC operon with the same flanking DNA region in 20 toxic Microcystis strains was seen as evidence against HGT, which was not expected to have always inserted the genes in the same, single chromosomal location (Tillett et al., 2001). Gene loss in nontoxic strains was suggested by consistent occurrence of mcyA and mcyB genes in the toxic Microcystis strains studied, whereas chimeric mcyB modules and nonconcordance of mcyABC sequence

Introduction

8

phylogenies with repetitive extragenic palindromic element and random amplifi ed polymorphic DNA fingerprints implied that recombination and HGT had also been involved in shaping the gene cluster (Mikalsen et al., 2003). Comparison of the mcy gene clusters from Microcystis and Planktothrix likewise did not support HGT of the entire gene cluster, but possibly only of a region of mcyA and mcyB that had a higher GC content and lower conservation than that of other mcy regions showing higher identity between Microcystis and Planktothrix (Christiansen et al., 2003). It was also hypothesized that the mcy genes of Microcystis and Planktothrix originated from nodularin biosynthesis genes by independent insertions of this mcyAB region, causing addition of the extra two amino acids present in microcystins but not in nodularins (Christiansen et al., 2003).

There are various methods for detecting HGT among microbial genomes (Ochman et al., 2000; Ragan, 2001). HGT can be suspected, e.g. when the sequence characteristics of a gene region such as GC content and codon usage patterns differ markedly from the ones of vertically transmitted DNA, or when a gene region includes plasmid-, phage-, or transposon-related sequences. Restricted phylogenetic distribution of a gene region among closely related strains and species and at the same time, high levels of sequence similarity to genes from distantly related taxa can also be a sign of HGT (Ochman et al., 2000). However, the strongest evidence for HGT is provided by the incongruence of well-supported phylogenetic trees, which indicates different genealogical histories for the genes compared. In contrast, sporadic distribution of genes caused by lineage-specifi c gene loss does

not interfere with congruence of the trees (Ragan, 2001).

1.5.1 Phylogenetic sequence analysisPhylogenetic studies are performed to reconstruct the genealogical relationships between species/genes and to estimate the time of divergence between them. This evolutionary history is depicted by phylogenetic trees (Li and Graur, 1991). Usually data from nucleotide or amino acid sequences are used to build trees. The most commonly used tree-building methods are distance, maximum parsimony (MP), and maximum likelihood (ML) methods (Nei and Kumar, 2000). Common to distance methods is that evolutionary distances (number of substitutions) are calculated for all pairs of taxa. Many methods, e.g. neighbor-joining or minimum evolution (ME) are available for determination of tree topology from the distance data. The MP method uses actual sequence data to compute the smallest number of substitutions that explains the evolutionary process for all potentially correct tree topologies. The topology that has the shortest tree length, i.e. requires the smallest number of substitutions, is then chosen as the best tree. The ML methods maximize the likelihood of observing a certain sequence set under specific substitution models for each topology and the tree topology with the highest maximum likelihood is chosen as the fi nal tree. The distance and ML methods use models with assumptions of nucleotide substitution patterns. Since any model can only be an approximation of reality, they can mislead tree construction. For this reason, model-free MP methods are believed to result in more reliable trees (Nei and Kumar, 2000).

To resolve the potential involvement of HGT in the evolutionary process

Introduction

9

by comparing phylogenetic trees, it is important that the tree topologies are reliable. The reliability of tree-construction methods can be improved by using sequences that are long enough so that the number of substitutions is not subject to large stochastic errors (Nei and Kumar, 2000). Topological errors can be caused by the use of outgroups that are either phylogenetically too distant or too close (Li and Graur, 1991). The reliability of tree topology is increased if the same topology is inferred with various methods. The statistical significance of the branching order can be evaluated with resampling, e.g. bootstrapping techniques. A robust topology is indicated with high bootstrap values for all interior branches (Nei and Kumar, 2000). The incongruence/congruence of the trees compared can also be tested, e.g. with a partition homogeneity test (Farris et al., 1995). To detect HGT in the evolutionary history of the gene of interest, its gene tree is usually compared with that of some conservative molecular marker. The most trusted marker is 16S rRNA, the evolutionary history of which is believed to be at least less affected by, if not immune to, the effects of HGT, interfering with its complex interactions with other molecules (Doolittle, 1999; Jain et al., 1999).

1.6 Microarrays in cyanobacterial research

In cyanobacterial research, DNA microarray technology has been mainly used for two purposes: to screen genome-wide changes in transcription profiles or to detect and identify cyanobacteria (diagnostic arrays). The use of microarrays has been enabled by the increasing number of sequences available from many genes and even entire genomes

(Burja et al., 2003). Today, the Genome Projects Database of the National Center for Biotechnology Information states that 26 cyanobacterial genomes have been completely and 32 partially sequenced (http://www.ncbi.nlm.nih.gov/genomes/lproks.cgi; July 17, 2007). However, none of the completed sequences are from microcystin-producing strains, and only two of the unfinished are from a microcystin-producing strain, Microcystis aeruginosa PCC 7806, and a nodularin-producing strain, Nodularia spumigena CCY9414. The cyanobacterium studied most by microarrays is Synechocystis sp. PCC 6803, the genome sequence of which was already released in 1996 and for which a commercial microarray is available (IntelliGene® Cyano CHIP Ver. 2.0; Takara Bio Inc.) (Burja et al., 2003).

The majority of microarray studies with cyanobacteria have used microarrays, e.g. to determine gene expression changes in response to various stress factors such as high light levels (Hihara et al., 2001), low temperature (Suzuki et al., 2001), desiccation (Katoh et al., 2004; Sato et al., 2004), or nutrient depletion (Ehira et al., 2003; Sato et al., 2004; Martiny et al., 2006), and to identify genes controlled, e.g. by the circadian clock (Kucho et al., 2004) or quality of light (Stowe-Evans et al., 2004; Steglich et al., 2006). Information on how cyanobacteria sense and respond to environmental stresses was also gained through the use of these microarrays (Marin et al., 2003; Murata and Suzuki, 2006; Yoshimura et al., 2007).

Diagnostic microarrays include sequences representing only one or a few genes from a variety of organisms and hence can be used to explore the biodiversity of samples. However, only a few cyanobacterial studies have been performed with diagnostic arrays (Burja

Introduction

10

et al., 2003). Transcriptional differences of genes putatively associated with saxitoxin biosynthesis between saxitoxin-producing and -nonproducing Anabaena circinalis strains were studied with a microarray including DNA fragments from 16S rRNA and various cyanotoxin biosynthesis genes (Pomati et al., 2006). Microarrays were used to detect the presence and expression of nifH in marine microbial mats (Moisander et al., 2006) and to investigate the spatial-temporal distribution of nitrogen-fi xing organisms, cyanobacteria among other bacteria, in water and sediment samples of Chesapeake Bay (Moisander et al., 2007). Microarrays were also developed for detection and identification of cyanobacterial groups with probes designed for 16S rRNA genes and used to explore the biodiversity of cyanobacteria in lake samples (Rudi et al., 2000; Matsunaga et al., 2001;

Castiglioni et al., 2004). The microarray method used by Castiglioni et al. (2004), which is also used in the present study, includes two separate steps (Gerry et al., 1999; Fig. 3). In the first step, the previously amplifi ed 16S rRNA genes are recognized in a ligation detection reaction (LDR) by two genus-specifi c probes. One probe (common) has a complementary ZipCode (cZipCode) sequence at its 3’ end and the other probe (discriminating) has a fluorescent dye, Cy3, attached to its 5’ end. The discriminating probe is named after the nucleotide(s) at its 3’ end, which is/are specifi c to the group of target sequences. Ligation requires adjacent attachment of the probes to the target sequence and perfect base pairing of the discriminating nucleotide(s). As a result of successful ligation, a molecule that has a Cy3 at one end and a cZipCode at the other is formed. In successive cycles of

Figure 3. Principle of ligation detection reaction (LDR) (Gerry et al., 1999) and hybridization used in DNA-chip detection of cyanobacterial phylogenetic groups (Castiglioni et al., 2004). In LDR, the discriminating and common probes attach to the target molecule (amplifi ed 16S rRNA gene) one after the other. Ligation of the probes by Pfu ligase requires that both probes are perfectly base-paired at the junction. In hybridization, the fl uorescent dye is addressed to the spot via recognition of the correct ZipCode by the cZipCode.

PCR productcZipCode

Common probeDiscriminating probe

POHCy3

ZipCode

Glass slideSpot

Hybridization

Ligation Detection Reaction

PCR productcZipCodeCy3

+ Pfu ligase

Cy3

Introduction

11

LDR, the same target is recognized again, leading to amplifi cation of the signal. The second step comprises hybridization, in which the cZipCodes find and attach to the corresponding ZipCode sequences printed on glass slides. However, the fl uorescent signal is detected only from probes that have found their target gene in LDR and been successfully ligated to each other. The array format used is referred to as universal, because the same set of ZipCodes/cZipCodes can be used irrespective of the probes attached to them (Gerry et al., 1999). LDR is especially suitable for differentiation of relatively similar sequences, e.g. 16S rRNA (Castiglioni et al., 2004), due to its ability to detect small-scale nucleotide variations (Gerry et al., 1999).

1.7 Detection of hepatotoxin producers by molecular methods

1.7.1 Indirect detectionDifferentiation of hepatotoxin-producing and -nonproducing cyanobacterial strains was already acknowledged as an important task at the beginning of development of molecular taxonomy methods. Hybridization of the genomic DNA digests with a repeated heptamer oligonucleotide differentiated the hepatotoxic Anabaena strains from Nostoc and neurotoxic Anabaena strains (Rouhiainen et al., 1995), but the hybridization patterns for hepatotoxic strains were heterogeneous. Later, phylogenetic studies with the most widely used 16S rRNA gene showed that clusters of the most important microcystin producers (Anabaena, Microcystis, and Planktothrix) (Neilan et al., 1997; Lyra et al., 2001; Gugger et al., 2002), always include both toxic and nontoxic strains hampering its use for discrimination. Even the more variable 16S-23S internal transcribed spacer (ITS) region was

not able to separate toxic and nontoxic Microcystis strains (Otsuka et al., 1999). Despite this, the ITS region was recently used in an attempt to differentiate toxic and nontoxic Microcystis colonies with denaturing gradient gel electrophoresis (DGGE) (Janse et al., 2004).

On the other hand, differentiation of nodularin-producing Nodularia spumigena strains seems to be possible with the use of either the 16S rRNA gene (Lehtimäki et al., 2000, Moffitt et al., 2001), the phycocyanin operon intergenic spacer (PC-IGS), or the 16S-23S ITS regions (Laamanen et al., 2001). With the restriction fragment length polymorphism (RFLP) analysis and sequencing of the 16S RNA gene, toxic Nodularia strains could be grouped separately from nontoxic Nodularia and other cyanobacterial species (Lehtimäki et al., 2000). Subsequently, PCR primers specifi c both for Nodularia and toxic Nodularia strains were designed for the 16S rRNA (Moffitt et al., 2001) and RFLP analysis of PC-IGS amplicons was developed for differentiation of toxic Nodularia filaments (Laamanen et al., 2001).

1.7.2 Biosynthetic genesAssociation of a phenotypic feature with a certain gene sequence always includes a risk of erroneous association. This risk becomes higher the more unrelated the feature and the gene are. Thus, the most reliable basis for detection of hepatotoxin producers would be the microcystin and nodularin biosynthetic genes. There are, however, strains that possess the biosynthetic genes, but do not (detectably) produce microcystins. Among Microcystis strains these inactive mcy genotypes are fairly rare (Nishizawa et al., 1999, 2000; Tillett et al., 2001; Kaebernick et al., 2001; Mikalsen et al., 2003; Via-Ordorika et al., 2004), but the proportion among

Introduction

12

Planktothrix strains seems to be higher (Kurmayer et al., 2004; Christiansen et al., 2006). In two Alpine lakes, the proportion of inactive Planktothrix-mcy genotypes was calculated to represent 5% and 21% of the population (Kurmayer et al., 2004). The failed detection of microcystin-producing Microcystis colonies was also reported (Via-Ordorika et al., 2004). Although important during study of individual strains, occurrence of false-positive or false-negative genotypes will not hamper environmental monitoring as long as their proportion among truly toxic strains does not form a majority. Table 1 lists studies in which the biosynthetic genes were used for the detection and identifi cation of potential microcystin or nodularin producers.

1.7.2.1 HybridizationHybridization methods have been used to detect the mcy genes in only a few studies. Southern hybridization was performed to explore the presence of mcy genes in microcystin-producing and -nonproducing Microcystis strains (Nishizawa et al., 1999, 2000). Dot blot hybridization with Microcystis-derived mcyB fragment also revealed the presence of homologous genes in microcystin-producing strains of other cyanobacterial genera: Anabaena, Nodularia, Nostoc, and Oscillatoria (Planktothrix) (Neilan et al., 1999). In addition, an RFLP in the mcyABC region of Microcystis strains was discovered by Southern hybridization, leading to recognition of the causative recombination event between adenylation domains of mcyB and mcyC (Mikalsen et al., 2003).

1.7.2.2 Conventional PCR The PCR has replaced the more labor-intensive hybridization methods and is currently used almost exclusively

for detection of potential microcystin or nodularin producers. Detection and identifi cation can be achieved in the PCR either with primers targeting one producer genus at a time or with ‘general’ primers that detect several genera simultaneously, complemented with a post-PCR analysis, such as sequencing or RFLP analysis, to differentiate the amplicons.

The majority of the PCR primers have been designed to detect microcystin-producing Microcystis and the targeted gene has most often been mcyA or mcyB (Table 1). With cultured Microcystis strains and isolated colonies, detection of these genes has been used, e.g. to explore how well the presence of biosynthetic genes correlates with the ability to produce microcystins (e.g. Tillett et al., 2001; Via-Ordorika et al., 2004; Saker et al., 2005a; Yoshida et al., 2005), to reveal the co-occurrence of strains with and without biosynthetic genes in lakes (Wilson et al., 2005; Yoshida et al., 2005), or to show that the proportion of microcystin-producing genotypes varies in different Microcystis morphospecies (Kurmayer et al., 2002; Via-Ordorika et al., 2004).

Primers targeting mcyA and mcyB were also used for detecting microcystin-producing Microcystis in fi eld samples, e.g. from catfi sh production ponds (Nonneman and Zimba, 2002) or lakes used for recreation and drinking water (Baker et al., 2002; Pan et al., 2002; Yoshida et al., 2005; Ouellette et al., 2006). The potential downfall of using primers designed originally to detect microcystin-producing Microcystis strains is the lack of detection of other toxin-producing genera in mixed populations of environmental samples. This was suggested by the differing results obtained with two primer pairs detecting mcyB and mcyD in Lake Oneida (Hotto et al., 2005) and lack of detection of mcyB

Introduction

13

and mcyD in a sample where microcystins were detected in Lake Erie (Rinta-Kanto et al., 2005). In both cases, microscopy had revealed presence of additional genera (Anabaena, Planktothrix) potentially capable of producing microcystins in the lakes studied.

Only one primer pair has been designed to specifi cally detect microcystin-producing Anabaena strains. Primers for Anabaena-mcyE showed high specifi city when tested among toxic and nontoxic Anabaena, Microcystis, Planktothrix, and Nostoc strains (Vaitomaa et al., 2003). High stringency of detection was also seen with Microcystis-specific primers designed to detect the same mcyE region (Vaitomaa et al., 2003). The level of specifi city observed enables identifi cation of the potential microcystin producer with high probability, even without sequencing of amplification products. Parallel use of the two primer pairs was used to determine which of the two genera was the potential microcystin-producer in Lake Agawam (Gobler et al., 2007). The Anabaena- and Microcystis-specifi c primers designed by Vaitomaa et al. (2003) and the Planktothrix-specific mcyE primers designed in this study (II) were used to confi rm Microcystis as the dominant microcystin producer in Lake Ontario samples (Hotto et al., 2007).

P r imer s des igned to de t ec t microcystin-producing Planktothrix were used to determine how abundant inactive mcyA genotypes are among P. agardhii and P. rubescens strains, and in Planktothrix populations in Lake Irrsee and Lake Mondsee (Kurmayer et al., 2004). Mbedi et al. (2005) evaluated four mcy gene and four intergenic regions for use as molecular markers with Planktothrix strains. Specifi c detection was achieved in both strains and lake water samples

with mcyE primers, which were the most suitable molecular marker for microcystin-producing Planktothrix.

The fi rst primers to detect nodularin-producing Nodularia strains, although designed for peptide and polyketide synthase genes before actual recognition of the biosynthetic genes, resulted in a relatively specific detection of toxic Nodularia strains (Moffitt and Neilan, 2001). However, primers later designed particularly for ndaF were highly specifi c for nodularin-producing Nodularia and suitable for detection both with conventional PCR and quantitative real-time PCR (qPCR) applications (Koskenniemi et al., 2007).

Only a few primer pairs have been designed for simultaneous detection of biosynthetic genes of different cyanobacterial genera (Table 1). The use of such general primers is especially valuable in studying environmental samples that could contain microcystin producers of several genera. Design of the fi rst general primer pair had to rely only on Microcystis-specifi c mcyB sequences, since at that time orthologous sequences from other microcystin-producing genera were not available. However, the primer pair could also amplify the gene region of the same size from microcystin-producing Anabaena, Nostoc, Oscillatoria (Planktothrix), and nodularin-producing Nodularia strains (Neilan et al., 1999). A general primer pair designed for mcyE could detect hepatotoxin-producing strains of the genera Anabaena, Microcystis, Phormidium, Planktothrix, Nostoc, and Nodularia (Jungblut and Neilan, 2006). However, when these general primers are used, cloning and sequencing of the PCR product is required for the final identifi cation of the producer organisms in environmental samples. Hisbergues et

Introduction

14

al. (2003) instead determined the origin of the general mcyA PCR product, using an RFLP that enabled differentiation between microcystin-producing Anabaena , Microcystis, and Planktothrix strains. The mcyA PCR and subsequent sequencing or RFLP analysis have been used, e.g. to identify the responsible microcystin producers in lake and river water samples (Hisbergues et al., 2003; Rinta-Kanto and Wilhelm, 2006; Hotto et al., 2007; Saker et al., 2007a). In addition, sequencing of the general mcyA (Hisbergues et al., 2003) and mcyE PCR (Jungblut and Neilan, 2006) products was used to detect and identify microcystin-producing Microcystis in several health food supplements (Saker et al., 2005b; 2007b). If identifi cation of the producer organism is not needed for some reason, the use of general primers offers a simple and sensitive method for monitoring water samples for the presence of potential hepatotoxin producers (Anjos et al., 2006; Mankiewicz-Boczek et al., 2006).

Reverse transcriptase (RT) -PCR applications have rarely been used, although detection of mcy- or nda-specific RNA would give proof of actively transcribed genes. This could be due to potential problems associated with analysis of RNA populations of single-copy genes, which can decrease the reliability of detection. RNA is more sensitive to digestion by nucleases than DNA and thus its protection under fi eld conditions is more challenging. Changes in gene expression levels and patterns could also be induced by conditions during sampling and transport to the laboratory. In addition, regulation of microcystin production could still occur at a later stage, as was suggested by an mcy transcript derived from a Microcystis strain not producing microcystins (Mikalsen et al.,

2003). Temporal variation in transcription of mcyE in Lake Agawam was revealed by RT-PCR analysis with Microcystis-specifi c primers. Higher expression during the summer months coincided with denser Microcystis populations, nutrient-replete conditions, and suppressed grazing by Daphnia sp. (Gobler et al., 2007).

1.7.2.3 Quantitative real-time PCR The main advantage of qPCR over conventional PCR is its ability to quantify the target gene copy numbers. Rapid and sensitive qPCR applications detecting hepatotoxin biosynthetic genes have been developed for Microcystis, Anabaena, Planktothrix, and Nodularia (Table 1). The methods used include both Taq nuclease and SYBR Green assays, in which fl uorescence originating from the breakdown of a labeled TaqMan probe or from the dye bound to double-stranded DNA is measured at each PCR cycle. The cycle at which fluorescence rises above the background level is called the threshold cycle (Ct). Both methods quantify the gene copy number of a sample by comparing its Ct to those of external standard dilutions with known concentrations of the gene. Differences in amplification efficiencies between standards and samples, which may contain PCR inhibitors, can compromise the reliability of quantifi cation. In addition, the sensitivity of detection can be infl uenced by the length of the PCR product. For example, a tenfold lower detection limit was attained, using an amplicon of 190 bp compared with an amplicon of 1369 bp (Furukawa et al., 2006).

SYBR Green-based assays have been used, e.g. to determine the seasonal variations in the proportion of Microcystis-mcyA genotypes in a lake in Japan (Yoshida et al., 2007). Additionally, the proportion

Introduction

15

of mcyA genotypes was positively correlated with nitrate concentrations, but not with temperature or ortho-phosphate concentrations in the lake (Yoshida et al., 2007). The copy numbers and relative proportions of both microcystin-producing Anabaena and Microcystis in two lakes in Finland were estimated with genus-specific mcyE primers (Vaitomaa et al., 2003). In the Baltic Sea samples, the copy numbers of Nodularia-ndaF were highest in the upper water layers, but present also in the deeper layers. In addition, a strong positive correlation between ndaF copy

numbers and nodularin concentrations was found (Koskenniemi et al., 2007). Taq nuclease assays were utilized in assessing variations in the proportion of Microcystis-mcyB genotypes from total Microcystis (PC-IGS) genotypes in a lake in Germany (Kurmayer and Kutzenberger, 2003). The ability of this assay to reliably provide quantitative estimates of mcyB was further confirmed by highly consistent results obtained in three different laboratories using different instruments (Schober et al., 2007).

Introduction

Met

hod

Gen

us/g

ener

a ta

rget

ed b

y th

e pr

imer

s/pr

obes

Gen

e/ge

nes t

hat

prim

ers o

r pr

obes

are

de

sign

ed to

det

ecta

Mai

n re

sults

reg

ardi

ng d

iffer

entia

tion

of to

xic

and

nont

oxic

stra

ins a

nd

dete

ctio

n/id

entifi

cat

ion

of to

xic

stra

ins

Ref

eren

ce

Sout

hern

hy

brid

izat

ion

Mic

rocy

stis

Ade

nyla

tion

dom

ains

of

mic

rocy

stin

synt

heta

se

gene

s A (m

cyA)

and

B

(mcy

B)

Ade

nyla

tion

dom

ains

(mcy

A, m

cyB)

pre

sent

in to

xic

Mic

rocy

stis

stra

ins (

6/8)

and

in

one

non

toxi

c st

rain

Nis

hiza

wa

et a

l., 1

999

Sout

hern

hy

brid

izat

ion

Mic

rocy

stis

mcy

D, m

cyE,

mcy

F,

mcy

Gm

cyD

, mcy

E, m

cyF,

mcy

G p

rese

nt in

toxi

c M

icro

cyst

is st

rain

s (3/

5) a

nd in

one

no

ntox

ic st

rain

Nis

hiza

wa

et a

l., 2

000

Dot

blo

t hy

brid

izat

ion

Anab

aena

, M

icro

cyst

is,

Nod

ular

ia, N

osto

c,

Osc

illat

oria

(P

lank

toth

rix)

mcy

B of

Mic

rocy

stis

Toxi

c st

rain

s of s

ever

al g

ener

a de

tect

ed w

ith th

e m

cyB

prob

e; a

ll to

xic

Anab

aena

(5

/12)

, Mic

rocy

stis

(5/7

), N

odul

aria

(3/4

), N

osto

c (1

/3),

Osc

illat

oria

(7/8

) stra

ins,

one

nont

oxic

Aph

aniz

omen

on (1

/4),

Mic

rocy

stis

, and

Osc

illat

oria

stra

in d

etec

ted;

A

naba

ena

prob

e de

tect

ed m

ainl

y th

e he

pato

toxi

c An

abae

na, N

odul

aria

, and

N

osto

c st

rain

s

Nei

lan

et a

l., 1

999

PCR

, se

quen

cing

Anab

aena

, M

icro

cyst

is,

Nod

ular

ia, N

osto

c,

Osc

illat

oria

(P

lank

toth

rix)

mcy

B (d

esig

ned

to

ampl

ify th

e fr

agm

ent

from

a b

road

rang

e of

to

xic

cyan

obac

teria

, al

thou

gh b

ased

on

Mic

rocy

stis

sequ

ence

s)

All

toxi

c An

abae

na (3

/8),

Mic

rocy

stis

(10/

16),

Nod

ular

ia (4

/5),

Nos

toc

(1/5

), O

scill

ator

ia (4

/5) s

train

s, an

d on

e no

ntox

ic L

yngb

ya (1

/1),

and

Mic

rocy

stis

stra

in

dete

cted

; non

toxi

c Ap

hani

zom

enon

(2),

Plec

tone

ma

(1),

Pseu

dana

baen

a (1

), Sy

nech

ococ

cus (

1) st

rain

s, or

a to

xic

Cyl

indr

ospe

rmop

sis s

train

not

det

ecte

d;

ampl

ifi ca

tion

also

with

blo

om sa

mpl

es (6

)

Nei

lan

et a

l., 1

999

PCR

Nod

ular

iaPe

ptid

e sy

nthe

tase

and

po

lyke

tide

synt

hase

Pept

ide

synt

heta

se se

quen

ce sp

ecifi

cally

am

plifi

ed fr

om a

ll to

xic

Nod

ular

ia st

rain

s (1

5) a

nd n

ot fr

om n

onto

xic

Nod

ular

ia (6

), An

abae

na (1

), C

ylin

dros

perm

opsi

s (1)

, Sy

nech

ocys

tis (1

) stra

ins;

pol

yket

ide

synt

hase

sequ

ence

am

plifi

ed w

ith a

ll to

xic

and

a no

ntox

ic N

odul

aria

, and

a to

xic

Mic

rocy

stis

stra

in; a

mpl

ifi ca

tion

with

a

bloo

m sa

mpl

e

Moffi t

t and

Nei

lan,

20

01

PCR

, se

quen

cing

Mic

rocy

stis

mcy

A (N

-m

ethy

ltran

sfer

ase

dom

ain)

mcy

A de

tect

ed in

all

toxi

c M

icro

cyst

is st

rain

s (18

/37)

, in

two

nont

oxic

stra

ins,

and

in to

xic

bloo

m sa

mpl

es (6

/8) f

rom

a p

ond

in A

ustra

lia; o

nly

PCR

pro

duct

s fro

m

stra

ins s

eque

nced

Tille

tt et

al.,

200

1

PCR

, se

quen

cing

Mic

rocy

stis

mcy

B To

xic

Mic

rocy

stis

det

ecte

d by

a n

este

d-PC

R a

ppro

ach

in 3

1% o

f catfi s

h pr

oduc

tion

pond

s in

USA

; lac

k of

det

ectio

n in

som

e po

nds w

ith lo

w a

mou

nt o

f tox

ins;

PC

R

frag

men

ts a

uthe

ntic

ated

by

sequ

enci

ng

Non

nem

an a

nd Z

imba

, 20

02

PCR

, se

quen

cing

Mic

rocy

stis

mcy

BA

ll to

xic

Mic

rocy

stis

stra

ins d

etec

ted

(18/

30);

nont

oxic

Ana

baen

a (2

), Ap

hani

zom

enon

(1),

toxi

c (2

), an

d no

ntox

ic O

scill

ator

ia (2

) stra

ins n

ot d

etec

ted;

m

cyB

also

det

ecte

d in

blo

om a

nd se

dim

ent s

ampl

es fr

om la

kes i

n C

hina

that

sh

owed

toxi

city

with

ELI

SA; d

etec

tion

from

ext

ract

ed D

NA

and

who

le c

ells

; som

e PC

R fr

agm

ents

aut

hent

icat

ed b

y se

quen

cing

Pan

et a

l., 2

002

Tabl

e 1.

Stu

dies

usi

ng g

enes

for m

icro

cyst

in/n

odul

arin

bio

synt

hesi

s to

dete

ct a

nd id

entif

y po

tent

ial t

oxin

pro

duce

rs a

mon

g st

rain

s or i

n en

viro

nmen

tal s

ampl

es.

The

tabl

e al

so in

clud

es o

rigin

al re

sults

of a

rticl

es I

– II

I.

PCR

Mic

rocy

stis

mcy

A (T

illet

t et a

l.,

2001

)D

etec

tion

of to

xic

Mic

rocy

stis

in sa

mpl

es (5

/12)

from

a re

serv

oir i

n A

ustra

liaB

aker

et a

l., 2

002

PCR

Mic

rocy

stis

mcy

B (fi

rst a

deny

latio

n do

mai

n; M

icro

cyst

is-

spec

ifi c)

mcy

B ge

ne d

etec

ted

dire

ctly

in c

olon

ies o

f Mic

rocy

stis

aer

ugin

osa

(73%

) and

M.

icht

hyob

labe

(16%

), bu

t not

in M

. wes

enbe

rgii

Kur

may

er e

t al.,

200

2

PCR

, qPC

R

(Taq

nuc

leas

e as

say)

Mic

rocy

stis

mcy

A, m

cyB

mcy

A de

tect

ed in

toxi

c M

icro

cyst

is st

rain

s (2)

but

not

in n

onto

xic

Mic

rocy

stis

(2)

and

Anab

aena

(2) s

train

s in

conv

entio

nal P

CR

; diff

eren

tiatio

n of

toxi

c (2

) and

no

ntox

ic (1

) Mic

rocy

stis

stra

ins b

y qP

CR

met

hod,

thre

e co

pies

of m

cyA

dete

cted

w

hen

cells

of t

oxic

Mic

rocy

stis

inoc

ulat

ed in

ster

ile w

ater

Foul

ds e

t al.,

200

2

Sout

hern

hy

brid

izat

ion,

PC

R, R

T-PC

R

Mic

rocy

stis

mcy

ABC

regi

onG

enet

ic v

aria

tion

in th

e re

gion

due

to re

com

bina

tion

betw

een

aden

ylat

ion

dom

ains

of

mcy

B1 a

nd m

cyC

; mcy

ABC

det

ecte

d in

all

toxi

c M

icro

cyst

is st

rain

s (10

/18)

, and

on

e no

ntox

ic M

icro

cyst

is st

rain

by

PCR

; mcy

B1 e

xpre

ssed

in th

e sa

me

toxi

c an

d no

ntox

ic st

rain

s

Mik

alse

n et

al.,

200

3

PCR

, res

trict

ion

frag

men

t len

gth

poly

mor

phis

m

(RFL

P)

Anab

aena

, M

icro

cyst

is, N

osto

c,

Plan

ktot

hrix

mcy

A (c

onde

nsat

ion

dom

ain;

gen

eral

pr

imer

s)

mcy

A de

tect

ed o

nly

in to

xic

Anab

aena

(6/8

), M

icro

cyst

is (1

0/24

), N

osto

c (1

/2) a

nd

Plan

ktot

hrix

(8/1

1) st

rain

s, bu

t not

in to

xic

or n

onto

xic

Nod

ular

ia (7

); An

abae

na-,

Mic

rocy

stis

- and

Pla

nkto

thri

x-m

cyA

diffe

rent

iate

d by

RFL

P an

alys

is o

f the

PC

R

prod

uct;

Mic

rocy

stis

- and

Pla

nkto

thri

x-sp

ecifi

c R

FLP

patte

rns i

dent

ifi ed

in tw

o La

ke W

anns

ee (G

erm

any)

sam

ples

His

berg

ues e

t al.,

200

3

qPC

R (T

aq

nucl

ease

ass

ay)

Mic

rocy

stis

mcy

Bm

cyB

geno

type

s of t

oxic

Mic

rocy

stis

qua

ntifi

ed in

env

ironm

enta

l sam

ples

; th

e pr

opor

tion

of m

cyB

geno

type

s var

ied

from

1%

to 3

8% o

f tot

al M

icro

cyst

is

popu

latio

n (P

C-I

GS

geno

type

); de

tect

ion

limit

less

than

5 m

cyB

cells

Kur

may

er a

nd

Kut

zenb

erge

r, 20

03

PCR

, qPC

R

(SY

BR

Gre

en)

Anab

aena

, M

icro

cyst

ism

cyE

(gen

us-s

pecifi c

pr

imer

pai

rs)

In P

CR

, Ana

baen

a-sp

ecifi

c pr

imer

s det

ecte

d m

cyE

only

in to

xic

Anab

aena

stra

ins

(7) a

nd M

icro

cyst

is-s

pecifi c

prim

ers o

nly

in to

xic

Mic

rocy

stis

(10)

stra

ins b

ut

not i

n to

xic

Plan

ktot

hrix

(6) a

nd N

osto

c (1

) stra

ins o

r non

toxi

c An

abae

na (7

), M

icro

cyst

is (3

), or

Pla

nkto

thri

x (2

) stra

ins;

toxi

c An

abae

na- a

nd M

icro

cyst

is-m

cyE

gene

s det

ecte

d an

d qu

antifi

ed

by q

PCR

in tw

o la

kes i

n Fi

nlan

d; d

etec

tion

limit

of

660

mcy

E co

pies

in a

reac

tion

mix

ture

or 4

0 –

400

copi

es m

l-1 la

ke w

ater

Vaito

maa

et a

l., 2

003

PCR

, se

quen

cing

Anab

aena

, M

icro

cyst

is,

Nod

ular

ia, N

osto

c,

Osc

illat

oria

(P

lank

toth

rix)

mcy

B (N

eila

n et

al.,

19

99)

mcy

B de

tect

ed in

18/

60 M

icro

cyst

is st

rain

s iso

late

d fr

om re

serv

oirs

in B

razi

l; co

rrel

atio

n w

ith m

icro

cyst

in d

etec

tion

by E

LISA

; co-

occu

rren

ce o

f tox

ic a

nd

nont

oxic

stra

ins d

etec

ted

Bitt

enco

urt-O

livei

ra,

2003

PCR

Plan

ktot

hrix

mcy

A (c

onde

nsat

ion

dom

ain)

mcy

A de

tect

ed in

49

P. ru

besc

ens s

train

s inc

ludi

ng 8

non

toxi

c st

rain

s, an

d in

9/2

3 P.

aga

rdhi

i stra

ins i

nclu

ding

one

non

toxi

c st

rain

; pro

porti

on o

f ina

ctiv

e m

cyA

geno

type

s in fi l

amen

ts w

as e

stim

ated

as 5

% in

Lak

e Ir

rsee

and

21%

in L

ake

Mon

dsee

(Aus

tria)

Kur

may

er e

t al.,

200

4

PCR

, se

quen

cing

Anab

aena

, M

icro

cyst

is,

Plan

ktot

hrix

, N

osto

c, N

odul

aria

mcy

D (β

-ket

oacy

l sy

ntha

se a

nd

acyl

trans

fera

se

dom

ains

; gen

eral

pr

imer

s), m

cyE

(ade

nyla

tion

and

phos

phop

ante

thei

ne

dom

ains

; gen

eral

pr

imer

s)

Sequ

ence

s det

ecte

d in

all

test

ed to

xic

stra

ins o

f Ana

baen

a (9

, mcy

E; 5

, mcy

D),

Mic

rocy

stis

(10;

5),

Plan

ktot

hrix

(7; 4

), N

osto

c (1

; 1),

and

Nod

ular

ia (3

; 3)

I (R

anta

la e

t al.,

20

04)

PCR

Mic

rocy

stis

mcy

A (H

isbe

rgue

s et

al.,

200

3), m

cyB

(Kur

may

er e

t al.,

200

2)

mcy

A or

mcy

B de

tect

ed in

2.5

% (3

/122

) of n

onto

xic

colo

nies

(fal

se-p

ositi

ves)

and

no

t det

ecte

d in

4.7

% (6

/128

) of t

oxic

col

onie

s (fa

lse-

nega

tives

) col

lect

ed fr

om 1

3 la

kes i

n Eu

rope

; mcy

gen

es d

etec

ted

in m

ost c

olon

ies o

f Mic

rocy

stis

aer

ugin

osa

and

M. b

otry

s (>7

5%),

in so

me

colo

nies

of

M. i

chth

yobl

abe

and

M. v

irid

is

(≤20

%),

but n

ot in

col

onie

s of M

. wes

enbe

rgii

Via

-Ord

orik

a et

al.,

20

04

PCR

Mic

rocy

stis

mcy

A (fi

rst a

deny

latio

n do

mai

n)m

cyA

dete

cted

in a

ll to

xic

Mic

rocy

stis

stra

ins (

12/6

1); m

cyA

dete

cted

and

qu

antifi

ed

in 3

/5 w

ater

sam

ples

from

Lak

e M

ikat

a (J

apan

) with

com

petit

ive

PCR

; de

tect

ion

limit

of 1

00 c

ells

ml-1

Yosh

ida

et a

l., 2

005

PCR

, se

quen

cing

Mic

rocy

stis

mcy

A (T

illet

t et

al.,

2001

), m

cyB

(Mik

alse

n et

al.,

200

3),

mcy

C (a

deny

latio

n do

mai

n), m

cyD

(acy

l ca

rrie

r pro

tein

and

β-

keto

acyl

synt

hase

do

mai

ns),

mcy

E (G

SA-

amin

otra

nsfe

rase

do

mai

n), m

cyG

(C-

met

hyltr

ansf

eras

e do

mai

n)

All

gene

s wer

e de

tect

ed o

nly

in to

xic

Mic

rocy

stis

stra

ins (

5/9)

; 2-3

mcy

gen

es

dete

cted

by

mul

tiple

x PC

R in

isol

ated

who

le-c

ell c

olon

ies;

PC

R fr

agm

ents

from

th

e st

rain

s aut

hent

icat

ed b

y se

quen

cing

Oua

hid

et a

l., 2

005

PCR

Mic

rocy

stis

mcy

B (N

onne

man

and

Zi

mba

, 200

2), m

cyD

(K

aebe

rnic

k et

al.,

20

00)

mcy

B de

tect

ed in

88%

and

mcy

D in

79%

of t

he L

ake

One

ida

(USA

) sam

ples

from

m

id-J

une

to O

ctob

er; p

oten

tial l

ack

of d

etec

tion

of to

xic

Anab

aena

sugg

este

dH

otto

et a

l., 2

005

PCR

, se

quen

cing

Plan

ktot

hrix

mcy

T, m

cyE,

mcy

A,

mcy

B; in

terg

enic

re

gion

s: m

cyTD

, m

cyEG

, mcy

HA,

mcy

CJ

mcy

B, m

cyE,

mcy

CJ,

and

mcy

HA

dete

cted

in a

ll to

xic

Plan

ktot

hrix

stra

ins (

23/4

7 or

22/

46) s

tudi

ed; m

cyA

dete

cted

14/

23 to

xic

stra

ins;

mcy

T an

d m

cyTD

als

o de

tect

ed in

19/

24 a

nd 9

/24

nont

oxic

stra

ins;

mcy

EG a

nd m

cyTD

show

ed le

ngth

va

riatio

n; m

cyE

used

to d

etec

t tox

ic P

lank

toth

rix

spec

ifi ca

lly in

sam

ples

from

two

lake

s in

Ger

man

y

Mbe

di e

t al.,

200

5

Met

hod

Gen

us/g

ener

a ta

rget

ed b

y th

e pr

imer

s/pr

obes

Gen

e/ge

nes t

hat

prim

ers o

r pr

obes

are

de

sign

ed to

det

ecta

Mai

n re

sults

reg

ardi

ng d

iffer

entia

tion

of to

xic

and

nont

oxic

stra

ins a

nd

dete

ctio

n/id

entifi

cat

ion

of to

xic

stra

ins

Ref

eren

ce

PCR

, qPC

R

(Taq

nuc

leas

e as

say)

Mic

rocy

stis

mcy

B (N

onne

man

and

Zi

mba

, 200

2), m

cyD

(K

aebe

rnic

k et

al.,

20

00)

mcy

B, m

cyD

det

ecte

d by

mul

tiple

x-PC

R in

Lak

e Er

ie (U

SA) b

loom

sam

ples

and

th

e ab

unda

nce

of to

xic

Mic

rocy

stis

est

imat

ed b

y m

cyD

cop

y nu

mbe

rs; d

etec

tion

limit

of 2

5 m

cyD

cop

ies i

n a

reac

tion/

5000

cop

ies l

-1 la

ke w

ater

; res

ults

indi

cate

a

pote

ntia

l lac

k of

det

ectio

n of

oth

er to

xic

cyan

obac

teria

l gen

era:

Ana

baen

a,

Plan

ktot

hrix

Rin

ta-K

anto

et a

l.,

2005

PCR

Mic

rocy

stis

ae

rugi

nosa

mcy

A (T

illet

t et a

l.,

2001

)m

cyA

dete

cted

in 9

/18

Mic

rocy

stis

stra

ins i

sola

ted

from

a ri

ver i

n Po

rtuga

l tha

t sh

owed

pre

senc

e of

mic

rocy

stin

s Sa

ker e

t al.,

200

5a

PCR

, se

quen

cing

Hep

atot

oxic

cy

anob

acte

ria;

Mic

rocy

stis

mcy

E (J

ungb

lut a

nd

Nei

lan,

200

6); m

cyA

With

bot

h pr

imer

pai

rs m

cy se

quen

ces s

pecifi c

to M

icro

cyst

is w

ere

foun

d in

12

food

supp

lem

ent s

ampl

es su

ppos

ed to

con

tain

non

toxi

c cy

anob

acte

rium

Ap

hani

zom

enon

fl os

-aqu

ae

Sake

r et a

l., 2

005b

PCR

Mic

rocy

stis

mcy

A (S

aker

et a

l.,

2005

b)m

cyA

dete

cted

in is

olat

es fr

om 1

2/14

lake

s in

Mic

higa

n (U

SA);

both

toxi

c an

d no

ntox

ic is

olat

es fr

om fo

ur la

kes

Wils

on e

t al.,

200

5

PCR

, se

quen

cing

Mic

rocy

stis

mcy

B (N

onne

man

and

Zi

mba

, 200

2), m

cyD

(K

aebe

rnic

k et

al.,

20

00)

mcy

B, m

cyD

det

ecte

d by

mul

tiple

x-PC

R in

toxi

c M

icro

cyst

is st

rain

s and

Lak

e Er

ie (U

SA) w

ater

sam

ples

; am

plifi

catio

n pr

oduc

ts n

ot d

etec

ted

in A

naba

ena

(4),

Plan

ktot

hrix

(2),

Osc

illat

oria

(1),

Nos

toc

(1),

Nod

ular

ia (1

), Pl

ecto

nem

a (1

), Sy

nech

ocys

tis (1

), Sy

nech

ococ

cus(

1) st

rain

s, an

d in

non

toxi

c M

icro

cyst

is (1

); m

cyB

and

mcy

D a

mpl

icon

s fro

m tw

o la

ke sa

mpl

es se

quen

ced

Oue

llette

et a

l 200

6

PCR

, se

quen

cing

An

abae

na,

Mic

rocy

stis

, N

odul

aria

, Nos

toc,

O

scill

ator

ia

(Pla

nkto

thri

x)

mcy

A (T

illet

t et a

l.,

2001

), m

cyB

(Nei

lan

et

al.,

1999

)

mcy

A an

d m

cyB

dete

cted

in tw

o is

olat

ed M

icro

cyst

is st

rain

s and

a b

loom

sam

ple

from

a la

ke in

Rom

ania

; det

ectio

n of

gen

es c

orre

late

d w

ith m

icro

cyst

in p

rodu

ctio

n B

oaru

et a

l., 2

006

qPC

R (S

YB

R

Gre

en)

Mic

rocy

stis

mcy

A (T

illet

t et a

l.,

2001

; new

reve

rse

prim

er d

esig

ned

for

shor

ter a

mpl

icon

)

mcy

A de

tect

ed in

all

toxi

c M

icro

cyst

is st

rain

s (9/

21);

10-f

old

low

er d

etec

tion

limit

with

prim

ers p

rodu

cing

an

ampl

icon

of 1

90 b

p in

size

com

pare

d w

ith p

rimer

s with

an

am

plic

on o

f 136

9 bp

; mcy

A al

so d

etec

ted

and

quan

tifi e

d in

sam

ples

from

two

lake

s in

Japa

n; d

etec

tion

limit

of 8

.8 c

ells

per

reac

tion

Furu

kaw

a et

al.,

200

6

PCR

Anab

aena

, M

icro

cyst

is,

Plan

ktot

hrix

, N

osto

c, N

odul

aria

(n

ot m

cyA)

mcy

A (H

isbe

rgue

s et

al.,

200

3), m

cyB

(Nei

lan

et a

l., 1

999)

, m

cyD

, mcy

E (I

)

All

gene

s det

ecte

d in

thre

e re

fere

nce

stra

ins (

one

toxi

c M

icro

cyst

is a

nd tw

o to

xic

Plan

ktot

hrix

); m

cyE

dete

cted

in a

ll (2

5), m

cyB

in 2

4/25

, mcy

D in

21/

25, a

nd

mcy

A in

20/

25 w

ater

sam

ples

from

thre

e la

kes i

n Po

lish;

no

post

-PC

R a

naly

sis f

or

prod

ucer

gen

us id

entifi

cat

ion

Man

kiew

icz-

Boc

zek

et

al.,

2006

PCR

, se

quen

cing

Anab

aena

, M

icro

cyst

is,

Plan

ktot

hrix

, N

osto

c, N

odul

aria

, Ph

orm

idiu

m

mcy

E/nd

aF

(am

inot

rans

fera

se

dom

ain;

gen

eral

pr

imer

s)

mcy

E/nd

aF d

etec

ted

in a

ll to

xic

Mic

rocy

stis

(6/1

1), A

naba

ena

(1/2

), N

odul

aria

(4

/5),

Nos

toc

(1/3

), Pl

ankt

othr

ix (2

/4),

and

Phor

mid

ium

(3/5

) stra

ins;

not

det

ecte

d in

non

toxi

c C

ylin

dros

perm

opsi

s (1)

, Syn

echo

cyst

is (1

), Le

ptol

yngb

ya (1

), Ly

ngby

a (3

), O

scill

ator

ia (2

), Sy

mpl

oca

(1),

or S

piru

lina

(1);

gene

s als

o de

tect

ed in

blo

om

sam

ples

(4) f

rom

Aus

tralia

and

Ital

y

Jung

blut

and

Nei

lan,

20

06

PCR

, se

quen

cing

Anab

aena

, M

icro

cyst

is, N

osto

c,

Plan

ktot

hrix

mcy

A (H

isbe

rgue

s et

al.,

2003

)To

xic

Mic

rocy

stis

- and

Pla

nkto

thri

x-re

late

d se

quen

ces o

btai

ned

from

Lak

e Er

ie (U

SA);

Mic

rocy

stis

-rel

ated

sequ

ence

s fro

m th

ree

sam

plin

g st

atio

ns, a

nd

Plan

ktot

hrix

-rel

ated

from

one

stat

ion

Rin

ta-K

anto

and

W

ilhel

m, 2

006

PCR

Anab

aena

, M

icro

cyst

is,

Plan

ktot

hrix

mcy

E (g

enus

-spe

cifi c

pr

imer

pai

rs; V

aito

maa

et

al.,

200

3; II

)

Spec

ifi c

dete

ctio

n of

toxi

c Pl

ankt

othr

ix st

rain

s am

ong

63 to

xic

and

nont

oxic

An

abae

na (9

/13)

, Aph

aniz

omen

on (0

/3),

Hap

alos

ipho

n (1

/0),

Lim

noth

rix

(0/2

), M

icro

cyst

is (9

/4),

Nod

ular

ia (6

/1),

Nos

toc

(1/1

), Ph

orm

idiu

m (1

/0),

Plan

ktot

hrix

(7

/4),

and

Syne

choc

occu

s (0/

1) st

rain

s tes

ted;

mic

rocy

stin

pro

duce

rs p

rese

nt in

91

% o

f the

70

lake

sam

ples

; sim

ulta

neou

s occ

urre

nce

of tw

o or

thre

e pr

oduc

er

gene

ra in

54%

of t

he la

ke sa

mpl

es

II (R

anta

la e

t al.,

20

06)

PCR

Anab

aena

, M

icro

cyst

is,

Nod

ular

ia, N

osto

c,

Plan

ktot

hrix

(O

scill

ator

ia)

mcy

B (N

eila

n et

al.,

19

99)

mcy

B de

tect

ed in

a sa

mpl

e fr

om a

rese

rvoi

r in

Bra

zil t

hat c

onta

ined

Mic

rocy

stis

an

d P.

aga

rdhi

i, an

d Ra

dioc

ystis

; pot

entia

l pro

duce

r org

anis

m n

ot id

entifi

ed

Anj

os e

t al.,

200

6

qPC

R (S

YB

R

Gre

en)

Mic

rocy

stis

mcy

AD

etec

tion

limit

of 2

5 m

cyA

copi

es; p

ropo

rtion

of m

cyA

geno

type

of t

otal

M

icro

cyst

is (P

C-I

GS)

var

ied

from

0.5

% to

35%

in L

ake

Mik

ata

(Jap

an) d

urin

g a

14-m

onth

stud

y pe

riod;

rela

tive

abun

danc

e of

mcy

A co

rrel

ated

pos

itive

ly w

ith

nitra

te c

once

ntra

tion,

no

corr

elat

ion

with

tem

pera

ture

and

orth

o-ph

osph

ate

Yosh

ida

et a

l., 2

007

PCR

, se

quen

cing

Anab

aena

, M

icro

cyst

is, N

osto

c,

Plan

ktot

hrix

mcy

A (H

isbe

rgue

s et

al.,

2003

)m

cyA

dete

cted

toge

ther

with

Mic

rocy

stis

16S

rRN

A g

ene

and

cyan

obac

teria

l PC

-IG

S fr

agm

ent i

n m

ultip

lex

PCR

in d

ieta

ry su

pple

men

ts p

rodu

ced

from

Ap

hani

zom

enon

fl os

-aqu

ae (1

2), b

ut n

ot in

thos

e pr

oduc

ed fr

om S

piru

lina

(12)

or

Nos

toc fl a

gelli

form

e (4

); se

quen

ces r

elat

ed to

Mic

rocy

stis

mcy

A or

Pla

nkto

thri

x m

cyA

Sake

r et a

l., 2

007b

PCR

, RFL

PAn

abae

na,

Mic

rocy

stis

, Nos

toc,

Pl

ankt

othr

ix

mcy

A (H

isbe

rgue

s et

al.,

2003

)m

cyA

pres

ent i

n 15

/26

wat

er sa

mpl

es fr

om la

kes a

nd ri

vers

in P

ortu

gal;

prod

ucer

or

gani

sm: M

icro

cyst

is (1

4) o

r Pla

nkto

thri

x (1

) ide

ntifi

ed w

ith R

FLP

of a

mpl

ifi ed

m

cyA

Sake

r et a

l., 2

007a

PCR

, qPC

R

(SY

BR

Gre

en),

sequ

enci

ng

Nod

ular

iand

aFPr

imer

s spe

cifi c

for t

oxic

Nod

ular

ia a

mon

g 63

toxi

c an

d no

ntox

ic A

naba

ena

(10/

3), A

phan

izom

enon

(0/1

), M

icro

cyst

is (9

/1),

Nod

ular

ia (2

0/12

), N

osto

c (1

/0),

Osc

illat

oria

(0/1

), an

d Pl

ankt

othr

ix (5

/0) s

train

s tes

ted

in P

CR

and

am

ong

20

stra

ins t

este

d in

qPC

R; d

etec

tion

limit

of 3

0 nd

aF c

opie

s ml-1

of B

altic

Sea

wat

er;

ndaF

cop

y nu

mbe

rs v

arie

d be

twee

n 30

and

45

000

copi

es m

l-1of

wat

er; s

trong

co

rrel

atio

n be

twee

n nd

aF c

opy

num

bers

and

nod

ular

in c

once

ntra

tion;

sequ

enci

ng

used

to c

onfi r

m sp

ecifi

city

Kos

kenn

iem

i et a

l.,

2007

PCR

, RT-

PCR

, se

quen

cing

Anab

aena

, M

icro

cyst

ism

cyE

(mod

ifi ed

from

Va

itom

aa e

t al.,

200

3)M

icro

cyst

is a

nd n

ot A

naba

ena

foun

d re

spon

sibl

e fo

r mic

rocy

stin

pro

duct

ion

in

Lake

Aga

wam

(USA

) by

mcy

E PC

R; e

xpre

ssio

n of

the

Mic

rocy

stis

-mcy

E ge

ne

was

freq

uent

in sa

mpl

es fr

om M

ay to

Aug

ust a

nd ra

re in

sam

ples

from

Sep

tem

ber

and

Oct

ober

; int

erac

tive

effe

cts o

f zoo

plan

kton

gra

zing

, nut

rient

load

ing

and

mcy

E ge

ne e

xpre

ssio

n on

blo

om d

ynam

ics

Gob

ler e

t al.,

200

7

Met

hod

Gen

us/g

ener

a ta

rget

ed b

y th

e pr

imer

s/pr

obes

Gen

e/ge

nes t

hat

prim

ers o

r pr

obes

are

de

sign

ed to

det

ecta

Mai

n re

sults

reg

ardi

ng d

iffer

entia

tion

of to

xic

and

nont

oxic

stra

ins a

nd

dete

ctio

n/id

entifi

cat

ion

of to

xic

stra

ins

Ref

eren

ce

qPC

R (T

aq

nucl

ease

ass

ay)

Mic

rocy

stis

mcy

B (K

urm

ayer

and

K

utze

nber

ger,

2003

)R

epro

duci

ble

resu

lts o

btai

ned

whe

n pr

opor

tion

of m

cyB

geno

type

s of t

otal

M

icro

cyst

is (P

C-I

GS)

in la

ke sa

mpl

es e

stim

ated

with

thre

e di

ffere

nt in

stru

men

ts in

th

ree

diffe

rent

labo

rato

ries

Scho

ber e

t al.,

200

7

PCR

, se

quen

cing

Mic

rocy

stis

, An

abae

na,

Plan

ktot

hrix

mcy

B (N

onne

man

an

d Zi

mba

, 200

2),

mcy

D (K

aebe

rnic

k et

al.,

200

0), m

cyA

(His

berg

ues e

t al.,

20

03),

mcy

E (V

aito

maa

et

al.,

200

3; II

)

Mic

rocy

stis

-spe

cifi c

mcy

B an

d m

cyD

prim

ers u

sed

for i

nitia

l scr

eeni

ng o

f sam

ple;

se

quen

cing

of m

cyA

PCR

pro

duct

, and

Ana

baen

a-, M

icro

cyst

is-,

and

Plan

ktot

hrix

-sp

ecifi

c m

cyE

prim

ers i

dent

ifi ed

Mic

rocy

stis

as t

he d

omin

ant m

icro

cyst

in p

rodu

cer

in L

ake

Ont

ario

(USA

) sta

tions

Hot

to e

t al.,

200

7

DN

A-c

hip,

qP

CR

(SY

BR

G

reen

)

Anab

aena

, M

icro

cyst

is,

Nod

ular

ia, N

osto

c,

Plan

ktot

hrix

mcy

E/nd

aF (g

ener

al

prim

ers:

I, g

enus

-sp

ecifi

c pr

imer

s:

Vaito

maa

et a

l., 2

003

and

II)

mcy

E/nd

aF re

gion

am

plifi

ed fr

om to

xic

cyan

obac

teria

l stra

ins (

Anab

aena

, M

icro

cyst

is, N

odul

aria

, Nos

toc,

and

Pla

nkto

thri

x ) i

dent

ifi ed

with

gen

us-s

pecifi c

pr

obe

pairs

by

DN

A-c

hip

anal

ysis

; 1 –

5 fm

ol o

f PC

R p

rodu

ct e

noug

h fo

r det

ectio

n by

the

prob

e pa

irs; D

NA

-chi

p re

sults

of s

ix la

kes i

n Fi

nlan

d va

lidat

ed w

ith q

PCR

ex

perim

ents

with

gen

us-s

pecifi c

prim

er p

airs

III (

Ran

tala

et a

l.,

2007

)

a or

igin

al re

fere

nces

for t

he p

rimer

s/pr

obes

are

pro

vide

d in

par

enth

eses

if th

ey d

iffer

from

the

refe

renc

e of

the

stud

y

22

2. AIMS OF THE STUDY

The main producers of microcystins include strains of the genera Microcystis, Anabaena, and Planktothrix, while nodularins are produced by strains of Nodularia. However, the toxic and nontoxic strains of these genera cannot be differentiated by conventional microscopy. Molecular methods based on the toxin biosynthetic genes are instead capable of detecting potential microcystin and nodularin producers. The fi rst aim of this study was to produce additional/novel sequence data on microcystin and nodularin biosynthetic genes, since publicly available data were restricted to a few Microcystis- mcy sequences when this study was initiated. Further aims were built upon the sequence data generated. One aim was to elucidate how microcystin production evolved in the cyanobacteria. Previously, two hypotheses had been presented to explain the observed sporadic distribution of toxin production: HGT of the biosynthetic genes or vertical inheritance combined with gene losses in some of the cyanobacterial lineages. Another aim was to develop detection methods based on the mcyE/ndaF genes capable of identifying all the main producers of microcystins and nodularins in environmental samples and associating them with environmental variables. This could be achieved only with molecular detection methods based on the biosynthetic genes.

The specifi c aims of this study were:

1. To investigate the evolutionary history of microcystin biosynthesis (I).

2. To explore the frequency and distribution of microcystin producers in lakes in Finland based on the presence of the mcyE gene (II).

3. To investigate the correlation between environmental factors and the occurrence of microcystin producers in lakes in Finland (II).

4. To develop a high-throughput method, a DNA-chip assay, for detection of microcystin-producing cyanobacteria in environmental samples (III).

Aims of the Study

23

3.1 Strains and environmental samples

Cyanobacterial strains maintained in the culture collection of K. Sivonen, University of Helsinki, Finland, were used throughout this study. DNAs extracted from these strains were utilized in the specifi city testing of the primers designed (I, II), as source material for sequencing (I), specifi city and sensitivity testing of DNA-chip probes (III), and construction of standards for qPCR applications (III). The strains used are listed in the respective articles.

Environmental water samples were collected from 70 lakes located in southern and central Finland. Oligotrophic, mesotrophic, eutrophic, and hypertrophic waters were represented among the lakes sampled. Composite

Table 2. Methods that were used in the study. The Roman numerals refer to the original articles in which the methods were applied.

Method ArticleSampling IIDetermination of physicochemical parameters from water samples IIDNA extraction from cyanobacterial strains or water samples I, IIPrimer design I, IIProbe design for LDR IIIPCR amplifi cation I, II, IIImcyD/mcyE gene detection by PCR

General primers I, IIGenus-specifi c primers (mcyE) for Anabaena, Microcystis, and Planktothrix II

mcyE-gene detection by quantitative real-time PCR SYBR Green: Genus-specifi c primers for Anabaena, Microcystis, and Planktothrix III

mcyE gene detection by the DNA-chip IIICloning ISequencing IPhylogenetic sequence analysis (ME, MP, ML) IToxin analysis by ELISA IIImage analysis of agarose gels IIOrdination (PCA), correlation, and regression analyses II

Materials and Methods

3. MATERIALS AND METHODS

water samples were taken from depths of 0-2 m. DNAs extracted from the water samples were used to study the frequency and composition of potential microcystin producers in the samples with conventional PCR (II). A subset of the lake samples was analyzed additionally with the DNA-chip and qPCR (III). The DNA-chip was furthermore applied to Baltic Sea samples. The environmental samples studied are identifi ed in the original articles (II, III).

3.2 Methods

3.2.1 Methods described in publicationsThe techniques and methods used in this study are listed in Table 2. A more detailed description of each method is provided in the original articles (I – III).

24

3.2.2 Additional optimization of the DNA-chip assayAdditional optimization of the DNA-chip assay was performed to determine the possibility and extent of preferential amplification in the PCR prior to LDR. Suspicion of potential preferential amplifi cation by the general mcyE primers was raised due to the slightly lower sensitivity of the DNA-chip (compared with qPCR) in detecting those potential microcystin producers that by qPCR were less abundant in lake samples (III). Mixes of microcystin-producing Microcystis PCC 7941, Anabaena sp. 90, and Planktothrix sp. 49, and mixes of nodularin-producing Nodularia sp. F8-1 and microcystin-producing Anabaena sp. 318 were used to represent toxic strains possibly co-occurring in lakes and in the Baltic Sea, respectively. Amplifi cation was performed with the general primer pairs mcyE-F2/R4 and mcyE-F2b/R4. Since mcyE-F2 (5’-GAAATTTGTGTAGAAGGTGC-3’) includes a 3’ mismatch to the ndaF sequence, mcyE-F2b (5’-TGAAATTTGTGTAGAAGGTG-3’) was designed to anneal one base upstream, where perfect base pairing with the primer and both mcyE and ndaF sequences was possible.

The PCR amplifi cations of the mixes (total of 20 ng DNA) were performed in 1 x DynaZyme II PCR buffer (Finnzymes) with 0.5 U of DyNAzyme II DNA polymerase (Finnzymes), 250 μM dNTPs (Finnzymes), and 0.5 μM of primers (Sigma-Genosys Ltd.) in a 40-μl final volume. The following protocol was used for the reactions: 95°C, 3 min; 35 x (94°C,

Materials and Methods

30 s; 53-57°C, 30 s; 72°C, 1 min); 72°C, 10 min. The PCR products were purifi ed with the GFX PCR DNA purification kit (Amersham) or the E.Z.N.A. Cycle-Pure kit (Omega) according to the manufacturers’ instructions. The size and concentration of the PCR products were analyzed with an agarose (1.5%) gel run and BioPhotometer (Eppendorf). The original annealing temperatures were 56°C for mcyE-F2/R4 and 57°C for mcyE-F2b/R4. Further experiments at temperatures 55°C and 53°C were carried out to determine the conditions for more equal amplifi cation.

LDRs with 25-50 fmol of mcyE PCR products, hybridization, washing, and scanning of arrays were performed as described in the article III. The scanned images were used to calculate the background-subtracted median signal intensities for the spots by GenePix Pro 5.1 microarray acquisition and analysis software (Axon Instruments). The threshold levels for detection were determined by calculating the average and standard deviation of signal intensities of blank spots (spots with no printed ZipCodes). The stringency of detection was increased by adding the standard deviation value two or fi ve times to the average signal intensity value. Correct detection of mcyE and ndaF sequences by the DNA-chip was evaluated with signals exceeding either threshold level, with the exception of the signals from spots corresponding to Planktothrix-mcyE probe pair I, which were excluded from the analysis.

25

4. RESULTS

4.1 Design of general and genus-specifi c primers for microcystin synthetase genes

4.1.1 General primers (I)Primer pairs that amplifi ed fragments of the microcystin synthetase genes from the most important microcystin producers were designed for the conservative regions of the mcyD (mcyD-F/R) and mcyE (mcyE-F2/R4) genes. These regions were located by comparing the mcyD and mcyE sequences of Anabaena and Microcystis strains

available at the time. The amplification range was then examined with both microcystin- and nodularin-producing and nontoxic strains. The mcyD-F/R and mcyE-F2/R4 primer pairs amplifi ed regions of approximately 820 bp and 810 bp, respectively, from the microcystin-producing Anabaena , Microcystis , Planktothrix, and Nostoc strains, and from the nodularin-producing Nodularia strains (Table 3). No amplifi cation products were detected with any of the nontoxic strains tested. This specificity of the general

Table 3. Results of specificity testing of the primer pair mcyE-F2/R4 with cyanobacterial strains.Strain mcyE-F2/R4b Microcystin/nodularin

productionc

Anabaena sp. strains90a, 202A1a, 202A2a, NIVA-CYA 83/1a, 66Aa, 299Ba, PH256, 315, 318

+ microcystin

86, 123, 14, 37, IC-1, PH133, 277, 299A, PCC 6309, PCC 7108, PCC 73105, PCC 9208

- -

Microcystis sp. strains205a, NIES 102a, NIES 89a, PCC 7806a, PCC 7941a, 98, GL260735, GL280646, Izancya M5, Izancya 25

+ microcystin

130, GL060916 - -269 (+) -Planktothrix sp. strains49a, NIVA-CYA 126/8a, NIVA-CYA 127a, NIVA-CYA 128/Ra, 97, 213, 226

+ microcystin

18, 45, 214, PCC 6304 - -Nostoc sp. strains152a, IO-102.1a + microcystin159 - -Nodularia sp. strainsHEM a, F8-1a, BY1a, AV3, GR8b, PCC 7804 + nodularinHKVV - -Other strainsAphanizomenon TRI 183, 202, PCC 7905 - -Limnothrix 007-A, 165-C - -Synechococcus GL150636 - -

a mcyD amplifi ed with the primer pair mcyD-F/Rb +, amplifi cation; (+), weak amplifi cation; -, no amplifi cationc -, no microcystin or nodularin produced

Results

26

primer pairs allowed differentiation of hepatotoxic and nonhepatotoxic strains as well as detection of potential hepatotoxin producers in environmental samples. In addition, sequencing of the amplifi cation products yielded novel sequence data of the biosynthetic genes of Planktothrix, Nostoc, and Nodularia. The same wide amplification range of microcystin and nodularin producers and specificity of amplification were detected when the modifi ed forward primer, mcyE-F2b, was used instead of mcyE-F2.

4.1.2 Genus-specifi c primers (II, III)Primers specific for Anabaena- and Microcystis-mcyE were designed and their specifi city tested already in a former study (Vaitomaa et al., 2003). In the present study, a Planktothrix-specifi c primer was designed in the same way as the other two primers with the help of sequence data generated with general mcyE primers. All genus-specifi c primers were used with the primer mcyE-F2. The specificity of the primer pair was tested with cyanobacterial strains. The primer pair amplified specifically the gene region only from the microcystin-producing Planktothrix strains, but not from hepatotoxin-producing Anabaena , Microcystis , Nostoc, Nodularia, Hapalosiphon and Phormidium strains or from nontoxic strains (II: Table 2). The use of this primer pair allowed identifi cation of one of the most important microcystin-producing cyanobacterial genera in freshwaters (II: Fig. 1). In addition to conventional PCR, this primer pair was optimized for use in qPCR applications, where it could be used to quantify the Planktothrix-specifi c mcyE gene copies present in the samples (III: Fig. 4).

4.2. Evolution of microcystin synthetase genes in cyanobacteria (I)

Sequences of the biosynthetic genes (mcyA , mcyD , and mcyE ) f rom representative strains of four microcystin-producing genera (Anabaena, Microcystis, Nostoc, and Planktothrix) and one nodularin-producing genus (Nodularia) were used to study the evolution of the mcy genes in cyanobacteria. Comparison of the phylogenetic trees constructed with biosynthetic gene sequences (mcyA, mcyD, mcyE) and housekeeping gene sequences (16S rRNA and rpoC1) revealed perfectly congruent trees with all construction methods used (ME, MP, ML) (I: Fig. 2). The tree topologies were well-supported in bootstrap analyses and there was no support for incongruence in a partition homogeneity test. In addition, the sequence divergence rate of the mcy gene set was comparable to that of the housekeeping gene set. These results indicated that the mcy genes had coevolved with housekeeping genes during the entire evolutionary history of microcystins. The results also showed that microcystins are of ancient origin and were present in the last common ancestor of a large number of cyanobacteria, after which the ability to produce microcystins has been lost repeatedly in some of the more derived lineages of cyanobacteria. In contrast, HGT as an explanation for the sporadic distribution of mcy genes among modern cyanobacteria was not supported by the results.

The congruence of the trees also suggested that the nda genes were derived from the mcy genes and hence nodularin could be considered as a variant of microcystins. Based on the structural

Results

27

differences of the two molecules, a hypothetical evolutionary scheme was proposed to explain the differences at the gene level (I: Fig. 1B). A partial deletion of the genes mcyA and mcyB, which encode module 2 of McyA and module 1 of McyB that incorporate the amino acids D-alanine1 and X2 into microcystins, would lead to formation of a pentapeptide instead of a heptapeptide. An additional mutation in module 1 of mcyA with a resulting change in substrate specifi city would lead to the replacement of Mdha in microcystins with Mdhb found in nodularins.

4.3 Frequency and composition of microcystin producers in lakes in Finland (II)

The frequency of the potential microcystin producers in 70 lakes in Finland was studied with both general and genus-

specific mcyE primers in conventional PCR. With general mcyE primers, 84% (59/70) of the lake samples contained potential microcystin producers. With the three genus-specific primer pairs, all of which amplify a fragment of about 250 bp, potential microcystin producers were detected in 91% (66/70) of the samples. Microcystis-specific mcyE was detected in 70% (49), Planktothrix-specifi c mcyE in 63% (44), and Anabaena-specific mcyE in 37% (26) of these lake samples, suggesting that at the time of sampling (mainly in mid-July) Microcystis was the most common microcystin producer in these lakes, followed by Planktothrix and Anabaena (II: Fig. 1).

The composition of potential microcystin producer populations in the lake samples was studied with the three genus-specifi c primer pairs. The presence of only one potential microcystin-

Figure 4. Proportion of lakes with different combinations of potential microcystin producers based on the presence of genus-specifi c mcyE genes in oligotrophic (total phosphorus (TP), < 10 μg l-1), mesotrophic (TP, 10-34 μg l-1), eutrophic (TP, 35-100 μg l-1), and hypertrophic (TP, > 100 μg l-1) lakes.

Oligotrophic (n = 19) Mesotrophic (n = 30)

Eutrophic (n = 17) Hypertrophic (n = 4)

no MC producersonly Microcystis

only Planktothrix

Microcystis and Anabaena

Microcystis and Planktothrix

Anabaena and Planktothrix

Microcystis, Anabaena andPlanktothrix

Results

28

producing genus was indicated in 37% (26/70), two genera in 30% (21/70), and all three producer genera in 24% (17/70) of the samples (II: Fig. 1). Thus, potential microcystin producers co-occurred in over half of the lakes. Among oligotrophic lakes (total phosphorus (TP) content: < 10 μg l-1), lakes with one microcystin-producing genus, Microcystis or Planktothrix, were most prevalent. Mesotrophic lakes (TP: 10-34 μg l-1) had the highest number of different producer combinations, with occurrence of Microcystis and Planktothrix together being the most common. In eutrophic (TP: 35-100 μg l-1) and hypertrophic (TP: > 100 μg l-1) lakes, the most widespread combination was the co-occurrence of Anabaena, Microcystis, and Planktothrix (Fig. 4). The proportion of three-producer lakes increased clearly under more TP-rich conditions from approximately 5% in oligotrophic to 75% in hypertrophic lakes. In addition, lakes that contained no microcystin producers were either oligotrophic or mesotrophic, while eutrophic and hypertrophic lakes always had potential microcystin producers.

4.4 Association of environmental variables with the occurrence of microcystin producers and microcystin concentration (II)

First, the lake samples (n = 58) were grouped, based on the Anabaena-, Microcystis, and Planktothrix-specific mcyE PCR results with principal component analysis (PCA) (Fig. 5). Seven groups were formed: lakes with no microcystin producers, lakes with only Microcystis, lakes with only Planktothrix, lakes with Microcystis and Planktothrix, lakes with Anabaena and Microcystis,

lakes with Anabaena and Planktothrix, and lakes with Anabaena, Microcystis, and Planktothrix. PC axis 1 separated lakes with no or one producer from lakes with several producers, while PC axis 2 separated lakes with Planktothrix from those with no Planktothrix. Second, correlation analysis was used to determine the correlations between the PC axes and the environmental variables (II: Table 2). The phytoplankton and cyanobacterial biomasses, microcystin concentration of the lake samples, pH of lake water, total nitrogen (TN) and chlorophyll-a (chl-a) concentrations, Microcystis, Aphanizomenon, and Oscillatoriales biomasses were signifi cantly (P < 0.05) correlated with PC axis 1. This suggested that the co-occurrence of multiple microcystin-producing genera was associated with greater cyanobacterial and phytoplankton biomasses, higher microcystin concentration, a more alkaline pH, and higher concentrations of TN and chl-a (Fig. 5). In addition, the correlations of TP (P = 0.06) and dissolved inorganic phosphorus (DIP) (P = 0.054) with PC axis 1 nearly attained the level of statistical significance (P < 0.05). The location and surface area of the lakes, TN/TP ratio, water color, Secchi depth, and Anabaena and Planktothrix biomasses did not signifi cantly correlate with PC axis 1 and hence were most likely not associated with the presence of multiple microcystin producers. None of the environmental variables correlated significantly with PC axis 2. The Secchi depth had the most signifi cant correlation (P = 0.095), implying that microcystin-producing Planktothrix was more commonly present in lakes with greater transparency (Fig. 5).

Results

29

Regression analyses were used to determine whether environmental variables could explain the presence of mcyE genes and microcystin concentration in the lake samples. A logistic regression model was used with a binomial response variable, the presence/absence of mcyE genes, and a linear regression model for

the analysis of a continuous-response variable, microcystin concentration. The models suggested that the occurrence probability of the mcyE genes rises with a higher pH, TN concentration, and greater surface area of the lake (II: Table 3, Fig. 4), and microcystin concentration with higher TP concentration (II: Table

Figure 5. PCA ordination of the lake samples (n = 58) based on the Anabaena-, Microcystis-, and Planktothrix-specifi c mcyE PCR results (presence/absence). The specifi city of the PCR (string vector) is indicated with a boxed genus name. The squares show the positions of the lake groups, which contain different combinations of mcyE genes. The number of lakes belonging to each group is in parentheses. The environmental variables correlating in a statistically signifi cant manner with PC axis 1 and most signifi cantly with PC axis 2 are indicated below or beside the corresponding axis. The arrows point out the direction of correlation. PC axes 1 and 2 explained 42.4% and 38.1% of the variation, respectively. The colors of the squares represent the mcyE gene combinations and are in accordance with Fig. 4. Ana, Mic, and Pla stand for Anabaena, Microcystis, and Planktothrix, respectively.

Results

no PCR (5)

Ana-mcyE + Mic-mcyE (4)

Ana-mcyE +Mic-mcyE +Pla-mcyE (12) Mic-mcyE +

Pla-mcyE (11)

Pla-mcyE (11)

Mic-mcyE (12)Microcystis

Planktothrix

Ana-mcyE +Pla-mcyE (3)

Anabaena

Secc

hi d

epth

PC 1

PC

2

TN, pH, Chl-a, cyanobacterial, phytoplankton, Microcystis,Aphanizomenon, Oscillatoriales biomasses, total and intracellularmicrocystin concentration

30

4, Fig. 4). The effects of both TN and TP showed parabolic responses (shown by the terms TN2 and TP2 in the respective models), where the occurrence probability of the mcyE genes and microcystin concentrations begin to drop with the highest TN and TP levels, respectively (II: Fig. 4).

4.5 Development of the DNA-chip to detect hepatotoxin producers (III)

4.5.1 Specifi city of the probesGenus-specifi c probe pairs were designed to recognize the mcyE or ndaF gene sequences of Anabaena, Microcystis, Planktothrix, Nostoc, and Nodularia in LDR/hybridization experiments ( = DNA-chip assay; Fig. 3). The specifi city of the probe pairs was tested with mcyE/ndaF PCR products (25 fmol) amplifi ed from the genomic DNA of hepatotoxic cyanobacterial strains as target, either separately or as an artifi cial mix. After the DNA-chip assay, signals were generated only from spots that corresponded to the probe pair/pairs under scrutiny (III: Figs. 1 and 2). Thus, every probe pair specifi cally detected the mcyE/ndaF sequence of the genus it was designed to target. As a result, the strict specifi city of the probes allowed identification of all the hepatotoxin-producing genera studied.

4.5.2 Sensitivity of the probesThe detection limit (sensitivity) for each of the probe pairs was determined, using a dilution series (0.2, 1, 5, 25, or 125 fmol) of the corresponding amplifi ed mcyE/ndaF gene product as target for the DNA-chip assay. Signals exceeding the threshold level set were detected at 0.2 fmol with the Anabaena-mcyE probes, at 1 fmol with Nostoc-mcyE, Planktothrix-mcyE, and

Nodularia-ndaF probes, and at 5 fmol with the Microcystis-mcyE probes. All probe pairs generated stronger signal intensities when 25 fmol of the PCR product was used as target. At 125 fmol, the signal intensities showed unexpected variation; they either increased (Nostoc- and Nodularia-specifi c probe pairs), remained approximately the same (Planktothrix- and Microcystis-specific), or decreased (Anabaena-specific) compared with the intensities at 25 fmol. For this reason, 25 fmol was chosen for use in the LDR experiments.

4.5.3 Validation of DNA-chip results with environmental samples Performance of the probe pairs was validated with 10 environmental DNAs extracted from lake and brackish water samples: 0tu33 Bloom and 0tu35 > 10 μm (Lake Tuusulanjärvi), 2ky11 (Lake Köyliönjärvi), 2vj26 (Lake Vanajanselkä), 2po39 (Lake Puujärvi), 2en69 (Lake Enäjärvi), and the depths at 3 m, 7 m, 18 m, and 30 m of the sample Cya04_2 (Baltic Sea). The DNAs from the lake samples were analyzed both with the DNA-chip assay (III: Fig. 3) and qPCR with Anabaena-, Microcystis-, and Planktothrix-specific primers (III: Fig. 4). The results of both analyses were compared to evaluate the ability of the DNA-chip to detect potential microcystin producers in lake samples. The DNA-chip results of the Baltic Sea samples (III: Fig. 5) were validated with previously published qPCR results (Koskenniemi et al., 2007).

In four of the six lake samples: 0tu33 Bloom, 0tu35 > 10 μm, 2po39 and 2en69, the DNA-chip and qPCR detected the same mcyE genes. In addition, the DNA-chip detected Planktothrix-mcyE

Results

31

Tabl

e 4.

Det

ectio

n of

mcy

E an

d nd

aF in

mix

es o

f cya

noba

cter

ial s

train

DN

A b

y D

NA

-chi

p as

says

afte

r PC

R w

ith

prim

er p

airs

mcy

E-F2

/R4

and

mcy

E-F2

b/R

4 at

diff

eren

t ann

ealin

g te

mpe

ratu

res.

Prim

er p

air

and

anne

alin

g te

mpe

ratu

reb

DN

A m

ixa

F2/R

456

°CF2

/R4

55°C

F2/R

453

°CF2

b/R

457

°CF2

b/R

455

°CF2

b/R

453

°CM

icro

cyst

is P

CC

974

1 –

Ana

baen

a 90

– P

lank

toth

rix

49:

80 -

10 -

10M

cnd

dnd

Mnd

M; (

A);

P10

- 80

- 10

M; A

; Pnd

ndM

; (A

); (P

)nd

M; A

; P10

- 10

- 80

(M);

Pnd

ndM

; Pnd

M; (

A);

P45

- 45

- 10

M; (

P)nd

ndM

ndM

; A; P

45 -

10 -

45(M

); P

ndnd

Mnd

M; (

A);

P10

- 45

- 45

(M);

PM

; A; P

M; A

; PM

; (P)

M; A

; PM

; A; P

33 -

33 -

33M

; PM

; A; P

M; A

; PM

M; A

; PM

; A; P

Nod

ular

ia F

8-1

– A

naba

ena

318:

90 -

10(N

); A

ndnd

N; (

A)

ndN

; A10

- 90

And

nd(N

); A

nd(N

); A

50 -

50A

ndnd

N; A

ndN

; A30

- 70

And

nd(N

); (A

)nd

N; A

70 -

30A

ndnd

Faile

dend

N; A

a Pro

porti

ons

(%) o

f DN

A fr

om M

icro

cyst

is, A

naba

ena,

and

Pla

nkto

thri

x, o

r fro

m N

odul

aria

and

Ana

baen

a st

rain

s us

ed fo

r m

ixes

b P

rimer

pai

r and

ann

ealin

g te

mpe

ratu

re u

sed

in P

CR

prio

r to

DN

A-c

hip

assa

y c S

igna

l exc

eedi

ng th

e m

ore

strin

gent

thre

shol

d le

vel f

rom

Mic

rocy

stis

-mcy

E (M

), An

abae

na-m

cyE

(A),

Plan

ktot

hrix

-mcy

E (P

), an

d N

odul

aria

-nda

F (N

) spo

ts; s

igna

l exc

eedi

ng th

e lo

wer

thre

shol

d le

vel i

s ind

icat

ed w

ith th

e sa

me

lette

r in

pare

nthe

ses

d nd,

ass

ay n

ot p

erfo

rmed

with

that

com

bina

tion

of p

rimer

pai

r and

tem

pera

ture

e F

aile

d D

NA

-chi

p as

say

Results

32

with the other (probe I) of the two probe pairs designed for this gene in both Lake Tuusulanjärvi samples, although it was not detected with the Planktothrix-mcyE probe pair II and qPCR. In the other two lake samples, 2ky11 and 2vj26, all three producers were detected with qPCR. In contrast, the DNA-chip detected only Anabaena- and Microcystis-mcyE in 2ky11, and Microcystis-mcyE in 2vj26. The mcyE genes that were undetected showed very low gene copy numbers in qPCR analysis, which most probably caused their failed detection by the DNA-chip. In the Baltic Sea samples, strong signals were detected from the Nodularia ndaF spots at each depth sampled (III: Fig. 5), despite the wide variation (max. 29,966 copies ml-1 at 7 m; min. 214 copies ml-1 at 30 m) in the ndaF gene copy numbers (Koskenniemi et al., 2007). In all samples, signals were also detected from spots corresponding to Planktothrix-mcyE probe pair I. Based on the results, Planktothrix-mcyE probe pair II apparently performed more specifi cally than Planktothrix-mcyE probe pair I, which was detected in every sample. In general, the probe pairs specifi c for the mcyE/ndaF genes of Anabaena, Microcystis, Planktothrix (probe II), and Nodularia reliably detected these potential microcystin/nodularin producers in complex environmental samples.

4.6 Additional optimization of the DNA-chip assay

We performed further optimization of the DNA-chip assay conditions (Table 4). The DNA mixes of freshwater cyanobacterial strains and Baltic Sea strains were first amplified with primer pairs mcyE-F2/R4 and mcyE-F2b/R4 at annealing temperatures of 56°C and 57°C, respectively. The freshwater mixes

indicated strong preferential amplifi cation with both primer pairs. Only Microcystis-mcyE was detected in every mix by the DNA-chip assays. Signals from the Anabaena-mcyE spots were detected only when the proportion of Anabaena DNA was 80% of the total DNA amount (20 ng) used for PCR. Signals from the Planktothrix-mcyE spots were detected more often and at higher intensity when primer pair mcyE-F2/R4 was used for amplification. In the Baltic Sea mixes, strong preferential amplification of Anabaena-mcyE with primer pair mcyE-F2/R4 was evident by the DNA-chip results. Signals from the Nodularia-ndaF spots were detected only when the proportion of Nodularia DNA was 90%. In contrast, both Nodularia-ndaF and Anabaena-mcyE were amplified in all mixes by primer pair mcyE-F2b/R4, making it the primer pair of choice for experiments with brackish water samples.

The effect of annealing temperature on amplifi cation by the two primer pairs was assessed with selected freshwater mixes at 55°C and 53°C. In each DNA-chip assay and with both primer pairs, amplifi cation products from Microcystis-, Anabaena-, and Planktothrix-mcyE were detected. Finally, all mixes were amplifi ed at 53°C. Primer pair mcyE-F2b/R4 was chosen for these assays, due to its usability both in fresh and brackish water samples. In every freshwater mix, all three mcyE genes were detected. Only when the proportion of Anabaena DNA was 10% (2 ng), did the average signal intensity of the Anabaena-mcyE spots not exceed the higher detection threshold. In all the Baltic Sea mixes, both Nodularia-ndaF and Anabaena-mcyE were detected at higher intensities than at the annealing temperature of 57°C. Only when the proportion of Nodularia DNA was 10%, did the average signal intensity

Results

33

of Nodularia-ndaF fail to exceed the higher detection threshold. The results suggested that by lowering the annealing temperature of the PCR step, the DNA-

chip assay can more sensitively detect potential microcystin and nodularin producers present less abundantly in complex environmental samples.

Results

34

5. DISCUSSION

Planktonic cyanobacteria frequently form blooms that present a problem for water users via drinking or recreation. The main threat for animal and human health is posed by the potential presence of cyanobacterial toxins, the most common of which are the hepatotoxins, microcystins and nodularins (Sivonen and Jones, 1999). They are synthesized in cyanobacterial cells by microcystin and nodularin synthetases encoded by the mcy and nda gene clusters, respectively (Tillett et al., 2000; Christiansen et al., 2003; Moffitt and Neilan, 2004; Rouhiainen et al., 2004). Since toxin-producing and -nonproducing cyanobacterial strains cannot be distinguished by conventional microscopy (Sivonen and Jones, 1999), molecular detection methods based on biosynthetic genes are needed.

5.1 Evolution of microcystin synthetase genes

In modern cyanobacteria, microcystin production and the presence of mcy genes is sporadic. For example, all the main producer genera (Anabaena, Microcystis, and Planktothrix) include both toxic and nontoxic strains (Sivonen and Jones, 1999). Congruence of the phylogenetic trees constructed with the housekeeping and mcy gene sets and similar sequence divergence levels between the datasets suggested that both gene sets have co-evolved throughout the entire evolutionary history of this toxin and that microcystins are of ancient origin. In other words, the biosynthetic genes were already present in the last common ancestor of a large number of cyanobacteria. Congruence of the phylogenetic trees also supported repeated gene loss in the more derived

lineages of cyanobacteria as a reason for the sporadic distribution. HGT, which has also been suggested to cause the patchy distribution (Moffitt and Neilan, 2001, 2004; Tillett et al., 2001; Mikalsen et al., 2003), gained no support in our study.

Our results with three mcy genes: mcyA, -D, and -E, indicated that these genes are very old and that there was no lateral transfer of mcy gene clusters between the genera studied. However, it is possible that parts of the mcy gene cluster are of more recent origin and may have been laterally transferred between strains. Recombination events were observed, particularly in the adenylation domains of peptide synthetase genes mcyA, -B, and C (Mikalsen et al., 2003; Tanabe et al., 2004; Kurmayer et al., 2005; Kurmayer and Gumpenberger, 2006), leading to altered substrate specificities and microcystin isoforms (Mikalsen et al., 2003; Kurmayer et al., 2005; Kurmayer and Gumpenberger, 2006). In Planktothrix strains, deletions (mcyB, mcyHA), and transposon-mediated insertions (mcyA, mcyD, mcyEG) that abolish microcystin production were identified (Christiansen et al., 2006). Unidentified mutations either in mcy gene or regulatory regions that prevent microcystin production were implied in both Microcystis (Nishizawa et al., 1999, 2000; Tillett et al., 2000; Kaebernick et al., 2001; Mikalsen et al., 2003; Via-Ordorika et al., 2004) and Planktothrix (Kurmayer et al., 2004; Christiansen et al., 2006). The proportion of inactive mcy genotypes was estimated to be relatively high, especially in Planktothrix (Kurmayer et al., 2004; Christiansen et al., 2006). Mutations that prevent microcystin production can be seen as intermediates in the continuing evolution of the gene region, eventually

Results

35

leading to total loss of the nonfunctional mcy genes from the genome and supporting gene loss as a major evolutionary force in the distribution of mcy genes among cyanobacterial strains (Christiansen et al., 2006).

The phylogenetic analyses also suggested that nda genes are closely related to and have been derived from the mcy genes of other filamentous heterocystous cyanobacteria (Anabaena and Nostoc). The close relationship was also seen in a similar phylogenetic tree constructed, using another mcyE gene region (Jungblut and Neilan, 2006). Sequencing of the total nda gene cluster of the hepatotoxic Nodularia strain NSOR10 (Moffitt and Neilan, 2004) corroborated our hypothesis on how the nda genes evolved from the mcy genes by pinpointing the deleted region of the mcyABC region. The ndaA gene contains two adenylation domains, the first of which is related to the first mcyA-adenylation domain and the second to the second adenylation domain of mcyB. The ndaA condensation domain between the two adenylation domains includes regions similar to both mcyA and mcyB condensation domains (Moffi tt and Neilan, 2004). The change in substrate specifi city of mcyA adenylation domain 1, hypothesized by us, was also confirmed. The recognition sequence of ndaA adenylation domain 1 suggested that it activated threonine (a precursor of Mdhb in nodularins) (Moffi tt and Neilan, 2004) instead of the serine (a precursor of Mdha in microcystins) activated by mcyA adenylation domain 1 of Anabaena, Microcystis, and Planktothrix (Tillett et al., 2000; Christiansen et al., 2003; Rouhiainen et al., 2004). Except for the deleted mcyAB region, a striking similarity in operon organization is apparent when the entire mcy gene cluster of Anabaena 90

(Rouhiainen et al., 2004) and the nda gene cluster of Nodularia NSOR10 (Moffi tt and Neilan, 2004) are compared (Fig. 3). This gives further support for the hypotheses that the nda genes evolved from the mcy genes (I; Moffi tt and Neilan, 2004), while independent acquisition of the mcyAB region by the other microcystin-producing genera (Christiansen et al., 2003) is unlikely. The evolutionary origin of the nda genes as well as the high similarity of the two toxin molecules suggest that nodularin could actually be regarded as a structural variant of the microcystins.

Age estimates of cyanobacterial existence differ greatly between ~3500 and ~1500 million years, based on fossil records and protein clocks, respectively (Schopf, 2000). Molecular fossils of membrane lipids that are characteristic for cyanobacteria were found in 2700 million-year-old rocks (Brocks et al., 1999). Fossi ls of akinete-forming cyanobacteria from 2000 million-year-old stromatolites (Amard and Bertrand-Sarfati, 1997) suggest that the genera Anabaena, Nostoc, and Nodularia and thus, the common ancestor of microcystin-producing cyanobacteria, are at least this old. This means that microcystins were most likely already synthesized before the rise of crown-eukaryotic lineages ~1500 million years ago (Wang et al., 1999), raising questions as to whether the toxic function of microcystins against modern zooplankton species (DeMott et al., 1991; Rohrlack et al., 1999, 2001) represents the primary function that evolved over 2000 million years ago. Did microcystins evolve instead as protection against other organisms? Cyanotoxins were suggested to play a role, e.g. in maintaining mutually benefi cial relationships with heterotrophic bacteria (Paerl, 1996).

Results

36

5.2 mcyE as a molecular marker for hepatotoxin-producing cyanobacteria

When this study was initiated in early 2002, only a few mcy sequences of Microcystis (Nishizawa et al., 1999, 2000; Tillett et al., 2000) were available in public databases. In addition to these sequences, the mcy gene cluster of Anabaena sp. 90 was being sequenced by our research group (Rouhiainen et al., 2004), giving us a unique opportunity to compare the sequences derived from these two distantly related cyanobacteria. This led to recognition of regions that could be utilized to design primers for detecting potential microcystin producers.

Of the many potentially usable mcy genes, mcyE was chosen as the fi rst target gene for primer design. This gene encodes McyE, a mixed polyketide synthase/peptide synthetase involved in the synthesis of Adda, and the activation and addition of D-glutamate into the microcystin molecule. These two amino acids are crucial to toxicity and vary less than do the other amino acids of the molecule (Sivonen and Jones, 1999). The peptide synthetase genes are structured as modules that usually contain at least adenylation, condensation, and thiolation domains. The adenylation and condensation domains of mcyE differed more clearly from the corresponding domains in the other mcy genes and other peptide synthetase genes present in cyanobacteria and other bacteria, diminishing the risk for false amplification of those other peptide synthetase gene sequences. Taken together, the mcyE gene region was believed to be a reliable molecular marker for the detection of microcystin producers.

Already one of the fi rst primer pairs designed for the mcyE gene, mcyE-F2/R4, fulfilled the expectations laid

on it. When tested with cyanobacterial strains, amplification was detected with microcystin-producing Anabaena, Microcystis, Planktothrix, and Nostoc, and nodularin-producing Nodularia (Table 3). The nontoxic or neurotoxic strains tested gave no amplification products. Based on these results, mcyE-F2/R4 earned its description as a ‘general’ primer pair. Such general primer pairs were also designed for the other mcy genes: mcyA (Hisbergues et al., 2003), mcyB (Neilan et al., 1999), mcyD (I), and mcyE (Jungblut and Neilan, 2006). Primers targeted to the condensation domain of mcyA (Hisbergues et al., 2003), however, could not detect nodularin-producing Nodularia that lack the corresponding gene region (Moffitt and Neilan, 2004), while primers targeted to regions of the biosynthetic genes other than the second module of mcyA or the fi rst module of mcyB can also detect the orthologous nda genes (Neilan et al., 1999; Jungblut and Neilan, 2006; I).

The majority of the first probes and primers were designed to detect specifi cally the mcy genes of Microcystis (Table 1), which probably refl ected both the importance of Microcystis as a toxin producer and availability of the gene sequence. However, in environmental samples more than one microcystin producer can co-occur (Vezie et al., 1997, 1998). Use of general primers cannot reveal the actual producer without either sequencing the amplification product (e.g. Neilan et al., 1999; Rinta-Kanto and Wilhelm, 2006; Jungblut and Neilan, 2006) or using another post-PCR analysis method, e.g. RFLP (Hisbergues et al., 2003). The fi rst sequence analyses showed that the mcyE region used in this study was highly similar (97.4-100%) among strains belonging to the same genus, but more diverse (74-88%) between

Results

37

different genera. Thus, this region could also provide a basis for the genus-level differentiation and direct identifi cation of the producer organisms. Genus-specific primers were designed for Anabaena-, Microcystis- (Vaitomaa et al., 2003), and Planktothrix-mcyE (II), and Nodularia-ndaF (Koskenniemi et al., 2007) genes. When used together, the main microcystin or nodularin producers could be identifi ed (Vaitomaa et al., 2003; Gobler et al., 2007; Hotto et al., 2007; II) or even quantifi ed using qPCR appl icat ions in f resh (Vaitomaa et al., 2003; III) and brackish water samples (Koskenniemi et al., 2007). Genus-specific primers for Anabaena-, Microcystis-, and Planktothrix-mcyE detected potential microcystin producers in 91% of 70 lakes in Finland compared with 84% that were detected by the general primer pair (II). Thus, genus-specific primers that amplified a shorter region (250 bp) than the general primer pair (810 bp) appeared to be slightly more effi cient. Genus-specifi c primer pairs may also give a more reliable result because they are not as likely to suffer from preferential amplifi cation as general primers.

The abil i ty to detect potential hepatotoxin producers is, however, not the only characteristic needed. Detection of the targeted gene region should be reliably associated with the feature in question. The evolutionarily stable mcyE gene region (I) is suitable for identification of potential microcystin and nodularin producers, which requires strong association between sequence variants and the genus of the producer. Primers designed for different regions of mcyE by us and others can reliably detect microcystin and nodularin producers both among laboratory strains and environmental samples (Vaitomaa

et al., 2003; Mbedi et al., 2005; Jungblut and Neilan, 2006; Mankiewicz-Boczek et al., 2006; Gobler et al., 2007; Hotto et al., 2007; Koskenniemi et al., 2007). The reliability of detection is challenged by the inactive mcy genotypes found in Microcystis (Nishizawa et al., 1999, 2000; Tillett et al., 2000; Kaebernick et al., 2001; Mikalsen et al., 2003; Via-Ordorika et al., 2004) and Planktothrix (Kurmayer et al., 2004; Christiansen et al., 2006). Recombinations in the mcyABC regions of Microcystis and Planktothrix (Mikalsen et al., 2003; Tanabe et al., 2004; Kurmayer et al., 2005; Kurmayer and Gumpenberger, 2006) and deletions/insertions of several gene regions in Planktothrix (Christiansen et al., 2006) can also cause problems for detection with sequence-specifi c primers and probes. Most variation and instability in the gene cluster appear to be sited in those regions that encode the peptide synthetases responsible for incorporation of the more variable amino acids of the microcystin. In contrast, such changes have only seldom been recognized from gene regions corresponding to the more stable parts of microcystins (Christiansen et al., 2006) and not at all in mcyE.

Detection of microcystin and nodu-larin producers by the methods based on mcyE is not restricted only to samples from lakes in Finland and the Baltic Sea, but can most probably be applied worldwide, due to the high intragenic similarity of mcyE sequences (I; Jungblut and Neilan, 2006). This was also demonstrated by successful amplifi cation (I) or identifi cation of target sequences in silico (III) from Finnish, Portuguese, French, Dutch, Russian, Canadian, and Japanese Microcystis strains as well as from Finnish, Danish, and Norwegian Anabaena strains.

Results

38

5.3 Potential microcystin producers in lakes in Finland

The frequency of potential microcystin producers detected by both general (84%) and genus-specific (91%) primers in the lake samples was very high. In previous studies, an average of 59% (range 10-92%) of cyanobacterial blooms have been hepatotoxic (Sivonen and Jones, 1999). A previous survey of lakes in Finland was conducted in the mid-1980s, including analysis of 188 cyanobacterial bloom samples (Sivonen et al., 1990). In that study, 29% of the samples were hepatotoxic. The overwhelming difference in proportions of hepatotoxic blooms and presence of potentially hepatotoxic cyanobacteria between the previous survey and this study (II) can at least partly be attributed to the nonsensitive mouse bioassay used to determine hepatotoxicity.

In our s tudy, the presence of potent ial ly microcyst in-producing Microcystis was indicated in 70% of the lakes, while microcystin-producing Anabaena and Plankto thr ix were detected in 37% and 63% of the samples, respectively. This is partly inconsistent with the previous survey, in which both Anabaena (78%) and Microcystis (69%) were commonly and Planktothrix (25%) less frequently found in hepatotoxic blooms by microscopy (Sivonen et al., 1990). The lower proportion of Anabaena in the present study could be explained by the early sampling time in mid-July, since Anabaena usually becomes more frequent in August and September when growth of non-nitrogen-fixing cyanobacterial genera is limited, due to nitrogen depletion (Rajaniemi-Wacklin et al., in press; Sommer et al., 1986). The higher proportion of Planktothrix in our

study could refl ect sampling of a deeper water column (0-2 m) instead of surface blooms. The proportion could have been even higher if the entire water column had been sampled. Planktothrix does not generally form surface blooms, but prefers growing under lower light intensities in deeper water layers (Lindholm and Meriluoto, 1991). The frequency of microcystin-producing Planktothrix could also have been overestimated by the PCR analysis. Some of the PCR results with environmental lake samples could have been false-positives, if Planktothrix-specific primers had detected unknown variants of mcyE or other peptide synthetase genes, or mcyE of Planktothrix strains unable to produce microcystins. The proportion of such inactive microcystin genotypes of Planktothrix was estimated to be relatively high (5% and 21%) in two Alpine lakes (Kurmayer et al., 2004).

Co-occur rence o f po ten t i a l ly microcystin-producing genera was frequent among the 70 lakes studied. More than one microcystin-producing genus was detected in 54% and all three genera were present in 24% of the samples. Eutrophication of the lakes appeared to favor the presence of microcyst in-producing genera . The proportion of the lakes, in which Anabaena, Microcystis, and Planktothrix were all detected, increased along with the TP concentration, and was 5%, 17%, 35%, and 75% in oligo-, meso-, eu-, and hypertrophic lakes, respectively. In PCA, the presence of multiple microcystin-producing genera was significantly associated with higher TN concentration as well as with other environmental factors that usually prevail during blooms, such as a more alkaline pH, higher chl-a concentration, and cyanobacterial and phytoplankton biomasses. In PCA, however, weaker correlations of TP (R =

Results

39

-0.25; P = 0.06) and DIP concentrations (R = -0.25; P = 0.054) with the presence of multiple microcystin producers failed to attain statistical signifi cance (P < 0.05). This could have been caused by the lower number of lake samples used in PCA (n = 58) than in PCR (n = 70). Additionally, a higher TN concentration together with a higher pH and larger lake surface area also predicted a greater probability of occurrence of mcyE genes in a logistic regression analysis. However, the response curve of the model was not linear and at the highest values of TN, the occurrence probability of the mcyE genes began to decline. In addition, the microcystin concentrations of the lake samples were predicted in the same way by TP. Similar nonlinear responses were detected between TN and TP concentrations and microcystin levels in 241 lakes in Missouri (Graham et al., 2004). High nitrogen and phosphorus concentrations were also associated with increased microcystin concentration in other environmental studies (Kotak et al., 1995, 2000; Lahti et al., 1997; Oh et al., 2001; Rolland et al., 2005). Instead of direct infl uence, nutrients were suggested to act indirectly by promoting cyanobacterial growth, which allows for a higher rate of microcystin production (Orr and Jones, 1998; Briand et al., 2005). In contrast, associations between the presence of mcy genes and environmental factors are still rarely reported. In a lake in Japan, the relative abundance of mcyA-containing Microcystis rose when surface nitrate concentrations increased (Yoshida et al., 2007). Nitrate-replete conditions were recently associated with periods of more active expression of Microcystis-specifi c mcyE in Lake Agawam (Gobler et al., 2007).

In general , the results showed that nutrient-rich waters offer suitable

conditions for the existence of multiple toxin producers. Thus, reduced nutrient loading could be of advantage in lowering the occurrence of microcystin producers in the lakes. However, changed conditions, e.g. diminished nitrogen or phosphorus concentrations or altered light climate, could benefit some previously nondominant species and cause them to form blooms instead (Lepistö et al., 1999; Jacquet et al., 2005). Information on all the potential microcystin producers existing in lakes could be useful in planning lake restoration. Conventional microscopy, which has traditionally been used to determine phytoplankton composition in lakes, cannot separate toxin-producing cyanobacterial species from nontoxic species. Since cyanobacterial strains produce various amounts of microcystins, the dominant species in a sample may not be the main microcystin producer. Therefore, the importance of highly toxic, but less abundant, species could be underestimated by microscopy (Sivonen and Jones, 1999). Employment of sensitive molecular methods such as PCR would provide additional information for decisions and actions to meet the specifi c needs of each lake.

5.4 DNA-chip as a detection method for hepatotoxin producers

The DNA-chip assay of the present study employed the LDR/universal microarray method (Gerry et al., 1999). The LDR is capable of detecting single nucleotide variations between highly similar sequences. In addition, similar thermodynamic behavior of ZipCodes and elevated hybridization temperature diminish nonspecif ic binding and background noise. The method has successfully been used to detect both

Results

40

low abundant point mutations and small insertions and deletions in human tissues (e.g. Favis et al., 2000; Fouquet et al., 2004), as well as to assess cyanobacterial 16S rRNA diversity (Castiglioni et al., 2004). Combining detection of potential hepatotoxin producers with mcyE /ndaF gene probes (III) with detection of cyanobacterial groups based on 16S rRNA probes (Castiglioni et al., 2004) would allow for comprehensive analysis of community structure by the DNA-chip. Applied to environmental monitoring, it could complement traditional microscopy, which cannot differentiate between toxic and nontoxic strains.

Performance of DNA-chips is dependent on the specificity, sensitivity, and resolution of detection (Zhou and Thompson, 2002; Bodrossy and Sessitsch, 2004). The genus-specifi c probes are the basis for the specifi city and sensitivity of the DNA-chip method used here. Results with cyanobacterial strains harboring the toxin synthetase genes showed that each probe pair specifically detected only its own target, even when only 1 – 5 fmol of the mcyE/ndaF PCR product was used in the LDR. High sensitivity and detection of even faint signals are attained not only by the PCR amplifi cation of the target but also by the reduced background signal level characteristic of the LDR/universal microarray method (Gerry et al., 1999). However, impurities such as humic acids that are commonly present in environmental samples may decrease the sensitivity of the method by reducing the effi ciency of PCR amplifi cation.

Resolution is dependent on the marker gene used for detection. The mcyE/ndaF region that was the basis for probe design is present in both mcy and nda gene clusters (Nishizawa et al., 2000; Tillett et al., 2000; Christiansen et al., 2003; Moffi tt

and Neilan, 2004; Rouhiainen et al., 2004). Since the DNA-chip assay represents a form of post-PCR analysis, the resolution relies mainly on the PCR step, i.e. the ability of the primers to amplify all the target sequences present in a sample. The present study had previously shown that the primer pair mcyE-F2/R4 was capable of amplifying the gene region from strains of various hepatotoxin-producing genera (I) and from environmental DNA (II). In general, comparison of the DNA-chip results of lake samples with qPCR results showed in general consistent detection of potential microcystin producers. However, in two lake samples (Lake Köyliönjärvi and Lake Vanajanselkä) the DNA-chip failed to detect Anabaena- and Planktothrix-mcyE genes that according to qPCR were present in lower quantities. In conventional PCR, both Anabaena-mcyE and Planktothrix-mcyE showed relatively strong amplification with genus-specific primers (II). Weaker detection in qPCR and lack of detection in the DNA-chip assays could in part have refl ected higher sensitivity for potential PCR inhibitors by the polymerase or the general primer pair used, respectively. One probable cause for failed detection in the DNA-chip assay, preferential amplifi cation by the general primer pair, was later investigated with mixes containing different combinations of Microcystis, Anabaena, and Planktothrix DNA (freshwater mixes), or Nodularia and Anabaena DNA (Baltic Sea mixes). Results with two general primer pairs, mcyE-F2/R4 and mcyE-F2b/R4, showed that Microcystis-mcyE was most reliably and Anabaena -mcyE most poor ly detected by the DNA-chip under the PCR conditions originally used. However, lowering the annealing temperature with both primer pairs equalized the detection, suggesting that performance of the DNA-

Results

41

chip, and thus also detection of microcystin producers in lake samples, could be further improved. In addition, improvement in equal detection of Nodularia-ndaF was achieved by the use of a modifi ed general primer pair mcyE-F2b/R4 at the lower annealing temperature. The PCR-induced bias could also be alleviated by performing several independent PCRs to increase the probability for amplification of the less abundant mcyE/ndaF genes present in samples.

The characteristics of the DNA-chip would be most advantageous when applied to large series of complex environmental samples. Currently, the mcyE/ndaF probes included in the assay target all the principal hepatotoxin producers, and the composition of toxin producers in eight samples could be revealed simultaneously with the DNA-chip. Co-occurrence of several producers, especially in freshwaters, is frequent (Vezie et al., 1997, 1998; Vaitomaa et al., 2003; Rinta-Kanto and Wilhelm, 2006; II). In contrast to the DNA-chip assay, separate reactions for each genus analyzed are needed for detection and identification with genus-specific primers (Vaitomaa et al., 2003;

II). RFLP analysis of mcyA (Hisbergues et al., 2003) provides a simple method for detecting and separating microcystin-producing Anabaena, Microcystis, and Planktothrix in lake samples. However, unlike the DNA-chip, it is not applicable to brackish water samples, where nodularin-producing Nodularia frequently occur (Sivonen and Jones, 1999). Use of the DNA-chip with Baltic Sea samples is further warranted by the discovery of microcystin-producing Anabaena in the Baltic Sea in addition to Nodularia (Halinen et al., 2007). The community composition of hepatotoxin-producers can also be revealed by cloning and subsequent sequencing of PCR products (Jungblut and Neilan, 2006). Although tedious and time-consuming if many samples are to be studied, it provides the only method that can reveal the presence of new, previously unknown mcy/nda variants, and thus new producer genera in environmental samples. However, due to the nature of ZipCode hybridization, new probe pairs can easily be appended to the DNA-chip assay if other important producer genera are discovered.

Results

42

6. CONCLUSIONS

Cyanobacterial mass occurrences (blooms) are widespread in fresh and brackish waters. They are usually formed by planktonic cyanobacterial species, including potential microcystin- or nodularin-producing species of the genera Anabaena, Microcystis, Planktothrix, and Nodularia. Co-occurrence of toxin-producing and nonproducing strains is common. Detection methods based on microcystin (mcy) and nodularin synthetase (nda) genes are needed for differentiation of hepatotoxin-producing and -nonproducing cyanobacteria strains, because they cannot be differentiated by conventional microscopy or even by use of other genes that are traditionally used for classifi cation, e.g. 16S rRNA genes.

Three methods were used in this study to detect microcystin- or nodularin-producing cyanobacteria in environmental samples. Primers to amplify the mcyE/ndaF genes were designed for use in conventional PCR and qPCR. In conventional PCR applications, use of general primers can reveal the presence of potential hepatotoxin producers. Use of genus-specifi c primers gives additional information on the identity of potential producers present. Conventional PCR is a relatively cheap and robust technology, but it only offers data on the presence and not the quantity of potential toxin producers. qPCR methods can instead be used to quantify the gene copy numbers and thus identify the dominant producer in samples. Identification of different microcystin producers present in environmental water samples by genus-specifi c primers requires separate reactions for each genus studied. This can prove laborious when large sample series are studied. The DNA-chip assay developed can detect all the

main microcystin and nodularin producers simultaneously and be applied to eight samples in parallel. These high-throughput characteristics make this method applicable to monitoring purposes. However, special and expensive equipment is needed for this method, thus restricting its use.

In addition to detecting and identifying potential microcystin producers in samples from lakes in Finland, the PCR data could be used to study which environmental variables were associated with the presence of mcyE genes and the respective microcystin producers. The molecular detection methods that were used to identify the potential producers, instead of laborious strain isolation with concomitant toxin analyses, made such an extensive comparison achievable. Although several environmental variables contributed to the results, the importance of nutrients was emphasized. The frequency of lakes with three microcystin producers increased along with the trophic status (total phosphorus concentration) of the lakes. In addition, the high total nitrogen concentration correlated signifi cantly with the presence of multiple microcystin-producing genera and predicted higher occurrence probability of mcyE genes in statistical analyses. The results suggested that the risk for toxic bloom development increases in eutrophic lakes and thus underlined the importance of protecting waters from eutrophication and reducing nutrients as part of lake restoration.

The sequence data of the biosynthetic genes produced in this study aided in elucidating the evolutionary history of microcystin synthesis in cyanobacteria. The sporadic distribution of microcystin-producing strains among and between genera is most probably the result of

Conclusions

43

vertical inheritance of the gene cluster from ancestral cyanobacteria combined with recurrent gene losses in some of the cyanobacterial lineages. Analyses showed that mcy genes are ancient. Comparison of the data from fossil records with phylogenetic trees of mcy genes suggested that these genes were already

present in an ancestor of all heterocystous cyanobacteria and thus are probably older than 2000 million years. For this reason, it is also highly possible that some genera previously believed to be nontoxic may have retained the mcy genes and are able to produce microcystins.

Conclusions

44

7. ACKNOWLEDGEMENTS

This work was carried out at the Department of Applied Chemistry and Microbiology, Division of Microbiology, University of Helsinki. The study was funded by the EU project MIDI-CHIP (EKV2-CT-1999-00026), Academy of Finland (Center of Excellence ‘Microbial Resources Research Unit’ and grants 46812, 53305, 214457 to Kaarina Sivonen), and Viikki Graduate School in Biosciences (VGSB). I would like to express my respect to Professor Mirja Salkinoja-Salonen, the head of the Division of Microbiology, for her ever-enthusiastic attitude towards science.

I warmly thank Docent Petri Auvinen and Dr. Rainer Kurmayer for reviewing this thesis and giving me invaluable suggestions for improving it.

This work was supervised by Academy Professor Kaarina Sivonen. Her open-mindedness, innovativeness, and fairness both as a scientist and a person have made me feel the luckiest graduate student ever for having a chance to grow as a researcher under her wings. I cannot fi nd words big enough to thank you – so I just say: THANK YOU, Kaarina!

The members of my support group: Christina, Per, and Petri, are thanked for their help and advice. I am especially grateful to Petri for his guidance concerning the DNA-chips. The coordinator of VGSB, Eeva Sievi, is warmly thanked for her help and always being so cheerful.

My MIDI-CHIP partners from Belgium (Annick, Stana, Christophe), Finland (Liisa, Pirjo K.), Great Britain (Peter, Richard), Italy (Gianluca, Ermanno, Bianca, Andrea, Stefano, Maria, Giovanni), Luxembourg (Lucien, Raphael), Poland (Malgorzata, Joanna), and the Czech Republic (Jiri, Jarka, Eliska) are thanked for the unforgettable meetings and fruitful collaboration. Pirjo W. is especially thanked for all the time spent together with planning, problem solving, report writing, and lab experiments.

All the coauthors are acknowledged for their valuable contribution to the papers. I give my special thanks to David and Jukka for their brilliance with phylogenetics and statistics, respectively.

All present and former members of the CYANO group are thanked for creating a most enjoyable working atmosphere. Pirjo, Eija, Jaana L., Anu, Katrianna, Katri, and Kaisa R. are warmly thanked for sharing a room and also their thoughts and ideas as well as joys and sorrows. I thank Matti and Lyudmila for their invaluable help with the lake samples. Leo – it is hard to express the respect I feel for you. Your experience, helpfulness, and sense of humor impress me beyond words. Christina – your devotion to helping everybody regardless of the nature of the problem and your ability to fi nd the silver lining in the darkest cloud are already reasons enough for more thanks than can fi t

Acknowledgements

45

on this page. However, teaching me how to crystallize the essence of my thoughts or, in other words, how to fi nd the “villakoiran ydin” is the real reason for my most heartfelt thanks to you. Last, but not least, I want to thank Tanja for introducing the CYANO group to me in the fi rst place and for encouragement during the writing of this thesis.

I want to address my warmest thanks to all my colleagues from the other groups. The scientifi c and recreational discussions and moments with you around the coffee table will not be forgotten. Special thanks to Shea, Janetta, Henrietta, and Mari for your friendship and evenings of knitting and good food. Janetta, thank you for the opportunity to vent my feelings during the numerous coffee breaks we have had together this year. The personnel of the Division of Microbiology is thanked for all their help with dishes, invoices, traveling etc. Without you life in the lab would have been so much harder. Kikka from the University Sports Service is thanked for her excellent training classes that helped me endure the physical strains of writing.

I have been blessed with many true friends. I wish to thank you all, and particularly Maija, Auli, Minna, and Tanja for the joyful moments I have had the privilege to spend with you and your families. You mean a world to me!

I want to thank my relatives, both ‘old’ and ‘new’, for their love and support. My parents, Leena and Ari, are thanked for helping me in countless ways and especially for never asking “How long is it still going to take you to graduate?”. My sister Pia is warmly thanked for the numerous car drives and dinners, and her help with the arrangements of the ‘karonkka’.

Janne, my dearest, you are my other (better) half, and without you I don’t feel myself complete. Your support and love have helped me fi nish off this thesis. Thank you for the refreshing walks hand in hand, listening to my troubles, relieving my doubts and fears, and hugging the stress away time after time.

Helsinki, November 2007

Acknowledgements

46

8. REFERENCES

Amard, B., and J. Bertrand-Sarfati. 1997. Microfossils in 2000 Ma old cherty stromatolites of the Franceville Group, Gabon. Precamb. Res. 81:197-221.

Anjos F. M., M.C. Bittencourt-Oliveira, M. P. Zajac, S. Hiller, B. Christian, K. Erler, B. Luckas, and E. Pinto. 2006. Detection of harmful cyanobacteria and their toxins by both PCR amplifi cation and LC-MS during a bloom event. Toxicon 48:239-245.

Baker, J. A., B. Entsch, B. A. Neilan and D. B. McKay. 2002. Monitoring changing toxigenicity of a cyanobacterial bloom by molecular methods. Appl. Environ. Microbiol. 68:6070-6076.

Beiko, R. G., T. J. Harlow, and M. A. Ragan. 2005. Highways of gene sharing in prokaryotes. Proc. Nat. Acad. Sci. USA 102:14332-14337.

Berg, O. G., and C. G. Kurland. 2002. Evolution of microbial genomes: sequence acquisition and loss. Mol. Biol. Evol. 19:2265-2276.

Bhattacharya, D., H. S. Yoon, and J. D. Hackett. 2003. Photosynthetic eukaryotes unite: endosymbiosis connects the dots. BioEssays 26:50-60.

Bianchi, T. S., P. Westman, and C. Rolff. 2000. Cyanobacterial blooms in the Baltic Sea: natural or human-induced? Limnol. Oceanogr. 45:716-726.

Bittencourt-Oliveira, M. C. 2003. Detection of potential microcystin-producing cyanobacteria in Brazilian reservoirs with a mcyB molecular marker. Harmful algae 2:51-60.

Boaru, D. A., N. Dragoş, M. Welker, A. Bauer, A. Nicoară, and K. Schirmer. 2006. Toxic potential of microcystin-containing cyanobacterial extracts from three Romanian freshwaters. Toxicon 47:925-932.

Bodrossy, L., and A. Sessitsch. 2004. Oligonucleotide microarrays in microbial diagnostics. Curr. Opin. Microbiol. 7:245-254.

Briand, J.-F., S. Jacquet, C. Flinois, C. Avois-Jacquet, C. Maisonnette, B. Leberre, and J.-F. Humbert. 2005. Variations in

the microcystin production of Planktothrix rubescens (Cyanobacteria) assessed from a four-year survey of Lac du Bourget (France) and from laboratory experiments. Microb. Ecol. 50:418-428.

Brocks, J. J., G. A. Logan, R. Buick, and R. E. Summons. 1999. Archean molecular fossils and the early rise of eukaryotes. Science 285:1033-1036.

Burja, A. M., S. Dhamwichukorn, and P. C. Wright. 2003. Cyanobacterial postgenomic research and systems biology. Trends Biotechnol. 21:504-511.

Carmichael, W. W. 1994. The toxins of cyanobacteria. Sci. Am. 270:78-86.

Carmichael, W. W., and R. Li. 2006. Cyanobacteria toxins in the Salton Sea. Saline Systems 2:5 (doi:10.1186/1746-1448-2-5).

Castenholz , R. W. 2001 . Genera l characteristics of the cyanobacteria, p. 474- 487. In D. R. Boone and R. W. Castenholz (ed.), Bergey’s manual of systematic bacteriology, 2nd edition. Springer-Verlag, New York, NY, USA.

Castiglioni, B., E. Rizzi, A. Frosini, K. Sivonen, P. Rajaniemi, A. Rantala, M. A. Mugnai, S. Ventura, A. Wilmotte, C. Boutte, S. Grubisic, P. Balthasart, C. Consolandi, R. Bordoni, A. Mezzelani, C. Battaglia, and G. De Bellis. 2004. Development of a universal microarray based on the ligation detection reaction and 16S rRNA gene polymorphism to target diversity of cyanobacteria. Appl. Environ. Microbiol. 70:7161-7172.

Christiansen, C., J. Fastner, M. Erhard, T. Börner, and E. Dittmann. 2003. Microcystin biosynthesis in Planktothrix: genes, evolution and manipulation. J. Bacteriol. 185:564-572.

Christiansen, C., R. Kurmayer, Q. Liu, and T. Börner. 2006. Transposons inactivate biosynthesis of the nonribosomal peptide microcystin in naturally occurring Planktothrix spp. Appl. Environ. Microbiol. 72:117-123.

DeMott, W. R., Q.-X. Zhang, and W. W. Carmichael. 1991. Effects of toxic cyanobacteria and purified toxins on the

References

47

survival and feeding of a copepod and three species of Daphnia. Limnol. Oceanogr. 36:1346-1357.

Dittmann, E., M. Erhard, M. Kaebernick, C. Scheler, B. A. Neilan, H. von Döhren, and T. Börner. 2001. Altered expression of two light-dependent genes in a microcystin-lacking mutant of Microcystis aeruginosa PCC 7806. Microbiol. 147:3113-3119.

Dittmann, E., K. Meißner, and T. Börner. 1996. Conserved sequences of peptide synthetase genes in the cyanobacterium Microcystis aeruginosa. Phycologia 35:62-67.

Dittmann, E., B. A. Neilan, M. Erhard, H. von Döhren, and T. Börner. 1997. Insertional mutagenesis of a peptide synthetase gene that is responsible for hepatotoxin production in the cyanobacterium Microcystis aeruginosa PCC 7806. Mol. Microbiol. 26:779-787.

Doolittle, W. F. 1999. Phylogenetic classifi cation and the universal tree. Science 284:2124-2128.

Ehira, S., M. Ohmori, and N. Sato. 2003. Genome-wide expression analysis of the responses to nitrogen deprivation in the heterocyst-forming cyanobacterium Anabaena sp. strain PCC 7120. DNA Res. 10:97-113.

Farris, J. S., M. Källersjö, A. G. Kluge, and C. Bult. 1995. Testing signifi cance of incongruence. Cladistics 10:315-319.

Favis, R., J. P. Day, N. P. Gerry, C. Phelan, S. Narod, and F. Barany. 2000. Universal DNA array detection of small insertions and deletions in BRCA1 and BRCA2. Nature Biotech. 18:561-564.

Finni, T., K. Kononen, R. Olsonen, and K. Wallström. 2001. The history of cyanobacterial blooms in the Baltic Sea. Ambio 30:172-178.

Foulds, I. V., A. Granacki, C. Xiao, U. J. Krull. A. Castle, and P. A. Horgen. 2002. Quantification of microcystin-producing cyanobacteria and E. coli in water by 5’-nuclease PCR. J. Appl. Microbiol. 93:825-834.

Fouquet, C., M. Antoine, P. Tisserand, R. Favis, M. Wislez, F. Commo, N. Rabbe,

M. F. Carette, B. Milleron, F. Barany, J. Cadranel, G. Zalcman, and T. Soussi. 2004. Rapid and sensitive p53 alteration analysis in biopsies from lung cancer patients using a functional assay and a universal oligonucleotide array: a prospective study. Clin. Cancer Res. 10:3479-3489.

Francis, G. 1878. Poisonous Australian lake. Nature 18:11-12.

Furukawa, K., N. Noda, S. Tsuneda, T. Saito, T. Itayama, and Y. Inamori. 2006. Highly sensitive real-time PCR assay for quantifi cation of toxic cyanobacteria based on microcystin synthetase A gene. J. Biosc. Bioeng. 102:90-96.

Gerry, N. P., N. E. Witowski, J. Day, R. P. Hammer, G. Barany, and F. Barany. 1999. Universal DNA microarray method for multiplex detection of low abundance point mutations. J. Mol. Biol. 292:251-262.

Gobler, C. J., T. W. Davis, K. J. Coyne, and G. L. Boyer. 2007. Interactive influences of nutrient loading, zooplankton grazing, and microcystin synthetase gene expression on cyanobacterial bloom dynamics in a eutrophic New York lake. Harmful Algae 6:119-133.

Graham, J. L., J. R. Jones, S. B. Jones, J. A. Downing, T. E. Clevenger. 2004. Environmental factors inf luencing microcystin distribution and concentration in the Midwestern United States. Water Res. 38:4395-4404.

Gugger, M., C. Lyra, P. Henriksen, A. Couté, J.-F. Humbert, and K. Sivonen. 2002. Phylogenetic comparison of the cyanobacterial genera Anabaena and Aphanizomenon . Int. J. Syst. Evol. Microbiol. 52:1867-1880.

Halinen, K., J. Jokela, D. P. Fewer, M. Wahlsten, and K. Sivonen. 2007. Direct evidence for production of microcystins by Anabaena strains from the Baltic Sea. Appl. Environ. Microbiol. 73:6543-6550.

Harada, K.-I., F. Kondo, and L. Lawton. 1999. Laboratory analysis of cyanotoxins, p.369-405. In I. Chorus and J. Bartram (ed.), Toxic cyanobacteria in water: a guide to their public health consequences, monitoring, and management. E & FN Spon, London, United Kingdom.

References

48

Hihara, Y., A. Kamei, M. Kanehisa, A. Kaplan, and M. Ikeuchi. 2001. DNA microarray analysis of cyanobacterial gene expression during acclimation to high light. Plant Cell 13:793-806.

Hisbergues, M., G. Christiansen, L. Rouhiainen, K. Sivonen, and T. Börner. 2003. PCR-based identif ication of microcystin-producing genotypes of different cyanobacterial genera. Arch. Microbiol. 180:402-410.

Honkanen, R. E., M. Dukelow, J. Zwiller, R. E. moore, B. S. Khatra, and A. L. Boynton. 1991. Cyanobacterial nodularin is a potent inhibitor of type 1 and type 2A protein phosphatases. Mol. Pharmacol. 40:577-583.

Hopwood, D. A., and D. H. Sherman. 1990. Molecular genetics of polyketides and its comparison to fatty acid synthesis. Annu. Rev. Genet. 24:37-66.

Hotto, A., M. Satchwell, and G. Boyer. 2005. Seasonal production and molecular characterization of microcystins in Oneida Lake, New York, USA. Environ. Toxicol. 20:243-248.

Hotto, A., M. Satchwell, and G. Boyer. 2007. Molecular characterization of potential microcystin-producing cyanobacteria in Lake Ontario embayments and nearshore waters. Appl. Environ. Microbiol. 73:4570-4578.

Isvánovics, V., H. M. Shafik, M. Présing, and S. Juhos. 2000. Growth and phosphate kinetics of the cyanobacterium, C y l i n d ro s p e r m o p s i s r a c i b o r s k i i (Cyanophyceae) in throughflow cultures. Freshwater Biol. 43:257-275.

Izaguirre, G., A.-D. Jungblut, and B. A. Neilan. 2007. Benthic cyanobacteria (Oscillatoriaceae) that produce microcystin-LR, isolated from four reservoirs in southern California. Water Res. 41:492-498.

Jacquet, S., J.-F. Briand, C. Leboulanger, C. Avois-Jacquet, L. Oberhaus, B. Tassin, B. Vinçon-Leite, G. Paolini, J.-C. Druart, O. Anneville, and J.-F. Humbert. 2005. The proliferation of the toxic cyanobacterium Planktothrix rubescens following restoration

of the largest natural French lake (Lac du Bourget). Harmful Algae 4:651-672.

Jain, R., M. C. Rivera, and J. A. Lake. 1999. Horizontal gene transfer among genomes: the complexity hypothesis. Proc. Natl. Acad. Sci. USA 96:3801-3806.

Janse, I., W. E. A. Kardinaal, M. Meima, J. Fastner, P. M. Visser, and G. Zwart. 2004. Toxic and nontoxic Microcystis colonies in natural populations can be differentiated on the basis of rRNA gene internal transcribed spacer diversity. Appl. Environ. Microbiol. 70:3979-3987.

Jochimsen, E.M., W. W. Carmichael, J. S. An, D. M Cardo, S. T. Cookson, C. E. Holmes, M. B. Antunes, D. A de Melo Filho, T. M. Lyra, V. S. T. Barreto, S. M. F. O. Azevedo, and W. R. Jarvis. 1998. Liver failure and death after exposure to microcystins at a hemodialysis center in Brazil. N. Engl. J. Med. 338:873-878.

Jungblut, A.-D., and B. A. Neilan. 2006. Molecular identifi cation and evolution of the cyclic peptide hepatotoxins, microcystin and nodularin, synthetase genes in three orders of cyanobacteria. Arch. Microbiol. 185:107-114.

Kaebernick, M., B. A. Neilan, T. Börner, and E. Dittmann. 2000. Light and the transcriptional response of the microcystin biosynthesis gene cluster. Appl. Environ. Microbiol. 66:3387-3392.

Kaebernick, M. , T. Rohrlack, K. Christoffersen, and B. A. Neilan. 2001. A spontaneous mutant of microcystin biosynthesis: genetic characterization and effect on Daphnia. Environ. Microbiol. 3:669-679.

Katoh, H., R. K. Asthana, and M. Ohmori. 2004. Gene expression in the cyanobacterium Anabaena sp. PCC7120 under desiccation. Microb. Ecol. 47:164-174.

Kleinkauf, H., and H. von Döhren. 1995. The nonribosomal peptide biosynthetic system – on the origins of structural diversity of peptides, cyclopeptides and related compounds. Antonie van Leeuwenhoek 67:229-242.

References

49

Koskenniemi, K., C. Lyra, P. Rajaniemi-Wacklin, J. Jokela, and K. Sivonen. 2007. Quantitative real-time PCR detection of toxic Nodularia cyanobacteria in the Baltic Sea. Appl. Environ. Microbiol. 73:2173-2179.

Kotak, B. G., A. K.-Y. Lam, E. E. Prepas, and S. E. Hrudey. 2000. Role of chemical and physical variables in regulating microcystin-LR concentration in phytoplankton of eutrophic lakes. Can. J. Fish. Aquat. Sci. 57:1584-1593.

Kotak, B. G., A. K.-Y. Lam, E. E. Prepas, S. L. Kenefi ck, and S. E. Hrudey. 1995. Variability of the hepatotoxin microcystin-LR in hypereutrophic drinking water lakes. J. Phycol. 31:248-263.

Kucho, K.-I., Y. Tsuchiya, Y. Okumoto, M. Harada, M. Yamada, and M. Ishiura. 2004. Construction of unmodifi ed oligonucleotide-based microarrays in the thermophilic cyanobacterium Thermosynechococcus elongatus BP-1: screening of the candidates for circadianly expressed genes. 79:319-329.

Kuiper-Goodman, T., I. Falconer, and J. Fitzgerald. 1999. Human health aspects, p.113-153. In I. Chorus and J. Bartram (ed.), Toxic cyanobacteria in water: a guide to their public health consequences, monitoring, and management. E & FN Spon, London, United Kingdom.

Kurmayer, R., G. Christiansen, J. Fastner, and T. Börner. 2004. Abundance of active and inactive microcystin genotypes in populations of the toxic cyanobacterium Planktothrix spp. Environ. Microbiol. 6:831-841.

Kurmayer, R., G. Christiansen, M. Gumpenberger, and J. Fastner. 2005. Genetic identification of microcystin ecotypes in toxic cyanobacteria of the genus Planktothrix. Microbiol. 151:1525-1533.

Kurmayer, R., E. Dittmann, J. Fastner, and I. Chorus. 2002. Diversity of microcystin genes within a population of the toxic cyanobacterium Microcystis spp. in Lake Wannsee (Berlin, Germany). Microb. Ecol. 43:107-118.

Kurmayer, R., and M. Gumpenberger. 2006. Diversity of microcystin genotypes among populations of the fi lamentous cyanobacteria Planktothrix rubescens and Planktothrix agardhii. Mol. Ecol. 15:3849-3861.

Kurmayer, R., and T. Kutzenberger. 2003. Application of real-time PCR for quantifi cation of microcystin genotypes in a population of the toxic cyanobacterium Microcystis sp. Appl. Environ. Microbiol. 69:6723-6730.

Laamanen, M. J., M. F. Gugger, J. M. Lehtimäki, K. Haukka, and K. Sivonen. 2001. Diversity of toxic and nontoxic Nodularia isolates (Cyanobacteria) and fi laments from the Baltic Sea. Appl. Environ. Microbiol. 67:4638-4647.

Lahti, K., J. Rapala, M. Färdig, M. Niemelä, and K. Sivonen. 1997. Persistence of cyanobacterial hepatotoxin, microcystin-LR in particulate material and dissolved in lake water. Wat. Res. 31:1005-1012.

Lehtimäki, J. C. Lyra, S. Suomalainen, P. Sundman, L. Rouhiainen, L. Paulin, M. Salkinoja-Salonen, and K. Sivonen. 2000. Characterization of Nodularia strains, cyanobacteria from brackish waters, by genotypic and phenotypic methods. Int. J. Syst. Evol. Microbiol. 50:1043-1053.

Lepistö, L., A. Räike, and O.-P. Pietiläinen. 1999. Long-term changes of phytoplankton in a eutrophicated boreal lake during the past one hundred years (1893-1998). Algol. Stud. 94:223-244.

Li, W.-H., and D. Graur. 1991. Fundamentals of molecular evolution. Sinauer Associates, Sunderland, MA, 284 pp.

Lindholm, T., and J. A. O. Meriluoto. 1991. Recurrent depth maxima of the hepatotoxic cyanobacterium Oscillatoria agardhii. Can. J. Fish. Aquat. Sci. 48:1629-1634.

Lyra, C., S. Suomalainen, M. Gugger, C. Vezie, P. Sundman, L. Paulin, and K. Sivonen. 2001. Molecular characterization of planktic cyanobacteria of Anabaena, Aphan i zomenon , Microcys t i s and Planktothrix genera. Int. J. Syst. Evol. Microbiol. 51:513-526.

References

50

MacKintosh, C., K. A. Beattie, S. Klumpp, P. Cohen, and G. A. Codd. 1990. Cyanobacterial microcystin-LR is a potent and specifi c inhibitor of protein phosphatases 1 and 2A from both mammals and higher plants. FEBS Lett. 264:187-192.

Mankiewicz-Boczek, J., K. Izydorczyk, Z. Romanowska-Duda, T. Jurczak, K. Stefaniak, and M. Kokocinski. 2006. Detection and monitoring toxigenicity of cyanobacteria by application of molecular methods. Environ. Toxicol. 21:380-387.

Marahiel, M. A., T. Stachelhaus, and H. D. Mootz. 1997. Modular peptide synthetases involved in nonribosomal peptide synthesis. Chem Rev. 97:2651-2673.

Marin, K., I. Suzuki, K. Yamaguchi, K. Ribbeck, H. Yamamoto, Y. Kanesaki, M. Hagemann, and N. Murata. 2003. Identification of histidine kinases that act as sensors in the perception of salt stress in Synechocystis sp. PCC 6803. Proc. Nat. Acad. Sci. USA 100:9061-9066.

Martin-Luna, B., E. Sevilla, J. A. Hernandez, M. T. Bes, M. F. Fillat, and M. L. Peleato. 2006. Fur from Microcystis aeruginosa binds in vitro promoter regions of the microcystin biosynthesis gene cluster. Phytochem. 67:876-881.

Martiny, A. C., M. L. Coleman, and S. W. Chisholm. 2006. Phosphate acquisition genes in Prochlorococcus ecotypes: evidence for genome-wide adaptation. Proc. Nat. Acad. Sci. USA 103:12552-12557.

Matsunaga, T., H. Nakayama, M. Okochi, and H. Takeyama. 2001. Fluorescent detection of cyanobacterial DNA using bacterial magnetic particles on a MAG-microarray. Biotechnol. Bioeng. 73:400-405.

Mbedi, S., M. Welker, J. Fastner, and C. Wiedner. 2005. Variability of the microcystin synthetase gene cluster in the genus Planktothrix (Oscillatoriales, Cyanobacteria). FEMS Microbiol. Lett. 245:299-306.

McElhiney, J., and L. A. Lawton. 2005. Detection of the cyanobacterial hepatotoxins microcystins. Toxicol. Appl. Pharmacol. 203:219-230.

Meißner, K., E. Dittmann, and T. Börner. 1996. Toxic and non-toxic strains of the cyanobacterium Microcystis aeruginosa contain sequences homologous to peptide synthetase genes. FEMS Microbiol Lett. 135:295-303.

Mikalsen, B., G. Boison, O. M. Skulberg, J. Fastner, W. Davies, T. M. Gabrielsen, K. Rudi, and K. S. Jakobsen. 2003. Natural variation in the microcystin synthetase operon mcyABC and impact on microcystin production in Microcystis strains. J. Bacteriol. 185:2774-2785.

Mira, A., H. Ochman, and N. A. Moran. 2001. Deletional bias and the evolution of bacterial genomes. Trends Genet. 17:589-596.

Moffitt, M. C., S. I. Blackburn, and B. A. Neilan. 2001. rRNA sequences reflect the ecophysiology and define the toxic cyanobacteria of the genus Nodularia. Int. J. Syst. Evol. Microbiol. 51:505-512.

Moffitt, M. C., and B. A. Neilan. 2001. On the presence of peptide synthetase and polyketide synthase genes in the cyanobacterial genus Nodularia. FEMS Microbiol. Lett. 196:207-214.

Moffitt, M. C., and B. A. Neilan. 2004. Characterization of the nodularin synthetase gene cluster and proposed theory of the evolution of cyanobacterial hepatotoxins. Appl. Environ. Microbiol. 70:6353-6362.

Mohamed, Z. A., H. M. El-Sharouny, and W. S. M. Ali. 2006. Microcystin production in benthic mats of cyanobacteria in the Nile River and irrigation canals, Egypt. Toxicon 47:584-590.

Moisander, P. H., A. E. Morrison, B. B. Ward, B. D. Jenkins, and J. P. Zehr. 2007. Spatial-temporal variability in diazotroph assemblages in Chesapeake Bay using an oligonucleotide nifH microarray. Environ. Microbiol. 9:1823-1835.

Moisander, P. H., L. Shiue, G. F. Steward, B. D. Jenkins, B. M. Bebout, and J. P. Zehr. 2006. Application of a nifH oligonucleotide microarray for profiling diversity of N2-fi xing microorganisms in marine microbial mats. Environ. Microbiol. 8:1721-1735.

References

51

Mur, L. R., O. M. Skulberg, and H. Utkilen. 1999. Cyanobacteria in the environment, p.15-40. In I. Chorus and J. Bartram (ed.), Toxic cyanobacteria in water: a guide to their public health consequences, monitoring, and management. E & FN Spon, London, United Kingdom.

Murata, N., and I. Suzuki. 2006. Exploitation of genomic sequences in a systematic analysis to access how cyanobacteria sense environmental stress. J. Exp. Bot. 57:235-247.

Nei, M., and S. Kumar. 2000. Molecular evolution and phylogenetics. Oxford University press, New York, NY, 333 pp.

Neilan, B. A., E. Dittmann, L. Rouhiainen, R. A. Bass, V. Schaub, K. Sivonen, and T. Börner. 1999. Nonribosomal peptide synthesis and toxigenicity of cyanobacteria. J. Bacteriol. 181:4089-4097.

Neilan, B. A., D. Jacobs, T. del Dot, L. L. Blackall, P. R. Hawkins, P. T. Cox, and A. E. Goodman. 1997. rRNA sequences and evolutionary relationships among toxic and nontoxic cyanobacteria of the genus Microcystis. Int. J. Syst. Bacteriol. 47:693-697.

Nishizawa, T., M. Asayama, K. Fuji, K.-I. Harada, and M. Shirai. 1999. Genetic analysis of the peptide synthetase genes for a cyclic heptapeptide microcystin in Microcystis spp. J. Biochem. 126:520-529.

Nishizawa, T., A. Ueda, M. Asayama, K. Fuji, K.-I. Harada, K. Ochi, and M. Shirai. 2000. Polyketide synthase gene coupled to the peptide synthetase module involved in the biosynthesis of the cyclic heptapeptide microcystin. J. Biochem. 127:779-789.

Nonneman, D., and P. V. Zimba. 2002. A PCR-based test to assess the potential for microcystin occurrence in channel catfish production ponds. J. Phycol. 38:230-233.

Ochman, H., J. G. Lawrence, and E. A. Groisman. 2000. Lateral gene transfer and the nature of bacterial innovation. Nature 405:299-304.

Oh, H.-M., S. J. Lee, J.-H. Kim, H.-S. Kim, and B.-D. Yoon. 2001. Seasonal variation and indirect monitoring of microcystin concentrations in Daechung Reservoir,

Korea. Appl. Environ. Microbiol. 67:1484-1489.

Oksanen, I., J. Jokela, D. P. Fewer, M. Wahlsten, J. Rikkinen, and K. Sivonen. 2004. Discovery of rare and highly toxic microcystins from lichen-associated cyanobacterium Nostoc sp. strain IO-102-I. Appl. Environ. Microbiol. 70:5756-5763.

Oliver, R. L., and G. G. Ganf. 2000. Freshwater blooms, p. 149-194. In B. A. Whitton and M. Potts (ed.), The ecology of cyanobacteria. Kluwer Academic, Publishers, Dordrecht, The Netherlands.

Orr, P. T., and G. J. Jones. 1998. Relationship between microcystin production and cell division rates in nitrogen-limited Microcystis aeruginosa cultures. Limnol. Oceanogr. 43:1604-1614.

Otsuka, S., S. Suda, R. Li, M. Watanabe, H. Oyaizu, S. Matsumoto, and M. M. Watanabe. 1999. Phylogenetic relationships between toxic and non-toxic strains of the genus Microcystis based on 16S to 23S internal transcribed spacer sequence. FEMS Microbiol. Lett. 172:15-21.

Ouahid, Y., G. Pérez-Silva, and F. F. del Campo. 2005. Identifi cation of potentially toxic environmental Microcystis by individual and multiple PCR amplifi cation of specific microcystin synthetase gene regions. Environ. Toxicol. 20:235-242.

Ouellette, A. J. A., S. M. Handy, and S. W. Wilhelm. 2006. Toxic Microcystis is widespread in Lake Erie: PCR detection of toxin genes and molecular characterization of associated cyanobacterial communities. Microb. Ecol. 51:154-165.

Paerl, H. W. 2000. Marine plankton, p. 121-148. In B. A. Whitton and M. Potts (ed.), The ecology of cyanobacteria. Kluwer Academic, Publishers, Dordrecht, The Netherlands.

Paerl, H. W. 1996. Microscale physiological and ecological s tudies of aquat ic cyanobacteria: macroscale implications. Microsc. Res. Tech. 33:47-72.

Pan, H., L. Song, Y. Liu, and T. Börner. 2002. Detection of hepatotoxic Microcystis strains by PCR with intact cells from both culture and environmental samples. Arch. Microbiol. 178:421-427.

References

52

Pomati, F., R. Kellmann, R. Cavalieri, B. P. Burns, and B. A. Neilan. 2006. Comparative gene expression of PSP-toxin producing and non-toxic Anabaena circinalis strains. Environ. Int. 32:743-748.

Ragan, M. A. 2001. Detection of lateral gene transfer among microbial genomes. Curr. Opin. Genet. Dev. 11:620-626.

Rajaniemi-Wacklin, P., A. Rantala, P. Kuuppo, K. Haukka, and K. Sivonen. Cyanobacterial community composition in shallow, eutrophic Lake Tuusulanjärvi studied by microscopy, strain isolation, DGGE and cloning, Algol. Stud., in press.

Ressom, R., F. S. Soong, J. Fitzgerald, L. Turczynowicz, O. El Saadi, D. Roder, T. Maynard, and I. Falconer. 1994. Health effects of toxic cyanobacteria (blue-green algae). National Health and Medical Research Council, Australian Government Publishing Service, Canberra.

Rinta-Kanto, J. M., A. J. A. Ouellette, G. L. Boyer, M. R. Twiss, T. B. Bridgeman, and S. W. Wilhelm. 2005. Quantifi cation of toxic Microcystis spp. during the 2003 and 2004 blooms in western Lake Erie using quantitative real-time PCR. Environ. Sci. Technol. 39:4198-4205.

Rinta-Kanto, J. M., and S. W. Wilhelm. 2006. Diversity of microcystin-producing cyanobacteria in spatially isolated regions of Lake Erie. Appl. Environ. Microbiol. 72:5083-5085.

Rohrlack, T., E. Dittmann, T. Börner, and K. Christoffersen. 2001. Effects of cell-bound microcystins on survival and feeding of Daphnia spp. Appl. Environ. Microbiol. 67:3523-3529.

Rohrlack, T., E. Dittmann, M. Henning, T. Börner, and J.-G. Kohl. 1999. Role of microcystins in poisoning and food ingestion inhibition of Daphnia galeata caused by the cyanobacterium Microcystis aeruginosa. Appl. Environ. Microbiol. 65:737-739.

Rolland, A., D. F. Bird, and A. Giani. 2005. Seasonal changes in composition of the cyanobacterial community and the occurrence of hepatotoxic blooms in the eastern townships, Québec, Canada. J. Plankton Res. 27:683-694.

Rouhiainen, L., K. Sivonen, W. J. Buikema, and R. Haselkorn. 1995. Characterization of toxin-producing cyanobacteria by using an oligonucleotide probe containing a tandemly repeated heptamer. J. Bacteriol. 177:6021-6026.

Rouhiainen, L., T. Vakkilainen, B. L. Siemer, W. Buikema, R. Haselkorn, and K. Sivonen. 2004. Genes coding for hepatotoxic heptapeptides (microcystins) in the cyanobacterium Anabaena strain 90. Appl. Environ. Microbiol. 70:686-692.

Rudi, K., O. M. Skulberg, R. Skulberg, and K. S. Jakobsen. 2000. Application of sequence-specifi c labeled 16S rRNA gene oligonucleotide probes for genetic profi ling of cyanobacterial abundance and diversity by array hybridization. Appl. Environ. Microbiol. 66:4004-4011.

Saker, M. L., J. Fastner, E. Dittmann, G. Christiansen, and V. M. Vasconselos. 2005a. Variation between strains of the cyanobacterium Microcystis aeruginosa isolated from a Portuguese river. J. Appl. Microbiol. 99:749-757.

Saker, M. L., A.-D. Jungblut, B. A. Neilan, D. F. K. Rawn, and V. M. Vasconcelos. 2005b. Detection of microcystin synthetase genes in health food supplements containing the freshwater cyanobacterium Aphanizomenon fl os-aquae. Toxicon 46:555-562.

Saker, M. L., M. Vale, D. Kramer, and V. M. Vasconcelos. 2007a. Molecular techniques for the early warning of toxic cyanobacteria blooms in freshwater lakes and rivers. Appl. Microbiol. Biotechnol. Epub ahead of print, DOI 10.1007/s00253-006-0813-8

Saker, M. L., M. Welker, and V. M. Vasconcelos. 2007b. Multiplex PCR for the detection of toxigenic cyanobacteria in dietary supplements produced for human consumption. Appl. Microbiol. Biotechnol. 73:1136-1142.

Sato, N., M. Ohmori, M. Ikeuchi, K. Tashiro, C. P. Wolk, T. Kaneko, K. Okada, M. Tsuzuki, S. Ehira, H. Katoh, S. Okamoto, H. Yoshimura, T. Fujisawa, A. Kamei, S. Yoshihara, R. Narikawa, T. Hamano, S. Tabata, and S. Kuhara. 2004. Use of segment-based microarray in

References

53

the analysis of global gene expression is response to various environmental stresses in the cyanobacterium Anabaena sp. PCC 7120. J. Gen. Microbiol. 50:1-8.

Schatz, D., Y. Keren, A. Vardi, A. Sukenik, S. Carmeli, T. Börner, E. Dittmann, and A. Kaplan. 2007. Towards clarifi cation of the biological role of microcystins, a family of cyanobacterial toxins. Environ. Microbiol. 9:965-970.

Schober, E., M. Werndl, K. Laakso, I . Korschineck, K. Sivonen, and R. Kurmayer. 2007. Interlaboratory comparison of Taq nuclease assays for the quantification of the toxic cyanobacteria Microcystis sp. J. Microbiol. Meth. 69:122-128.

Schopf, J. W. 2000. The fossil record: tracing the roots of the cyanobacterial lineage, p. 13-35. In B. A. Whitton and M. Potts (ed.), The ecology of cyanobacteria: their diversity in time and space. Kluwer Academic Publishers, Dordrecht, The Netherlands.

Sivonen, K., and G. Jones. 1999. Cyanobacterial toxins, p. 41-111. In I. Chorus and J. Bartram (ed.), Toxic cyanobacteria in water: a guide to their public health consequences, monitoring, and management. E & FN Spon, London, United Kingdom.

Sivonen, K. , K. Kononen, W. W. Carmichael, A. M. Dahlem, K. L. Rinehart, J. Kiviranta, and S. I. Niemelä. 1989. Occurrence of the hepatotoxic cyanobacterium Nodularia spumigena in the Baltic Sea and structure of the toxin. Appl. Environ. Microbiol. 55:1990-1995.

Sivonen, K., S. I. Niemelä, R. M. Niemi, L. Lepistö, T. H. Luoma, and L. A. Räsänen. 1990. Toxic cyanobacteria (blue-green algae) in Finnish fresh and coastal waters. Hydrobiol. 190:267-275.

Sommer, U., Z. M. Gliwicz, W. Lampert, and A. Duncan. 1986. The PEG*-model of seasonal succession of planktonic events in fresh waters. Arch. Hydrobiol. 106:433-471.

Spoof, L. 2005. Microcystins and nodularins, p. 15-39. In J. Meriluoto and G. A. Codd

(ed.), TOXIC: cyanobacterial monitoring and cyanotoxin analysis. Åbo Akademi University Press, Åbo, Finland.

Stachelhaus, T., H. D. Mootz, and M. A. Marahiel. 1999. The specificity-conferring code of adenylation domains in nonribosomal peptide synthetases. Chem. Biol. 6:493-505.

Steglich, C., M. Futschik, T. Rector, R. Steen, and S. W. Chisholm. 2006. Genome-wide analysis of light sensing in Prochlorococcus. J. Bacteriol. 188:7796-7806.

Stowe-Evans, E. L., J. Ford, and D. M. Kehoe. 2004. Genomic DNA microarray analysis: identification of new genes regulated by light color in the cyanobacterium Fremyella diplosiphon. J. Bacteriol. 186:4338-4349.

Suzuki, I. Y. Kanesaki, K. Mikami, M. Kanehisa, and N. Murata. 2001. Cold-regulated genes under control of the cold sensor Hik33 in Synechocystis. Mol. Microbiol. 40:235-244.

Tanabe, Y., K. Kaya, and M. M. Watanabe. 2004. Evidence for recombination in the microcystin synthetase (mcy) genes of toxic cyanobacteria Microcystis spp. J. Mol. Evol. 58:633-641.

Tillett, D., E. Dittmann, M. Erhard, H. von Döhren, T. Börner, and B. A. Neilan. 2000. Structural organization of microcystin biosynthesis in Microcystis aeruginosa PCC7806: an integrated peptide-polyketide synthetase system. Chem. Biol. 7:753-764.

Tillett, D., D. L. Parker, B. A. Neilan. 2001. Detection of toxigenity by a probe for the microcystin synthetase A gene (mcyA) of the cyanobacterial genus Microcystis: comparison of toxicities with 16S rRNA and phycocyanin operon (Phycocyanin intergenic spacer) phylogenies. Appl. Environ. Microbiol. 67:2810-2818.

Tonk, L., P. M. Visser, G. Christiansen, E. Dittmann, E. O. F. M. Snelder, C. Wiedner, L. R. Mur, and J. Huisman. 2005. The microcystin composition of the cyanobacterium Planktothrix agardhii changes toward a more toxic variant with increasing light intensity. Appl. Environ. Microbiol. 71:5177-5181.

References

54

Utkilen, H., and N. Gjølme. 1995. Iron-stimulated toxin production in Microcystis aeruginosa. Appl. Environ. Microbiol. 61:797-800.

Vaitomaa, J., A. Rantala, K. Halinen, L. Rouhiainen, P. Tallberg, L. Mokelke, and K. Sivonen. 2003. Quantitative real-time PCR for determination of microcystin synthetase E copy numbers for Microcystis and Anabaena in lakes. Appl. Environ. Microbiol. 69:7289-7297.

Vezie, C., L. Brient, K. Sivonen, G. Bertru, J.-C. Lefeuvre, and M. Salkinoja-Salonen. 1997. Occurrence of microcystin-containing cyanobacterial blooms in freshwaters of Brittany (France). Arch. Hydrobiol. 139:401-413.

Vezie, C., L. Brient, K. Sivonen, G. Bertru, J.-C. Lefeuvre, and M. Salkinoja-Salonen. 1998. Variation of microcystin content of cyanobacterial blooms and isolated strains in Lake Grand-Lieu (France). Microb. Ecol. 35:126-135.

Via-Ordorika, L., J. Fastner, R. Kurmayer, M. Hisbergues, E. Dittmann, J. Komarek, M. Erhard, and I. Chorus. 2004. Distribution of microcystin-producing and non-microcystin-producing Microcystis sp. in European freshwater bodies: detection of microcystins and microcystin genes in individual colonies. System. Appl. Microbiol. 27:592-602.

Vieira, J. M. S., M. T. P. Azevedo, S. M. F. O. Azevedo, R. Y. Honda, and B. Corrêa. 2003. Microcystin production by Radiocystis fernandoi (Chroococcales, Cyanobacteria) isolated from a drinking water reservoir in the city of Belém, PA, Brazilian Amazonia region. Toxicon 42:709-713.

von Döhren, H., R. Dieckmann, and M. Pavela-Vrancic. 1999. The nonribosomal code. Chem. Biol. 6:R273-R279.

Wang, D. Y.-C., S. Kumar, and S. B. Hedges. 1999. Divergence time estimates for the early history of animal phyla and the origin of plants, animals and fungi. Proc. R. Soc. Lond. B 266:163-171.

Welker, M., and H. von Döhren. 2006. Cyanobacterial peptides - Nature’s own combinatorial biosynthesis . FEMS Microbiol. Rev. 30:530-563.

Whitton, B. A., and M. Potts. 2000. The ecology of cyanobacteria. Kluwer Academic, Publishers, Dordrecht, The Netherlands.

Wicks, R. J., and P. G. Thiel. 1990. Environmental factors affecting the production of peptide toxins in floating scums of the cyanobacterium Microcystis aeruginosa in a hypertrophic African reservoir. Environ. Sci. Technol. 24:1413-1418.

Wilson, A. E., O. Sarnelle, B. A. Neilan, T. P. Salmon, M. M. Gehringer, and M. E. Hay. 2005. Genetic variation of the bloom-forming cyanobacterium Microcystis aeruginosa within and among lakes: implications for harmful algal blooms. Appl. Environ. Microbiol. 71:6126-6133.

Yoshida, M., T. Yoshida, Y. Takashima, N. Hosoda, and S. Hiroishi. 2007. Dynamics of microcystin-producing and non-microcystin-producing Microcystis populations is correlated with nitrate concentration in a Japanese lake. FEMS Microbiol. Lett. 266:49-53.

Yoshida, M., T. Yoshida, Y. Takashima, R. Kondo, and S. Hiroishi. 2005. Genetic diversity of the toxic cyanobacterium Microcystis in Lake Mikata. Environ. Toxicol. 20:229-234.

Yoshimura, H., S. Okamoto, Y. Tsumuraya, and M. Ohmori. 2007. Group 3 sigma factor gene, sigJ, a key regulator of desiccation tolerance, regulates the synthesis of extracellular polysaccharide in cyanobacterium Anabaena sp. strain PCC 7120. DNA Res. 14:13-24.

Zhou, J., and D. K. Thompson. 2002. Challenges in applying microarrays to environmental studies. Curr. Opin. Biotech. 13:204-207.

References